Volume 20 Issue 4

International Food and Agribusiness Management Review

Official Journal of the International Food and Agribusiness Management Association

Volume 20 Issue 4 2017

EVERY GENERATION NEEDS ITS LEADERS. As a farmer-owned cooperative, we help lead the agricultural industry by promoting the most innovative, responsible farming practices available today. And we want to make sure that the next generation of leaders is well trained for tomorrow. That’s why we proudly support youth leadership programs like FFA. It’s also why we’ve invested in scholarships and countless land-grant university initiatives across the country. At CHS, we know every generation needs its leaders. And every potential leader needs a little help becoming one.

Visit chsinc.com/stewardship to learn more about our ongoing commitment to leadership.

©2017 CHS Inc.

International Food and Agribusiness Management Review

Editorial Staff Executive Editor

Gerhard Schiefer, University of Bonn, Germany

Regional Managing Editors Asia, Australia, and New Zealand

Derek Baker, UNE, Australia Kim Bryceson, University of Queensland, Australia Kevin Chen, IFPRI-Bejing, China Jeff Jia, University of Exeter, United Kingdom Nicola M. Shadbolt, Massey University, New Zealand

Europe

Pegah Amani, Technical Institute of Sweden, Sweden Vera Bitsch, Technical University of Munich, Germany Laura Carraresi, University of Bonn, Germany Alessio Cavicchi, University of Macerata, Italy Hans De Steur, Ghent University, Belgium Loic Sauvee, UniLaSalle, Beauvais, France Cristina Santini, University San Raffaele, Italy Jacques Trienekens, Wageningen University, The Netherlands

North America

Ram Acharya, New Mexico State University, USA Yuliya Bolotova, Clemson University, USA Michael Gunderson, Purdue University, USA William Nganje, North Dakota State, USA R. Brent Ross, Michigan State University, USA Aleksan Shanoyan, Kansas State University, USA David Van Fleet, Arizona State University, USA Nicole Olynk Widmar, Purdue University, USA Cheryl Wachenheim, North Dakota State University, USA

South America

Aziz da Silva Júnior, Federal University of Vicosa, Brazil Jose Vincente Caixeta Filho, University of Sao Paulo, Brazil Emilio Morales, University of New England, Australia

Africa

Ajuruchukwu Obi, University of Fort Hare, South Africa Nick Vink, Stellenbosch University, South Africa

Editorial Board Filippo Arfini, Universita’ di Parma, Italy Stefano Boccaletti, Universita’ Cattolica, Italy Michael Boehlje, Purdue University, USA Dennis Conley, University of Nebraska - Lincoln, USA Francis Declerck, ESSEC Business School, France Jay Lillywhite, New Mexico State University, USA Woody Maijers, INHOLLAND University, The Netherlands

Marcos Fava Neves, FEA / USP / PENSA, Brazil Onno Omta, Wageningen University, The Netherlands Hernán Palau, Buenos Aires University, Argentina Christopher Peterson, Michigan State University, USA Thomas Reardon, Michigan State University, USA Mary Shelman, (Retired) Harvard Business School, USA Johan van Rooyen, University of Stellenbosch, South Africa

The IFAMR (ISSN #: 1559-2448) is published quarterly and the archived library is available at http://www.ifama.org. For copyright and publishing information, please contact: Marijn van der Gaag, Administrative Editor Wageningen Academic Publishers • P.O. Box 220 6700 AE Wageningen • The Netherlands • Tel: +31 317 476511 Fax: +31 317 453417 • Email: ifamr@wageningenacademic.com • Web: http://www.wageningenacademic.com/loi/ifamr.

International Food and Agribusiness Management Review Volume 20 Issue 4, 2017

TABLE OF CONTENTS 1.

2.

3.

E-commerce in agri-food sector: a systematic literature review

439

A template for sustainable food value chains

461

How do pricing and the representation of price affect consumer evaluation of nursery products? A conjoint analysis

477

Management support for agricultural enterprises: a case study for a fruit-producing company

493

Governance structures and coordination mechanisms in the Brazilian pork chain – Diversity of arrangements to support the supply of piglets

511

Shaping food systems towards improved nutrition: a case study on Tuscan Bread Protected Designation of Origin

533

Socioeconomic impacts of innovative dairy supply chain practices – The case of the Laiterie du Berger in the Senegalese Sahel

553

Comparative advantage and export potential of Thai vegetable products following the integration into the ASEAN Economic Community

575

Yiwu Zeng, Fu Jia, Li Wan, and Hongdong Guo

Elena Monastyrnaya, Gwenola Yannou Le Bris, Bernard Yannou, and Gaëlle Petit

Zhiwen Zhu, Bridget Behe, Patricia Huddleston, and Lynnell Sage

4.

Leonardo Ensslin, Vinícius Dezem, Ademar Dutra, Sandra R. Ensslin, and Karine Somensi

5.

Franco M. Martins, Jacques Trienekens, and Onno Omta

6.

Francesca Galli, Francesca Venturi, Fabio Bartolini, Oriana Gava, Angela Zinnai, Sanmartin Chiara, Gianpaolo Andrich, and Gianluca Brunori

7.

Abdrahmane Wane, Jean-Joseph Cadilhon, and Mamadou Yauck

8.

Pheesphan Laosutsan, Ganesh P. Shivakoti, and Peeyush Soni

International Food and Agribusiness Management Review

i

Table of contents

9.

Volume 20, Issue 4, 2017

The impact of Mexican competition on the U.S. strawberry industry

591

10. Is dairy complex a solution to milk safety? A comparison of farmers’ perceived and realized food safety effects

605

Dong Hee Suh, Zhengfei Guan, and Hayk Khachatryan

H. Holly Wang, Hailong Yu, and Binglong Li

International Food and Agribusiness Management Review

ii

OPEN ACCESS International Food and Agribusiness Management Review Volume 20 Issue 4, 2017; DOI: 10.22434/IFAMR2016.0156

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0156 - Tuesday, August 22, 2017 12:29:21 PM - IP Address:24.21.169.207

Received: 29 September 2016 / Accepted: 26 February 2017

E-commerce in agri-food sector: a systematic literature review REVIEW ARTICLE Yiwu Zenga, Fu Jiab, Li Wanc, and Hongdong Guo

d

aPhD

Candidate and dProfessor, China Academy of Rural Development, Zhejiang University, Hangzhou, 310058 Zhejiang, China P.R.

bSenior

cPhD

Lecturer, Business School, University of Exeter, Streatham Court, Rennes Drive, EX4 4PU Exeter, United Kingdom

Candidate, Polytechnic Department of Engineering and Architecture, University of Udine, Via delle Scienze 206, 33100 Udine, Italy

Abstract This paper aims to synthesize findings in the Agri-food E-commerce (AE) field through a systematic literature review and propose a number of future research directions based on the gaps identified from the review.

There has been a general increase in the number of publications, indicating that AE research has elicited more and more interest from scholars in different countries and across multiple disciplines. We have identified a number of themes and made sense of them by developing an integrated conceptual model, which consists of two parts: one for AE adoption at a firm level and one for AE development at a regional level. Furthermore, we recommend that more emphasis should be put on the regional development modes of AE and their impact in the developing world, as the practice is evolving rapidly in some developing countries such as China.

Keywords: agriculture, food, e-commerce, smallholder market access, China.

JEL code: M31, Q13, Q18, R11 Corresponding author: guohongdong@zju.edu.cn

Š 2017 Zeng et al.

439

Zeng et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0156 - Tuesday, August 22, 2017 12:29:21 PM - IP Address:24.21.169.207

1. Introduction There has been much evidence that e-commerce offers an important opportunity for cost reduction and demand enhancement. Although the characteristics of some agri-food products present a few challenges for those wishing to market products through e-commerce, there is still much optimism about the potential success of e-commerce in agriculture (Leroux et al., 2001). The high level of fragmentation in the food supply chain reinforces the expectation for Agri-food E-commerce (AE) (Montealegre et al., 2007). The provision of food builds on a vertical chain of subsequent production, service and trading processes that span from the production of agricultural inputs to the delivery of final food products to consumers. AE means introducing e-strategy into the interaction and trading activities between participants in the food sector and changing the configuration and relationships at various stages and linkages of the food supply chain (Fritz et al., 2004; Giustiniano and Fratocchi, 2002).

In the developing world, smallholders in agriculture are considered disadvantaged in the agri-food supply chain and income growth poses a common and critical issue for policymakers (Wiggins et al., 2010). It is important for smallholders to successfully gain access to the market; however, they face many difficulties in this area. Due to their small scale, unit transaction costs are high in almost all transactions (Poulton et al., 2010). The pervasive imperfections of markets in the developing countries, such as lack of information on prices and technologies, lack of connections to established market actors, distortions or absence of input and output markets, and credit constraints, often make it very difficult for smallholders to take advantage of market opportunities (Markelova et al., 2009). To help smallholders address the inefficiencies and barriers to market access, two main approaches have been put forward. The first concerns taking collective action by establishing farmer organizations such as agricultural cooperatives (Hazell et al., 2010; Markelova et al., 2009). Acting collectively, smallholders may be in a better position to bargain with buyers and intermediaries, reduce the procurement cost of inputs, and obtain more market information and policy support. The second approach is to promote contract arrangements between smallholders and agribusiness firms (Abebe et al., 2013; Guo and Jolly, 2008; Key and Runsten, 1999). In contract farming, smallholders arrange their production and sell the primary products to processing or distribution firms at a prior agreed price according to the signed contract.

In recent years, more and more smallholders in developing countries such as China have begun to sell agriproducts directly to consumers via online shops in a third party trade platform. It is increasingly clear that e-commerce has become a new and effective way of helping smallholders to gain access to the market. By adopting e-commerce, smallholders can sell most of their products at a higher price than before because of the elimination of the price squeeze from intermediaries and the marketing constraints of information asymmetry (Zeng et al., 2016). China’s practice presents some new and exciting issues for AE research.

Noticeably, there are some authors who have just reviewed a small number of articles related to AE in the literature review section of their empirical works (Batte and Ernst, 2007; De Koster, 2002; Liu et al., 2011; Machfud and Kartiwi, 2013; Zapata et al., 2013). We have only found one literature review paper on this topic but it focuses on e-business models in the agri-food supply chain, which is narrow in scope, and it is dated (Van der Vorst et al., 2002). However, to the best of our knowledge, no systematic literature review comprehensively examining AE has been carried out to date. With the fast development of AE practice, the aim of the paper is therefore to identify the key themes in this field and shed light on future research directions.

2. Research method A systematic literature review method, considered as a necessary step in structuring a research field and forming an integral part of any research conducted (Easterby-Smith et al., 2010), is adopted. Compared to a traditional narrative method of reviewing literature, a systematic review employs a more transparent paper-selection process, which enhances rigor and thoroughness, and reduces the effects of researcher bias (Tranfield et al., 2003). Fink (2005) also argues that a systematic literature review is a systematic and reproducible design for International Food and Agribusiness Management Review

440

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0156 - Tuesday, August 22, 2017 12:29:21 PM - IP Address:24.21.169.207

Zeng et al.

Volume 20, Issue 4, 2017

identifying and evaluating an existing body of scholarly works. To increase the reliability of the research, the systematic literature review process of going through the literature search, the selection of studies, data extraction, and thematic synthesis, were independently carried out by two of the researchers/co-authors, who then compared notes and reached agreement on the selection and coding of papers (identifying themes) to increase inter-coder reliability (Duriau et al., 2007). Since there is no systematic knowledge system for AE, we didn’t have any prior framework to guide this coding process. Rather, we allowed the themes to emerge from the papers reviewed. To develop the conceptual model, due to the limited studies of AE, we used some case examples from industry reports (e.g. Ali Research) and other secondary sources (e.g. Internet news coverage) to complement the findings of the literature review.

The literature analyzed here comprises peer-reviewed English language papers published in academic journals and conference proceedings. Given the small numbers, books and book chapters in this field were excluded from the review. In terms of data collection, Scopus database (www.scopus.com), regarded as the largest database with respect to peer-reviewed literature and international publishers, was used. It also delivers the most comprehensive overview of the world’s research outputs in the field of science, technology, medicine, art & humanities, and social sciences. Two search strings of e-commerce and agri-food related keywords were first identified. The keywords related to e-commerce include: ‘e-commerce’ OR ‘e-business’ OR ‘e-market* (e-markets, e-marketing and e-marketplace)’ OR ‘electronic commerce’ OR ‘electronic business’ OR ‘electronic market*’. And the relevant keywords for agri-food sector are: ‘agri* (agriculture, agricultural, agribusiness and agri-food)’ OR ‘farm* (farming and farmer)’ OR ‘agro* (agro-based and agro-food)’ OR ‘food’. By combining the two search strings, there were 24 combinations input in the ‘Article Title, Abstract, Keywords’ domains of Scopus. We selected the subject areas of ‘Economics, Econometrics and Finance’; ‘Business, Management and Accounting’; ‘Social Sciences’; ‘Decision Sciences’ at Scopus. Initially, 310 hits were obtained, including 129 journal articles and 181 conference papers.

A first-round selection was made based on titles and abstracts in order to decide whether or not the full paper should be read for further analysis. We tried to be comprehensive in the selection so only two overarching exclusion criteria were applied. First, papers not directly related to AE, or papers discussing food and nonfood e-commerce at the same time but putting more emphasis on the latter, were excluded. Second, any technical articles on topics such as leveraging radio frequency identification technology and convergent mobile technologies were excluded. As a result, 80 papers remained after the initial selection round.

Then the second-round selection was made by screening the full text of the articles and assessing their quality.

Papers not written in English or translated by poor-quality translation software or just mentioning AE in a descriptive way, were also excluded from the review. Finally, 64 papers, including 41 journal articles and 23 conference papers, remained for review.

3. Descriptive analysis Descriptive details of the papers identified were extracted and analyzed according to the distribution of publication across the period, countries and areas, the quality of journals, and research methodologies (Supplementary Table S1).

3.1 Distribution of publications by year of publication As shown in Figure 1, the time period of the publications is between January 2000 and December 2015.

This starting point (2000) represents the beginning of the e-business century. We can see that the number of publications increased over time, reaching a peak in 2011 with 8 papers. The general trend in journal articles then stabilized. Most conference papers were presented after 2008.

International Food and Agribusiness Management Review

441

Zeng et al.

Volume 20, Issue 4, 2017

9

conference paper

Number of publications

journal article

7 6 5 4 3 2 1 0

20 00 20 02 20 03 20 03 20 04 20 05 20 06 20 07 20 08 20 09 20 10 20 11 20 12 20 13 20 14 20 15

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0156 - Tuesday, August 22, 2017 12:29:21 PM - IP Address:24.21.169.207

8

Year

Figure 1. Distribution of publications per year.

3.2 Distribution of publications across countries As shown in Figure 2, there appears to be some predominance from China (around 33% of the 64 selected papers and most of them are conference papers). USA is ranked second with 12 contributions (19%) and all are journal articles. Contributions were made mainly by the developed countries but we saw some from the new industrialized countries, i.e. South Korea and Taiwan. China, Brazil and Malaysia are the only three contributing countries from the developing world.

3.3 Distribution of publications by journals The 41 journal articles identified were classified, according to the journals in which they are published, under six disciplines (Table 1). ‘Agricultural Economics and Agribusiness and Management’ contributed the most (14 papers; 34%) to this topic. ‘Information Management and Electronic Commerce’ and ‘Operation and Supply Chain Management’ are both the second largest disciplines contributing nine articles (22%). Three articles were published by ‘Marketing’ journals; two papers published in the ‘Tourism Management’ journals and four papers in other disciplines (i.e. Sector Studies, Entrepreneurship and Technology Management), showing that the topic is cross-disciplinary in nature.

journal article

conference paper

Mixed Korea Brazil Singapore France Netherlands Malaysia Germany Greece UK Canada Australia Taiwan Italy USA China 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 Number of publication

Figure 2. Distribution of publications across countries.

International Food and Agribusiness Management Review

442

Zeng et al.

Volume 20, Issue 4, 2017

Table 1. Distribution by disciplines and journals.

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0156 - Tuesday, August 22, 2017 12:29:21 PM - IP Address:24.21.169.207

Area/journal

No. of articles

A – Agricultural Economics, Agribusiness and Management 14 Agribusiness 1 Agricultural and Resource Economics Review 1 American Journal of Agricultural Economics 3 International Food and Agribusiness Management Review 4 Journal of Agricultural and Resource Economics 1 Journal of Food Products Marketing 2 Journal of International Food & Agribusiness Marketing 1 Review of Agricultural Economics 1 B – Information Management and Electronic Commerce 9 International Journal of Business Information Systems 2 International Journal of Electronic Commerce 1 International Journal of Information Science and Management 1 International Journal of Security and Its Applications 1 Information and Management 1 Journal of Digital Information Management 1 Journal of Global Information Technology Management 1 Journal of Global Information Management 1 C – Operation and Supply Chain Management 9 Annals of Operation Research 1 International Journal of Business Performance Management 1 International Journal of Engineering Business Management 1 International Journal of Logistics: Research and Applications 1 International Journal of Physical Distribution and Logistics Management 1 Supply Chain Management 2 Journal on Chain and Network Science 1 The International Review of Retail, Distribution and Consumer Research 1 D – Marketing 3 Industrial Marketing Management 1 Journal of Business and Industrial Marketing 1 Journal of Interactive Marketing 1 E – Tourism Management 2 International Journal of Tourism Research 1 Journal of Vacation Marketing 1 F – Others (Sector Studies, Entrepreneurship and Technology Management) 4 British Food Journal 1 Environment and Planning A 1 International Journal of Entrepreneurship and Innovation Management 1 Technology in Society 1 1

SNIP1 2014

SSCI2 2015

0.834 0.572 1.194 1.021 0.801 0.419 0.591 1.134

X

0.628 1.738 0.085 0.980 2.087 0.296 0.259 0.715 1.124 0.260 0.722 0.858 1.509 1.842 0.368 0.600 1.579 0.911 2.185 1.560 0.959 0.690 1.277 0.543 0.601

X X

X

X X X

X X

X X X

X

SNIP = Source Normalized Impact per Paper, provided by Scopus. The larger value means the higher quality. Value n≥1 means that the journal is on or above average for its field and vice versa.

2 SSCI = Social Science Citation Index.

International Food and Agribusiness Management Review

443

Zeng et al.

Volume 20, Issue 4, 2017

In order to identify the quality of these journals, the value of Source Normalized Impact per Paper (SNIP) indicator in 2014 and whether the journal is covered by Thomson Reuters’ Social Science Citation Index (SSCI) in the 2015 version are presented in Table 1. In terms of SNIP indicator, only twelve journals have a numerical value larger than one (accounting for about 36% of the journals) and the mean value of all is 0.938. Journals with an impact factor of 1 or above are considered good quality in social science, so the overall quality of the journals is not high. On the other hand, only thirteen journals (less than 40% of the journals) are included in the list of SSCI. 3.5 Distribution according to the research methodologies The articles identified were coded according to their research methodologies into five categories as suggested

by Seuring and MĂźller (2008). Figure 3 presents the results. The survey appears to be the prevalent research method employed with 30 papers (almost half of the papers identified). 21 papers are of a rather theoretical or conceptual nature. The third most popular methodology is the case study with eight contributions. There are also four representations demonstrating new ideas by mathematical modelling. Only one selected paper is devoted to AE research via a literature review (Van der Vorst et al., 2002). For the survey-based studies, the sample size is not large on the whole, and for the conceptual-based studies, the theoretical foundation is often missing (Supplementary Table S1).

4. Thematic findings The main themes identified from the review are presented in this section and include factors affecting the firm-level adoption of AE, the firm-level adoption of AE, the firm-level performances of AE, factors affecting the regional development of AE, the regional development modes of AE, and the regional development impacts of AE. 4.1 Factors affecting the firm-level adoption of agri-food e-commerce The factors affecting the adoption of AE by the agribusiness firms identified are one of the major themes in existing research (Table 2). The influencing factors are broadly classified as internal and external dimensions.

Many internal factors related to a firm’s characteristics such as leader traits and business patterns are indicated as salient (Henderson et al., 2004, 2005; Janom and Zakaria, 2010; Kalaitzandonakes et al., 2003; Machfud and Kartiwi, 2013; Molla et al., 2010; Ng, 2005). Molla et al. (2010) highlight the fact that technology competence, financial commitment, perceived environmental e-readiness and organizational size are

35 30

30 Number of publications

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0156 - Tuesday, August 22, 2017 12:29:21 PM - IP Address:24.21.169.207

3.4 Distribution according to the quality of journals

25 20

21

15 10

8

5 0

4 1

Conceptuals

Case studies

Surveys

Models

Reviews

Figure 3. Research methodologies applied.

International Food and Agribusiness Management Review

444

Zeng et al.

Volume 20, Issue 4, 2017

Table 2. Factors affecting the Agri-food E-commerce (AE) adoption and the AE development.

Groups

Internal factors

External factors

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0156 - Tuesday, August 22, 2017 12:29:21 PM - IP Address:24.21.169.207

Adoption by agribusiness firms

Technology competence Financial commitment Perceived environment e-readiness Firm size Personal traits Perceived benefits Follow-up services Resources availability Target market segment Market scope Nature of products or services Organization structure and culture Types of business strategy Overall development of AE Sales channels Service quality Cognition and quality of peasants E-commerce talents Branding and standardization Industry structure Product complexity High-touch nature of transactions Network effects Competition among e-market

Traction trust and control in supply networks Strategic partners’ influence Competitions Market trends Industrial contexts Government

Logistics costs Customs fees Technological changes Accessing international market Accessing rural areas Consumers’ switch rate Urbanization rate

influential factors that directly affect e-business use. Technology context referring to firm’s investment and accumulation of e-enabling technologies (e.g. handling transaction security and confidentiality) is identified as an important factor by some authors (Janom and Zakaria, 2010; Kalaitzandonakes et al., 2003; Ng, 2005).

As for the organizational size, Molla et al. (2010) find that the smaller the size of a horticulture firm, the greater the extent of e-business use; however, Henderson et al. (2005) provide an opposite result, reporting that Internet strategies are more likely to be adopted by larger firms with a global scope. This mixed result shows that firm size has complex effects on adoption and may be a regulated or mediating variable rather than an independent variable. The factor of personal traits for small and medium agribusiness firms, such as educational level, entrepreneurial characteristics, business experience, feeling and attitude to e-commerce, is also proposed (Janom and Zakaria, 2010; Machfud and Kartiwi, 2013). Generally, perceived benefits can provide incentives for firms’ e-commerce adoption. For example, Henderson et al. (2004) disclose that firms perceiving greater logistics and inventory management gains are more engaged in e-commerce activities.

But the lack of follow-up services (product returns) is the biggest barrier to e-commerce as perceived by agricultural producers in Midwestern US (Kalaitzandonakes et al., 2003).

Through case analysis of Australian agribusiness firms, Ng (2005) develops a preliminary model for the selection of Business to Business e-business, including internal influencing factors such as resources availability, target market segment and market scope, nature of products or services, organizational structure and culture and types of business strategy and external influencing factors such as strategic partners’ influence, competitions and market trends.

E-commerce provides support for vertical coordination processes in food supply networks. However, with e-business there is a need to build on the communication of safeguards for trust and control to influence the transaction decision as a prerequisite for e-business adoption in food networks (Fritz and Canavari, 2008).

International Food and Agribusiness Management Review

445

Zeng et al.

Volume 20, Issue 4, 2017

O’Keeffe (2001) proposes that those firms in the perishable food industries who have already traveled some way down the partnership route will find such Business to Business e-commerce opportunities much more attainable than those who have not. Other external factors e.g. industrial contexts (regulatory environment, market structure, supporting industries) are also mentioned by authors such as Hsiao (2007) and Janom and Zakaria (2010).

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0156 - Tuesday, August 22, 2017 12:29:21 PM - IP Address:24.21.169.207

4.2 The firm-level adoption of agri-food e-commerce ■■ The firm-level strategies/models of agri-food e-commerce There are five typical e-commerce models commonly adopted nowadays: Business to Business (B2B, a model that facilitates the transaction between businesses); Business to Customer (B2C, a model that facilitates businesses selling directly to customers); Customer to Customer (C2C, innovative ways to allow customers to interact with each other); Customer to Business (C2B, a model that facilitates customers offering products or services to businesses or that enables customers to name their own prices for a specific product or service); and Online to Offline (O2O, a model that draws potential customers from online channels to physical stores) (Zhang and Ma, 2015). A few scholars have made comparative analysis of these models in the agri-food sector and evaluated their developing trends (Chou, 2011; Gui and He, 2011; Huang et al., 2009; Yang and Wang, 2015; Zhang and Huang, 2015). As for the application of B2B in AE, supermarkets will have a distinct advantage in the future, due to the rapid development of supermarket chains in China (Chou, 2011; Gui and He, 2011). Gui and He (2011) point out that there will be further development opportunities for B2C in AE with the improvement of consumption levels and the change in consumption patterns in China. There are two barriers to C2C application in China: first, the variety of agricultural products for sale online are very limited at present; second, most farmers do not yet have the skills to operate an online shop. Chou (2011) proposes that small agri-product processing enterprises (as customers) can apply the C2B model because it can reduce procurement costs. Huang et al. (2009) conducted a study on the C2B model applied to food souvenir products in Taiwan, and found that there are still some potential problems such as the high stock levels of products and raw materials and the complex ordering processes. They therefore developed a model that introduces a common platform which sends the order information to the factory, raw material suppliers and the shop at the same time and arranges the delivery to customers directly from the factory rather than from the shop. Zhang and Huang (2015) and Yang and Wang (2015) are in favor of the O2O community model, a new model of online-offline interaction, supported by logistics of cold chain with intelligent terminal pickup and delivery to residential communities.

■■ The firm-level tactics of agri-food e-commerce About 40% of the papers identified are centered on the firm-level tactics of AE and they can be divided into three groups: tactics of agricultural firm websites, tactics of third party e-commerce platforms, and tactics of offline supply chain management.

Many agribusiness firms conduct e-commerce by establishing and operating their own websites. Some authors put forward tactics of firm websites based on evaluating the website quality or identifying success factors for website-based e-commerce. Information is quite an important component of a firm’s website.

For example, Internet sites provided by agricultural input suppliers offer extraneous information including not only current market price, but also crop market forecasts, market trend analyses, technical information on weed identification, online interaction with experts and so on (Just and Just, 2006). It seems that simply establishing a point of presence on the web is not enough (Volpentesta and Ammirato, 2007). It is necessary to improve message clarity and content accuracy, provide information on technical subjects, and update information on the website on time (Andreopoulou et al., 2008; Bodini and Zanoli, 2011; Ernst and Hooker, 2006; Yu and Zhao, 2014). Communication is also a key aspect of a firm’s website. Two-way communication between both parties in the transaction is critical to successful individual relationships (Ernst and Hooker, 2006). The website designers should put much more emphasis on enhancing customer communication International Food and Agribusiness Management Review

446

Zeng et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0156 - Tuesday, August 22, 2017 12:29:21 PM - IP Address:24.21.169.207

services, especially allowing their visitors to communicate with the company or via ‘chat rooms’ with virtual or real person customer service (Andreopoulou et al., 2008; Tsekouropoulos et al., 2012). Yu and Zhao (2014) provide empirical evidence that service quality is the most important factor affecting the internet portal selections of online consumers. Other tactics like providing visitors with the possibility to tour the website as well as to be connected via this site with other websites, enabling the realization of commercial transactions, improving efficiency of navigation, paying attention to the optimization of search engines, increasing the exposure of websites, and pouring more money into Internet marketing, are also mentioned by some authors (Andreopoulou et al., 2008; Bodini and Zanoli, 2011; Yi and Rong, 2010).

Electronic commerce platforms are provided by a third party service supplier supporting online transactions between sellers and buyers. Some trade platforms are established by private firms, while others are funded partly or entirely by governments. There were 85 AE platforms/portals in the US and EU in 2000, but only 25 remained active in 2002. Fritz et al. (2004) identify a range of strategic development best practices of the successful platforms, which include initiating cooperation with other platforms, gaining support from major market participants, improving trading functionalities, and expanding value-added services. Yang et al. (2008) and Lu et al. (2011) suggest that the local government should perfect the function of online transaction, and continuously improve the functional innovation, technological innovation and management innovation of the public platform. A study on the US-based electronic commerce platform called MarketMaker shows that encouraging producers to update their site profiles frequently, specifically their contact information and attributes and the availability of their products, is needed for further development (Zapata et al., 2013).

Firms should bear in mind that e-commerce requires a strong business plan, a sound business structure and a carefully managed supply chain (Montealegre et al., 2007). Ha and Hong (2005) propose a supply chain partner selection process system for e-collaboration and apply it to the food supply chain. Bacarin et al. (2008) construct a framework for contract e-negotiation in agricultural supply chains. A strategic decision-support framework is designed by Stritto and Schiraldi (2013) for food and beverage e-supply chain management.

Hung et al. (2010) propose an organic e-farming system based on three business value models which they call value chain, value shop, and value network, for the purpose of creating high-end agricultural businesses in the era of the Internet. Van der Vorst et al. (2002) argue that the value chain business-web is likely to become widespread in food supply chains due to the importance of tracking and tracing of products related to food safety and environmental issues and the presence of quality certification programs in these supply chains. Some authors also put forward an information-sharing mechanism in the agricultural products supply chain enabled by e-commerce (Li and Gao, 2011). With the available information technology, it is feasible for companies in food supply chains to implement systematic planning and response measures.

Kinsey (2000) gives a brief introduction to cooperative planning, forecasting and replenishment, a system that involves manufacturers and retailers each using real-time data to forecast future demand, share their forecasting information and agree to deliver and to receive merchandise on a prearranged schedule. Survey results indicate that full participation in e-marketplaces requires enterprises to integrate their internal and external supply chain activities and share strategic information (Eng, 2004). It is also important to have suppliers and distributors show enthusiasm and sincerity in their services (Chen et al., 2014). Logistics distribution and home delivery are also discussed. Online organic home delivery may be the most successful type of online food retailing; for agricultural product logistics it is advisable to implement a quick response distribution task management tool; Internet-only food-retailing companies should fulfil orders by at least having warehouses; building a comprehensive information platform is required for agricultural products logistics (Cai et al., 2015; De Koster, 2002; Murphy, 2003; Zhang and Zhang, 2013).

4.3 The firm-level performances of agri-food e-commerce Financial, online marketing and cross-border market entry performance are reported here. Motiwalla et al.

(2005) present an intra- and inter-industry financial performance analysis and their comparative results show that the electronic business (EB) companies did better than their non-EB counterparts in the post-EB period and the food, beverages and tobacco sector performed better than the retail and consumer products sectors.

International Food and Agribusiness Management Review

447

Zeng et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0156 - Tuesday, August 22, 2017 12:29:21 PM - IP Address:24.21.169.207

An evaluation index system of agri-products’ online marketing performance was designed by Jiang and Wang (2009), and included 12 aspects: return on net assets, growth rate of sales profit, customer satisfaction, inventory turnover ratio, safety reliability, web page hits, website social popularity, consumer loyalty, average retention time per visitor, function comprehensiveness, information update frequency, and style uniqueness.

This provides a basis for studying how to improve network marketing performance for agribusiness firms.

It is generally accepted that e-commerce can provide small and medium enterprises and bigger companies alike with the same access to global markets. But Giustiniano and Fratocchi (2002) find that the Internet is yet to become a tool for promoting the internationalization process for Small and medium-sized enterprises.

By analyzing the data collected from wine and olive industries in Italy, three reasons for this are revealed: first, most of the firms only approached the Internet in 2000 and have not yet developed a significant level of competence with respect to this new strategy; second, the companies seem not to have implemented Custom Relationship Management tools; finally, Internet has been considered merely as an adjunctive distribution channel.

4.4 Factors affecting the regional development of agri-food e-commerce Some papers discuss the factors affecting the regional development of AE (Table 2). The factors from within the agri-food industry affecting the regional development of AE are seen as internal factors, and those from outside agriculture are defined as external factors. Three dominant internal factors are proposed by Leroux et al. (2001) including agri-food industry structure, product complexity, and the high-touch nature of transactions (i.e. building close relationships with clients). Mueller (2001) argues that two internal causes will drive the evolution of e-commerce: network effects (i.e. the value of good changes increases because of the increasing number of people using the good changes) and competition among markets for patronage and trading volume. Both Gan et al. (2011) and Jiong et al. (2013) applying AHP method reveal that the most important factors to the AE development are logistics costs, cognition and information quality of peasants, sales channels and service quality. Wu et al. (2010) identify two additional key factors, i.e. talents, branding and standardization in agri-products: severe shortage of e-commerce talents poses a barrier to the regional development of AE, and the construction of branding and standardization in agricultural products should be speeded up. In addition, more investment is needed for the expansion of AE; however, the exact timing of an investment is a key decision. To address this, Wang et al. (2015) propose an evolutionary discounted cash flow model showing that the optimal investment time depends on the consumers’ switch rate from the physical store to e-store and the urbanization rate. Other external factors such as customs fees, technological changes and accessing international markets are also put forward by some scholars (Boyd et al., 2003; Hobbs et al., 2003; Mueller, 2001).

4.5 The regional development modes of agri-food e-commerce Some academics propose new development modes for the regional agri-food industry based on e-commerce.

Wang (2010) presents a government-driven mode in which governments assume an intermediary role between farmers and businesses thanks to their credibility and authority. In addition, the government not only assumes the responsibility of infrastructure investment and the improvement of e-commerce market regulation but also provides public services (e.g. information delivery, production guidance, and training) to farmers and agri-food enterprises. Kang et al. (2010) propose another type of government-driven mode, in which governments establish regional agricultural e-commerce centers, public platforms and information database affiliated to the government. Zhao and Tian (2014) identify a service provider driven mode in which a local service provider is responsible for bringing the products of individual farmers and agricultural cooperatives together, putting them in consolidated warehouses and helping sell products via an e-commerce platform. The cooperative-driven mode indicates that the agricultural cooperatives buy in and sell farmers’ products on their own e-commerce portals and facilitate the two-way information flow between farmers and consumers (Liu and Li, 2011; Liu et al., 2011).

International Food and Agribusiness Management Review

448

Zeng et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0156 - Tuesday, August 22, 2017 12:29:21 PM - IP Address:24.21.169.207

4.6 The regional development impacts of agri-food e-commerce The rapid development in e-commerce can bring about some changes in the demand for labor. An input-output model is used by Schluter and Lee (2005) to examine the effects of e-commerce boom on the demand for high-skilled and low-skilled workers and they find that food and agricultural trade has reduced low-skilled labor demand in the US because the New Economy emphasizes knowledge-based labor practices. Huang (2006) conducts a study on the revitalization of the leisure farm industry by implementing an e-commerce strategy and finds that the revolution of e-commerce in leisure agriculture does impact traditional leisure farm businesses and challenges all traditional travel suppliers. All leisure farm owners should improve their service ability and build marketing alliances through e-commerce development in order to seize resource and achieve performance advantages (Huang, 2006). Based on a survey, Baourakis et al. (2002) tried to examine the impact of e-commerce on agri-food marketing in Crete and found that most of the cooperatives and firms gained little benefit as they only used the Internet for e-mail communication rather than for transactions or other important activities, i.e. online banking, bill-paying, B2B trading, or supply chain management.

5. Discussion 5.1 Descriptive findings Generally, there was an upward trend in the number of publications from 2000 to 2015, showing that AE research has attracted more and more attention from academics. However, the number and sub-optimal quality of the selected papers shows that academic research lags behind the rapid development of AE practice.

In terms of the distribution across countries, there was some predominance from China and the USA. However, around 85% of the articles written by Chinese authors were only presented at international conferences, with few citations. In contrast, all of the papers published by American scholars were journal articles of relatively higher quality, the majority of which were published before 2007.

By analyzing the distribution of publications across disciplines, agricultural economics and agribusiness management occupies a dominant position in terms of number of publications and is followed by operations management (OM) and information systems (IS). However, the impact factor of AE journals is generally low.

It is envisaged that there is great potential for OM and IS scholars to contribute to this topic. We are aware that the literature on non-agricultural e-commerce is increasing in IS literature. With the fast development of AE practice, it is expected that many IS researchers will be attracted to this field. It is also advisable for scholars researching agricultural supply chain from an OM/Supply Chain Management background to enter into this promising field of research.

As for the research methodologies used, there are 21 papers adopting a theoretical or conceptual approach without empirical evidence. But the theoretical foundation is often missing. Case studies and modelling seem to be under-represented. It is proposed that more case studies should be carried out to explore this field.

Furthermore, none of the selected papers carried out a literature review specifically about AE. 5.2 Development of an integrated model To synthesize the thematic findings and signpost some future research directions, we developed a conceptual model based on the literature gaps and industry reports on AE practice in China where people have witnessed a rapid development of AE in recent years. The Chinese central and local governments have paid unprecedented attention to the AE development through policy and public investment. As presented in Figure 4, the model comprises two basic parts: one is built around AE adoption at a firm-level, and the other looks at development modes at a regional level. The relationship between them is reciprocal: the firm-level adoption provides a micro-level foundation for the regional level AE development, and the regional development can lead to the widening and deepening of firm-level adoption and multiplying effects.

International Food and Agribusiness Management Review

449

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0156 - Tuesday, August 22, 2017 12:29:21 PM - IP Address:24.21.169.207

Zeng et al.

Volume 20, Issue 4, 2017

Factors •Internal factors e.g. Resources availability; Perceived benefits; Follow-up services; Technology competence; ...

•External factors e.g. Competitions’ influence; Market trends; Industrialcontexts; Government; ...

Adoption of AE (firm-level) Adopters •Agribusiness firms •E-commerce firms (platform suppliers; service suppliers) •Agricultural Cooperatives (including farmer associations) •Individualfarmers(forsellingproducts)

Performances

Strategies

•Financial

Factors •Internal factors e.g. Industry structure; Product complexity; Network effects; E-commerce talents; ...

•External factors e.g. Logistics costs; Customs fees; Technological changes; Urbanization rate; ...

Development modes of AE (regional level)

•B2Bmodel •B2Cmodel •C2Cmodel •C2Bmodel •O2Omodel

Tactics •Online marketing Firm websites; Trade platforms; Online shops •Offline supply chain management Procurement; Production; Processing; Logistics/delivery

•County-level AE mode Government driven mode Service provider driven mode Rural entrepreneur driven mode •Village-level AE mode Smallholders driven mode Cooperatives driven mode

•Online marketing •Operational

•Cross-border market entry

Impacts •Agriculture •Peasants e.g.Income; Consumption; Mindset; ...

•Rural economy/society e.g.Economic growth; Supporting industries; Poverty alleviation; Employment; Left-behind generation; ...

Figure 4. A conceptual model of agri-food e-commerce (AE). Models: B = Business; C = Customer; O2O = Online to Offline.

At a firm-level (upper part of Figure 4), there are three boxes: factors affecting AE adoption, AE adoption and associated performance. Within the adoption of AE box, adopters indicate the subjects that adopt or drive the adoption of AE and tend to be agribusiness firms, e-commerce firms (e.g. Alibaba), agricultural cooperatives and individual farmers. These adopters can choose various business strategies/models (e.g.

B2C), which determine the tactics (i.e. online marketing and offline supply chain) used. Online marketing consists of a number of tactics which vary for firm websites, trade platforms or online shops. After deciding a business model, the AE adopter faces the issue of configuring its offline supply chain, including but not confined to procurement, production, processing, logistics and delivery.

The influencing factors are classified as internal (e.g. resource availability) and external dimensions (e.g. government policy) and affect the level of AE adoption at a firm level. We propose four approaches to evaluating the performance of AE adoption, including online marketing, operation, finance, and cross-border market entry. Marketing and financial performance are found in the reviewed papers. Operational performance is included because it is affected by AE adoption, for example on-time delivery. Cross-border market entry is proposed as a performance measure because there is an emerging phenomenon of cross-border e-commerce which changes the traditional internationalization process of firms.

As for the regional level (lower part of Figure 4), we identify two development modes of regional AE from industry reports and news coverage on China’s AE practice: county-level mode and village-level mode. We summarize that county-level AE refers to a mode in which a whole county1 is committed to AE development of normally a few agricultural products (e.g. pecan), carrying out strategic planning and prioritizing financial allocation to help the product gain access to national or even international markets. Three county-level modes are identified based on the drivers of the AE development in the initial stage: government-driven mode, service provider driven mode, and e-businessmen driven mode.

1

China has four levels of local governments, i.e. province, county, township and village.

International Food and Agribusiness Management Review

450

Zeng et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0156 - Tuesday, August 22, 2017 12:29:21 PM - IP Address:24.21.169.207

The village-level AE mode refers to a kind of AE development pattern emerging in certain villages in China, where quite a number of smallholders or agricultural cooperatives located within their villages sell their agriproducts via Internet, which is the main source of income to the villages (AliResearch, 2015). The concept of Taobao Village first appeared in the news in 2009 and Alibaba Group officially adopted it in 2013. The term Taobao County emerged in 2014. Both concepts indicate those villages and counties that meet a certain criterion set by Alibaba selling their products through the Taobao e-portal. The very first batch of Taobao Villages were recognized and presented with a plaque by Alibaba in December 2013.

Again, the influencing factors are classified as internal (e.g. industry structure) and external dimensions (e.g. technological change), which affect the level of AE development and the choice of development modes in a region. Based on China’s ‘San Nong’2 policy, the impacts of regional AE development can be classified as follows: agriculture, peasants, and rural development.

5.3 Future research directions ■■ Four types of adopters We identify four types of adopters at the firm-level, i.e. agribusiness firms, e-commerce firms, agricultural cooperatives, and individual farmers. For the factors affecting the adoption of e-commerce, existing scholars only focus on one type of adopter, i.e. the agribusiness firms, while ignoring other types. Some Chinese e-commerce companies/service providers, such as Alibaba, began to put a lot of emphasis on exploiting the agricultural e-markets in rural areas. For instance, Alibaba tries to promote the sale of agricultural products from rural to urban areas by establishing Rural Taobao Service Stations (‘Cun Tao’ in Chinese) in many villages. According to the statistics of AliResearch (an internal research institute established by Alibaba), 9,278 such stations were established by December 9th, 2015 (AliResearch, 2015). The number of farmer cooperatives in China has increased fast since the enactment of the Specialized Farmers Cooperatives Law in 2007, and some cooperatives have tried to adopt e-strategies in recent years. In addition, more and more Chinese smallholders sell their agri-products directly to consumers online. It presents a trend for the development of C2C e-commerce in agriculture. Hence, much more research should be devoted to other types of adopters in addition to agricultural firms.

■■ Implementation of online and offline marketing The implementation of online marketing and offline supply chain management for various adopters is still not well understood. More research is needed to understand the best practices of online marketing. Most of the existing literature is focused on the success factors of the firm websites and only four papers discuss those of trade platforms, while few articles focus on the tactics of online shops. In China, more and more agribusiness firms tend to open online shops on third party trade platforms because their own websites do not have transaction functions and/or the desired number of visits (Cui et al., 2014). Some agricultural cooperatives and individual farmers are also active in opening and operating their own online shops. With the rapid increase in the number of online shops, the competition in online marketing among the adopters is considerably fierce (Zeng et al., 2015). Therefore, making online shops attract visitors, in order to survive and set a foot in the hyper competitive e-marketplace and then look for further expansion is a critical issue that scholars should investigate in the near future. Another important tactic is how to improve the supply chain efficiency after AE adoption. It is expected that, after the adoption of e-commerce, many changes will have to be made in every process of the food supply chain including procurement, production, processing, logistics and delivery. However, few articles make a systematic and comprehensive analysis of the tactics of every process of the whole food supply chain enabled by e-commerce for agribusiness firms. This is an obvious gap which researchers could address. For greater performance, the adopters should be aware of

2

China’s ‘San Nong’ is concerned about agriculture, peasants and rural development. All three words have an initial character of ‘Nong’ in Chinese.

International Food and Agribusiness Management Review

451

Zeng et al.

Volume 20, Issue 4, 2017

the need to align online marketing with offline supply chain management and adjust one or both constantly according to the dynamic interaction between them.

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0156 - Tuesday, August 22, 2017 12:29:21 PM - IP Address:24.21.169.207

■■ Quantifying performances of agri-food e-commerce There is a lack of a commonly-used index to quantify the performances of AE activities and potential links between specific tactics and performances at a firm level. We have identified four categories of AE performance: online marketing, operation, finance, and cross-border market entry at a firm level. Operational performance weighs the benefits that firms have gained from offline food supply chain management after the adoption of e-commerce and can be evaluated by many aspects such as production costs, lead time, inventory stock-out, delivery reliability and the flexibility of the process (Devaraj et al., 2007). Financial performance is an ultimate goal and the most important trigger for firms that wish to implement e-commerce programs. It is logical to infer that positive online marketing performance and operational performance leads to positive financial performance (Jiang and Wang, 2009; Motiwalla et al., 2005). Cross-border market entry performance represents the market access capability and cost reduction of firms in the internationalization process with the help of e-commerce. On the one hand, e-commerce allows retailers to conduct global sourcing and marketing (Zwass, 1998); on the other hand, the cross-border e-tailers need to address some challenges, such as custom procedures, cultural adaption, and dispute resolution (Sinkovics et al., 2007; Stylianou, 2008). In addition to the unified index systems and measurements to evaluate the performances of AE activities, there is a clear research need to establish the potential links between specific AE tactics and enhanced online marketing, operational, financial and cross-border market entry performance, so as to provide more accurate and practical measures for organizations to adopt e-commerce.

■■ Regional development modes of agri-food e-commerce Much emphasis should be put on the emerging phenomenon of regional development modes of AE. China’s best practices in recent years have provided a great opportunity for the further development of AE in the developing world and have presented new opportunities for AE research. This is evidenced by the emergence of county-level and village-level modes.

There are three kinds of county-level modes identified, which we can illustrate with live examples from industry reports or news coverage. Cheng County, a small county in the mountainous area in Gansu Province in China, is a representative sample for the government-driven mode. Without the leadership of the county governor, Li Xiang, it would be hard to imagine AE developing so rapidly there (Wei, 2014). Cheng County is famous for some local agri-products such as pecan, garlic, honey and pork. However, before the advent of AE, these products were only sold to local agricultural product market or traders at low prices. In June 2013, Li tried to sell pecans via a micro-blog online and attracted many consumers, successfully helping some local peasants solve the problem of market access. Then the county government decided to promote the development of network marketing by enacting policies and taking measures such as engaging in public relationships with media, providing training to peasants, supporting the establishment of Cheng County E-commerce Association, building Longnan E-commerce Incubation Park and Cheng County E-commerce Logistics Park, and signing a cooperation contract with Alibaba for setting up Rural Taobao Service Stations in the villages. The AE development was so successful that there were 676 online shops with an annual revenue of 370 million RMB3 by the end of 2015 and Cheng County was named the ‘National Comprehensive Demonstration County of Introducing E-commerce into Rural Areas’ by the Ministry of Finance and the Ministry of Commerce in July, 2015.

Tongyu County, a county in Jilin Province in North China, developed another mode called service provider driven mode. Before the adoption of AE, Tongyu County was an underdeveloped county due to its geographical location and resource restraints. In recent years, with the help of a service provider called Yunfeihewu 3

RMB = Chinese yuan. 1 Chinese yuan = 0.145285 USD, calculated on the basis of the exchange rate on April 13, 2017.

International Food and Agribusiness Management Review

452

Zeng et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0156 - Tuesday, August 22, 2017 12:29:21 PM - IP Address:24.21.169.207

E-commerce Company (YEC), Tongyu County has made great progress. YEC was established in October 2013, as a local service provider responsible for selling local agricultural products online. YEC put much emphasis on building a well-known brand called ‘San He Mu’ and provided ample guidance to local agricultural producers on how to standardize production and offer higher prices than before for purchasing agri-products from farms. YEC applied the unified package, put on their own-brand labels, and sold the products online via Taobao platform through its professional marketing team. Tongyu’s practices provide great evidence of how to develop AE in less developed areas without any prior knowledge of e-commerce (Mo, 2015). In this case, the government assumed a supportive role.

Suichang County, a county in Zhejiang Province in China, promotes AE development through a mode called rural entrepreneur driven mode, which Avgerou and Li (2013) call netpreneurer. Suichang County has a superior ecological environment and produces high-quality local agri-products such as yam, camellia oil, alpine vegetables, free range chicken, and edible fungi. However, due to the small production scale, the peasants’ position remains weak in the supply chains. In 2005, several smallholders began trying to sell pecans through online shops spontaneously and attracted more and more participants. In March 2010, Suichang E-businessmen Association (SEA) was established thanks to the strong push of Pan Dongming, an excellent entrepreneur who returned to Suichang County from Shanghai after quitting a highly paid job. He then became the president of SEA. As a non-profit organization, SEA is set up to provide public services (e.g. growing technique advice, training, pooling common resources, information sharing and negotiation with logistics companies) and ensures industry self-regulation. Being registered with SEA, two enterprises called Suiwang E-commerce Company (SEC) and Ganjie E-commerce Company (GEC) were founded in 2011 and in 2013 respectively. Pan became the principal business partner of both, providing strategic advice. SEC has a profound influence on promoting AE development in Suichang County, serving as a comprehensive service provider similar to YEC in Tongyu County. GEC set up Ganjie E-commerce Stations in villages helping peasants purchase/sell products online. Under the guidance of SEA, AE in Suichang County is developing fast, making Suichang County well-known throughout China.

We have also identified two different types of village-level AE modes: smallholder-dominated and cooperativeoriented mode. Noticeably, more and more Chinese smallholders are selling their agricultural products directly to urban consumers, mostly through e-shops at Taobao’s online portal. When the number of these smallholders in the same village reaches a certain level, Alibaba company recognizes this village as an AgroTaobao Village. It is a new kind of industry cluster based on C2C e-commerce and located in rural areas, which has rarely been seen in other countries (Zeng et al., 2015).

In some villages, given the difficulties of AE adoption by local individual peasants, farmer cooperatives (e.g.

Akesu Feng Guo Agricultural E-commerce Cooperatives in Xinjiang, Hua Sheng Xiang Grape Cooperatives in Shandong Province) represent a different mode from the smallholders driven mode. In a cooperative-driven mode, peasant e-tailers collaborate through co-operatives and sell products online to non-local wholesalers in bulk or consumers piecemeal through the co-operative or third party e-portal. Chen et al. (2015) summarize three approaches to implementing e-commerce by farmer cooperatives: releasing product information on the public agricultural websites; self-operating online shops; and entrusting third party e-commerce trade platforms.

For the regional AE development modes, there is an imminent need for researchers to keep observing and summarizing the best practices and theorizing them. In practice, successful experience and policies from the best practices should be summarized and recommended to other countries, particularly those in the developing world. Due to the exploratory and under-researched nature of the topic, a grounded theory case study method may be applied in future research reporting the ‘best practice’ case studies, including the mechanisms behind the complex phenomena, which help theorize at a later stage. Some typologies of different modes may be developed.

International Food and Agribusiness Management Review

453

Zeng et al.

Volume 20, Issue 4, 2017

■■ Impacts of agri-food e-commerce development

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0156 - Tuesday, August 22, 2017 12:29:21 PM - IP Address:24.21.169.207

Few papers were found focusing on the impacts of AE development. Future researchers should first of all develop and validate the measures for impact and then focus on empirically testing the impacts of the AE development. China’s AE practices have revealed that AE development has many positive economic and social impacts. Zeng et al. (2016) argue that e-commerce is becoming a new engine in solving China’s ‘San Nong’ issues. Based on the tenets of ‘San Nong’, there are three aspects of agriculture, peasants, rural societies or communities.

Agriculture is the foundation of the national economy but still the least developed industry in China. The modernization of agriculture is a critical issue for policy makers. The emerging best practices show that e-commerce can be an effective approach. With the help of e-commerce, agri-products can be sold at higher prices, while getting rid of the constraints of space and time (Mo, 2015). The online marketing competition also compels producers to improve the quality of agri-products and the efficiency of the food supply chain.

The determining factor for the improvement of competitiveness lies in offline production and supply chain capability in the long run (Zeng et al., 2016).

The peasant issue is a central one within the ‘San Nong’ framework, having attracted much attention from academics, due to the fact that the economic and social gap between peasant class and non-peasant class is getting wider. In recent years, some peasants have become richer and happier by selling products in their own online shops (AliResearch, 2015; Zeng et al., 2016). More and more peasants are learning how to shop online, and their mindsets towards Internet and information are gradually changing (AliResearch, 2014; Zeng et al., 2015).

E-commerce can also be very useful for promoting the development of the rural economy and society in such facets as economic growth, poverty alleviation and employment. According to the report on e-commerce and poverty alleviation published recently, 832 nationally-underdeveloped counties sold products via Alibaba’s platforms for a total of 12 billion RMB yuan in 2014 (Zhang and Jiang, 2015). It is estimated that there were more than 70 thousand households or 280 thousand individuals (on average four employees per household) in employment in the 212 Taobao Villages by late December, 2014 (AliResearch, 2014). In addition, e-commerce can give an impetus to its supporting industries like network infrastructure, road reconstruction, logistics, finance, education, technology and online services (Cai et al., 2015; Cui et al., 2014; Zeng et al., 2015). For example, with the emergence of Taobao Village, logistics companies seeking economies of scale establish many new distribution hubs and collection points in rural areas. In addition, AE development in rural areas can attract migrant workers in cities back to their hometowns and promote local employment, helping to resolve many social problems such as the left-behind generation (AliResearch, 2014).

For the positive effects of the regional AE development on the ‘San Nong’ mentioned above, it is expected that future researchers could make an effort to explore the possible mechanisms that may lead to the positive impact on San Nong with sound theoretical foundation. For example, the positive impact on farmers’ income of AE adoption may result from different channels such as the sales growth, investment in production and operation, profitability and the increase in absolute working hours. These promoting mechanisms based on rigorous analyses and when well-understood can provide meaningful guidance for the future development of AE. On the other hand, rigorous empirical designs and analyses based on a large sample size are necessary to improve our understanding of the more accurate effects of AE. To draw valid causal inferences without estimate bias caused by endogeneity, some more rigorous methods such as propensity score matching can be used to study the adoption and effects of AE at the household or community level.

Every coin has two sides. It is also necessary for researchers to identify the potential negative effects of AE development and find out how to cope with them. For example, some opinion leaders in China are concerned by the fact that peasants have to invest too much money in the operation of their e-shops in the form of advertisement, and the competition for ranking shops on the Taobao platform is increasingly fierce International Food and Agribusiness Management Review

454

Zeng et al.

Volume 20, Issue 4, 2017

(Zhang and Ma, 2015). This situation has eroded the profits of e-shop owners and even forced some to exit the market. Furthermore, there are some challenges in implementing AE, i.e. the lack of AE talents, the imperfections of the logistics system and the low degree of agri-food standardization (Mo, 2015; Wu et al., 2010). Finding solutions to these problems will be a daunting task for future researchers.

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0156 - Tuesday, August 22, 2017 12:29:21 PM - IP Address:24.21.169.207

6. Conclusions It has been indicated that the adoption and development of e-commerce is an innovative way of influencing food systems and market access for smallholders. As far as we are aware, this may be the first study to systematically review the literature on AE. Based on research at Scopus, 41 journal articles and 23 conference papers were finally identified. Descriptive analysis results show that there has been a general increase in the number of publications since 2000, indicating that AE research has attracted more and more interest from scholars in different countries and across multiple disciplines. The 64 articles identified were coded and summarized into a number of themes. We contribute to the agricultural e-commerce research by developing a conceptual model based on a systematic review and some industry reports, and proposing a number of future research directions. In particular, the regional development modes of AE and their impacts on agriculture, peasants, and rural economy and society pose significant agri-food policy implications for the developing countries.

Market access for smallholders is a long-term topic in agricultural economics literature, and e-commerce is an eye-opening experience and revolutionary change for smallholders, since it could disintermediate middlemen including traders, distributors, and retailers removing information asymmetry. Policymakers in developing countries may learn from the best practices in China, and more research is needed to understand the regional development of AE.

Supplementary material Supplementary material can be found online at https://doi.org/10.22434/IFAMR2016.0156.

Table S1. List of the selected papers and demographic information.

Acknowledgements This study was supported by two National Science Foundation of China grants (Grand No. 71473218 and No.71673244) and Newton Caldas Institutional Link grant (Grand No. 172727857) funded by the British Council.

References Abebe, G.K., J. Bijman, R. Kemp, O. Omta and A. Tsegaye. 2013. Contract farming configuration: smallholders’ preferences for contract design attributes. Food Policy 35: 14-24.

AliResearch. 2014. Report on China’s Taobao village (2014). Presented at the summit forum on the 2014th China’s Taobao Village. Available at: http://tinyurl.com/kymcsgs (in Chinese).

AliResearch. 2015. Report on China’s Taobao village (2015). Presented at the summit forum on the 2015th China’s Taobao Village. Available at: http://tinyurl.com/ldhvnj3 (in Chinese).

Andreopoulou, Z., G. Tsekouropoulos, T. Koutroumanidis, M. Vlachopoulou and B. Manos. 2008. Typology for e-business activities in the agricultural sector. International Journal of Business Information Systems 3: 231-251.

Avgerou, C. and B. Li. 2013. Relational and institutional embeddedness of web-enabled entrepreneurial networks: case studies of netrepreneurs in China. Information Systems Journal 23: 329-350.

Bacarin, E., R.M. Madeira and C.B. Medeiros. 2008. Contract e-negotiation in agricultural supply chains.

International Journal of Electronic Commerce 12: 71-97.

Baourakis, G., M. Kourgiantakis and A. Migdalas. 2002. The impact of e-commerce on agro-food marketing: the case of agricultural cooperatives, firms and consumers in Crete. British Food Journal 114: 580-590.

International Food and Agribusiness Management Review

455

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0156 - Tuesday, August 22, 2017 12:29:21 PM - IP Address:24.21.169.207

Zeng et al.

Volume 20, Issue 4, 2017

Batte, M.T. and S. Ernst. 2007. Net gains from net purchases? Farmers’ preferences for online and local input purchases. Agricultural and Resource Economics Review 36: 84-94.

Bodini, A. and R. Zanoli. 2011. Competitive factors of the agro-food e-commerce. Journal of Food Products Marketing 17: 241-260.

Boyd, S.L., J.E. Hobbs and W.A. Kerr. 2003. The impact of customs procedures on business to consumer e-commerce in food products. Supply Chain Management 8: 195-200.

Cai, Y., Y. Lang, S. Zheng and Y. Zhang. 2015. Research on the influence of e-commerce platform to agricultural logistics: an empirical analysis based on agricultural product marketing. International Journal of Security and Its Application 9: 287-296.

Chen, C., C. Liang and H. Yueh. 2014. The business policy and strategy of online marketing: a case study of the Sinyi Township Farmers Association in Taiwan. International Conference on Business and Information. Available at: https://tinyurl.com/m366x6k.

Chen, D., D. Qiu, H. Guo and Q. Liang. 2015. Comparative analysis on the e-commerce modes implemented by cooperatives. China Farmer Cooperatives 7: 25-26 (in Chinese).

Chou, H. 2011. A comparative analysis of e-commerce patterns of agricultural products in China. International Conference on Business Management and Electronic Information. Available at: http://tinyurl.com/ lekada7.

Chu, J., M. Arce-Urriza, J.J. Cebollada-Calvo and P.K. Chintagunta. 2010. An empirical analysis of shopping behavior across online and offline channels for grocery products: the moderating effects of household and product characteristics. Journal of Interactive Marketing 24: 251-268.

Cui, L., L. Wang and J. Wang. 2014. Empirical analysis on the factors of social innovation in promoting the development of Taobao village: taking the city of Lishui in Zhejiang as example. China Rural Economy 31: 50-60 (in Chinese).

De Koster, R.B.M. 2002. Distribution structures for food home shopping. International Journal of Physical Distribution and Logistics Management 32: 362-380.

Devaraj, S., L. Krajewski and J.C. Wei. 2007. Impact of e-business technologies on operational performance: the role of production information integration in the supply chain. Journal of Operations Management 25: 1199-1216.

Duriau, V.J., R.H. Reger, M.D. Pfarrer. 2007. A content analysis of the content analysis literature in organization studies: research themes, data sources, and methodological refinements. Organizational Research Methods 10: 5-34.

Easterby-Smith, B.M., R. Thorpe and A. Lowe. 2010. Management research: an introduction. Sage Publications, London, UK. Eng, T.Y. 2004. The role of e-marketplaces in supply chain management. Industrial Marketing Management 33: 97-105.

Ernst, S. and N.H. Hooker. 2006. Signaling quality in an e-commerce environment: the case of an emerging e-grocery sector. Journal of Food Products Marketing 12: 11-25.

Fink, A.G. 2005. Conducting research literature reviews, the internet to the paperback. Sage Publications, London, UK. Fritz, M. and M. Canavari. 2008. Management of perceived e-business risks in food-supply networks: e-trust as prerequisite for supply-chain system innovation. Agribusiness 24: 355-368.

Fritz, M., T. Hausen and G. Schiefer. 2004. Developments, development directions of electronic trade platforms in US and European agri-food markets: impact on sector organization. International Food and Agribusiness Management Review 7: 1-21.

Gan, W., T. Zhang and Y. Zhu. 2011. On countermeasures of promoting agricultural products’ e-commerce in China. IFIP Advances in Information and Communication Technology. Available at: http://tinyurl. com/l4vrrzf.

Giustiniano, L. and L. Fratocchi. 2002. The virtual internationalisation process of Italian SMEs in the food industry. International Journal of Business Performance Management 4: 231-247.

Gui, L. and Z. He. 2011. Analysis of agricultural products e-commerce models based on supply chain management. International Conference on E-Business and E-Government. Available at: http://tinyurl. com/kmfv94q.

International Food and Agribusiness Management Review

456

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0156 - Tuesday, August 22, 2017 12:29:21 PM - IP Address:24.21.169.207

Zeng et al.

Volume 20, Issue 4, 2017

Guo, H. and R.W. Jolly. 2008. Contractual arrangements and enforcement in transition agriculture: theory and evidence from China. Food Policy 33: 570-575.

Ha, S.H. and G.H. Hong. 2005. Selecting supply partners for e-collaboration in supply chains. IFIP Advances in Information and Communication Technology. Available at: http://tinyurl.com/mg2ed93.

Hazell, P., C. Poulton, S. Wiggins and A. Dorward. 2010. The future of small farms: trajectories and policy priorities. World Development 38: 1349-1361.

Henderson, J., F. Dooley and J. Akridge. 2004. Internet, e-commerce adoption by agricultural input firms.

Review of Agricultural Economics 26: 505-520.

Henderson, J., F. Dooley, J. Akridge and A. Carerre. 2005. Adoption of internet strategies by agribusiness firm. International Food and Agribusiness Management Review 8: 42-61.

Hobbs, J.E., S.L. Boyd and W.A. Kerr. 2003. To be or not to B-2-C: e-commerce for marketing specialized livestock products. Journal of International Food and Agribusiness Marketing 14: 7-20.

Hsiao, R.L. 2007. Misaligned market: the importance of industry context in technology-mediated exchanges.

Journal of Global Information Management 15: 69-87.

Huang, L. 2006. Rural tourism revitalization of the leisure farm industry by implementing an e-commerce strategy. Journal of Vacation Marketing 12: 232-245.

Huang, T.C., T.J. Lee and K.H. Lee. 2009. Innovative e-commerce model for food tourism products.

International Journal of Tourism Research 11: 595-600.

Hung, C.L., T.Y. Yu and C.H. Huang. 2010. Incorporating business value models into organic e-farming system. International Conference on Management of Innovation and Technology. Available at: http:// tinyurl.com/lqg5qzf.

Janom, N. and M.S. Zakaria. 2010. The development of B2b e-commerce readiness assessment model for SMEs: identification of barriers using AHP method. International Journal of Information Science and Management 2: 61-75.

Jiang, H. and F. Wang. 2009. Analysis of influencing factors on performance evaluation of agricultural products network marketing based on AHM. Conference on Power Electronics and Intelligent Transportation System. Available at: http://tinyurl.com/loyynbe.

Jiong, M., L. Xu, Q. Huang and C. Li. 2013. Research on the e-commerce of agricultural products in Sichuan Province. Journal of Digital Information Management 11: 97-101.

Just, D.R. and R.E. Just. 2006. Information exchange, distributional implications of price discrimination with Internet marketing in agriculture. American Journal of Agricultural Economics 88: 882-899.

Kalaitzandonakes, N., J. Kaufman and X. Wang. 2003. Firm entry through e-commerce in the U.S. agricultural input distribution industry. Journal on Chain and Network Science 3: 123-133.

Kang, J., L. Cai and H. Li. 2009. The development model electronic commerce of regional agriculture. IFIP Advances in Information and Communication Technology. Available at: http://tinyurl.com/lgt757q.

Key, N. and D. Runsten. 1999. Contract Farming, smallholders, and rural development in Latin America: the organization of agroprocessing firms and the scale of outgrower production. World Development 27: 381-401.

Kinsey, J. 2000. A fast, leaner, supply chain: new uses of information technology. American Journal of Agricultural Economics 82: 1123-1129.

Leroux, N., S. Wortman and D. Mathias. 2001. Dominant factors impacting the development of businessto-business (B2B) e-commerce in agriculture. International Food and Agribusiness Management Review 4(2): 205-218.

Li, Z. and Y. Gao. 2011. Information sharing pattern of agricultural products supply chain based on e-commerce.

International Conference on E-Business and E-Government. Available at: http://tinyurl.com/lbeycem.

Liu, W. and H. Li. 2011. Research on the questions of countryside electronic-commerce and the pattern of innovation. International Conference on Management and Service Science. Available at: http:// tinyurl.com/nyetlvo.

Liu, F., W. Tang, Y. Zhang and H. Makoto. 2011. Construction of the agricultural products e-commerce mode linked by rural economic cooperation organization: through two Japanese cases study. International Conference on Business Management and Electronic Information. Available at: http://tinyurl.com/ mqhg4jn.

International Food and Agribusiness Management Review

457

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0156 - Tuesday, August 22, 2017 12:29:21 PM - IP Address:24.21.169.207

Zeng et al.

Volume 20, Issue 4, 2017

Lu, G., P. Lu and C. Liu. 2011. Studying on construction programs of the platform of primary products marketing. International Conference on Advances in Information and Communication Technology.

Available at: http://tinyurl.com/lndf78o.

Machfud, A.K. and M. Kartiwi. 2013. E-commerce adoption by Indonesian small agribusiness: reconsidering the innovation-decision process model. International Conference on Information and Communication Technology for the Muslim World. Available at: http://tinyurl.com/kbv9pt2.

Markelova, H., R. Meinzen-Dick, J. Hellin and S. Dohrn. 2009. Collective action for smallholder market access. Food Policy 34: 1-7.

Mo, W. 2015. The ‘internet plus’ of a naturally-poor county for poverty alleviation and development.

Publishing House of Electronic Industry, Beijing, CN, China. (in Chinese) Molla, A., K. Peszynki and S. Pittayachawan. 2010. The use of e-business in agribusiness: investigating the influence of e-readiness and OTE factors. Journal of Global Information Technology Management 13: 56-78.

Montealegre, F., S. Thompson and J.S. Eales. 2007. An empirical analysis of the determinants of success of food and agribusiness e-commerce firms. International Food and Agribusiness Management Review 10(1): 61-81.

Motiwalla, L., M.R. Khan and S. Xu. 2005. An intra- and inter-industry analysis of e-business effectiveness.

Information and Management 42: 651-667.

Mueller, R.A.E. 2001. E-commerce and entrepreneurship in agricultural markets. American Journal of Agricultural Economics 83: 1243-1249.

Murphy, A.J. 2003. (Re)solving space, time: fulfilment issues in online grocery retailing. Environment and Planning 35: 1173-1200.

Ng, E. 2005. An empirical framework developed for selecting B2B e-business models: the case of Australian agribusiness firms. Journal of Business, Industrial Marketing 20: 218-225.

O’Keeffe, M. 2001. Myths, realities of e-commerce in the perishable foods industries: unleashing the power of reputation, relationship Assets. Supply Chain Management 6: 12-15.

Picot-Coupey, K., E. Huré, G. Cliquet and C. Petr. 2009. Grocery shopping and the internet: exploring French consumers’ perceptions of the ‘hypermarket’ and ‘cybermarket’ formats. The International Review of Retail, Distribution and Consumer Research 19: 437-455.

Poulton, C., A. Dorward and J. Kydd. 2010. The future of small farms: new directions for services, institutions, and intermediation. World Development 38: 1413-1428.

Schluter, G. and C. Lee. 2005. The ‘New Economy’ and efficiency in food market system: a complement or a battleground between economic classes? Technology in Society 27: 217-228.

Seuring, F. and M. Müller. 2008. From a literature review to a conceptual framework for sustainable supply chain management. Journal of Cleaner Production 16: 1699-1710.

Sinkovics, R.R., Y. Yamin and M. Hossinger. 2007. Cultural adaption in cross-border e-commerce: a study of German companies. Journal of Electronic Commerce Research 8: 221-235.

Stritto, G.D. and M.M. Schiraldi. 2013. A strategy oriented framework for food and beverage e-supply chain management. International Journal of Engineering Business Management 5: 1-12.

Stylianou, P. 2008. Online dispute resolution: the case for a treaty between the United States and the European Union in resolving cross-border e-commerce disputes. Syracuse Journal of International Law and Commerce 36: 117-143.

Tranfield, D., D. Denyer and P. Smart. 2003. Towards a methodology for developing evidence-informed management knowledge by means of systematic review. British Journal of Management 14: 207-222.

Tsekouropoulos, G., Z. Andreopoulou, A. Seretakis, T. Koutroumanidis and B. Manos. 2012. Optimising e-marketing criteria for customer communication in food, drink sector in Greece. International Journal of Business Information Systems 9: 1-25.

Van der Vorst, G.A.J., S. Van Dongen, S. Nouguier and R. Hilhorst. 2002. E-business initiatives in food supply chains; definition and typology of electronic business models. International Journal of Logistics: Research and Applications 5: 119-138.

Volpentesta, A.P. and S. Ammirato. 2007. Evaluating web interfaces of B2C e-commerce systems for typical agrifood products. International Journal of Entrepreneurship and Innovation Management 7: 74-91.

International Food and Agribusiness Management Review

458

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0156 - Tuesday, August 22, 2017 12:29:21 PM - IP Address:24.21.169.207

Zeng et al.

Volume 20, Issue 4, 2017

Wang, W. 2010. Agricultural e-commerce application mode in China. International Conference on Information Management and Engineering. Available at: http://tinyurl.com/n4a5yvl.

Wang, Z., D. Yao and X. Yue. 2015. E-business system investment for fresh agricultural food Industry in China. Annals of Operations Research: 1-16.

Wei, Y. 2014. Surpassing the e-commerce economy from county-level e-commerce. New Agriculture 41: 25-27 (in Chinese).

Wiggins, S., J. Kirsten and L. Llambí. 2010. The future of small farms. World Development 38: 1341-1348.

Wu, Y., S. Huang, Q. Chen and Y. Peng. 2010. Opinions about the development of agricultural e-commerce in the New Socialism Countryside Construction. International Conference on E-Business and E-Government. Available at: http://tinyurl.com/l4lthmd.

Xu, L. and L. Wu. 2009. The application of e-commerce in food traceability system: based on the analysis of consumers’ behavior of online searching for traceability information. International Conference on Management and Service Science. Available at: http://tinyurl.com/ma77ujx.

Yang, Q., B. Feng and P. Song. 2008. Research on agricultural e-commerce public trade platform system.

Conference on International Federation for Information Processing. Available at: http://tinyurl.com/ knwhchm.

Yang, J. and Z. Wang. 2015. Research on the application of e-commercial modes for agro-food in mainland China: O2O and B2C. International Conference on Logistics, Informatics and Service Sciences.

Available at: http://tinyurl.com/kv2r9nf.

Yi, F. and X. Rong. 2010. E-commerce applications of agricultural enterprises in Guangdong Province: an empirical analysis of the websites. International Conference on E-Business and Information System Security. Available at: http://tinyurl.com/k77tnpg.

Yu, P. and D. Zhao. 2014. Effect of website quality factors on the success of agricultural products B2C e-commerce. International Conference on Advances in Information and Communication Technology.

Available at: http://tinyurl.com/n56phe5.

Zapata, S.D., C.E. Carpio, O. Isengildina-Massa and R.D. Lamie. 2013. The economic impact of services provided by an electronic trade platform: the case of MarketMaker. Journal of Agricultural and Resource Economics 38: 359-378.

Zeng, Y. and H. Guo. 2016. The formation mechanism of Agro-Taobao village: a multiple-case study. Issues in Agricultural Economy 35: 90-97 (in Chinese).

Zeng, Y., D. Qiu and H. Guo. 2016. Report on the development of China’s agro-Taobao village. In: The modes and cases of China’s internet plus Sannong, edited by H. Guo. Zhejiang University Press, Hangzhou, China (in Chinese).

Zeng, Y., D. Qiu, Y. Shen and H. Guo. 2015. Study on the formation of Taobao village: taking Dongfeng village and Junpu village as examples. Economic geography 35: 90-97 (in Chinese).

Zhang, R. and Z. Jiang. 2015. Report on e-commerce and poverty alleviation (2015). Social Sciences Academic Press, Beijing, CN, China (in Chinese).

Zhang, G. and T. Ma. 2015. Mobile E-commerce. Posts and Telecom Press, Beijing, CN, China (in Chinese).

Zhang, C. and W. Zhang. 2013. Study on quick response distribution task management in agricultural products logistics based on e-commerce. International Asia Conference on Industrial Engineering and Management Innovation: Core Areas of Industrial Engineering. Available at: http://tinyurl.com/ktjpjfx.

Zhang, Y. and L. Huang. 2015. China’s e-commerce development path and mode innovation of agricultural product based on business model canvas method. Wuhan International Conference on E-Business.

Available at: http://tinyurl.com/m3sub5c.

Zhao, Z. and Y. Tian. 2014. Discussion about agricultural e-commerce situation and optimization. International Conference on Management Science and Engineering. Available at: http://tinyurl.com/mm68vlh.

Zwass, V. 1998. Structure and macro-level impacts of electronic commerce: from technological infrastructure to electronic marketplaces. In: Emerging information technologies, edited by K.E. Kendall. Sage Publications, Thousand Oaks, CA, USA.

International Food and Agribusiness Management Review

459

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0156 - Tuesday, August 22, 2017 12:29:21 PM - IP Address:24.21.169.207

OPEN ACCESS International Food and Agribusiness Management Review Volume 20 Issue 4, 2017; DOI: 10.22434/IFAMR2015.0061

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0061 - Tuesday, August 22, 2017 12:31:15 PM - IP Address:24.21.169.207

Received: 6 May 2015 / Accepted: 21 March 2017

A template for sustainable food value chains RESEARCH ARTICLE Elena Monastyrnaya a, Gwenola Yannou Le Brisb, Bernard Yannouc, and Gaëlle Petitd aPhD

student, Laboratoire Genie Industriel, CentraleSupélec, Université Paris-Saclay, 3 rue JoliotCurie, 91192 Gif-sur-Yvette, France; Department of Environmental Systems Science, Swiss Federal Institute of Technology, ETH Zurich, 8092 Zurich, Switzerland; Institute of Humanities, Social Sciences and Technologies, Tomsk Polytechnic University, 30 Avenue Lenin, 634050 Tomsk, Russia bAssociate

Professor, Laboratoire Genie Industriel, CentraleSupélec, Université Paris-Saclay, 3 rue Joliot-Curie, 91192 Gif-sur-Yvette, France; UMR GENIAL, AgroParisTech, INRA, Université Paris-Saclay, 91300 Massy, France cProfessor,

Laboratoire Genie Industriel, CentraleSupélec, Université ParisSaclay, 3 rue Joliot-Curie, 91192 Gif-sur-Yvette, France

dPhD

student, UMR GENIAL, AgroParisTech, INRA, Université Paris-Saclay, 91300 Massy, France

Abstract This paper proposes a template to assist food value chain actors in their collaborative efforts to develop common sustainable strategies and business models. Inspired by the simplicity of the Business Model Canvas, the template reintroduces sustainable practices as a support for management solutions for sustainable food value chains. The template requires cooperation between actors and stakeholders and comprises three steps: (1) identification of needs for sustainability; (2) development of value chain practices aimed to deliver sustainable value, and assignment of responsibilities to actors for these practices; and (3) formulation of a sustainable value proposition. The template also allows a simple graphical representation of sustainability in value chains, which helps improve communication between actors, and allows stakeholders to be kept informed. The template is applied to a sustainable pork value chain to illustrate how it captures various aspects of sustainability in the pork industry.

Keywords: sustainability, food value chains, collaboration, business strategy, business model JEL code: L21 Corresponding author: elena.monastyrnaya@usys.ethz.ch

© 2017 Monastyrnaya et al.

461

Monastyrnaya et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0061 - Tuesday, August 22, 2017 12:31:15 PM - IP Address:24.21.169.207

1. Introduction It is estimated that to meet the nutritional needs of the world’s population and ensure food security in, 2050, food production will have to be increased by 70% (Bruinsma, 2011). Meeting this target raises concerns about the ecological impact of food systems (Ericksen, 2008), food production significantly contributing to the human-induced environmental footprint (FAO, 2014a; Garnett, 2013). For example, in the European Union more than 20% of various environmental impacts are attributed to food (European Commission, 2006). The social implications of enhanced production also have to be considered, food systems providing income and social well-being for over one billion people worldwide (FAO, 2012). Intensified ecological and social stress underscores the necessity for a balanced, sustainable development of food systems based on the conservation, protection, and enhancement of natural ecosystems, and on the protection and improvement of livelihoods and social well-being of people engaged in food production (FAO, 2014a). Ultimately, sustainability can be seen as a prerequisite to ensuring food security in the long term (Berry et al., 2015).

Although the development of sustainable policies is often considered to be a national-level agenda (FAO, 2014a), it is not only scientists, experts, or global agencies who recognize the need for a global shift towards sustainable development: food businesses and value chains are already experiencing increasing pressure from stakeholders and government regulations, urging them to seek ways to be more sustainable and responsible regarding their activities (Bloemhof et al., 2015; Soosay et al., 2012; Wognum et al., 2011). Many authors argue that the advancement of a whole value chain towards sustainability is often more successful than unconnected actions by its individual actors (Lassale-de Salins et al., 2014; Porter and Kramer, 2010). Such a value chain-based vision calls for new business models that allow alignment of stakeholder demands with activities of value chain actors in order to formulate a sustainably sound value proposition.

In the food industry, existing tools and approaches can assist specific aspects of modelling sustainability, such as identification of sustainable value, mostly with the aim of measuring it (FAO, 2013; Maloni and Brown, 2006), development of a conceptual view (Lassale-de Salins et al., 2014), or can provide extensive guidelines that require specific expertise to facilitate sustainable implementation (FAO, 2014a,b; M4P, 2008).

Applying tools, each designed for other purposes, to create a business model for sustainable food value chains would lead to a complex, time-consuming, and possibly inconsistent process, whereas the need for aligned performance requires mechanisms to facilitate collaboration among value chain actors.

In this light, we address the following research issue: how can we help actors in their collaborative efforts to create a vision and develop a model for sustainable food value chains? We pursued three objectives: (1) identify a vision and means for sustainability in the food industry, and in food value chains in particular; (2) develop a tool that assists the collaboration-based business modelling process, and definition of strategies for sustainable food value chains; and (3) illustrate the application of the proposed tool to an existing value chain.

This paper is in four parts. After an introduction, Part 2 discusses the evolution of sustainable thinking in business modelling, together with tools and approaches to assist sustainable modelling in food value chains.

It concludes with a proposition for a template that synthesizes a vision of sustainability in food value chains to help actors in a food value chain rethink their strategies and business models in terms of sustainability. In Part 3, the application of this model to an existing pork value chain is demonstrated. Finally, Part 4 presents a discussion and conclusion on the advantages and limitations of the proposed template and case study.

International Food and Agribusiness Management Review

462

Monastyrnaya et al.

Volume 20, Issue 4, 2017

2. Literature review 2.1 Sustainable thinking in business modelling

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0061 - Tuesday, August 22, 2017 12:31:15 PM - IP Address:24.21.169.207

■■ Conventional business model Teece (2010) defines a business model as aiming to identify a market segment, formulate a relevant value proposition, and define how a business creates and delivers value to its customers while generating a profit.

A business model serves to represent business strategy and ultimately assist in developing tactics (CasadesusMasanell and Ricart, 2010). The Business Model Canvas (BMC) developed by Osterwalder and Pigneur (2010) is a concise graphical representation of generic business models. BMC is composed of nine building blocks that encompass the above aspects of business models: (1) business infrastructure (key partners, key activities, key resources, and cost structure); (2) market segments and customers (customer relationships, channels, customer segments, and revenue streams); and (3) value proposition.

BMC is a popular tool among managers because it allows both a description of existing businesses and the development of new ones. In addition, owing to its simplicity in visual representation, BMC is commonly used for communicating business strategy (Frick and Ali, 2013). However, with the rise of sustainable thinking in management science, the conventional understanding of business models such as BMC has appeared ill-suited to sustainable business modelling, because it tends to focus primarily on customer value (Bocken et al., 2013), whereas sustainable thinking requires consideration of a wider range of stakeholders.

Accordingly, a new vision has emerged, incorporating sustainability into the conventional understanding of business, and leading on to a conceptual transformation of business models.

■■ Stakeholder perspective The broad understanding of business sustainability is underpinned by the concept of the Triple Bottom Line (Elkington, 1999), which implies that sustainability requires the value proposition to be extended beyond a demand from potential customers. According to Triple Bottom Line, a sustainable value proposition incorporates interests of stakeholders representing three pillars of sustainability: business, society, and nature (Hart and Milstein, 2003). In food systems, an extended overview on the three pillars of TLB is represented by ‘sustainable dimensions’, along with more detailed indicators. For example, Maloni and Brown (2006) propose eight dimensions for food sustainability: animal welfare, biotechnology, health and safety, labor and human rights, procurement, fair trade, community, and environment. The United Nations Food and Agriculture Organization (FAO, 2013) proposes 118 indicators clustered in four main dimensions: environmental integrity, economic resilience, social well-being, and good governance. The sustainable dimensions typically represent an overview on relevant aspects of sustainability in the food sector, while indicators aim to assess sustainability both quantitatively and qualitatively.

The large number of indicators is explained by the diversity of sustainability issues, which depend on different contexts of specific value chains (animal or plant production, geographical contexts, institutional set-ups, etc.). Although many of the dimensions and indicators overlap, they differ overall, being based on the broad practical experience of the experts who propose them. Hence it is practically impossible to propose a definitive set of indicators that will fit all food value chains (FAO, 2013). These disparities create difficulties for actors operating within a specific food system context to implement sustainability. Hence the identification of the ‘hotspots’ for sustainability and adjustment of indicators and dimensions relies on the analysis of the specific food value chain, which can be based on literature reviews or direct communication with stakeholders.

International Food and Agribusiness Management Review

463

Monastyrnaya et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0061 - Tuesday, August 22, 2017 12:31:15 PM - IP Address:24.21.169.207

■■ Value chain perspective Embedding sustainable thinking into conventional business arrangements requires more than just considering the interests of society and nature; it calls for the transformation of a firm’s business models (Stubbs and Cocklin, 2008). Many authors argue that isolated efforts of individual firms is not sufficient for such transformation, and that sustainability is instead achieved through aligned performance of value chain members (Boons, 2012; Lassale-de Salins et al., 2014; Soosay and Hyland, 2015). Some inconsistency is observed in discussions on sustainability: certain authors use the term ‘supply chain’, whereas others prefer ‘value chain’. The meaning of ‘supply chain’ and ‘value chain’ become intertwined in such a way that it is often practically impossible to separate them. In this paper, the authors have opted to use the term ‘value chain’ to refer to groups of organizations that exchange materials, financing, and information, as well as collaborate in the medium and long terms. However, in the references to works of other scholars, the terms initially used are cited: the matters referred to can to a large extent be assumed to be applicable to both supply and value chains.

Humphrey and Memedovic (2006) state that the food business is increasingly dominated by value chain relationships, in which retailers and branded marketers exercise vertical coordination. Such collaboration is typically associated with power imbalances (Hartmann, 2011). The power imbalances can negatively influence collaboration and trust between actors (Kähkönen, 2014). Meanwhile, collaboration is seen as a key principle to improve sustainable performance in the food value chains (Touboulic and Walker, 2015; Varsei et al., 2014). Although bigger players have more sources and power to drive sustainability in the value chains (Hartmann, 2011), a lack of collaboration caused by power imbalances may have negative effects on overall sustainable performance in the value chain (Touboulic and Walker, 2015). Thus sustainability thinking implies reconsidering chains in the sense of building a new form of communication and coordination between their actors. It can be assumed that a sustainable value results from a synergistic effect of the contributions of value chain actors who share a willingness and vision for sustainability together with a common sustainable strategy. The necessity for redefining productivity in value chains is clearly expressed by Porter and Kramer (2010) through the concept of Shared Value. Hence the practices applied in the chains are seen as a source for the creation of sustainable value. Beske et al. (2014) summarize practices commonly used in sustainable supply chain management and discussed in the scientific literature, placing them in five categories. These are ‘strategic orientation’, which indicates a company’s determination to address sustainable values, ‘continuity’, representing the structural aspect of permanent relationships between actors, ‘collaboration’, which describes the technical and logistical alignment of activities and information flows, ‘risk management’, concerning the mitigation of possible risks, including those related to external pressure from stakeholders, and ‘proactivity’, indicating collaboration with stakeholders and openness to changes (Table 1).

These practices are aimed to enhance productivity of supply chains directly or indirectly in order to achieve greater sustainable effects and keep a company profitable (Beske et al., 2014). This comprehensive overview on practices to achieve sustainability opens up powerful perspectives for the development of new business models, strategies and tactics for sustainable value chains especially if gathered into a simple tool. We go on to discuss the existing tools for managers that assist in building sustainable food value chains.

■■ Tools to support development of business models for sustainable food value chains The FAO (2014b) have proposed a guiding approach to analyzing sustainability in food value chains and developing strategies and plans to improve sustainability across value chain activities. The approach falls in line with the principles of the Shared Value concept, combining an analysis of the stakeholders’ needs and expectations with practices to achieve greater sustainable results. However, the approach calls for positive impacts at larger scales, covering entire product sub-sectors, and so targets the behaviors of large firms, institutional changes and policies, in what is a largely top-down view (FAO, 2014b: vii, 52). At the same time, the scientific literature highlights the importance of sustainable actions at micro-levels (Hart and Milstein, 2003; Porter and Kramer, 2010; Varsei et al., 2014), which requires awareness and pro-activity from individual International Food and Agribusiness Management Review

464

Monastyrnaya et al.

Volume 20, Issue 4, 2017

Table 1. Sustainable supply chain practices (Beske et al., 2014).

Strategic orientation Continuity

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0061 - Tuesday, August 22, 2017 12:31:15 PM - IP Address:24.21.169.207

Collaboration

Risk management

Pro-activity

• Supply chain management • Triple bottom line • Long-term relationship • Partner development • Partner selection • Joint development • Technical integration • Logistic integration • Enhanced communication • Individual monitoring • Pressure group management • Standards and certification • Learning • Stakeholder management • Innovation • Life cycle assessment

value chains and their actors. In recent years, increasing attention has thus been paid to tools able to assist managers in defining strategy and sustainable business modelling at organization and chain level (Bocken et al., 2013; Stubbs and Cocklin, 2008; Varsei et al., 2014). However, these approaches have been developed primarily for the general context of sustainability, and not for the specific context of the food sector.

In the field of food sustainability research, a number of findings can be applied for this purpose. In general, approaches tend to focus on one particular pillar of sustainability – environmental (Pelletier, 2015; Soosay et al., 2012), or social (M4P, 2008), or on particular practices that can enhance sustainability, e.g. collaboration (Wognum et al., 2011). An integrated yet theoretical approach is proposed by Lassale-de Salins et al. (2014) to implement all three aspects of sustainability, referred to as ‘extended demand’ in supply chain management.

The sustainable dimensions and indicators discussed earlier (FAO, 2013; Maloni and Brown, 2006) provide general directions for sustainable performance in the food industry, and so are suitable for operational purposes.

To support such operationalization in food chains, Heikkurinen et al. (2012) propose a visual framework to place sustainable indicators in the supply chain context. However, since indicators define what sustainably is, rather than how to create it, they are best suited to measuring sustainability and/or to specifying value propositions, and are therefore insufficient to cover all aspect of business modelling, which ideally should include the identification of a market segment, and formulation of a value proposition, as well as defining how a business creates and delivers value to its customers while generating a profit.

Each of the above tools can be used to support particular aspects of business modelling. However, their disconnected use makes managers’ tasks more complicated, and is liable to hamper information flow, thereby jeopardizing communication between actors, especially important in the context of the food value chains.

To date, to our best knowledge, there is no tool for assisting the design of business modelling and defining business strategies in a sustainable food value chain that is comparable to the BMC for simplicity. We shall now introduce a template aimed to assist value chain actors in defining a sustainable strategy and business model.

2.2 A template for sustainable food value chains We set out to develop a template to assist practitioners in developing business models for sustainable food value chains. Based on our literature review, we argue that the following considerations should underpin the template: International Food and Agribusiness Management Review

465

Monastyrnaya et al.

■■ ■■ ■■

■■ ■■ ■■

The template should embrace the aspects of a conventional business model: i.e. customer demand, architecture of a value chain, and a value proposition.

In addition to the customer demand, it should include the interests of stakeholders representing all three pillars of sustainability.

The template should allow the demands of stakeholders to be aligned to the architecture of a value chain so as to formulate a valid sustainable value proposition.

The structure of the template should allow the roles of different value chain actors along the process of creation of shared sustainable value to be demonstrated, in line with the concept of Shared Value (Porter and Kramer, 2010).

Since stakeholders, sustainable practices, or indicators can vary depending on the context, the structure of the template should allow for flexibility, providing clear yet easily adjustable directives.

The template should provide simple and easily understandable directions that allow improved communication among value chain actors.

The tool should have a concise graphical representation, similar to that proposed by BMC, to favor efficient communication of sustainability performance to other stakeholders.

Figure 1 presents a tool based on all the above considerations – a template for sustainable food value chains.

The template graphically represents the process of alignment of value chain practices to a sustainable value proposition, and ultimately aims to help actors identify new areas and solutions for the sustainable development of their value chains. Following Figure 1, detailed guidelines for the use of the template are provided.

We emphasize that the aim of this template is not to define what sustainability is or to provide set-in-stone directions on how it should be built, but to provide general yet comprehensive guidelines for the creation of a business model for a sustainable food value chain. Above all, the template calls for a three-step collaborative action plan that includes communication with stakeholders and cooperation between value chain actors.

It is through the commitment of stakeholders and value chain actors that the tool is turned from a general conceptual vision into a practical context-specific tool.

Identification of solutions

VC actors Customers Environment Workshops and discussion between value chain actors

Figure 1. Template for sustainable food value chains. VC = value chain.

International Food and Agribusiness Management Review

466

Retail

Processing

Value chain activities

Society

Literature analysis, interview with stakeholders

Proposition of value Step 3

Production

Pro-activity

Risk management

Collaboration

Sustainable practices Continuity

Stakeholders

Step 2 Value destroyed or missed

Step 1

Assignment of responsibilities

Formulation of sustainable value

Identification of stakeholder demands

Data Design process sources

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0061 - Tuesday, August 22, 2017 12:31:15 PM - IP Address:24.21.169.207

■■

Volume 20, Issue 4, 2017

Monastyrnaya et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0061 - Tuesday, August 22, 2017 12:31:15 PM - IP Address:24.21.169.207

The Template for Sustainable Food Value Chains (Figure 1) includes a structured three-step procedure for developing a sustainability-oriented strategy for food value chains.

In the first step, the template is used to lay a foundation for developing a problem-oriented sustainable strategy and business model. This step involves the collection of information about stakeholders’ needs in order to define the main sustainability-related problems that require solving by the value chain actors. It is recommended that the information is collected by at least two project managers – representatives of different activities of the value chain – to mitigate possible biases. The project managers should carefully identify stakeholders of the value chain representing different dimensions of sustainability. Stakeholders can be identified through either literature review or interviews with experts. The list of stakeholders is further used by the project managers to analyze stakeholder needs or, in other terms, as suggested by Bocken et al. (2013), what is destroyed or missing value in the respective food industry. Identification of value issues requires closer interaction with stakeholders, i.e. it should occur through communication, surveys or interviews with them. The dimensions and indicators for sustainability in the food industry discussed in 2.1 (FAO, 2013; Maloni and Brown, 2006) can be used to make more detailed tentative check lists so as to identify gaps and opportunities for sustainable value, with the added consideration that stakeholders may raise new issues, rarely mentioned by scholars. The issues detected are then placed in Step 1 of the template.

The second step assumes that sustainability should not be based on a single decision-maker, but requires the collaborative effort and consent of the value chain actors (Lassale-de Salins et al., 2014). This second step requires representatives from each value chain activity to gather for a meeting in the form of an open discussion or workshop. During the meeting, destroyed and missed values identified in Step 1 are first introduced by the project managers to the value chain representatives. Following this introduction, value chain representatives and project managers propose potential solutions to compensate for the missed or destroyed values. The template allows the positioning of solutions to existing problems in a matrix-like pattern. In the default version of the template (Figure 1), the authors propose practices summarized by Beske et al.

(2014) (Table 1) (except for ‘strategic orientation’, which characterizes the whole process represented by the template), with the added consideration that more specific actions are likely to be developed by value chain actors during the discussions and workshops. Responsibilities for identified practices are then assigned in such a way that the actors have a clear understanding of what their role is in the process of sustainable value creation. The idea of this representation belongs to Heikkurinen et al. (2012), who initially used it for measuring the sustainable performance of a value chain.

In the third step, once the agreement is reached (solutions are designed and responsibilities are assigned), the proposition of sustainable value for the food value chain should be formulated. It is not necessary to separately identify different types of value for the final value proposition – it is often not possible to distinguish between them (e.g. an environmental value can be at the same time a social value, and this is indeed most often the case). The third step allows a reasonable value proposition to be constructed that corresponds to both the real needs of stakeholders and the architecture of the value chain.

Once all three steps are completed, the template provides a firm grounding for mission, strategy and tactics throughout the value chain. It can be used to identify and specify changes necessary for improved sustainable performance. The final version of the template (including all three steps) can additionally be used for general marketing purposes, and for communication with consumers and other stakeholders. Our next part illustrates the application of the template to the case of a pork value chain in France.

International Food and Agribusiness Management Review

467

Monastyrnaya et al.

Volume 20, Issue 4, 2017

3. Case study: a pork value chain in France

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0061 - Tuesday, August 22, 2017 12:31:15 PM - IP Address:24.21.169.207

3.1 Introduction to a case study For reasons of business confidentiality, we could not conduct a direct application of the template to a food value chain. This case study illustrates the ability of the template to summarize the sustainability-relevant problems of a mainstream French pork value chain. In addition, the actions conducted by an existing value chain already working on the sustainability of its activities are collected and represented in Part 2 of the template to highlight the nature of the solutions mobilized. This process allows the verification of the relevance of sustainable practices proposed by Beske et al. (2014), which are used in the template for the identification of actions aimed to improve sustainability in food value chains. Finally, the template is used to highlight the stakeholder needs not addressed in the case study sustainable value chain. The identification of these gaps could help managers update the sustainable strategy of the value chain if needed. What we sought to define here is the ability of the template to synthesize the sustainability issues of a value chain, to position solutions addressing these challenges, and to highlight contributions to both impacts and improvements of each actor in the value chain. The ability to account for these aspects is central in the Sustainable Food Value Chains.

The following subsections provide a general overview on the French pork sector and on the pork value chain case. The data were obtained through a literature review and through interviews with experts.

The case study follows the design of the template proposed in 2.2 This analysis was carried out by us, acting as neutral experts. In the first step (3.2), a brief overview of the French pork industry is presented in order to identify sustainability-related issues. In 3.3 we describe the sustainable performance of a value chain.

Finally, 3.4 reassembles and recapitulates the information introduced in 3.2 and 3.3 from the perspective of the template for sustainable food value chains (Figure 2), and introduces a detailed overview on the sustainable value created by the pork value chain.

3.2 Overview on the pork sector in France The European pork production sector comprises different organizations. First, breeders can be independent (producing or buying piglets) and work alone. Second, they can be independent but associated in producer groups. This type of association improves their negotiating capacities for feed buying and the sale of pigs.

Third, they can be integrated by an industrial organization that can be a feed producer or a downstream actor in meat processing. When a breeder is integrated he becomes the employee responsible for fattening the animals supplied by the cooperative or industrial organization. Roguet and Rieu (2011) report that pork production differs markedly from one European country to another: in Germany and the Netherlands, the vertical coordination between breeders and their partners is poorly standardized and not well documented; in Denmark, about 90% of the pork production is by breeders who are members of the Danish Crown cooperative, a commercial company; in Spain, between 85 and 90% of the farmers in charge of the pork fattening step are integrated (Daridan and Gil, 2007).

French pork production is represented by two main types of actors: (1) independent actors (who make up the lowest proportion of the whole production sector) and (2) agricultural cooperatives or industrial groups that in general include suppliers of animal feed and other agricultural inputs. In 2009, 93% of pork production was by breeders who were members of the 56 existing producer groups (Roguet and Rieu, 2011). In 2010, only 34% of the pork meat output was produced by integrated breeders (Nicourt and Cabaret, 2014). As of 2015, total French pig stock was 13.3 million; the sector has been facing competition from the Netherlands and Spain (Agreste, 2016; Roussillon and Legendre, 2015). The Netherlands and Spain have lowered operating costs through new automated and more productive installations and lower labor costs. Poor integration of French channels also limits the possibilities of coordination and strategic alignment between actors from different links in the value chain (Roussilon and Legendre, 2015). This weak integration results from a desire for independence in parts of the farming communities, which ultimately contributes to the non-homogeneity International Food and Agribusiness Management Review

468

Monastyrnaya et al.

Volume 20, Issue 4, 2017

Step 1

Step 2

Value chain actors Society Customers Environment Employees

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0061 - Tuesday, August 22, 2017 12:31:15 PM - IP Address:24.21.169.207

2. Inconsistent capacities of actors: in particular capacities of slaughter and needs of distributors are not well aligned.

3. Specialization of producers (animal breeding without cropping, or monoculture farming) increase their dependency on international markets and their exposure to price fluctuations.

4. Breeders are exposed to price volatility on primary product markets and on their products with, in all cases, low margins.

5. Uncertainties related to risks in the food industry hamper investment capacities of actors.

6. Uneven distribution of financial risks across the value chains. Since some actors are more exposed to the risks than others, the capacity of the whole value chain to sustain market price fluctuations is reduced.

7. Lack of trust regarding origin of meat and breeding conditions.

8. Limiting the use of antibiotics.

9. Modern animal feed contributes to a too-high omega-6/omega-3 ratio.

10. National demand is focused on certain pork parts, while other parts are considered of little value.

11. Conditions of breeding can induce a too-high stress for the animal (living conditions too far from the natural mode, lack of space, problems of temperature, etc.).

12. Pollution of water and soil through excessive use of manure and other agro-inputs in the production of feedstuffs for animals.

13. Carbon emissions related to transport including that of foodstuffs for animals.

14. Deforestations related to cultivation of soya in the Americas.

15. Environmental and social impacts of animal production create a negative image of this profession in society. This makes farmers especially vulnerable, both economically and psychologically.

Figure 2. Model of a case of a sustainable pork value chain in France.

International Food and Agribusiness Management Review

469

The cooperative of retailers

Production

Input supply

Pro-activity

Risk management

Value destroyed or missed 1. People living close to livestock farms could be disturbed by odor and noise pollutions.

Collaboration

Continuity

Stakeholders

Solution proposed by the case study value chain

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0061 - Tuesday, August 22, 2017 12:31:15 PM - IP Address:24.21.169.207

Monastyrnaya et al.

Volume 20, Issue 4, 2017

of the production capacities between the different activities of the value chain (i.e. production, processing and retail). Difficulties in coordination and risk-sharing between actors of the pork industry reduce their financial ability to withstand changes in market demands. Furthermore, less time spent preparing meals and changing diets has led to the concentration of domestic demand on certain parts of pig carcasses (approximately 30% of a carcass is used for raw meat; the rest is processed into different meat-based products (e.g. sausages) or used in medicine, or for other products such as glue, energy production based on fat, feed, etc. (European Commission, 2005). This disqualification of some cheap cuts has economic consequences, and entails sales work by slaughterhouses to ensure the economic stability of their activity. On the positive side, pork has a competitive advantage, its sale price being lower than that of beef or lamb. Hence pork could be an economically attractive alternative in the food budgets of many households. Nevertheless, in recent years, the sales of pork meat in France have been decreasing (94 kg carcass equivalent kgce/capita (kilo carcass equivalent/capita) sold in 1988 against 86 kgce/inhab. in 2014) (FranceAgriMer, 2015).

To conclude, farmers are exposed to high economic risks, caught between price volatility for agricultural raw materials, and unstable, low prices of the pork market. In addition, the environmental and social impacts of livestock production foster a bad image of breeders in civil society. This is a factor of psychological distress among part of the profession. Overall, pork production is associated with various environmental and social impacts that can be summarized as follows: ■■ Water and soil pollution (by manure in the case of overuse and by fertilizers and pesticides used in crops).

■■ Carbon emissions associated with animal feed.

■■ Deforestation in the Americas associated with soy crops.

■■ Noise and olfactory pollution close to breeding farms.

3.3 The case of a sustainable pork value chain The present study uses the case of a sustainable pork value chain. The objective of this case study is to clarify how the template for sustainable food value chains can help identify the relevance of the proposed business model to the real needs of stakeholders. Our study aims to identify how the elements communicated on this ‘sustainable’ value chain form part of a strategy that seems logical with regard to the sustainability issues identified for the French pork industry. The case described here is a real French case, but the commercial names of the organizations are not disclosed. For the same reasons, only published data from the literature are used here to describe the case.

The value chain described in this experiment was created in 2014 by a contract for a partnership lasting at least three years between a cooperative of producers (referred to in the text as CPX) and a French cooperative of distribution (referred to in the text as CDY). This contract governs a common commitment to a set of specifications, volumes and prices. CPX was established in 1889 and is present in an area equivalent to a quarter of France. It provides 14,000 jobs, and its activities cover the plant and animal sectors. The conditions for farmers to be members of the cooperative are their geographical presence in the territory of CFX and their adherence to its values. CPX includes approximately 20,000 farmers, and voluntary members. Part are engaged in an intensive ecological farming approach. The sector that we consider in this example includes approximately 180 farmers and committed volunteers in a procedure designed to provide a more sustainable farming and distribution process without being organic. According to the communication of the agricultural cooperative commitment, this process involves various aspects for volunteer breeders. The specification involves GMO-free feeds for the second age feed and no soy for fattening feed. This fattening stage meets the criteria of pork load specifications defined by the French Heart Association Bleu-Blanc-Coeur. This specification was defined to meet four objectives: ■■ Improve the nutritional quality of meat by increasing its omega-3 content (through introducing linseed mix in animal feed).

■■ Reduce emissions into the environment by structuring the diversity of animal feed to allow rotation of crops on soils that limit the needs for nitrogen and phosphorus chemical inputs.

International Food and Agribusiness Management Review

470

Monastyrnaya et al.

■■

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0061 - Tuesday, August 22, 2017 12:31:15 PM - IP Address:24.21.169.207

■■

Volume 20, Issue 4, 2017

Develop the autonomy of farmers by enabling them to reclaim the link between crop land and farms.

Reduce use of antibiotics.

All the constraints that these specifications entail require investment by farmers and hence raise operating costs. Economic studies had already shown that in France, pig feed represents on average 60% of the cost of breeding. After the farming step, pigs are slaughtered and butchered in the slaughterhouse and two butcheries that are subsidiaries of the cooperative. The products from the butcheries are shared among three types of use: ■■ raw meat parts (10 references) marketed in the form of trays ready to go; ■■ sausages (4 references); ■■ parts that do not fit into the above two categories, and are sold without special labelling along with pork meat from conventional farms.

Product reference management is conducted jointly by CPX and CDY, and aims to optimize the economic performance of the value creation chain. The economic support of the upstream phase is ensured by the payment to the farmers of a gain of 8-12 cents per kilogram compared with the price set each week for French pork. The investments made by farmers are facilitated by the cooperative, an assistance made possible by the commitment for supply over 3 years binding both partners.

3.4 Model for the case of a sustainable pork value chain The analysis of the information published on the Internet and press together with discussions conducted with the value chain actors were used to fill out the template (Figure 2) in order to recreate a model for the case of a sustainable pork value chain. The main issues on sustainability (or concerns of different types of stakeholders in the pork industry), revealed in the overview presented in 3.1, are summarized in Step 1 of the template. Step 2 of the template includes the practices used in the pork value chain (3.2), and highlights issues they target in green. Finally, blue in the last three columns indicates who the actors responsible for implementation of these practices are. In the template, the cooperative of retailers is referred as ‘R’, and the cooperative of producers is represented by two activities – input supply ‘IS’ and production ‘P’, to highlight the importance of agricultural inputs to the implementation of sustainability.

In what follows, the actions implemented by the pork value chain in response to the sustainability issues are detailed. This synthesis provides further insights into the relevance of the proposed practices to the real needs of stakeholders.

■■ Issue 1: people living in the proximity of livestock farms could be disturbed by odour and noise pollutions.

There is no public mention of any action conducted to modify this situation directly or indirectly.

However, the partnership secures those breeders most inclined to invest in the renovation of buildings: sound insulation and optimized ventilation to minimize odour.

■■ Issue 2: inconsistent capacities of actors: in particular production capacities of slaughter and needs of distributors are not well aligned Having contracts binding the agricultural cooperative and the distributor over a period of three years, favours investment when it is necessary for breeding. The longer-term relationship can also make it possible to jointly seek how to make more gainful use of the parts of the carcass in least demand.

■■ Issue 3: specialization of producers (animal breeding without cropping, or monoculture farming) increase their dependency on international markets and their exposure to price fluctuations The introduction of locally grown crops helps reduce farmers’ dependence on agricultural commodity prices. This solution is made possible by the acceptance of a purchase price of the carcass above the standard price of the pig.

■■ Issue 4: breeders are exposed to price volatility on primary product markets and on their products, with low margins in all cases

International Food and Agribusiness Management Review

471

Monastyrnaya et al.

■■

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0061 - Tuesday, August 22, 2017 12:31:15 PM - IP Address:24.21.169.207

■■

■■

■■

■■

■■

■■

■■

■■ ■■ ■■

Volume 20, Issue 4, 2017

Contractual agreements offer guarantees in terms of compensation for carcasses, even though these guarantees are limited because the price is indexed on average prices. This is a form of collaboration aimed at increasing the quality of the product.

Issue 5: uncertainties related to risks in the food industry hamper the investment capacities of actors Help is provided by the payment of a surplus per kilo, which influences the decision-making for the investment.

Issue 6: uneven distribution of financial risks across the value chains. Since some actors are more exposed to the risks than others, the capacity of the whole value chain to sustain market price fluctuations is reduced The proposal of a differentiated offer compared to competitors through a commitment to product quality, and the sharing of profits underpinned by financial contracting encourage a reduction of financial risks for all players Issue 7. lack of trust regarding origin of meat and breeding conditions Full traceability of livestock products, a plentiful meat offer on the market, and information on livestock conditions contribute to ensuring the transparency of the information transmitted to consumers.

Issue 8. limiting the use of antibiotics This is a voluntary action by the actors in response to the perception of their use as dangerous by consumers and civil society. This action is also a source of cost reduction.

Issue 9. modern animal feed contributes to a too-high omega-6/omega-3 ratio The change in animal feed makes it possible to modify the omega-3/omega-6 ratios to help redress the imbalance observed in epidemiological surveys.

Issue 10: national demand is focused on certain pork parts, while other parts are considered of little value In the communication developed around the value chain described, there are no solutions to this issue Issue 11: conditions of breeding can induce too-high stress for the animal (living conditions too far from the natural mode, lack of space, problems of temperature, etc.)

In the communication developed around the value chain described, there are no solutions to this issue Issue 12: pollution of water and soil through excessive use of manure and other agro-inputs in the production of feedstuffs for animals In the communication developed around the value chain described, there are no solutions to this issue Issue 13. carbon emissions related to transport including that of foodstuffs for animals In the communication developed around the value chain described, there are no solutions to this issue Issue 14. deforestations related to cultivation of soya The integration of locally grown foods helps reduce this effect.

Issue 15. environmental and social impacts of animal production create a negative image of this profession in society. This makes farmers especially vulnerable, both economically and psychologically.

Communicating on actions for society and consumers helps improve the image of upstream agriculture.

It is an upgrading factor for producers.

Overall, the template for sustainable food value chains demonstrates that the performance aspects of this case of a sustainable pork value chain fall in line with the sustainability issues identified for the French pork industry. This means that the sustainable proposition corresponds to the expectations of main stakeholders in the pork value chain. A second observation is that for issues 10, 11, 12, 13, no solutions were found. This might indicate room for improvement, but the available sources of information may have been insufficient to enable us to identify solutions.

4. Discussion and conclusions The case study described here was intended to verify the capacity of the template to synthesize the sustainability issues of a food value chain. The different hotspots identified in the survey on the sustainability issues of the French pork sector are stated in the first part of the template in relation with the stakeholders concerned.

Facing these issues, the actions conducted by a value chain acting to improve its sustainable performance International Food and Agribusiness Management Review

472

Monastyrnaya et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0061 - Tuesday, August 22, 2017 12:31:15 PM - IP Address:24.21.169.207

are found in the columns created in relation with the proposition for the sustainability of the food value chains. Finally, the matrix-based representation allows easy identification of the issues not dealt with, and the actors contributing to value creation. This then makes it possible for the value chains actors concerned to define a new sustainable strategy to improve the performance of their activities.

Importantly, the template does not make it possible either to rank the issues listed in the first column, or to measure the efficacy of the responses provided by the solutions implemented (whether they solve a small part of the problem, a large part, or all of it). However, the issue of prioritizing could be addressed by a cost-benefit analysis of the proposed solutions. The relevance of the proposed solutions requires tests and measurements for the progressive constitution of a balance score card to monitor the levels of performance achieved.

To summarize, the tool is not intended to provide means to directly measure the sustainable performance of food value chains. The added value of its use lies in the template, which (1) enables stakeholder driven process of defining sustainable strategy and business model; (2) encourages collaboration between value chain actors; and (3) offers a synthesis, i.e. a single-picture representation, of the major issues in a particular value chain, and the potential and existing forms of action that can be mobilized (as defined in column headings of the template: continuity of relationships, contractualization, etc.). The impact of the right-hand side of the template, which indicates ‘who’ in the value chain supports the efforts to improve the sustainability of the value chain, could not be tested in our case. The expected effect of this information is to facilitate the sharing of the economic value between the actors according to their contribution to the final value of the product. Hence it is important to find out whether this representation influences the economic negotiations of the actors.

The next step will be to test the template in a real situation with different actors of a food value chain. This step will enable us to determine whether this common representation offered by the template promotes collaboration between the actors. To test this aspect, the ideal situation would be to use it in a case where the value chain actors are about to start cooperating for sustainability.

References Agreste. 2016. Agreste conjoncture – Porcins, enquête cheftel 2015, Résultats Français et Européens.

Agreste, Montreuil-sous-Bois, France.

Berry, E.M., S. Dernini, B. Burlingame, A. Meybeck and P. Conforti. 2015. Food security and sustainability: can one exist without the other? Public health nutrition 18: 1-10.

Beske, P., A. Land and S. Seuring. 2014. Sustainable supply chain management practices and dynamic capabilities in the food industry: a critical analysis of the literature. International Journal of Production Economics: 131-143.

Bloemhof, J.M., J.G.A.J. Van der Vorst, M. Bastl and H. Allaoui. 2015. Sustainability assessment of food chain logistics. International Journal of Logistics Research and Applications 18: 101-117.

Bocken, N.M.P., S.W. Short, P. Rana and S. Evans. 2013. A value mapping tool for sustainable business modelling. Corporate Governance 13: 482-497.

Boons, F. 2012. Sustainable innovation, business models and economic performance: an overview. Journal of Cleaner Production 45: 1-8.

Bruinsma, J. 2011. Looking ahead in world food and agriculture: perspectives to 2050. In: Looking ahead in world food and agriculture: perspectives to 2050, edited by P. Conforti. Food and Agriculture Organization of the United Nations, Rome, Italy.

Casadesus-Masanell, R. and J.E. Ricart. 2010. From strategy to business models and onto tactics. Long Range Planning 43: 195-215.

Daridan, D. and J.M. Gil. 2007. La production porcine Espagnole, entre croissance et consolidation. Journées Recherche Porcine 39: 301-310.

International Food and Agribusiness Management Review

473

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0061 - Tuesday, August 22, 2017 12:31:15 PM - IP Address:24.21.169.207

Monastyrnaya et al.

Volume 20, Issue 4, 2017

Elkington, J. 1999. Cannibals with forks: triple bottom line of 21st century business. Capstone Publishing Ltd, North Mankato, MN, USA. Ericksen, P.J. 2008. Conceptualizing food systems for global environmental change research. Global Environmental Change 18: 234-245.

European Commission. 2005. Reference Document on Best Available Techniques in the Slaughterhouses and Animal By-Products Industries. Available at: http://tinyurl.com/mup2ezx.

European Commission. 2006. EUR 22284 Environmental Impact of Products (EIPRO) – Analysis of the life cycle environmental impacts related to the final consumption of the EU-25. Available at: http:// tinyurl.com/dxwqd58.

Food and Agriculture Organization of the United Nations (FAO). 2012. FAO Statistical Yearbook. World Food and Agriculture, Rome, Italy.

Food and Agriculture Organization of the United Nations (FAO). 2013. Sustainability assessment of food and agricultural system: indicators. Rome, Italy. Available at: http://tinyurl.com/k77fmhu.

Food and Agriculture Organization of the United Nations (FAO). 2014a. Building a common vision for sustainable food and agriculture. Principles and approaches. Available at: http://tinyurl.com/kqq7uu4.

Food and Agriculture Organization of the United Nations (FAO). 2014b. Developing sustainable food value chains – guiding principles. Available at: http://www.fao.org/3/a-i3953e.pdf.

FranceAgriMer. 2015. Consommation des produits carnés en 2014. Available at: http://tinyurl.com/n2eu8s6.

Frick, J. and M.M. Ali. 2013. Business model canvas as tool for SME. IFIP Advances in Information and Communication Technology 415: 142-149.

Garnett, T. 2013. Food sustainability: problems, perspectives and solutions. Proceedings of the Nutrition Society 72: 29-39.

Hart, S.L. and M.B. Milstein. 2003. Creating sustainable value. Academy of Management Executive 17: 56-67.

Hartmann, M. 2011. Corporate social responsibility in the food sector. European Review of Agricultural Economics 38: 297-324.

Heikkurinen, P., L. Jalkanen, K. Järvelä and M. Järvinen. 2012. Corporate responsibility in the food chain: the criteria and indicators. In: Proceedings of the 6th International European Forum on System Dynamics and Innovation in Food Networks, edited by. U. Rickert and G. Schiefer. Universität Bonn-ILB Press, Bonn, Innsbruck-Igls, Austria, pp. 653-666.

Humphrey, J. and O. Memedovic. 2006. Global value chains in the agrifood sector. Available at: http:// tinyurl.com/m4xjbcb.

Kähkönen, A.-K. 2014. The influence of power position on the depth of collaboration. Supply Chain Management 19: 17-30.

Lassale-de Salins, M., G. Bertoluci and A. Chapdaniel. 2014. Managing sustainability in supply chains: the sustainable demand-supply chain approach, a proposal for a pragmatic approach in the food sector.

Available at: http://tinyurl.com/ldqef3j.

M4P. 2008. Making value chains work better for the poor: a toolbook for practitioners of value chain analysis, version 3. M4P, Phnom Pehn, Cambodia.

Maloni, M.J. and M.E. Brown. 2006. Corporate social responsibility supply chain: an application in the food industry. Journal of Business Ethics 68: 35-52.

Nicourt, C. and J. Cabaret. 2014. Ni patrons ni ouvriers: le cas des eleveurs intégrés. La Nouvelle Revue du travail 5.

Osterwalder, A. and Y. Pigneur. 2010. Business model generation: a handbook for visionaries, game changers, and challengers. John Wiley and Sons, Hoboken, NJ, USA. Pelletier, N. 2015. Life cycle thinking, measurement and management for food system sustainability.

Environmental Science and Technology 49: 7515-7519.

Porter, M. and M. Kramer. 2010. Creating shared value. Harvard Business Review January-Fe: 2-17.

Roguet, C. and M. Rieu. 2011. Les groupements d’éleveurs de porcs en france: une organisation originale.

Available at: http://tinyurl.com/maztyr4.

Roussillon, M.A. and V. Legendre. 2015. Adaptation de l’offre à la demande de produits de porc en France.

Journées Recherche Porcine 47: 191-196.

International Food and Agribusiness Management Review

474

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0061 - Tuesday, August 22, 2017 12:31:15 PM - IP Address:24.21.169.207

Monastyrnaya et al.

Volume 20, Issue 4, 2017

Soosay, C., A. Fearne and B. Dent. 2012. Sustainable value chain analysis – a case study of Oxford landing from ‘vine to dine’. Supply Chain Management 17: 68-77.

Soosay, C.A. and P. Hyland. 2015. A decade of supply chain collaboration and directions for future research.

Supply Chain Management 20: 613-630.

Stubbs, W. and C. Cocklin. 2008. Conceptualizing a sustainability business model. Organization and Environment 21: 103-327.

Teece, D.J. 2010. Business models, business strategy and innovation. Long Range Planning 43: 172-194.

Touboulic, A. and H. Walker. 2015. Love me, love me not: a nuanced view on collaboration in sustainable supply chains. Journal of Purchasing and Supply Management 21: 178-191.

Varsei, M., C. Soosay, B. Fahimnia and J. Sarkis. 2014. Framing sustainability performance of supply chains with multidimensional indicators. Supply Chain Management 19: 242-257.

Wognum, P.M., H. Bremmers, J.H. Trienekens, J.G.A.J. Van der Vorst and J.M. Bloemhof. 2011. Systems for sustainability and transparency of food supply chains – Current status and challenges. Advanced Engineering Informatics 25: 65-76.

International Food and Agribusiness Management Review

475

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0061 - Tuesday, August 22, 2017 12:31:15 PM - IP Address:24.21.169.207

OPEN ACCESS International Food and Agribusiness Management Review Volume 20 Issue 4, 2017; DOI: 10.22434/IFAMR2017.0003

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2017.0003 - Tuesday, August 22, 2017 12:31:57 PM - IP Address:24.21.169.207

Received: 2 January 2017 / Accepted: 20 March 2017

How do pricing and the representation of price affect consumer evaluation of nursery products? A conjoint analysis RESEARCH ARTICLE Zhiwen Zhu a, Bridget Beheb, Patricia Huddlestonc, and Lynnell Saged aAssociate

Professor, Business School, Huaiyin Institute of Technology, 89 Beijing Bei Road, Huaian, Jiangsu 223001, China; Visiting research scholar, Department of Horticulture, College of Agriculture and Natural Resources, Michigan State University, 1066 Bogue St., East Lansing, MI 48824, USA bProfessor

and dResearch Assistant, Department of Horticulture, College of Agriculture and Natural Resources, Michigan State University, 1066 Bogue St., East Lansing, MI 48824, USA

cProfessor,

Department of Advertising and Public Relations, College of Communication Arts and Sciences, Michigan State University, 404 Wilson Road, East Lansing, MI 48824, USA

Abstract This study investigates how pricing and the representation of price (location of price sign, area of price number on price sign, extrinsic cue on price sign) influence consumer evaluation of nursery products. An incomplete orthogonal set of levels for each factor was used in a 35 factorial conjoint analysis. A web-based survey was conducted in March 2015. The results show the four factors account for 50.24% of the variance in plant preference, the other factor (plant type) accounts for the remaining 49.76%. Relative importance decreases from price level (16.15%) to cue (12.27%) to area (11.04%) and location (10.79%). Consumer’s purchase intention varies by demographics. The price level-purchase intention relationship also depends on cue. Benefit cue is more likely to offset the adverse effect of a price increase on purchase intention than feature cue, and the price level-purchase intention relationship for plants with feature/benefit cue appears to be inverted U shape.

Keywords: nursery product, pricing, representation of price, perceived value, conjoint analysis JEL code: D03, M31, Q13 Corresponding author: zhiwen@msu.edu

Š 2017 Zhu et al.

477

Zhu et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2017.0003 - Tuesday, August 22, 2017 12:31:57 PM - IP Address:24.21.169.207

1. Introduction Although the green industry has grown slowly in recent years, it remains an important contributor to national, state, and local economies. In 2013, it contributed 0.72% of US gross domestic product and 1.11% of total workforce employment (Hodges et al., 2015). US garden retailers have been facing growing competition from mass merchandisers entering the garden market during the last two decades (IBISWorld, 2015). Meanwhile, the nursery industry as a whole was confronted with consumers having an increasing number of alternatives for their discretionary spending other than purchasing plants. Additionally, the population aged 50 and over, the industry’s largest market, grows slowly (Colby and Ortman, 2015).

So far, extensive research efforts have made to better understand consumers and their attitude and purchase behavior with nursery products. Researchers have investigated consumer preferences for the price and intrinsic attributes of the plant (Behe et al., 1999; Brascamp, 1996; Kelley et al., 2001; Mason, et al., 2008; Wollaeger et al., 2015) and the container (Kelley et al., 2001; Nambuthiri et al., 2015). Researchers also have examined the variation of preferences by the characteristics of the consumer (Behe et al., 2014, 2015, 2016). However, other than price level, only limited research has investigated the relevance of marketing factors such as branding (Behe et al., 2016; Collart et al., 2010), in consumer evaluation of nursery products.

Lack of this stream of literature, coupled with a long-standing need by retailers for this information, may contribute to the relatively slow growth of the nursery industry during the last decade. For this reason, the goal of the current study is to provide marketing professionals with insight and nursery retailers with information that improves the perceived value of nursery products. We hypothesize that perceived value of nursery products is a function of pricing, the representation of price, and the intrinsic attributes of the plant. We use a conjoint methodology to quantify the relative importance of these factors and attributes. This information will help producers decide whether to invest in product improvements (intrinsic attributes) or in marketing (extrinsic cues) to improve value perceptions. We also postulate that perceived value of nursery products may vary by the demographic characteristics of the consumer, such as age and/or gender. Furthermore, we investigate the potential interactions between marketing and consumer factors.

2. Conceptual framework and hypotheses development 2.1 Perceived quality A model first proposed by Dodds and Monroe (1985), and then improved by Zeithaml (1988), is adapted to ground our hypotheses (Figure 1). Early authors used the word ‘quality’ to refer to explicit features (i.e. properties or characteristics) of an object as perceived by a subject (e.g. Russell, 1912). This tendency to infer quality from specific attributes was termed as ‘surrogate-based preference forming behavior’ (Olshavsky, 1985). Attributes that signal quality were dichotomized into intrinsic and extrinsic cues (e.g. Hersleth et al., 2015; Szybillo and Jacoby, 1974). Intrinsic cues refer to the physical attributes of the product. In a plant, intrinsic cues would include such attributes as flower or leaf color, plant or leaf shape, and growing condition.

A plant producer cannot change the intrinsic attributes without changing the nature of the product. We use plant type as a proxy for all of the intrinsic attributes of the plant in the current study. Extrinsic cues are related to the product physically. In other words, they are exogenous to the product. For example, brand, safety certification and origin are all extrinsic cues (Grunert et al., 2015).

H1: Consumer’s purchase intention (PI) is related to the plant type of the plant.

Under conditions of imperfect information, consumers may perceive price to be a signal of inherent product quality. The price-perceived quality relationship has received much attention in behavioral price research (Somervuori, 2012). Nevertheless, the relationship has not been made clear yet (Boyle and Lathrop, 2009; Monroe and Krishnan, 1985; Monroe et al., 2015). An increasing number of researchers have recognized that the price-perceived quality studies should focus on the conditions under which price information is International Food and Agribusiness Management Review

478

Zhu et al.

Volume 20, Issue 4, 2017

Demographic factors Age

Gender

Marketing factors

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2017.0003 - Tuesday, August 22, 2017 12:31:57 PM - IP Address:24.21.169.207

Price level

Perceived price

Perceived sacrifice

Perceived quality

Perceived value

Location of price sign Area of price

Extrinsic cue

Purchase

Intrinsic attributes Plant type

Figure 1. The determinants of perceived value and purchase intention. likely to signal quality (Lalwani and Forcum, 2016; Olshavsky, 1985; Peterson and Wilson, 1985). There exist three types of factors affecting the use of price as a quality signal.

First, the price-perceived quality relationship may vary by product category. In categories with small price variation among products, the consumer may not attribute higher quality to products that cost only a few dollars more than those of competitors. Similarly, in categories with small quality variation among brands, price may be deemed only as a kind of loss or sacrifice, whereas in categories with significant quality variation among brands (such as cellular phones), price may function also as a signal to quality (Monroe and Krishnan, 1985).

H2: Consumer’s PI is negatively related to the price level of the plant.

Second, the price-perceived quality relationship may vary by the characteristics of the consumer. One such characteristic is the price awareness of the consumer: a consumer who does not know the reference prices in the market obviously is unable to use price to infer quality (Cheng and Monroe, 2013). Another characteristic is consumer’s knowledge about the product (Lambert, 1972). If a consumer has not enough product knowledge to judge the variation in quality, he may tend to use price and other extrinsic cues. Lalwani and Forcum (2016) showed that consumers high (vs low) in power distance belief have a greater tendency to use price to judge quality because they have a greater need for structure, which makes them more likely to discriminate between products and rank them based on price.

Third, other information available to the consumer may moderate the price-perceived quality relationship.

When other intrinsic and extrinsic cues to quality, such as the colors, brand names and advertisement, are highly accessible, the consumer may be more likely to use those cues than price (Monroe and Krishnan, 1985).

H3: The extrinsic cue and price level have an interactive effect on consumer’s PI. International Food and Agribusiness Management Review

479

Zhu et al.

Volume 20, Issue 4, 2017

H3.1: Consumers are more likely to purchase low priced plants than medium and high priced plants when there are no highly accessible extrinsic cues.

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2017.0003 - Tuesday, August 22, 2017 12:31:57 PM - IP Address:24.21.169.207

H3.2: Consumers are as likely to purchase low priced plants as medium and high priced plants when there are highly accessible extrinsic cues.

Which type of cue is more important in signaling quality to the consumer? While allocating limited resources to improve consumer perception of quality, producers and retailers face two alternatives: investing in product improvements (intrinsic cues) or in marketing efforts (extrinsic cues). Therefore, an answer to this question would help them make the more effective decision. However, extant research shows conflicting evidence of the importance for intrinsic (Acebrón and Dopico, 2000; Darden and Schwinghammer, 1985) and extrinsic cues (Richardson et al., 1994; Sawyer et al., 1979). Zeithaml (1988) indicated that consumers depend on intrinsic cues more than extrinsic cues under pre-purchase circumstances when intrinsic cues are search attributes (rather than experience attributes), and consumers depend on extrinsic cues more than intrinsic cues under initial purchase circumstances when intrinsic cues are not accessible. In most cases, plant type is both an accessible and a search attribute for consumers shopping at the garden market.

H4: The plant type is the most important factor that determines consumer’s PI. 2.2 Perceived value Figure 1 delineates that perceived value mediates the relationship between quality and purchase. As Olshavsky (1985) suggested, not all consumers want to buy the highest quality item in every category. Instead, quality appears to be factored into the implicit or explicit valuation of a product by many consumers (Zeng et al., 2011). A given nursery product may be of high quality, but if the consumer does not want the product, does not have enough money to buy it, or does not want to spend the amount required, its value will not be perceived as being as high as that of a nursery product with lower quality but a more affordable price. In other words, consumers may obtain more value from the lower quality plants because the low costs compensated for the reduction in quality.

Zeithaml (1988) defined perceived value as ‘the consumer’s overall assessment of the utility of a product based on perceptions of what is received and what is given.’ According to this definition, value perception involves a tradeoff between the benefit and cost components to get a product (Dan, 2008; Zeng et al., 2011).

The benefit components include important intrinsic attributes, extrinsic attributes, perceived quality, and other relevant high level abstractions. The cost components of perceived value refer to the perceived price and sacrifice (Völckner, 2008).

Some intrinsic attributes of nursery products, other than those signaling quality, could provide value to consumers. For instance, flavor is one important intrinsic attribute for edible plant (e.g. cilantro). Most food preparers know which flavors of edible plants their family would eat; only those flavors were considered to be acceptable to the family and therefore to have value. In addition to perceived quality and these intrinsic attributes, other higher level abstractions contributed to perceptions of value. An important higher level abstraction for plant is convenience. Some consumers do not want to water and take care of the plants very often. For this reason, cactus would add more value than begonia. These intrinsic and extrinsic lower level attributes add value through the higher level abstraction of convenience.

Extrinsic attributes often serve as ‘value signals’ and can substitute for active tradeoff between benefits and costs (Richardson et al., 1994). To judge from the nursery product category, cognitive resources of the consumer are limited. Therefore, we conjecture that rather than carefully having tradeoff between costs and benefits, most consumers tend to use cues, often extrinsic cues (signals such as ‘need little care and water,’ ‘color all the summer’), in value perceptions.

International Food and Agribusiness Management Review

480

Zhu et al.

Volume 20, Issue 4, 2017

H5: Consumer’s PI is related to the extrinsic cue of the plant.

H5.1: Consumers are more likely to buy the plant with intrinsic attributes printed on its price sign (feature cue) than without.

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2017.0003 - Tuesday, August 22, 2017 12:31:57 PM - IP Address:24.21.169.207

H5.2: Consumers are more likely to buy the plant with higher level abstractions printed on its price sign (benefit cue) than without.

2.3 Objective price, perceived price and sacrifice Figure 1 depicts the compositions of price: objective pricing (the actual price level) of a plant, perceived price and sacrifice of consumers. According to Jacoby and Olson (1977), the price perceived by consumers is often not the objective price. For instance, some consumers may note that the price of begonia is exactly $1.99 for a container, but others may perceive the price only as ‘cheap’ or ‘expensive’. Still others may not notice the price at all. From the consumer’s perspective, some scholars in the pricing field conceptualized the price as ‘what is given up or sacrificed’ to acquire a product (Monroe and Krishnan, 1985).

A growing body of research supports this difference between objective and perceived price (Allen et al., 1976; Gabor and Granger, 1961). Studies show that consumers are not always aware of the actual price of a product. Instead, they perceive price in ways that are relevant to themselves (Zeithaml, 1982, 1984). Levels of consumer attention, awareness, and knowledge of prices appear to be considerably lower than necessary for consumers to have accurate internal reference prices for many products (Cheng and Monroe, 2013; Dickson and Sawyer, 1986). Zeithaml and Berry (1987) revealed that price awareness varies by demographics, consumers who are female, married, older, and unemployed have the highest levels of awareness. Consumer attention to price is likely to be higher for expensive, durable goods and services than for cheap goods and services.

An additional factor contributing to the gap between actual and perceived price is the representation of price. An increasing number of marketing researchers have been investigating this issue (Monroe, 2003; Thomas and Morwitz, 2009). For example, research has revealed how the digital sequence in a price affects consumer evaluation of the price via a left-digit anchoring effect (Thomas and Morwitz, 2005), right-digit effect (Coulter and Coulter, 2007), or price precision effect (Thomas et al., 2010). Researchers also have shown how fonts affect consumer evaluation of price, whether in larger or smaller sizes (Coulter and Coulter, 2005) or in different colors (Puccinelli et al., 2013). Other researchers have indicated that perceived price is a function of the display location of the sale price relative to the original price, displaying the smaller sale price to the right (vs left) to the original price makes consumers easier to initiate the subtraction task, a phenomenon they refer to as the ‘subtraction principle’ (Biswas et al., 2013).

The space-based approach to visual attention assumes that, the probability of sampling features will be highest near the center of the item and drop off exponentially as distance from the center of the item increases (Van Oeffelen and Vos, 1982, 1983). The assumption that information about features are distributed over space is similar to assumptions made by Wolford (1975), Ratcliff (1981), and Ashby et al. (1996) to account for spatial factors in visual tasks. The assumption can be articulated in terms of the receptive fields of feature detectors in visual cortex: If an item falls in the center of a receptive field, the detector will respond strongly to it. If the item falls near the edge or on the edge of a receptive field, the detector will respond less strongly (Logan, 1996). Therefore, we conjecture that consumer’s PI is related to the location where the price sign is placed on the plant (hereinafter referred to as ‘location’) and the area where the price number is situated at the price sign (hereinafter referred to as ‘area’).

H6.1 (H6.2): Consumer’s PI is highest when the price sign (price number) is at the center location (middle area).

H7.1 (H7.2/H7.3): Consumer’s age moderates the price level-PI (location-PI/area-PI) relationship.

International Food and Agribusiness Management Review

481

Zhu et al.

Volume 20, Issue 4, 2017

H8.1 (H8.2/H8.3): Consumer’s gender moderates the price level-PI (location-PI/area-PI) relationship.

3. Method

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2017.0003 - Tuesday, August 22, 2017 12:31:57 PM - IP Address:24.21.169.207

3.1 Research design and stimuli In order to answer our research questions and test hypotheses, we adopted a conjoint design for this study.

The conjoint design consists of showing respondents a manipulated set of digitally modified plant displays, varying several factors. Conjoint analysis is then conducted to determine how much each of these factors contributes to overall PI of respondent. By analyzing how respondents score these displays, we can also compute the implicit valuation (part-worth utilities) of each level for each of the factors making up the displays.

We manipulated five factors (sign location, area on sign where price was displayed, type of cue, plant type, and price) for the plant displays. The first factor is the location of the price sign. We defined three levels for this factor: left, center and right. The second factor is the area of the price locating at the price sign, for which we defined three levels: top, middle and bottom. The third factor, the cue, was manipulated at three levels: none, feature (describing a feature or obvious product attribute on the sign, such as ‘novel shapes’) and benefit cue (describing on the sign a benefit to the consumer, such as ‘needs little care and water’). The fourth factor is the plant type, for which we defined three levels: ornamental (begonia), both ornamental and edible (cactus), and edible (cilantro). Finally, the fifth factor, the price, was also manipulated at three levels for the same size container holding the plant: low ($0.99), medium ($1.99) and high ($2.99).

We used SPSS orthoplan (IBM, Chicago, IL, USA) to generate an incomplete factorial, orthogonal design of 16 plant displays (Figure 2). Each display is a combination of choosing one level for each of the five factors (location, area, cue, plant type and price level). In the orthogonal design, the factor levels are selected such that, for each pair of the factors (say, X and Y), the high level of X appears equally often in displays that have a high level of Y as in displays that have a low level of Y, and vice versa. This design is highly efficient for part-worth utility estimation, and is thus frequently adopted in extant literatures (e.g. Behe et al., 1999, 2016; Brascamp, 1996).

Figure 2. Stimuli to determine the overall purchase intention.

International Food and Agribusiness Management Review

482

Zhu et al.

Volume 20, Issue 4, 2017

3.2 Measures The sixteen displays were presented to participants in a randomized sequence, for which participants were instructed to indicate their likelihood to buy on an 11-point Juster scale (0=no chance; 10=certain) (Brennan and Esslemont, 1994; Juster, 1966). Finally, we measured participants’ demographics. For multi-item measures, the construct score was calculated by averaging the different items.

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2017.0003 - Tuesday, August 22, 2017 12:31:57 PM - IP Address:24.21.169.207

3.3 Data collection and sample The survey was developed on a Qualtrics platform (Provo, UT, USA) and responses were generated from the panel maintained by GMI Lightspeed (New York, NY, USA) in March, 2015. We excluded incomplete questionnaires from our analyses and conducted quality checks by rejecting respondents who scored all 16 displays in the same way, or provided the same answer to reverse-scaled items. The final sample consists of 718 unique respondents (43.18% male; 56.82% female; mean age: 43.73 years) (Table 1).

Table 1. Summary statistics.

Variables

Number of Mean Variables respondents (SD) or frequency (%) Age (years) 718 43.73 (16.65) Ethnicity 18-29 217 30.22% White/Caucasian 30-49 238 33.15% African American ≥50 263 36.63% Hispanic Gender Asian Male 310 43.18% Native American Female 408 56.82% Other Adults (≥19) in household 718 2.51 (1.12) Highest level of education 1 99 13.79% Less than high School 2 331 46.10% High School/GED 3 173 24.09% Some college 4 74 10.31% 2-year college degree 5 27 3.76% 4-year college degree 6 10 1.39% Master’s degree 7 2 0.28% Doctoral degree (Ph.D.) ≥8 2 0.28% Professional degree (JD, MD) Children in household 718 0.71 (1.03) Household income ($) 0 434 60.45% Less than 19,999 1 122 16.99% 20,000-39,999 2 112 15.60% 40,000-59,999 3 42 5.85% 60,000-79,999 4 5 0.70% 80,000-99,999 5 2 0.28% 100,000-119,999 ≥6 1 0.14% 120,000-139,999 Residence location 140,000-159,999 Metropolitan 149 20.75% 160,000-179,999 Suburban 416 57.94% 180,000-199,999 Rural 153 21.31% 200,000 or more Prefer not to answer

International Food and Agribusiness Management Review

483

Number of Mean respondents (SD) or frequency (%) 587 54 54 41 8 9

81.75% 7.52% 7.52% 5.71% 1.11% 1.25%

11 109 154 83 268 74 4 15

1.53% 15.18% 21.45% 11.56% 37.33% 10.31% 0.56% 2.09%

58 142 136 103 93 54 36 25 13 9 18 31

8.08% 19.78% 18.94% 14.35% 12.95% 7.52% 5.01% 3.48% 1.81% 1.25% 2.51% 4.32%

Zhu et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2017.0003 - Tuesday, August 22, 2017 12:31:57 PM - IP Address:24.21.169.207

Of the total 718 participants, referring to the classification of other authors (Behe et al., 2016; Douglas, 1991; Howe and Strauss, 2009), 30.22% were classified as Gen Y (18-29 years old), 33.15% as Gen X (30-49 years old), and 36.63% as Boomers (≥50 years old). The mean number of adults (≥19 years old) in the household was 2.51 and the mean number of children 0.71; 57.94% lived in a suburban area and 81.75% were white/ Caucasian. More than one third had earned a 4-year college degree. Median household income was in the $60,000 to 79,999 range. The demographic characteristics of this sample are generally consistent with other samples of plant purchasers or gardeners (e.g. Behe et al., 2016; Mason et al., 2008).

4. Results and discussion 4.1 Conjoint analyses We used SPSS software version 19.0 (IBM, Chicago, IL, USA) to conduct the conjoint analyses. The price level was set to be linear, while all other factors were set to be discrete. Of the 718 respondents, 47 responded with the same PI for all 16 displays and were excluded from conjoint analysis. For each respondent, the part-worth utility for each level of each factor was computed to indicate the relative importance of each factor. The part-worth utility and relative importance means of each factor were calculated by averaging the individual scores over each sample. The correlation between the actual and predicted preferences was calculated for each respondent and tested for statistical significance (Table 2). In each group, the correlations are significant (Pearson’s R≥0.97; Kendall’s τ≥0.85), indicating good fits (Murtagh and Heck, 2012).

Table 2 shows the relative importance percentages for each factor (Panel A) and the part-worth utility means for each factor level (Panel B) over the total sample and across the age and the gender groups. Higher partworth means indicate a stronger PI. The range of the part-worth means (highest to lowest) for each factor provides a measure of how important the factor is to affect respondent’s PI. Factors with larger part-worth ranges play a more important role than those with smaller ranges. The importance percentages were derived by calculating the utility range for each factor respectively and dividing it by the sum of the utility ranges for all factors. Consequently, the percentages sum up to 100.

Plant type is the most important factor affecting respondent’s PI (49.76%), consistent with other studies (e.g. Behe et al., 2016; Mason et al., 2008). This result supports the H4. When considering the three levels, in support of H1, we find that the part-worth utilities of begonia (0.84), cilantro (-0.03) and cactus (-0.81) are in descending order and the differences between them are all significant (P<0.001). This indicates that respondents are more likely to buy begonia than cilantro, and cilantro than cactus. The relative importance percentage of price and cue is 16.15 and 12.27% respectively. In support of H2, PI is negatively affected by price (β=-0.50, P<0.001), and the part-worth utility of feature cue is lower than that of benefit cue (Mfeature=-0.20, Mbenefit=0.08; P<0.001), indicating benefit cue is more persuasive than feature cue.

Perhaps counterintuitively, we find feature cue is less likely to improve PI than none cue (Mfeature=-0.20, Mnone=0.11; P<0.001), and there exists no significant difference between benefit cue and none cue (Mbenefit=0.08, Mnone=0.11; P>0.10). These results support H5, but not H5.1 and H5.2. With the relative importance percentage being 10.79 and 11.04% respectively, location and area are relatively less important than other factors. However, when considering the three levels, we find that placing price sign at the center location of the plants is most likely to improve PI, whereas placing price sign at the right location of the plants is least likely to improve PI (Mleft=-0.02, Mcenter=0.07, Mright=-0.05; P<0.001). Meanwhile, printing price number on the middle area of the price sign is most likely to improve PI, whereas printing price number on the top area of the price sign is least likely to improve PI (Mtop=-0.06, Mmiddle=0.04, Mbottom=0.02; P<0.001).

These results support H6.1 and H6.2.

We followed the same methods as Behe et al. (2016) to calculate the monetary value for each factor level.

The range in part worth utility means is 1.65 units (0.84 for begonia and -0.81 for cactus), which is equal to $2 (equidistant range from low price to high price). Therefore, each unit of utility scores equals $1.21.

International Food and Agribusiness Management Review

484

Zhu et al.

Volume 20, Issue 4, 2017

Table 2. Relative importance means and utility.1,2,3 Factor

Panel A: relative importance means

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2017.0003 - Tuesday, August 22, 2017 12:31:57 PM - IP Address:24.21.169.207

Age4

Location Area Cue Plant Price Factor

Level

Gender

Gen X (N=128)

Gen Y (N=287)

Boomer (N=256)

Male (N=287)

Female (N=384)

9.54 10.79 12.65 50.54 16.48

12.21 12.36 12.87 46.67 15.90

9.72 9.65 11.38 52.93 16.32

12.28 12.62 12.59 46.29 16.23

9.67 9.87 12.02 52.35 16.09

Gender

Gen X (N=128)

Area

(1) Left

Gen Y (N=287)

0.04 (0.08)

0.01 (0.07)

(2) Center

-0.01 (0.09)

(3) Right

-0.03 (0.09)

(1) Top

-0.04 (0.08)

0.07 (0.08) δ23*** -0.08 (0.08) δ31† -0.04 (0.07)

(2) Middle

10.79 11.04 12.27 49.76 16.15

Panel B: part-worth means (SE) Age

Location

Total (N=671)

0.03 (0.09)

Boomer (N=256)

Male (N=287)

Female (N=384)

-0.07 (0.12) δ12*** 0.09 (0.14) δ23* -0.02 (0.14)

-0.04 (0.09) δ12* 0.07 (0.10) δ23* -0.04 (0.10)

-0.00 (0.08)

-0.09 (0.12) δ12*** 0.10 (0.14) δ23** -0.02 (0.14)

-0.07 (0.09)

Total (N=671)

0.06 (0.09) δ23*** -0.06 (0.09)

-0.02 (0.08) δ12*** 0.07 (0.10) δ23*** -0.05 (0.10)

-0.05 (0.08) δ12** 0.06 (0.09)

-0.06 (0.08) δ12*** 0.04 (0.10)

-0.02 (0.08) 0.01 (0.10) ΔYB* -0.00 (0.09) 0.02 (0.10) (3) Bottom 0.01 (0.09) 0.06 (0.08) 0.05 (0.10) ** * δ δ31† δ31 31 Cue (1) None 0.11 (0.08) 0.03 (0.07) 0.21 (0.12) 0.10 (0.09) 0.13 (0.08) 0.11 (0.08) δ12*** δ12***ΔYB*** δ12*** δ12*** δ12*** δ12*** (2) Feature -0.14 (0.09) -0.17 (0.08) -0.25 (0.14) -0.15 (0.10) -0.23 (0.09) -0.20 (0.10) δ23* δ23*** δ23*** δ23***ΔMF* δ23*** δ23*** (3) Benefit 0.03 (0.09) 0.04 (0.14) 0.06 (0.10) 0.10 (0.09) 0.08 (0.10) 0.14 (0.08) ΔXY† δ31*** δ31*ΔYB* 0.66 (0.07) 1.08 (0.12) 0.66 (0.09) 0.98 (0.08) 0.84 (0.08) Plant (1) Begonia 0.78 (0.08) δ12***ΔYB* δ12*** δ12***ΔMF* δ12*** δ12*** δ12*** (2) Cactus -0.72 (0.09) -0.80 (0.08) -0.88 (0.14) -0.65 (0.10) -0.81 (0.10) -0.93 (0.09) δ23** δ23*** δ23*** δ23***ΔMF† δ23*** δ23*** (3) Cilantro -0.06 (0.09) 0.14 (0.08) -0.01 (0.10) -0.04 (0.09) -0.03 (0.10) -0.20(0.14) *** *** *** *** δ31*** δ31***ΔYB† δ δ δ δ31 31 31 31 Price (1) $0.99 -0.48 (0.07) -0.41 (0.06) -0.62 (0.11) -0.40 (0.08) -0.58 (0.07) -0.50 (0.08) (2) $1.99 -0.96 (0.14) -0.81 (0.12) -1.23 (0.21) -0.80 (0.16) -1.15 (0.15) -1.00 (0.15) (3) $2.99 -1.44 (0.22) -1.22 (0.18) -1.85 (0.32) -1.20 (0.24) -1.73 (0.22) -1.50 (0.23) β -0.48 (0.04) -0.41 (0.02) -0.62 (0.03) -0.40 (0.03) -0.58 (0.02) -0.50 (0.02) *** *** Δ ** *** ***Δ ** *** *** YB MF *** *** *** *** *** Correlation Pearson’s R 0.98 0.99 0.98 0.97 0.99 0.98*** Kendall’s τ 0.86*** 0.95*** 0.85*** 0.86*** 0.91*** 0.86*** 1 ***, **, * and † denote significance level at 0.1, 1, 5 and 10%, respectively; 47 of the 718 respondents responded with the same purchase intention for all 16 displays and were therefore excluded from conjoint analysis.

2 δ = (i) – (j), i, j ∈ {1,2,3}. ij 3Δ XY = Gen X – Gen Y; ΔYB = Gen Y – Boomer; ΔBX = Boomer – Gen X; ΔMF = Male – Female.

4 Gen X = 30-49 years old; Gen Y = 18-29 years old; Boomer ≥ 50 years old.

International Food and Agribusiness Management Review

485

Zhu et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2017.0003 - Tuesday, August 22, 2017 12:31:57 PM - IP Address:24.21.169.207

Feature cue detracts 24.2 cents from the perceived value of the product; whereas benefit cue and none cue add 9.68 and 13.31 cents in value respectively. This result does not support H5.1 and H5.2, and is different from that of Leclerc and Little (1997) that showed feature cue would help to persuade customer to make purchase decision. Left and right location detract 2.42 and 6.05 cents respectively, whereas center location adds 8.47 cents in value. Top area detracts 7.26 cents, whereas middle and bottom area add 4.84 and 2.42 cents in value respectively. Although these results indicate that location and area are two relatively less important factors than the plant type and the price level, they are still considerable in the value perception of the respondents.

Next, we investigated whether these relative importance and utility scores vary by age and gender. We conducted separate conjoint analyses for the three sets of groups respectively. Table 2 shows the part-worth utility means for each level, and the relative importance means of the five factors for each group. For each group, plant, price and cue are the three most important factors, and the relative importance means of them are in descending order. For the Boomer, this is followed by the location than the area. However, for other groups, the cue is followed by the area than the location in terms of relative importance.

We found a few subtle differences in part-worth utility means by age and gender (Table 2). Firstly, for each group, begonia is the most preferred plant and cactus the least. However, Boomers (vs Gen Y) and females (vs males) are more likely to buy begonia (MBoomer=1.08, MGenY=0.66, P<0.05; MFemale=0.98, MMale=0.66, P<0.05). Females were slightly less likely to buy cactus than males (MFemale=-0.93, MMale=-0.65, P<0.10), and Gen Y is slightly more likely to buy cilantro than Boomers (MGenY=0.14, MBoomer=-0.20, P<0.10).

Secondly, Boomers (vs Gen Y) and females (vs males) are more sensitive to price while making purchase decision (βBoomer=-0.62, βGenY=-0.41, P<0.01; βFemale=-0.58, βMale=-0.40, P<0.01). These results support H7.1 and H8.1.

Thirdly, although the feature cue is least preferred for each group, it is relatively more persuasive for males (vs females) respondents (MMale=-0.15, MFemale=-0.23, P<0.05). Meanwhile, Boomers (vs Gen Y) respondents are more likely to buy product with no cue (MBoomer=0.21, MGenY=0.03, P<0.001). Gen Y is more likely to buy product with a benefit cue than the Boomers and Gen X (MGenY=0.14, MBoomer=0.04, MGenX=0.03, P<0.05). In addition, it is worth noting that the preference for benefit cue to none cue varies by the age of respondent. For Boomers, no cue is preferred to a benefit cue (Mnone=0.21, Mbenefit=0.04, P<0.001), whereas for Gen Y, a benefit cue is preferred to no cue (Mnone=0.03, Mbenefit=0.14, P<0.05).

Finally, inconsistent with H7.2 and H8.2, the part-worth utilities for each level of the location do not vary by age or gender. However, the Boomer respondents are more likely to buy plants with the price number in the middle area of the sign than the Gen Y (MBoomer=0.10, MGenY=-0.02, P<0.05). These results support H7.3 and H8.3.

4.2 The interaction of price and cue With the scores for different displays as the dependent measure, a 3 (location)×3 (area)×3 (price)×3 (cue)×3 (plant) repeated measure analysis of variance (ANOVA) was used to test the hypotheses 3. In support of H3, A significant ‘cue×price’ term clearly reveals this interactive effect (F (3,11469)=4.80, P<0.01). Table 3 indicates that, in support of H3.1, consumers are more likely to purchase low priced plants than medium and high priced plants when there are no accessible extrinsic cues (Mlow=7.05, Mmedium=5.10, Mhigh=5.10; P<0.001). However, H3.2 is only partially supported. Table 3 shows that, consumers are as likely to purchase low priced plants as high priced plants (Mlow=5.65, Mhigh=5.54; P>0.10), whereas less likely to purchase low priced plants than medium priced plants (Mlow=5.65, Mmedium=6.31; P<0.001) when there are accessible feature cues. When there are accessible benefit cues, consumers are less likely to purchase low priced plants than medium and high priced plants (Mlow=5.78, Mmedium=6.43, Mhigh=6.20; P<0.001).

International Food and Agribusiness Management Review

486

Zhu et al.

Volume 20, Issue 4, 2017

Table 3. Comparison of means by price and cue.1 Plant

Price

Low Medium High

None Feature Benefit Low None Feature Benefit Medium None Feature Benefit High None Feature Benefit

1 ***, **

N

Mean (SD)

a-b

5,744 2,872 2,872 5,744 2,872 2,872 5,744 2,872 2,872 2,872 1,436 1,436 1,436 718 718 1,436 718 718

6.58 (3.37)a 5.04 (3.47)b 5.77 (3.45)c 6.38 (3.55)a 5.73 (3.38)b 5.48 (3.33)c 6.07 (3.50)a 5.79 (3.44)b 6.04 (3.45)c 7.05 (3.39)a 5.65 (3.57)b 5.78 (3.59)c 5.10 (3.36)a 6.31 (3.27)b 6.43 (3.28)c 5.10 (3.32)a 5.54 (3.29)b 6.20 (3.29)c

a-c

F, P

b-c

1.54***

0.81*** -0.73*** F(2,11485)=204.02, P<0.0000

0.65***

0.90***

0.25*

F(2,11485)=75.66, P<0.0000

0.29***

0.03

-0.26*

F(2,11485)=7.02, P<0.0009

1.40***

1.27*** -0.13

-1.21*** -1.33*** -0.12

-0.44*

F(2,5741)=106.49, P<0.0000

F(2,2869)=52.87, P<0.0000

-1.10*** -0.66*** F(2,2869)=26.45, P<0.0000

and * denote significance level at 0.1, 1 and 5% based on Bonferroni test, respectively.

As shown in Figure 3, inconsistent with H5.1 and H5.2 and , when price level is low, displays without cues are evaluated more favorably than those displays with feature or benefit cue (Mnone=7.05, Mfeature=5.65, Mbenefit=5.78; F(2,5741)=106.49, P<0.001). The low price is convincing enough information that consumers are not influenced by more cues (e.g. feature or benefit). Conversely, in support of H5.1 and H5.2, displays without cues are evaluated less favorably than displays with feature or benefit cues when price level is medium (Mnone=5.10, Mfeature=6.31, Mbenefit=6.43; F(2,2869)=52.87, P<0.001) or high (Mnone=5.10, Mfeature=5.54, Mbenefit=6.20; F(2,2869)=26.45, P<0.001). In other words, for moderate and higher prices, consumers are influenced by feature and benefit cues to help convince them to buy the product.

7.0

none

6.5 Purchase intention

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2017.0003 - Tuesday, August 22, 2017 12:31:57 PM - IP Address:24.21.169.207

Begonia Cactus Cilantro

Cue

feature

benefit benefit

6.0 feature

benefit feature

5.5 none

none 5.0

low

medium

high

Price

Figure 3. The interaction effect of price (low vs medium vs high) and cue (none vs feature vs benefit) on purchase intention.

International Food and Agribusiness Management Review

487

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2017.0003 - Tuesday, August 22, 2017 12:31:57 PM - IP Address:24.21.169.207

Zhu et al.

Volume 20, Issue 4, 2017

Further indications of the impact of cue can be observed by comparing particular conditions involving feature cue with analogous conditions involving benefit cue. Referring to Table 3, we cannot find a pattern of results in which the scores of displays with benefit cue are invariably higher than the scores of displays with feature cue at a given level of price. Concretely, at the high price level, the displays with benefit cue are rated higher than those same displays with feature cue (Mfeature=5.54, Mbenefit=6.20; P<0.001; Bonferroni test). Whereas there is no significant difference between feature cue and benefit cue at the low (Mfeature=5.65, Mbenefit=5.78; P>0.15) or medium price level (Mfeature=6.31, Mbenefit=6.43; P>0.15). These results held for both plants, suggesting that benefit cue could be more likely to offset the adverse effect of price increase on PI than feature cue.

In addition, the results show a significant main effect for price on participant reactions in the expected direction that price negatively influences PI (Mlow=6.38, Mmedium=5.73, Mhigh=5.48; F(2,11485)=75.66, P<0.001). However, This result held only for displays without cue (Mlow=7.05, Mmedium=5.10, Mhigh=5.10; F(2,5741)=241.52, P<0.001). Whereas there existed a likely invert U relationship between price and PI for displays with feature (Mlow=5.65, Mmedium=6.31, Mhigh=5.54; F(2,2869)=11.47, P<0.001) or benefit cue (Mlow=5.78, Mmedium=6.43, Mhigh=6.20; F(2,2869)=9.55, P<0.001).

5. Conclusions The major results are summarized in Table 4. All hypotheses, except H7.2 and H8.2, are supported, with the exception that H3.2 is partially and H5.1 and H5.2 are conditionally supported. Among the five factors tested here, the representation of the total price has an impact not less than price per se. This is a new finding in the horticulture marketing literature. Price has been deemed as a determinant of purchase decision in most economics and marketing literatures. Here, we provide some evidence that the representation of price is, at least, as important as price per se. Therefore, producers and retailers should pay more attention to how and where to present the price to help positively influence consumersâ&#x20AC;&#x2122; purchase decision.

We find subtle differences by age and gender. Boomers (vs Gen Y) and females (vs males) are more price sensitive while making purchase decision. The Boomer consumers are more likely to buy the plants with price number on the middle area of the price sign than the Gen Y. We have not found that the location-PI relationship varies by age or gender. These data suggest that producers and retailers should use different price presentation strategies while entering these segments of market.

In addition, we find benefit cue is more likely to offset the adverse effect of price increase on PI than feature cue, and the price level PI relationship for plants with feature or benefit cue appears to be invert U shape.

We also find that consumerâ&#x20AC;&#x2122;s PI for plants with feature or benefit cue relative to none cue is dependent on price level. The low priced plants without cues are more favorable than that with cues, while the medium Table 4. Major results.

Hypothesis

Support

Hypothesis

Support

Hypothesis

Support

Hypothesis

Support

H1 yes H4 yes H6.2 yes H8.2 no H2 yes H5 yes H7.1 yes H8.3 yes H3 yes H5.1 depends on H7.2 no H3.1 yes H5.2 price level H7.3 yes H3.2 yes (partially) H6.1 yes H8.1 yes Other findings 1. The preference for each type of the plant varies by age and gender.

2. The preference for none cue varies by age, for feature cue varies by gender, for benefit cue varies by age.

3. Benefit cue is more likely to offset the adverse effect of price increase on purchase intention (PI) than feature cue.

4. Invert U shape price level-PI relationship for displays with feature or benefit cue.

International Food and Agribusiness Management Review

488

Zhu et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2017.0003 - Tuesday, August 22, 2017 12:31:57 PM - IP Address:24.21.169.207

or high priced plants without cues are less favorable than that with cues. We explain that low price and accessible extrinsic cues provide two conflicting signals to quality simultaneously, which lead consumers to be skeptical of the quality of the plant and therefore reduce their evaluation. However, this explanation needs more direct evidence.

We used an incomplete factorial, orthogonal design to collect data, which enabled us to analyze by using fewer groups of sample. However, one major limitation of orthogonal design is that it often confounds interactions (Green and Srinivasan, 1990), so that we can only use it to estimate main effects. Therefore, we can only draw a tentative conclusion about the interactive effect of price and cue by ANOVA in this study. Future research should further investigate this effect. Moreover, respondents may process live plants differently from plant pictures, even though the pictures were of high quality. Future research could consider using live plants as stimuli instead. Finally, more research is needed to reveal consumer’s visual attention to the representation of price and the extrinsic feature/benefit cues. That information may help us to better understand the underlying mechanism that the representation of price and extrinsic cues affects consumer perception of value.

Acknowledgments This work was supported by the Jiangsu Overseas Research & Training Program for University Prominent Young & Middle-aged Teachers and Presidents, and the Social Science Foundation of Jiangsu Province of China, Projects No. 14GLD001.

References Acebrón, L.B. and D.C. Dopico. 2000. The importance of intrinsic and extrinsic cues to expected and experienced quality: an empirical application for beef. Food Quality and Preference 11: 229-238.

Allen, J.W., G.D. Harrell and M.D. Hutt. 1976. Price awareness study. Food Marketing Institute, Washington, WA, USA. Ashby, F.G., W. Prinzmetal, R. Ivry and W.T. Maddox. 1996. A formal theory of feature binding in object perception. Psychological Review 103: 165-192.

Behe, B.K., M. Bae, P.T. Huddleston and L. Sage. 2015. The effect of involvement on visual attention and product choice. Journal of Retailing and Consumer Services 24: 10-21.

Behe, B.K., B.L. Campbell, H. Khachatryan, C.R. Hall, J.H. Dennis, P.T. Huddleston and R.T. Fernandez.

2014. Incorporating eye tracking technology and conjoint analysis to better understand the green industry consumer. Hortscience 49: 1550-1557.

Behe, B.K., P. Huddleston and L. Sage. 2016. Age cohort influences brand recognition, awareness, and likelihood to buy vegetable and herb transplants. Hortscience 51: 145-151.

Behe, B., R. Nelson, S. Barton, C. Hall, C.D. Safley and S. Turner. 1999. Consumer preferences for geranium flower color, leaf variegation, and price. Hortscience 34: 740-742.

Biswas, A., S. Bhowmick, A. Guha and D. Grewal. 2013. Consumer evaluations of sale prices: Role of the subtraction principle. Journal of Marketing 77: 49-66.

Boyle, P.J. and E.S. Lathrop. 2009. Are consumers’ perceptions of price – quality relationships well calibrated?

International Journal of Consumer Studies 33: 58-63.

Brascamp, W. 1996. Evaluation and measurement of consumer preferences for outdoor ornamental plants.

PhD thesis, Massey University, Palmerston North, New Zealand.

Brennan, M. and D. Esslemont. 1994. The accuracy of the Juster scale for predicting purchase rates of branded, fast-moving consumer goods. Marketing Bulletin 5: 47-52.

Cheng, L.L. and K.B. Monroe. 2013. An appraisal of behavioral price research (part 1): price as a physical stimulus. AMS Review 3: 103-129.

Colby, S.L. and J.M. Ortman. 2015. Projections of the size and composition of the US population: 2014 to 2060. US Census Bureau. Available at: http://tinyurl.com/ohxe279.

International Food and Agribusiness Management Review

489

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2017.0003 - Tuesday, August 22, 2017 12:31:57 PM - IP Address:24.21.169.207

Zhu et al.

Volume 20, Issue 4, 2017

Collart, A.J., M.A. Palma and C.R. Hall. 2010. Branding awareness and willingness-to-pay associated with the Texas Superstar™ and Earth-Kind™ brands in Texas. Hortscience 45: 1226-1231.

Coulter, K.S. and R.A. Coulter. 2005. Size does matter: the effects of magnitude representation congruency on price perceptions and purchase likelihood. Journal of Consumer Psychology 15: 64-76.

Coulter, K.S. and R.A. Coulter. 2007. Distortion of price discount perceptions: The right digit effect. Journal of Consumer Research 34: 162-173.

Dan, A. 2008. Predictably irrational: the hidden forces that shape our decisions. HarperCollins Publishers, New York, USA. Darden, W.R. and J.K. Schwinghammer. 1985. The influence of social characteristics on perceived quality in patronage choice behavior. In: Perceived quality, edited by J. Jacoby and J. Olson. Lexington Books, Lexington, MA, USA, pp. 161-172.

Dickson, P.R. and A.G. Sawyer. 1986. Point-of-purchase behavior and price perceptions of supermarket shoppers. Marketing Science Institute. Available at: http://tinyurl.com/maym95a.

Dodds, W.B. and K.B. Monroe. 1985. The effect of brand and price information on subjective product evaluations. Advances in Consumer Research 12: 85-90.

Douglas, C. 1991. Generation X: tales for an accelerated culture. St. Martin’s Press, New York, NY, USA. Gabor, A. and C.W.J. Granger. 1961. On the price consciousness of consumers. Journal of the Royal Statistical Society, Series C 10: 170-188.

Green, P.E. and V. Srinivasan. 1990. Conjoint analysis in marketing: new developments with implications for research and practice. Journal of Marketing 54: 3-19.

Grunert, K.G., S.M. Loose, Y. Zhou and S. Tinggaard. 2015. Extrinsic and intrinsic quality cues in Chinese consumers’ purchase of pork ribs. Food Quality and Preference 42: 37-47.

Hersleth, M., E. Monteleone, A. Segtnan and T. Næs. 2015. Effects of evoked meal contexts on consumers’ responses to intrinsic and extrinsic product attributes in dry-cured ham. Food Quality and Preference 40: 191-198.

Hodges, A.W., C.R. Hall, M.A. Palma and H. Khachatryan. 2015. Economic contributions of the green industry in the United States in 2013. Horttechnology 25: 805-814.

Howe, N. and W. Strauss. 2009. Millennials rising: the next great generation. Vintage Books, New York, NY, USA. IBISWorld. 2015. Nursery and garden stores in the US: market research report (Vol. 2016). Available at: http://tinyurl.com/lhlmlub.

Jacoby, J. and J.C. Olson. 1977. Consumer response to price: an attitudinal, information processing perspective.

In: Moving ahead with attitude research, edited by Y. Wind. American Marketing Association, Chicago, USA, pp. 73-97.

Juster, F.T. 1966. Consumer buying intentions and purchase probability: an experiment in survey design.

Journal of the American Statistical Association 61: 658-696.

Kelley, K.M., B.K. Behe, J.A. Biernbaum and K.L. Poff. 2001. Consumer preference for edible-flower color, container size, and price. Hortscience 36: 801-804.

Lalwani, A.K. and L. Forcum. 2016. Does a dollar get you a dollar’s worth of merchandise? The impact of power distance belief on price-quality judgments. Journal of Consumer Research 43: 317-333.

Lambert, Z.V. 1972. Price and choice behavior. Journal of Marketing Research 9: 35-40.

Leclerc, F. and J.D. Little. 1997. Can advertising copy make FSI coupons more effective? Journal of Marketing Research 34: 473-484.

Logan, G.D. 1996. The CODE theory of visual attention: an integration of space-based and object-based attention. Psychological Review 103: 603-649.

Mason, S.C., T.W. Starman, R.D. Lineberger and B.K. Behe. 2008. Consumer preferences for price, color harmony, and care information of container gardens. Hortscience 43: 380-384.

Monroe, K.B. 2003. Behavioral foundations for pricing management. In: Pricing: making profitable decisions (3rd ed.). McGraw-Hill/Irwin, Burr Ridge, IL, USA, pp. 102-127.

Monroe, K.B. and R. Krishnan. 1985. The effect of price on subjective product evaluations. In: Perceived quality, edited by J. Jacoby and J. Olson. Lexington Books, Lexington, MA, USA, pp. 209-232.

International Food and Agribusiness Management Review

490

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2017.0003 - Tuesday, August 22, 2017 12:31:57 PM - IP Address:24.21.169.207

Zhu et al.

Volume 20, Issue 4, 2017

Monroe, K.B., V. Rikala and O. Somervuori. 2015. Examining the application of behavioral price research in business-to-business markets. Industrial Marketing Management 47: 17-25.

Murtagh, F. and A. Heck. 2012. Multivariate data analysis (Vol. 131). Springer Science and Business Media, New York, NY, USA. Nambuthiri, S., A. Fulcher, A.K. Koeser, R. Geneve and G. Niu. 2015. Moving toward sustainability with alternative containers for greenhouse and nursery crop production: a review and research update.

Horttechnology 25: 8-16.

Olshavsky, R.W. 1985. Perceived quality in consumer decision making: an integrated theoretical perspective.

In: Perceived quality, edited by J. Jacoby and J. Olson. Lexington Books, Lexington, MA, USA, pp. 3-29.

Peterson, R. and W. Wilson. 1985. Perceived risk and price-reliance schema and price-perceived-quality mediators. In: Perceived quality, edited by J. Jacoby and J. Olson. Lexington Books, Lexington, MA, USA, pp. 247-268.

Puccinelli, N.M., R. Chandrashekaran, D. Grewal and R. Suri. 2013. Are men seduced by red? The effect of red versus black prices on price perceptions. Journal of Retailing 89: 115-125.

Ratcliff, R. 1981. A theory of order relations in perceptual matching. Psychological Review 88: 552-572.

Richardson, P.S., A.S. Dick and A.K. Jain. 1994. Extrinsic and intrinsic cue effects on perceptions of store brand quality. Journal of Marketing 58: 28-36.

Russell, B. 1912. The problems of philosophy. Oxford University Press, London, UK. Sawyer, A.G., P.M. Worthing and P.E. Sendak. 1979. The role of laboratory experiments to test marketing strategies. Journal of Marketing 43: 60-67.

Somervuori, O. 2012. Essays on behavioral pricing. PhD thesis, Aalto University, Helsinki, Finland.

Szybillo, G.J. and J. Jacoby. 1974. Intrinsic versus extrinsic cues as determinants of perceived product quality.

Journal of Applied Psychology 59: 74-78.

Thomas, M. and V. Morwitz. 2005. Penny wise and pound foolish: the left-digit effect in price cognition.

Journal of Consumer Research 32: 54-64.

Thomas, M. and V.G. Morwitz. 2009. The ease-of-computation effect: the interplay of metacognitive experiences and naive theories in judgments of price differences. Journal of Marketing Research 46: 81-91.

Thomas, M., D.H. Simon and V. Kadiyali. 2010. The price precision effect: evidence from laboratory and market data. Marketing Science 29: 175-190.

Van Oeffelen, M.P. and P.G. Vos. 1982. Configurational effects on the enumeration of dots: Counting by groups. Memory and Cognition 10: 396-404.

Van Oeffelen, M.P. and P.G. Vos. 1983. An algorithm for pattern description on the level of relative proximity.

Pattern Recognition 16: 341-348.

Völckner, F. 2008. The dual role of price: decomposing consumers’ reactions to price. Journal of the Academy of Marketing Science 36: 359-377.

Wolford, G. 1975. Perturbation model for letter identification. Psychological Review 82: 184-199.

Wollaeger, H.M., K.L. Getter and B.K. Behe. 2015. Consumer preferences for traditional, neonicotinoidfree, bee-friendly, or biological control pest management practices on floriculture crops. Hortscience 50: 721-732.

Zeithaml, V.A. 1982. Consumer response to in-store price information environments. Journal of Consumer Research 8: 357-369.

Zeithaml, V.A. 1984. Issues in conceptualizing and measuring consumer response to price. Advances in Consumer Research 11: 612-616.

Zeithaml, V.A. 1988. Consumer perceptions of price, quality, and value: a means-end model and synthesis of evidence. Journal of Marketing 52: 2-22.

Zeithaml, V.A. and L. Berry. 1987. The time consciousness of supermarket shoppers. Texas A&M University Working Paper. Texas A&M University, College Station, TX, USA. Zeng, F., Z. Yang, Y. Li and K. Fam. 2011. Small business industrial buyers’ price sensitivity: do service quality dimensions matter in business markets? Industrial Marketing Management 40: 395-404.

International Food and Agribusiness Management Review

491

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2017.0003 - Tuesday, August 22, 2017 12:31:57 PM - IP Address:24.21.169.207

OPEN ACCESS International Food and Agribusiness Management Review Volume 20 Issue 4, 2017; DOI: 10.22434/IFAMR2016.0152

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0152 - Tuesday, August 22, 2017 12:38:00 PM - IP Address:24.21.169.207

Received: 15 September 2016 / Accepted: 1 March 2017

Management support for agricultural enterprises: a case study for a fruit-producing company CASE STUDY Leonardo Ensslina, Vinícius Dezem b, Ademar Dutraa, Sandra R. Ensslinc, and Karine Somensid aProfessor

and bResearcher, University of South Santa Catarina (Unisul), Av. José Acácio Moreira, 787-Dehon, Tubarão, SC 88704-900, Brazil

cResearcher

and dStudent, Federal University of Santa Catarina (UFSC), Campus Reitor João David Ferreira Lima, Trindade, Florianópolis, SC 88040-900, Brazil

Abstract Over the last few decades, agricultural production has improved its productivity significantly. This improvement has notably focused on the genetic development of plants and on equipment technology. Nevertheless, its managers have not monitored this growth in the competitiveness of the sector. Based on this context, this study has a central question: how can performance evaluation, from its decision-making support perspective, improve the management process of an agricultural company? In order to answer this, the purpose of creating a decision-making model to support the management decisions of an agricultural company is taken into consideration. Seven strategic objectives were identified, operationalized by 57 performance indicators, for the levels of reference set by the decision maker; fourteen performance indicators are at a compromising level, showing the need for intervention. With the model created, it was possible to have a picture of this situation and provide a process to propose improvement actions.

Keywords: performance evaluation, fruit production, MCDA-C, multi-criteria JEL code: M11 Corresponding author: vinidezem@yahoo.com.br

© 2017 Ensslin et al.

493

Ensslin et al.

Volume 20, Issue 4, 2017

1. Introduction

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0152 - Tuesday, August 22, 2017 12:38:00 PM - IP Address:24.21.169.207

Agricultural productive systems use sophisticated techniques to correlate human, natural, industrial and economic resources. This is undertaken to meet the demand for food in today’s highly competitive and exacting market in terms of environmental and social sustainability (Bronnmann and Asche, 2016; Figueiredo Junior et al., 2016; López et al., 2008; Neves and Chaddad, 2012).

In these contexts, performance evaluation systems must be used to generate useful information for managers.

Such systems enable managers to foresee the consequences of potential decisions on the aspects they regard as critical to the success of the enterprise (De Barros et al., 2009; Neves et al., 2015).

A great deal of scientific effort has been dedicated over the years to ensuring the well-balanced and sustainable development of the agricultural sector. Such efforts have given rise to the genetic engineering of species and the development of cost-effective equipment that is better adapted to its purpose. In the area of management, science has offered performance evaluation systems, which, though greatly based on mathematical calculations, have not met the needs of managers in this area in practical terms. Recent studies have shown that general management models do not provide the enterprise-specific detail required in such cases. Despite their success in other academic fields, such as physics and mathematics, these models ultimately frustrate managers and confuse researchers (Roy, 1993). Researchers such as Roy (1993), Landry (1995), Dantsis et al. (2010), Scott et al. (2015) and Keeney (1992) found that decision-making environments in areas with rapid development, such as the agricultural sector in recent years, have found a competitive edge in the singularities of their physical context and managers’ values and preferences. Therefore, attempts to use systems to support the management of agricultural enterprises that have been developed using information gathered from outside the specific decision-making context of this sector have often not been satisfactory.

Over the last few years, the mid-western region of the Brazilian State of Santa Catarina has shown conditions suitable for growing fruit in terms of its climatic characteristics, the suitability of its soils, the state’s favourable logistics system and, mainly, the profiles of its entrepreneurs. One such fruit producer, located in the western region of the state, is Sitio do Vale. It is a young company, characterised by the technology used in its fruit production; these are determining factors in how it grows distinctive products of a higher standard than its competitors. The company’s productive potential is currently low regarding market demands. This necessitates the use of a management tool to help improve the company’s production performance without negatively impacting product quality.

As this is an agricultural context undergoing rapid technological development, and highly influenced by its managers’ choices and involving many actors, with multiple conflicting and poorly defined objectives with unique characteristics, the use of general models is not advisable. The Multi-Criteria Decision AidingConstructivist methodology (MCDA-C) is therefore used as an instrument of intervention as it can deal with complex and conflicting contexts (Ensslin et al., 2012, 2015; Lacerda et al., 2011; Tasca et al., 2012; Zamcopé et al., 2010).

By understanding the aforementioned context, this study addresses the following research question: how can a constructivist model improve an agricultural company’s management process in terms of its decisionmaking? This study aims to answer this question by creating a model that can support the decision-making processes of the management of an agricultural company, using its managers’ perceptions to: 1. identify those aspects (performance indicators, criteria) regarded as critical for the company’s performance and the creation of scales used to measure performance and show levels of reference; 2 show the profile of the current performance level (status quo), taking into account the criteria (indicators) set for the decision-making context; 3. suggest actions to improve indicators at a compromising performance level.

International Food and Agribusiness Management Review

494

Ensslin et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0152 - Tuesday, August 22, 2017 12:38:00 PM - IP Address:24.21.169.207

This study is justified by its uniqueness, importance and viability (Castro, 1977). It is unique because no other articles were found in the literature searched that address management in agricultural companies or take into account performance evaluation as a system to support decision-making and the specifics of the context and authors involved in the management process.

By the end of this study, a model will have been created, which allows managers to be familiar with: the critical factors for the success of the company; the current performance level in each of these factors, including which factors present compromising characteristics and which provide a competitive edge; the current goal of each indicator; and how to use the process available to improve strategic actions. All of this information helps managers broaden their understanding of financial and non-financial consequences for their organisation in decision-making processes.

In addition to this introductory section, this study has four more sections: theoretical framework, methodology of the research, case study, and final considerations.

2. Theoretical framework In this section, the research axes are presented: (1) constructivist performance evaluation and (2) management in agricultural production, taking its development and recently published articles into consideration.

2.1 Constructivist performance evaluation Performance evaluation and its indicators have gone hand in hand with humankind since the very first signs of life in groups of human beings. Its documented beginnings are found in the Tratactus de Computis et Scripturis do Summa de arithmetica, geometrica, proportioni et proportionalita by Frei Luca Bartolomeo de Paccioli (1494), containing descriptions of Venetian merchantsâ&#x20AC;&#x2122; accounting methods (Ensslin et al., 2015).

Over the course of the eighteenth-century Industrial Revolution, Francis Bacon (1620) added an experimental scientific characteristic to performance evaluation, initially to operations and quality control purposes.

This suggested that processes could improve competitive conditions based on reducing production costs.

Performance evaluation was further consolidated in the twentieth century with mass production in the industrial segment and scientific administration taking shape within academia. In this context, realistic models gained importance and were successful in such areas as physics, mathematics and business administration.

Constructivist approaches began to be used in recent decades, as proposed by Landry (1995), Roy (1993), and Keeney (1992), helping researchers and managers in their professional and personal capacities. Such studies have made use of scientific knowledge to perfect and innovate commercial methods and courses of action.

The multiplicity of definitions of performance evaluation and the existing knowledge gaps regarding the decision-making support perspective prompted Ensslin et al. (2010: 130) to propose the following conceptualization for performance evaluation as an instrument to support decision making: Performance evaluation is characterized as the process which aims to build knowledge in the decision maker regarding a specific context which he/she intends to evaluate, by means of activities that from the perception of the decision maker him/herself identify, organize and measure, both ordinally and cardinally, integrate and allow to see the impact of actins and their management (Ensslin et al., 2010: 130).

According to this perception, performance evaluation is defined as a management tool conceived to build, establish and disseminate knowledge so that it is possible to monitor and improve the context in which a decision maker performs their managerial function (Ensslin et al., 2014). This is the perception proposed for the type of management described in this study, as presented in Figure 1.

International Food and Agribusiness Management Review

495

Ensslin et al.

Volume 20, Issue 4, 2017

Model to performance evaluation Identify what is important (criteria)

Establish standards for reference

Process for performance evaluation

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0152 - Tuesday, August 22, 2017 12:38:00 PM - IP Address:24.21.169.207

Identify how to measure the scales

Feedback

Measure the performance Form judgment

Management Subjective

Objective

Figure 1. Process of performance evaluation (adapted from Ensslin et al., 2007).

By creating a performance evaluation model with a focus on decision-making support, a process is used that consists of phases that integrate the subjectivity and objectivity of the decision-making context by structuring the addressed problem by in turn identifying the main goals and concerns of the decision makers.

Performance scales and levels are then defined to measure such concerns, which are based on the following statement by Lord Kelvin: When you can measure what you are speaking about, and express it in numbers, you know something about it; but when you cannot measure it, when you cannot express it in numbers, your knowledge is of a meagre and unsatisfactory kind: it may be the beginning of knowledge, but you have scarcely, in your thoughts, advanced to the stage of science (Thomson, 1968: 53).

This is the main responsibility of performance development systems; this study will develop its model based on this perspective using the MCDA-C. 2.2 Management in agricultural production Over the last years, fruit production has been centralized in family-based farming â&#x20AC;&#x201C; small productive units with low technical development (Bronnmann and Asche, 2016). As large urban centers started to form, demand for high quantities of food emerged, providing favorable conditions for investment in management in the agricultural sector.

A variety of management tools are used in food production, among which performance evaluation stands out (De Barros et al., 2009; Dantsis et al., 2010; LĂłpez et al., 2008). In a recent bibliometric analysis, a significant fraction of the literature on this topic was found to include prominent use of the multi-criteria methodology. Table 1 shows a summary of these works, highlighting: (1) objectives; (2) methodology used; and (3) results achieved.

Table 1 shows a tendency towards multi-criteria research in agricultural management, as the present authors regard the analysis of subjective and objective aspects as necessary elements in managersâ&#x20AC;&#x2122; decision-making contexts. In line with the information in Table 1, the present authors used realistic models (normativist, descriptivist) to aid agricultural managers. According to Roy (1993), the use of the normativist/descriptivist path describes realities independently of decision makers and other human actors. This approach tends to impoverish the managerial reality and decision-making process. The formulation of a problem cannot be International Food and Agribusiness Management Review

496

Ensslin et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0152 - Tuesday, August 22, 2017 12:38:00 PM - IP Address:24.21.169.207

Table 1. Summary of the multi-criteria studies (adapted from Bibliographic analysis, 2016).

Article

Objective

Model used

Results

Multi-metric evaluation of leaf wetness models for large-area application of plant disease models (Bregaglio et al., 2011).

Evaluate irrigation models and their impact on large farming areas, based on the comparison of six models.

A methodological approach to assess and compare the sustainability level of agricultural plant production systems. (Dantsis et al., 2010).

Evaluate and compare the level of sustainability of certain Greek productive systems, taking into account three sustainability pillars: Environmental, Economic and Social.

Test the hypothesis of superiority of the production of organic olive trees with conventional systems in the south of Spain.

The models evaluated had their application in the US and Italy, and their results were compared with a multi-criteria model, which evaluated irrigation capacity and simulated the results of pathogen infection in plants.

21 individual indicators were generated with integration in a single score system by means of Multi-attribute Value Theory.

The classification model and classification and regression tree achieved the best performance in most of the conditions; the authors suggest the adoption of decisionmaking support in future research.

The results show the particularities of the regions studied regarding sustainability. The continuity of the research is suggested, with the same indicators in other regions.

Despite the existence of ideological tendencies, the results show that organic systems achieve better performance. According to the authors, there are still conflicting issues between environment and production.

3 different scenarios were presented to the management board of Porto Marghera, which considered using technologies to treat the soil, identifying the best use of the land.

A systemic comparative assessment of the multifunctional performance of alternative olive systems in Spain within an Analytic Hierarchy Process (AHP)extended framework (LĂłpez et al., 2008).

Using multiple indices to evaluate scenarios for the remediation of contaminated land: the Porto Marghera (Venice, Italy) contaminated site (Critto and Agostini, 2009).

This work suggests a set of indicators used to evaluate contaminated sites.

The work classifies and compares conventional and organic pig production, evaluating parameters such as nitrogen, ammonia, greenhouse gases and radius of odour discomfort.

Improve management capacity Energy evaluation and and decision making in economic performance of investments regarding banana banana cropping systems in Guadeloupe (De Barros et al., cropping in Guadeloupe.

2009).

Conventional versus alternative pig production assessed by multi-criteria decision analysis (DegrĂŠ et al., 2007).

A systemic analysis was performed, using multiple criteria, examining such questions as economy, technique, culture and environment. To this end, the AHP technique was adopted.

The DESYRE system was used (DEcision Support sYstem for REhabilitation), which is software composed of 6 different modules that are integrated into the final decision-making support module where scenarios and possible solutions are presented.

The multi-criteria model used compared the performance of the processes, which were evaluated by a jury composed of 16 experts.

The six main production systems were compared by means of a multi-criteria model which considered factors such as use of the soil, disease control, and healthy environmental practices.

International Food and Agribusiness Management Review

497

The organic production achieved the best evaluated performance; however, the variability of performance was a highlight.

The results indicate that banana cropping in Guadeloupe presents low environmental performance and improving performance would cause costs to be higher; therefore, subsidies from the government should be granted.

Ensslin et al.

Volume 20, Issue 4, 2017

treated independently of the relationship between an individual and reality. These findings led Roy (1993) to recommend the use of the constructivist path for such scientific contexts.

3. Research methodology

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0152 - Tuesday, August 22, 2017 12:38:00 PM - IP Address:24.21.169.207

For this study, the applied research methodology will be presented in two stages. The first stage refers to the methodological framework of the research; the second shows the instrument of intervention, the MCDA-C, using the Brazilian fruit producer Sitio do Vale to create a performance evaluation model.

3.1 Methodological framework According to Richardson (2008), the methodological framework comprises the following factors: ■■ Nature of the research: this study is characterized by its applied nature, as a case study, aiming to understand and address a real phenomenon by creating a performance evaluation model for the decision-making process of a certain company in the agricultural sector.

■■ Nature of the objective: it is characterized by being exploratory, as it expands knowledge in the decision maker involved in the process, aiming to develop the knowledge of a certain field of expertise and, from it, obtain a set of criteria which may be transformed into management performance indicators of Sitio do Vale, taking into account the personal perception and values of the decision maker himself.

■■ Problem approach: a qualitative-quantitative approach is considered, so that the qualitative aspects that occurred when the decision maker’s concerns were identified, during the stage of structuring the model, especially in the development of Primary Elements of Evaluation (PEEs) and Cognitive Maps, can be evaluated. Next, the quantitative stage takes place by developing ordinal scales of performance indicators.

■■ Data collection: data collection comprised primary and secondary data (Richardson, 2008). The primary data comes from the observations made when the status quo profile of the performance indicators took place. The secondary data is obtained from the analyses of documents and financial statements.

3.2 MCDA-C as an instrument of intervention The instrument of intervention chosen for the creation of the model was the MCDA-C, (Ensslin et al., 2010). This methodology was chosen because it meets the requirements for the creation of the model. The requirements taken into consideration were that: (1) decision makers wish to improve their understanding of a problem; (2) decision makers wish to present critical factors for the success of business management; and (3) decision makers want specific details of their environment taken into account.

One of the principles of the MCDA-C consists of incorporating objective and subjective elements that are present in the decision-making process (Ensslin et al., 2001). In management contexts, objectivity and subjectivity are inherent in the decision-making process, therefore situations that involve decision making need to be analyzed based on both of those elements (Bana e Costa, 1993; Micheli and Mari, 2014). Thus, the MCDA-C shifts the focus of the analysis from being ontological (knowing how reality is) to epistemological (expressing how reality is understood or perceived) (Micheli and Mari, 2014).

Drawing on the concept of constructivism, as proposed by Roy (1993), this study is based on the recognition that a decision maker must expand his or her understanding of the consequences of their decisions regarding the aspects that they deem to be important and, through constructed knowledge, evaluate these aspects and recommend improvements without imposing on the rationalism of objectivity (De Moraes et al., 2010; Ensslin et al., 2010; Lacerda et al., 2011; Roy, 1994; Skinner, 1986).

The MCDA-C process of intervention operates in a systemic and systematic way through three sequential and interactive stages: structuring, evaluation and recommendation (Bana e Costa, 1993; Ensslin et al., 2017; Lacerda et al., 2014; Longaray and Ensslin, 2015).

International Food and Agribusiness Management Review

498

Ensslin et al.

Volume 20, Issue 4, 2017

The first stage (structuring) focuses on understanding the problem in accordance with the decision maker’s perceptions. The main objective at this stage is to help actors to identify, characterize and organize the relevant factors in the decision-making process. Soft approaches to operational research structuring, such as Cognitive Mapping (Ackermann and Eden, 1998) or the Soft Systems Methodology (Checkland and Scholes, 1999) can be used at this stage to elicit knowledge and engage decision makers. This procedure is followed by an elaboration of a hierarchical structure that represents the decision maker’s judgments (Bana e Costa, 1993; Lacerda et al., 2014; Longaray and Ensslin, 2015). Finally, the process involves the construction of ordinal scales to measure each criteria and sub-criteria of the model (Ensslin et al., 2013).

The evaluation stage, as Bana e Costa (1999) explains, involves clarifying potential choices through the application of mathematical methods. These methods assist in modelling and aggregating decision makers’ preferences. The development of the evaluation model should provide decision makers with a tool to understand the different consequences of the alternative decisions for each criterion (Ensslin et al., 2001; Lacerda et al., 2014; Longaray and Ensslin, 2015).

The Macbeth method – Measuring Attractiveness by a Categorical Based Evaluation Technique (Bana e Costa et al., 2012a,b) is used at this stage to transform the ordinal scales into cardinal scales and to help the decision maker to establish the taxes between criterions.

The last stage of the MCDA elaborates on the recommendations. This stage involves a discussion of the possible actions that could help the decision maker improve the performance of the assessed object. These actions are specific for each case and are established following the analysis of the performance profiles.

This analysis identifies the criteria that the decision maker is expected to meet to improve performance. The sensitivity analysis can be performed whenever the effects of any variation in the model parameters are to be tested (Bana e Costa et al., 1999). As observed in the two previous stages, the recommendations stage does

Structuring stage

Evaluation stage

Recommendations stage

1

Contextualization

2

Hierarchical structure of value

33

Hierarquic Value Structure Construction of descriptors

4

Independence analysis

5

Construction of values functions

6

Determination of compensation rates

7

Sensitivity analysis

8

Formulation of recommendations

Figure 2. Stages of performance evaluation (adapted from Ensslin et al., 2001).

International Food and Agribusiness Management Review

499

Formulation of recommendations

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0152 - Tuesday, August 22, 2017 12:38:00 PM - IP Address:24.21.169.207

The method proposes to build knowledge in those involved in the process that will be reflected in a performance evaluation model. The decisions made from this model are based on what are believed to be the decisions most suitable for the given situation (Roy, 1993). By considering this information, the methodology is composed of three stages: (1) structuring; (2) evaluation; and (3) recommendation, as shown in Figure 2.

Ensslin et al.

Volume 20, Issue 4, 2017

not have a prescriptive character either (Roy, 1993; Roy and Vanderpooten, 1996). The recommendations derived during the final stage as well as those that originate during the process are a result of the learning generated due to participation in the construction of the model (Roy, 1993).

The implementation of the phases of the MCDA-C will be detailed in the description of the case study. The structuring and recommendation stages were addressed together.

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0152 - Tuesday, August 22, 2017 12:38:00 PM - IP Address:24.21.169.207

4. Case study This section shows how a performance evaluation model is structured for the management process of a certain company in the agricultural sector, supported by the value systems, interests and preferences of the decision makers. This takes into account stakeholders’ requirements in accordance with decision makers’ perceptions.

4.1 Identification of the decision-making context The western region of the Brazilian State of Santa Catarina is characterized by the strength of its family farming enterprises. Several varieties of food are produced in small agricultural units. Infrastructure and production volumes are limited as products are essentially handmade but control over the environmental impacts and the quality of the products is extensive, thus generating a great competitive differential.

In this context, we address the agricultural company Sitio do Vale. It faces several challenges, the greatest of which is establishing a competitive position based on management tools that enable it to improve its performance according to the particularities of its operating context.

Actors are the people involved in the decision-making process; they are classified as decision makers, stakeholders and facilitators (Ensslin et al., 2001) The models developed, as influenced by the MCDA-C, are structured based on the values and preferences of decision makers; therefore, it is important to highlight the importance of the appreciation of the actors’ subjectivity by valuing the internal and external particularities of the context, and the motivation and preferences of the decision makers (Roy, 1993).

■■ the actors involved in the question of improving the management process, to be presented, are as follows: decision makers – manager and business partner at Sitio do Vale; ■■ stakeholders: company employees; ■■ the ones influenced: clients; ■■ facilitator: Vinicius Dezem.

The title established was ‘Model to the Management Support of Sitio do Vale’.

■■ Primary elements of evaluation By contextualizing the question of improving the management process, the MCDA-C, in its structuring stage, continues to identify the PEEs. They represent the concerns, wishes and motivations associated with the values and objectives of the decision maker in a particular decision-making context. In the present case, the PEEs were identified by means of open interviews with the decision maker, where he was encouraged to detail the problem he faced (Ensslin et al., 2001; Lacerda et al., 2014; Longaray and Ensslin, 2015).

Each PEE can generate more than one concept as sometimes more than one objective can be related to the element of concern; for example, ‘Plant diseases’ represented three preoccupations: (1) to control diseases; (2) to prevent diseases; and (3) to have processes in place to do so. After interactions with decision makers, 98 PEEs were identified. Table 2 shows the PEEs obtained.

The PEEs allow subjective and context-dependent concerns or objectives to be identified in such a way that makes the values of interest explicit.

International Food and Agribusiness Management Review

500

Ensslin et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0152 - Tuesday, August 22, 2017 12:38:00 PM - IP Address:24.21.169.207

Table 2. Primary elements of evaluation.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

Knowledge Weather conditions Market Work force Input Distance Investment Brand Products Cost accounting Society Partnerships Maintenance Use Weather/time Insecurity Cash flow

33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49

18 19 20 21 22 23 24 25 26 27 28 29 30 31 32

Long term relationship Legalization Control Costs Production Product mix Productivity Rent Innovate Performance Motivation Planning Package Election Competition

50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

Delivery Incentives Develop Debts Guarantees Payroll Shipping Payment Default Surpluses Sustainable Understanding Leadership Disagreements Liquidity Plant diseases (to control them) Plant diseases (to prevent them) Plant diseases ( to have processes to deal with them) Human diseases Goals Motivation Commitment Sustainable business Soils Suppliers Purchasers Marketing Advertising Focus Brand Diversification Balance Points Cost-effectiveness Profitability

66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82

Monitoring of results Risk Security Access Market trends Visits Appearance Side activities Bills Expenses Communication Environmental strategies Investors Employees Challenges Update

83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98

Superiority Crop growing techniques Point of sale Industrialization New cultures Focus Plagues Weekend Opportunities Performance Research Extension Recession External environment Internal environment Decision making

The MCDA-C methodology recommends expanding understanding by identifying the direction of preference represented by each PEE as well as its psychological opposite to understand the minimum degree of acceptability of the underlying goal. This evolutionary form of presenting the PEE is called a concept or action-oriented concept (Ensslin et al., 2010; De Moraes et al., 2010).

It is important to highlight that each PEE can develop more than one concept; this occurs when more than one objective is related to the concern element. Table 3 shows the first five concepts associated with the first five PEEs mentioned above. In each concept where an ellipsis (...) is used, this should be interpreted as ‘it is preferable to’ or ‘instead of’ (De Moraes et al., 2010; Lacerda et al., 2011) Based on the initial understanding of the decision maker in addition to the knowledge built, with the identification of the PEEs and development of the concepts, the facilitator prompts the decision maker to define the major areas of concern indicated by candidates for Fundamental Points of View (FPVs). These are perceived by the decision maker as necessary and sufficient for the management of the context of the study.

International Food and Agribusiness Management Review

501

Ensslin et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0152 - Tuesday, August 22, 2017 12:38:00 PM - IP Address:24.21.169.207

Table 3. Concept of the five first primary elements of evaluation.

N°

Primary elements of evaluation

Concept (implied objective)1

1

Knowledge

2

Storms

3

Market

4

Workforce

5

Inputs

Build knowledge in those involved in productive processes...run tests and experiments, wasting time and money.

Improve prevention systems and create stocks in order to have supplies in case of incidents ... to be completely unprepared with unreliable systems.

Expand in and keep secure market ... having products and not knowing where to sell them.

Have qualified workforce, capable of developing related activities with fair pay ... work overload and unfinished work.

Quality and availability of necessary inputs for the activities developed, at right prices ... to obtain products found in distant areas.

1 An

ellipsis (...) should be interpreted as ‘is preferable to’ or ‘instead of’.

To progress with the sequence of this process, the facilitator exhaustively tests the FPVs, arranging concepts into their respective areas of concern. If some concepts cannot be grouped accordingly, a suitable FPV is analyzed together with the decision maker; any areas of concern without corresponding concepts should be excluded.

The resulting structure is called a Hierarchical Structure of Values and the areas that describe them are composed of the family of FPVs of the context, as shown in Figure 3.

The next stage of structuring consists of arranging these concepts into a hierarchy and organizing them in terms of their influence relations. In order to achieve this, cognitive maps are used, as proposed by Montibeller et al. (2008).

■■ Maps of mean-end relationships, hierarchical structure of values and descriptors At the very beginning, the analyst must try to understand ‘what the problem is’ from the perspective of the actors involved in a given situation. To aid in this understanding, many tools have been developed by researchers in the field to schematically represent the construction of the problematic situation. In this study,

Model to support the Sitio do Vale Management

Production

Market

Finances

People

Logistics

Strategy

5;9;15; 19;20;22; 23;24;35; 47;48;49; 55;62;78; 83;84;86; 89;97.

3;6;8;12; 18;26;30; 31;32;54; 56;57;58; 59;61;70; 71;72;76; 77;85;91; 93;94;95; 96.

7;10;14; 17;21;25; 36;37;38; 40;46;63; 64;65;74; 75;79.

1;4;11; 28;43;41; 44;45;50; 52;53;60; 80;82;90.

33;39;69.

16;27;29; 42;51;66; 67;73;81; 87;88;92; 98.

Infrastructure 2;13;68.

Figure 3. Areas of concern for the evaluation of the Management Process of Sitio do Vale. The numbers in the figure correspond to the primary elements of evaluation numbers of Table 2.

International Food and Agribusiness Management Review

502

Ensslin et al.

Volume 20, Issue 4, 2017

the technique identified as being the most appropriate was the Cognitive Maps proposed by Eden (1988, 1983) for its adequacy in the structuring of multicriteria models. These maps depict causal and influential relationships.

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0152 - Tuesday, August 22, 2017 12:38:00 PM - IP Address:24.21.169.207

According to the decision maker’s perspective, the concepts grouped into areas of concern, called FPVs, are used to start the process of creating maps of mean and end relationships. The purpose of these maps is to have a better understanding of each FPV so that they can be operationalized and measured (Montibeller et al., 2008).

This step consists of constructing a hierarchy of concepts and establishing influence links. Mean and end relationships are created by requesting that the decision maker talks about each concept and explains why it is important and can be resolved (Eden, 1988, 1983). Thus, when mapping focuses on the ends, the decision maker explains his system of values through the higher-level hierarchical concepts. The mapping of the means also provides a set of potential actions through subordinate hierarchical concepts. Following the establishment of concept hierarchy, the connections between concepts are made using influence relationships.

This procedure facilitates the creation of chains of concepts and, within them, the creation of new concepts to justify lines of argument, from strategic concepts to the most operational ones. Figure 4 shows one of the cognitive maps for FPV1: ‘Production – Product Area’.

Once all maps have been constructed, the MCDA-C methodology proposes, in order to continue its process of constructing the understanding, that the structure of influence relations be converted into a hierarchical structure of value. This incorporates the understanding of the preferred judgments of the decision maker in the model under construction (Keeney, 1992).

One concern to be considered is that initial maps should be tested to represent aspects of the context in order to be: essential, controllable, complete, measurable, operational, isolable, non-redundant, concise and understandable (Keeney, 1992).

35 – To have better products than the competitors’, create production with a competitive edge, adding value to the end product. To be just one more product in the market.

9 – Quality and diversity with the competitors’ competitive edge... a single product in a saturated market.

100 – To have products with size, color and smell which are better than those from the competitors.... To lose competitiveness.

5 – Quality and availability of the necessary inputs for the activities developed at right prices.

103 – To have products which meet fruit high standards...

Not recognized by quality.

22 – To have suitable and good quality (volume) in the development of process, balance between production and demand... lack of products, not meeting demand; lack of products in certain times of the year.

101 – To have a structured supplier selection process...

To have high cost inputs.

102 – To search for permanent clients. To have unbalanced demand.

Figure 4. Map of the mean and end relationships for the strategic objective: ‘Production – Product Area’.

An ellipsis (...) should be interpreted as ‘is preferable to’ or ‘instead of’.

International Food and Agribusiness Management Review

503

Ensslin et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0152 - Tuesday, August 22, 2017 12:38:00 PM - IP Address:24.21.169.207

Maps of mean and end relationships were created for all of the strategic objectives of the top-down hierarchical structure. The MCDA-C, in its process of expanding knowledge about the evaluated context, suggests that the structures of a causal relationship be transformed and transferred to the Hierarchical Structure of Values, as created and shown in Figure 4. For such, each of the cognitive maps, associated with each of the FPVs, has its concepts grouped into clusters representing subareas of concern to be addressed. These clusters are transferred to the Hierarchical Structure of Values, where they are called Elementary Points of View (EPVs).

After the transition stage was developed, it was possible to create the Hierarchical Structure of Values, where the EPVs are decomposed into EPVs and SubEPVs to enable them to be measured, resulting in: ■■ 53 EPV of the 3rd level; ■■ 11 EPVs of the 2nd level; ■■ 20 EPVs of the 1st level; ■■ 7 FPVs. The hierarchical structure of values, also called the tree structure, contains all of the points of view, which are branched until the moment it is possible to measure them; from this level on, the EPVs are operationalized by ordinal measurement scales called descriptors.

The next stage to be developed for each aspect identified in the hierarchical structure is the creation of descriptors. These are ordinal scales that indicate the direction in which the decision maker’s preference goes in relation to each item (Bortoluzzi et al., 2014; Dezem, 2015). According to Bortoluzzi et al. (2010: 12), ‘The measurement scale of each descriptor associates the decision maker’s abstract values with one or more physical properties of the objectives in the context’. Once each descriptor’s scales are created, the level of reference for the scales are set; such levels are called ‘Compromising’ (representing the performance evaluated by the decision maker as insufficient to keep competitive), ‘Market’ (representing acceptable performance corresponding to market standards) and ‘Excellence’ (representing a competitive edge in the market). At the end of this stage, the following quantities of descriptors in each FPV were identified: ■■ production: 12 criteria; ■■ market: 15 criteria; ■■ finances: 8 criteria; ■■ people: 7 criteria; ■■ logistics: 3 criteria; ■■ strategy: 7 criteria; ■■ infrastructure: 5 criteria.

With a total of 57 indicators in the Hierarchical Structure of Values, aiming at a better understanding of the descriptors and their measurement levels, Figure 5 represents the FPV Production and the descriptors of the EPV Product, with their levels of reference.

4.2 Performance profile To create the current performance profile of Sitio do Vale, in order to meet the second specific objective of this study, the performance in each of the 57 criteria was identified by means of primary data collection, during day-to-day activities, by means of management reports made available by the company and interviews with the decision maker. Figure 6 shows the same EPV, with a dotted line, which represents the profile of the descriptor evaluated.

After identifying the performance of the criteria, it is possible to determine which points of the process in question exceed the market performance level (good level), are at the market level (between good and neutral) or are below the market level (below and neutral). In the criteria where the performance level is below neutral, improvement actions must be promoted with efforts geared towards performance improvement (Ensslin et al., 2001).

International Food and Agribusiness Management Review

504

Ensslin et al.

Volume 20, Issue 4, 2017

Product

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0152 - Tuesday, August 22, 2017 12:38:00 PM - IP Address:24.21.169.207

Quality

Excellence Market Compromising

Competitive edge

Standardization

Number of competitive edge elements existing in the products in comparison with competitors

Percentage of the products regarded as ‘out of spec’

Continuous demand

Selection Percentage of inputs acquired by means structured supplier selection process

Percentage of products sold to ‘permanent client’

5 or more

5% or less

100%

80%

4

10%

90%

70%

3

15%

80%

60%

2

20%

70%

50%

1 or none

30% or more

50% or less

40% or less

Figure 5. Descriptors and levels of reference for the EPV Product – FPV Production. EPV = elementary point of view; FPV = fundamental point of view.

Product Quality

Excellence Market Compromising

Competitive edge

Standardization

Number of competitive edge elements existing in the products in comparison with competitors

Percentage of the products regarded as ‘out of spec’

Selection Percentage of inputs acquired by means structured supplier selection process

Continuous demand Percentage of products sold to ‘permanent client’

5 or more

5% or less

100%

80%

4

10%

90%

70%

3

15%

80%

60%

2

20%

70%

50%

1 or none

30% or more

50% or less

40% or less

Figure 6. Performance profile EPV1 Product – FPV Production. EPV = elementary point of view; FPV = fundamental point of view.

In Figure 6, the results achieved are presented, taking an Elementary Point of View into account. This aims to measure the performance level of the product; it is possible to see that one out of the four factors analyzed is at a compromising level, and the other three factors are at a market level.

International Food and Agribusiness Management Review

505

Ensslin et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0152 - Tuesday, August 22, 2017 12:38:00 PM - IP Address:24.21.169.207

With the second specific objective of this study at hand, which aims to show the performance profile of the management process of Sitio do Vale, the following was identified: ■■ In 3 aspects, the performance level exceeds the level of excellence.

■■ In 40 aspects, the performance level is at a market level.

■■ In 14 aspects, the performance level is at a compromising level.

For these 14 aspects at a compromising level, improvement actions were developed, as seen in Table 4.

5. Final considerations In recent decades, productivity in agricultural production has improved significantly. This improvement has notably focused on the genetic engineering of plants and advancements in equipment technology. Nevertheless, managers have not monitored growth in the competitivity of the sector.

By analysing the existing literature on this topic, an opportunity to perform a study was identified, using performance evaluation from the decision-making support perspective as a management tool for a particular company in the agricultural sector. In Bregaglio et al. (2011), the authors present a multi-criteria model and suggest new studies, taking into account a constructivist approach (López et al., 2008). They also present the multi-criteria approach in their work, highlighting the importance of taking into account the complexity of the given context when validating the models proposed.

Drawing on this theoretical foundation, the purpose of this study was to create a model that supports the decision-making process of the management of an agricultural company. This aim was met by achieving the following specific objectives: 1. identifying the aspects that show the performance of the process, regarding evaluation criteria (indicators), and creating scales to measure performance and show its levels of reference; 2. showing the profile of the current performance level (status quo), taking into account the criteria (indicators) set for the management process of Sitio do Vale; and 3. suggesting a course of action, structured for the improvement of indicators at a compromising performance level.

By facing the first specific objective, after interviewing the decision maker and contextualizing the problem, 57 PEEs were identified, which were expanded by means of action-oriented concepts, grouped into areas of concern, better understood by means of concepts, and then arranged into maps of mean and end relationships to clarify their strategic, tactical and operational contribution.

Table 4. Performance profile EPV1 Product – FPV Production.1 Descriptor

Action

Selection percentage of the inputs acquired by means of a supplier selection structured process Expected result Necessary resources Person in charge Commencement date End date Follow-up process frequency How to follow up Person in charge of follow-up process

Create a supplier database, taking into account cost-effectiveness of each input Have reliable suppliers, who offer a fair price for the inputs Time the manager takes to perform searches and create a database Manager February February Every week Number of registered suppliers Manager and business partner

1

EPV = elementary point of view; FPV = fundamental point of view.

International Food and Agribusiness Management Review

506

Ensslin et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0152 - Tuesday, August 22, 2017 12:38:00 PM - IP Address:24.21.169.207

These maps were grouped into clusters and subclusters, which were named according to what the decision maker associated with the set of concepts therein. They were then transferred to the Hierarchical Structure of Values to respectively compose the FPVs and the EPVs. The subclusters that compose the most extreme EPVs were used to support the process that identified the most suitable descriptor. The performance levels of these ordinal scales were classified as Excellence, Market and Compromising by the identification of reference levels classified as Good and Neutral.

Next, performance was presented, from global, strategic, tactical and operational levels, in each scale where the current situation was and presented by means of a graph and numbers. All of this information enabled the decision maker to be familiar with the aspects presenting weak and/or strong performance levels and their corresponding consequences at strategic and global levels, meeting the second specific objective of this study.

The decision maker’s participation in the whole process ensured that, on one hand, everything being developed corresponded with his perceptions and represented his values and preferences; on the other hand, his confidence in the created model helped him to use it in order to make his management stronger and more transparent.

He thereby felt more comfortable justifying his choices and showing how his process was developed.

The final specific objective was achieved when improvement actions were proposed for the fourteen identified descriptors at a compromising level; essentially, such actions rely on a performance process, which develops in a pragmatic manner.

Therefore, the general purpose of this study was met, with the creation of a performance evaluation model for the management process of Brazilian fruit producer Sitio do Vale, developed from the values and preferences of the company’s team manager.

Thus, the use of the MCDA-C methodology as the research instrument is justified for confusing environments involving multiple actors, with conflicting and partially set objectives. This study was based on representative studies such as Dantsis et al. (2010), López et al. (2008) and De Barros et al. (2009), and supported such studies by using performance evaluation in a manner which had not been used previously, taking into account decision-making process support.

Taking into account the assumptions of performance evaluation, as a tool to support decisions, there were scientific contributions for the management of the Sitio Vale agricultural enterprise. These highlighted the incorporation of the constructivist approach and establishing a structured management process capable of measuring the objective and subjective elements present in decision making. It is important to highlight the knowledge built in the decision maker, which fostered an appropriate positioning in the decision-making context, putting into practice strategies and actions consistent with the needs of the company.

By being graphically and ordinally aware of the situation analyzed, the decision-maker was provided with information with which they could improve their company. This gave them confidence regarding which factors to target for improvement actions, and to what extent they should be addressed in order to develop the business.

As number of research limitations must be acknowledged. The model herein presented is specific to a company in the agricultural sector; therefore, its direct application, without being adapted to a new context (other companies) is not recommended. The MCDA-C process used is, however, general and can be used in different contexts. Also, the model created takes the decision maker’s perceptions into account when dealing with his work team and managed context, which makes the model legitimate for this decision maker in this context.

In this sense, the following areas are suggested for future research: (1) adapting and applying the model created herein to other companies of the addressed sector, with other decision makers; (2) continuity in the creation process of the model, regarding evaluation, which corresponds with the methodology used International Food and Agribusiness Management Review

507

Ensslin et al.

Volume 20, Issue 4, 2017

(MCDA-C); and (3) monitoring the management of the performance of the process when faced with the improvement suggestions proposed in this study.

The model developed to aid decision makers is specific to Sitio Do Vale, but the constructivist process used is general and can be used to develop models to help other companies to monitor and improve their performance.

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0152 - Tuesday, August 22, 2017 12:38:00 PM - IP Address:24.21.169.207

References Ackermann, F. and C. Eden. 2011. Making strategy: mapping out strategic success. Sage, Thousand Oaks, CA, USA. Bacon, F. 1620. Novum organum. PF Collier and Son, New York, NY, USA. Bana e Costa, C.A. 1993. Três convicções fundamentais na prática do apoio à decisão. Pesquisa Operacional 13: 9-20.

Bana e Costa, C.A., J.M. De Corte and J.C. Vansnick. 2012b. MACBETH. Journal of Information Technology and Decision Making 11: 359-387.

Bana e Costa, C.A., L. Ensslin, E.C. Corrêa and J.C. Vansnick. 1999. Decision support systems in action: integrated application in multicriteria decision aid process. European Journal of Operational Research 113: 315-335.

Bana e Costa, C.A., R. Lopez and B. Baets. 2012a. The MACBETH approach for multi-criteria evaluation of development projects on cross-cutting issues. Annals of Operations Research 199: 393-408.

Bortoluzzi, S.C., S.R. Ensslin, L. Ensslin and L.C. Chaves. 2014. Indicadores de desempenho propostos em pesquisas nacionais e internacionais para avaliar redes de pequenas e médias empresas (PMEs).

Espacios 35: 18.

Bortoluzzi, S.C., S.R. Ensslin, M.V.L. Lyrio and L. Ensslin. 2010. Proposta de um modelo multicritério de avaliação de desempenho econômico-financeiro para apoiar decisões de investimentos em empresas de capital aberto. Accounting and Management 3: 92-100.

Bregaglio, S., M. Donatelli, R. Confalonieri, M. Acutis and S. Orlandini. 2011. Multi metric evaluation of leaf wetness models for large-area application of plant disease models. Agricultural and Forest Meteorology 151: 1163-1172.

Bronnmann, J. and F. Asche. 2016. The value of product attributes, brands and private labels: an analysis of frozen seafood in Germany. Journal of Agricultural Economics 67: 231-244.

Castro, C.M. 1977. A prática da pesquisa. McGraw-Hill, São Paulo, Brazil.

Checkland, P. and J. Scholes. 1999. Soft systems methodology in action: include a 30 years retrospective.

Wiley, New York, NY, USA. Critto, A. and P. Agostini. 2009. Using multiple indices to evaluate scenarios for the remediation of contaminated land: the Porto Marghera (Venice, Italy) contaminated site. Environmental Science and Pollution Research 16: 649-662.

Dantsis, T., C. Douma, C. Giourga, A. Loumou and E. Polychronaki. 2010. A methodological approach to assess and compare the sustainability level of agricultural plant production systems. Ecological Indicators 10: 256-263.

De Barros, I., J.M. Blazy, G.S. Rodrigues, R. Tournebize and J.P. Cinna. 2009. Emergy evaluation and economic performance of banana cropping systems in Guadeloupe (French West Indies). Agriculture, ecosystems and environment 129: 437-449.

De Moraes, L., R. Garcia, L. Ensslin, M.J. da Conceição and S.M. de Carvalho. 2010. The multicriteria analysis for construction of benchmarkers to support the clinical engineering in the healthcare technology management. European Journal of Operational Research 200: 607-615.

Degré, A., C. Debouche and D. Verheve. 2007. Conventional versus alternative pig production assessed by multicriteria decision analysis. Agronomy for sustainable development 27: 185-195.

Dezem, V. 2015. Modelo Construtivista para apoiar a gestão: o caso do processo de atendimento e negócios de uma agência bancária. Administration Master’s Dissertation. University of South Santa Catarina, Tubarão, Brazil.

Eden, C. 1988. Cognitive mapping. European Journal of Operational Research n. 36: 1-13.

International Food and Agribusiness Management Review

508

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0152 - Tuesday, August 22, 2017 12:38:00 PM - IP Address:24.21.169.207

Ensslin et al.

Volume 20, Issue 4, 2017

Eden, C., S. Jones and D. Sims. 1983. Messing about in problems. Pergamon, Oxford, UK. Ensslin, L., A. Dutra, S.R. Ensslin, L.C. Chaves and V. Dezem. 2015. Research process for selecting a theoretical framework and bibliometric analysis of a theme: illustration for the management of customer service in a bank. Modern Economy 6: 782-796.

Ensslin, S.R., L. Ensslin, F. Back and R. Tadeu de Oliveira Lacerda. 2013. Improved decision aiding in human resource management: a case using constructivist multi-criteria decision aiding. International Journal of Productivity and Performance Management 62: 735-757.

Ensslin, S.R., L. Ensslin, L.C. Chaves, M.G. Gabriel and C.C. Oliveira. 2014. Seleção e análise de conteúdo de um portfólio de artigos sobre a avaliação do desempenho logístico. Espacios 35: 21.

Ensslin, L., S.R. Ensslin and A. Dutra. 2007. Avaliação de desempenho: objetivos e dimensões. Avaliação de Políticas Públicas. Governo do Estado de Santa Catarina, Florianópolis, Brazil.

Ensslin, L., S.R. Ensslin, A. Dutra, N.A. Nunes, and C. Reis. 2017. BPM governance: a literature analysis of performance evaluation. Business Process Management Journal 23: 71-86.

Ensslin, L., S.R. Ensslin and G.C. Pacheco. 2012. Um estudo sobre segurança em estádios de futebol baseado na análise bibliométrica da literatura internacional. Perspectivas em Ciência da Informação 17: 71-91.

Ensslin, L., E. Giffhorn, S.R. Ensslin, S.M. Petri and W.B. Vianna. 2010. Avaliação do desempenho de empresas terceirizadas com o uso da metodologia multicritério de apoio à decisão-construtivista.

Pesquisa Operacional 30: 125-152.

Ensslin, L., N.G. Montibeller and S.M. Noronha. 2001. Apoio à decisão: metodologias para estruturação de problemas e avaliação multicritério de alternativas. Insular, São Paulo, Brazil.

Figueiredo Junior, H.S., M.P.M. Meuwissen, J. Amaral Filho and A.G.J.M. O Lansink. 2016. Evaluating strategies for honey value chains in brazil using a value chain structure – conduct – performance (SCP) framework. International Food and Agribusiness Management Review. 19 (3): 225-250.

Keeney, R.L. 1992. Value-focused thinking: a path to creative decision making. Harvard University Press, Cambridge, UK. Lacerda, R.T.O., L. Ensslin and S.R. Ensslin. 2011. A performance measurement framework in portfolio management: a constructivist case. Management Decision 49: 1-15.

Lacerda, R.T.D.O., L. Ensslin and S.R. Ensslin. 2014. Research opportunities in strategic management field: a performance measurement approach. International Journal of Business Performance Management 15: 158-174.

Landry, M. 1995. A note on the concept of problem. Organization Studies 16: 315-343.

Longaray, A.A. and L. Ensslin. 2015. Use of multi-criteria decision aid to evaluate the performance of trade marketing activities of a Brazilian industry. Management and Organizational Studies 2: 15-31.

López, P.C., J. Calatrava-Requena and T. De-Haro-Giménez. 2008. A systemic comparative assessment of the multifunctional performance of alternative olive systems in Spain within an AHP-extended framework. Ecological Economics 64: 820-834.

Micheli, P. and L. Mari. 2014. The theory and practice of performance measurement. Management Accounting Research 25: 147-156.

Montibeller, G., V. Belton, F. Ackermann and L. Ensslin. 2008. Reasoning maps for decision aid: an integrated approach for problem-structuring and multi-criteria evaluation. Journal of the Operational Research Society 59: 575-589.

Neves, M.F. and F.R. Chaddad. 2012. The benefits of sugarcane chain development in Africa. International Food and Agribusiness Management Review 15(1): 159-165.

Neves, M.F., E.S. Simprini and F.R. Valerio. 2015. Management level evaluation in distribution channels of agricultural inputs. Available at: http://tinyurl.com/khopyrt.

Richardson, R.J. 2008. Pesquisa social: métodos e técnicas. Atlas, São Paulo, Brazil.

Roy, B. 1993. Decision science or decision-aid science? European Journal of Operational Research 66: 184-203.

Roy, B. 1994. On operational research and decision aid. European Journal of Operational Research 73: 23-26.

Roy, B. and D. Vanderpooten. 1996. The European school of MCDA: emergence, basic features and current works. Journal of Multi-Criteria Decision Analysis 5: 22-38.

International Food and Agribusiness Management Review

509

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0152 - Tuesday, August 22, 2017 12:38:00 PM - IP Address:24.21.169.207

Ensslin et al.

Volume 20, Issue 4, 2017

Scott, J., W. Ho, P.K. Dey, S. Talluri. 2015. A decision support system for supplier selection and order allocation in stochastic, multi-stakeholder and multi-criteria environments. International Journal of Production Economics 166: 226-237.

Skinner, W. 1986. The productivity paradox. Harvard Business Review 75: 41-45.

Tasca, J.E., L. Ensslin and S.R. Ensslin. 2012. A avaliação de programas de capacitação: um estudo de caso na administração pública. Revista de Administração Pública 46: 647-675.

Thomson, W. and L.J. Kelvin. 1968. Popular lectures and addresses in Bartlett Bartlett’s Familiar Quotations.

Little Brown, Boston, MA, USA. Zamcopé, F.C., L. Ensslin, S.R. Ensslin and A. Dutra. 2010. Modelo para avaliar o desempenho de operadores logísticos: um estudo de caso na indústria têxtil. Gestão y Produção 17: 693-705.

International Food and Agribusiness Management Review

510

OPEN ACCESS International Food and Agribusiness Management Review Volume 20 Issue 4, 2017; DOI: 10.22434/IFAMR2016.0064

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0064 - Tuesday, August 22, 2017 12:32:27 PM - IP Address:24.21.169.207

Received: 6 March 2016 / Accepted: 18 March 2017

Governance structures and coordination mechanisms in the Brazilian pork chain – Diversity of arrangements to support the supply of piglets RESEARCH ARTICLE Franco M. Martins

a,b,

Jacques Trienekensc, and Onno Omtac

aPhD

Student, and cProfessor, Management Studies Group, Social Sciences Department, Wageningen University, Hollandseweg 1, 6706 KN Wageningen, the Netherlands

bResearcher,

Brazilian Research Agricultural Corporation (EMBRAPA), BR 153, Km 110. P.O. Box 21, Zip Code 89 715 899, SC Concórdia, Brazil

Abstract This paper depicts the main coordination mechanisms (CMs) included in governance structures used to support the supply of piglets in the Brazilian Pork Chain (BPC). Furthermore, it analyses how and why actors use plural forms of coordination to support similar transactions. Based on the literature and an exploratory study carried out in the BPC, we propose a framework to analyse how price, volume, quality and resource allocation are coordinated in a transaction. This paper builds on transaction cost economics in two ways.

First, it shows that to arrange a transaction, a buyer may set CMs in distinct positions within the markethierarchy continuum. Second, it shows that actors use plural CMs with different counterparties in similar transactions. We found four explanations for plural governance: market fluctuations, bargaining power of suppliers, stricter coordination and quality, and the exchange context.

Keywords: pork supply chain, coordination mechanisms, plural governance JEL codes: D230, L24, L220 Corresponding author: franco.martins@embrapa.br

© 2017 Martins et al.

511

Martins et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0064 - Tuesday, August 22, 2017 12:32:27 PM - IP Address:24.21.169.207

1. Introduction The literature on transaction cost economics theory (TCE) has paid little attention to the complexity of coordination mechanisms (CMs) that underlie governance structures (GSs) (Wever, 2012). Researchers have used different GSs, ranging within a continuum from market (‘buy’) to integration (‘make’) to explain coordination in food chains (Gellynck and Molnar, 2009; Raynaud et al., 2005; Schulze et al., 2007; Wever et al., 2010). However, a GS (e.g. contract) may incorporate CMs – such as quality, price, investments and volume – that may be located at different points in this continuum (Wever, 2012). Examining these CMs in an integrated way, supports more refined insights into how a GS coordinates different aspects of the exchange.

Next, the use of plural GSs (Bradach and Eccles, 1989; Ménard, 2013), to support transactions with different counterparties within a same supplying context, has attracted the interest of scholars. This organisational diversity, which in part contradicts the principle of the efficient alignment (Williamson, 1991), is largely present in different sectors. Technological uncertainty, development of mutual (supplier-buyer) skills, monitoring difficulties and strategies to handle problems in coordination are examples of explanations for this development (Heide, 2003; Ménard, 2013; Miranda and Chaddad, 2014; Mols et al., 2012; Parmigiani, 2007).

Brazil is the fourth largest global producer and exporter of pork. In terms of quality, the Brazilian Pork Chain (BPC) meets, predominantly, public regulations, which are sufficient to supply the internal market and the majority of importer countries. In addition, BPC meets specific requirements set by domestic buyers and importers. Although BPC shows little diversity in quality standards, chain actors use many types of GSs, combining different CMs, to support pig production. These characteristics fit an interesting object of research in TCE. The goal of this paper is to analyse the heterogeneity of CMs and GSs used to support transactions between farmers and buyers in BPC. It includes analysing how and why chain actors use plural forms of combined CMs in similar exchange relationships. The next section presents a theoretical discussion and the research questions. Section 3 describes the research methods. Section 4 describes the elaboration of a modified framework of CMs and the main characteristics of coordination in the BPC. Section 5 presents case studies on the complexity of CMs and use of plural forms of governance. Section 6 discusses the results reflecting on the literature. Finally, section 7 presents the conclusions.

2. Governance in food chains TCE poses three different attributes to which the problem of selecting a matching governance structure is paramount: asset specificity, uncertainty and measurement difficulties (Ghosh and John, 1999; Rindfleisch and Heide, 1997).

Asset specificity regards investments made to fit the requirements of a particular agreement, which lose their value if used in another relationship. For instance, a processor concerned with a strict quality requirement, may set, in contracts, price incentives for suppliers to invest in specific resources (e.g. facilities, computer controlled feeding). However, if one of these suppliers uses these resources in transactions not driven by the same standards, the returns decrease. Therefore, a GS (e.g. a contract) may include a safeguard to protect the investments against opportunistic behaviour (Klein, 1996).

Uncertainty stems from the environment and behaviour of transaction parties. Environmental uncertainty raises the transaction costs of adaptation and coordination (Ghosh and John, 1999; Rindfleisch and Heide, 1997; Williamson, 2008). Examples of uncertainty are changing customer requirements and information on quality (Martinez, 2012), market conditions (Heyder et al., 2010), public regulations and their enforcement (Ménard and Valceschini, 2005; Williamson, 2008; Zylbersztajn and Farina, 1999). To handle uncertainty a processor may use a governance structure specifying, for instance, standards and mechanisms of control on processes and inputs used by suppliers.

International Food and Agribusiness Management Review

512

Martins et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0064 - Tuesday, August 22, 2017 12:32:27 PM - IP Address:24.21.169.207

Measurement difficulties regard the complexity inherent in monitoring a transaction according to a desired performance. Ghosh and John (1999) define it as the degree to which the value of an actor’s contribution is not verifiable by ex post inspection of an output. This complexity poses difficulties in aligning incentives and may cause loss of value in the transaction. For example, to facilitate control on a credence attribute, which is not possible to verify in visual inspections (i.e. food safety), a buyer of livestock may provide a farmer with specific inputs (e.g. GMO free feed). On the one hand it facilitates control of farming processes. On the other hand, it increases the costs incurred by the buyer to produce and deliver these inputs (i.e. selecting feedstuff supplier, logistics).

2.1 Governance choices Coase (1937) launched the discussion on forms to support transactions by pointing out markets and internal organisation (hierarchies) as alternative arrangements used to produce a good at comparable (transaction) costs. With this rationale, decision makers would use a firm only if it produced at lower costs than market prices. Over time, hybrid GSs that range between market and hierarchy began to be analysed by TCE scholars (Ménard, 2004; Sauveé, 2013; Williamson, 1991). Parties to a transaction rely on hybrids to cope with the risks that accrue from the market on the one hand, and to reduce the costs of internal organisation on the other. For Williamson (1991), hybrids are intermediary forms of control where parties remain autonomous but become mutually dependent to some extent. Ménard (2004) adds that in hybrids, parties to a transaction rely on a ‘little help’ from the price system to make an exchange but do not unify ownership of resources.

As examples of hybrids, the author describes ‘franchising, collective trademarks, partnership, cooperatives, networks, alliances and contracts’.

TCE literature has presented different typologies of GSs, used to support transactions in food chains.

Gellynck and Molnár (2009), depicted product, chain level and country-specific characteristics of GSs used in European food chains. Raynaud et al. (2005) use six types of GSs following a hierarchical sequence – Spot market, Relational contract, Relational contract with approved partner, Formal written contract, Equity based contract and Vertical integration – to analyse the alignment between quality and GSs. Schulze et al.

(2007) present a typology of GSs used in pork chains: Spot market, Long-term, Relationships, Marketing contracts, Production contracts, Farming contracts and Vertical integration.

2.2 Coordination mechanisms GSs differ from one another in aspects such as formality, duration, resource allocation, quality requirements and monitoring. Therefore, comparing GSs for their cost efficiency (Williamson, 1991) has not been sufficient to depict more clearly which aspects each alternative (GS) coordinates. A GS is, indeed, a combination of CMs (Foss, 2002; Grandori, 1997) used to control different aspects of the exchange. For example, to support transactions with suppliers, buyers may use contracts (i.e. a GS) including standards for inputs and processes. To support compliance with such standards, the buyer may implement CMs such as monitoring schemes, grades of quality and price incentives (Boger, 2001; Martinez, 2012; Martinez and Zering, 2004).

Examining CMs included in a GS refines the understanding of how such GS supports an exchange and helps to distinguish, more clearly, different GSs used to support similar transactions (Grandori, 1997). However, the literature lacks integrated analyses on how these mechanisms jointly make up GSs. First, some studies focus on only one mechanism. Second, little attention is given to the fact that different aspects underlying a GS may be coordinated differently (i.e. by more hierarchical or market-like settings). Third, there is no exploration of how interactions between CMs affect coordination (Wever, 2012).

To fill these gaps, Wever (2012) proposed a framework that includes four CMs: Price, Volume, Quality and Investments. These CMs may assume different positions within the market-hierarchy continuum (Table 1).

To illustrate this, let us take an example of a transaction between a farmer and a processor. The farmer delivers the input with amounts defined in each transaction and prices set in a reference market. In addition, International Food and Agribusiness Management Review

513

Martins et al.

Volume 20, Issue 4, 2017

Table 1. Typology of contractual coordination mechanisms (adapted from Wever, 2012).

Coordination Variables mechanisms Price

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0064 - Tuesday, August 22, 2017 12:32:27 PM - IP Address:24.21.169.207

Volume

Quality

Investments

• Setter • Duration • Criteria • Duration • Amount • Specification • Setter • Monitor • Types • Sources

Values Market ‹–––––––––––––––––––––––––––––––––––––––––› Hierarchy Spot price with/ without fixed bonus Spot volume

Reference market price with/without variable bonus Fixed volume with min/max deviations

Fixed forward price with/without variable bonus Fixed volume

Spot market specifications/ Public framework No (external) investments used

Third party quality Counterparty coordination quality coordination Debt security Convertible debt security

Internal price with/ without variable bonus Internal volume

Internal quality coordination Equity security

the processor adds a bonus based on a specific standard. The buyer monitors farming processes to check compliance. In the end, no investments are required. This simple example shows that a GS may be a more complex arrangement than is assumed in the discretionary perspective (Raynaud et al., 2005; Schulze et al., 2007; Williamson, 1991).

Analyses of GSs may involve even more complexity. Firms may apply plural forms of governance (settings of CMs) to support similar transactions. This topic is discussed in the following section.

2.3 Plural forms As long as TCE has centred attention on identifying the most cost-efficient mode of organisation solution (Williamson, 1991), empirical evidence and literature have demonstrated that companies use more than one GS to support similar exchange relationships (Bradach and Eccles, 1989; Heide, 2003; Ménard, 2013; Mols et al., 2012; Parmigiani, 2007). Bradach and Eccles (1989) consider plural forms as ‘arrangements where distinct organisational control mechanisms are operated simultaneously for the same function by the same firm’. Ménard (2013) explains that actors rely on plural forms: ‘for a class of transactions dealing with the same activity and within the same institutional and competitive environment, a party uses simultaneously different modes of governance or relies simultaneously on substantially different types of contracts’. Plural governance takes place in supply (Heide, 2003; Parmigiani, 2007) and distribution (Bradach, 1997; Hendrikse and Jiang, 2011) relationships. This paper focuses on the first.

Studies have indicated that the combination of internal production and outsourcing can function as a source of knowledge and may increase the performance of buyers and suppliers. For Heide (2003), this combination serves as a selective strategy used when quality is difficult to assess and customised products are at stake. In addition, internal production helps a buyer to develop the skills to monitor suppliers. On the other hand, it ‘enables suppliers to self-select into a buyer relationship’ because they learn how to signal information to buyers. Miranda and Chaddad (2014), in line with the view on mutual learning, argue that a firm depends on its capabilities and resources to be effective in measuring the quality attributes of an input and to define the GSs used to support the procurement. For Mols et al. (2012) internal production, combined with outsourcing, enables buyers to assess the skills, facilities and quality control systems that suppliers use. In addition, combining internal and external supply moderates uncertainties in volumes, technology and specificity of assets and works as a safeguard for the termination of the relationship. Parmigiani (2007) also found that the use of plural GS can be beneficial for buyers and suppliers. For the author, the factors that drive the adoption of plural GSs are technological and performance uncertainty, scope economies and expertise of buyers and suppliers.

International Food and Agribusiness Management Review

514

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0064 - Tuesday, August 22, 2017 12:32:27 PM - IP Address:24.21.169.207

Martins et al.

Volume 20, Issue 4, 2017

Ménard (2013) explained that the principle of efficient alignment (Williamson, 1991) does not explain why actors set plural GSs to support an exchange. The author discussed drivers for plural governance found in literature (e.g. innovation, benchmarking, and credibility for the termination of a relationship) and proposed an integrated framework with three groups of explanations: ambiguity with respect to asset specificity, monitoring complexity and strategising. Ambiguity relates to difficulties an actor faces to, ex ante, evaluate the benefits that can be seized from transactions supported by distinct GSs. Therefore, an actor may use plural GSs to compare their respective advantages. Monitoring complexity relates to uncertainties an actor has in identifying an adequate way to monitor the transaction. It occurs, for instance, when a buyer deals with suppliers who use different technologies to produce the same input and each technology demands a distinct monitoring mechanism. Finally, strategising develops when a party faces difficulties in implementing the form of coordination that best fits his/her business view (i.e. cost advantages, reputation for quality) and is forced to implement another type of GS to support a part of the supply or distribution. For instance, suppliers may use bargain power to prevent buyers from controlling processes and/or inputs used in production.

2.4 Research questions This paper aims to depict and analyse the heterogeneity of GSs and underlying CMs used to support the supply of piglets in the BPC. It includes analysing how and why actors use different CMs and apply these in differentiated ways (plural forms) in similar transactions. The literature discussed above and an exploratory study conducted in the BPC enabled us to propose a modified framework to analyse different CMs and GSs used by chain actors. To achieve the goals of this study the following research questions were set: RQ1: which GSs and CMs are predominantly used to support the supply of piglets in the BPC? RQ2: how do distinct CMs differentiate in their position within the market-hierarchy continuum in GSs used to support the supply of piglets in the BPC? RQ3: why do actors rely on distinct CMs (plural forms) to support similar exchange relationships in the supply of piglets in the BPC?

3. Data collection Data were collected by means of semi-structured interviews (n=41) with representatives of the sector and the institutional environment (Table 2), between September 2014 and April 2015. The set of buyers include small, medium and big firms and cooperatives that together maintain the lion’s share of the domestic markets as well as the exports. For example, in 2014, the IOFs and cooperatives included in the sample together slaughtered, respectively, 45 and 19% of Brazilian production. The retailer and the information technology company are leaders in their respective sectors. Moreover, the two feed companies supply important firms and cooperatives in the BPC. The average interview duration was 86 minutes. The main topics of the interviews were quality and coordination. Regarding quality, interviewees were asked about aspects such as their view on quality standards (e.g. buyers requirements, regulations) developments, virtues and bottlenecks. Questions on coordination included the characteristics of CMs used to support production. They also included interviewees’ opinions on the strengths and bottlenecks of these relationships. The contents of the interviews were arranged in reports. The field research also relied on sectorial documents, buyers’ annual reports and manuals (good practices) and regulatory information.

International Food and Agribusiness Management Review

515

Martins et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0064 - Tuesday, August 22, 2017 12:32:27 PM - IP Address:24.21.169.207

Table 2. Interviews settings.

Type of interviewees, organisations and number of interviews (n)

n

Buyers1: IOFs, Coops, MIs

Famers associations (6 state and 1 local and the national association) Slaughterhouse associations (2 state and the national association) Information technology Retailer Feed/feedstuff companies Government: Agricultural Ministry and Brazilian Agricultural Research Corporation. (EMBRAPA) 1

Interview duration Min

Max

21

48

130

7

74

118

3

104

240

1 1 2

71 72 88

91

6

38

83

Interviewee function

States where interviews took place

Directors and managers in production, quality, exports, owners.

Presidents, executive directors, consultant

Rio Grande do Sul, Santa Catarina, Paraná, Goiás, Brasília, Mato Grosso.

Rio Grande do Sul, Santa Catarina, Paraná, Goiás, Minas Gerais and Brasília Rio Grande do Sul and Santa Catarina

Vice-President, executive directors Santa Catarina São Paulo Owner, technical adviser Staff of the Ministry areas: animal health, livestock production, foreign affairs, inspection service; researcher on animal health

Owner (Director) Development of Meat Supply Rio Grande Do Sul and Santa Catarina.

Santa Catarina and Brasília.

IOF = Investor owned firms; Coops = Cooperatives; MI = Mini integrations.

4. The Brazilian pork chain Between 2011 and 2015, Brazil had a share of 3% (3.3 million tons) of global pork production and 8.4% (590 kilo tons) of the exports (USDA, 2016). The commercial herd accounts for 1,600,000 sows and 39,000,000 pigs in the rearing stages (ABCS, 2015). The main importers of Brazilian pork in 2014 were Russia (38%), Hong Kong (22.6%), Angola (10.7%), Singapore (6.6%) and Uruguay (4.2%) (MAPA, 2015). Brazil’s most important regions of production are the south, the south-east and the mid-west. These regions comprise respectively 61, 21 and 16.5% of the Brazilian herd in terms of housed sows (ABCS, 2015).

Overall, BPC meets public standards and specific customer requirements. The Brazilian Ministry of Agriculture, Livestock and Food Supply (MAPA, 2015) sets the public regulations on animal health, food safety, and animal welfare. State level (environment) agencies set specific rules for the licensing of pig production.

These standards are sufficient to meet international standards mediated by the World Trade Organization. In addition, some importers require standards on substances used in the feed (e.g. Russia; China) and sanitary status of regions of production (e.g. Japan). Furthermore, to address their policies on quality, buyers set their own standards (e.g. biosecurity, genetics, welfare) to be met by farmers.

There are retailers that set requirements and carry out inspections over chain stages to accredit suppliers.

Overall, actors do not use specialised quality management systems (Wever et al., 2010) such as Protected Designation of Origin (PDOs), Protected Geographical Indication (PGIs), Traditional Speciality Guarantee, organic production, differentiated retail schemes and regional production adopted in Europe (Becker and Staus, 2009; Bonneau et al., 2011; Grunert et al., 2011; Trienekens et al., 2009; Verbeke et al., 2010). Indeed, actors use an array of GSs and underlying CMs to handle a non-diverse set of standards. This research identified five general types of supplier arrangements in the BPC: Spot Market, Mini-Integrations, Singular Cooperatives, Central Cooperatives and Investor Owned Companies.

International Food and Agribusiness Management Review

516

Martins et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0064 - Tuesday, August 22, 2017 12:32:27 PM - IP Address:24.21.169.207

Spot market (SM) arrangements support informal agreements with a low level of coordination. Transactions based purely on market mechanisms are rarely used in BPC. In this research, SM represents the exchanges in which the farmer has supplier agreements with different buyers. In these arrangements, farmers meet baseline public regulations and supply, mainly, local butcheries and slaughterhouses and other farmers.

Mini Integrations (MI) are arrangements coordinated by big producers or middlemen by means of formal or informal agreements with pig famers. In these transactions, the integrators may allocate feed and technical support in production, depending on the farming stage and type of agreement. MIs meet public standards and supply local and national slaughterhouses. MIs deliver pigs to different buyers by means of spot markets and/or contracts.

Singular Cooperatives (SC) produce by means of contracts with farmers that are also cooperative members.

In these agreements, the cooperative provides technical support, monitors production and set prices based on quality. Piglet farmers normally use resources sold or approved by the SC. These main customers of SCs are regional and national retailers. Some SCs export with baseline or stricter standards.

Central Cooperatives (CC) are big organisations (i.e. food companies) that hold affiliated cooperatives. To arrange the supply of pigs, the affiliated cooperatives use contracts. However, these contracts are established with farmers that are member of these cooperatives. The CCs set the quality standards member cooperatives use to produce pigs. Furthermore, CCs slaughter all production from their affiliates and deliver the pork products. National retailers and exports (with baseline and stricter standards) are the main channels to which CCs deliver pork.

Investor Owned Firms (IOF) arrange their supply by means of contracts with farmers. However, these companies use more hierarchical mechanisms in these contracts. For instance, firms focus on allocating feed and animals in all production stages. IOFs deliver pork to the national market and export with baseline and stricter standards.

In summary, these five arrangements use the same baseline requirements of quality. However, the major part of production (estimates based on data from SIPS, personal communication) meets stricter requirements and is reliant on coordination supported by contracts (Table 3). The following sections explain and illustrate CMs buyers use to support the supply of piglets in the BPC. Chain actors normally arrange production in a ‘three site’ system, with the rearing stages in different locations.

Weaning and farrowing are the main systems chain actors use to raise piglets. In the first, the piglets are born and raised until they reach a weight between 7 and 8 kg. Then the piglets are transferred to nurseries where they reach a weight between 22-25 kg. In the farrowing system, piglets are born and raised until they are 22-25 kg. Finally, piglets are delivered to fattening farms where they are raised until the slaughter (100-125 kg). In the ‘wean to finish’ model, pigs enter the farm at 7-8 kg and are raised until slaughter. The Table 3. Supplying arrangements in the Brazilian pork chain.

Main characteristics

Spot market

Mini integrations

Singular cooperatives

Central cooperatives

Investor owned firms

Main types of agreements Spot market and informal agreements Predominant standards Public

Informal agreements and contracts Public

Contracts

Contracts

Contracts

Public export

Public export – strict 17

Public export – strict 52

Production share (%)

24

7

International Food and Agribusiness Management Review

517

Martins et al.

Volume 20, Issue 4, 2017

‘farrowing to finish’ system includes three stages in a farm. This system is little used but is still adopted to supply spot markets or even cooperatives.

5. Results

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0064 - Tuesday, August 22, 2017 12:32:27 PM - IP Address:24.21.169.207

5.1 A modified framework of coordination mechanisms (based on the Brazilian pork chain) Wever’s (2012) framework comprises CMs on Price, Volume, Quality and Investments and respective variables (Table 1). In this research, interviews with managers of different types of organisations in BPC and literature on GSs used in the pork sector (Boger, 2001; Martinez, 2012; Martinez and Zering, 2004; Miele and Waquil, 2007; Schulze et al., 2007), enabled us to refine set of variables and values underlying CMs. ■■ Price mechanisms As mentioned in section 2, Wever (2012) uses the variables Price Setter, Price Term and Price Criteria to explain coordination on prices. Price Setter refers to the actors that set the prices – Centralised Markets, Reference Markets, Parties to a Transaction and Internal Prices. Findings in BPC fit these values1. For example, actors normally use reference markets (e.g. prices set by a slaughterhouses association within a region) or parties to a transaction (i.e. buyers) to set prices. To distinguish this CM from the criteria that affect bonus or penalties we changed its name to Base Price Reference.

To specify values for the variable Price Term Wever (2012) uses Short Term (i.e. until 10 days), Medium/ Long term (i.e. longer than 10 days) and Indefinitely (i.e. no termination date is fixed). Findings in the BPC fit these values. Actors normally set prices for the Short Term or Indefinitely. As some buyers review the prices paid to farmers periodically (e.g. twice a year), this variable can be refined with an upper limit of six months for the value Middle Long Term. To explain Price Criteria Wever (2012) uses the values No Bonus Component, Variable Bonus and Fixed Bonus. These values look limited if compared to the array of mechanisms actors may combine to define a bonus. In the BPC actors use different aspects of productivity and quality to reward compliance. Therefore, the following criteria are included in the framework: Fixed Bonus, Bonus on Productivity, Bonus on Checklist, Production Costs Sheet, Performance Comparison, Penalty for Weight Deviation and Bonus on Carcass Quality.

The Fixed Bonus is a pre-agreed premium that a farmer receives for commitment to the agreement, regardless of his performance. The Bonus on Productivity rewards aspects such as rates of mortality and feed conversion in exchanges in which farmers use animals and feed allocated by the buyer. Performance comparison is a mechanism that compares the productivity of a farmer with a threshold defined in the agreement. This threshold may be, for instance, the performance of other buyers classified in categories (e.g. top, average and tail).

Actors use the Production Costs Sheet as a reference with which to negotiate. To set the costs and prices for pigs, parties define an expected productivity based on the technology (e.g. equipment, practices) and price of inputs farmers use (e.g. feed, electricity). Buyers use the Bonus on Checklist to reward compliance with specific requirements. These items may include issues on animal health (e.g. biosecurity facilities), food safety (e.g. silo, pipes), animal welfare (e.g. equipment, handling), environment (e.g. water treatment) and documentation. Penalty for Weight Deviation is a mechanism buyers use to incentivise farmers to deliver pigs within a weight range. Carcass Quality is a mechanism based on fat/meat percentage and the presence of injuries in the pigs.

1 In countries such as the Netherlands and Germany buyers use (spot) market prices to define base prices for finished pigs (Schulze et al., 2007; Wognum et al., 2009). However, penalties are applied, for instance, if pigs present lesions. In the USA, and Canada companies use market prices and bonuses based on carcass quality (Martinez, 2012; Martinez and Zering, 2004; Saab and Neves, 2009). PDOs and PGIs use specific reference markets to set prices (Wever, 2012). Because BPC is a special case in which chain actors use more segmented schemes to organise production, this supply chain presents diverse settings of coordination.

International Food and Agribusiness Management Review

518

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0064 - Tuesday, August 22, 2017 12:32:27 PM - IP Address:24.21.169.207

Martins et al.

Volume 20, Issue 4, 2017

The aforementioned mechanisms were used to refine the values of Price Criteria in the framework (Figure 1). First, Bonus Criteria is more appropriate because these mechanisms relate to incentives. To define the values, it was necessary to combine mechanisms in distinct groups. A criterion called Productivity includes Bonus on Productivity, Performance Comparison and Production Costs Sheet. It implies that a transaction in which one or more of these mechanisms is used meets the criterion Productivity. The criterion Pig Quality includes Bonus on Carcass Quality and Penalty for Weight Deviation. Finally, the criterion Process Quality includes items used in Bonus on Checklist. Afterwards the values were set in sequence within the market-hierarchy continuum. The first value does not include bonus or penalty. The second includes only the pre-agreed fixed bonus. In the third, Pig Quality or Productivity works as an incentive. The other values combine aspects of quality and productivity.

■■ Volume Wever (2012) uses the variables Volume Term and Amount Specification to explain coordination on Volume.

The values for Volume Term are Short Term, Medium/Long Term, and Indefinitely. These values fit periods used in BPC. For instance, to handle market fluctuations, some buyers use spot markets with deliveries valid for the Short Term. Buyers that use contracts normally set terms for Indefinitely. For Amount Specification Wever uses the values No Amount Specified, Base Volume with Allowed Deviations and Fixed Amount.

Usually, transactions in BPC fit the last two values. However, regardless of the type of transaction, the amount needs to be specified. Thus, this value is changed to Specified per Order.

■■ Quality Wever (2012) uses the values Public Actor, Third Party, Party to a Transaction, Intra Company to explain both variables of Quality Setter and Quality Monitor. However, more than one actor may set or monitor the standards. In BPC, public standards cover all transactions. Nevertheless, there are buyers that add requirements to address the demands of customers. Furthermore, a third party could add and monitor its own standards regardless of the existence of other requirements. Therefore, following the logic developed to set the values on Bonus Criteria the variables Quality Setter and Quality Monitor are refined with six distinct values: (1) Public Actor; (2) Public Actor and Third Party;(3) Public Actor, Third Party and Party to the Transaction; (4) Public Actor and Party to the Transaction; (5) Public Actor, Third Party and Internal Setting; and (6) Public Actor and Internal Setting.

■■ Resources allocation Wever (2012) uses the variables Monetary Benefits/Risks, Non-Monetary Benefits/Risks and Source of the Investment to explain coordination on Investments. These variables are related to allocation of financial capital.

However, transactions within the pork sector are reliant on allocation of resources actors use in production to meet contractual clauses (Schulze et al., 2007). In addition, understanding the allocation of resources used in production facilitates the interpretation of values of other CMs. For instance, a buyer that provides a critical resource (e.g. feed) to be used by his suppliers, may set a bonus for those suppliers that use this resource more efficiently. Therefore, the name Resources Allocation is more appropriate when designating this CM. Two variables explain Resources Allocation: Critical Resources and Buyer’s Support Resources.

Examples of Critical Resources are the feed and the animals farmers use. These resources can be allocated

No bonus

Fixed agreed bonus

Pig Quality or Productivity

Pig Quality and Process Quality

Productivity combined with Pig Quality or with Process Quality

Productivity, Pig Quality and Process Quality

Market

Internal incentives

Hierarchy

Figure 1. Bonus criteria values.

International Food and Agribusiness Management Review

519

Martins et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0064 - Tuesday, August 22, 2017 12:32:27 PM - IP Address:24.21.169.207

by the buyer or by the farmer. If the farmer allocates the resource, the buyer may require the farmer to use resources that meet specific standards. For example, to increase control of quality and productivity a buyer may allocate or recommend the standards of the feed and or genetics. Therefore, the values proposed for Critical Resources are: (1) Resources are Not Allocated Nor Approved by the Buyer; (2) Farmer uses Feed or Animals Approved by the Buyer; (3) Farmer uses Feed and Animals Approved by the Buyer; (4) Resources are Partially Allocated by The Buyer (feed or animals); (5) Resources are Totally Allocated by the Buyer; and (6) Resources are Used Internally.

Buyer’s Support Resources include technical support in production, implementation of projects and the use of information technology to support farming management. Technical Support in Production is the technical advice buyers give on production. Support in Projects is the support buyers give when a farmer sets up a new farm, makes renovations, up-scales or acquires equipment. By doing this, buyers help farmers to get credit to invest and exert more control over the standards used in projects. Support with Information technology (IT) consists of schemes in which farmers use software for farming management and exchange information with the buyer. Buyers use this information to guide farmers on how to improve their processes.

The proposed values for Buyer’s Support Resources are: (1) No Buyers Resources are Allocated; (2) Buyer Gives Technical Support in Production According to Suppliers Request; (3) Buyer Gives Regular Technical Support in Production; (4) Buyer Gives Regular Technical Support in Production and in Projects; (5) Buyer Gives Regular Technical Support in Production, in Projects and Information Exchange; and (6) Support is Used Internally.

Bonus Criteria, Critical Resources and Buyer’s Support Resources illustrate differences in how the arrangements identified in the research coordinate transactions (Table 4).

Singular and Central Cooperatives do not differentiate for characteristics of the relationship between the farmer and the buyer and are therefore in the same group. IOFs use CMs close to the hierarchy. IOFs normally allocate feed, animals and support in production, projects and information exchange. Thus, the bonus is reliant on the productivity and quality of processes. Cooperatives do not allocate feed and animals but give support in production and projects. Normally, quality (i.e. weight) of pigs affects the bonus. MIs and SMs use more market-like CMs. MIs provide technical support but it is normally less regular than in cooperatives and IOFs. SM farmers hire technical support or rely on advice given by feed companies.

Table 4. Coordination mechanisms used in the supply of piglets.

Arrangements

Variables1

Investor owned firms

BC CR SR BC CR SR BC CR SR BC CR SR

Cooperatives

Mini integrations

Spot markets

1

Market ‹––––––––––––––––––––––––––––––––––––––› Hierarchy X X X

X

X

X

X X

X

X X X

BC= Bonus criteria; CR = Critical resources; SR = Support resources.

International Food and Agribusiness Management Review

520

Martins et al.

Volume 20, Issue 4, 2017

5.2 Coordination mechanisms underlying a governance structure – a case study

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0064 - Tuesday, August 22, 2017 12:32:27 PM - IP Address:24.21.169.207

In this section we present case studies to analyse the complexity of CMs included in GSs. First, CMs underlying a GS used by a cooperative is analysed. Afterwards, we present three different cases and analyse how and why individual buyers use plural CMs (and GSs) in the same supply context. Information on production organisation and respective explanations were collected in the interviews with the managers of the firms and cooperatives.

■■ A case study on the complexity of coordination mechanisms – Singular cooperative This section presents a case study to explore the complexity of CMs included in GSs used to support the supply of piglets in the BPC. Case A is a SC, located in Rio Grande do Sul State, in the south of Brazil. The cooperative produces pork, dairy and poultry products. To produce pigs, Cases A uses contracts with 20 weaning, 60 nursery and 200 fattening farmers. The slaughters were estimated at about 290,000 heads in 2014 (SIPS, personal communication). The cooperative delivers pork products to national markets and exports that meet baseline public standards. To set base prices of piglets Case A uses weekly quotations arranged by the National Supply Company (CONAB) in the region. Furthermore, Case A sets a targeted weight for piglets (i.e.

8 kg) and establishes a penalty if the weight deviates from this value. Price mechanisms assume the values Reference Market for Base Price Reference, Short Term for Price Term, and Pig Quality for Bonus Criteria.

The contracts specify the number of sows (i.e. volume) of a farm for Indefinitely. The amount may vary due to occasional problems (e.g. mortality) or when the farm size changes, in accordance with Case A’s demands. Thus, it fits the value Fixed Amount. As Case A meets only public regulations, the Quality Setter value is a Public Actor. Case A’s technicians monitor production regularly. Thus, the variable Quality Monitor assumes the value Party to the Transaction. Farmers buy the feed produced by Case A and acquire sows with genetics that meet the standards the cooperative recommends. Therefore, the variable Critical Resources assumes the value Farmer Uses Feed and Animals Approved by the Buyer. Technicians give technical support in production and the cooperative supports farmers in procedures to get credit for new projects. In addition, farmers use software to exchange information on production with Case A. Thus, the variable Support Resources assumes the value Regular Technical Support on Production, in Projects and IT Based Management. The CMs assume different positions within the market-hierarchy continuum. These values are highlighted in bold in Table 5.

Settings of Base Price Reference, Price Term and Quality Setter assume market-like values. However, by setting a defined weight for piglets as a bonus and monitoring the processes, Case A refines the coordination of the costs of the supply of piglets and quality. Coordination on Volume is extremely hierarchical. The use of approved feed and genetics makes the allocation of Critical Resources assume an intermediary level. However, the complete set of Support Resources offered by the cooperative makes the transaction more hierarchical.

Case A is only an illustration of the complexity of CMs underlying a GS. Other combinations of values can be identified in other contexts. For example, a buyer that allocates feed and animals in the exchange may set incentives based on productivity and quality. It means that Bonus Criteria and Resources Allocation assume more hierarchical values. A Public Actor and a Party (buyer) may set the standards and a Third Party could be the monitor. Others buyers could set volumes in market-like arrangements and require farmers to use approved critical resources.

Combinations of CMs may fit types of GSs known in literature and used in pork chains. In the BPC, the most known types of GSs used to purchase pigs and piglets are partnership and selling and buying contracts (Miele and Waquil, 2007). In what follows, we present cases on plural governance.

International Food and Agribusiness Management Review

521

Martins et al.

Volume 20, Issue 4, 2017

Table 5. Coordination mechanisms in Case A.1 Coordination mechanisms

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0064 - Tuesday, August 22, 2017 12:32:27 PM - IP Address:24.21.169.207

Variables Price Base price reference

Term

Bonus criteria

Volume Term

Amount

Quality Setter

Monitor

Values Market

• Centralized market • Reference market • Party to the transaction Hierarchy • Internal price Market • Short term • Medium-Long Term Hierarchy • Indefinitely Market • No bonus • Fixed agreed bonus • Pig quality or productivity • Pig quality and process quality • Productivity combined with pig quality or process quality • Productivity, pig quality and process quality Hierarchy • Internal incentives Market

• Short term • Medium-long term Hierarchy • Indefinitely Market • Specified per order • Base volume with allowable deviations Hierarchy • Fixed amount (based on internal demand) • Public actor • Public actor and third party • Public actor, third party and party to the transaction • Public actor and party to the transaction • Public actor, third party and internal setting Hierarchy • Public actor, and internal setting Market • Public actor • Third party • Third party and party to the transaction • Party to the transaction • Third party and internal monitoring Hierarchy • Internal monitoring Market

Resources allocation Critical Market resources

• Resources are not allocated nor approved by the buyer • Farmer uses feed or animals approved by the buyer • Farmer uses feed and animals approved by the buyer • Resources are partially allocated by the buyer (feed or animals) • Resources are totally allocated by the buyer (feed and animals) Hierarchy • Feed and animals are used internally Buyer’s support Market • No buyer’s resources are allocated resources • Technical support in production according to suppliers request • Regular technical support in production • Regular technical support in production and in projects • Regular technical support in production, in projects and information exchange Hierarchy • Support is used internally

1 Bold

values: the coordination mechanisms assume different positions within the market-hierarchy continuum.

International Food and Agribusiness Management Review

522

Martins et al.

Volume 20, Issue 4, 2017

5.3 Cases on plural forms This section presents three cases to explain the use, by one buyer, of different CMs to support similar transactions (Table 6). The identification of GSs follow the nomenclature used in BPC.

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0064 - Tuesday, August 22, 2017 12:32:27 PM - IP Address:24.21.169.207

■■ Case B – Investor owned firm In the BPC, driven by cost efficiency and strict quality, IOFs rely on quasi-integration GSs to support production (Table 4). However, some firms combine these GSs with less strict mechanisms. Case B is an IOF that leads production, slaughtering, processing and exports of pork in Brazil. The firm owns branches in the main regions of production in Brazil. The slaughters range from 8 to 9 million pigs a year. To produce piglets, the firm uses about 345,000 sows in the weaning (26%) and farrowing system (74%). This case study focuses on the CMs included in contracts with farrowing farmers (the values assumed by this firm in the framework of CMs are highlighted in bold in Table 7).

In the partnership contract, the firm sets base prices based on expected productivity and costs associated with the technology used in production. This fits the value Party to the Transaction for the variable Base Price Reference. The firm reviews the base price twice a year depending on prices of inputs farmers use to produce. It implies the value ‘Medium/Long Term’ for Price Term. The firm sets a bonus based on the number of weaned piglets per sow; feed conversion and mortality, weight of the piglets and a checklist based on biosecurity, practices and documentation. These settings fit the value Productivity, Pig Quality and Process Quality for the variable Bonus Criteria.

In loan contracts, the firm uses base prices based on the criteria used in the partnership contract. However, famers produce or buy the feed used in production. Thus, fluctuations in grain prices (e.g. maize) affect base prices in the Short Term (i.e. weekly). The Bonus Criteria are reliant on the productivity of sows, penalties for weight deviation and performance comparison. This fits the value Productivity Combined with Pig Quality or Process Quality.

Buying and selling contracts set base prices based on Reference Market (e.g. in Santa Catarina, the association of slaughterhouses surveys prices used by pork processors). These prices are subject to variations within the Short Term and the bonus is based on Pig Quality (i.e. weight at 22-24 kg).

The terms for the arrangement on volumes are Indefinitely in the partnership and lending contracts. Farmers follow Case B’s production plan in both agreements. Volumes in buying and selling contracts are subject to changes in the Short Term. Plurality does not hold in coordination of quality. The whole production meets baseline standards set by the MAPA, a Public Actor, and by Case B – Party to The Transaction. The firm monitors suppliers in the three contracts. These GSs differentiate in allocation of Critical Resources. In the partnership contract, Case B allocates feed and sows. It fits the value Resources Are Totally Provided Table 6. Plural governance structures used in the case studies.

Type of arrangement Type of transaction Volume Types of governance structures and participation

1

Case B

Case C

Case D

Investor owned firm Farrowed piglets 235,000 sows Partnership: 68% Lending: 24%

Cooperative Farrowed piglets 23,000 sows Centralised production: 78% Buying and selling: 22%

Buying and selling: 8%

–

Cooperative Weaned piglets 41,000 sows Buying and selling: 28% Buying and selling with stricter CMs1: 55% Lending: 17%

CMs = coordination mechanisms.

International Food and Agribusiness Management Review

523

Martins et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0064 - Tuesday, August 22, 2017 12:32:27 PM - IP Address:24.21.169.207

Table 7. Plural forms of coordination mechanisms used to support the supply of piglets in an IOF.1,2 Variables

Values

Price Base price reference

Market

Term Bonus criteria

Volume Term Amount

Quality Setter

Monitor

• Centralised market • Reference market (BS 8%) • Party to the transaction (P 68%; L 24%) Hierarchy • Internal price • Short term (BS 8%; L 24%) Market • Medium-long term (P 68%) Hierarchy • Indefinitely Market • No bonus • Fixed agreed bonus • Pig quality or productivity (BS 8%) • Pig quality and process quality • Productivity combined with pig quality or process quality (L 24%) • Productivity, pig quality and process quality (P 68%) Hierarchy • Internal incentives Market

• Short term • Medium-long term Hierarchy • Indefinitely (P 68%; L 24%; BS 8%) Market • Specified per order (BS 8%) • Base volume with allowable deviations Hierarchy • Fixed amount (based on internal demand) (P 68%; L 24%) • Public actor • Public actor and third party • Public actor, third party and party to the transaction • Public actor and party to the transaction (P 68%; L 24%; BS 8%) • Public actor, third party and internal setting Hierarchy • Public actor, and internal setting Market • Public actor • Third party • Third party and party to the transaction • Party to the transaction (P 68%; L 24%; BS 8%) • Third party and internal monitoring Hierarchy • Internal monitoring Market

Resources allocation Critical Market resources

Buyer’s support resources

1 2

• Resources are not allocated nor approved by the buyer • Farmer uses feed or animals approved by the buyer • Farmer uses feed and animals approved by the buyer (BS 8%) • Resources are partially allocated by the buyer (feed or animals) (L 24%) • Resources are totally allocated by the buyer (feed and animals) (P 68%) Hierarchy • Feed and animals are used internally Market • No buyer’s resources are allocated • Technical support in production according to supplier’s request • Regular technical support in production (BS 8%) • Regular technical support in production and in projects • Regular technical support in production, in projects and information exchange (P 68%; L 24%) Hierarchy • Support is used internally

P = partnership; L = lending; BS = buying and selling.

Bold values: the coordination mechanisms assume different positions within the market-hierarchy continuum.

International Food and Agribusiness Management Review

524

Martins et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0064 - Tuesday, August 22, 2017 12:32:27 PM - IP Address:24.21.169.207

by The Buyer. In the lending contract, Resources Are Partially Provided by The Buyer because the firm allocates only the sows. In the buying and selling contracts, Case B does not allocate feed and sows but recommends standards on these resources. The firm provides Regular support in production, projects and IT based information exchange in the partnership and lending contracts. In the Buying and selling contract, Case B provides technical support to farmers.

In summary, this case illustrates how a buyer combines different values of CMs to support the supply of the same input in relationships with different suppliers. The reasons why the buyer uses plural CMs and GSs piglets are now presented.

The first reason revealed by the manager was the need to handle market fluctuations. In Partnership contracts, coordination in volume is not flexible enough to respond to fluctuations in the short term. Furthermore, a notice period of 6 months is required if a party wants to terminate the agreement. Thus, to have this flexibility, the firm uses less strict contracts to support a part of the supply. However, these agreements support a volume that exceeds the level aimed at by the firm.

One of Case B’s branches situated in Goiás, in the mid-west of Brazil, absorbs the whole production of farrowed piglets supported by loan contracts within the firm. In the late 1990s the main players in the pork (and poultry) sectors decided to expand their activities in that region. One important driver for this expansion was to reduce production costs. Unlike in the south, maize is produced in large-scale proprieties in the midwest. As pig production was non-existent, the firm had to use incentives for farmers to produce in that area.

Farmers were required to install large-scale farms with up-to-date technology and produce their own feed.

Moreover, the company supported farmers to obtain credit for the large investments that were made. At the time of this research, driven by its policy on food safety, traceability and efficiency, Case B was aiming to shift these contracts to the partnership model. However, to implement this change, the company faced resistance from farmers. Because farmers produced their own feed and obtained cost advantages with this, there were not willing to lose this autonomy. Furthermore, these farmers concentrate the supply in the region and maintain an association from which they receive constant technical assistance and managerial advice to support their decision making. In addition, many of these farmers run other businesses. Therefore, these conditions give farmers the bargaining power that Case A faces to negotiate contractual changes.

In another branch, located in the state of Paraná, Case B faces similar problems. Case B purchases about 30% of the volume of piglets by means of selling and buying contracts and aims to change these contracts to the partnership model. These farmers also perceive cost advantages in producing their own feed. However, in this region, the farmers are surrounded by potential buyers, who make it known they are available to set contracts fitting the current buying and selling model.

Another development, not depicted in Table 5, is the implementation of farms that meet EU regulations on animal welfare. The firm uses partnership contracts to support production that meets this standard. As these projects demand high investments, the price is set to cover the production costs, investments and ensure an interest rate. Currently, this CM covers about 15% of piglet production within the company. The firm aims to cover the whole supply with the EU standard until 2026.

■■ Case C – Singular cooperative Case C is a SC located in Paraná State, in the south of Brazil. This cooperative delivers grain, pork and dairy products. Along with two other cooperatives, Case C recently made an investment in a new slaughterhouse.

This plant absorbs the production of the three cooperatives and delivers pork with a common brand. Case C delivers 8,400 pigs per month (31% of the slaughtered volume).

The cooperative produces piglets in three central farms and by means of contracts with farmers (Table 6).

These farmers also conduct the fattening stage and are cooperative members. Recently, Case C increased International Food and Agribusiness Management Review

525

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0064 - Tuesday, August 22, 2017 12:32:27 PM - IP Address:24.21.169.207

Martins et al.

Volume 20, Issue 4, 2017

investments in the central farms to respond to the demand from the new slaughterhouse. One of the central farms (5,000 sows) is a new investment focused on the reduction of sanitary risks and EU standards on animal welfare (e.g. housed in collective crates and fed by computer-controlled systems). The cooperative aims to extend the use of these standards in the contracted production in the future. In the framework, the central production fits the value Feed and animals are used internally. The contracted production fits the value Farmers use feed and animals approved by the buyer. Furthermore, Case C provides regular support in production, supports implementation of projects and IT based information exchange. The manager pointed out two reasons for keeping these two GSs. The first is using central production to respond to the growth strategy driven by the new slaughterhouse and meeting the required quality. The second is to use the new farm, and its very process of adaptation, as a source of knowledge to facilitate the adoption of the same standards by the (contracted) farmers.

■■ Case D – Affiliated cooperative Case D is a cooperative affiliated to a central one, located in Santa Catarina State, in the south of Brazil. Case D delivers 3,753 fattened pigs per day. This volume represents 21% of slaughters of the central cooperative.

In 2014, the slaughters of the central cooperative were estimated at about 4,000,000 of pigs, representing 12.4% of Brazilian production (SIPS, personal communication). The central delivers pork that meets public regulations and stricter requirements of importers. Case D concentrates production in the mid-west of Santa Catarina. However, there are production areas in the south-east of the state and in Rio Grande do Sul. Case D arranges the supply of piglets by means of contracts with 160 farmers. Two developments illustrate the use of plural CMs in this case (Table 6).

The conventional agreement used to support the supply piglets fits the characteristics of buying and selling contracts discussed in the IOF case. This type of GS addresses the cooperative’s view, which is not to allocate Critical Resources to the production of piglets and incentivise farmers to produce. However, Case D carries out, in cooperation with an IT company, a programme designed to increase farming productivity. To access this programme, which is voluntary, farmers need to accept special conditions that, in comparison with the conventional contract, imply stricter coordination. In the framework, the allocation of Critical Resources fits the values met in the conventional contract – Farmers Use Feed and Animals Approved by the Buyer.

Regardless of their participation in the programme, all farmers need to, at least, acquire premixes from the cooperative. However, within the programme, farmers need to use only feed produced by the cooperative.

In addition, farmers need to use software to exchange information with Case D. Furthermore, the frequency of visits to support and monitor production is higher in this programme. Participation in regular technical meetings with other farmers and the board is also mandatory. The framework does not address variations in the frequency of regular support and monitoring and participation in technical meetings. However, it detects the difference in the use of IT-based information support. With respect to Support Resources, the program fits the value Buyer gives regular technical support on production, projects and information exchange. Case D does not use price incentives in this programme yet, but plans to apply these in the future. Increasing productivity is the benefit at stake. For instance, in 2014 the number of piglets weaned per sow has increased by 1.14.

The production manager explained that this programme is designed to make farmers more competitive. It also works as a channel to increase the sales of feed (maize) that the cooperative produces.

The second development that accounts for plurality, within the buying and selling agreements, is the use of a checklist. Unlike in the conventional contracts, the cooperative applies a bonus based on Process Quality in contracts with 26 farmers (9,000 sows) that produce in the south-east of Santa Catarina. The manager explained that other companies in that region use this incentive in contracts with their piglet suppliers. It forces Case D to adopt the same incentive.

International Food and Agribusiness Management Review

526

Martins et al.

Volume 20, Issue 4, 2017

6. Discussion Based on TCE theory, this exploratory research identified the main CMs used in different types of supplying arrangements in the BPC. A modified framework was elaborated to illustrate and explain, first, the complexity of CMs included in a single GSs and, second, the reasons why buyers use plural GSs and underlying CMs to support the supply.

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0064 - Tuesday, August 22, 2017 12:32:27 PM - IP Address:24.21.169.207

6.1 Complex coordination The results show that chain actors may combine CMs that assume different positions within the markethierarchy continuum. Case A, for instance, provided a detailed illustration of this complexity. The cooperative combines market-like and hybrid values for CMs of Price, Quality and Resources Allocation. Coordination on Volume, however, is subject to hierarchical coordination. These results and those found in three other cases, corroborate with Weverâ&#x20AC;&#x2122;s (2012) assumption on the use of CMs with different levels of control in one GS. The modified framework of CMs refines the model suggested by Wever in different aspects. With respect to Price mechanisms, it includes values that explain why a bonus is used. In terms of Quality, the framework addresses the fact that more than one actor may set standards. The same holds for monitoring. Finally, mechanisms on Resources Allocation include values that are related to resources applied in production. It brings the focus of analysis to what is involved in the exchange and produces interesting insights about the interaction between the CMs. For instance, the framework may show that a buyer that allocates critical resources may set price incentives so that the supplier uses these resources efficiently. Alternatively, if a buyer does not allocate critical resources, he/she may set incentives based on quality compliance. Overall, these findings are in line with TCE, meaning that actors aim to set coordination in the most efficient way (Williamson, 1991). However, the results demonstrate that each GS is made up of CMs which assume different positions in the market-hierarchy continuum. This refines the perspective that sees GSs as discretionary solutions (Raynaud et al., 2005; Schulze et al., 2007; Williamson, 1991).

The framework can be used to analyse different combinations of CMs. However, it presents some limitations.

For example, to make the framework flexible we grouped different elements in the value Productivity. It includes mechanisms such as performance comparison, mortality and feed conversion and number of piglets per sow. It implies that a GS which includes at least one of these mechanisms fits a value where Productivity has an effect in the framework. The values Pig Quality and Process Quality have similar characteristics. In addition, Case D showed that a buyer may refine the coordination of a transaction by increasing the frequency of inspections or asking suppliers to attend technical meetings. These are examples of elements that can be refined or included in the framework according to the interest of managers or scholars. Furthermore, the framework, along with the literature, supported the analyses of plural forms of governance.

6.2 Plural forms Literature has explored the phenomenon of plurality by analysing dual internal-contracted production (Heide, 2003; Parmigiani, 2007). This paper shows that in BPC, overall, actors allocate resources (e.g. technical advice, feed, animals) to support hybrids GSs. With respect to the explanations proposed by MĂŠnard (2013), the results do not correspond to monitoring difficulties. All cases show that buyers hold the expertise that is necessary to support and monitor suppliers in terms of efficiency and quality (Miranda and Chaddad, 2014).

Case D implemented a productivity programme that illustrates the allocation of internal resources (expertise) to improve the performance of suppliers and explains the coexistence of different CMs. In Case C, one of the reasons for the implementation of a central farm, meeting EU standards, is to produce knowledge for both the cooperative board and farmers. The results obtained in these cases (Table 8), corroborates the perspective that internal and contracted production improve their capabilities and performance (Heide, 2003; Mols et al., 2012; Parmigiani, 2007).

International Food and Agribusiness Management Review

527

Martins et al.

Volume 20, Issue 4, 2017

Table 8. Drivers of plural forms of governance in the case studies.

Explanations

Cases

Governance structures used to complement the supply

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0064 - Tuesday, August 22, 2017 12:32:27 PM - IP Address:24.21.169.207

Market fluctuations B Absence of alternative suppliers B Incentives offered by competitors B, D (other buyers)

Investments to meet stricter quality requirements Context of exchange

The buyer uses a contract with coordination mechanisms that are less strict than usual.

The firm (B) uses a contract with coordination mechanisms that are less strict than usual.

The cooperative (D) uses a contract with coordination mechanisms that are stricter than usual.

B, D, C The firm (B) and cooperative (D) use contracts with stricter control and incentives.

The cooperative (C) produces in a central farm.

D Cooperative (D) sets quality programme with stricter coordination mechanisms with voluntary adherence.

Another finding that fits the perspective of increased coordination as a driver of plural governance is the use of incentives for investments that meet specific quality standards in Case B (EU standards). We lack clear definitions as to the extent to which a quality requirement shifts the type of transaction at stake. However, if we maintain that plurality holds in the perspective of similar transactions, this incentive corroborates the view on the improvement of processes (Parmigiani, 2007). This mechanism itself fits the principle of the alignment with transaction attributes (Williamson, 1991).

The results obtained in this paper do not show ambiguity regarding returns on specific assets as MĂŠnard (2013) explains. All buyers were shown to have clear views on the forms of coordination that best fit their demands.

Furthermore, these actors hold the necessary expertise to produce efficiently and to meet the desired quality.

The IOFs, for instance, focus on strict mechanisms, by allocating critical and support resources and setting incentive mechanisms to ensure the supply driven by cost efficiency and quality. Cooperatives, however, pursue performance by handling organisational constraints. For example, as farmers are also owners of the business, cooperatives find it more difficult to enforce contractual sanctions. Thus, compared to IOFs, the settlement of strict coordination in cooperatives demands more dialogue and complex decision-making.

The cases C and D demonstrated how characteristics found in specific transaction contexts (i.e. cooperatives) explain plural governance. First, the productivity programme is also a channel to which the coop markets the feed. Second, the cooperative has a clear view in not allocating critical resources in the transactions but aims to incentivise farmers to produce more efficiently. Third, the voluntary nature of the productivity programme addresses the fact that the board cannot oblige farmers to undergo strict coordination. Therefore, the need to conduct the changes in a gradual fashion illustrates how the exchange context may trigger plural governance. Case D also supports this view. The project of a central farm, working as a source of knowledge that farmers will use in their farms, is a joint decision between the board and the farmers.

Findings in Cases B and C corroborate the strategizing view of MĂŠnard (2013). These cases illustrate the difficulties of a buyer to coordinate a transaction as desired, which pushes coordination towards plural forms.

In Case B, the firm has a clear view on using less strict arrangements (i.e. buying and selling contract) to support some of the supply with more flexibility to handle market fluctuations. However, the volume of production that is supported by these (looser) mechanisms exceeds the level targeted by the company. In contrast, Case D needs to use, with a group of suppliers, a strict incentive mechanism that is not usual in its coordination policy. In both cases, farmers perceive sustained advantages in keeping the CMs in the current fashion. First, other buyers are available to keep the mechanisms that are currently used by some of the suppliers of Case B and Case D. Second, farmers (Case B) are organised in an association, have other businesses and hold the whole volume the firm purchases. It gives them bargaining power to negotiate contractual changes with the buyer.

International Food and Agribusiness Management Review

528

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0064 - Tuesday, August 22, 2017 12:32:27 PM - IP Address:24.21.169.207

Martins et al.

Volume 20, Issue 4, 2017

In summary, this research identified four groups of explanations for the use of plural forms: the need to handle market fluctuations, bargaining power of suppliers, the organisational context of exchange, and the need to implement stricter quality standards. Regarding Ménard’s (2013) assumptions, the results fit only the strategizing view. Handling market fluctuations is seen by the IOF as a reason to support some of the supply using less strict coordination (i.e. buying and selling contracts). Bargaining power, on the other hand, explains why the firm faces difficulties in enhancing the volume of supply supported by stricter coordination (i.e. partnership contracts). In line with this, in Case B and Case D, incentives offered by potential buyers suggest that competition triggers plural governance. Hence, these support the strategizing view in two dimensions. First, they illustrate the difficulties an actor may face in coordinating a transaction as desired.

Second, they provide examples of (plural) CMs used to overcome such difficulties.

In this paper, the analysis of plural coordination is limited by its exploratory nature and by few case studies.

However, the proposed framework includes CMs identified by means of interviews with companies that hold the lion’s share of pig production in Brazil. It makes our results representative. The application of the framework should be extended to other food chains. It could bring further insights into its validity and refinements to fit specific contexts.

7. Conclusions This research has corroborated assumptions on the complexity of CMs that underlie a GS (Wever, 2012) and the use of plural GSs (Bradach and Eccles, 1989; Heide, 2003; Ménard, 2013). First, this paper identified the main CMs used to support the supply of piglets in the BPC. Second, the framework of CMs provides more precise definitions about what is coordinated by a GS than so far provided in the literature (Gellynck and Mólnar, 2009; Raynaud, 2005; Schulze et al., 2007).

The framework of CMs elaborated in this paper supports a comprehensive analysis of plural forms of governance used in supply relationships in the BPC. Predominantly, actors rely on different hybrid arrangements in which they allocate critical and/or support resources, set price incentives so that suppliers use these resources efficiently and/or meet quality requirements.

With respect to the framework Ménard (2013) proposes to analyse plural GSs, the results obtained in this paper do not fit the assumptions on monitoring complexity and ambiguity. In the BPC, buyers hold the capabilities and resources that are necessary to coordinate transactions (Miranda and Chaddad, 2014) and have a clear view on how to seize returns from these relationships. Furthermore, the results show that, to improve coordination and quality, actors (cooperatives) use in combination with basic GSs, stricter CMs that feed the exchange relationship with better knowledge, efficiency and quality compliance (Heide, 2003; Mols et al, 2012; Parmigiani, 2007).

This paper offers interesting insights into the assumption of strategizing proposed by Ménard. Regardless of the (possible) endogenous causality among the aspects we found – bargaining power of suppliers and the incentives given by competitors – these factors contribute to the explanation of why plural forms are used.

The studies in the cooperatives show that the organisational context in which the transactions are embedded may affect coordination and result in plural governance. This variable could be tested in further analysis on production organisation in cooperatives.

Finally, TCE theory was useful for supporting the elaboration of the framework used to analyse the main CMs and GSs that chain actors use to organise transactions in the BPC. However, in line with the literature on plural governance, the results show that TCE does not offer sufficient explanations. Combining organisational (capabilities, competences) and neoclassical theories (competition), may offer a more comprehensive approach to addressing the phenomena of plurality. This paper has contributed with additional explanations to be examined in these fields.

International Food and Agribusiness Management Review

529

Martins et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0064 - Tuesday, August 22, 2017 12:32:27 PM - IP Address:24.21.169.207

References Becker, T. and A. Staus. 2009. European food quality policy: the importance of geographical indications, organic certification and food quality insurance schemes in European countries. Available at: http:// tinyurl.com/krgdkjs.

Boger, S. 2001. Quality and contractual choice: a transaction cost approach to the Polish hog market. European Review of Agricultural Economics 28: 241-261.

Bonneau, M., E. Antoine-Ilari, C. Phatsara, D. Brinkmann, M. Hviid, M.G. Christiansen, E. Fàbrega, P. Rodríguez, L. Rydhmer, I. Enting, K. De Greef, H. Edge, J.-Y. Dourmad and S. Edwards. 2011.

Diversity of pig production systems at farm level in Europe. Journal on Chain and Network Science 11: 115-135.

Bradach, J. 1997. Using the plural form in the management of restaurant chains. Administrative Science Quarterly 42: 276-303.

Bradach, J.L. and R.G. Eccles. 1989. Price, authority, and trust: from ideal types to plural forms, Annual Review of Sociology 15: 97-118.

Brazilian Association of Pig Farmers (ABCS). 2015. Panorama setorial [sector panorama]. Available at: http://tinyurl.com/k6lq42w.

Coase, R.H. 1937. The nature of the firm. Economica 4: 386-405.

Foss, N., 2002. Introduction: new organizational forms – Critical perspectives. International Journal of the Economics of Business 9: 1-8.

Gellynck, X. and A. Molnar. 2009. Chain governance structures: the European traditional food sector. British Food Journal 111: 762-775.

Ghosh, M. and G. John. 1999. Governance value analysis and marketing strategy. Journal of Marketing 63: 131-145.

Grandori, A. 1997. Governance structures, coordination mechanisms and cognitive models. The Journal of Management and Governance 1: 29-47.

Grunert, K.G., W. Verbeke, J.O. Kugler, F. Saeed and J. Scholderer. 2011. Use of consumer insight in the new product development process in the meat sector. Meat Science 89: 251-258.

Heide, J.B. 2003. Plural governance in industrial purchasing. Journal of Marketing 67: 18-29.

Hendrikse, G. and T. Jiang. 2011. An incomplete contracting model of dual distribution in franchising.

Journal of Retailing 87: 332-344.

Heyder, M., T. Hollmann-Hespos and L. Theuvsen. 2010. Agribusiness firm reactions to regulations: the case of investments in traceability systems. International Journal on Food System Dynamics 2: 133-142.

Klein, B. 1996. Why hold-ups occur: the self-enforcing range of contractual relationships. Economic Inquiry 34: 444-463.

MAPA. 2016. Brazilian Ministry of Agriculture, Livestock and Food Supply. Available at http://indicadores. agricultura.gov.br/agrostat/index.htm.

Martinez, S.W. 2012. Pork quality and the role of marketing contracts: a case study of the US pork industry.

British Food Journal 114: 302-317.

Martinez, S.W. and K. Zering. 2004. Pork quality and the role of market organization. USDA – Agricultural Economic Report 835. Available at: http://tinyurl.com/kt54xps.

Ménard, C. 2004. The economics of hybrid organizations. Journal of Institutional and Theoretical Economics 160: 345-376.

Ménard, C. 2013. Plural forms of governance: where do we stand? Managerial and Decision Economics 34: 124-139.

Ménard, C. and E. Valceschini. 2005. New institutions for governing the agri-food industry. European review of Agricultural Economics 32: 421-440.

Miele, M. and P.D. Waquil. 2007. Estrutura e dinâmica dos contratos na suinocultura de Santa Catarina: um estudo de casos múltiplos. Estudos Econômicos, 37: 817-847.

Miranda, B.V. and F.R. Chaddad. 2014. Explaining organizational diversity in emerging industries: the role of capabilities. Journal on Chain and Network Science 14: 171-188.

International Food and Agribusiness Management Review

530

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0064 - Tuesday, August 22, 2017 12:32:27 PM - IP Address:24.21.169.207

Martins et al.

Volume 20, Issue 4, 2017

Mols, N.P., J.R. Hansens, and A.R. Villadsen. 2012. Plural governance: the effect of internal production on supplier performance. Industrial Marketing Management 41: 874-885.

Parmigiani, A. 2007. Why do firms both make and buy? An investigation of concurrent sourcing. Strategic Management Journal 28: 285-311.

Raynaud, E., L. Sauvee and E. Valceschini. 2005. Alignment between quality enforcement devices and governance structures in the agro-food vertical chains. Journal of Management and Governance 9: 47-77.

Rindfleisch, A. and J.B. Heide. 1997. Transaction cost analysis: past, present, and future applications. Journal of Marketing 61: 30-54.

Saab, M.S.M. and M.F Neves. Pork chains in Brazil and Canada: a comparison. In: European pork chains – Diversity and quality challenges in consumer-oriented production and distribution, edited by J. Trienekens, B. Petersen, N. Wognum; and D. Brinkmann, Wageningen Academic Publishers, Wageningen, the Netherlands.

Sauveé, L. 2013. Hybrid governance: sketching discrete alternatives. Journal on Chain and Network Science 13: 1-9 Schulze, B., A. Spiller and L. Theuvsen. 2007. A broader view on vertical coordination: lessons from German pork production. Journal on Chain and Network Science 7: 35-53.

Trienekens, J., B. Petersen, N. Wognum and D. Brinkmann (eds.). 2009. European pork chains – Diversity and quality challenges in consumer-oriented production and distribution, Wageningen Academic Publishers, Wageningen, the Netherlands.

United States Department of Agriculture (USDA). 2016. Foreign agricultural service. Available at http:// apps.fas.usda.gov/psdonline/psdHome.aspx.

Verbeke, W., F.J.A. Pérez-Cueto, M.D. De Barcellos, A. Krystallis and K.G. Grunert, 2010. European citizen and consumer attitudes and preferences regarding beef and pork. Meat Science 84: 284-292.

Wever, M. 2012. Chain-wide consequences of transaction risks and their contractual solutions: managing interdependencies in differentiated agri-food supply chains. PhD Thesis, Wageningen University, Wageningen, the Netherlands. Available at: http://tinyurl.com/kav9huq.

Wever, M., N. Wognum, J. Trienekens and O. Omta. 2010. Alignment between chain quality management and chain governance in EU pork supply chains: a transaction-cost-economics perspective. Meat Science 84: 228-237.

Williamson, O.E. 1991. Comparative economic-organization – the analysis of discrete structural alternatives.

Administrative Science Quarterly 36: 269-296.

Williamson, O.E. 2008. Outsourcing: transaction cost economics and supply chain management. Journal of Supply Chain Management 44: 5-16.

Wognum, N., J. Trienekens, M. Wever, J. Vlajic, J. Van der Vorst, O. Omta, J. Hermansen and T.L.T. Nguyen, 2009. Organisation, logistics and environmental issues in the European pork chain. In: European pork chains – Diversity and quality challenges in consumer-oriented production and distribution, edted

by J. Trienekens, B. Petersen, N. Wognum; and D. Brinkmann. Wageningen Academic Publishers, Wageningen, the Netherlands.

Zylbersztajn, D. and E.M.M.Q, Farina. 1999. Strictly coordinated food-systems: exploring the limits of the co-Asian firm. International Food and Agribusiness Management Review 2: 249-265.

International Food and Agribusiness Management Review

531

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0064 - Tuesday, August 22, 2017 12:32:27 PM - IP Address:24.21.169.207

OPEN ACCESS International Food and Agribusiness Management Review Volume 20 Issue 4, 2017; DOI: 10.22434/IFAMR2015.0174

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0174 - Tuesday, August 22, 2017 12:39:39 PM - IP Address:24.21.169.207

Received: 30 August 2015 / Accepted: 7 April 2017

Shaping food systems towards improved nutrition: a case study on Tuscan Bread Protected Designation of Origin CASE STUDY Francesca Galli a, Francesca Venturib, Fabio Bartolinic, Oriana Gavad, Angela Zinnaie, Sanmartin Chiarab, Gianpaolo Andrichf, and Gianluca Brunorig aResearch

Fellow, cAssociate Professor, dPhD student, and gFull Professor, Department of Agriculture, Food and Environment, University of Pisa, Via del Borghetto 80, 56124 Pisa, Italy bResearcher, eAssociate

Professor, and fFull Professor, Department of Agriculture, Food and Environment, Interdepartmental Research Center Nutrafood-Nutraceuticals and Food for Health, University of Pisa, Via del Borghetto 80, 56124 Pisa, Italy

Abstract The concern for the quality of food, its composition and contribution towards nutrition and health is widespread among public and private food system actors. The increasing interest in locally integrated supply chains leads to reconsider the configuration of the food system in relation to sustainability and health outcomes.

This article focuses on the relationship between processing practices and nutritional value in the wheat-tobread sector, illustrated by a case study on the Tuscan Bread Protected Designation of Origin (PDO). By adopting a food system perspective, the case study shows how the different actors have mobilized to respond to multiple drivers of change. A mixed research method approach is adopted to illustrate the relationship between processing practices and nutritional value outcomes: practice-based indicators for each step of the chain are complemented with performance-based indicators of the chemical, physical and sensorial profile of Tuscan Bread PDO. Furthermore, the implications on food system governance of a differentiation strategy based on territorial origin and enhanced nutrition are discussed.

Keywords: food system, wheat-to-bread chain, nutrition, territoriality, governance JEL code: Q1, Q13, Q18, Q02 Corresponding author: francesca.galli@for.unipi.it

Š 2017 Galli et al.

533

Galli et al.

Volume 20, Issue 4, 2017

1. Introduction

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0174 - Tuesday, August 22, 2017 12:39:39 PM - IP Address:24.21.169.207

A concern for nutritional qualities of food, composition and contribution towards physical health and wellbeing, is widely spread among academics, policy, health professionals and practitioners. The increase of non-communicable diseases – including obesity, type II diabetes and a range of cardiovascular diseases – is recognized as being strongly related to diets of poor nutritional quality. Diseases related to unbalanced nutrition and, more broadly, unsustainable diets, raise concerns also in relation to the impacts on public healthcare costs (Popkin, 2006; Stuckler and Nestle, 2017; Wang et al., 2011).

Improving nutrition and enhancing sustainability is the core of several Sustainable Development Goals (SDGs), to which United Nations have recently committed. Specifically, SDG 2 calls for ‘ending hunger, achieve food security and improve nutrition, and promote sustainable agriculture’. The global transition towards more nutritionally balanced and higher quality diets is a basic requirement for food systems’ sustainability (see Auestad and Fulgoni, 2105 for a review). The concerns on the healthiness of diets and nutritional value of food has driven food producers to adapt processing technologies to compete over food’s health attributes (Nestle, 2013) although it has been noted that this does not necessarily result in improved healthiness of marketed products (Golan and Unnevehr, 2008). Nutritional information has also become a terrain of competition on the market (Mancino et al., 2008; Verbeke, 2008). Authors indicate that investments and actions on the production side should be coupled with parallel investments on the demand side to increase the requests for nutrient-dense food (Jones and Ejeta, 2016).

A productive, diverse, ecologically and socially sustainable agricultural system has been recognized as crucial for shaping healthy diets and improving human nutrition (Jones and Ejeta, 2016). Agriculture plays an essential role in providing a diversity of nutrient dense foods to meet dietary recommendations for health. The agricultural sector is increasingly considered in relation to the wider food system’s upstream and downstream activities, such as input supply, transformation, packaging and storage, logistics and retail (Venturi et al., 2016a). There is an increasing need for integration between the various stages of the supply chain, both to encounter the demand side pressure for healthier food and to reduce the risks on raw material supply, associated with different shocks and stresses (e.g. prices, quantities, sanitary requirements). A food system approach to food and nutrition security has contributed to broaden the view on the links among actors, resources and on capturing food systems’ multiple outcomes (Ericksen, 2008; Ingram, 2011). The objectives of different categories of food systems’ stakeholders may be divergent in relation to food and nutrition security and sustainability outcomes (Galli et al., 2016). It occurs that agricultural systems have achieved increased productivity to the detriment of the safeguard food nutritional value and sustainability (Mozaffarian and Ludwig, 2010).

Recently, the spread of locally integrated – and shorter – supply chains has raised public attention on the origin of the raw material, as well as on the ingredients and processing methods in relation to food’s nutritional value (Aprile et al., 2016; Brunori et al., 2016). Moreover, longer transportation and storage times are perceived as negatively affecting the nutritional content of food (Caputo et al., 2013). Despite the perception that local, or shorter chains, have a supposed ability to benefit consumers with healthier foods – compared to mainstream food chains – the relationship between the configuration of the food system and the contribution to health is still widely debated, especially in socio-economic analysis (Bogomolova et al., in press; Brunori and Galli, 2016).

This paper addresses the relationship between food system and nutritional outcomes by focusing on the following research questions: how can the supply chain mobilize to pursue improved nutritional value, compared to conventional processes? And, what are the implications for food system actors?

To address such wide questions, we adopt an inter-disciplinary approach to understand the interplays among changes in product quality, nutritional value, agri-food chain and rural development. A case study on the wheat-to-bread sector illustrates the method adopted.

International Food and Agribusiness Management Review

534

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0174 - Tuesday, August 22, 2017 12:39:39 PM - IP Address:24.21.169.207

Galli et al.

Volume 20, Issue 4, 2017

Bread holds a central role in diets worldwide, wheat is traded globally and bread manufacturing often develops locally. Furthermore, bread and cereal products are at the center of a contrasted debate on sustainability and on nutritional value more specifically (see Galli et al., 2015 for a review). Cereals are staple foods, that provide a major source of carbohydrates, proteins, B-vitamins and minerals1. A growing body of evidence suggests that regular consumption of cereals may have a role in the prevention of chronic diseases2. At the same time, cereal processed products contribute to a considerable proportion of sodium intake; therefore, manufacturers are encouraged to reduce the sodium content of foods, such as breakfast cereals and breads (McKevith, 2004). The mechanisms by which cereals convey beneficial effects on health are multifactorial and are related to, among other indicators, the micronutrient and fiber content and their glycemic index.

There is mixed evidence in the literature on the relationship between the supply chain practices and the nutritional value of cereal bread. During all steps of bread making, complex biochemical and physical transformations occur, affect and are affected by the flour constituents and different substances are used in processing to optimize their functionality (Venturi et al., 2013a, 2016b). The quality of component cereals, storage, processing and conservation impact on the nutritional value of bread (Dewettinck et al., 2008). As with other food processing, the challenge in fermenting cereals lies in the ability to combine good sensory quality with demonstrated nutritional and health benefits (Jones et al., 2015; McKevith, 2004; Venturi et al., 2013b). Consumer’s perception of bread quality is, to some extent, determined by sensory and health attributes and the perception about nutrition and health can be influenced by written information (Dewettinck et al., 2008; Hellyer et al., 2012; Mancino et al., 2008; Pohjanheimo et al., 2010). The wheat-to-bread sector is rather opaque as grains provenance is often unknown, whereas traceability and communication to the consumer imply a shift towards higher quality and identity preservation of wheat (i.e. varietal traceability) (Barling et al., 2009). Traceability systems that track physical entities along the chain are increasingly used to meet a range of regulatory and commercial objectives, including growing ethical concerns on the content and nutritional value of ‘mass-produced’ bread and specific health risks for humans (e.g. the danger from mycotoxins, fungal infections of grain, etc.). Other studies reinforce the recognition among stakeholders about a public anxiety on the healthiness of bread (Sharpe et al., 2008; Jones et al., 2015)3.

This contribution draws from such contested debate to address how practices along the wheat to bread chain shape nutritional value of bread, by illustrating a case study. The Tuscan Bread Protected Designation of Origin4 (PDO) supply chain shows how integration, localization and the enhancement of nutritional value can be prioritized as a basis for differentiation on the market.

The background in Section 2 introduces the food system concept which highlights the opportunities and implications for improving or worsening food and nutrition security. Materials and methods (Section 3) 1

Nutrient composition of bread cereals is 50-80% carbohydrate, proteins (8-12%), lipids (1.5-7.0%) and micronutrients, and a range of phytochemicals beneficial to health (McKevith, 2004). The composition of the dry matter of wheat varies depending on soil, climate and genetic variations between wheat types. Components of the wheat grain include bran and germ. Bran, the outer coating or ‘shell’, is rich in B-vitamins and minerals. Cereals are rich in phosphorus, calcium, magnesium, potassium, zinc and copper, while the level of sodium – before processing – is relatively low. Vitamins such as thiamine (B1), riboflavin (B2), pantothenic acid, inosotol, P-aminobenzoic acid, folic acid and vitamin B6 are also distributed throughout the wheat grain.

2 Epidemiological studies indicate a strong correlation between the consumption of whole grain cereals and the decrease of diseases like cardiovascular disease, type-2 diabetes, obesity, metabolic syndrome as well as some types of cancer (Larsson et al., 2005; Mellen et al., 2008). This evidence encouraged the Food and Drug Administration to approve the first health claim for whole grains in 1997 and since then, countries and organizations around the globe are increasingly including them in their dietary recommendations.

3 A recent social media current – particularly strong in the USA – poses grain staples, especially wheat, under attack in popular press books (e.g. protein intensive Paleo Diets, in the book by Cordain (2010)) and drives to reduce the consumption of gluten by eliminating wheat, grains and in certain cases carbohydrates. Beyond the concern on overweight and obesity related to an excessive amount of cereals in the diet, the other widespread issue is related to the allergenic potential of wheat. It should be recalled that the spread of extreme conceptions of ‘healthy food’ fail to recognize the importance of calories intake and balance within the diet and encourage the – sometimes unnecessary – exclusion of whole food groups without appropriate support and professional advice (Jones, 2015).

4 Designations of origin are names that identify a product from a given territory, testifying a link between a quality, reputation or characteristic of the product and its geographical origin. Protected Denominations of Origin entails the name of a region, a specific place or, in exceptional cases, a country, used to describe an agricultural product or a foodstuff: (1) originating in that region, specific place or country; (2) possessing quality or characteristics which are essentially or exclusively due to a particular geographical environment with its inherent natural and human factors; and (3) the production, processing and preparation of which take place in the defined geographical area.

International Food and Agribusiness Management Review

535

Galli et al.

Volume 20, Issue 4, 2017

indicates how the case study analysis was developed. The cast study described in Section 4 articulates three sub-sections: the first presents the drivers of change, the second gives an overview of the activities developed at different levels of the food system, and the third focuses on the performance aspects. A discussion follows in Section 5, to highlight the implications for actors and the interplays between changes in product quality, nutritional value, agri-food chain practices and rural development.

Agriculture and food studies have progressively shifted towards systemic approaches, which allow to widen the scope from supply chains linear links to the complex relations among actors and resources, including food systems’ main outcomes (Cordell et al., 2009; Ericksen, 2008; Garnett, 2013; Hammond and Dubé, 2012; Ingram, 2011). A food system approach allows to identify opportunities and implications for improving or worsening food security including nutrition. Moreover, it offers a lens to gain a more comprehensive understanding of the factors of change impacting upon the food system (within and beyond the food chain), the multiplicity of actor that may contribute to achieving outcomes, and to increase awareness on the dynamics and the effects of actions, for the anticipation of possible unexpected consequences.

Figure 1 presents the key elements and activities of the food system, based on Ingram, (2011) and Ericksen, (2008).

Food systems comprise natural and human made assets (i.e. environment, infrastructures, etc.), institutions (i.e. cognitive, normative and regulative) and activities (from input provision to consumption and disposal), that deliver outcomes (i.e. food and nutrition security, health, socio-economic growth and equity and environmental sustainability) (HLPE, 2014; UNEP, 2016).

The bottom part of Figure 1 indicates that there is a bi-directional relationship between food system’s activities and outcomes. On one side, food systems influence consumers’ choices, collective dietary patterns and this has implications for health: for example, food processing, storage and logistics determine food safety conditions, and this increases or decreases exposure to illness. The way food system activities affect the

External factors (globalisation of markets, political instability, climate) Natural assets (land, biodiversity, natural resources)

Human assets (traditions, technologies, know-how) Food demand

Food supply Inputs (seeds, wheat)

Consumption (food environment)

Storage Processing (milling, baking) Distribution, Logistics

Managing losses and waste

Cognitive, normative, regulative institutions

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0174 - Tuesday, August 22, 2017 12:39:39 PM - IP Address:24.21.169.207

2. Theoretical background: a food system approach

Outcomes Food and nutrition security, environmental, socio-economic outcomes

Figure 1. Food systems’ elements (adapted from Ingram (2011) and Ericksen (2008)).

International Food and Agribusiness Management Review

536

Galli et al.

Volume 20, Issue 4, 2017

nutritional quality of foods, their affordability, accessibility and acceptability within the ‘food environment’, is the focus of increasing research (Downs and Fanzo, 2016; Hawkes et al., 2012; Waterlander et al., in press).

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0174 - Tuesday, August 22, 2017 12:39:39 PM - IP Address:24.21.169.207

On the other side, nutrition and health outcomes, (that depend on dietary choices and consumer purchases), may drive the way food production, processing, storage, trade and retailing are developed.

The ‘food environment’ concept is central in this regard: it encompasses the physical, economic, political and socio-cultural surroundings, the opportunities and conditions that influence food choices and nutritional status, mediated through food preferences and knowledge (Swinburn et al., 2014). Information affects consumers’ knowledge, attitudes, food choices and, ultimately, dietary behavior. Food choice can be influenced by a wide range of factors, among which health nutritional labeling, sensory features, ethical concerns and affordability. The actual drivers of purchase behavior and their effects on dietary patterns are still widely debated (Campos et al., 2011; Cowburn and Stockley, 2005; Grunert and Wills, 2007).

The central section of Figure 1 shows the supply chains’ activities that shape how foods are produced, processed, distributed and marketed, consumed and disposed of. The way food is processed along the chain can affect nutrition and diets both positively and negatively by creating both ‘entry’ and ‘exit’ points for nutrition (Hawkes et al., 2012). Food chains processes can help preserve existing nutritional value (e.g. safe handling conditions of food) or can contribute to improving the nutritional value being by introducing micronutrients (e.g. through fortification) or can lead to decreasing nutrients, by removal from the original raw material (e.g. grain germ, which contains several nutrients, is generally removed from white flour as it is highly perishable) or by adding substances partly associated with diet-related non-communicable diseases (e.g. sodium or preservatives) (Augustin et al., 2016; Mozaffarian et al., 2016). Value chains represent a potential way to leverage agriculture to improve nutrition, particularly regarding traditional value chains for micronutrient rich foods (Gelli et al., 2015).

Normative and regulative institutions impact on food systems (right hand side of Figure 1) and play a key role in balancing supply and demand for nutritious food. Institutions have the responsibility to address the several factors of change (top of Figure 1) that challenge contemporary food systems, including the survival of the agricultural sector. Agriculture, which contributes critically to the nutritional value of food, is increasingly facing land abandonment and urbanization trends, environmental degradation, retailing markets concentration, price transmission along the chains and, not least, policy developments (including agricultural policy) (FAO, 2017).

Public involvement in nutrition outcomes can be broadly distinguished in three categories: (1) interventions to increase food supply of nutritious food, (particularly where availability is lacking); (2) interventions to change food demand (particularly where nutritious food is available, but is not consumed); and (3) interventions to enhance value chain organization and performance, when both demand and supply for nutritious food exist.

In the last case, public action may focus on optimizing the nutrient flow along the value chain (Allen et al., 2016). This leads to considering the role of policy in supporting valorization strategies developed by private food chain actors, through different instruments and incentives. Private actors in food value chains range from vertically integrated multinational corporations to farmers, individual entrepreneurs that transport, store, aggregate or sell food, whose incentives and goals should be aligned. Because goals for nutrition, sustainability, and economic development will not always be complementary, interventions will need to manage trade-offs and constraints to meet the multiple goals.

3. Materials and methods The research was designed as a case study. According to Yin (2003: 2) ‘the distinctive need for case studies arises out of the desire to understand complex social phenomena’ such as organizational processes, for example. Case studies are indicated as a preferred strategy when ‘how’ or ‘why’ questions are posed, when the investigator has little control over events, and when the focus is on a contemporary phenomenon within International Food and Agribusiness Management Review

537

Galli et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0174 - Tuesday, August 22, 2017 12:39:39 PM - IP Address:24.21.169.207

a real-life context (Yin, 1981:59, 2003: 2). The Tuscan Bread PDO case is rich in information and data is readily available from multiple sources. For this reason, the research questions are addressed by integrating a qualitative phase and quantitative evidence (Eisenhardt, 1989: 534-535) to develop a mixed research, conducted by inter-disciplinary teams, in which multiple skills are exploited (Paluck, 2010; Starr, 2004).

The first part of the work consists in analyzing the challenges affecting the wheat-to-bread chain (i.e. factors of change), framing the context in which the strategies undertaken by food system actors, take place. The analysis integrates desk based data and grey literature, participation to public events, a set of in depth interviews to key informants and a focus group. The key informants addressed were food system stakeholders covering, but not limited to, all stages of the supply chain: the agricultural consortium responsible for collecting wheat, a wheat genetics’ professor that contributed to developing the PDO specification, the milling company, three bakers, the director of the PDO consortium, a quality manager of a large retailing company, a regional policy maker responsible for managing the ‘integrated supply chain’ project and providing funding based on Rural Development Program. These have been interviewed repeatedly, between September 2014 and December 2016. A focus group in March 2015 was organized to discuss the performance of the Tuscan Bread PDO chain, compared to other more local and global bread chains (Galli et al., 2016).

The second part of the work focuses on the linkage between food system activities and nutritional outcomes, assessed by developing performance based indicators and practice based indicators. As defined in the Sustainability Assessment for Food and Agriculture ‘performance based indicators focus on the results of compliance with an objective and measure the performance of an operation, identify trends and communicate results; practice based indicators focus on prescribing the necessary tools and systems required for best practices to be in place. The latter indicators are process rather than outcome oriented’ (FAO, 2013:48).

3.1 Practice based indicators For each stage of the supply chain a set of practices that impact on the nutritional value of wheat, flour and bread are identified and discussed. The Tuscan Bread PDO practices are compared and benchmarked in relation to a comparable conventional industrial bread. The relationships between the phases of the bread production chain and the nutritional value were investigated based on academic literature analysis, grey literature and in-depth interviews with stakeholders and experts.

3.2 Performance based indicators The characterization of nutritional parameters for Tuscan Bread PDO is developed through chemical and sensorial analysis. As done for practice based indicators, the empirical data from chemical and sensory analysis of the Tuscan Bread PDO is compared to a generic white bread (Supplementary Materials S1).

The chemical analysis entails: (1) in laboratory sourdough bread production by using flour according to the PDO specification; (2) chemical characterization of sourdough and conventional bread, by calculating concentrations of the main fermentative metabolites produced in the sourdough during the storage time and in the bread samples, after cooking; (3) sensorial analysis of bread (crust and crumb). For details see Zinnai et al. (2013) for industrial white bread and Venturi et al. (2013a) for Sourdough Tuscan Bread; and (4) Statistical analysis, to evaluate the statistical significance of the data obtained.

4. Case study: Tuscan Bread Protected Designation of Origin 4.1 Context: the challenges and the drivers of change Despite the cereals market being complex and highly globalized, due to the supply variations in the main production areas worldwide, overall production yields have increased over the past decades in many countries of the world, including Italy (FAO Stat, 2015). The steady increase in productivity has guaranteed production being higher than consumption, driving grain prices steadily downward. At the global level supply is greater International Food and Agribusiness Management Review

538

Galli et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0174 - Tuesday, August 22, 2017 12:39:39 PM - IP Address:24.21.169.207

than demand, while this does not hold for Italy, where national production is insufficient for processing needs. Italy is among the largest importers from Canada, France, East Europe, USA. The Tuscan Bread PDO case study develops in Tuscany, whose regional (and national) cereal market is affected by economic, social and environmental crisis. The cereal sector in Tuscany (in central Italy) is suffering increased price volatility on international markets, and the recent changes in the Common Agricultural Policy support (namely the transition from coupled payments to single farm payments, which aim to orient farmers’ business decisions to market signals). The economic sustainability of cereal farms is at risk, as the prices on the market hardly cover the production costs. The sharp contraction in the number of farms and cultivated areas over time clearly documents such crisis5.

Looking at downstream phases of the wheat-to-bread chain, wheat storage represents another weakness in the region, because it is fragmented on the territory and not conducive to product differentiation, resulting in increased operation and transportation costs. The milling of soft wheat in Tuscany is generally done by small-medium firms, fragmented over the territory. Regional transformation is not closely linked to supplies from Tuscany, although there are established links between some mills and their territorial supply basins.

The reasons for the use of supplies from outside the region are linked to price convenience and uniformity of batches, in relation to the requirements on the quality of raw material for baking purposes. In the current context, producer organizations and consortia play a key role in increasing the level of integration with the other actors of the chain (mills, pasta factories, bakeries, feed mills, etc.) and to pursue the valorization of a regional product, both through the concentration of the supply in qualitative homogeneous lots and by exploiting the peculiarities of the regional production system.

Industrially produced bread necessitates of specific standard characteristics of flour, made of grain blends whose provenance is often unknown (either national or continental, but also global) to the detriment of transparency within the supply chains. Moreover, the bread market in Italy is undergoing a deep process of change. At demand level, bread and pasta real prices have undergone a reversal over time: up to 1970s pasta’s price was higher than bread price per kg, while in the following years bread’s price became higher (3 euro for bread versus 1.5 of pasta, on average per kg). Despite the overall decline in bread consumption, the baking industry has resisted the crisis in different ways among the market segments: consumption of ‘table bread’, which refers to traditional freshly baked bread, is generally decreasing. The sector of bread substitutes, (i.e. dry breads and breadsticks, sandwich bread, unleavened bread and the segment of gluten free) has recently shown the highest increase rates.

In recent years, a re-localization trend has spread across Italy, as a possible strategy to revitalize the cereal sector through the enhancement of authenticity, quality and traceability of wheat. Bread is an important part of the cultural and social fabric associated with Tuscan territory, agriculture, gastronomy, and bread recipes reflect the specific culinary tradition of each Tuscan province. Over time, Tuscan Bread has maintained the characteristic of being baked without salt, to the point that this is now one of its distinctive features, and produced with sourdough (a dough containing a lactobacillus culture in symbiotic combination with yeasts) as a fermenting agent, which gives the final product a special aroma and flavor (Malandrin et al., 2015).

The proliferation of spontaneous initiatives promoting local baking, as well as the establishment of PDO and Protected Geographical Indications (PGI) for bread are indications the re-localization process developed by bridging the gap between producers, processors and consumers6. However, the protection of the integrity of traditional products does not prevent traditional food producers from the necessity to innovate (Vanhonacker et al., 2013). Moreover the effectiveness of PDO and PGI labels have been questioned, also by academic 5

In 2010, the Italian census of agriculture recorded 16,571 farms (23% of total number in Tuscany) active in grain production, about a half compared to the previous decade, while the contraction of cultivated areas is about one-third. The average area per farm dedicated to cereals is about 10.5 hectares, rising to 13.32 for durum wheat and dropping to just over 5 hectares for soft wheat.

6 Seven different Italian bread types, including Tuscan Bread PDO, are registered in the PDO/PGI catalogue, indicating the variety of traditional recipes and the linkage between wheat production, milling, bread baking and territorial specificities.

International Food and Agribusiness Management Review

539

Galli et al.

Volume 20, Issue 4, 2017

literature (Carbone et al., 2014). Innovation is a strategic task especially for small and medium enterprises: in the bread sector, the most interesting dynamics concern the ability to offer the right balance between taste, authenticity, healthiness, and at the same time adaptability to new lifestyles (including packaging).

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0174 - Tuesday, August 22, 2017 12:39:39 PM - IP Address:24.21.169.207

4.2 Actions by food system actors The process of recognition of Tuscan Bread as a PDO (according to Regulation 510/2006 (EC, 2006)) started in 2002, thanks to the initiative of a group of bakers, farmers’ associations, a milling industry and other local stakeholders, whilst the financial support was provided by the regional administration on agricultural policy funds. The aim was to promote and protect Tuscan Bread, codifying the original recipe and its related product specifications, including the use of varieties of wheat traditionally grown within the region. Product specifications strictly determine processing and features of the final product. Some crucial aspects differentiate a PDO bread supply chain from a conventional one: (1) wheat cultivated in Tuscany must belong to a set of soft wheat varieties, allowed in specific proportions; (2) flour must include the wheat germ; (3) sourdough leavening is compulsory; (4) no salt can be included in the recipe; (5) the final weight has to range between 0.45 and 1.10 kg; and (6) the Consortium packaging must be used. Product specifications strictly determine processing and features of the final product (see Figure 2 for an image of bread as available on the consortium website).

A ‘Consortium of Promotion for the Valorization of Sourdough Tuscan Bread’ was established in 2004 aiming to obtain the PDO recognition. The process lasted 13 years, and was completed in 2016. The nutritional ‘premium’, (further examined in the following paragraphs) was among the crucial aspects, together with the link to the regional pedo-climatic conditions, that allowed for the PDO recognition.

Figure 2. Tuscan Bread Protected Designation of Origin.

International Food and Agribusiness Management Review

540

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0174 - Tuesday, August 22, 2017 12:39:39 PM - IP Address:24.21.169.207

Galli et al.

Volume 20, Issue 4, 2017

Instrumental to the success of the initiative, was an ‘integrated supply chain project’ submitted and granted

by Tuscany Regional Administration (within the Rural Development Plan (RDP), 2006-2012) to enhance regional wheat production. The project, championed by a milling company as ‘channel captain’, establishes a formal relationship among actors of the regional supply chain, above all to encourage farmers recovering the production of traditional and native soft wheat varieties (in line with the requirements of the PDO specifications). The RDP measures included farm investment support to update technologies to the requirements of the agronomic protocol. The protocol defines the agronomic planning of farming operations and inputs that must be used to ensure homogeneity and quality of soft wheat. Funding for research and development was granted to the Universities of Pisa and Florence, to develop respectively innovative bread making technologies and agronomic protocols to facilitate and consolidate the manufacturing process based on objective parameters (such as markers or identifiers of chemical composition of bread). The commercial agreement defines production quantities and premium price ranges for farmers and coordinates stakeholders to the common aim of ensuring the quality of the final product, a prerequisite of the commercial and economic success of the initiative.

Figure 3 identifies the main actors involved in the Tuscan Bread PDO project, which mirrors the structure of the regional integrated supply chain. The project joins around 60 farmers: in 2016, out of 80,000 tons of soft wheat produced overall in Tuscany, 5,000 tons of soft wheat for PDO bread were planted, using certified seeds for traceability purposes. Price of wheat is anchored to the reference commodity market price (which is currently 180 euro per ton on the Bologna market (ISMEA, 2017)), to which a premium price is added (a range of minimum, average and maximum prices is set in the supply chain contract). For example, in 2016, soft wheat was paid 270 euro per ton (250 euro in 2015) to the consortium, (which retains approximately 30 euro per ton and then pays the farmers). 5,000 tons of wheat allow to produce 5,000 tons of bread (15% is lost in milling and 15% is recovered in baking, approximately). The integrated supply chain project ends in 2016, however the contract can be renovated year by year.

Collection and storage centers involved in the project invested in the expansion of storage capacity, endowment of equipment necessary to detect the quality of the wheat stored and the adoption of conservation techniques to ensure a healthier product. The farmers involved in the project stock their product at a Consortium (located in center of Tuscany, in Siena), which also provides extension services. Recently a second storage center in the north of Tuscany joined the Consortium. farmer farmer farmer farmer farmer farmer farmer farmer farmerfarmer

University of Florence

farmer

University of Pisa baker

baker

baker baker

baker baker

farmerfarmer

Consortium of Promotion

Pam Coop Tirreno

farmer Agricultural Consortium Siena

baker baker

baker

farmer

farmer

baker baker

farmer

farmer

farmer farmer

farmer farmer

farmer farmer farmer

farmer

farmer farmer

farmer

farmer

farmer farmer farmer farmer farmer

farmer

farmer

farmer

farmer

farmer

Miller

farmer

farmer

farmer

baker baker Small retailers (regional) baker

farmer

farmer

baker

Conad

farmer

farmer

farmer

farmer farmer

Figure 3. The Tuscan Bread Protected Designation of Origin integrated supply chain project members.

International Food and Agribusiness Management Review

541

Galli et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0174 - Tuesday, August 22, 2017 12:39:39 PM - IP Address:24.21.169.207

The Integrated Supply Chain Project encompasses one milling company, that’s also the project leader and among the promoters of the PDO recognition. After the PDO success, two other milling plants have joined the Consortium. The average price of flour for PDO bread purposes is approximately 500 euro per ton, which is higher than average price of white flour for baking (390 euro per ton (ISMEA, 2017)).

Twenty-five bakeries located all over Tuscany, mostly medium scale, are also included in the project. Bakers joining the Consortium have changed through the years, as not all of them were willing to follow such a long term and uncertain project. Moreover, the baking process for Tuscan Bread PDO determines higher costs of production with respect to more conventional bread, discouraging some of the bakers. Although these costs are hardly quantifiable, they are linked to longer baking times, which impact on labor, and to the organization of spaces for the sourdough leavening process, which impact on structures. All these conditions have led to a self-selection of larger size bakers, able to comply with PDO specifications, with the very small and artisanal farmers leaving the Consortium over the years. It should be noted that the Tuscan Bread PDO chain is defined as ‘industrial’ and as such it can exploit the economies of scale to afford the costs related to the search of quality, harder to achieve by small and artisanal production.

Because the larger size bakers remained involved through the PDO recognition process, large scale distribution represents the main market channel for Tuscan Bread PDO. The Consortium was able to bargain specific bread surplus management agreements. Normally, large scale distribution returns the unsold, left over bread, to the reference bakeries, at the end of the day: this has strong impact on the price of bread and the competitiveness of the sector, the profitability of bakers and not least, on waste. The Consortium activated ad hoc agreements with the main large distributors, by setting the rule that Tuscan Bread PDO should not be returned to bakers, relieving them from disposal of surplus bread.

Tuscan Bread PDO is currently sold on the market with the Consortium label at a minimum price of 3.5 euro per kilo compared to a regional average price of 1.94 euro per kilo of fresh bread (ranging between 1 and 4 euros, data from Italian Ministry of Economic Development, June 2015). Bestsellers’ products are large loaves, baked in wood ovens, and cut over the counter, where the customer can actually see the PDO label.

Traceability contributes to the prevention of contamination and adulteration, but it is not intended specifically for the end user. Full traceability of the Tuscan Bread PDO supply chain requires certification of seeds, dedicated storage structures for wheat and flour, to ensure wheat separation and identity preservation.

Traceability implementation entails the adoption of an IT-system by companies involved in the project, which allows to collect and centrally manage data on incoming raw materials and outgoing products for each operator involved in the production chain. The PDO quality label contributes by definition to the transparency towards consumers and enables farmers and producers to communicate about origin (the geographical distance between the supply chain stages within Tuscany regional borders is at maximum 100 km) and other features, irrespective of the number of intermediaries in the supply chain. Beyond delivering a message on the value-added quality of the production process, the ingredients and the baking processes, the PDO guarantees that an official control over the supply chain is implemented by one of the bodies recognized by the Italian Ministry of Agriculture. The marketing strategy of the Consortium is strongly pushing, beyond territoriality, on the nutritional qualities of the product, (i.e. low gluten, sourdough leavening, and preservation of organoleptic features).

4.3 Outcomes ■■ Performance and practice based indicators of Tuscan Bread Protected Designation of Origin Table 1 presents parameters related to the chemical-physical characterization of Tuscan Bread PDO in comparison to a widespread industrial white bread obtained through a fast leavening process. Additional data, referring to mold spoilage, staling process as well as nutritional value of both sourdough bread and industrial white bread, are shown in Table 1.

International Food and Agribusiness Management Review

542

Galli et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0174 - Tuesday, August 22, 2017 12:39:39 PM - IP Address:24.21.169.207

Table 1. Performance parameters of Sourdough Tuscan Bread and generic industrial bread.

Parameters

Sourdough Tuscan Bread

Generic industrial bread

pH Lactic acid Acetic acid • Mold spoilage1

3.4-4.6 0.4-0.8% 0.1-0.3% Protected from deterioration due to the antimicrobial substances produced by lactic acid bacteria Slower, because the acidifying lactic acid slows down the starch retrogradation • Reduced glycemic index: GI=50.

• Production of exopolysaccharides (EPS) with probiotic action.

• No salt.

• Increased bioavailability of amino acids.

• Degradation of phytic acid which forms complexes with certain ions difficult to assimilate.

• Increased solubility of the fibers.

• Good tolerance by celiac towards sourdough products.

Production of D/L-lactic and acetic acids and secondary products (aldehydes, ketones, esters, etc.)

5.3-5.8 0.005-0.040% 0.005-0.040% Sensitive to contamination by molds and bacteria Faster because of the reduced concentration of lactic acid • Glycemic index: GI=100.

• Reduced production of EPS with probiotic action.

• Presence of salt: 0.8 g/100g.

• Reduced bioavailability of amino acids.

• Presence of phytic acid.

• Reduced solubility of fibers.

• Staling1 • Nutritional value1

Flavor and taste 1

Reduced flavor complexity and savoriness

Secondary data, adapted from Venturi et al. (2013a, 2016b).

Tuscan Bread PDO is characterized by a lower pH and a higher concentration of lactic and acetic acid than the industrial white bread. In this context, we can also explain the reduced mold spoilage as well as the slower staling process generally observed in sourdough bread, compared to industrial white bread. As reported in previous papers (Venturi et al., 2013a, 2016b), the PDO bread is characterized by a significantly higher production of secondary products that allow to obtain an enhanced flavor complexity and a higher savoriness than a widespread industrial white bread.

Table 2 identifies a set of practices of the bread making process that have potential implications on the whole quality as well as on the nutritional value of bread, and illustrates them with the Tuscan Bread PDO chain compared to the ones commonly adopted in standard white bread supply chains.

Table 2. Supply chain practices and impact on nutrition.

Tuscan Bread PDO1

Supply chain stage

Practice

1. Seed and wheat cultivation

Choice of soft wheat varieties Old/ancient wheat varieties traditionally harvested in Tuscany Choice of flour Flour includes wheat germ

2. Milling

3. Mixing, Choice of leavening method fermentation and baking 4. Distribution Communication strategy and transparency all along the chain 1

Standard industrial white bread No limitation about flour origin and composition

Sourdough leavening

White flour without inclusion of wheat germ Use of improvers

Full traceability of bread along the wheat to bread chain

Provenance and quality of grains and flour often unknown

PDO = Protected Designation of Origin.

International Food and Agribusiness Management Review

543

Galli et al.

Volume 20, Issue 4, 2017

■■ Seed and wheat cultivation

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0174 - Tuesday, August 22, 2017 12:39:39 PM - IP Address:24.21.169.207

The suitability of a variety of wheat for breadmaking is determined mainly by its genetic make-up. A wheat variety is more suitable for bread-making when the ability of its proteins to form the dimensional networks of gluten during kneading is greater. Environmental factors, such as nitrogen fertilization, water and temperature influence protein content. By contrast, protein quality is largely under genetic control (Callejo et al., 2015).

The choice of wheat varieties to be utilized for bread-making greatly influence the bread quality as well as its nutritional value. It has been recognized that ancient varieties of wheat have a higher genetic variability, thus contributing to enhance biodiversity (Dinelli et al., 2009). In Italy, old landraces and varieties, which underwent intensive breeding programs at the beginning of the twentieth century, were replaced after the Second World War with the modern semi-dwarf and high-yielding cultivars (Rascio et al., 2015).

Current interest in the health benefit of grain consumption has led to an increased focus on the phytochemical content of different grains and grain varieties. Ancient wheats have been recognized to offer unique nutraceutical values for their peculiar content in health-beneficial phytochemicals (Dinelli et al., 2009). Although no significant differences among investigated cultivars were detected in relation to the amounts of total phenolic and flavonoid compounds, the qualitative phytochemical profile between old and modern varieties was remarkably diverse. The peculiar and varied phytochemical profile of investigated old wheat genotypes confirmed that ancient grains may represent a rich source of genetic diversity, especially with regard to functional properties (Dinelli et al., 2009).

As regards the indication of wheat varieties admitted for Tuscan bread PDO production, the selected ones by ‘Consortium of Tuscan Bread Sourdough’ represent wheat varieties traditionally harvested in Tuscany, to guarantee a strong linkage between the Tuscan bread and its territory of origin7. These can be mixed with more recent varieties to obtain a flour characterized by the needed technological features.

■■ Milling The milling of grain to produce white flours reduces the levels of phytates as well as the levels of many nutrients present in different parts of the wheat kernel. The wheat germ, is a rich source of B-vitamins, oil, vitamin E and lipids, composed of essential fatty acids, mainly palmitic and linoleic acids. However, germ is discarded during industrial milling because the fat is liable to become rancid during storage. Eventually, the re-addition of wheat germ into the white flour allows to recover part of the nutrients lost. Wheat germ is a high nutritive by-product of the flour industry (Gomez et al., 2012). It is characterized by a high protein content, mainly in the form of albumins and globulins, and a balanced amino acid composition, which repairs the defects of cereal proteins. The germ also provides six times more minerals, seven times more fat and fifteen times more sugars than does white flour without germ. Antioxidants in wheat germ are useful to prevent cardiovascular diseases and cancer. Therefore, wheat germ offers an appropriate medium to convey these benefits to human diet (Gomez et al., 2012).

Generally, wheat germ has a considerable fat content of around 10 g/100 g, and a significant quantity of bioactive molecules related to its role as part of the embryo of a new plant (Gomez et al., 2012). Among these bioactive molecules, the lipases and the lipoxygenases can hydrolyze lipids and initiate the oxidative rancidity process. This fact and the high unsaturated fatty acid content of wheat germ lead to the very short shelf-life of raw germ (Sjovall et al., 2000) or even flours with germ. Because of the limited shelf life, the utilization of flour including wheat germ is generally avoided to produce widespread industrial white bread.

7

The selected varieties are listed as: red wheat varieties (≤80% of total): Centauro, Bilancia, Serio, Verna and Pandas. White wheat varieties (≤50% of total): Mieti, Mec, Marzotto and Bolero.

International Food and Agribusiness Management Review

544

Galli et al.

Volume 20, Issue 4, 2017

Based on the indications provided by Tuscan Bread PDO Consortium, the flour utilized for bread production must be white flour including wheat germ, in order to obtain a foodstuff characterized by a sensorial expression strongly linked to the bread traditionally produced in Tuscany, but showing an increased nutritional value.

This is allowed by the technological facilities of the milling plant involved in the integrated supply chain project, which extracts the germ from the kernel at the beginning of the process and re-adds it subsequently.

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0174 - Tuesday, August 22, 2017 12:39:39 PM - IP Address:24.21.169.207

■■ Mixing, fermentation and baking Sourdough is a mixture of flour (mainly wheat or rye) and water, fermented with lactic acid bacteria and yeasts, which are responsible for its capacity to leaven a dough, while contemporarily and unavoidably acidifying it (De Vuyst et al., 2014). In the modern bakery technology, sourdough represents an alternative to the use of baker’s yeast (although bakers often use a combination of both leavening agents) to manufacture a variety of products (Minervini et al., 2014) such as bread, crackers, snacks, pizza and sweet baked goods, because it offers many advantages over baker’s yeast: enhanced flavor, prolonged shelf-life, improved dough structure and increased nutritional value of the leavened baked good (Arendt et al., 2007; Katina et al., 2005).

Sourdough processes can be used to modify levels of bioactive compounds, however, not much data is available (Katina et al., 2005). Sourdough fermentation has been reported to increase folate content, decrease tocopherol and tocotrienol content, and decrease or increase thiamin content depending on the process (Katina et al., 2005). Thus, sourdough fermentation can both increase or decrease the levels of bioactive compounds depending on the nature of the compound and the type of the sourdough process. The presence of yeast seems to favor the formation of folates and thiamin. Formation of acidity can both increase levels of bioactive compounds (such as total amount of phenolic compounds) or decrease levels of some compounds (such as thiamin, ferulic acid dehydrodimers, tocopherols and tocotrienols). The degradation of phytate, already discussed above, has repeatedly been reported in sourdough processes (Katina et al., 2005).

Phytates can chelate and bind minerals, resulting in insoluble complexes that may lead to a decrease in mineral absorption and bioavailability, and therefore the removal of phytates from baked goods has long been considered desirable. Evidence demonstrating a diverse range of benefits to health and wellbeing is now accumulating. Sourdough fermentation has been shown to decrease the amount of phytate in wholegrain cereals. Many studies have indicated that phytate hydrolysis during dough fermentation significantly enhances the bioavailability of minerals including calcium, copper, magnesium, zinc and iron. The results reported

by Buddrick et al. (2014), indicate that the proving time had the most significant impact on the phytate reduction which might enhance mineral bioavailability of the final products.

According to the indications provided by Consortium, sourdough leavening is compulsory in Tuscan Bread production. The operating conditions related to the sourdough utilization and storage have been defined and bakers must readapt the organization of the production process to be able to comply with such operating conditions. This represents a challenge for those bakers who rely only on bakers’ yeast fermentation, which allows standard leavening, flavor and shorter production times. Comparing to the industrial bread production, it is worth noticing that mechanization, large scale production and increased consumer demand for high quality, convenience and longer shelf life have created the need for functional food additives such as emulsifiers and anti-staling agents to achieve desired quality in bread. Addition of emulsifiers is particularly important for large scale, industrial bread baking as these impart greater dough strength to withstand machine handling, improve rate of hydration, improve crumb structure, improve slicing characteristic, improve gas holding capacity and extend shelf life (Moayedallaie et al., 2010; Mondal et al., 2008).

■■ Storage and distribution The use of sourdough has been reported to have positive effects on bread staling, thanks to the role of lactic acid bacteria (Arendt et al., 2007). One such effect is an improvement in loaf specific volume, which is associated with a reduction in the rate of staling, as has been demonstrated by a reduction in crumb softness International Food and Agribusiness Management Review

545

Galli et al.

Volume 20, Issue 4, 2017

loss for sourdough breads during storage. A decrease in the staling rate as measured by differential scanning calorimetry has also been reported for breads containing sourdough8.

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0174 - Tuesday, August 22, 2017 12:39:39 PM - IP Address:24.21.169.207

In traditional bread making processes, such as that indicated for Tuscan Bread Sourdough, the use of sourdough appears as the main step in order to allow to extend the product shelf life in a ‘natural’ way, despite the operating conditions required for sourdough leavening being quite expensive in terms of time and labor.

The breadmaking industry has undergone important changes in the last decades because of mechanization for increasing production and consumer demands. The mechanization of the breadmaking process involves dough rheology changes and the frozen technologies also require recipe modification for reducing the freezing damage. Consumers demand products of better quality and longer shelf-life. The search for solutions to meet those requirements has been parallel to the development of different additives and technological aids that modify dough rheology and improve bread quality. Additives and technological aids are extensively used in the baking industry for improving dough machinability in the case of emulsifiers and enzymes, bread characteristics by using enzymes, hydrocolloids, emulsifiers and to extend the shelf life of the resulting products (Gomez et al., 2004).

5. Discussion The cases study shows that a quality differentiation process – based on the valorization of traditions and territory, coupled with the improvement of nutritional value – does not concern isolated actors but activates the food system at different levels, leading to systemic changes. The crisis of the regional cereal sector has acted as a driving force towards the innovation of the wheat-to-bread chain, supported by the policy maker, who has enabled structural investments, remuneration of farmers, and higher quality and nutrition for consumers.

Coordination and governance of the food chain are key aspects in relation to outcomes, including nutrition: integrating the supply chain by establishing an explicit connection (i.e. a contract) among partners has led to a change in the overall governance structure.

According to literature on the governance of global value chains, ‘internal’ governance is referred to the relations within the chain. Governance as ‘drivenness’ is the process of organizing activities, allocating resources and distribution of gains, and determining inclusion or exclusion of other firms (Gereffi, 1994; Wilkinson, 2006). The leading role of the Consortium is crucial in the organization of the Tuscan Bread PDO chain, since the beginning to the end of a very lengthy PDO recognition process. The Consortium also plays a power rebalancing role, for example when it engages in the negotiation with retailers, (i.e. to avoid surplus bread return as a responsibility of bakers).

Other governance approaches refer to coordination (Gereffi, 2005; Gibbon et al., 2008) and normalization, as a process of re-aligning a given practice to follow a norm (Gibbon and Ponte, 2008). Drawing from conventions theory, ‘quality conventions’ are mutual expectations that include, but are not limited to, institutions (Ponte, 2009): when price alone cannot evaluate quality, economic actors adopt other conventions to assess quality (i.e. trust, reputation, certification, etc.). Conventions of quality thus become instruments of coordination of the supply chain, as they provide the necessary knowledge base and determine the behavior of actors. A quality convention based on health and regional identity helps to set standards that regulate the stability of raw material and information flows, preserves the territorial identity of a product, avoids the exploitation of local resources, and fairly distributes costs and benefits among involved stakeholders (Brunori et al., 2016).

8

It has been noted, however, that the anti-staling effect found for sourdough is dependent from the particular strain performing the fermentation, and that this effect involves dynamics other than those associated with the degree of acidification. Starch molecules can be affected by enzymes produced by lactic acid bacteria, causing a variation in the retrogradation properties of the starch. This in turn slows the rate of staling. Additionally, proteolysis of gluten subunits has also been proposed. Additionally, proteases aid the liberation of water associated with the protein network thus allowing for an increased alpha amylase activity. The acidic conditions and proteases associated with sourdough play a role in reducing the staling rate of bread (Arendt et al., 2007).

International Food and Agribusiness Management Review

546

Galli et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0174 - Tuesday, August 22, 2017 12:39:39 PM - IP Address:24.21.169.207

‘External’ governance refers to the distribution of duties and rights between the firms and stakeholders in a broader sense (i.e. including civil society and institutions) (Sacconi, 2006). When considering performance in relation to nutrition, ‘extended’ governance assumes a key role. Public administrations, producers’ consortia, third party certifiers and research actors all contribute to the definition of the regulatory context, enforce quality controls and can exert pressure on firms to frame competition on nutrition performance.

Governance change can be interpreted by referring to the analytics of governmentality, suggested in relation to sustainability by Spence and Rinaldi (2014) and Dean (2009). In the Tuscan PDO case, transformation and appraisal of its consequences do not take place just within one organization, but across the organizational boundaries of the supply chain and the wider food system. Overall, the analytics (fields of visibility, techne, episteme) contribute to the ‘identity formation’ of the different stakeholders of the chain (Spence and Rinaldi, 2014: 43). The ‘field of visibility’, as a condition to competitiveness and reputation building, develops around the link to territory, improved nutritional value and increased transparency for the consumer. The field of visibility corresponds to distinctive practices along the chain with economic and technological implications (i.e. ‘techne’ of governmentality). The most striking example is the sourdough leavening process, which improves nutrition and requires the re-organization of the production process, and impacts on labor working hours and structures. Similarly, the inclusion of the wheat germ into flour, which enhances sensorial expression and maintains part of the nutritional value contained into the wheat kernel, needs a technological adaptation of industrial facilities of the milling plant, and the management of stocks for the optimal and safe conservation of flour. In relation to transparency, full traceability is a necessary tool which envisages the design and development of an information system to support companies involved in the project, to collect and centrally manage data on incoming raw materials and outgoing products for each operator involved in the supply chain and identification of any chemical identifiers (i.e. markers) capable of distinguishing unambiguously the finished product.

Furthermore, the PDO specifications, that codify quality among producers and convey a message on the origin, expresses at once the plans and the know-how, codified and made available to a community (i.e. the ‘episteme’) (Spence and Rinaldi, 2014: 43). The PDO specification also represents a challenge, in relation the know-how required to re-discover and uptake traditional production practices that have been lost, due to modern continuous baking processes (which led to the removal of sourdough leavening for bread for example). A training for bakers is currently being organized by the Consortium, as a necessary action to allow the compliance with PDO specifications, by an increasing number of bakeries in the region. This training addresses particularly the handling and management of sourdough in accordance to PDO specifications, and with the support of research institutions.

6. Conclusions The Consortium of Tuscan Bread has embarked on a long-term process of innovation to compete with conventional industrial white bread: the territorial differentiation, gained through the PDO labelling, is reinforced by enhancing nutritional value. This process is not linear and addresses multiple levels of the food system, aiming at delivering a healthier bread and encountering consumers’ preference for localness and traditional products.

The food system perspective adopted enables to shed light on the multiple levels involved in the transition towards an improved nutrition: the revitalization of the soft wheat sector, triggered by supply chain actors and supported by policy that allows farmers to receive a higher price for wheat; the recognition of the differential quality to processors, through a tool as the PDO label; the increased transparency in a very opaque market, through traceability.

The success of the differentiation strategy adopted by the stakeholders relies on the integration of the supply chain stakeholders and requires a change in the governance. This allows the realization of an interplay among price, taste and nutritional value which would not take place without an explicit supply chain agreement.

International Food and Agribusiness Management Review

547

Galli et al.

Volume 20, Issue 4, 2017

The regional scale provides a favorable context to balance territorial identity, artisanal know-how and industrial scale. It is a sort of ‘laboratory’ to experiment with standardization of ‘artisanal’ practices for the attainment of economies of scale and eventually reach wider markets (e.g. standardization of the sourdough leavening process).

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0174 - Tuesday, August 22, 2017 12:39:39 PM - IP Address:24.21.169.207

The Tuscan Bread PDO enhancement potential will be fully visible after some time from the recent PDO recognition. Only then, a complete assessment of the sustainability of the initiative will be possible.

Acknowledgements The research reported in this paper was funded by the EU within the project Global and local food chain assessment: a multidimensional performance-based approach (GLAMUR, 7 th framework program, grant agreement no. 311778). Furthermore, this work was supported by the Tuscany Region PIF Project INNOVAPANE ‘Processi innovativi per la produzione del Pane Toscano a lievitazione naturale (DOP)’ (D.D. n. 2260-12.06.2013). The views expressed in this paper are solely those of the authors. We thank the anonymous reviewers for the useful comments and insights.

Supplementary material Supplementary material can be found online at https://doi.org/10.22434/IFAMR2015.0174.

Materials S1. Chemical analysis.

References Allen, S., A. de Brauw and A. Gelli. 2016. Harnessing value chains to improve food systems international food policy research institute. Global Food Policy Report. International Food Policy Research Institute, Washington, WA, USA, 49-55.

Aprile, M.C., V. Caputo and R.M. Nayga Jr. 2016. Consumers’ preferences and attitudes toward local food products. Journal of Food Products Marketing 22: 19-42.

Arendt, E.K, L. Am Ryan and F. Dal Bello. 2007. Impact of sourdough on the texture of bread. Food microbiology 24:165-174.

Auestad, N. and V.L. Fulgoni. 2015. What current literature tells us about sustainable diets: emerging research linking dietary patterns, environmental sustainability, and economics. Advances in Nutrition 6:19-36.

Augustin, M.A., M. Riley, R. Stockmann, L. Bennett, A. Kahl, T. Lockett, M. Osmond, P. Sanguansri, W. Stonehouse, I. Zajac and L. Cobiac. Role of food processing in food and nutrition security. Trends in Food Science & Technology 56: 115-125.

Barling, D., R. Sharpe and T. Lang. 2009. Traceability and ethical concerns in the UK wheat –bread chain: from food safety to provenance to transparency. International journal of agricultural sustainability 7:261-278.

Bogomolova, S., A. Loch, L. Lockshin and J. Buckley. In press. Consumer factors associated with purchasing local versus global value chain foods. Renewable Agriculture and Food Systems. Doi: http://doi. org/10.1017/S1742170516000375.

Brunori, G. and F. Galli. 2016. Sustainability of local and global food chains: introduction to the special issue. Sustainability 8: 765.

Brunori, G., F. Galli, D. Barjolle, R. van Broekhuizen, L. Colombo, M. Giampietro, J. Kirwan, T. Lang, E. Mathijs, D. Maye, K. de Roest, C. Rougoor, J. Schwarz, E. Schmitt, J. Smith, Z. Stojanovic, T. Tisenkopfs and J-M. Touzard. 2016. Are local food chains more sustainable than global food chains?

Considerations for Assessment Sustainability 8: 1-27.

Buddrick, O., O.A.H. Jones, H.J. Cornell and D.M. Small. 2014. The influence of fermentation processes and cereal grains in wholegrain bread on reducing phytate content. Journal of Cereal Science 59: 3-8.

International Food and Agribusiness Management Review

548

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0174 - Tuesday, August 22, 2017 12:39:39 PM - IP Address:24.21.169.207

Galli et al.

Volume 20, Issue 4, 2017

Callejo, M.J., M.E. Vargas-Kostiuk and M. Rodríguez-Quijano. 2015. Selection, training and validation process of a sensory panel for bread analysis: influence of cultivar on the quality of breads made from common wheat and spelt wheat. Journal of Cereal Science 61: 55-62.

Campos, S., J. Doxey and D. Hammond. 2011. Nutrition labels on pre-packaged foods: a systematic review.

Public health nutrition 14: 1496-1506.

Caputo, V., R.M. Nayga and R. Scarpa. Food miles or carbon emissions? Exploring labelling preference for food transport footprint with a stated choice study. 2013. Australian Journal of Agricultural and Resource Economics 57: 465-482.

Carbone, A., J. Caswell, F. Galli and A. Sorrentino. 2014. The performance of protected designations of origin: an ex post multi-criteria assessment of the Italian cheese and olive oil sectors. Journal of Agricultural and Food Industrial Organization 12: 12.

Cordain, L. 2010. The paleo diet revised: lose weight and get healthy by eating the foods you were designed to eat. Houghton Mifflin Harcourt, Boston, MA, USA. Cordell, D., J.O. Drangert and S. White 2009. The story of phosphorus: global food security and food for thought. Global environmental change 19: 292-305.

Cowburn, G. and L. Stockley. 2005. Consumer understanding and use of nutrition labelling: a systematic review. Public health nutrition 8: 21-28.

De Vuyst, L., S. Van Kerrebroeck, H. Harth, G. Huys, H-M. Daniel, and S. Weckx. 2014. Microbial ecology of sourdough fermentations: diverse or uniform? Food microbiology 37: 11-29.

Dean, M. 2009. Governmentality: power and rule in modern society. SAGE, London, UK. Dewettinck, K., F. Van Bockstaele, B. Kühne, D. Van de Walle, T.M. Courtens and X. Gellynck. 2008.

Nutritional value of bread: influence of processing, food interaction and consumer perception.

Journal of Cereal Science 48: 243-257.

Dinelli, G., A.S. Carretero, R. Di Silvestro, I. Marotti, S. Fu, S. Benedettelli, L. Ghiselli and A. Fernández Gutiérrez. 2009. Determination of phenolic compounds in modern and old varieties of durum wheat using liquid chromatography coupled with time-of-flight mass spectrometry. Journal of Chromatography 1216: 7229-7240.

Downs, S. and J. Fanzo. 2016. Managing value chains for improved nutrition. good nutrition: perspectives for the 21st century. Karger, Basel, Switzerland, pp. 45-59.

Eisenhardt, K.M. 1989. Building theories from case study research. Academy of management review 14: 532-550.

Ericksen, P.J. 2008. Conceptualizing food systems for global environmental change research. Global Environmental Change 18: 234-245.

European Commission (EC). 2006. Council Regulation (EC) No 510/2006 of 20 March 2006 on the protection of geographical indications and designations of origin for agricultural products and foodstuffs. Official Journal of the European Union L 93: 12-25.

FAO. 2013. Sustainability assessment of food and agriculture systems (SAFA). Guidelines, indicator and tool. Available at: http://tinyurl.com/lngtb5m.

FAO. 2017. The future of food and agriculture – Trends and challenges. FAO, Rome, Italy.

FAO Stat. 2015. Statistical Pocket Book. Food and Agriculture Organization of the United Nations, Rome, Italy. Available at: http://www.fao.org/3/a-i4691e.pdf.

Galli, F., F. Bartolini and G. Brunori. 2016. Handling diversity of visions and priorities in food chain sustainability assessment. Sustainability 8: 305.

Galli, F., F. Bartolini, G. Brunori, L. Colombo, O. Gava, S. Grando and A. Marescotti. 2015. Sustainability assessment of food supply chains: an application to local and global bread chains in Italy. Agricultural and Food Economics 3: 21.

Garnett, T. 2013. Food sustainability: problems, perspectives and solutions. Proceedings of the Nutrition Society 72: 29-39.

Gelli, A., C. Hawkes, J. Donovan, J. Harris, S. Allen, A. de Brauw, S. Henson, N. Johnson, J. Garrett and D. Ryckembusch. 2015. Value chains and nutrition: a framework to support the identification, design, and evaluation of interventions. IFPRI Discussion Paper 01413. International Food Policy Research Institute, Washington DC, WA, USA. International Food and Agribusiness Management Review

549

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0174 - Tuesday, August 22, 2017 12:39:39 PM - IP Address:24.21.169.207

Galli et al.

Volume 20, Issue 4, 2017

Gereffi, G. 1994. The organization of buyer-driven global commodity chains: how US retailers shape overseas production networks. In: Commodity chains and global capitalism, edited by G. Gereffi and M. Korzeniewicz. Praeger, Westport, CT, USA, pp. 93-122.

Gereffi, G., J. Humphrey and T. Sturgeon. 2005. The governance of global value chains. Review of International Political Economy 12: 78-104.

Gibbon, P., J. Bair and S. Ponte. 2008. Governing global value chains: an introduction. Economy and Society 37: 315-338.

Gibbon, P. and S. Ponte. 2008. Governing value chains: from governance to governmentality? Economy and Society 37: 365-392.

Golan, E. and L. Unnevehr. 2008. Food product composition, consumer health, and public policy: introduction and overview of special section. Food Policy 33: 465-469.

Gómez, M., S. Del Real, C.M Rosell, F. Ronda, C.A. Blanco and P.A. Caballero. 2004. Functionality of different emulsifiers on the performance of breadmaking and wheat bread quality. European Food Research and Technology 219: 145-150.

Gómez, M., J. González and B. Oliete. 2012. Effect of extruded wheat germ on dough rheology and bread quality. Food and Bioprocess Technology 5: 2409-2418.

Grunert, K.G. and J.M. Wills. 2007. A review of European research on consumer response to nutrition information on food labels. Journal of Public Health 15: 385-399.

Hammond, R.A. and L. Dubé. 2012. A systems science perspective and transdisciplinary models for food and nutrition security. Proceedings of the National Academy of Sciences 109: 12356-12363.

Hawkes, C., S. Friel, T. Lobstein and T. Lang. 2012. Linking agricultural policies with obesity and noncommunicable diseases: a new perspective for a globalising world. Food Policy 37: 343-353.

Hellyer, N.E., I. Fraser and J. Haddock-Fraser. 2012. Food choice, health information and functional ingredients: an experimental auction employing bread. Food Policy 37: 232-245.

HLPE. 2014. Food losses and waste in the context of sustainable food systems. A report by the High-Level Panel of Experts on Food Security and Nutrition of the Committee on World Food Security, Rome.

Available at: http://www.fao.org/3/a-i3901e.pdf.

Ingram, J. 2011. A food systems approach to researching food security and its interactions with global environmental change. Food Security 3: 417-431.

ISMEA. 2017. Settore cereali. Sheda di settore. Available at: http://tinyurl.com/m2s7rlk.

Jones, A.D. and G. Ejeta. 2016. A new global agenda for nutrition and health: the importance of agriculture and food systems. Bulletin of the World Health Organization 94: 228.

Jones, J.M., J. Adams, C. Harriman, C. Miller and J.W. Van der Kamp. 2015. Nutritional impacts of different whole grain milling techniques: a review of milling practices and existing data. Cereal Foods World 60: 130-139.

Katina, K., E. Arendt, K-H. Liukkonen, K. Autio, L. Flander and K. Poutanen. 2005. Potential of sourdough for healthier cereal products. Trends in Food Science & Technology 16: 104-112.

Larsson, S.C., E. Giovannucci, L. Bergkvist and A. Wolk. 2005. Whole grain consumption and risk of colorectal cancer: a population-based cohort of 60 000 women. British journal of cancer 92: 1803-1807.

Malandrin, V., A. Rossi, L. Dvortsin and F. Galli. 2015. The evolving role of bread in the tuscan gastronomic culture. In: Gastronomy and culture, edited by K. Csobán and H.E. Könyves, University of Debrecen, Litográfia Nyomda, Hungary.

Mancino, L., F. Kuchler and E. Leibtag. 2008. Getting consumers to eat more whole-grains: the role of policy, information, and food manufacturers. Food Policy 33: 489-496.

McKevith, B. 2004. Nutritional aspects of cereals. Nutrition Bulletin. British Nutrition Foundation, London, UK. Mellen, P.B., T.F. Walsh and D.M. Herrington. 2008. Whole grain intake and cardiovascular disease: a metaanalysis. Nutrition, Metabolism and Cardiovascular Diseases 18: 283-290.

Minervini, F., M. De Angelis, R. Di Cagno and M. Gobbetti. 2014. Ecological parameters influencing microbial diversity and stability of traditional sourdough. International journal of food microbiology 171: 136-146.

International Food and Agribusiness Management Review

550

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0174 - Tuesday, August 22, 2017 12:39:39 PM - IP Address:24.21.169.207

Galli et al.

Volume 20, Issue 4, 2017

Moayedallaie, S., M. Mirzaei and J. Paterson. 2010. Bread improvers: comparison of a range of lipases with a traditional emulsifier. Food Chemistry 122: 495-499.

Mondal, A., and A.K. Datta. 2008. Bread baking–a review. Journal of Food Engineering 86: 465-474.

Mozaffarian, D. and D.S. Ludwig. Dietary guidelines in the 21st century – a time for food. Journal of the American Medical Association 304: 681-682.

Nestle, M. 2013. Food politics: how the food industry influences nutrition and health. University of California Press, Berkeley, CA, USA. Paluck, E.L. 2010. The promising integration of qualitative methods and field experiments. The Annals of the American Academy of Political and Social Science 628: 59-71.

Pohjanheimo, T., R. Paasovaara, H. Luomala and M. Sandell. 2010. Food choice motives and bread liking of consumers embracing hedonistic and traditional values. Appetite 54: 170-80.

Ponte, S. 2009. Governing through quality: conventions and supply relations in the value chain for South African wine. Sociologia Ruralis 39: 236-257.

Popkin, B.M. 2006. Global nutrition dynamics: the world is shifting rapidly toward a diet linked with noncommunicable diseases. The American Journal of Clinical Nutrition 84: 289-298.

Rascio, A., V. Picchi, J.P. Naldi, S. Colecchia, G. De Santis, A. Gallo, E. Carlino, R. Lo Scalzo, and L. De Gara. 2015. Effects of temperature increase, through spring sowing, on antioxidant power and health-beneficial substances of old and new wheat varieties. Journal of Cereal Science 61: 111-118.

Sacconi, L. 2006. A social contract account for CSR as an extended model of corporate governance (I): Rational bargaining and justification. Journal of Business Ethics 68: 259-281.

Sharpe, R., D. Barling and T. Lang. 2008. Ethical traceability in the UK wheat-flour-bread chain. In Ethical traceability and communicating food. Springer, London, UK. Sjövall, O., T. Virtalaine, A. Lapveteläinen and H. Kallio. 2000. Development of rancidity in wheat germ analyzed by headspace gas chromatography and sensory analysis. Journal of Agricultural and Food Chemistry 48: 3522-3527.

Spence, L.J. and L. Rinaldi. 2014. Governmentality in accounting and accountability: a case study of embedding sustainability in a supply chain. Accounting, Organizations and Society 39: 433-452.

Starr, M.A. 2014. Qualitative and mixed-methods research in economics: surprising growth, promising future. Journal of Economic Surveys 28: 238-264.

Stuckler, D. and M. Nestle. 2017. Big food, food systems, and global health. Food insecurity and disease: prevalence, policy, and politics. Apple Academic Press, Waretown, NJ, USA, pp. 231-240.

Swinburn, B., C. Dominick and S. Vandevijvere. 2014. Benchmarking food environments: experts’ assessments of policy gaps and priorities for the New Zealand Government. Auckland. University of Auckland, Auckland, New Zealand.

UNEP. 2016. Food systems and natural resources. A report of the working group on food systems of the international resource panel. Available at: http://tinyurl.com/ms2bef3.

Vanhonacker, F., B. Kühne, X. Gellynck, L. Guerrero, M. Hersleth and W. Verbeke 2013. Innovations in traditional foods: impact on perceived traditional character and consumer acceptance. Food Research International 54: 1828-1835.

Venturi, F., G. Andrich, C. Sanmartin and A. Zinnai. 2013a. The kinetics of fermentations in sourdough bread stored at different temperature and influence on bread quality. Journal of Bioprocessing and Biotechniques 3: 2.

Venturi, F., G. Andrich, M.F. Quartacci, C. Sanmartin, L. Andrich and A. Zinnai. 2013b. A kinetic method to identify the optimum temperature for β-glucanase activity. South African Journal of Enology and Viticulture 34(2): 281-286.

Venturi, F., C. Sanmartin, I. Taglieri, A. Nari, G. Andrich and A. Zinnai. 2016b. Effect of baking process on artisanal sourdough bread-making: A technological and sensory evaluation. Agrochimica 60(3): 223-234.

Venturi, F., C. Sanmartin, I. Taglieri, Y. Xiaogou, G. Andrich and A. Zinnai. 2016a. The influence of packaging on the sensorial evolution of white wine as a function of the operating conditions adopted during storage. Agrochimica 60(2): 150-160.

International Food and Agribusiness Management Review

551

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0174 - Tuesday, August 22, 2017 12:39:39 PM - IP Address:24.21.169.207

Galli et al.

Volume 20, Issue 4, 2017

Verbeke, W. 2008. Impact of communication on consumersâ&#x20AC;&#x2122; food choices. Proceedings of the Nutrition Society 67: 281-288.

Wang, Y.C., K. McPherson, T. Marsh, S.L. Gortmaker and M. Brown. 2011. Health and economic burden of the projected obesity trends in the USA and the UK. The Lancet 378: 815-825.

Waterlander, W.E., C.N. Mhurchu, H. Eyles, S. Vandevijvere, C. Cleghorn, P. Scarborough, B. Swinburn and J. Seidell. In press. Food futures: developing effective food systems interventions to improve public health nutrition. Agricultural Systems. DOI: https://doi.org/10.1016/j.agsy.2017.01.006.

Wilkinson, J. 2006. Fish: a global value chain driven onto the rocks. Sociologia Ruralis 46: 139-153.

Yin, R.K. 1981. The case study crisis: some answers. Administrative science quarterly 26:58-65.

Yin, R.K. 2003. Case studies research: design and methods. Sage, Thousand Oaks, CA, USA. Zinnai, A., F. Venturi, C. Sanmartin and G. Andrich. 2013. Changes in physicochemical and sensory characteristics of fresh bread rolls maintained in different storage conditions. Agrochimica 56(3): 77-89.

International Food and Agribusiness Management Review

552

OPEN ACCESS International Food and Agribusiness Management Review Volume 20 Issue 4, 2017; DOI: 10.22434/IFAMR2015.0218

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0218 - Tuesday, August 22, 2017 12:33:35 PM - IP Address:24.21.169.207

Received: 7 December 2015 / Accepted: 29 March 2017

Socioeconomic impacts of innovative dairy supply chain practices – The case of the Laiterie du Berger in the Senegalese Sahel RESEARCH ARTICLE Abdrahmane Wane a, Jean-Joseph Cadilhonb, and Mamadou Yauckc aSenior

bSenior

Drylands Economist, CIRAD-PPZS and ILRI-PIL, International Livestock Research Institute, P.O. Box 30709, 00100 Nairobi, Kenya

Agricultural Economist, Policy, Institutions and Livelihoods Program, International Livestock Research Institute, P.O. Box 30709, 00100 Nairobi, Kenya

cStatistician

Economist, Pole on Pastoralism and Drylands (PPZS), CIRAD Delegation Régionale, 37 Avenue Jean XIII, BP 6189 Dakar-Etoile, Dakar, Senegal; PhD student, Laval University, 2185 avenue Chapdelaine, G1V 1M9 Québec, Canada

Abstract This study analyzes the Laiterie Du Berger (LDB)’s milk supply chain and its contribution to strengthening the food security and socioeconomic resources of Senegalese Sahelian pastoral households. Porter’s value chain model is used to characterize the innovations introduced by the LDB dairy in its milk inbound logistics and supplier relationships. A socioeconomic food security index and qualitative data are used to assess the dairy’s supply chain’s contribution to strengthen smallholder households’ livelihoods. Data for this research were obtained through individual surveys, focus groups and in-depth interviews of LDB managers and milk suppliers. Results show that milk income contributes significantly to household food security. Suppliers who stabilize their dairy income between rainy and dry seasons, diversify income sources and have larger herds are more likely to remain food secure. The LDB innovations contribute by helping herders access biophysical and economic resources, leading to better livestock feed and household food security.

Keywords: innovation, dairying, food security, pastoralism, value chains JEL code: D13, O33, O35, O54, Q12, Q13 Corresponding author: awane@cirad.fr

© 2017 Wane et al.

553

Wane et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0218 - Tuesday, August 22, 2017 12:33:35 PM - IP Address:24.21.169.207

1. Introduction The livestock system in Senegal is mainly dominated by traditional activities. These cannot be measured solely in quantitative or monetary terms because they also have significant non-market drivers, which may be as or more important than market drivers (Wane et al., 2014). Livestock keeping occupies 30% of the population. Overall, 90% of rural households own livestock while 52% of urban households also own animals. Of the three main livestock sub-systems in Senegal, the pastoral system in northern Senegal (a region called ‘Ferlo’) is considered the most traditional in this Sahelian environment. It occupies close to one-third of the national territory. Livestock densities in this extensive itinerant system are low: between 2 and 10.6 TLU1 per km2 (De Haan, 2016).

Like the dairy sectors of other West African countries, Senegal is confronting many changes in terms of milk production and imports. Several mini-dairies have emerged in the past 25 years; these are primarily located in rural areas and supported by NGOs. Farming practices are changing: use of crop by-products as feed; breeding of crossbred animals and forage crops; settling of animals; new suburban intensive dairy farms.

These changes have increased milk productivity and sales. However, this emerging local dairy development cannot compensate for Senegal’s increasing imports of milk powder. In 2010 nearly 60% of the country’s total demand for milk of 421 million litres was covered by imports, which represented a value of USD 166.2 million (Duteurtre and Corniaux, 2013). The country has become structurally dependent on foreign markets for milk. Although milk powder imports allow urban populations to access cheap dairy products and the dairy processing sector to grow, they also compete with local milk production. Although local milk is more expensive to source because of diseconomies of scale, dairies are showing a growing interest in supplying local milk because it allows them to produce dairy products more closely aligned to local consumers’ tastes, and thus achieve higher added value (Duteurtre and Corniaux, 2013).

It is in this context that the Laiterie Du Berger (LDB) was created in 2006 as a modern dairy plant collecting milk in pastoral areas of Northern Senegal. Its largest challenge has been to address the seasonality of rains, and thus fodder, the determining factor for milk production in the Ferlo. The business increased quickly from 200 farmer suppliers at the beginning to more than 800 suppliers in 2010; the volume of milk collected has more than quadrupled (Parisse, 2012). The development of modern retailing and agro-industries in developing countries has had an important impact on the livelihoods of the smallholder farmers who supply large-scale enterprises like the LDB (Reardon et al., 2003). Setting up agri-food value chains that are inclusive of smallholder farmers requires changes in business models from the buyer but also major transformations of the farm management models and livelihoods by suppliers (Vorley et al., 2009). Can the LDB be considered as a socially motivated enterprise? The LDB website brands the company as an agribusiness firm that is developing strong corporate social responsibility by helping its pastoralist suppliers while continuing to respect sound financial standards in a competitive dairy market2.

Consequently, this article aims to assess the contribution of the LDB and its modern supply chain management to strengthening the food security and socioeconomic resources of pastoral dairy households in the Ferlo.

Although there are already many theoretical and empirical contributions on the topic of value chains inclusive of smallholder farmers in developing countries, their study scope is relatively wide (crops, forestry, fair trade, governance and food safety) and none have specifically covered the livestock sub-sector and more particularly, Sub-Saharan pastoral systems. The main contribution of this article to the agribusiness literature is to highlight the consequences of modern dairy supply practices introduced by LDB on the livelihood choices, food security and market orientation of pastoral milk suppliers. Section 2 of this paper describes the local context and organization of the LDB dairy plant’s supply chain. Section 3 discusses the conceptual framework, data and method to analyze the impacts of the LDB’s supply chain on its suppliers. Section 4 1 TLU (Tropical Livestock Unit) aggregates different livestock species, based on 250 kg live weight: 1 TLU is equivalent to 1 camel, 0.7 TLU is 1 cow, 0.1 TLU equals 1 sheep or goat, and 0.01 TLU represents 1 chicken.

2 https://lalaiterieduberger.wordpress.com.

International Food and Agribusiness Management Review

554

Wane et al.

Volume 20, Issue 4, 2017

presents the main findings from this research and Section 5 concludes by considering the research gaps, proposing future research directions, and suggesting recommendations for agribusiness development.

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0218 - Tuesday, August 22, 2017 12:33:35 PM - IP Address:24.21.169.207

2. Context of the Laiterie du Bergerâ&#x20AC;&#x2122;s dairy business Pastoralism in the African Sahel is a production system and a livelihood strategy confronted with risks, uncertainties and opportunities. This situation is also valid in the Senegalese Ferlo. Pastoralists and agropastoralists combine market and non-market inputs to produce livestock products. These individuals also diversify their livelihoods by producing crops. The household productions are consumed within the household, sold or stocked. This economic activity occurs in a context of various changes, which impact actors in isolation or simultaneously, sequentially or occasionally. The major shock element comes from extreme weather changes with high variability of annual rainfall and temperatures. The herders attempt to address these spatiotemporal variations through mobility, leading their herds to areas where there is still grass. This itinerant livelihood remains the main strategy in their uncertain environment. In addition, herders are increasingly subjected to other shocks: price volatility of food and animal feed at national and international levels (Wane et al., 2009, 2014); diseases due to vaccination programs barely achieving the 80% coverage recommended by the World Organization for Animal Health and Animal Diseases (Kaboret, 2010); uncontrolled human and animal demography (TourĂŠ et al., 2013); and social transformations. All of these shocks make pastoral economic activities and livelihoods more vulnerable and jeopardize the ability of these marginalized populations to be resilient.

As the main source of food in the Sahel, livestock contributed an average of 38% to agricultural Gross domestic product in the 2000s (Ly et al., 2010). The increasing demand for meat and milk in West Africa is seen as an opportunity if various stakeholders can collectively develop the resilience of this system (Ickowicz et al., 2012). However, efforts to achieve food security are strongly constrained by socioeconomic factors such as poverty, low productivity, unfair marketing relationships, human and animal demographics, lack of institutions and infrastructure. Negative biophysical trends such as climate variability or pressure on natural resources are further constraints to the sustainability of the systemâ&#x20AC;&#x2122;s food security.

The LDB began operating a private dairy plant in Richard Toll City, in northern Senegal (Figure 1), to collect and add value to milk from local herders and help to meet the increasing demand for milk products in the country. In the context of Sahelian pastoral systems, setting up a supply chain for a modern dairy plant is far from impact-neutral. This impact is materialized by changes towards more efficient dairy production and

Pasture during wet season (September 2010)

Seasonality Pasture during dry season (February 2011)

Figure 1. Location of the Laiterie du Berger in Senegal in a context of inter-annual climate variability (adapted from Parisse, 2012).

International Food and Agribusiness Management Review

555

Wane et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0218 - Tuesday, August 22, 2017 12:33:35 PM - IP Address:24.21.169.207

sourcing. In particular, the procurement process has to be thought out carefully because milk production by traditional pastoral herders is not market-driven.

The LDB collects milk from pastoralist campsites located in the arid wilderness around the city. To limit the transportation time of fresh milk on dirt roads, the dairy has encouraged milk producers to become partially sedentary. The main non-written contractual link between the LDB and its pastoralist milk suppliers is developed around a package of transactions on milk production with collaterals provided by the LDB to secure its milk supply. In this context of market and environmental uncertainties (Wane et al., 2014), the main innovations generated by the LDB consist of (1) settling dairy farmers within a 50 km radius of the dairy; (2) organizing six daily milk collection routes; (3) providing animal feed through a check-off recovered on future milk sales; and (4) providing technical support through development partners to farmers on milking hygiene, dairy herd nutrition, veterinary advice, protected areas for grazing and water wells.

3. Conceptual frameworks, data and quantitative research methodology 3.1. Conceptual frameworks ■■The generic value chain model helps characterize the LDB’s supply chain innovations From the perspective of the LDB, the challenges posed by the creation of a dedicated supply chain to source local milk from pastoral herders can be envisaged through the model of a firm’s generic value chain (Porter, 1985: 37). A company applying this generic value chain model to improve its inbound logistics needs to reinforce the support activities that will allow its staff to deal with supply challenges and help its suppliers deliver the raw materials the firm needs in sufficient quantity, and to an appropriate quality standard. These support activities encompass procurement (finding suppliers and organizing the supply chains), technology development (innovating in production, information and management processes to remain competitive), human resources management (training company staff and suppliers to put innovations into practice to reach the supply objective), and firm infrastructure (the company’s support systems that allow it to run and pay its staff and suppliers). The results section reviews the LDB’s innovations in dairy supply chain management along the lines of this generic value chain model for the primary activity of inbound logistics.

■■A more holistic model is needed to understand the contribution of the LDB’s activities to the sustainability of the wider pastoral system However, this study of the LDB’s contribution to restructure complex traditional pastoral systems also needs to address whether the dairy’s innovations are impacting on individual supplier households and the sustainability of the traditional pastoral system within which they live. Therefore, a complementary conceptual framework would consist in analyzing the sustainability of the dairy’s business and supply chain models. For the very particular context of Sahelian pastoral systems, Ayantunde et al. (2011) focused on West and Central Africa and invited researchers to define and delineate the scale of analysis (household, community and region) and its time horizon (mid- and long-term). Lambert-Derkimba et al. (2015) have merged various sustainability assessment approaches and, in accordance with findings by Rey-Valette et al. (2008) and Gerber et al. (2009), used the three classical pillars of sustainable development (economic, social and environmental) to integrate these within three fields of analysis of a pastoral system within its territory (Figure 2).

The first field of analysis concerns the ‘availability of resources’ in the territory and considers that the sustainability of a farming system depends on the dynamics of available resources, which enable the functionality of the production activity and enable households to survive. The second field of analysis concerns the ‘properties of the system’ and includes factors that allow access to resources as well as the potential reactions of a system to external shocks. Finally, the third field concerns ‘extended sustainability’, which considers the positive or negative impacts of pastoral systems on the components of the territory. In this study’s context of the LDB’s dairy supply chain, the processor has started a contractual relationship with International Food and Agribusiness Management Review

556

Wane et al.

Volume 20, Issue 4, 2017 External shocks

INNOVATION?

Territory

Other components of the territory

Pastoral System A (PS A)

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0218 - Tuesday, August 22, 2017 12:33:35 PM - IP Address:24.21.169.207

2. Properties of the system

PS B PS C

Impacts on territory 3. Extended sustainability

Vulnerability

Accessibility

1. Resource availability Resources for pastoral systems

Figure 2. Pastoral system in its territory (Lambert-Derkimba et al., 2015). supplier households. It is thus relevant to take the herder household as the unit of study to reveal the linkages between availability of resources and properties of the system. Elaborating further along the framework by Lambert-Derkimba et al. (2015) of pastoral systems, studying the socioeconomic sustainability of households translates into understanding how householdsâ&#x20AC;&#x2122; socioeconomic resources match their livelihood choices in terms of food security: allocation of dairy products between household consumption and market sales, purchase of food from outside using dairy income (Figure 3).

Fields of analysis

Principles Biophysical resources

Availability in the territory

Household level

Organisational resources Socioeconomic resources Relationships between farming system and society Spatio-temporal organisation

Properties of the system

Matching

Social capital Household livelihood Vulnerability

Extended sustainability

Local development Environment

Figure 3. Fields and principles of pastoral systemsâ&#x20AC;&#x2122; sustainability at household level (Lambert-Derkimba et al., 2015)

International Food and Agribusiness Management Review

557

Wane et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0218 - Tuesday, August 22, 2017 12:33:35 PM - IP Address:24.21.169.207

■■Gender roles in African pastoral households Boogaard et al. (2015) have reviewed the literature on gender roles within livestock keeping households in Africa. A household consists of diverse members with different characteristics, perspectives and influence, and who make different decisions; all these components determine the allocation of resources among household members (Haddad et al., 1997). Thus, household decisions – such as when to use livestock for home consumption, when to sell livestock and how to use the money – strongly influence the way livelihood assets are put to use within livelihood strategies. Assets are often owned by individual household members instead of being pooled, as defined by intra-household allocation rules (Haddad et al., 1997). As such, men and women within the household can own or have access rights to different assets, and assets may be unequally distributed within a household (Doss, 2013; Huss-Ashmore, 1996; Meinzen-Dick et al., 2011).

For these reasons, women’s ownership or access rights to livestock, livestock products and their resulting income, should not be considered as given. Women’s access rights to livestock also vary with the social status of the individual: Buhl and Homewood (2000) showed how power in decision making within the household changed over time for women according to their age and status in Fulani herder families. Younger women, second and third wives or daughters had less freedom in decision making over assets than older women, first wives and mothers in law.

■■Research questions to be answered and overall research methodology In light of this article’s objective to assess the contribution of the LDB’s milk supply chain to strengthening the livelihood and socioeconomic resources of pastoral smallholders in the Senegalese Sahel, one can use both Porter’s (1985) generic value chain model to understand how the LDB’s milk supply chain organization contributes to adding value for the firm while reviewing the same supply chain arrangements through the lens of the conceptual framework for pastoral systems by Lambert-Derkimba et al. (2015) to identify changes to supplier households’ livelihoods and socioeconomic resources. This article therefore attempts to answer the following research questions: Q1: do the LDB’s innovative milk supply chains add value to the company’s products?

Q2: do the LDB’s innovative milk supply chains contribute to improve the livelihoods of pastoral herder households?

Q3: do the LDB’s innovative milk supply chains help build up the socioeconomic resources needed by the herder households to sustain their pastoral system?

In a Sub-Saharan African context, the implementation of modern processing plants in the agricultural sector has generally been viewed as an innovation similar to technology introduction and has been empirically studied in terms of adoption in accordance with Griliches’ (1957) seminal economic perspective. Subsequently, more rigorous approaches based on innovative statistical tools have used regression models following a logistic law (LOGIT model) or a Gaussian law (probit model), which provide similar results in experiments involving with-and-without group comparisons (Negatu and Parikh, 1999). However, these models are criticized for their lack of discernment of the adoption failures due to technology or innovation availability or access problems, particularly in countries facing gaps in technology and innovation dissemination (Mulubrhan et al., 2012).

In this paper, we use these classical impact assessment methodologies in a very broad sustainability framework to reflect the complexity of Sahelian pastoral systems characterized by the strong interaction between production, social and cultural aspects. Thus, the model proposed by Lambert-Derkimba et al. (2015) (Figure 3) is used as a starting point to define the successive steps of analysis for the contribution of the LDB’s supply chain practices on strengthening the sustainability of the Ferlo’s pastoral system measured at the level of herder households. To implement this framework, we analyze the changes undertaken by

International Food and Agribusiness Management Review

558

Wane et al.

Volume 20, Issue 4, 2017

the dairy’s pastoralist suppliers on their milk production practices, the milk production destination and the diversification of income sources between dry and rainy season in order to supply raw milk to the LDB. 3.2. Data sources

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0218 - Tuesday, August 22, 2017 12:33:35 PM - IP Address:24.21.169.207

To answer the research questions stated above, both qualitative and quantitative data were used.

The viewpoints of the LDB were collected through in-depth interviews of its CEO and of the general manager of the Richard Toll processing plant, who was in charge of the raw milk supply chain at the time of field research. These two in-depth interviews were conducted in May 2014. They were meant to gather information on the business strategy of the LDB, the managers’ viewpoints on the organization of its raw milk supply chain, and their assessments on their supply chain’s and overall business performance.

Additional primary data on the social aspects of sustainability were collected through individual qualitative surveys of 70 milk suppliers to LDB from January 2013 to January 2014. The supplier surveys were conducted on the Rosso and the Mouda milk delivery routes established by the LDB; both these routes were among the first to be part of the LDB’s raw milk supply chain. The sample was divided equally: 35 suppliers interviewed on the Rosso route and 35 on the Mouda route. Suppliers to be interviewed were chosen by simple random selection among the list of suppliers along both routes. Suppliers comprised both herder households and cooperatives supplying raw milk to the LDB. Respondents were interviewed in the Fulani language, which is spoken by two of the authors. The qualitative surveys were meant to gather information on the suppliers’ pastoral practices, their relationship with the LDB (in particular, access to LDB services and technical assistance), and their perception of their level of income (for households).

Two focus group discussions were also organized with women from pastoral households involved in this supply chain in May 2014. One focus group was held with ten women producing milk in a fixed settlement called Niassanté of the Dièri region. The second focus group was held with seven women and two men producing milk in Ngoudompe village in the Walo region, which is located in an irrigated perimeter closer to Richard Toll City. Both focus groups comprised individuals supplying milk to the LDB and others who had never supplied or who had stopped supplying milk to the dairy. Because all the researchers conducting the interviews were men, the two focus group discussions were organized through the head of the villages who gathered participants according to the researchers’ sample requirements (mainly women, all types of social status, suppliers and non-suppliers to the dairy). In both cases, the focus groups were held in the presence of men related to the women being interviewed. The discussions nonetheless allowed all participants to engage by prompting the shier women in the groups after the men and the older women had expressed themselves.

These group discussions were held in the local language, which is spoken by one of the authors. The focus group discussions encouraged participants to discuss their cattle herding practices, decisions concerning milk allocation for household consumption or for sale, milk marketing conditions, the relationship with LDB milk collectors and technical staff, and the household decisions on the use of the money from milk sales.

The quantitative data gathered for this research came from existing databases collected from 445 households of LDB milk suppliers by a research consortium grouping IFPRI, CIRAD and GRET (Bernard et al., 2015).

The households surveyed were self-selected as the volunteer participants to a supply contract research experiment linking regular supply of 0.5 L of raw milk per cow per day to the LDB over five days of the week in exchange for free access at milk collection points to an iron-fortified milk product targeted to the supplier’s children aged 2 to 5. The consortium undertook two visits of the same volunteer supplier households in January 2013 and January 2014 to capture the evolution of variables over the calendar year. Out of the 445 households surveyed in 2013, 437 repeated the survey the following year. The questionnaire collected information on the household’s wealth, demographics and milk production. Mothers in the households were interviewed on their child feeding practices, living conditions and their individual milk production enterprise. Additional information linked to these households was gained through milk container level data

International Food and Agribusiness Management Review

559

Wane et al.

Volume 20, Issue 4, 2017

from the LDB supply chain: level of relationship between herders and LDB, milk quantities sold, resulting milk income and whether herders belonged to milk cooperatives (Table 1).

3.3. Quantitative research methodology

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0218 - Tuesday, August 22, 2017 12:33:35 PM - IP Address:24.21.169.207

■■Food security index Access to food through dairy income generation constitutes a central parameter of our analysis. A food security index (FSI) was tabulated as one indicator of the livelihood of producer households according to LambertDerkimba et al. (2015)’s sustainability conceptual framework adapted to Sahelian pastoral systems. Our FSI is based on the Household Food Insecurity Assessment Scale defined by USAID from recommendations by FAO. IFPRI has also adopted this approach in its ongoing research on nutrition aspects in the Ferlo (Coates et al., 2007). The FSI was developed using a Multiple Correspondence Analysis from 18 questions related to Table 1. Descriptive statistics for the households included in the IFPRI-CIRAD-GRET nutrition database (Bernard et al., 2015).

Values Container level data Female container head Number of children on contract Number of cows listed in contract Collective container Milk production from December 9, 2012 (pre-study) Total weekly-milk delivered to LDB1 (liters) Container delivered at least once in the past week (%) # of days delivered milk in the past week Household level data Female household head Age of household head Household head has any schooling Household size Number of children 0-5 years Owns or manages land =1 if HH2 member is responsible for milk container =1 if HH member fills other milk containers Total number of milk containers HH is responsible for or fills Number of lactating cows Number of cows that were milked yesterday Liters of milk collected yesterday Liters of milk collected in a typical day (dry) Liters of milk collected in a typical day (rainy) % of income from – Milk (dry) % of income from – Milk (rainy) % of milk sold to LDB (dry) % of milk sold to LDB (rainy) % of milk sold to local market (dry) % of milk sold to local market (rainy) Number of years affiliated with LDB 1 2

0.24 4.05 3.77 0.13

381 385 385 385

22.59 0.96 6.29

385 385 385

0.19 49 0.04 8.73 1.99 0.51 0.83 0.16 1.05 6.53 6.38 5.96 4.17 12.69 25.43 55.91 55.72 64.03 3.55 3.8 4.75

437 436 437 437 437 436 437 437 437 436 435 431 435 435 433 433 407 434 407 434 437

LDB = Laiterie Du Berger.

HH = Household.

International Food and Agribusiness Management Review

560

n

Wane et al.

Volume 20, Issue 4, 2017

food security including, for example, eliminating certain types of food from the household diet, a reduction in the number of meals and a reduction of the quantities consumed. Household groups were characterized according to their degree of food insecurity through an Ascending Hierarchical Classification.

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0218 - Tuesday, August 22, 2017 12:33:35 PM - IP Address:24.21.169.207

If we consider Q qualitative variables chosen for the index, let us define: đ?&#x2018;&#x2039;đ?&#x2018;&#x2039;đ?&#x2018;&#x2013;đ?&#x2018;&#x2013; (đ?&#x2018;&#x2014;đ?&#x2018;&#x2014;đ?&#x2018;&#x2014;đ?&#x2018;&#x2014;đ?&#x2018;&#x2014;đ?&#x2018;&#x2014;) = {

1 đ?&#x2018;&#x2013;đ?&#x2018;&#x2013;đ?&#x2018;&#x2013;đ?&#x2018;&#x2013; đ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;&#x17D;đ?&#x2018;&#x17D;đ?&#x2018;&#x17D;đ?&#x2018;&#x17D; đ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;&#x2122;đ?&#x2018;&#x2122;đ?&#x2018;&#x2122;đ?&#x2018;&#x2122;đ?&#x2018;&#x2122;đ?&#x2018;&#x2122;đ?&#x2018;&#x2122;đ?&#x2018;&#x2122;đ?&#x2018;&#x2122;đ?&#x2018;&#x2122;đ?&#x2018;&#x2122;đ?&#x2018;&#x2122;đ?&#x2018;&#x2122;đ?&#x2018;&#x2122;đ?&#x2018;&#x153;đ?&#x2018;&#x153;đ?&#x2018;&#x153;đ?&#x2018;&#x153; đ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Łđ?&#x2018;Łđ?&#x2018;Łđ?&#x2018;Łđ?&#x2018;Łđ?&#x2018;Łđ?&#x2018;Łđ?&#x2018;Łđ?&#x2018;Łđ?&#x2018;Łđ?&#x2018;Łđ?&#x2018;Łđ?&#x2018;Łđ?&#x2018;Łđ?&#x2018;Łđ?&#x2018;Łđ?&#x2018;Łđ?&#x2018;Łđ?&#x2018;Ł (1) 0 đ?&#x2018;&#x2013;đ?&#x2018;&#x2013;đ?&#x2018;&#x2013;đ?&#x2018;&#x2013; đ?&#x2018;&#x203A;đ?&#x2018;&#x203A;đ?&#x2018;&#x203A;đ?&#x2018;&#x203A;đ?&#x2018;&#x203A;đ?&#x2018;&#x203A;

and W (j,q) is the weight of level j for variable q

The FSI is defined, for a household i, as follows: đ?&#x2018;&#x203A;đ?&#x2018;&#x203A;

1 đ??šđ??šđ??šđ??šđ??šđ??šđ?&#x2018;&#x2013;đ?&#x2018;&#x2013; = â&#x2C6;&#x2018; â&#x2C6;&#x2018; đ?&#x2018;&#x160;đ?&#x2018;&#x160;(đ?&#x2018;&#x2014;đ?&#x2018;&#x2014;, đ?&#x2018;&#x17E;đ?&#x2018;&#x17E;)đ?&#x2018;&#x2039;đ?&#x2018;&#x2039;đ?&#x2018;&#x2013;đ?&#x2018;&#x2013; (đ?&#x2018;&#x2014;đ?&#x2018;&#x2014;, đ?&#x2018;&#x17E;đ?&#x2018;&#x17E;) (2) đ?&#x2018;&#x201E;đ?&#x2018;&#x201E; đ?&#x2018;&#x17E;đ?&#x2018;&#x17E;=1 đ?&#x2018;&#x2014;đ?&#x2018;&#x2014; â&#x2C6;&#x2C6; đ??˝đ??˝đ?&#x2018;&#x17E;đ?&#x2018;&#x17E;

Where Jq is the number of levels for variable q.

We computed a standardized index to facilitate interpretation: đ??šđ??šđ??šđ??šđ??šđ??š â&#x2C6;&#x2019;min(đ??šđ??šđ??šđ??šđ??šđ??š )

đ?&#x2018;&#x2013;đ?&#x2018;&#x2013; đ?&#x2018;&#x2013;đ?&#x2018;&#x2013; đ??šđ??šđ??šđ??šđ??šđ??šđ?&#x2018;&#x2013;đ?&#x2018;&#x2013;â&#x2C6;&#x2014; = max(đ??šđ??šđ??šđ??šđ??šđ??š )â&#x2C6;&#x2019; min(đ??šđ??šđ??šđ??šđ??šđ??š ) đ?&#x2018;&#x2013;đ?&#x2018;&#x2013;

đ?&#x2018;&#x2013;đ?&#x2018;&#x2013;

(3)

Knowing that a suitable index must respect a hierarchy, we ensured that the First Axis Ordinal Consistency was well reflected. This result helped us define the weight of each component of the index. To do so, let G1 (j,q) be the coordinates of level j for variable q on the first axis and Îť1 the eigenvalue. The weight of the index is then defined as follows: 4) đ?&#x2018;&#x160;đ?&#x2018;&#x160;(đ?&#x2018;&#x2014;đ?&#x2018;&#x2014;đ?&#x2018;&#x2014; đ?&#x2018;&#x2014;đ?&#x2018;&#x2014;) =

đ??şđ??ş1 (đ?&#x2018;&#x2014;đ?&#x2018;&#x2014;đ?&#x2018;&#x2014;đ?&#x2018;&#x2014;đ?&#x2018;&#x2014;đ?&#x2018;&#x2014;

(4)

â&#x2C6;&#x161;đ?&#x153;&#x2020;đ?&#x153;&#x2020;1

Introducing this term into Equation 2, the FSI becomes: đ?&#x2018;&#x203A;đ?&#x2018;&#x203A;

1 đ??şđ??ş1 (đ?&#x2018;&#x2014;đ?&#x2018;&#x2014;, đ?&#x2018;&#x17E;đ?&#x2018;&#x17E;) đ??šđ??šđ??šđ??šđ??šđ??šđ?&#x2018;&#x2013;đ?&#x2018;&#x2013; = â&#x2C6;&#x2018; â&#x2C6;&#x2018; đ?&#x2018;&#x2039;đ?&#x2018;&#x2039;đ?&#x2018;&#x2013;đ?&#x2018;&#x2013; (đ?&#x2018;&#x2014;đ?&#x2018;&#x2014;, đ?&#x2018;&#x17E;đ?&#x2018;&#x17E;) đ?&#x2018;&#x201E;đ?&#x2018;&#x201E; â&#x2C6;&#x161;đ?&#x153;&#x2020;đ?&#x153;&#x2020; 1 đ?&#x2018;&#x17E;đ?&#x2018;&#x17E;=1 đ?&#x2018;&#x2014;đ?&#x2018;&#x2014; â&#x2C6;&#x2C6; đ??˝đ??˝

(5)

đ?&#x2018;&#x17E;đ?&#x2018;&#x17E;

To evaluate the index, we acted in accordance with Kiâ&#x20AC;&#x2122;s (2005) approach, which consists in defining classes for the index and comparing the distribution of the variables throughout those classes. For instance, if we consider a privative variable, its degree should be reduced if we move from one quartile of the index to another. In our case, this resulted in four household groups of food security, which were labeled as follows: Group 1:â&#x20AC;&#x2122;insecureâ&#x20AC;&#x2122; because households of this group suffered food insecurity, Group 2:â&#x20AC;&#x2122;poorly secureâ&#x20AC;&#x2122; because they occasionally suffered food insecurity, Group 3:â&#x20AC;&#x2122;secureâ&#x20AC;&#x2122; because they rarely suffered food insecurity and Group 4:â&#x20AC;&#x2122;highly secureâ&#x20AC;&#x2122; because they never suffered food insecurity. The distribution of households interviewed across the FSI can then be calculated (Table 2).

Table 2. Statistical information on the food security index, calculated from IFPRI-CIRAD-GRET data base on nutrition of 445 Laiterie Du Berger suppliers.

Groups

Class size

Proportion (%) Mean

Standard deviation Minimum Maximum

Group 1: â&#x20AC;&#x2DC;insecureâ&#x20AC;&#x2122; Group 2: â&#x20AC;&#x2DC;poorly secureâ&#x20AC;&#x2122; Group 3: â&#x20AC;&#x2DC;secureâ&#x20AC;&#x2122; Group 4: â&#x20AC;&#x2DC;highly secureâ&#x20AC;&#x2122; Total

121 138 108 78 445

27 31 24 18 100

0.08 0.05 0.07 0.08 0.27

0.16 0.38 0.61 0.92 0.47

International Food and Agribusiness Management Review

561

0.00 0.27 0.50 0.77 0.00

0.27 0.49 0.76 1.00 1.00

Wane et al.

Volume 20, Issue 4, 2017

â&#x2013; â&#x2013; Calculating the probability of changing food security status

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0218 - Tuesday, August 22, 2017 12:33:35 PM - IP Address:24.21.169.207

The analysis of milk income was performed using a First-order Markov Chain; this was supported by income mobility indices to highlight the links between dairy income stability and food security. A Markov Chain is a finite states process. A Markov Chain is also a stochastic process with a limited memory; its state at time t depends on its state at time tâ&#x20AC;&#x201C;1. This property can be translated into the following equation: đ?&#x2018;&#x192;đ?&#x2018;&#x192;(đ?&#x2018;&#x2039;đ?&#x2018;&#x2039;đ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ą= đ?&#x2018;&#x2014;đ?&#x2018;&#x2014;|đ?&#x2018;&#x2039;đ?&#x2018;&#x2039;đ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ą= đ?&#x2018;&#x2013;đ?&#x2018;&#x2013;đ?&#x2018;Ąđ?&#x2018;Ą , đ?&#x2018;&#x2039;đ?&#x2018;&#x2039;đ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ą= đ?&#x2018;&#x2013;đ?&#x2018;&#x2013;đ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ą, â&#x20AC;Ś , đ?&#x2018;&#x2039;đ?&#x2018;&#x2039;0 = đ?&#x2018;&#x2013;đ?&#x2018;&#x2013;0) = đ?&#x2018;&#x192;đ?&#x2018;&#x192;(đ?&#x2018;&#x2039;đ?&#x2018;&#x2039;đ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ą= đ?&#x2018;&#x2014;đ?&#x2018;&#x2014;|đ?&#x2018;&#x2039;đ?&#x2018;&#x2039;đ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ą= đ?&#x2018;&#x2013;đ?&#x2018;&#x2013;đ?&#x2018;Ąđ?&#x2018;Ą ) = đ?&#x2018;?đ?&#x2018;?đ?&#x2018;&#x2013;đ?&#x2018;&#x2013;đ?&#x2018;&#x2014;đ?&#x2018;&#x2014; (đ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ą

(6)

Where pij(t) is the transition probability from state i to state j at time t. The transition probabilities define the transition matrix, which has the following properties: đ?&#x2018;?đ?&#x2018;?đ?&#x2018;&#x2013;đ?&#x2018;&#x2013;đ?&#x2018;&#x2013;đ?&#x2018;&#x2013; (đ?&#x2018;Ąđ?&#x2018;Ą) â&#x2030;Ľ 0 đ?&#x2018;&#x201C;đ?&#x2018;&#x201C;đ?&#x2018;&#x201C;đ?&#x2018;&#x201C;đ?&#x2018;&#x201C;đ?&#x2018;&#x201C;đ?&#x2018;&#x201C;đ?&#x2018;&#x17D;đ?&#x2018;&#x17D;đ?&#x2018;&#x17D;đ?&#x2018;&#x17D;đ?&#x2018;&#x17D;đ?&#x2018;&#x17D;đ?&#x2018;&#x17D; (đ?&#x2018;&#x2013;đ?&#x2018;&#x2013;đ?&#x2018;&#x2013; đ?&#x2018;&#x2013;đ?&#x2018;&#x2013;đ?&#x2018;&#x2013;

(7)

â&#x2C6;&#x2018; đ?&#x2018;?đ?&#x2018;?đ?&#x2018;&#x2013;đ?&#x2018;&#x2013;đ?&#x2018;&#x2013;đ?&#x2018;&#x2013; (đ?&#x2018;Ąđ?&#x2018;Ą) = 1 đ?&#x2018;&#x201C;đ?&#x2018;&#x201C;đ?&#x2018;&#x201C;đ?&#x2018;&#x201C;đ?&#x2018;&#x201C;đ?&#x2018;&#x201C;đ?&#x2018;&#x201C;đ?&#x2018;&#x17D;đ?&#x2018;&#x17D;đ?&#x2018;&#x17D;đ?&#x2018;&#x17D;đ?&#x2018;&#x17D;đ?&#x2018;&#x17D;đ?&#x2018;&#x17D; đ?&#x2018;&#x2013;đ?&#x2018;&#x2013; đ?&#x2018;&#x2014;đ?&#x2018;&#x2014;

In this study, we consider a homogenous Markov Chain, i.e.: (8)

đ?&#x2018;&#x192;đ?&#x2018;&#x192;(đ?&#x2018;Ąđ?&#x2018;Ą) = đ?&#x2018;&#x192;đ?&#x2018;&#x192;đ?&#x2018;&#x192;đ?&#x2018;&#x192;đ?&#x2018;&#x192;đ?&#x2018;&#x192;đ?&#x2018;&#x192;đ?&#x2018;&#x192;đ?&#x2018;&#x192;đ?&#x2018;&#x192;đ?&#x2018;&#x192;đ?&#x2018;&#x192;đ?&#x2018;&#x192;đ?&#x2018;&#x192;đ?&#x2018;&#x192;đ?&#x2018;&#x192;đ?&#x2018;&#x192;đ?&#x2018;&#x192;đ?&#x2018;&#x192;đ?&#x2018;&#x192;

We are interested in knowing the transition situation of households between the wet season and dry season. In this case, we consider the total income of households divided in four classes or states. The income mobility indices were calculated using the transition matrix or Markov chain from the variable â&#x20AC;&#x2DC;dairy incomeâ&#x20AC;&#x2122;, which was divided into four quartiles that represented the four groups of households previously defined. Then, we determined the transition probabilities from one income group to another between the rainy season and the dry season. We calculated a Shorrock index (Îź1nor) to indicate whether the households are mobile in terms of income. Therefore, a certain hierarchy is considered between the states. Our approach is based upon the fact that moving from class 1 to class 2 between the seasons is a relative improvement in terms of income, whereas moving from class 2 to class 1 is a relative degradation. The movements in the matrix are synthesized by mobility indices. The Shorrock index calculates the overall mobility in the Chain: đ?&#x153;&#x2021;đ?&#x153;&#x2021;1 =

1 â&#x2C6;&#x2018;(1 â&#x2C6;&#x2019; đ?&#x2018;?đ?&#x2018;?đ?&#x2018;&#x2014;đ?&#x2018;&#x2014;đ?&#x2018;&#x2014;đ?&#x2018;&#x2014; ) đ?&#x2018;&#x203A;đ?&#x2018;&#x203A; đ?&#x2018;&#x203A; đ?&#x2018;&#x203A;

(9)

đ?&#x2018;&#x2014;đ?&#x2018;&#x2014;

The standardized Shorrock index is given by the formula: đ?&#x153;&#x2021;đ?&#x153;&#x2021;1đ?&#x2018;&#x203A;đ?&#x2018;&#x203A;đ?&#x2018;&#x203A;đ?&#x2018;&#x203A;đ?&#x2018;&#x203A;đ?&#x2018;&#x203A; = 1 â&#x2C6;&#x2019;

đ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ą(đ?&#x2018;&#x192;đ?&#x2018;&#x192;đ?&#x2018;&#x192; đ?&#x2018;&#x203A;đ?&#x2018;&#x203A;

(10)

Where tr(P) represents the trace of the transition matrix P. We then estimated the householdsâ&#x20AC;&#x2122; income improvement or degradation through adequate indicators (Îźimp and Îźdeg, respectively) and analyzed the direction of change of the income mobility indices.

The improvement index is given by: đ?&#x153;&#x2021;đ?&#x153;&#x2021;đ?&#x2018;&#x2013;đ?&#x2018;&#x2013;đ?&#x2018;&#x2013;đ?&#x2018;&#x2013;đ?&#x2018;&#x2013;đ?&#x2018;&#x2013;

đ?&#x2018;&#x203A;đ?&#x2018;&#x203A;đ?&#x2018;&#x203A;đ?&#x2018;&#x203A;

đ?&#x2018;&#x203A;đ?&#x2018;&#x203A;

1 = â&#x2C6;&#x2018; â&#x2C6;&#x2018; đ?&#x2018;?đ?&#x2018;?đ?&#x2018;&#x2013;đ?&#x2018;&#x2013;đ?&#x2018;&#x2013;đ?&#x2018;&#x2013; đ?&#x2018;&#x203A;đ?&#x2018;&#x203A; đ?&#x2018;&#x203A; đ?&#x2018;&#x203A;

(11)

đ?&#x2018;&#x2013;đ?&#x2018;&#x2013;đ?&#x2018;&#x2013;đ?&#x2018;&#x2013; đ?&#x2018;&#x2014;đ?&#x2018;&#x2014;đ?&#x2018;&#x2014;đ?&#x2018;&#x2014;đ?&#x2018;&#x2014;đ?&#x2018;&#x2014;đ?&#x2018;&#x2014;

The degradation index is defined by the formula: đ?&#x153;&#x2021;đ?&#x153;&#x2021;đ?&#x2018;&#x2018;đ?&#x2018;&#x2018;đ?&#x2018;&#x2018;đ?&#x2018;&#x2018;đ?&#x2018;&#x2018;đ?&#x2018;&#x2018;

đ?&#x2018;&#x203A;đ?&#x2018;&#x203A;đ?&#x2018;&#x203A;đ?&#x2018;&#x203A;

đ?&#x2018;&#x203A;đ?&#x2018;&#x203A;

1 = â&#x2C6;&#x2018; â&#x2C6;&#x2018; đ?&#x2018;?đ?&#x2018;?đ?&#x2018;&#x2013;đ?&#x2018;&#x2013;đ?&#x2018;&#x2013;đ?&#x2018;&#x2013; đ?&#x2018;&#x203A;đ?&#x2018;&#x203A; đ?&#x2018;&#x203A; đ?&#x2018;&#x203A;

(12)

đ?&#x2018;&#x2014;đ?&#x2018;&#x2014;đ?&#x2018;&#x2014;đ?&#x2018;&#x2014; đ?&#x2018;&#x2013;đ?&#x2018;&#x2013;đ?&#x2018;&#x2013;đ?&#x2018;&#x2013;đ?&#x2018;&#x2013;đ?&#x2018;&#x2013;đ?&#x2018;&#x2013;

International Food and Agribusiness Management Review

562

Wane et al.

Volume 20, Issue 4, 2017

â&#x2013; â&#x2013; Regression model for the determinants of food security

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0218 - Tuesday, August 22, 2017 12:33:35 PM - IP Address:24.21.169.207

The determinants of food security were identified using an ordered probit multinomial regression model.

The idea behind this approach is to know how the income mobility movements (transition matrix) and other household characteristics impact on the food security of the households between seasons.

The Ordered Multinomial approach is developed on usual regression techniques to explain a variable of interest by other variables. In this case, the variable to be explained is the food security index: a qualitative variable with more than two levels and a natural order between them. Assume that Y is a qualitative variable with m+1 levels. The model is defined by: 0 đ?&#x2018;&#x2013;đ?&#x2018;&#x2013;đ?&#x2018;&#x2013;đ?&#x2018;&#x2013; đ?&#x2018;&#x152;đ?&#x2018;&#x152;đ?&#x2018;&#x2013;đ?&#x2018;&#x2013;â&#x2C6;&#x2014; â&#x2030;¤ đ?&#x2018;?đ?&#x2018;?1 1 đ?&#x2018;&#x2013;đ?&#x2018;&#x2013;đ?&#x2018;&#x2013;đ?&#x2018;&#x2013; đ?&#x2018;?đ?&#x2018;?1 â&#x2030;¤ đ?&#x2018;&#x152;đ?&#x2018;&#x152;đ?&#x2018;&#x2013;đ?&#x2018;&#x2013;â&#x2C6;&#x2014; â&#x2030;¤ đ?&#x2018;?đ?&#x2018;?2 đ?&#x2018;&#x152;đ?&#x2018;&#x152;đ?&#x2018;&#x2013;đ?&#x2018;&#x2013;đ?&#x2018;&#x2013;= { â&#x20AC;Śâ&#x20AC;Śâ&#x20AC;Ś đ?&#x2018;&#x161;đ?&#x2018;&#x161;đ?&#x2018;&#x161; đ?&#x2018;&#x2013;đ?&#x2018;&#x2013;đ?&#x2018;&#x2013;đ?&#x2018;&#x2013; đ?&#x2018;&#x152;đ?&#x2018;&#x152;đ?&#x2018;&#x2013;đ?&#x2018;&#x2013;â&#x2C6;&#x2014; â&#x2030;Ľ đ?&#x2018;?đ?&#x2018;?đ?&#x2018;&#x161;đ?&#x2018;&#x161;

(13)

With cj+1 â&#x2030;Ľ cj and:

đ?&#x2018;&#x152;đ?&#x2018;&#x152;đ?&#x2018;&#x2013;đ?&#x2018;&#x2013;â&#x2C6;&#x2014; = đ?&#x2018;&#x2039;đ?&#x2018;&#x2039;đ?&#x2018;&#x2013;đ?&#x2018;&#x2013; đ?&#x203A;˝đ?&#x203A;˝ đ?&#x203A;˝ đ?&#x203A;˝đ?&#x203A;˝đ?&#x2018;&#x2013;đ?&#x2018;&#x2013; đ?&#x153;&#x20AC;đ?&#x153;&#x20AC;đ?&#x2018;&#x2013;đ?&#x2018;&#x2013; ~ (đ?&#x2018;&#x2013;đ?&#x2018;&#x2013;đ?&#x2018;&#x2013;đ?&#x2018;&#x2013;đ?&#x2018;&#x2013;đ?&#x2018;&#x2013;) (0, đ?&#x153;&#x17D;đ?&#x153;&#x17D;đ?&#x153;&#x20AC;đ?&#x153;&#x20AC;2 )

(14) đ?&#x2018;&#x2013;đ?&#x2018;&#x2013; đ?&#x2018;&#x2013; đ?&#x2018;&#x2013;đ?&#x2018;&#x2013; đ?&#x2018;&#x2013; đ?&#x2018;&#x2013; đ?&#x2018;&#x2013;đ?&#x2018;&#x2013; Îľi could follow the logistic law (logit model) or the Gaussian law (probit model). In this study, Y represents the FSI with its four levels. We are interested in estimating the probability that an individual belongs to a definite level of the food security index: đ?&#x2018;?đ?&#x2018;?đ?&#x2018;&#x2014;đ?&#x2018;&#x2014;đ?&#x2018;&#x2014;đ?&#x2018;&#x2014; đ?&#x2018;&#x2039;đ?&#x2018;&#x2039;đ?&#x2018;&#x2013;đ?&#x2018;&#x2013; đ?&#x203A;˝đ?&#x203A;˝ đ?&#x2018;?đ?&#x2018;?đ?&#x2018;&#x2014;đ?&#x2018;&#x2014; đ?&#x2018;&#x2039;đ?&#x2018;&#x2039;đ?&#x2018;&#x2013;đ?&#x2018;&#x2013; đ?&#x203A;˝đ?&#x203A;˝ â&#x2C6;&#x2019; ) â&#x2C6;&#x2019; đ??šđ??š đ??š â&#x2C6;&#x2019; ) đ?&#x153;&#x17D;đ?&#x153;&#x17D;đ?&#x153;&#x20AC;đ?&#x153;&#x20AC; đ?&#x153;&#x17D;đ?&#x153;&#x17D;đ?&#x153;&#x20AC;đ?&#x153;&#x20AC; đ?&#x153;&#x17D;đ?&#x153;&#x17D;đ?&#x153;&#x20AC;đ?&#x153;&#x20AC; đ?&#x153;&#x17D;đ?&#x153;&#x17D;đ?&#x153;&#x20AC;đ?&#x153;&#x20AC; đ?&#x2018;&#x2014;đ?&#x2018;&#x2014; đ?&#x2018;&#x2014; đ?&#x2018;&#x2014;đ?&#x2018;&#x2014; đ?&#x2018;&#x2014; đ?&#x2018;&#x2014; đ?&#x2018;&#x2014;đ?&#x2018;&#x2014; đ?&#x2018;?đ?&#x2018;?0 = â&#x2C6;&#x2019;â&#x2C6;&#x17E; đ?&#x2018;&#x17D;đ?&#x2018;&#x17D;đ?&#x2018;&#x17D;đ?&#x2018;&#x17D;đ?&#x2018;&#x17D;đ?&#x2018;&#x17D;đ?&#x2018;&#x17D; đ?&#x2018;?đ?&#x2018;?đ?&#x2018;&#x161;đ?&#x2018;&#x161;đ?&#x2018;&#x161;đ?&#x2018;&#x161; = â&#x2C6;&#x17E;

đ?&#x2018;&#x192;đ?&#x2018;&#x192;(đ?&#x2018;&#x152;đ?&#x2018;&#x152;đ?&#x2018;&#x2013;đ?&#x2018;&#x2013; = đ?&#x2018;&#x2014;đ?&#x2018;&#x2014;) = đ??šđ??š đ??š

(15)

Where F is the distribution function of the logistic or the Gaussian law. In this case, the probit and logit models provide similar results. In this study, we chose a probit model with the Gaussian law, which is more commonly used in social science (Powers and Xie, 2000: 215).

The explanatory variables (Table 3) of the probit model were chosen based on the combination between primary data collected during our fieldwork with milk suppliers, semi-structured in-depth interviews with managers of the LDB and the IFPRI-CIRAD-GRET databases. Multicollinearity tests showed no correlation between variables used in the model.

The Markov Chain convergence test also showed that our model satisfies all required hypotheses. The interpretations are based on the marginal effects of an ordered probit and probability calculations. The main quantitative changes that we can highlight a priori from the implementation of the LDBâ&#x20AC;&#x2122;s milk supply chain management are intra-annual because we based our observations on a one-year database.

International Food and Agribusiness Management Review

563

Wane et al.

Volume 20, Issue 4, 2017

Table 3. Descriptive statistics for the independent variables used to explain herder households’ probability of being in a given class of the food security index, calculated from data from IFPRI-CIRAD-GRET database on nutrition of 445 Laiterie Du Berger suppliers.

Independent variables

Percentage of total sample

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0218 - Tuesday, August 22, 2017 12:33:35 PM - IP Address:24.21.169.207

Income mobility

Sources of income Number of years supplying milk

Number of livestock heads

Deterioration Stability Improvement 1 or 2 More than 2 1-4 4-5 5-6 6-8 1-24 24-45 45-80 ≥80

69 30 1 72 28 3 14 19 64 24 25 25 26

4. Interactions between Laiterie Du Berger’s raw milk supply chain and supplier households’ food security and socioeconomic resources 4.1. The Laiterie Du Berger’s innovative supply chains secure good quality raw milk supplies to the dairy ■■The LDB’s raw milk supply chain was purpose-built for the Senegalese Ferlo The interviews with the LDB managers revealed that, having decided that the LDB would purposefully supply raw milk from pastoralist herders of the Ferlo, the company had no choice but to develop the supply chains that would allow this goal to materialize. The LDB collects milk from pastoralist campsites located in the arid wilderness around Richard Toll City. To limit the transportation time of raw milk on dirt roads to below two hours, the dairy has encouraged milk producers to become partially sedentary. The permanent settlements harboring the herders’ dairy cows are located along six milk collection routes radiating up to 50 km away from the dairy plant.

The dairy has coopted some young men from the herder settlements and has helped them invest in motorbikes allowing them to become private milk collectors who operate the six collection routes on a daily basis. The collectors ride their motorbikes trailing a cart with plastic buckets belonging to the dairy, which contain the milk. Each bucket contains the milk of one individual herder, a household or cooperative, allowing traceability of the milk back to each individual supplier.

With the dairy striving to source more local milk for its processing plant, the LDB is actively encouraging other agro-entrepreneurs to start semi-intensive dairy farms closer to the city. This would allow the LDB to enlarge its supplier base of local producers while making sourcing easier as these peri-urban producers can deliver milk to the processing plant by themselves.

■■The LDB has introduced technological and process innovations in its relationship with its suppliers The main non-written contractual link between the LDB and its pastoralist milk suppliers is developed around a package of transactions on milk production in exchange of financial, technological and training collaterals provided by the LDB to secure its milk supply.

International Food and Agribusiness Management Review

564

Wane et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0218 - Tuesday, August 22, 2017 12:33:35 PM - IP Address:24.21.169.207

As part of its supply stabilization strategy, the LDB has developed and implemented various services to increase the milk production of pastoralists. The most innovative service was the provision of animal feed through a check-off recovered on future milk sales. Because they wanted to stabilize their dairy incomes, suppliers were motivated to change many of their production practices. Thus, 58% of LDB’s suppliers interviewed report and characterize changes in their milk production system (Figure 4).

In particular, 39% of the milk suppliers declare having increased their milk production. Moreover, 29% of suppliers put a greater focus on the quality of the milk produced than before, and 14% confirm changes in animal nutrition through the use of feed supplement. Individual qualitative interviews with herders and the focus groups have uncovered that the changes in milk productivity are predominantly related to the technical support of LDB and the income incentive from increasing milk sales to the dairy. Producers who did not supply the dairy were less likely to implement the technical innovations.

■■Human resources management and capacity development by the LDB The LDB has also invested in developing the capacities of its own staff, collectors and suppliers to put the innovations into practice. Thanks to partnerships with local and international NGOs specialized in agricultural development, the LDB’s suppliers have benefited from training on milking hygiene and dairy herd nutrition.

They have also received veterinary advice and learned how to protect areas for grazing and water wells from itinerant livestock to sustain their forage and water resources.

To reach the women who are the traditional dairy livestock keepers in these highly patriarchal pastoralist communities, the LDB managers indicated that the trainings were first delivered to the men, who would then allow the trainers’ access to the communities’ women to replicate the training in favor of those who would likely make most use of it.

39%

29%

14% 7%

Milk production Greater Change in Less milkincreasing vigilance animal nutrition processing on quality

6%

5%

Greater More cows vigilance on kept in the animal health encampment

Figure 4. Changes in production practices by milk suppliers (number of respondents implementing the non-exclusive changes and % of total sample), calculated from data from IFPRI-CIRAD-GRET database on nutrition of 445 Laiterie Du Berger suppliers.

International Food and Agribusiness Management Review

565

Wane et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0218 - Tuesday, August 22, 2017 12:33:35 PM - IP Address:24.21.169.207

■■Modifying company systems and infrastructure to adjust to local sociocultural practices The LDB has had to modify its accountancy and milk supply chain to accommodate the practices and customs of its local pastoralist suppliers. The check-off system for the animal feed has led the LDB’s supply manager and accountants to monitor both feed purchase and milk sales from each individual supplier in order to calculate their monthly negative or positive balance. The close relationships developed by the LDB’s supply manager and individual suppliers have led him to consent credit to some suppliers whose overall monthly check-off balance was negative, but who had to be seen bringing some milk income back to the household, thus allowing the male heads of households to save face back in the village.

The interview with LDB managers revealed that the dairy had even made its supply chain less efficient in order to keep good relationships with its suppliers. Indeed, the polygamous nature of households among traditional pastoralist herders resulted in several wives producing milk under the same household supplier contract. However, it transpired that each individual wife had access and control to their own animals and did not want to pool the milk from their cows with the milk from the other wives’ cows. So the male heads of households who had signed the supply contract with the dairy were asking for individual buckets for each one of their wives. To accommodate these special requests from its suppliers, the LDB was issuing many individual buckets with a capacity of 10 liters to individual women producers within the same household, thus increasing its own transaction costs to process all these containers and making the collectors travel with buckets containing only a few liters of milk. These inefficiencies were nonetheless judged a prerequisite to develop their suppliers’ trust in the LDB and encourage sales of milk.

Overall, according to the LDB managers, the supply chain arrangements, technological innovations, human resources management, and infrastructure changes implemented by the LDB seem to have contributed to increase the quantity and quality of milk supplied, thus adding to the value creation by the dairy.

4.2. The Laiterie Du Berger’s innovative milk supply chains have positive and negative effects on supplier households’ livelihoods depending on their income status In this study, the main indicators used to measure household livelihood are food security and income stabilisation between seasons.

■■Suppliers face complex choices in terms of food security practices In the traditional pastoral cattle production system of the Ferlo, only 0.5% of milk produced was sold due to a lack of viable market opportunities (Wane et al., 2009). Thus, a large portion of the milk available was intended for feeding calves, while another was used for own-consumption by pastoral households in the form of fresh and processed milk (butter and curdled milk). The appearance of the LDB has changed the milk use habits for 75% of its suppliers. Own-consumption has been reduced for 51% of suppliers during the entire year and for 33% of households in the dry season to increase the share of milk that is marketed (Figure 5).

In comparison, our qualitative interviews show that own-consumption remains very widespread among non-suppliers, who continue to drink or process for their own use 74% of the milk they produce. Before the arrival of the LDB, herders offered their dairy products for sale on the main road (informal market). This random marketing process has declined with the appearance of the LDB, particularly for its suppliers: the LDB has become the sole outlet of the milk produced for 75% of the dairy’s suppliers. This explains why 15% of LDB suppliers report a fall in market sales: these producers have chosen to sell most of their milk to the dairy directly.

With the monthly payment of milk sales from the LDB, and the additional check-off system that can lead some suppliers actually owing money to the dairy for feed, LDB suppliers can be seen as actually more cash-strapped than they used to be when they marketed some milk surplus on the informal market. Due International Food and Agribusiness Management Review

566

Wane et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0218 - Tuesday, August 22, 2017 12:33:35 PM - IP Address:24.21.169.207

Fall in on-farm consumption in dry season 33%

Fall in on-farm consumption throughout the year 51%

Fall in market sales throughout the year 12% Fall in market sales in dry season 3%

Less available milk for calves 1%

Figure 5. Changes in milk outlet of Laiterie Du Berger (LDB) suppliers, calculated from data from IFPRICIRAD-GRET database on nutrition of 445 LDB suppliers. to this lack of monetary resources, 77% of the LDB suppliers deprive themselves of the staple foods they usually consume. Although 33% report that this situation rarely occurs, more than half (55%) experience this occasionally and 12% often (Figure 6).

Another strategy to cope with the lack of money to buy food is to forego a meal. Nearly half (49%) of LDB suppliers interviewed have had to reduce the number of meals per day during the four weeks prior to the surveys. Among these, 9% had encountered this situation often, whereas half have encountered it occasionally and 41% rarely.

To address food security issues, it is useful to consider the quantities of food consumed per person. The IFPRI-CIRAD-GRET questionnaire lists household members who have been forced to reduce the quantity of food they previously ate. The results show that the majority (64%) of the LDBâ&#x20AC;&#x2122;s milk suppliers needed to reduce the quantity of food consumed. Of these, 13% encountered this situation often, whereas 57% did so occasionally and 30% rarely.

often 12% rarely 33%

occasionally 55%

Figure 6. Share of Laiterie Du Berger (LDB) suppliers who have to deprive themselves of staple food, calculated from data from IFPRI-CIRAD-GRET database on nutrition of 445 LDB suppliers.

International Food and Agribusiness Management Review

567

Wane et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0218 - Tuesday, August 22, 2017 12:33:35 PM - IP Address:24.21.169.207

■■ The income status of LDB suppliers is largely dependent on their ability to keep delivering milk during the dry season The second indicator of supplier household livelihoods used in this study is income stabilization. Crossing the income mobility indices with the food security status typology of pastoral households highlights the importance of income stabilization between the seasons in explaining the households’ food security status.

Overall, the value of the Shorrock index calculated for the entire sample (μ1nor=0.62) indicates that households are relatively mobile within food security groups: their food security status tends to change between dry and wet seasons in a given year for the better or for the worse (Table 4).

Groups 1 and 2 are more likely to observe degradation than improvement of their relative income (μimp≤μdeg).

Thus, it is difficult for these groups to maintain their relative level of income between seasons. In groups 3 and 4, there is more income improvement than degradation (μimp>μdeg). These households appear to find a means to stabilize their incomes between dry and wet seasons. In fact, despite the significant decrease in dairy revenues in the dry season, groups 3 and 4 likely manage to stabilize their overall revenue by selling a portion of their herds.

If we consider ‘Group 1: insecure’, income mobility and herd size are key factors that explain the food insecurity of these households (Table 5). The probability of being in the food insecure group decreases significantly by 0.21 when income mobility moves from deterioration to improvement. Thus, the stability Table 4. Income mobility indices and food security groups of Laiterie Du Berger (LDB) milk suppliers, calculated from data from IFPRI-CIRAD-GRET database on nutrition of 445 LDB suppliers.

Food security groups

Income mobility groups

Group 1 – ‘insecure’ Group 2 – ‘poorly secure’ Group 3 – ‘secure’ Group 4 – ‘highly secure’ Total

Shorrock index (μ1nor)

Improvement (μimp)

Degradation (μdeg)

0.58 0.59 0.70 0.63 0.62

0.25 0.32 0.61 0.55 0.41

0.53 0.47 0.32 0.29 0.42

Table 5. Ordered probit results on marginal effects of various variables on household food security, calculated from data from IFPRI-CIRAD-GRET database on nutrition of 445 Laiterie Du Berger suppliers.1 Food security groups Income mobility

Sources of income Number of years supplying milk

Number of livestock heads

1

Deterioration Stability Improvement 1 or 2 >2 1-4 4-5 5-6 6-8 1-24 24-45 45-80 ≥80

Insecure

Poorly secure Secure

Highly secure

Reference -0.07 -0.21*** Reference -0.04 Reference -0.02 0.04 -0.05 Reference -0.08 -0.02 -0.14**

Reference -0.08 -0.16*** Reference -0.14** Reference -0.16** 0.02 -0.11 Reference 0.03 -0.11 -0.07

Reference 0.08 0.13*** Reference 0.02 Reference 0.13** -0.10** -0.02 Reference 0.09 0.05 0.24***

Levels of statistical significance: *** = 1%; ** = 5%; * = 10%.

International Food and Agribusiness Management Review

568

Reference 0.07 0.24*** Reference 0.16** Reference 0.04 0.04 0.18*** Reference -0.04 0.07 -0.03

Wane et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0218 - Tuesday, August 22, 2017 12:33:35 PM - IP Address:24.21.169.207

of dairy income between the dry and rainy season brought by being a regular supplier of the LDB plays an important function in the food security strategies of pastoral households.

In ‘Group 2: poorly secure’ households, income mobility, number of years supplying milk and number of income sources appear to be the most important factors. The odds of being in this poorly food secure group also decreased very significantly by 0.16 when household income went from degradation to improvement.

The seniority in milk supply is also a determining factor; in fact, adding one additional year of supplying milk from the reference group of ‘less than four years supplying milk’ decreases the probability of being ‘poorly secure’ by 0.16. This result is all the more relevant as focus group discussions uncovered that women with the most experience of supplying milk to markets continue to supply milk in the dry season and therefore, benefit from dairy income despite the more difficult production conditions.

In ‘Group 3: secure’, income stability, number of income sources and number of years supplying milk are also the main determining factors. Moving from the reference income degradation to income improvement increases the probability of being food secure by 0.24 at the 1% statistically significant level. Seniority in supplying milk is a key factor of food security; from less than four to at least six years of supplying milk, the probability of being ‘secure’ increases by 0.18. Thus, the oldest suppliers of LDB have a significant chance of not experiencing food deprivation.

Being in ‘Group 4: very secure’ in food depends significantly on income mobility, seniority in milk supplying and herd size. Moving from income degradation to improvement increases by 0.13 the probability of being ‘very secure’. It is worth noting that seniority in milk supplying reveals a double trend. When moving from less than four years of supplying milk to more years, the probability of being in this food security group increases by 0.13. However, a move from the reference of less than four years of milk supply to 5-to-6 years decreases the chances of belonging to this ‘very secure’ food security group by 0.1. The mixed effect of milk supply seniority can be explained by the fact that most of the relatively new suppliers to the LDB are also in this ‘very secure’ group: already relatively food-secure pastoralist households have spotted this new income opportunity of supplying milk to the LDB. By choosing to channel more of their milk production to the dairy, these households can increase their household income and thus purchase increasingly more varied foods, thus improving their food security status when they move from the new supplier status to 4-to-5 years of supplying dairy. However, this overall increase in income also leads households to choose to spend it on non-food items such as clothing or improving their living conditions, to the detriment of food security. The focus group discussions with women who supply the dairy also showed that mothers were selling the majority of the milk they produced rather than retaining a portion of this nutritious foodstuff for their children, as they previously did when pastoralists had no market outlet for their milk. These livelihood decisions could contribute to a decrease in the household’s food security status in the longer term; thus, this explains the negative sign of the coefficients when ‘very secure’ farmers become established suppliers of the LDB. 4.3. Supplying milk to the Laiterie Du Berger seems to improve herders’ socioeconomic resources ■■LDB suppliers have a greater diversification of income sources In the Ferlo, 98% of herders’ incomes are related to the marketing of ruminants (Wane et al., 2009). But in the LDB’s milk supply area, there are different sources of income for pastoral households (Table 6).

Despite livestock sales during the dry season remaining the most important source of income (61%), there is an emerging trend of milk sales constituting an increasing share of household income. In the rainy season milk income constitutes more than half (56%) of overall household income. Milk is increasingly becoming a new opportunity for income generation, whereas in the past, pastoralists were forced to recapitalize by selling a portion of their herd to obtain cash. This finding also links being a milk supplier to the LDB with the potential to keep increasing one’s herd size rather than having to sell animals in times of financial need.

The results of the ordered probit model (Table 5) showed that increasing the number of income sources and

International Food and Agribusiness Management Review

569

Wane et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0218 - Tuesday, August 22, 2017 12:33:35 PM - IP Address:24.21.169.207

Table 6. Components (%) of Laiterie Du Berger (LDB) suppliers’ overall household income in the dry and rainy seasons (adapted from IFPRI-CIRAD-GRET database on nutrition of LDB 445 suppliers).

Livestock sales Milk sales Crop production Other sources of income/self-employment Wage labor Others (rent, transfers, donations)

Dry season

Rainy season

61 25 6 4 2 2 100%

34 56 4 3 1 1 100%

the number of cattle heads had a significant positive impact on improving the food security status of the already more food secure households.

■■The LDB has become a facilitator for linking family farmers to competitive markets The qualitative interviews reveal that the LDB has increased the market orientation of pastoral cattle herders.

The milk suppliers to the LDB are price-takers who adapt to the conditions set by the dairy plant. The price system arising from the relationship between the LDB and milk suppliers does not always correspond to the relative scarcity of milk and the optimal resource allocation by herder households. Pastoral households now respond to milk market opportunities by allocating more of their milk produced to sales. These opportunities are reflected in particular by the existence of market outlets for milk produced beyond what is needed to feed calves and the household members, mainly in the wet season.

The LDB’s role in facilitating farmers’ access to markets is also reflected by its supply of animal feed and loan grants when pastoral investment strategies were previously based on self-financing (Wane, 2005).

Credit advances for animal feed have always been the cornerstone and the strength of the LDB. However, the qualitative data gathered from milk suppliers indicate that they generally consider the quantities of feed received as insufficient. Despite the apparently advantageous conditions, these suppliers also deplore the high cost of these feed supplements.

Similarly, the dairy plant has removed obstacles previously faced by pastoralists to access the complex and competitive markets of livestock products. The facilitation of the marketing of pastoral products contributes to the herder households’ evolution from a primarily subsistence production logic to an increasing use of markets, which leads to a change of productive strategies (Barrett, 2008). However, the pastoralists remain subject to uncertainty in their productive activities, to the combined effects of prices and taxes on their decisions and to the conditions of access to other market players (Duteurtre et al., 2010; Wane, 2005). To remove these market access constraints, different institutional initiatives could be used based around collective action (Markelova et al., 2009) particularly through producer organizations, market standards or partnerships.

■■The LDB is a catalyst in the partial restructuring of pastoral mobility Because monetary incentives are not the most important ones for pastoral herders, it is equally essential to analyze the possible impacts of the LDB milk supply chain on the pastoral practices of its suppliers, as noted by Cesaro (2009).

The majority of milk suppliers continue to use geographical mobility as a strategy for cattle herd management.

However, this traditional itinerant lifestyle is partially modified in its general organization for milk suppliers of the LDB. The most radical change for herders stems from the desire to continue to supply milk in the dry season by maintaining many of the dairy cows in a sedentary encampment.

International Food and Agribusiness Management Review

570

Wane et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0218 - Tuesday, August 22, 2017 12:33:35 PM - IP Address:24.21.169.207

This change translates into the splitting of the herd and to a change in the social organization of mobility.

Our qualitative individual interviews and the focus group discussion held in the pastoral settlement concur in identifying that women and children now remain on the sedentary encampment with the dairy cows, whereas the men move to other locations with the remainder of the herd in search of pastureland. Similarly, the pace and magnitude of transhumance have been modified by the herders’ strategy to remain within the dairy’s milk collection area. Nonetheless, pastoral mobility remains the principal coping strategy of pastoralists who live in an uncertain biophysical context.

5. Conclusions Using the generic value chain model (Porter, 1985), this study has shown that the innovative raw milk supply chain developed by the LDB in Northern Senegal has allowed the dairy to increase its number of pastoralist suppliers, and the quantity and quality of the milk they sold to the processing plant. We have also studied the changes brought by this new supply chain from the viewpoint of the pastoralist households using a conceptual framework on the sustainability of pastoral systems (Lambert-Derkimba et al., 2015). Our findings suggest that by contributing to stabilizing suppliers’ dairy incomes in the dry season, diversifying income sources and enabling households to keep capitalizing into substantial livestock herds, the LDB’s milk supply chains could have played an important role in securing some of its supplier households’ food security. Supplier households in the Ferlo that had focused on the regularity of their milk supply within a calendar year and over several years seemed to have witnessed an improvement in their food security and overall access to socioeconomic resources. Households placing milk sales to the dairy as their preferred source of stable income had likewise seemed to improve their livelihoods.

However, the quantitative findings from this research are limited by the cross-sectional data featuring only one calendar year of observations. In a context of great environmental and market variability, as highlighted in the introduction, this limited data set does not allow to conclude on the LDB’s new supply chain as a cause of its suppliers’ evolution in sustainability. Further investigations on this topic should use longitudinal data covering several years of observations from the herders. This could contribute to describing better the complex tendencies that accompany innovation or technology introduction.

Nevertheless, the combined use of cross-sectional quantitative and qualitative data suggests that the LDB’s innovative raw milk supply chain does contribute to strengthening the food security and socioeconomic resources of its supplier households. The new dairy marketing outlet that appeared with the LDB has helped provoke profound changes in the local dairy production system, with women and children now remaining in semi-permanent encampments with the producing dairy cows, where they receive animal feed from the dairy to sustain their cows’ milk production, whereas the men of the community continue their pastoral practices, moving their herds of bulls and non-lactating cows to new areas as needed in search of water and forage.

The findings from this research have implications for other agro-processors interested in developing local milk supply chains in traditional pastoralist drylands environments so as to tap this large and still mobile potential milk reservoir. The LDB’s example shows how innovations in supply chain management and business relationships tailored to smallholder herders have allowed the LDB to secure a good quality supply of milk across the year despite the natural trough in milk production during the dry season when forage becomes scarce. However, to ensure that this new access to milk markets and the income opportunities it brings do not destabilize the livelihoods of traditional pastoralist communities, additional training targeting women milk suppliers through development partners should also cover the strategies that will help protect the food security and welfare of the more vulnerable pastoralist household members, who have no say in how the new dairy income is used.

International Food and Agribusiness Management Review

571

Wane et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0218 - Tuesday, August 22, 2017 12:33:35 PM - IP Address:24.21.169.207

Acknowledgements We would like to express our special gratitude and thanks to the Consortium of the International Food Policy Research Institute (IFPRI), The French Centre for Agricultural Research and International Cooperation (CIRAD) and GRET, an international development NGO governed by French law, for allowing us to use their databases from the study area. We also express thanks to partners from the Laiterie du Berger (LDB) and pastoralists for their time in in-depth discussions that helped us better understand various and complex dynamics. This work was undertaken as part of the CGIAR Research Program on Policies, Institutions, and Markets (PIM) led by IFPRI. Funding support for this study was provided by CIRAD, the Small Grants Program of the CGIAR Standing Panel on Impact Assessment and the CGIAR Research Program on Policies, Institutions, and Markets. The opinions expressed here belong to the authors, and do not necessarily reflect those of LDB, PIM, IFPRI, or CGIAR.

References Ayantunde, A.A., J. De Leeuw, M.D. Turner and M. Said. 2011. Challenges of assessing the sustainability of (agro)-pastoral systems. Livestock Science 139: 30-43.

Barrett, C. 2008. Smallholder market participation: concepts and evidence from Eastern and Southern Africa.

Food Policy 33: 299-317.

Bernard, T., M. Hidrobo, A. Le Port and R. Rawat. 2015. Health benefits in contract farming and their impacts on production: evidence from the dairy sector in Senegal. IFPRI, Dakar, Senegal. Available at: http://tinyurl.com/mavdzwy.

Boogaard, B.K., E. Waithanji, E.J. Poole and J.-J. Cadilhon. 2015. Smallholder goat production and marketing: a gendered baseline study from Inhassoro District Mozambique. NJAS –Wageningen Journal of Life Sciences 74-75: 51-63.

Buhl, S. and K. Homewood. 2000. Milk selling among Fulani women in Northern Burkina Faso. In: Rethinking Pastoralism in Africa, edited by D.L. Hodgson. Ohio University Press, Athens, OH, USA. Cesaro, J.D. 2009. Mobilité pastorale et accès aux marchés: le cas des éleveurs du forage de Niassanté, MSc Thesis, Université Paris, Montpellier, France.

Coates, J., A. Swindale and P. Bilinsky. 2007. Household food insecurity access scale (HFIAS) for measurement of food access: indicator guide (V.3). Academy for Educational Development, Washington, WA, USA. De Haan, C. 2016. Prospects for livestock-based livelihoods in Africa’s Drylands. World Bank Studies.

World Bank, Washington, WA, USA. Doss, C. 2013. Data needs for gender analysis in agriculture. IFPRI Discussion Paper 01261. IFPRI, Washington, WA, USA. Duteurtre G., V. Alary, V. Ancey, C. Corniaux, P.N. Dieye, D. Gautier, O. Ninot and F. Vatin. 2010. Accès aux marchés et développement de l’élevage en Afrique : la construction sociale des liens marchands , communication aux 4èmes journées. SFER-INRA-CIRAD, Rennes, Agro-campus Ouest, les 9 et 10 décembre 2010, 18 p. Available at: http://tinyurl.com/mqywvfx.

Duteurtre, G. and C. Corniaux. 2013. Etude relative à la formulation du programme d’actions détaillé de développement de la filière lait en zone UEMOA. Rapport définitif. UEMOA, Ouagadougou, Burkina Faso and CIRAD, Dakar, Sénégal.

Gerber, M., L. Astigarraga, C. Bockstaller, J.L. Fiorelli, N. Hostiou, S. Ingrand, M. Marie, W. Sadok, P. Veysset, R. Ambroise, J. Peigné, S. Plantureux and X. Coquil. 2009. Le modèle Dexi-SH pour une évaluation multicritère de la durabilité agro-écologique des systèmes d’élevage bovins laitiers herbagers. Innovations Agronomiques 4: 249-252.

Griliches, Z. 1957. Hybrid corn: an exploration in the economics of technological change. Econometrica 25: 501-522.

Haddad, L., J. Hoddinott and H. Alderman. 1997. Intrahousehold resource allocation in developing countries: models, methods and policy. Johns Hopkins University Press, Baltimore, MA, USA. Huss-Ashmore, R. 1996. Livestock, nutrition, and intrahousehold resource control in Uasin Gishu District, Kenya. Human Ecology 24: 191-213.

International Food and Agribusiness Management Review

572

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0218 - Tuesday, August 22, 2017 12:33:35 PM - IP Address:24.21.169.207

Wane et al.

Volume 20, Issue 4, 2017

Ickowicz, A., V. Ancey, C. Corniaux, G. Duteurtre, R. Poccard-Chappuis, I. Touré, E. Vall and A. Wane. 2012.

Crop – livestock production systems in the Sahel – increasing resilience for adaptation to climate change and preserving food security. In: Proceeding of the FAO–OECD Workshop on Building Resilience for Adaptation to Climate Change in the Agriculture sector. FAO-OECD, Rome, Italy.

Kaboret. 2010. Appuis institutionnels à la mise en œuvre de la stratégie régionale de renforcement des services vétérinaires et préparation à l’accès des viandes sahéliennes aux marchés des pays d’Afrique du Nord. Phase d’étude de faisabilité et de programmation des activités du futur projet. Rapport final. CILSS, Niamey, Niger.

Ki, J.B., S. Faye and B. Faye. 2005. Pauvreté multidimensionnelle au Sénégal: une approche non monétaire par les besoins de base. PMMA Working Paper N° 2005-05. Available at: http://ssrn.com/abstract=985928.

Lambert-Derkimba, A., C. Aubron, A. Ickowicz, I. Touré and C.H. Moulin. 2015. An innovative method to assess the sustainability of pastoral systems in their territories (PSSAF). Revue d’Elevage et de Médecine Vétérinaire dans les Pays Tropicaux 68: 135-142.

Ly, C., U. Pica-Ciamarra and J. Otte. 2010. A dual-track approach to livestock development: economic rationales and institutional bottlenecks in West Africa. FAO Regional Office for Africa, Accra, Ghana and FAO Animal Production and Health Division, Rome, Italy.

Markelova, H., R. Meinzen-Dick, J. Hellin and S. Dohrn. 2009. Collective action for smallholder market access. Food Policy 34: 1-7.

Meinzen-Dick, R., N. Johnson, A. Quisumbing, J. Njuki, J. Behrman, D. Rubin, A. Peterman and E. Waithanji.

2011. Gender, assets, and agricultural development programs: a conceptual framework. CAPRi Working Paper No. 99. IFPRI, Washington, WA, USA. Mulubrhan, A., A. Solomon and S. Bekele. 2012. Welfare impacts of maize – pigeon pea intensification in Tanzania. Agricultural Economics 43: 27-43.

Negatu, W. and A. Parikh. 1999. The impact of perception and other factors on the adoption of agricultural technology in the Morert and Jiru Woreda (district) of Ethiopia, Agricultural Economics 21: 205-216.

Parisse, P. 2012. Developing local dairy production: the Laiterie du Berger, Senegal, FACTS Reports 6.

Available at: http://factsreports.revues.org/2271.

Porter, M.E. 1985. The competitive advantage: creating and sustaining superior performance. Free Press, New York, NY, USA. Powers, D.A. and Y. Xie. 2000. Statistical methods for categorical data analysis. Academic Press, San Diego, CA, USA. Reardon, T., C.P. Timmer, C.B. Barrett and J. Berdégué. 2003. The rise of supermarkets in Africa, Asia and Latin America. American Journal of Agricultural Economics 85: 1140-1146.

Rey-Valette, H., O. Clément, J. Aubin, S. Mathé, E. Chia, M. Legendre, D. Caruso, O. Milolasek, J. P. Blancheton, J. Slembrouck, A. Baruthio, F. René, P. Levang, P. Morissens, and J. Lazard. 2008. Guide to the co-construction of sustainable development indicators in aquaculture. CIRAD, Montpellier, France.

Touré, I., A. Ickowicz, A. Wane, I. Garba and P. Gerber. 2013. Atlas of trends in pastoral systems in the Sahel 1970-2012. SIPSA. FAO, Rome, Italy and CIRAD, Dakar, Senegal. Available at: http://tinyurl. com/nxxavo7.

Vorley, B., M. Lundy and J. MacGregor. 2009. Business models that are inclusive of small farmers. In: Agro-Industries for Development, edited by C.A. Da Silva, D. Baker, A.W. Shepherd, C. Jenane and S. Miranda-da-Cruz. FAO, Rome, Italy, UNIDO, Vienna, Austria and CABI, Cambridge, MA, USA, pp. 186-222.

Wane, A. 2005. Marchés de bétail du Ferlo (Sahel sénégalais) et comportements des ménages pastoraux.

Colloque SFER. SFER, Montpellier, France. Available at: http://agritrop.cirad.fr/529547.

Wane, A., I. Touré and V. Ancey. 2009. Pastoralisme et recours aux marchés – cas du Sahel Sénégalais (Ferlo). Cahiers Agricultures 19: 1-7.

Wane, A., I. Touré and A. Ickowicz. 2014. Changing environment and market behaviours of Sahelian herders.

6th All Africa Conference on Animal Agriculture (AACAA), 27–30 October 2014, Nairobi, Kenya.

Available at: http://tinyurl.com/msswmah.

International Food and Agribusiness Management Review

573

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2015.0218 - Tuesday, August 22, 2017 12:33:35 PM - IP Address:24.21.169.207

OPEN ACCESS International Food and Agribusiness Management Review Volume 20 Issue 4, 2017; DOI: 10.22434/IFAMR2016.0029

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0029 - Tuesday, August 22, 2017 12:35:02 PM - IP Address:24.21.169.207

Received: 18 February 2016 / Accepted: 5 December 2016

Comparative advantage and export potential of Thai vegetable products following the integration into the ASEAN Economic Community RESEARCH ARTICLE Pheesphan Laosutsana, Ganesh P. Shivakotib, and Peeyush Soni

c

aPhD

Student, bVisiting Professor and cAssociate Professor, Department of Food, Agriculture and Bioresources, School of Environment, Resources and Development, Asian Institute of Technology, P.O. Box 4, Klong Luang, 12120 Pathumthani, Thailand

Abstract International trade, which is the exchange of capital, goods and services across international borders or territories, has contributed to the rapid global economic growth in recent decades. In Southeast Asia, the establishment of the Association of Southeast Asian (ASEAN) Free Trade Area and the implementation of the ASEAN Economic Community have benefited Thai entrepreneurs and other member countries’ as nearly all import restrictions are removed and market entry barriers lowered. The ASEAN is an organization of countries in Southeast Asia set up to promote cultural, economic and political development in the region; and comprises 10 member states: Brunei Darussalam, Cambodia, Indonesia, Laos, Malaysia, Myanmar, the Philippines, Singapore, Thailand and Vietnam. Specifically, this research has explored the possible economic impacts of trade liberalization and improved connectivity on Thailand’s exportation of 23 vegetable product groups to the ASEAN member states (AMSs) using the Revealed Comparative Advantage and Revealed Symmetric Comparative Advantage indices based on the 2009-2013 datasets. In the analysis, the indices were applied to the 14 fresh and 9 preserved vegetable product groups from 15 countries (i.e. AMSs, Australia, China, Japan, South Korea, the USA) to determine their comparative advantages. The dendrogram was used to cluster the countries with regard to their ability to export the fresh and preserved vegetable products. In addition, the Boston Consulting Group matrix was utilized to determine the relative market positions of the Thai fresh and preserved vegetables. The analysis results identified four each of the Thai fresh and preserved vegetable product groups with high comparative advantage in the ASEAN market.

Keywords: ASEAN, RCA, RSCA, comparative advantage, BCG, exporting, integration JEL code: F13, F19, F49, F69 Corresponding author: soni.ait@gmail.com

© 2017 Laosutsan et al.

575

Laosutsan et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0029 - Tuesday, August 22, 2017 12:35:02 PM - IP Address:24.21.169.207

1. Introduction The global economy of recent decades has been characterized by rapid growth1, driven in part by the exponential increase in international trade. At the same time, the international trade boom has been fueled by advancing technology and the concerted efforts of nations to eliminate trade barriers2. Integration into the world economy has proven a powerful means for countries to realize economic growth, development and poverty reduction3, in addition to the improved overall standards of living. Thus, some developing countries have opened their own economies while others have deliberately opted to limit the access to their markets (IMF, 2001).

Thai vegetables are of economic significance in the country’s international trade of agricultural products.

Revenues from vegetable exports contribute enormously to the country’s agricultural sector and economic stability.4 Due to the Kingdom’s location in the tropical zone, most Thai vegetables can be grown all year round (Cheyroux, 2003). In addition, the Thai government has embarked on a plan to transform Thailand into ‘the kitchen of the world’5. Nevertheless, Thai vegetable products are facing various obstacles, including competitive offerings, diverse consumer preferences between regions, and trade and non-trade restrictions (Prachason, 2009).

Between 2011-2013, the Thai export value of vegetables decreased at an annual rate of 1.18% due to the decline in demand from its major importers, such as the EU, Japan, Taiwan and the UK. However, Thailand’s vegetable export to certain ASEAN countries, including Malaysia and Singapore, had increased. Thus, the ASEAN countries present an opportunity to expand the trade (Anon, 2014). As a free trade area, the 10-member countries expect the ASEAN Free Trade Area (FTA) to ease the flow of goods and services in the region. Thailand, an ASEAN member, would reap considerable benefits from the freer regional trade as the production costs would be lower.

In essence, the benefits of the ASEAN FTA include the reduction, or even removal, of tariffs between the member states, the reduced product prices due to increased competition, and possible investment opportunities in the market. On the other hand, the major drawbacks of the trade agreement entail the risks of an influx of cheap imports flooding local markets and the poor governance in some member countries. In fact, the reduced tariffs have negatively impacted Thailand’s agricultural sector due to a greater number of imported agricultural goods.

With the increase in global and regional FTAs, the Kingdom’s ‘vegetables industry’, which refers to the production, distribution and sale of agricultural and food products, must refine its current strategy to remain relevant and competitive (Urata, 2004). In Southeast Asia, the ASEAN FTA has benefitted Thai businesses with the increase in exports. By the same token, other member countries of the FTA are provided with opportunities to enter the Thai market. Furthermore, several of the FTA partner countries are capable of offering the same products at comparable, or competitive, price and quality. Thai operators are thus required to be prepared for the intensified competition.

It is thus imperative to analyze the competitiveness of Thai vegetables exports and identify the vegetable commodities that can compete in the export market. The results would be of particular use to future free

1 The

world has become a much smaller place over the past two decades. International trade has grown twice as fast as worldwide income during this period. Spurred by advances in information technology, a growing share of this trade is in services rather than merchandise, especially among rich countries (Dollar and Kraay, 2001).

2 Trade barriers are government-induced restrictions on international trade. Economists generally agree that trade barriers are detrimental and decrease overall economic efficiency, which can be explained by the theory of comparative advantage (Anon, 2011).

3 Economic integration is the unification of economic policies between different states through the partial or full abolition of tariff and non-tariff restrictions on trade present among them prior to their integration.

4 An economy with fairly constant output growth and low and stable inflation would be considered economically stable.

5 The campaign is aimed at promoting Thai food and products.

International Food and Agribusiness Management Review

576

Laosutsan et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0029 - Tuesday, August 22, 2017 12:35:02 PM - IP Address:24.21.169.207

trade negotiations between Thailand and other ASEAN member countries on vegetable products, as well as to the promotion of vegetable products with export potential.

This research has attempted to assess, based on the 2009-2013 datasets, the economic impacts of trade liberalization and improved connectivity on Thailand’s exportation of vegetables to the ASEAN member states (AMSs) using the Revealed Comparative Advantage (RCA) and Revealed Symmetric Comparative Advantage (RSCA) indices. In the analysis, the indices were applied to the 14 fresh (Harmonized System (HS) codes 0701-0714) and 9 preserved (HS codes 2001-2009) vegetable product groups from 15 countries (i.e. AMSs, Australia, China, Japan, South Korea, the USA) to determine the comparative advantages among the countries under investigation. The Boston Consulting Group (BCG) matrix was also utilized to determine the relative market positions of the Thai fresh and preserved vegetables and to subsequently formulate the corresponding strategic plans and actions. Moreover, the dendrogram was used to cluster the countries with regard to their ability to export the fresh and preserved vegetable products.

The organization of the rest of the paper is as follows: Section 2 chronicles the ASEAN trade liberalization and reviews the global and ASEAN trade of vegetable products. Section 3 details the research data and analytical methods. Section 4 discusses the research findings and the guideline recommendations to raise the competitiveness of Thai vegetable products. The concluding remarks are provided in Section 5.

2. Trade liberalization and international trade of vegetable products 2.1 Thailand and ASEAN trade liberalization The Association of Southeast Asian Nations (ASEAN)6 was first established in 1967 in accordance with the Bangkok Declaration, comprising five founding member nations: Indonesia, Malaysia, Philippines, Singapore and Thailand. Brunei, Vietnam, Laos, Myanmar and Cambodia were later admitted into the grouping respectively in 1984, 1995, 1997, 1998 and 1999. In 2007 a formal charter for ASEAN was ratified by the 10 AMSs and came into effect the following year.

ASEAN seeks to promote the socioeconomic progress and regional stability through cooperation in banking, trade, technology, agriculture, industry and tourism. In 1992, the bloc members agreed to create a regional common market, i.e. the ASEAN FTA, which became effective in 1993. With the implementation of the ASEAN FTA, tariffs among the ASEAN nations, particularly among the six oldest ASEAN members, are greatly reduced. According to Ken (2014), three major impacts of an FTA are the reduction of tariffs on goods, the lowering of barriers to trade in services, and a time-cost saving arising from shared logistics. The bloc members have also pledged to collaborate to promote foreign investments in the region. In addition, the grouping has reached free trade agreements with China (2004), South Korea (2006), Japan (2008), Australia, New Zealand and India (2009).

Thailand is Southeast Asia’s second largest economy with gross domestic product (GDP) of approximately US$ 365 billion in 20147. Given the importance of exports to the country’s GDP, the Kingdom has been spearheading trade liberalization in the region and facilitation with the rest of the world (Kawai and Wignaraja, 2011). In addition, Thailand has been a strong advocate for ASEAN’s regional economic integration, which has taken shape since the inception of ASEAN Free Trade Agreement in 1992. The liberalization process under the ASEAN Economic Community (AEC) is giving Thailand huge opportunities for the expansion of market and production (Nguyen, 2014).

Overall, ASEAN has been on the right track to eliminating all tariffs as it has progressed to a level of 80% with regard to elimination of intra-regional tariffs on goods. Certain items remain on sensitive lists of each 6 Overview ASEAN. Association 7 The

of Southeast Asian Nations. Retrieved 15 February 2015. populations and GDP of 10 ASEAN member countries available on the IMF’s world economic outlook database 2014.

International Food and Agribusiness Management Review

577

Laosutsan et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0029 - Tuesday, August 22, 2017 12:35:02 PM - IP Address:24.21.169.207

member country, e.g. coffee beans, copra, potatoes and cut flowers in the case of Thailand. Nonetheless, the tariffs on the sensitive-listed items are to be reduced to zero by the end of 2015.

Even with a complete elimination of tariffs, non-tariff barriers (NTBs) to trade, e.g. quotas and licenses, are still widely deployed despite the requirements of the member countries to scale down and totally remove the NTBs under the free trade agreement8. One such NTB example is the inconsistent and unreasonable labeling and packaging requirements that impede free trade across ASEAN. NTBs could undermine the economic integration process and the realization of the AEC by 2015. Economic integration will continue but will be ‘limited to economies which are able to address the NTBs and the supply-side capacity constraints. It will also be limited to highly integrated production networks.’ The private sector should involve in identifying NTBs and subject all non-tariff measures (NTMs) to a ‘compliance review’ in order to ensure that they are transparent and non-discriminatory and in order to minimize trade restrictiveness (Sanchita et al., 2013).

Due to its geographical location, the region is poised to be the world’s supplier of fruits and vegetables, which are a necessary part of human diet and a source of earnings for the region. In addition, countries in the northern hemisphere, e.g. European countries and the US, cannot grow vegetables in the winter and thereby provides ASEAN with an enormous economic opportunity. In fact, Thailand’s trade in vegetables within the region, although gradually rising during 2010-20149, was relatively small vis-à-vis with countries outside the region, e.g. Japan, the EU and the USA. It is thus anticipated that the full implementation of the AEC in 2015 would significantly increase the trade and investment activity in agricultural products within the ASEAN region.

The EU, US and Japan are currently the major export markets of Thai vegetables, while ASEAN is an attractive market with good prospects for vegetable products from Thailand (Anon, 2012). The implementation of the ASEAN FTA has gradually reduced import tariffs to 0% for most member countries since January 1 2010, with the ASEAN newcomers (i.e. Cambodia, Laos, Myanmar and Vietnam) also gradually lowering their respective import tariffs. The import tariff reductions in turn lower the costs of exports and thereby contribute to an increase in exports to these countries (Calvo Pardo et al., 2009).

In addition, greater benefits from the ASEAN economic integration could be realized with increased outbound investment by Thai businesses. Currently, a majority of Thai small and medium-sized enterprises (SMEs) have failed to fully explore opportunities in the region. Rather, they need to broaden their perspective and be looking for allies, partners and connections in ASEAN so that they have more of a competitive advantage once the full integration takes place.

Within ASEAN, trade liberalization will stimulate the output of each member country in accordance with their comparative advantage. Nonetheless, since trade liberalization tends to increase the output of capitalintensive goods more than that of labor-intensive goods, the less-developed countries within the region likely reap a smaller share of benefits vis-à-vis their more developed counterparts. In addition, a significantly larger proportion of gains from an FTA would be captured by the physical means of production than by the labor, further exacerbating the wide income gap between high-income and low-income households within ASEAN (Danupon et al., 2009).

2.2 The aftermath of Thai-Chinese Bilateral Trade Agreement Under the global free trade agreement, China and India are two formidable exporting countries of vegetables and fruit, while countries in Southeast Asia and South Asia are the net importers (Tingjun and Thomas, 2006).

Notwithstanding, China at the same time needs to import inputs from multiple sources and thus ASEAN,

8 Given the very slow progress in identifying NTBs from among the NTMs, a critical step is to subject all existing NTMs to a compliance review to ensure that they are transparent, non-discriminatory, and minimizes trade restrictiveness. (Chia, 2013).

9 The United Nations Commodity Trade Statistics Database (UN COMTRADE).

International Food and Agribusiness Management Review

578

Laosutsan et al.

Volume 20, Issue 4, 2017

with its proximity to China, could serve as a strategic source of natural-resource-based and intermediate inputs under the FTA.

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0029 - Tuesday, August 22, 2017 12:35:02 PM - IP Address:24.21.169.207

According to Sarasin (2011), China was a country with the highest production of vegetables in 2005, accounting for as much as 60% of the worldâ&#x20AC;&#x2122;s vegetables production. This was followed by India at 9%, USA 4%, Turkey 3%, and other countries with an aggregate of 24%.

For Thai vegetable products exports, China is the countryâ&#x20AC;&#x2122;s principal customer, with the 2013 export value of over 500 million US dollars, contributing to almost half the export value. By comparison, the combined value of Thai vegetables exports to the other nine ASEAN member countries was surprisingly small, accounting for a mere 6.38% of the total vegetables export value of US$ 700 million annually. However, trade in vegetable products among ASEAN grew by 11.7% per annum (Anon, 2014).

Despite the Thai-Chinese free trade agreement, a number of Chinese local administrative governments have however imposed NTBs in the form of, e.g. local taxes, quality inspection, hygiene standards, plant diseases and insects. In addition, the conditions and terms of trade are specified by Chinese entrepreneurs (Sandee, 2013). On the contrary, no import restrictions exist for vegetables and fruits from China to Thailand under the Thai-Chinese FTA, placing Thailand in a disadvantageous position (Sally, 2007).

The situation is exacerbated if the vegetables and fruits produced by Thai and Chinese agriculturists are similar, e.g. onions and garlic, and destined for the same buyer countries. When faced with this situation, most Thai farmers switch to other non-competitive crops (Benja and Kingkorn, 2004). In addition, many Chinese vegetables can be grown all year round and are much lower in production costs. Thus, the entry of Chinese vegetables into the Thai market has driven down the price of locally produced vegetables. It is believed that more Thai fruits and vegetables with Chinese substitutes would inevitably suffer from the adverse impact of a future price slump.

2.3 World trade for fresh and chilled vegetables and processed vegetables As illustrated in Supplementary Figure S1, the EU and APEC countries collectively commanded the global vegetables exports in 2011, with their respective export of fresh vegetables of 35 and 36%; and the processed vegetables exports of 41 and 28% of the global total vegetables exports. On the other hand, the shipments of vegetable products from ASEAN of the same period were 4 and 5% of the global total vegetables exports for fresh and processed vegetables, respectively. In Supplementary Figure S2, the principal importers of vegetables products in 2011 were the EU and G7 countries. In addition to being an important exporter of vegetables products, the EU was a major consumer of fresh and processed vegetables with the import values of 36 and 37% of the global total vegetables exports. Although the main export markets of vegetables products are in the EU and the USA, ASEAN is still a lucrative market with good prospects for Thai vegetables products due to their geographical proximity and similarity in diets.

2.4 ASEAN trade for fresh and chilled vegetables and processed vegetables As shown in Supplementary Tables S2 and S3, trade in vegetable products among ASEAN member countries has been on an upward trend. Upon closer inspection, the fresh vegetables exports from Cambodia and Laos increased significantly in value by positioning their offerings as organically produced vegetables, while most Thai vegetables exports are preserved vegetables due to the availability of processing technology and added value. On the other hand, Brunei and Indonesia are two principal importers of vegetables products in the ASEAN grouping with little exports because of less favorable geographical location and high domestic demand.

International Food and Agribusiness Management Review

579

Laosutsan et al.

Volume 20, Issue 4, 2017

3. Research methodology

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0029 - Tuesday, August 22, 2017 12:35:02 PM - IP Address:24.21.169.207

This research has employed the RCA index developed by Béla Balassa (1965)10. The RCA index is defined as a ratio of certain export products of a country’s overall exports to the world in relation to a country’s total exports to total world exports (Vollrath and Thomas, 1991). In other words, it measures a country’s export of a particular commodity in relation to its total exports and to the corresponding export performance of a set of countries. In addition, the RCA index is applicable to certain sectors or commodities. In this research, the RCA index is concerned with the regional ASEAN bloc. The RCA index can be expressed as: RCAij =

(Xij / Xit)

(Xnj / Xnt)

(1)

where RCAij is the comparative advantage index of export product j of country i, Xij is the export value of product j of country i, Xit is the total export value of country i, Xnj is the export value of product j of n countries (or ASEAN), and Xnt is the total export of n countries (or ASEAN).

The RCA index values in this research are interpreted in accordance with the classifications of RCA index values in Hinloopen and Marrewijk (2008), where RCA>1 means that the proportion of a country’s (country i) export of product j to the total export of country i is greater than the proportion of exports of that product (product j) in the ASEAN market to the total export of the ASEAN countries (i.e. country i has a comparative advantage in exporting product j in the ASEAN market). In other words, when the RCA is greater than 1, the country under consideration has a revealed comparative advantage in the sector.

RCA<1 means that the proportion of country i’s export of product j to the total export of country i is less than the proportion of exports of that product (product j) in the ASEAN market to the total export of the ASEAN countries (i.e. country i has a comparative disadvantage in exporting product j in the ASEAN market). In other words, when the RCA is less than 1, the country has a revealed comparative disadvantage in the sector.

RCA=1 means that the proportion of country i’s export of product j to the total export of country i is equal to the proportion of exports of that product (product j) in the ASEAN market to the total export of the ASEAN countries. In other words, country i has neither comparative advantage nor disadvantage in exporting product j in the ASEAN market.

To address the issue of asymmetry, the RSCA, which has a measurement range of -1 to 1, was applied to the RCA index values in the second stage of the analysis.

RSCAij =

(RCAij – 1) (RCAij + 1)

(2)

The RSCAij index magnitude is -1≤RCAij≤1. Specifically, an RSCAij index beyond zero or close to {+1} indicates that product j of country i possesses a comparative advantage over its competitors. Conversely, an RSCAij index below zero or approaching {–1} shows that product j of country i is comparatively disadvantageous. In Tables 3 and 4, a plus sign (+) indicates a revealed symmetric comparative advantage and a minus sign (–) indicates a revealed symmetric comparative disadvantage.

In addition to the analysis of comparative advantages, this research has determined the strategic market positions of the Thai fresh and processed vegetable products using the BCG matrix. The BCG matrix is a 10 The RCA is an index used in international economics for calculating the relative advantage or disadvantage of a certain country in a certain class of goods or services as evidenced by trade flows (Widodo, 2009).

International Food and Agribusiness Management Review

580

Laosutsan et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0029 - Tuesday, August 22, 2017 12:35:02 PM - IP Address:24.21.169.207

framework to evaluate the strategic position of a business/brand portfolio and its potential.11 The matrix classifies a business portfolio into four categories based on industry attractiveness (growth rate of that industry) and competitive position (relative market share), i.e. star, question mark, cash cow and dog. The general idea is to classify positions of products along the two dimensions (Wind and Mahajan, 1981). These two dimensions reveal the likely profitability of the business portfolio in terms of cash needed to support that unit and cash generated by it. The aim of the BCG analysis is to provide an insight into which products/ brands to make further investment and which to pursue divestment.

The implementation of the BCG matrix requires undertaking the following five steps: first, select a products (i.e. fresh and preserved vegetables), and, second, define the market (i.e. the ASEAN market). Next, calculate the relative market share by dividing the market share of the product by the market share of biggest competitor.

Then, determine the relevant market growth rate by referring to the average revenue growth of market leaders in the vegetable industry. Finally, identify the strategic position of the product or business unit in the BCG matrix with a circle, taking into account the relative market share and growth. Furthermore, there are four strategic courses of action that could be pursued by a company after classification: harvest (for a business/ product that falls in the cash cow category), divest (for dogs and some question marks), maintain and invest (for cash cows and stars) (Kotler, 2003).

Furthermore, to reduce the data complexity and provide conceptual simplifications, this research has utilized the dendrogram to cluster the 15 countries under study by their respective ability to export the fresh (HS 0701-0714) and processed (HS 2001-2009) vegetable products. Generally, clustering lumps together objects that share some observed qualities, or divides a set of objects into separate classes whose boundaries reflect differences in the membersâ&#x20AC;&#x2122; observed qualities. Specifically, the dendrogram is a tree-like structure whose branches terminate at the objects being clustered, and the lengths of its branches indicate differences within clusters being merged or partitioned (Krippendorff, 1980).

3.1 Data The current bilateral or multilateral (e.g. World Trade Organization) trade negotiations require the use of the HS codes. Thus, this research has grouped and categorized the products to be analyzed into 23 standard items as suggested in the Harmonized System: HS code of the Global Trade Atlas12. Specifically, the vegetables under study are divided into two groupings by HS codes: the fresh or chilled vegetables grouping, consisting of 14 product groups (HS 0701-0714) and the preserved or processed vegetables grouping, consisting of nine product groups (HS 2001-2009) (Supplementary Table S1). In addition, the import and export datasets of interest belong to the 2009-2013 period of 15 countries (i.e. the AMSs, Australia, China, Japan, South Korea and the USA).

3.2 Analysis In the RCA analysis, the RCA indices were calculated on a yearly basis (2009-2013) for each country under study (15 countries). A positive RCA value reveals a comparative advantage while a negative outcome indicates a comparative disadvantage with regard to a particular commodity of a country. To mitigate the issue of asymmetry associated with RCA, the RSCA was applied to the RCA index values for further analysis, where a positive RSCA result reveals a comparative advantage while a negative outcome indicates a comparative disadvantage. The countries were then clustered by their respective ability to export the fresh and processed vegetable products using the dendrogram. In addition, the BCG matrix was utilized to determine the strategic market positions.

11 The

BCG model is based on the product life cycle theory that can be used to determine the level of priority to be given in the product portfolio of a business unit (Fleisher and Bensoussan, 2003).

12 The Harmonized System is a multipurpose international product nomenclature developed by the World Customs Organization, which classifies items into 21 categories, 97 groups and over 5,000 commodities. The classification system has been adopted by all member countries of the World Trade Organization for international trade.

International Food and Agribusiness Management Review

581

Laosutsan et al.

Volume 20, Issue 4, 2017

4. Results and discussion 4.1 The Revealed Comparative Advantage and Revealed Symmetric Comparative Advantage results

■■ Revealed Comparative Advantage of fresh or chilled vegetables (HS 0701-0714) In Table 1, the overall average RCA of fresh or chilled vegetables of Thailand between 2009-2013 was 0.61, which is less than 1, with an average annual growth rate of -1.41%. The result showed that Thailand has a competitive disadvantage in the fresh or chilled vegetables in the ASEAN market and relies little on the income from the export of fresh or chilled vegetables in this market. By comparison, several other countries exhibit a competitive advantage in exporting fresh or chilled vegetables in the ASEAN market, including Cambodia (RCA=1.38), Laos (RCA=8.25), Malaysia (RCA=1.56), Myanmar (RCA=21.72), Vietnam (RCA=2.18), Australia (RCA=2.69) and China (RCA=8.04).

Even though the overall RCA indicates Thailand’s comparative disadvantage in the export of fresh or chilled vegetables to the ASEAN market, the individual RCA results reveal the comparative advantage for Thailand (RCA>1) in four groups: 0703, 0710, 0711, 0714. These four groups are tomatoes, frozen vegetables, vegetables provisionally preserved, and manioc, arrowroot salem. Interestingly, Malaysia has a comparative advantage in almost all fresh or chilled vegetables listed, while some countries are in an unfavorable position of no comparative advantage in any fresh or chilled vegetables groups, such as Singapore.

Table 1. The average Revealed Comparative Advantage indexes of vegetable products belonging to Harmonized System codes 0701-0714 based on the 2009-2013 data of the 15 countries (adapted from The United Nations Commodity Trade Statistics Database (UN COMTRADE: https://comtrade.un.org)).

Country

Code 0701 0702 0703 0704 0705 0706 0707 0708 0709 0710 0711

Brunei Cambodia Indonesia Laos Malaysia Myanmar Philippines Singapore Thailand Vietnam Australia China Japan South Korea USA

1.01 0.00 2.58 0.00 1.16 0.11 0.00 0.95 0.21 0.48 26.19 22.00 0.01 0.08 9.22

0.00 0.29 0.00 0.00 0.00 0.00 0.19 0.53 2.29 0.00 0.01 72.07 4.86 1.76 1.52 0.00 4.50 0.00 0.00 1.42 0.00 0.01 0.13 0.16 0.10 1.13 0.08 0.49 5.17 0.38 0.49 0.42 2.50 0.25 21.00 13.04 0.00 0.00 0.01 0.00 0.00 0.03 0.01 0.03 0.02

0.00 0.05 0.00 0.00 0.44 1.05 0.01 26.16 4.36 2.67 0.00 0.06 0.00 0.00 0.20 0.34 0.16 0.03 0.06 1.90 4.20 29.45 1.83 22.62 0.00 0.01 0.00 0.04 1.04 0.15

0.00 0.00 0.08 0.01 5.05 0.00 0.00 0.00 0.13 0.03 0.08 0.00 0.00 0.00 0.00

0.00 0.19 1.46 0.69 3.75 0.53 0.00 0.09 0.25 0.09 0.09 5.22 0.00 0.00 0.00

0.00 0.07 0.33 0.04 2.73 0.02 0.08 0.29 0.68 4.11 0.97 1.52 0.04 0.33 0.31

0712 0713 0714

0.05 0.03 0.00 0.00 0.01 0.01 11.85 0.04 0.03 21.76 1.91 0.06 1.20 0.50 0.88 0.00 0.99 0.05 0.31 51.83 1.99 0.96 1.80 0.10 0.27 0.22 0.01 0.41 60.30 0.04 0.00 0.23 2.63 0.01 0.00 0.32 0.14 0.46 0.03 0.11 1.24 2.54 0.39 0.48 1.09 0.55 6.09 3.72 0.05 6.52 1.05 0.12 0.38 3.81 0.08 3.19 4.68 128.05 0.92 2.00 0.03 0.00 0.19 0.00 0.16 0.01 0.00 0.18 0.00 0.01 0.69 1.10 8.02 0.48 0.00

International Food and Agribusiness Management Review

582

Average year 2009-2013 Growth rate (% p.a.)

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0029 - Tuesday, August 22, 2017 12:35:02 PM - IP Address:24.21.169.207

The RCA is measured by comparing the ratio of total exports of goods in a country with a proportion of such products in the market. However, the RCA can only reveal whether the country has a comparative advantage in the export of goods. In other words, the RCA results of different countries are not for comparison purposes and thus it is difficult to infer from the results with respect to the competitive advantages between countries.

0.07 1.38 0.67 8.25 1.56 21.72 0.32 0.14 0.61 2.18 2.69 8.04 0.02 0.07 0.57

0.00 12.09 -4.74 40.30 1.25 -4.31 8.28 3.47 -1.41 12.19 5.01 -5.17 5.94 8.92 7.16

Laosutsan et al.

Volume 20, Issue 4, 2017

â&#x2013; â&#x2013; Revealed Comparative Advantage of preserved or processed vegetables (HS 2001-2009)

Further analysis indicates that Thailand, in fact, has a comparative advantage (RCA>1) in only four preserved or processed vegetables groups: 2005, 2007, 2008 and 2009. It is also shown that some other countries have an advantage in the export of other preserved or processed vegetable groups. In particular, the US has a comparative advantage in almost all preserved or processed vegetables. On the contrary, other countries are in a position with no comparative advantage in any preserved or processed vegetables, including Brunei, Cambodia, Myanmar, Vietnam, Japan and South Korea.

To mitigate the issue of asymmetry associated with the RCA, the RSCA was utilized in the subsequent stage of the analysis and the results presented in Tables 3 and 4, where a positive sign denotes an RSCA>0, indicating a comparative advantage and a negative sign for otherwise. The RSCA results revealed eight vegetable products (identical to the RCA analysis) with comparative advantage (four each for the fresh and preserved vegetable groups under study): onion, garlic and leeks (HS 0703, RSCA=0.06); frozen vegetables (HS 0710, RSCA=0.11); provisionally preserved vegetables (HS0711, RSCA=0.44); manioc, arrowroot salem (yams), etc. (HS 0714, RSCA=0.04); preserved vegetable (excluding frozen) (HS 2005, RSCA=0.42); jam, fruit jellies and marmalades (HS 2007, RSCA=0.37); preserved fruits (HS 2008, RSCA=0.29); and unfermented fruit and vegetable juices (HS 2009, RSCA=0.52).

Country

Brunei Darussalam Cambodia Indonesia Laos Malaysia Myanmar Philippines Singapore Thailand Vietnam Australia China Japan South Korea USA

Code 2001

2002

2003

2004

2005

2006

2007

2008

2009

0.00 0.00 1.05 0.00 1.13 0.97 0.09 1.12 0.85 0.71 1.99 1.63 0.23 0.42 1.03

0.03 0.84 0.03 0.00 1.06 0.15 0.00 1.63 0.63 0.57 1.42 43.86 0.00 0.00 16.11

0.00 0.00 0.44 0.00 0.63 0.17 0.27 1.74 0.40 0.98 0.06 60.92 0.05 0.06 1.31

0.00 0.00 0.21 0.08 4.16 0.01 0.02 0.29 0.26 0.35 2.27 1.31 0.15 0.03 82.84

0.17 0.00 0.01 0.00 1.78 0.16 0.14 0.53 2.45 0.31 0.30 2.66 0.08 0.27 2.67

0.00 0.00 0.11 1.14 1.83 0.04 0.42 1.10 0.95 0.14 0.18 6.84 0.01 0.02 0.88

0.08 0.00 0.13 0.00 1.12 0.08 5.77 0.28 2.16 2.33 2.25 0.32 0.01 0.01 0.97

0.23 0.00 0.40 2.23 0.79 0.14 8.56 0.50 1.82 0.39 0.38 2.80 0.03 0.08 1.05

0.52 0.04 0.25 0.00 1.27 0.00 1.40 0.27 3.14 0.59 1.32 0.08 0.03 0.04 1.70

International Food and Agribusiness Management Review

583

Growth rate (% p.a.)

Table 2. The average Revealed Comparative Advantage indexes of vegetable products belonging to Harmonized System codes 2001-2009 based on the 2009-2013 data of the 15 countries (adapted from The United Nations Commodity Trade Statistics Database (UN COMTRADE: https://comtrade.un.org)).

Average year 2009-2013

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0029 - Tuesday, August 22, 2017 12:35:02 PM - IP Address:24.21.169.207

In Table 2, the overall average RCA of preserved or processed vegetables of Thailand between 2009-2013 was 2.23, with an average annual growth rate of 0.22%. The result showed that Thailand has a competitive advantage in the preserved or processed vegetables in the ASEAN market and relies considerably on the income from the export of preserved or processed vegetables in this market. Other countries with a competitive advantage in exporting preserved or processed vegetables in the ASEAN market include Malaysia (RCA=1.17), the Philippines (RCA=4.79), China (RCA=2.53) and the USA (RCA=2.88).

0.28 0.00 0.02 0.00 0.27 -0.55 0.99 -11.87 1.17 -1.16 0.10 0.00 4.79 6.29 0.45 -0.71 2.23 0.22 0.57 -24.08 0.79 -3.79 2.53 1.91 0.04 2.02 0.09 -1.01 2.88 5.68

Laosutsan et al.

Volume 20, Issue 4, 2017

Table 3. The Revealed Symmetric Comparative Advantage (RSCA) results for vegetable products belonging to Harmonized System codes 0701-0714.1

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0029 - Tuesday, August 22, 2017 12:35:02 PM - IP Address:24.21.169.207

Country

Code

Brunei Darussalam Cambodia Indonesia Laos Malaysia Myanmar Philippines Singapore Thailand Vietnam Australia China Japan South Korea USA 1 A positive

07

0701 0702 0703 0704 0705 0706 0707 0708 0709 0710 0711 0712 0713 0714

– + – + + + – – – + + + – – –

+ – + – + – – – – – + + – – +

– – – – + – – – – – – – – – –

– – – – + + + – + + – + – – –

– – + + + – – – – – + + – – –

– – – – + – – – – – + + – – +

– – + + + – – – – + + + – – –

– – – – + – – – – – – – – – –

– – + – + – – – – – – + – – –

– – – – + – – – – + – + – – –

– – + – + – – – + – + + – – –

– + – – – – – – + + – + – – +

– – – – + – + – – + – + – – +

– – – – – + – – – – + – – – –

– + – + – – – – + + – + – – –

sign denotes an RSCA>0 and a negative sign otherwise.

Table 4. The Revealed Symmetric Comparative Advantage (RSCA) results for vegetable products belonging to Harmonized System codes 2001-2009.1 Country

Code

Brunei Darussalam Cambodia Indonesia Laos Malaysia Myanmar Philippines Singapore Thailand Vietnam Australia China Japan South Korea USA 1 A positive

20

2001

2002

2003

2004

2005

2006

2007

2008

2009

– – – – + – + – + – – + – – +

– – + – + – – + – – + + – – +

– – – – + – – + – – + + – – +

– – – – – – – + – – – + – – +

– – – – + – – – – – + + – – +

– – – – + – – – + – – + – – +

– – – + + – – + – – – + – – –

– – – – + – + – + + + – – – –

– – – + – – + – + – – + – – +

– – – – + – + – + – + – – – +

sign denotes an RSCA>0 and a negative sign otherwise.

Out of the 23 vegetable product groups under study, the eight aforementioned Thai vegetable product groups are poised to compete for a larger share of the ASEAN market. Nonetheless, to fulfill that ambition, a comprehensive long term strategic plan must be formulated. In addition, the competitiveness of Thai vegetable products should be further enhanced through the adoption of the guideline recommendations proposed in this research.

International Food and Agribusiness Management Review

584

Laosutsan et al.

Volume 20, Issue 4, 2017

4.2 The clustering results

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0029 - Tuesday, August 22, 2017 12:35:02 PM - IP Address:24.21.169.207

The SPSS (IBM, Armonk, NY, USA) dendrograms using the Ward Method13 of the fresh (HS 0701-0714) and preserved (HS 2001-2009) vegetable product groupings are presented in Figures 1 and 2, respectively, in which every case (i.e. country) has been brought into one single cluster.

The results identify two clusters (i.e. the clusters without and with comparative advantage) each under the two vegetable product groupings. Under the fresh vegetable product grouping, the cluster with comparative disadvantage consists of Brunei, Japan, South Korea, the Philippines, Singapore, Indonesia, Thailand and the USA, whereas the cluster with comparative advantage consists of Cambodia, Vietnam, Australia, Laos, Myanmar, China, Malaysia (Figure 1). Meanwhile, under the preserved vegetable grouping, the cluster with 13 SPSS

Dendrogram and Wardâ&#x20AC;&#x2122;s Hierarchical Clustering Method: in this method all possible pairs of clusters are combined and the sum of the squared distances within each cluster is calculated, and 0-25 is the standard scale of the dendrogram. This is then summed over all clusters and the combination that gives the lowest sum of squares is chosen. This method tends to produce clusters of approximately equal size, which is not always desirable because of the sensitivity to outliers. Despite this, it is one of the most popular methods, along with the average linkage method (Rosie, 2007).

HIERARCHICAL CLUSTER ANALYSIS Dendogram using Ward method, rescaled distance cluster combine for cluster country groups inability to export fresh vegetable products CASE 0 5 10 15 20 25 Label +---------+---------+---------+---------+---------+ Brunei Japan Korea South Philippines Singapore Indonesia Thailand USA Cambodia Vietnam Australia Laos Myanmar China Malaysia

Figure 1. Clustering of country-groups by the ability to export fresh vegetable products (Harmonized System 0701-0714).

HIERARCHICAL CLUSTER ANALYSIS Dendogram using Ward method, rescaled distance cluster combine for cluster country groups inability to export fresh vegetable products CASE 0 5 10 15 20 25 Label +---------+---------+---------+---------+---------+ Japan Korea South Brunei Cambodia Laos Indonesia Myanmar Singapore Vietnam Malaysia Australia Thailand USA Philippines China

Figure 2. Clustering of country-groups by the ability to export processed vegetable products (Harmonized System 2001-2009).

International Food and Agribusiness Management Review

585

Laosutsan et al.

Volume 20, Issue 4, 2017

the comparative disadvantage comprises Japan, South Korea, Brunei, Cambodia, Laos, Indonesia, Myanmar, Singapore, Vietnam, while that with the comparative advantage consists of Malaysia, Australia, Thailand, USA, the Philippines and China (Figure 2).

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0029 - Tuesday, August 22, 2017 12:35:02 PM - IP Address:24.21.169.207

■■ Fresh and chilled vegetables For the fresh vegetable product grouping, the analysis outcomes indicate that the countries with the comparative disadvantage with regard to the export of fresh vegetables to the ASEAN market are Brunei, Japan, South Korea, the Philippines, Singapore, Indonesia, Thailand and the USA. The finding could be attributed to the high domestic demand of these countries and to the inadequate government support on export. Meanwhile, the countries with the comparative advantage include Cambodia, Vietnam, Australia, Laos, Myanmar, China and Malaysia. In fact, several ASEAN countries, including Laos and Vietnam, have instituted the policy to promote the production and exportation of quality fresh vegetables.

■■ Preserved and processed vegetables For the preserved vegetables, Malaysia, Australia, Thailand, USA, the Philippines and China possess the considerable potential to supply to the ASEAN market. This is attributable to the deployment of modern technology and scientific methods in the production.

4.3 Boston Consulting Group market position The BCG matrices relevant to the fresh (HS 0701-0714) and preserved (HS 2001-2009) vegetable products of the 15 countries under study are respectively illustrated in Figures 3 and 4. It could be observed that most ASEAN countries are in either the ‘dog’ or ‘question mark’ quadrant of the matrix, suggesting that their exports of vegetables experience a dual-low dilemma in which both the market share and market growth are low. This phenomenon gives rise to the subsequent poor profitability and unfavorable export prospects.

To counter, these less competitive countries must explore new markets for their vegetable products and at the same time increase the degree of penetration of the current markets. Operational streamlining is also imperative so as to lower the costs.

Figure 3. The Boston Consulting Group (BCG) matrix for fresh vegetable products (Harmonized System 0701-0714) of the 15 countries for years 2004-2013. AU = Australia; CN = China; ID = Indonesia; KH = Cambodia; LA = Laos; MY = Malaysia; MM = Myanmar; PH = Philippines; SG = Singapore; TH = Thailand; US = United States; VN = Vietnam.

International Food and Agribusiness Management Review

586

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0029 - Tuesday, August 22, 2017 12:35:02 PM - IP Address:24.21.169.207

Laosutsan et al.

Volume 20, Issue 4, 2017

Figure 4. The Boston Consulting Group (BCG) matrix for preserved vegetable products (Harmonized System 2001-2009) of the 15 countries for years 2004-2013. AU = Australia; CN = China; ID = Indonesia; KH = Cambodia; LA = Laos; MY = Malaysia; MM = Myanmar; PH = Philippines; SG = Singapore; TH = Thailand; US = United States; VN = Vietnam.

In Figure 3, the Thai fresh and chilled vegetables from the years 2004 2013 were in the dog quadrant of the matrix, with a market share of 0.04 and market growth of 16.41. With its market share of 6.88 and market growth of 18.48, China is in an enviable position with both high relative market share and a high market growth rate. While China has to adopt a strategy to maintain and protect the market share, other players could implement the segmentation, targeting and positioning to expand their shares of the market (Supplementary Table S4). Meanwhile, faced with a slow market growth rate and a small market share, Thailand could increase the export volume through the exploration of new untapped markets and the increased penetration of the current markets.

Figure 4 illustrates the BCG matrix for the preserved vegetable products (HS 2001-2009) of the same 15 countries. The Thai preserved vegetables during the years 2004-2013 were in the question mark quadrant of the matrix, with a market share of 0.25 and market growth of 17.83. This is considerably lower when compared with China whose market share and market growth were 1.88 and 26.40 (Supplementary Table S4). Interestingly, the relative market share and market growth of Thailand and the US are in the same range.

To increase the export of preserved vegetables, Thailand must be innovative in the development of products and might consider forging an alliance with overseas entrepreneurs as a way to explore new markets. Borrowing certain initiatives from other countriesâ&#x20AC;&#x2122; trade policies could be useful for Thailand. For example, China has a policy to support the domestic consumption while Thailand relies heavily on export to drive its economy.

In addition, Chinaâ&#x20AC;&#x2122;s agricultural development plan emphasizes the reform of the agricultural sector using innovative and modern technologies. At the same time, the Chinese government is implementing policies to maintain the quality of agricultural land and increase crop yields to achieve food security. Thailand could thus incorporate certain effective practices in the formulation of the countryâ&#x20AC;&#x2122;s 5-year agricultural development plan.

4.4 Non-tariff barriers between ASEAN countries Even with a complete removal of the trade tariffs, exporting countries are nevertheless faced with the imposition of NTBs by other countries, especially those for agricultural products. The current NTBs can be classified into eight main groups: (1) anti-dumping; (2) counter-vailing duty; (3) safeguard; (4) sanitary and phytosanitary; (5) technical barrier to trade; (6) environment; (7) labor; and (8) others, e.g. purchasing International Food and Agribusiness Management Review

587

Laosutsan et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0029 - Tuesday, August 22, 2017 12:35:02 PM - IP Address:24.21.169.207

by state, import monopolies, rule of origin. Within the ASEAN grouping, Thailand needs to deal with the following NTBs from other member nations: Indonesia: the main NTB to doing business in the archipelago country of Indonesia is an over-abundance of Chinese intermediaries, through whom business negotiations and transactions are undertaken. In addition, the country has recently closed down half of the ports from originally eight to four ports, resulting in an overcrowded traffic and long queues at the remaining ports, causing damage to the perishables imports. The situation is anticipated to exacerbate with the full integration of the AEC. With regard to food products imports, the country requires that the production and distribution of foods strictly adhere to the Halal regulations.

Malaysia and Singapore: Thailand must enter talks with neighboring Malaysia on the logistics issue that would allow freight trucks and trains from Thailand to access and pass through its borders with minimal friction. In addition, Thai exporters are facing discriminatory treatments by many Malaysian intermediaries/ importers, making it challenging for Thai products to reach Malaysian consumers. On the other hand, the access to Singaporeâ&#x20AC;&#x2122;s market is open but the competition is fierce. The Singaporeans have high purchasing power and demand good quality products; thus, the export of Thai vegetables to the city-state must be of superior quality.

Cambodia and Vietnam: a majority of Cambodians perceive that Thai vegetable products are replete with insecticides and hazardous chemicals. To address the issue and establish a strong foothold in Cambodia, Thai entrepreneurs could either export organic produce to or invest in organic farming using modern technologies in neighboring Cambodia. Meanwhile, in the case of Vietnam, the government sector has established a clear policy on the future directions of agricultural development and encouraged the adoption of the Good Agricultural Practice system. Furthermore, in contrast to Thailand, the Vietnamese government has in place procedures and measures to assist agriculturists to improve their production yields, to support local import and export operators with trade information and advice, and to set up a central market system for vegetable products.

Myanmar and Laos: numerous entry barriers to the Myanmar and Lao markets await Thai exporters of vegetable products, including poor logistics, inefficient domestic distribution systems, government interventions, as well as cultural differences. In the case of Myanmar, the country has a policy that promotes the development of the local agricultural sector by foreign investments through the adoption and deployment of modern production and processing technologies to meet the domestic demand. Despite Myanmarâ&#x20AC;&#x2122;s joint-venture requirement for a foreign investment, the opportunities are plentiful for Thailand to invest in the agribusiness and expand the market in Myanmar. For Laos, its government encourages investment in agribusiness that deploys modern, environmentally-friendly production and processing technologies, thereby representing an opportunity for Thai organic vegetable products. Lao laborers are nonetheless limited in both quantity and quality.

The Philippines: despite positive attitudes among Filipinos toward vegetable products from Thailand, the distribution of goods to more than 7,000 islands and the decrepit infrastructure pose an enormous challenge to Thai exports.

In general, the full integration of the AEC is beneficial to Thailand with regard to a market expansion opportunity in which the consumer base increases from currently 67 million to over 600 million. Besides, the elimination of trade tariffs and NTBs would positively contribute to a larger export market. Nevertheless, the AEC brings with it the fiercer intra-regional competition, in particular, which is relevant to this research, the vegetable products with either comparable or inferior comparative advantage relative to those of other AMSs. To reap the benefits of the full integration of the AEC, this research has thus put forward the following recommendations to systematically improve Thai vegetable productsâ&#x20AC;&#x2122; competitiveness among the ASEAN member countries.

International Food and Agribusiness Management Review

588

Laosutsan et al.

Volume 20, Issue 4, 2017

1. Export entrepreneurs must fully understand the terms and conditions of the trade agreements and duly comply with the terms and conditions.

2. The government agencies must gain a thorough understanding of the NTBs and NTMs of all trading partners and formulate the action plans in response to these non-tariff trade strategies.

3. Informal collaboration and alliance with importers (i.e. buyers) of Thai vegetable products should be promoted to mitigate the impacts of NTB and NTM.

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0029 - Tuesday, August 22, 2017 12:35:02 PM - IP Address:24.21.169.207

5. Conclusions This research has evaluated the competitiveness in 14 fresh and 9 preserved vegetable products exports of Thailand in relation to Australia, China, Japan, South Korea, the USA and nine ASEAN member countries using the RCA and RSCA indices. The study found eight vegetable product groups (four each for fresh and preserved vegetable product groups) with a comparative advantage vis-Ă -vis those of the other countries in the study. In addition, Thailand seems to be enjoying more benefit from the preserved or processed vegetable products since these products require certain levels of technology, which some rival ASEAN countries lack.

Thus, the country should focus more on the preserved or processed vegetable products for export to the AEC. To enhance the competitiveness of the fresh vegetable products, Thailand should attach greater importance to the quality since Thai products are known more for quality than quantity (low prices). Furthermore, the issues of overreliance on the chemical insect repellants and chemical residues on the products should be tackled.

The research findings also show that most ASEAN countries are in either the dog or question mark quadrant of the BCG matrix, suggesting that their exports of vegetables experience a dual-low dilemma in which both the market share and the market growth are low. Moreover, the findings reveal that Cambodia, Vietnam, Australia, Laos, Myanmar, China and Malaysia possess the comparative advantage with regard to the fresh and chilled vegetables export in the ASEAN market. For the preserved and processed vegetables, Malaysia, Australia, Thailand, US, the Philippines and China possess the considerable potential to supply to the ASEAN market with such agriculture products.

Essentially, the full integration of the AEC would benefit the Thai fruits and vegetables sector due to a larger market with a combined population over 600 million, the enhanced regional cooperation, the improved economies of scale, and dynamism of the bloc members. Moreover, the integration would bring about the improved or innovative forms of supply-chain coordination, thereby further facilitating the bilateral and multilateral trade activities between the member states.

Supplementary material Supplementary material can be found online at https://doi.org/10.22434/IFAMR2016.0029.

Table S1. Harmonized System nomenclature.

Table S2. The intra-ASEAN exports of fresh and chilled vegetables (HS code 07) from 2007-2014.

Table S3. The intra-ASEAN exports of preserved and processed vegetables (HS code 20) from 2007-2014.

Table S4. The Boston Consulting Group matrix (BCG) analysis of the fresh (HS 0701-0714, BCG07) and preserved (HS 2001-2009, BCG20) of the 15 countries for the years 2004-2013.

Figure S1. Proportions of global exports of fresh and processed vegetables in 2011 (by economic bloc).

Figure S2. Proportions of global imports of fresh and processed vegetables in 2011 (by economic bloc).

References Anon. 2011. What is trade barrier? Definition and meaning. BusinessDictionary. Available at: http://tinyurl. com/7dxtcq3.

Anon, 2012. K-Econ analysis. Kasikorn Research Center. Available at: http://tinyurl.com/my8u9qv.

Anon, 2014. Thailand foreign Agricultural Trade Statistics 2013. Office of Agricultural Economics, Bangkok, Thailand.

International Food and Agribusiness Management Review

589

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0029 - Tuesday, August 22, 2017 12:35:02 PM - IP Address:24.21.169.207

Laosutsan et al.

Volume 20, Issue 4, 2017

Balassa, B. 1965. Trade liberalization and ‘revealed’ comparative advantage. The Manchester School of Economics and Social Studies 33: 99-123.

Benja, S., and N. Kingkorn, 2004. Thai-Chinese free trade agreement: who benefits? In: Free trade agreements: impact in Thailand. FTA Watch, Muang, Thailand, pp. 77-113.

Calvo Pardo, H. F., C.L. Freund and E. Ornelas. 2009. The ASEAN free trade agreement: impact on trade flows and external trade barriers. Available at: http://tinyurl.com/lxfwa2f.

Cheyroux, B. 2003. Fruits and vegetables in Thailand’s rice bowl: the agricultural development of poldered raised bed systems in the Damnoen Saduak area. Perspectives on social and agricultural change in the Chao Phraya Delta In: Thailand’s rice bowl, edited by F. Molle and T. Srijantr. White lotus press, Bangkok, Thailand, pp. 157-176.

Chia, S. 2013. The ASEAN economic community: progress, challenges, and prospects. Available at: http:// tinyurl.com/kvzxdn6.

Danupon, A., J.P. Gander, R. Somchai and E. Stephen. 2009. ASEAN FTA, distribution of income, and globalization. Journal of Asian Economics 20: 327-335.

Dollar, D., and A. Kraay. 2001. Trade, growth, and poverty. World Bank, Development Research Group, Macroeconomics and Growth. Available at: http://tinyurl.com/kj3x54c.

Fleisher, C S. and B.E. Bensoussan. 2003. Strategic and competitive analysis: methods and techniques for analyzing business competition. Prentice Hall, Upper Saddle River, NJ, USA, pp. 457.

Hinloopen, J. and C. Marrewijk. 2008. Empirical evidence of the Hillman condition for revealed comparative advantage: 10 stylized facts. Applied Economics 40: 2313-2328.

International Monetary Fund (IMF). 2001. Global trade liberalization and the developing countries. Available at: http://tinyurl.com/6woruwj.

Kawai, M. and G. Wignaraja. 2011. Asia’s free trade agreements: how is business responding? Edward Elgar Publishing, Cheltenham, UK. Ken, I. 2014. Impact of liberalization and improved connectivity and facilitation in ASEAN. Journal of Asian Economics 35: 2-11.

Kotler, P., 2003. Strategic marketing management, 4th edition. Prentice Hall, New York, NY, USA. Krippendorff, K. 1980. Clustering. In: Multivariate techniques in human communication research, edited

by P.R. Monge and J.N. Cappella. Academic Press, New York, NY, USA, pp. 259-308.

Nguyen, N. 2014. Thai workforce-Ready for ASEAN Economic Community 2015. University of the Thai Chamber of Commerce, Krung Thep Maha Nakhon, Thailand.

Prachason, S. and F.T.A Watch. 2009. Impact of FTAs on agriculture: issues in food security and livelihood.

In: IDEAs-GSEI-ITD Asian regional workshop on ‘Free Trade Agreements: towards inclusive trade policies in post-crisis Asia’, Bangkok, Thailand, pp. 8-9.

Rosie, C. 2007. Statistic: cluster analysis. Loughborough University, Loughborough, UK. Sally, R. 2007. Thai trade policy: from non-discriminatory liberalisation to FTAs. The World Economy 30: 1594-1620.

Sanchita, B., M. Jayant, S. Rodolfo and L. Omkar. 2013. The ASEAN economic community: a work in progress.

Asian development bank, Institute of Southeast Asian Studies, ISEAS publishing, Singapore, pp. 15.

Sandee, S., 2013. Thai agricultural products in China: problems and prospects. Thai data center business, Shanghai, China.

Sarasin, V. 2011. Vegetable industry will be the next future if development was the right way, the farmers are rich and wealthy nation. Available at: http://tinyurl.com/nyg78sc.

Tingjun, P. and L. Thomas. 2006. An economic analysis of the impacts of trade liberalization on Asian dairy market. Food Policy 31: 249-259.

Urata, S. 2004. Towards an East Asia free trade area. Available at: http://tinyurl.com/kwqdagq.

Vollrath, T.L. 1991. A theoretical evaluation of alternative trade intensity measures of revealed comparative advantage. Weltwirtschaftliches Archiv 127: 263-80.

Widodo, T. 2009. Comparative Advantage: theory, empirical measures and case studies. Review of Economic and Business Studies 4: 57-82.

Wind, Y. and V. Mahajan. 1981. Designing product and business portfolios. Harvard Business Review 59: 155-165.

International Food and Agribusiness Management Review

590

OPEN ACCESS International Food and Agribusiness Management Review Volume 20 Issue 4, 2017; DOI: 10.22434/IFAMR2016.0075

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0075 - Tuesday, August 22, 2017 12:36:02 PM - IP Address:24.21.169.207

Received: 17 March 2016 / Accepted: 14 April 2017

The impact of Mexican competition on the U.S. strawberry industry RESEARCH ARTICLE Dong Hee Suha, Zhengfei Guan b, and Hayk Khachatryanc aAssistant

Professor, Department of Food and Resource Economics, Korea University, 145 Anam-ro, Seongbuk-gu, 02841 Seoul, Republic of Korea

bAssistant

Professor, Gulf Coast Research and Education Center & Food and Resource Economics Department, University of Florida, Gainesville, FL 32611, USA

cAssistant

Professor, Mid-Florida Research and Education Center & Food and Resource Economics Department, University of Florida, Gainesville, FL 32611, USA

Abstract This paper models the U.S. strawberry market and examines how increasing imports from Mexico affect the prices and shipment values of California and Florida winter strawberries. The Synthetic Inverse Demand System is used to quantify the impact of Mexican shipments on the prices of strawberries. The estimation results indicate that market prices are responsive to supply from each of the three sources, suggesting an integrated, competitive national market. The simulation results suggest that rapidly growing Mexican shipments will cause large losses to the U.S. strawberry industry, posing challenges to the sustainability and survival of the industry, particularly that of the Florida industry. Policy implications and recommendations for the industry are discussed.

Keywords: competitiveness, Mexican competition, NAFTA, strawberry market, sustainability, synthetic inverse demand system JEL code: Q11, Q13, Q18 Corresponding author: guanz@ufl.edu

Š 2017 Suh et al.

591

Suh et al.

Volume 20, Issue 4, 2017

The North American Free Trade Agreement (NAFTA) has enabled free movement of commodities in North America and created greater market integration between the U.S. and Mexican produce industries.

Geographic proximity and lower cost of production have greatly boosted Mexican exports to the U.S. under NAFTA. In recent years, growing imports from Mexico have created great challenges to the U.S. domestic produce industry. The literature has pointed out that rapidly growing imports have the potential to displace domestic production (Burfisher et al., 2001; Young, 1988). Many domestic produce sectors, such as tomatoes, cucumbers, bell peppers, and strawberries, have found it difficult to compete with produce imported from Mexico (Asci et al., 2016; Wu et al., in press; Zahniser et al., 2015). This study focuses on the strawberry industry to highlight the increasing competition and its impact on the U.S. domestic industry.

As a high-value fruit crop, the total U.S. production value of strawberries amounted to 2.8 billion dollars in 2014, which was more than two times higher than that of fresh tomatoes, one of the highest valued vegetable crops (Wu et al., in press). According to the National Agricultural Statistics Service (NASS) of the U.S. Department of Agriculture (USDA), approximately 3.0 billion pounds of strawberries were produced in 2014 (USDA/NASS, 2015). The leading strawberry-producing states are California and Florida. The total amount of strawberries produced in the two states account for about 98% of total U.S. production. In 2014, California produced nearly 2.8 billion pounds of strawberries from 41,500 acres. Florida produced approximately 0.2 billion pounds of strawberries from 10,900 acres.

In addition to the production of California and Florida, Mexico is another major supplier of strawberries in the U.S. market. According to the Foreign Agricultural Service of the USDA, the imported strawberries from Mexico account for about 95% of total imported strawberries in the U.S. market. In 2014, about 300 million pounds were imported from Mexico between November and April. The three suppliers compete in the winter strawberry market. Figure 1 shows the seasonal differences in strawberry production across the three competitors. Florida produces only in the winter season, while California produces year round. Californiaâ&#x20AC;&#x2122;s winter production is mainly in the southern region. Mexican production is mainly in the winter season, similar to that of Florida. The average market shares of the three competitors over 2010-2014 during the winter months (December through March) were 35, 39 and 26% for California, Florida, and Mexico, respectively.

California

Florida

Mexico

300

250

200 Million pounds

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0075 - Tuesday, August 22, 2017 12:36:02 PM - IP Address:24.21.169.207

1. Introduction

150

100

50

0

Jan

Feb

Mar

Apr

May

Jun

Jul

Aug

Sep

Oct

Figure 1. Average monthly shipments of strawberries, 2010- 2014.

International Food and Agribusiness Management Review

592

Nov

Dec

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0075 - Tuesday, August 22, 2017 12:36:02 PM - IP Address:24.21.169.207

Suh et al.

Volume 20, Issue 4, 2017

In recent years, the U.S. strawberry industry has become increasingly concerned about the strong competition from Mexico. Mexico has surpassed Florida as the largest supplier of winter strawberries in the U.S. market since 2012. In a period of 10 years (2004-2014), imports from Mexico increased fourfold, creating tremendous pressure on Florida growers. As a result, the production value of Florida strawberries slumped from 370 million dollars in 2010 to 201 million dollars in 2012 (USDA/NASS, 2013). The competition from Mexico, along with labor shortages and increasing production costs (e.g. Baker, 2004; Carter et al., 2005; Goodhue et al., 2005; Johnson, 2014; Norman, 2005), is posing a great challenge to the U.S. winter strawberry industry.

In Florida, the largest U.S. winter strawberry producing state, the labor cost of domestic strawberries is about $9,000 per acre, which accounts for about 40% of farm-gate sales (Guan et al., 2015). Moreover, it is increasingly difficult to find enough labor in the harvesting season as more Mexican immigrants are returning to Mexico due to the increased employment opportunities in the Mexican economy (Taylor et al., 2012) and stricter immigration policies.

Admittedly, the U.S. producers have comparative advantages in breeding technology and have better infrastructure and extension services. But the growing production capacity of the Mexican industry has kept putting pressure on the U.S. strawberry industry. Over the years, the Mexican government has been promoting and subsidizing its horticulture industry, which has intensified since 2009 with the introduction of its strategic project for protected agriculture (Victoria et al., 2011). In 2013, Mexico proposed to further double the production capacity of its strawberry industry in the coming years (Guan et al., 2015). A significant increase in Mexican production capacity will pose further challenges to the U.S. winter strawberry industry, particularly the Florida industry.

In the literature, economic analyses that focus on strawberries are limited. Wu et al. (2015) identified the optimal yield distribution over the season to maximize profit for Florida growers given California and Mexicoâ&#x20AC;&#x2122;s supply pattern, providing information to support breeders in developing cultivars of more economic value to growers. Lee and Kennedy (2016) conducted a partial equilibrium analysis to study the trade creation and diversion effect of NAFTA in the strawberry market. The present paper investigates the impact of imports on the U.S. strawberry market and industry sustainability. The contribution of this paper is twofold. First, the paper models the effects of shipments of California, Florida, and Mexican strawberries on shipping prices the industry receives. Second, this paper further quantifies how growing imports from Mexico will affect the U.S. strawberry industry by simulating shipping prices and shipment values (market shares) of U.S. strawberries under different growth scenarios. This information is then used to assess the loss caused to the industry under these scenarios.1 The empirical findings in this paper will provide strawberry producers and policy makers with important insights on the challenges and the sustainability of the U.S. strawberry industry. The case of the strawberry industry will also shed light on the impact of Mexican competition on the U.S. fresh produce industry under the NAFTA. The paper is organized as follows. The next section presents the Synthetic Inverse Demand System (SIDS) approach used in this study, and discusses its application to the U.S. strawberry market. The following sections present data descriptions and estimation results of the scale elasticities and price flexibilities, followed by simulations of the effects of Mexican shipments on the prices and shipment values of U.S. domestic strawberries. The final section concludes and discusses the sustainability and the future of the U.S. strawberry industry.

1

The U.S. is a large importer of strawberries; its exports are small relative to the total imports and its total production. Export to Mexico accounts for roughly 1% of the U.S. production (or 10% of U.S. total export) over the last few years according to the U.S. Department of Commerce statistics.

In this study, we focus on fresh strawberries. There are processed or frozen strawberries, but the market share is small and economically insignificant compared to fresh strawberries.

International Food and Agribusiness Management Review

593

Suh et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0075 - Tuesday, August 22, 2017 12:36:02 PM - IP Address:24.21.169.207

2. The synthetic inverse demand system In general, inverse demand systems are considered suitable to estimate the demand for fresh food due to its perishable nature (Brown et al. 1995; Chambers and McConnell, 1983; Eales and Unnevehr, 1993, 1994; Grant et al., 2010; Huang, 1988; Matsuda, 2005; Park and Thurman, 1999; Park et al., 2004). In most inverse demand systems, quantities supplied are considered to be predetermined by production at the market level.2 Strawberries are highly perishable with a limited shelf life. After harvest, they are sorted and stored in cold rooms at 0-10 °C, and then sold within approximately 7-10 days to meet the commercially-acceptable quality (Ayala-Zavala et al., 2004; Hernandez-Munoz et al., 2008). In Florida, freshly picked strawberries are usually shipped within 24 hours. For the demand analysis of perishable strawberries, the SIDS developed by Brown et al. (1995) is used to examine the responsiveness of the prices of perishable strawberries to the changes in quantities. In particular, the SIDS is used to assess the effects of the shipments of California, Florida, and Mexican strawberries on the prices in terms of scale elasticities and price flexibilities. The SIDS nests different inverse demand systems and allows for hypothesis testing among systems in empirical applications.

This section briefly presents the SIDS approach.

Following Brown et al. (1995), we denote p=(p1,...,pn) as the vector of nominal prices, q=(q1,...,qn)’ as the vector of quantities consumed, m=p’q as the total expenditure or income, and π=(π1,..., πn)’≡p/m as the normalized price vector. A consumer is assumed to maximize the utility, u=u(q1,...,qn) subject to the budget constraint, m=p’q. The compensated inverse demand function is derived from a distance function, d(u,q) where u is the utility level and q is a consumption bundle of n commodities. The distance function is assumed linearly homogeneous, concave, non-decreasing in quantities, and decreasing in utility, which determines whether quantities decrease or increase to reach a specific utility level. Differentiating the distance function with respect to quantity yields πi =

∂d(u,q) = πi (u,q) (1) ∂qi

for i=1,...,n. Totally differentiating Equation 3 yields dπi =

n ∂π ∂πi i du + Σ qj (2) ∂u ∂q j=1 j

for i=1,...,n. In Equation 2, the first term represents the scale effects, and the second term represents the Antonelli substitution effects. When we define q* as a reference bundle so that q=kq* where k is a positive scalar, the first term becomes πi =

∂lnπ i n Σ s dlnqj (2a) ∂lnk j=1 j

where si=πiqi is the expenditure or budget share of commodity i (see Brown et al. (1995) for detailed derivation). Multiplying Equation 2 by qi yields the Rotterdam Inverse Demand System (RIDS) proposed

by Barten and Bettendorf (1989) as n

sidlnπ i = αidlnQ + Σ αijdlnqj

(3)

j=1

for i=1,...,n where dlnQ≡ Σ j s j dlnqj is the Divisia volume index, and the parameters, αi and αij, represent the scale and substitution effects, respectively. The regularity conditions are imposed on these parameters: adding up (Σ iαi=–1 and Σ iαij=0), homogeneity (Σ jαij=0), and symmetry (αij=αji).

2 The perishable nature of strawberries allows us to regard the quantities produced as the quantities available to be consumed in the market. When the quantities supplied are considered to be predetermined, the prices of strawberries are determined by the quantities demanded by consumers.

International Food and Agribusiness Management Review

594

Suh et al.

Volume 20, Issue 4, 2017

Other parameterizations also generate different inverse demand systems from the RIDS. Let dlnP≡ Σ i s i lnpi denote the Divisia price index so that dlnm=dlnP+dlnQ. Adding sidlnQ to both sides of Equation 3 yields the Laitinen-Theil Inverse Demand System (LTIDS) as

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0075 - Tuesday, August 22, 2017 12:36:02 PM - IP Address:24.21.169.207

s i dln

n

( pP ) = β dlnQ + Σ α dlnq (4) i

i

j=1

ij

j

for i=1,...,n where βi=si+αi. Equation 4 follows from the relationship of dlnπ i +dlnQ=dlnp i –dlnP. The LTIDS is a variant of the RIDS with αi= βi–si.

In addition, the differential form of the linear approximation of the Almost Ideal Inverse Demand System (AIIDS) is derived by adding si(dlnqi–dlnQ) to both sides of Equation 4. Since si(dlnp i +dlnq i –dlnm)=ds i , the AIIDS proposed by Eales and Unnevehr (1994) is written as ds i =β i dlnQ+

n

Σ β ijdlnqj (5)

j=1

for i=1,...,n. In Equation 5, βij=αij+si(δij–sj) where δij denotes the Kronecker delta, which is equal to unity if i=j and zero otherwise. The AIIDS is a variant of the RIDS with αi=βi–si and αij=βij–si(δij–sj).

Lastly, another differential inverse demand system is obtained by subtracting sidlnQ from both sides of Equation 5 so that n

ds i –s i dlnQ=α i dlnQ+ Σ β ij dlnq j (6) j=1

for i=1,...,n. Equation 6 is referred to as the Rotterdam Almost Ideal Inverse Demand System (RAIIDS) that has the RIDS scale effects and the AIIDS quantity effects (Brown et al., 1995).

Based on the approach of Brown et al. (1995), the SIDS is developed to nest the RIDS, the LTIDS, the AIIDS, and the RAIIDS. Since the four alternative differential inverse demand systems have identical rightside variables, the SIDS is written as n

s i dlnπ i =(e i –d 1 s i )dlnQ+ Σ [e ij –d 2 s i (δ ij –s j )]dlnqj (7) j=1

for i=1,...,n. Equation 7 is constructed by the weighted average of the systems so that e i =(1–d 1 )α i +d 1 β i and e ij =(1–d 2 )α ij +d 2 β ij . In Equation 7, ei and eij are parameters to be estimated and used to calculate scale elasticities and price flexibilities. The economic regularity conditions require that the parameters satisfy adding up (Σ iei=-1=d1 and Σ ieij=0), homogeneity (Σ jeij=0), and symmetry (eij=eji) conditions. The alternative forms of the differential inverse demand systems are retrieved by restricting d1 and d2. The SIDS becomes the RIDS when (d1,d2)=(0,0), the LTIDS when (d1,d2)=(1,0), the AIIDS when (d1,d2)=(1,1), and the RAIIDS when (d1,d2)=(0,1). The SIDS nests these four different inverse demand systems and allows hypothesis tests among the systems in empirical applications (Brown et al., 1995). In our empirical application, the shippingpoint prices and market shares of strawberries supplied by each strawberry industry are used to construct the dependent variable, while the quantities of strawberries shipped by each strawberry industry are used for the explanatory variables. In addition, the Divisia volume index used for the explanatory variable is constructed by the sum of the market share times each quantity volume.

International Food and Agribusiness Management Review

595

Suh et al.

Volume 20, Issue 4, 2017

3. Empirical analysis

Data on prices and quantities of fresh strawberries from California, Florida, and Mexico were obtained from the Agricultural Marketing Service of the USDA. The quantities of strawberries used in this analysis represent shipment volumes measured in million pounds, and their prices indicate shipping-point prices measured in dollars per pound (Table 1). The data for California strawberries include the shipments from Santa Maria, Orange and San Diego, and Oxnard Districts, while those for Florida are predominantly the shipments from central Florida. The data for Mexican strawberries represent the cross-border shipments from Mexico. As shown in Table 1 and Figure 1, Florida and Mexico have lower prices than California because of the heavy competition between them in the winter season.

The data studied in this analysis covers the sixteen-week winter production period, from the second week in December through the fourth week in March for 2010-2014. The period between December and March covers the peak harvesting and marketing period of winter strawberries, particularly for Florida and Mexican strawberries (Figure 2). To account for seasonality, we take differences between observations in a 16-week cycle in Equation 7 (Brown et al., 1995). In addition, the variables are tested for unit roots, cointegration, and structural breaks but we found no statistical evidence of unit roots, cointegration, or structural breaks in the variables.

Table 1. Descriptive statistics of weekly data by source, Dec. 2010-Mar. 2014 (data provided by Agricultural Marketing Service from the U.S. Department of Agriculture; https://www.ams.usda.gov).

Quantity (million pounds)

Price (dollars per pound)

Variables

Mean

Std. dev.

Min

Max

California shipment Florida shipment Mexico shipment California price Florida price Mexico price

11.195 12.299 7.729 2.039 1.682 1.660

11.199 6.400 3.050 0.627 0.774 0.684

0.163 0.783 2.284 1.305 0.800 0.844

53.961 24.197 16.170 3.709 3.550 3.250

California

Florida

Mexico

6

8

10

60 50

Million pounds

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0075 - Tuesday, August 22, 2017 12:36:02 PM - IP Address:24.21.169.207

3.1 Data and estimation results

40 30 20 10 0

1

2

3

Dec.

4

5

7 Jan.

9

11

12

13

Feb.

Figure 2. Average weekly shipments of strawberries, 2010-2014.

International Food and Agribusiness Management Review

596

14

15

Mar.

16

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0075 - Tuesday, August 22, 2017 12:36:02 PM - IP Address:24.21.169.207

Suh et al.

Volume 20, Issue 4, 2017

The system specified in Equation 7 is conditional on the expenditure on strawberries. Following the multistage budgeting approach, we assume separability of utility so that U.S. consumers allocate total expenditure among groups of commodities, strawberries being one of them (Seale et al., 1992, 2003; Yang and Koo, 1994). Within the group of strawberries, U.S. consumers further select products from different sources. Since qualities of agricultural products vary with production regions, we differentiate strawberries shipped from different sources and construct three equations for (1) California; (2) Florida; and (3) Mexico. Distinguishing the supplying sources also allows for varying effects on prices, which may result due to different degrees of market integration or segmentation for strawberries shipped from different suppliers. Dropping the equation for Mexican strawberries to avoid the singularity of the variance-covariance matrix, we estimate the SIDS using the Iterated Seemingly Unrelated Regression (Zellner, 1962). Homogeneity and symmetry are imposed to improve the predictive power of the demand system (Kastens and Brester, 1996).

Table 2 reports the results of the likelihood-ratio (LR) tests for the nested systems. The LR tests are used to compare the SIDS with the nested demand systems. The test results show that the RIDS, the AIIDS, and the RAIIDS are rejected against the SIDS. The LTIDS is not rejected in favor of the SIDS at conventional significance levels. Accordingly, the SIDS is chosen to obtain accurate estimates for scale elasticities and price flexibilities. The estimation results of the SIDS are reported in Table 3. The log-likelihood value is 198.08 and the estimates for d1 and d2 are 0.95 and 0.15, respectively, which means that our model is different from the RIDS but close to the LTIDS. The estimated parameters of the equation for Mexican strawberries and their associated standard errors are calculated by the adding-up restrictions.

Table 2. Likelihood-ratio (LR) test statistics for nested systems.1,2 Systems

d1

d2

Log-likelihood values

LR test statistics

RIDS LTIDS AIIDS RAIIDS

0 1 1 0

0 0 1 1

155.171 196.471 167.052 135.903

85.82 3.22 62.06 124.36

1

The Likelihood-Ratio test statistic follows a chi-squared distribution.

RIDS = Rotterdam Inverse Demand System; LTIDS = Laitinen-Theil Inverse Demand System; AIIDS = Almost Ideal Inverse Demand System; RAIIDS = Rotterdam Almost Ideal Inverse Demand System..

2

Table 3. Iterated Seemingly Unrelated Regression estimates of parameters for the Synthetic Inverse Demand System.1 Nesting parameters California price Florida price Mexico price d1 d2 Log-likelihood 1

0.949 (0.078) 0.150 (0.091) 198.082

Scale parameters Price parameters 0.027 (0.024) -0.037 (0.042) -0.041 (0.020)

California

Florida

Mexico

0.016 (0.017)

-0.014 (0.012) 0.029 (0.019)

-0.002 (0.007) -0.015 (0.008) 0.017 (0.014)

Numbers in parentheses are standard errors.

International Food and Agribusiness Management Review

597

Suh et al.

Volume 20, Issue 4, 2017

3.2 Scale elasticities Using the estimates of the SIDS, we calculate scale elasticities (ε) that represent the extent to which strawberry prices respond to proportional changes in strawberry shipments. The scale elasticity of strawberries shipped from source i is calculated by

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0075 - Tuesday, August 22, 2017 12:36:02 PM - IP Address:24.21.169.207

ε i=

ei –1 si

(8)

where ei is the parameter estimated from Equation 7, and s i is the sample mean of the share of strawberries shipped from source i. The scale elasticity shows the percentage change in the shipping-point prices in response to a 1% increase in the aggregate shipments of strawberries. Since homothetic preferences require that all expenditure elasticities be equal to one, strawberry is considered scale flexible (inflexible) when a scale elasticity is greater (less) than -1.

The estimated scale elasticities are calculated using Equation 8 and reported in Table 4. All the estimates are negative and statistically significant at 1% significance level, showing that an increase in the shipment scale reduces strawberry prices. The estimates represent that a 1% increase in the aggregate shipments of strawberries will result in decreases in the prices of California, Florida, and Mexican strawberries by 0.87, 1.05 and 1.11%, respectively. When evaluated at the sample mean of the data, the results imply reductions of 1.78, 1.76, and 1.84 cents per pound in the prices of California, Florida, and Mexican strawberries, respectively. The scale elasticity of -1 represents that the market share of shipment value is constant when the scale changes. Our results imply that the shipment value of California strawberries slightly decreases with respect to an increase in the scale, whereas those of Florida and Mexican strawberries slightly increase with increased shipment scale.

3.3 Price flexibilities Price flexibilities (f) represent the percentage changes in strawberry prices induced by a 1% change in strawberry shipments (Brown et al., 1995). While the compensated price flexibility (f ij* ) of strawberry i with respect to strawberries shipped from a source j is calculated by f *ij=e ij /s i –d 2 (δ ij –s j ), the uncompensated price flexibility (fij) is fij=f *ij+s j ε i (9) where δij is the Kronecker delta that equals one if i=j and s i is the sample mean of the share of strawberries shipped from source i. The own-price flexibilities represent the percentage change in the price of strawberry of source i when its own shipment changes by 1%. The cross-price flexibilities represent the percentage change in the price of strawberry of source i when the shipment from source j changes by 1%. They are gross quantity-substitutes (quantity-complements) if the cross-price flexibility is negative (positive).

Table 4. Scale elasticities and price flexibilities.1 Scale elasticities California price Florida price Mexico price 1

-0.873 (0.042) -1.046 (0.045) -1.107 (0.037)

Price flexibilities California

Florida

Mexico

-0.364 (0.024) -0.355 (0.021) -0.351 (0.017)

-0.317 (0.026) -0.421 (0.030) -0.428 (0.023)

-0.193 (0.019) -0.270 (0.022) -0.329 (0.043)

Numbers in parentheses are standard errors; all estimates are statistically significant at the 1% significance level.

International Food and Agribusiness Management Review

598

Suh et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0075 - Tuesday, August 22, 2017 12:36:02 PM - IP Address:24.21.169.207

3.4 Price responses to shipments from different sources Table 4 presents the estimated price flexibilities based on Equation 9. The diagonal elements show the own-price flexibilities, and the off-diagonal elements show the cross-price flexibilities. The estimated ownprice flexibilities are negative and statistically significant at the 1% significance level. The results indicate that the prices of strawberries are not very flexible to own-shipment changes. The own-price flexibility of Florida strawberries is the greatest (-0.42), but the absolute values are less than 1. When evaluated at the sample mean of the data, the Florida strawberry price decreases by 5.76 cents per pound with respect to a 1-million-pound increase in the weekly shipment. In addition, the own-price flexibilities of California and Mexican strawberries show that a 1% increase in own shipments leads to 0.36 and 0.33% reductions in the prices of California and Mexican strawberries, respectively. That is, when California and Mexico increase their shipments by 1 million pounds, the corresponding own prices will decrease by about 6.63 and 7.07 cents per pound, respectively.

Moreover, the estimated cross-price flexibilities are all negative and statistically significant at the 1% significance level, suggesting substitutable, competitive relationships among the strawberries shipped from California, Florida, and Mexico. The estimated cross-price flexibilities are inflexible, indicating prices are relatively less sensitive to shipments from other sources. The low cross-price flexibilities may be attributed to the geographical market segmentation and/or product differentiation. For instance, an increase in California shipments reduces the prices of Florida and Mexican strawberries by 0.36 and 0.35%, respectively, implying that a 1-million-pound increase in California shipments will reduce the prices of Florida and Mexican strawberries by 5.33 and 5.21 cents per pound, respectively, when evaluated at the sample mean. Similarly, an increase in Florida shipments reduces the prices of California and Mexican strawberries by 0.32 and 0.43%, respectively, implying the prices of California and Mexican strawberries will decrease by 5.26 and 5.78 cents per pound in response to a 1-million-pound increase in Florida shipments. While the effects of California shipments on Florida and Mexican prices are very similar, the effect of Florida shipments on Mexican prices are greater than that of California shipments because Florida strawberries compete mainly with Mexican strawberries in the market during the winter season. Furthermore, the estimated cross-price flexibilities show that a 1% increase in Mexican shipments reduces the prices of California and Florida strawberries by 0.19 and 0.27%, respectively. That is, the prices of California and Florida strawberries will decrease by 5.09 and 5.88 cents per pound, respectively, with respect to a 1-million-pound increase in weekly Mexican shipments. The effects of Mexican shipments on the prices of California and Florida are significant.

3.5 Impact of growing imports on the U.S. strawberry industry Given the U.S. strawberry industryâ&#x20AC;&#x2122;s concerns about growing imports from Mexico, it is worth evaluating the potential impact of increasing Mexican shipments on the future of the U.S. strawberry industry under different growth scenarios. Specifically, we investigate how increasing Mexican shipments affect the prices and shipment values of domestic strawberries if the Mexican production capacity and shipments grow by 25, 50, and 100%, respectively. To analyze the impact, we calculate the point estimates of price flexibilities using the average shares of weekly shipment values over the sample period. In Tables 5 and 6, we present the simulated weekly prices and shipment values of California and Florida strawberries under the three scenarios, assuming California and Florida producers maintain their shipment levels. Note that our simulation is based on the static analysis on the U.S. strawberry industry that does not consider potential industry responses or adjustments to increasing Mexican shipments that could occur over time (for example reducing acreage and shipments or adopting new technologies); in other words, we disentangle trade impact holding non-trade factors constant, and show the potential losses for the U.S. strawberry industry if Mexican shipments increase.

Table 5 presents how much Mexican shipments lead to changes in the prices and shipment values of California strawberries in each week from December to March. In the baseline scenario, the shipments of California strawberries grow from December to March, while their prices diminish over the period. Due to the increasing shipments in this period, California has greater shipment values in March than in December.

International Food and Agribusiness Management Review

599

Suh et al.

Volume 20, Issue 4, 2017

Table 5. Weekly effects of Mexican shipments on prices and shipment values of California strawberries.1 Month/ Week

Baseline

Scenario 1

Scenario 2

Scenario 3

Quantity Price Value (million ($/lb.) (million $) lbs.)

Price Value (million ($/lb.) $)

Price Value (million ($/lb.) $)

Price Value (million ($/lb.) $)

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0075 - Tuesday, August 22, 2017 12:36:02 PM - IP Address:24.21.169.207

Dec.

1 3.868 2.936 11.357 2.766 10.700 (-0.66) 2.596 10.042 (-1.31) 2.257 8.728 (-2.63) 2 4.155 3.034 12.606 2.880 11.964 (-0.64) 2.725 11.322 (-1.28) 2.416 10.038 (-2.57) 3 2.332 2.663 6.210 2.521 5.878 (-0.33) 2.378 5.546 (-0.66) 2.093 4.881 (-1.33) 4 2.676 2.235 5.981 2.116 5.663 (-0.32) 1.997 5.345 (-0.64) 1.759 4.708 (-1.27) Jan. 5 3.518 2.388 8.400 2.252 7.921 (-0.48) 2.115 7.441 (-0.96) 1.843 6.483 (-1.92) 6 4.790 2.224 10.652 2.094 10.027 (-0.62) 1.963 9.402 (-1.25) 1.702 8.153 (-2.50) 7 6.807 2.113 14.382 1.977 13.456 (-0.93) 1.841 12.529 (-1.85) 1.568 10.676 (-3.71) 8 6.337 1.937 12.278 1.819 11.530 (-0.75) 1.701 10.782 (-1.50) 1.465 9.286 (-2.99) Feb. 9 7.557 1.828 13.816 1.728 13.060 (-0.76) 1.628 12.304 (-1.51) 1.428 10.792 (-3.02) 10 10.004 1.813 18.141 1.726 17.268 (-0.87) 1.639 16.394 (-1.75) 1.464 14.648 (-3.49) 11 10.694 1.649 17.636 1.582 16.919 (-0.72) 1.515 16.203 (-1.43) 1.381 14.769 (-2.87) 12 11.697 1.662 19.436 1.604 18.767 (-0.67) 1.547 18.098 (-1.34) 1.433 16.761 (-2.67) Mar. 13 14.265 1.682 23.987 1.629 23.238 (-0.75) 1.577 22.489 (-1.50) 1.472 20.991 (-3.00) 14 23.088 1.546 35.705 1.501 34.655 (-1.05) 1.456 33.606 (-2.10) 1.365 31.507 (-4.20) 15 32.076 1.478 47.405 1.434 45.990 (-1.41) 1.390 44.575 (-2.83) 1.301 41.746 (-5.66) 16 35.259 1.434 50.548 1.393 49.117 (-1.43) 1.352 47.685 (-2.86) 1.271 44.821 (-5.73) Sum 308.539 296.15 (-12.39) 283.76 (-24.78) 258.99 (-49.55) 1 Scenarios 1 through 3 present the simulated prices and shipment values when Mexican shipments increase by 25, 50 and 100%, respectively.

The simulation results presented in scenarios 1 through 3 show the impact of Mexican shipments on prices in December is greater, which occurs when California supply is relatively low in the market. In scenarios 1 and 2, the reduced prices will reduce the total shipment values of California strawberries by 12.39 and 24.78 million dollars, respectively. That is, the California industry will lose 4.01 and 8.03% of its total shipment value if Mexican shipments increase by 25 and 50%, respectively (Table 7). When Mexico doubles the shipments as in scenario 3, it will cause a total loss of 49.56 million dollars for the California strawberry industry between December and March (i.e. 16.06% of the total shipment value). To put it in perspective, assuming an average yield of 4,000 flats (32,000 lbs) per acre, a rough yield estimate for California winter fresh strawberries, farm revenue will be reduced by $2,213, $4,426, and $8,852 per acre under the three scenarios, respectively, which represent significant losses for strawberry growers (Table 7).3 Table 6 reports the impacts of Mexican shipments on the prices and shipment values of Florida strawberries.

In the baseline scenario, the shipments of Florida strawberries gradually grow, hitting the peak in the last week of February. The shipment values peak in the third week of February. Under scenarios 1 and 2, the shipment values decrease by 20.16 and 40.33 million dollars, respectively. That is, the Florida industry will lose 6.87 and 13.74% of the total shipment value. Assuming a typical average yield of 3,000 flats (24,000 lbs) per acre for Florida strawberries, scenarios 1 and 2 will result in a revenue loss of $2,460 and $4,919 per acre, respectively (Table 7). The simulation results suggest that Mexican shipments have higher effects on the Florida strawberry industry than on the California industry due to the fact that Florida and Mexico have the same production window and supply pattern.

3

The per-acre loss estimates are calculated using price differences between the baseline scenario and corresponding scenarios and assuming a yield distribution that follows the pattern of the aggregate industry shipments over the season (see Table 5).

International Food and Agribusiness Management Review

600

Suh et al.

Volume 20, Issue 4, 2017

Table 6. Weekly effects of Mexican shipments on prices and shipment values of Florida strawberries.1 Month/ Week

Baseline

Scenario 1

Scenario 2

Scenario 3

Quantity Price Value (million ($/lb.) (million $) lbs.)

Price Value ($/lb.) (million $)

Price Value ($/lb.) (million $)

Price Value ($/lb.) (million $)

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0075 - Tuesday, August 22, 2017 12:36:02 PM - IP Address:24.21.169.207

Dec.

1 4.369 2.988 13.053 2.744 11.988 (-1.07) 2.500 10.923 (-2.13) 2.012 8.793 (-4.26) 2 6.367 3.020 19.228 2.806 17.867 (-1.36) 2.592 16.506 (-2.72) 2.165 13.784 (-5.44) 3 7.606 2.238 17.018 2.069 15.737 (-1.28) 1.901 14.456 (-2.56) 1.564 11.894 (-5.12) 4 8.443 2.081 17.573 1.925 16.256 (-1.32) 1.769 14.939 (-2.63) 1.458 12.306 (-5.27) Jan. 5 9.168 1.909 17.505 1.758 16.115 (-1.39) 1.606 14.725 (-2.78) 1.303 11.945 (-5.56) 6 11.673 1.863 21.740 1.711 19.970 (-1.77) 1.559 18.199 (-3.54) 1.256 14.658 (-7.08) 7 10.696 1.878 20.089 1.708 18.272 (-1.82) 1.538 16.456 (-3.63) 1.199 12.822 (-7.27) 8 12.389 1.613 19.977 1.476 18.284 (-1.69) 1.339 16.591 (-3.39) 1.066 13.205 (-6.77) Feb. 9 15.089 1.425 21.501 1.317 19.878 (-1.62) 1.210 18.254 (-3.25) 0.995 15.006 (-6.49) 10 15.701 1.363 21.393 1.271 19.961 (-1.43) 1.180 18.529 (-2.86) 0.998 15.665 (-5.73) 11 17.754 1.316 23.358 1.242 22.043 (-1.31) 1.167 20.728 (-2.63) 1.019 18.098 (-5.26) 12 22.009 1.019 22.421 0.969 21.336 (-1.09) 0.920 20.250 (-2.17) 0.821 18.079 (-4.34) Mar. 13 20.235 1.066 21.563 1.017 20.578 (-0.99) 0.968 19.592 (-1.97) 0.871 17.622 (-3.94) 14 17.577 1.071 18.824 1.020 17.927 (-0.90) 0.969 17.030 (-1.79) 0.867 15.235 (-3.59) 15 11.854 1.027 12.169 0.965 11.440 (-0.73) 0.904 10.712 (-1.46) 0.781 9.254 (-2.91) 16 5.856 1.034 6.058 0.965 5.653 (-0.41) 0.896 5.247 (-0.81) 0.758 4.437 (-1.62) Sum 293.469 273.30 (-20.17) 253.14 (-40.33) 212.80 (-80.67) 1 Scenarios 1 through 3 present the simulated prices and shipment values when Mexican shipments increase by 25, 50 and 100%, respectively.

Table 7. Total industry loss and reduction in net return per acre.

California

Scenario 1 (25%) Scenario 2 (50%) Scenario 3 (100%)

Florida

Total loss (million $)

Reduction in net return ($/acre)

Total loss (million $)

Reduction in net return ($/acre)

12.39 (4.01%) 24.78 (8.03% 49.56 (16.06%)

2,213 4,426 8,852

20.16 (6.87%) 40.33 (13.74%) 145.29 (49.51%)

2,460 4,919 8,923

When Mexican shipments are doubled in scenario 3, the Florida total loss due to price difference over the sample period would be 80.67 million dollars (Table 6), which is 27.49% of the total shipment value.

However, under scenario 3, the actual loss will be larger. When the shipping prices are consistently less than the marginal costs, producers will give up picking and abandon strawberries in the field. The average marginal cost of harvesting, packing, cooling, and selling strawberries in California is about 50 cents per pound (Daugovish et al., 2011). An industry survey we conducted in 2012-2013 indicates that the average marginal cost in Florida is about 77 cents per pound. Assuming these costs increase by 3% per year over the next five years, the marginal costs will be 58 and 89 cents per pound at the end of the period for California and Florida, respectively. In this case, California strawberry producers would continue to harvest and ship strawberries despite reduced shipment values. However, Florida strawberry producers would stop harvesting when the market price falls below the harvest threshold of 89 cents per pound, because the price is not enough to recover the cost of harvesting, packing, cooling and selling. Thus, the shipment values from the last week of February to the last week of March will become zero; the production season is shortened by five weeks.

This will further reduce revenues by 64.63 million dollars in scenario 3. Accounting for this additional loss, the total reduction in the Florida shipment value will amount to 145.29 million dollars, which is about half International Food and Agribusiness Management Review

601

Suh et al.

Volume 20, Issue 4, 2017

the value of the current industry. The lost shipment value (minus the cost saving due to reduced harvest and shipment) will translate into an average reduction of $8,923 per acre in net return. As the industry is already struggling to break even (making zero profit) under the current market condition, the extra losses caused by the increased Mexican supply will pose serious challenges to the sustainability and survival of the Florida strawberry industry.

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0075 - Tuesday, August 22, 2017 12:36:02 PM - IP Address:24.21.169.207

4. Conclusions and discussions This study aims to shed light on the impact of Mexican competition on the U.S. strawberry industry. The study develops a strawberry market model and examines the effects of the shipments of Mexican strawberries on the prices and shipment values of the U.S. winter strawberries. The estimated price flexibilities suggest that Mexican shipments significantly affect the prices of California and Florida strawberries. In particular, the simulation results indicate that further expansion in Mexican production capacity will cause severe losses to U.S. growers, decreasing the profitability and sustainability of the industry. The empirical methods used in our study could be applied to other produce or to other countries to examine the potential impact of competition and the increasing market shares of competitors on prices and shipment values.

The findings in our study provide a clear indication of the challenges and difficulties the U.S. strawberry industry is facing. There have been heated debates on the trade policy in light of NAFTA. The industry has been exploring options to ensure industry sustainability, including changes in trade policy. The industry's main argument for change is that Mexican production has been ‘unfairly subsidized’ and that the produce industry has been the ‘sacrificial lamb’ in the trade deal.

Given the large losses found in this study, improved production and marketing are necessary for the domestic industry to remain viable under NAFTA. The efforts to reduce costs are critical for the strawberry industry, especially for the Florida industry, to survive. This calls for advancements in production technologies.

In particular, introduction of mechanical harvesting could substantially reduce costs and increase the competitiveness of the U.S. industry. It may take time and a large investment to develop a mechanical harvesting system. However, the cost reductions that come with mechanization could effectively neutralize Mexico’s competitive advantage in labor cost, thus creating a level playing field between Mexico and the U.S. The bed and mulch production system adopted in the 1970’s and 1980’s was a major innovation over the last few decades. Mechanical harvesting potentially could be another major event in technological innovation.

It could not only reduce cost but also address the serious labor shortage problem in the industry. Besides labor-saving technologies, developing new and superior varieties could also help growers differentiate in the generic commodity market to alleviate the impact of competition. Unlike in the apple market where varieties are usually labeled distinctly with recognizable differences in size, appearance and taste, strawberries are usually not labeled by variety and are generally treated as generic commodity. The U.S. strawberry industry is investing in research and development seeking to differentiate their products from competitors. Florida strawberry industry is taking further measures to limit Mexican access to new varieties Florida is developing.

However, successful product differentiation of a generic commodity may require institutional changes to ensure effective coordination within the industry in branding and labeling as well as regulation of quality standards. In summary, the industry may need significant changes in technology, marketing, and industrial organization to effectively compete in the marketplace.

References Asci, S., J.L. Seale, O. Gulcan and J.J. Van Sickle. 2016. U.S. and Mexican tomatoes: perceptions and implications of the renegotiated suspension agreement. Journal of Agricultural and Resource Economics 41: 138-160.

Ayala-Zavala, J.F., S.Y. Wang, C.Y. Wang and G.A. Gonzalez-Aguilar. 2004. Effect of storage temperatures on antioxidant capacity and aroma compounds in strawberry fruit. LWT-Food Science and Technology 37: 687-695.

International Food and Agribusiness Management Review

602

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0075 - Tuesday, August 22, 2017 12:36:02 PM - IP Address:24.21.169.207

Suh et al.

Volume 20, Issue 4, 2017

Baker, G.A. 2004. California strawberry production and methyl bromide. The International Food and Agribusiness Management Review 7: 65-69.

Barten, A.P. and L.J. Bettendorf. 1989. Price formation of fish: an application of an inverse demand system.

European Economic Review 33: 1509-1525.

Boriss, H., H. Brunke, and M. Kreith. 2006. Commodity profile: strawberries. Agricultural Issues Center, University of California, Brown, M.G., J. Lee and J.L. Seale. 1995. A family of inverse demand systems and choice of functional form. Empirical Economics 20: 519-530.

Burfisher, M.E., S. Robinson and K. Thierfelder. 2001. The impact of NAFTA on the United States. The Journal of Economic Perspectives 15: 125-144.

Carter, C.A., J.A. Chalfant, R.E. Goodhue, F.M. Han and M. DeSantis. 2005. The methyl bromide ban: economic impacts on the California strawberry industry. Applied Economic Perspectives and Policy 27: 181-197.

Chambers, R.G. and K.E. McConnell. 1983. Decomposition and additivity in price dependent demand systems. American Journal of Agricultural Economics 65: 596-602.

Daugovish, O., K.M. Klonsky and R.L. De Moura. 2011. Sample costs to produce strawberries, South Coast â&#x20AC;&#x201C; Ventura County, Oxnard Plain 2011. University of California Cooperative Extension, Oakland, CA, USA. Eales, J.S. and L.J. Unnevehr. 1993. Simultaneity and structural change in U.S. meat demand. American Journal of Agricultural Economics 75: 259-268.

Eales, J.S. and L.J. Unnevehr. 1994. The inverse almost ideal demand system. European Economic Review 38: 101-115.

Goodhue, R.E., S.A. Fennimore and H.A. Ajwa. 2005. The economic importance of methyl bromide: does the California strawberry industry qualify for a critical use exemption from the methyl bromide ban?

Applied Economic Perspectives and Policy 27: 198-211.

Grant, J.H., D.M. Lambert and K.A. Foster. 2010. A seasonal inverse almost ideal demand system for North American fresh tomatoes. Canadian Journal of Agricultural Economics 58: 215-234.

Guan, Z., F. Wu, F. Roka and A. Whidden. 2015. Agricultural labor and immigration reform. Choices 30: 1-9.

Hernandez-Munoz, P., E. Almenar, V.D. Valle, D. Velez and R. Gavara. 2008. Effect of chitosan coating combined with postharvest calcium treatment on strawberry (FragariaĂ&#x2014;ananassa) quality during refrigerated storage. Food Chemistry 110: 428-435.

Huang, K.S. 1988. An inverse demand system for U.S. composite foods. American Journal of Agricultural Economics 70: 902-908.

Johnson, R. 2014. The U.S. Trade situation for fruit and vegetable products. CRS Report. Congressional Research Service, Washington, WA, USA. Kastens, T.L. and G.W. Brester. 1996. Model selection and forecasting ability of theory-constrained food demand systems. American Journal of Agricultural Economics 78: 301-312.

Lee, Y. and L. Kennedy. 2016. Trade creation and diversion under NAFTA: the North American strawberry market. 2016 Annual Meeting, July 31-August 2, 2016, Boston. Available at: http://tinyurl.com/ mfp7htc.

Matsuda, T. 2005. Forms of scale curves and differential inverse demand systems. American Journal of Agricultural Economics 87: 786-795.

Norman, C.S. 2005. Potential impacts of imposing methyl bromide phaseout on U.S. strawberry growers: a case study of a nomination for a critical use exemption under the Montreal Protocol. Journal of Environmental Management 75: 167-176.

Park, H. and W. Thurman. 1999. Interpreting inverse demand systems: a primal comparison of scale flexibilities and income elasticities. American Journal of Agricultural Economics 81: 950-958.

Park, H., W.N. Thurman and J.E. Easley. 2004. Modeling inverse demands for fish: empirical welfare measurement in Gulf and South Atlantic fisheries. Marine Resource Economics 19: 333-351.

Seale, J.L., M.A. Marchant and A. Basso. 2003. Imports versus domestic production: a demand system analysis of the U.S. red wine market. Review of Agricultural Economics 25: 187-202.

International Food and Agribusiness Management Review

603

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0075 - Tuesday, August 22, 2017 12:36:02 PM - IP Address:24.21.169.207

Suh et al.

Volume 20, Issue 4, 2017

Seale, J.L., A.L. Sparks and B.M. Buxton. 1992. A Rotterdam application to international trade in fresh apples: a differential approach. Journal of Agricultural and Resource Economics 17: 138-149.

Taylor, J.E., D. Charlton and A. Yúnez-Naude. 2012. The end of farm labor abundance. Applied Economic Perspectives and Policy 34: 587-598.

USDA/NASS. 2013. Noncitrus fruits and nuts: 2012 summary. National Agricultural Statistics Service. U.S. Department of Agriculture, Washington, WA, USA. USDA/NASS. 2015. Noncitrus fruits and nuts: 2014 summary. National Agricultural Statistics Service. U.S. Department of Agriculture, Washington, WA, USA. Victoria, N.G., O. van der Valk and A. Elings. 2011. Mexican Protected Horticulture: Production and market of Mexican protected horticulture described and analysed. Report GTB-1126. Wageningen UR, Wageningen, the Netherlands and Dutch Agricultural Economic Research Institute, The Hague, the Netherlands.

Wu, F., Z. Guan and D.H. Suh. In press. The effects of tomato suspension agreements on market price dynamics and farm revenue. Applied Economic Perspectives and Policy.

Wu, F., Z. Guan and V. Whitaker. 2015. Optimizing yield distribution under biological and economic constraints: Florida strawberries as a model for perishable commodities. Agricultural Systems 141: 113-120.

Yang, S.R. and W.W. Koo. 1994. Japanese meat import demand estimation with the source differentiated AIDS model. Journal of Agricultural and Resource Economics 19: 396-408.

Young, R.R. 1998. NAFTA’s impact on the North American agriculture – A logistician’s perspective. The International Food and Agribusiness Management Review 1: 13-24.

Zahniser, S., S. Angadjivand, T. Hertz, L. Kuberka and A. Santos. 2015. NAFTA at 20: North America’s free-trade area and its impact on agriculture. Economic Research Service. U.S. Department of Agriculture, Washington, WA, USA. Zellner, A. 1962. An efficient method of estimating seemingly unrelated regressions and tests for aggregation bias. Journal of the American Statistical Association 57: 348-368.

International Food and Agribusiness Management Review

604

OPEN ACCESS International Food and Agribusiness Management Review Volume 20 Issue 4, 2017; DOI: 10.22434/IFAMR2016.0160

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0160 - Tuesday, August 22, 2017 12:37:03 PM - IP Address:24.21.169.207

Received: 28 September 2016 / Accepted: 13 February 2017

Is dairy complex a solution to milk safety? A comparison of farmers’ perceived and realized food safety effects RESEARCH ARTICLE H. Holly Wanga,b, Hailong Yu c, and Binglong Lid aEconomist,

Institute of Agricultural Resources and Regional Planning, Chinese Academy of Agricultural Sciences, No.12 Zhongguancun South St., Haidian District, Beijing, China P.R.

bProfessor,

Department of Agricultural Economics, Purdue University, West Lafayette, IN 47907, USA cPostdoctoral

fellow, China Institute for Rural Studies, Tsinghua University, Haidian District, Beijing 100084, China P.R.

dProfessor,

College of Economics and Management, China Agricultural University, Haidian District, Beijing 100083, China P.R.

Abstract This study explores the major reasons for Chinese small dairy farms to accept the new organization structure, dairy complex (DC), and the discrepancies between the actual effect and farmers’ perceptions. Our results show that the frequency of milk refusal, herd scale and farmers’ age contribute to farmers’ decision in accepting DCs, while sale price and disease prevention do not have significant influence. Independent farmers’ perceived effects and the actual effects of DCs in improving raw milk safety and price are consistent, but there exists misperception of DCs’ effect in disease prevention.

Keywords: dairy complex, food safety, perceived and actual effects, misperception JEL code: Q12, Q18 Corresponding author: yuhailong110@126.com

© 2017 Holly Wang et al.

605

Holly Wang et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0160 - Tuesday, August 22, 2017 12:37:03 PM - IP Address:24.21.169.207

1. Introduction The 2008 infamous scandal that Chinese tainted infant formula killed and sickened thousands of babies shocked the world and impacted the Chinese dairy industry dramatically (Ortega et al., 2012). Illegal chemicals are added to milk by dairy producers and milk collectors to make the milk appear to have high protein component, but can cause severe health problems to consumers. One important factor leading to such milk safety problems is the highly fragmented production sector (Calvin et al., 2006; Ortega et al., 2011), which is a common situation in developing countries. The Chinese government has taken a series of regulatory measures that call for stronger control over milk producer, collectors, and processors. One strategy has been to encourage small-scale farmers to move their cows into the communal dairy complex (DC) in each village.

Dairy complexes, yangzhixiaoqu in Chinese, is also translated as farming communities, cooperative dairy farms, or cow hotels. They are usually co-funded by local governments, agribusiness firms and individuals and centrally planned, managed and constructed under the guidance of local governments, according to the ‘White paper on China dairy’ (Li, 2014). In general, DCs are constructed outside villages and managed by a single manager who owns the DC or is hired to manage it. The small, backyard dairy producers in the village or nearby are encouraged to check their cows into the DC. In a DC, individual farmers still own their cows and assume the primary responsibilities of feeding and providing other inputs. The cows are milked in the DC-owned milking parlor, and vaccination and breeding are managed by the DC. Small sample of milk is kept from each farmer until the whole batch of milk is accepted by the processor after inspection. If rejection occurs due to quality or safety reasons, the causes can be traced to individual farmers who will be held responsible, with the help of the samples. The government provides subsidies and uses other measurements to promote DCs in an attempt to provide a transitional model as the dairy industry shifts gradually into modern large scale production.

Farmers are allowed to join a DC on a voluntary basis, although the government promotes hard. Many farmers are reluctant to join, because entry is not free and some DCs are far away from farmers’ homes that they can’t commute several times a day to feed, milk and take care of the cows (Mo et al., 2012). Most importantly, a large number of individual farmers consider that DCs are simply gathering cows into a centralized location with little effect on improving raw milk safety, raw milk price, and cow health.

DC policy is one of the most important production polices after the 2008 scandal and has significant impact on the development of China’s dairy industry (Jia et al., 2012). There exists rich literature about farmers’ understandings and perceptions of a new program and their effect on farmers’ willingness to participate in it (Buckley et al., 2012; Leftley and Mapfumo, 2006; Patt et al., 2010; Vanslembrouck et al., 2002; Wossink and Wenum, 2003). Most of these studies investigated the effects of farmers’ evaluation or the adoption of the program by using their subjective attitude toward a public good, say the environment, or their own perceptions of program attributes directly as explanatory variables in their models, and thus the accuracy of the perception measurement is crucial in such methods. Given that farmers’ answers to the perception questions in surveys are often inaccurate and even endogenous, studies that avoid using such explanatory variables are needed.

Further, there exist a limited number of studies on Chinese dairy production and policy and they mainly focus on impact of policy on farmers’ production (Jia et al., 2012; Mo et al., 2012; Zhong et al., 2014). Little analysis has been found trying to explain small scale farmers’ motivation to move their cows into DCs from the milk safety and quality angles. There is also hardly any literature on the actual milk safety and quality effect comparison between independent farms and DCs. In this paper, we conduct an empirical analysis on the Chinese DCs in the dairy production to explore whether DCs can help improve the milk safety and quality. We first check the farmers’ willingness to join the DCs, which is fundamental for DCs to be the milk safety solution. To avoid using inaccurate perceptions International Food and Agribusiness Management Review

606

Holly Wang et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0160 - Tuesday, August 22, 2017 12:37:03 PM - IP Address:24.21.169.207

as independent variables directly, we employ measurements of independent small scale farmers’ actual factors to explain their willingness to enter the DCs. The linkage between such actual factors outside DCs and the willingness to enter DCs implies farmers’ perception and expectations of the DCs’ effect on these factors. Milk safety measurements are considered among these factors. Then we also present the real effect of DCs on milk safety, economic and ecological factors. Finally, we will give some suggestion about how to improve the effectiveness of policy.

Whether the DC policy can be implemented smoothly depends not only on local governments’ executive power but also farmers’ acceptance, while the latter mainly depends on their endowments and comprehension of DCs’ actual effects. This paper, using the randomly survey data in the main production areas in China, aims to fill this gap. In this study, we will compare farmers’ perceived and actual effects of DCs on the improvement of raw milk safety, raw milk price, and cow health, the three primary goals of the government’s DC promotion policy.

2. Data and variables definition A survey was conducted in the summer of 2012 by graduate student from China Agricultural University in a face to face interview method for dairy farms in three major dairy production areas of China that supply milk to leading dairy processors. Ten villages were chosen in Tangshan district of Hebei province, seven villages in Weifang district of Shandong province, and six villages were chosen in Shuangcheng district of Heilongjiang Province. In each village we randomly selected dairy farmers from a name list with the help of the local government. The questionnaire was pre-tested in Tangshan before the formal survey started. Finally, we obtained 164 observations, among which, 48 are currently independent farms, and 116 are those who used to be independent farms but have moved to DCs. The proportion of farms in DCs in our sample is 71%, much higher than independent farms, which can attribute to the strong DC promotion by local governments.

Since we are interested in evaluating farmers’ attitude toward DCs, we ask a question explicitly to the independent farmers whether they are willing to enter (WTE), and assign value 1 to the binary variable WTE if the answer is yes. This question was not asked to farmers already in DCs, because they are not the program target and also because they are mostly willing if not all.

The three primary goals of the DC policy are the improvement of raw milk safety, raw milk price received by farmers, and cow health, and thus, we focus on farmers’ actual effects of DCs on these three aspects. The variables, RefFreq, Sprice and DisPre, representing the frequency of raw milk being rejected in the previous year, the sale price of raw milk in the current year, and the dummy indicator for farms with the presence of some disease this year, respectively, are included in the model as the key independent variables. Notice, all these three variables are actual instead of perceptional. We expect the variable RefFreq and DisPre positively affect the WTE to a DC, because farmers with larger frequency of refusal and the occurrence of cow diseases will be more likely to check their cows into DCs, as the cows are managed with better techniques in terms of disease prevention and sanitation, and better guidance in feed nutrition. Sprice is expected to be negative, because farmers experience low prices may count on DCs to have more market power and better quality reputation to receive better price. Notice, there is no government regulation on raw milk price in China, and farmers can sell their milk in open market to small collectors, and to processors with and without a contract.

In order to capture farmers’ production and external monitoring situation, we also include variables about herd-scale and governments’ supervision. The variable HerdScale is the number of cows a farmer has, and it is reasonable to assume that larger (no more than a few dozens of cows) dairy farms are more likely to join, because they can benefit more after checking into DCs and they are the government’s targets (Mo et al., 2012). The variable GovSup is a dummy variable and equals to 1 if local government official rigorously examines the safety of raw milk and sanitation of production environment in previous three months. Farmers who are supervised rigorously are more likely to enter the DCs, because it is harder and harder for small independent farms to meet the continuously improvement of sanitary requirement. In addition, farmers’ International Food and Agribusiness Management Review

607

Holly Wang et al.

Volume 20, Issue 4, 2017

demographic characteristics variables, age, gender and education level are also included. The definition and descriptive statistics of variables in the model are present in Table 1.

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0160 - Tuesday, August 22, 2017 12:37:03 PM - IP Address:24.21.169.207

Same questions except WTE are asked to farmers in the DCs to capture the factual differences in milk safety and price. The descriptive statistics will be discussed in the result section because they contribute to the results.

Additional questions are asked to farmers in DCs to explore their behavioral and environment changes. Twelve questions are asked. Behavioral questions include, after moving into the DC, whether they have changed their production style, whether they use better technologies, whether they pay more attention to milk safety measurements, whether they obtain higher quality feed, whether they experience reduced refusal frequencies, whether they improve milk protein content, whether they receive higher price, whether the price is more stable, whether they reduce production cost, and whether they receive dramatic profit increase. Questions about environmental change include whether they have better production environment, and whether they receive better disease prevention support. In order to obtain more accurate responses, farmers are asked to give one or two examples if the answer is yes to a question and we bench mark their answers with the examples provided. Corresponding variables are assigned the value one to the yes answers and zero to the no answers.

3. Method We first conduct t-tests for the three factual variables, RefFreq, Sprice, and DisPre, between the farms in and outside DCs. The null hypothesis is that the means of the two populations for each variable is the same.

The alternative hypothesis can be directional in one-tailed tests for variables if we have a clear expectation of the signs, or non-directional in two-tailed tests for those no sign is expected at a priori. These results give us indications that whether DCs can improve milk safety.

We then employ a logistic regression model to check the impact from actual factors of milk safety, profitability and other economic attributes on farmers’ willingness to enter the DC. These impacts on WTE reflexes farmers’ perception of the corresponding factors can help them in DCs. For example, if a particular factor contributes to WTE positively, it means that independent farmers with higher level of this factor are more Table 1. Descriptive statistics of variables in the model.1 Variables WTE3

Definition

Independent farms

Willingness to enter a DC, equals 1 if the farmer is willing, 0 otherwise RefFreq Frequency of raw milk refusal in the previous year Sprice Sale price of raw milk in the current year DisPre Occurrence of disease, =1, if there are some diseases this year, 0 otherwise GovSup Government supervision, =1, if it is very rigorous, 0 otherwise HerdScale The number of cows the farmer has Age Years of farmer’s age Male Gender dummy, male=1, female=0 Edu Years of farmer’s schooling

DC farms

t-test2

Mean SD

Max

Min

Mean

0.44

0.50

1

0

N/A

N/A

0.35

0.79

3

0

0.34

-0.07

2.97 0.75

0.05 0.44

3.05 1

2.8 0

3.08 0.29

5.38*** -5.9***

0.68

0.47

1

0

0.84

1.96*

16.25 41.6 0.83 7.31

10.1 8.34 0.38 2.48

60 59 1 12

5 27 0 0

31.44 49.44 0.69 7.82

6.92*** 6.0*** -2.07** 1.14

1

Two refusals caused by inappropriate behavior and disobeying the dairy complex’s regulation rather than safety or quality issues are excluded.

2 **, and *** indicate statistically significant at 5 and 1%, respectively.

3 WTE = willing to enter.

International Food and Agribusiness Management Review

608

Holly Wang et al.

Volume 20, Issue 4, 2017

likely to join a DC. If we know this factor is a ‘bad’ one, then, it implies farmers expect the factor can be reduced in a DC. This way farmers’ perception and expectations are examined. We expect the good factors to be negative and bad factors to be positive intuitively.

Logistic regression is a common method for binary dependent variables.

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0160 - Tuesday, August 22, 2017 12:37:03 PM - IP Address:24.21.169.207

Y* = Χβ + e (1) Y*is a latent variable measuring farmers perceived benefits of entering the DCs, if Y* > 0, farmers will express willingness to enter, and WTE=1. X is the matrix of explanatory variables, as explained in Table 1.

P(Y* > 0) = P (Χβ + e > 0 ) = F(Χβ)

(2)

We assume the error term in Equation 1 is distributed as the logistic function, then: F(Χβ) = P =

1 eΧβ =

1 + e –Χβ 1+e Χβ

(3)

where P is the probability that the observed variable, WTE takes value 1. Thus, we can get the marginal effect of each variable: ∂P = P (1 – P) βj from Equation 3. ∂Xj Our third analysis is to study the behavioral change before and after farmers entering DCs, and their perceived condition differences. This analysis is only conducted for farmers already in DCs.

4. Results Factual results from comparison From Table 1 we see that the mean of WTE is 0.44 among independent farmers, which means 56% of independent farmers prefer to feed their cows in their backyards, rather than checking into DCs. This value indicates that DCs are not accessible or conveniently ready to some farmers who are willing to join, and it also indicates that for the remaining independent farmers, a large portion are unwilling.

The mean of RefFreq in independent farms is 0.35, only slightly higher than their counterpart in DCs, 0.34.

Giving that some farms receive 3 refusals in a year, the majority of farms, about 70% or more have never been rejected. The safety of raw milk in DCs is a little better than independent farms, but not very significant as shown in the t-tests. Notice, the rejections are not all due to safety reasons such as bacteria measures, but also due to nutritional quality reasons such as protein content. The latter is individually determined as farmers in DCs still manage their feed and feeding operation. Another reason is that DCs receive more rigorous government supervisions as shown next, and such supervisions include more careful inspection at the processing firm receiving end resulting in refusals. A third reason is that the statistics are from two different samples. There may be a self-selection effect that farmers who had historically higher refusal frequency tend to join DCs. This factor alone cannot conclude whether DCs have advantages in milk safety.

Farmers in DCs receive significantly higher prices than the independent ones, although by a small margin, 7 cents or about 2%. The raw milk variation is not big among the three provinces across farmers. The likelihood of disease occurrence is significantly lower for cows in DCs and outsides, with the former being only 40% of the latter. Healthy cows produce safer and better milk. This factor shows the advantages of DCs.

International Food and Agribusiness Management Review

609

Holly Wang et al.

Volume 20, Issue 4, 2017

The mean of GovSup in DCs and independent farms is 0.84 and 0.68 respectively, that farms in DCs receive significantly more rigorous government supervisions. The average of HerdScale in DCs is 31.44, much larger than independent farms at 16.25, ranging from 5 to 60. This result corresponds well to our expectation that larger farms are more likely to check into DCs. Demographic variables show that farmers entering DCs are significantly older, for about 8 years, are more likely to be females. The average education levels are about seven years for both groups.

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0160 - Tuesday, August 22, 2017 12:37:03 PM - IP Address:24.21.169.207

Results of perceived effects The regression results for WTE are presented in Table 2. These results confirm that the milk refusals made by processors contribute to farmers’ decision of moving to DCs. The marginal effect of RefFreq is 0.705, indicating that if the number of refusal increases by 1 in a year, the probability of the farmer’s willingness to enter a DC will increase by 0.71. This is a strong support that farmers do link the DCs with better milk quality and safety and resulting fewer refusals. The variables Sprice and DisPre are insignificant. This means, although the signs are as expected, there is no significant evidence showing that independent farmers connect DCs with higher raw milk prices or low chance for cow disease, which meets our expectations. Or in other words, DCs’ effects on improving sale price and disease prevention as shown in Table 1 have not been perceived by farmers.

Other factors like production scale and farmers’ age also contributed to such decisions. One more cow an independent farmer owns will lead to 0.079 increases in her/his probability of entering the DC. This result is accordance with our expectation and that in Table 1. Chinese farmers live in villages with farm houses closely constructed to one another, which makes it difficult for their casual barns in backyards to hold more than a few cows. As the number increases, they need to consider the space in DCs. Local governments also target the relatively larger farms for the DC promotion. The younger independent farmers are more likely to enter the DCs than their older counterparts. This is different to the age comparison between the independent farms and those already in DCs. They are from different samples, and there is no strong economic reason for one way or the other.

Table 2. Logit model: factors influencing farmers’ willingness to send their cows to dairy complexes.1 Variable

Coefficient2

z-statistic

Constant RefFreq Sprice DisPre GovSup Age Male Edu HerdScale No. of observations

13.748 2.834** -6.038 0.904 2.441 -0.151* 3.340 -0.126 0.318*** 48

0.317 2.196 -0.434 0.295 1.249 -1.653 1.022 -0.469 2.681

1

Marginal effect on WTE3 probability 0.705 -1.503 0.215 0.512 -0.037 0.553 -0.031 0.079

McFadden R-squared: 0.62; LR statistic: 40.88. The partial effect are calculated when we set variables to their means. and *** indicate statistically significant at 10, 5 and 1%, respectively.

3 WTE = willing to enter.

2 *, **,

International Food and Agribusiness Management Review

610

Holly Wang et al.

Volume 20, Issue 4, 2017

Results of realized effect

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0160 - Tuesday, August 22, 2017 12:37:03 PM - IP Address:24.21.169.207

Using the survey data from producers who used to operate on their own farms but have moved to DCs, we can observe real effects of DCs on the same group of people through the changes they experience. Questions about whether they experience particular changes are asked to these farmers. The sample average of the binary variable represents the percentage of yes answers for each question. We conducted t-test to check each mean against zero. The results are listed in Table 3.

Results show that all estimates are significant, except for the cost reduction question. 47% of DC farmers have changed they production style, and adopted better technology. Over 90% of them paid more attention to food safety of raw milk, used higher quality feed, and improve their milk protein content. 85% experienced a lower milk refusal frequency. However, only 27% of farmers in DCs realized their sale price increased consistent to the small price difference reported by farmers in and outside DCs in Table 1, although 58% of them received more stable prices. So the DCs’ effect of improving sale price is very limited. DCs do not help lower feed cost, because farmers still manage their own feed independently, not taking advantage of DCs’ collective bargaining power in feed purchasing, and they also use better quality feed. Therefore, only 21% realized that their profit has increased dramatically. 91% of DC farms have better production environment and better disease prevention support in DCs. These are actual changes experienced by the same people before and after moving to DCs. Although we don’t have third party record for these variables rather stated by farmers themselves, we have good faith about their honesty and accuracy as we take extra measures or validating their answers during the face to face interview.

Comparison of perceived and realized effects We can conclude from the Logit model that independent small farmers do link the DCs with better milk safety and quality. At the same time, 85% of once independent farmers realized that their frequency of milk refusal decreased after moving into the DCs. So it is consistent in the perceived and realized effects of DCs in improving milk safety.

The effects of DCs on improving milk price and disease prevention are not perceived by most of the independent farmers from the results of the Logit model. Although 58% farmers agree with DCs effects on stabilizing Table 3. Realized effects of dairy complexes from famers currently operating in dairy complexes who used to be independent.1 Variables

Average

t-value2

Change of production style Using better technology Paying more attention to food safety of raw milk Obtaining higher quality feed Reduction of refusal frequency Improved quality with higher protein content Receiving higher price Price is more stable Profit increased dramatically Cost reduction Having better production environment Better disease prevention Number of observations

0.47 0.47 0.96 0.91 0.85 0.94 0.27 0.58 0.21 0.03 0.91 0.91 116

10.18*** 10.18*** 50.53*** 34.91*** 25.88*** 42.32*** 6.48*** 12.54*** 5.48*** 1.75 34.91*** 34.91***

1

Null hypothesis of t-test is μ=0; alternative hypothesis is μ≠0. indicate statistically significant at 1%.

2 ***

International Food and Agribusiness Management Review

611

Holly Wang et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0160 - Tuesday, August 22, 2017 12:37:03 PM - IP Address:24.21.169.207

price, only 27% expressed that their price increased significantly. That is a critical issue we have found in the survey, that most farmers’ perception of DCs lack of effects on improving price is also consistent with its actual effects. This may hamper farmers’ willingness to enter DCs as they need to pay extra space cost without receiving higher price. The only economic gain is through refusal reduction.

Disease prevention is an important issue, which directly contributes to milk quality and safety. Most of independent farmers worry about the disease prevention in DCs. They do not trust the DCs have great performance in disease prevention, which has also been supported by Logit model. However, 47% of DC farmers expressed that they had changed their production style after moving into DCs, and 91% agreed that they had better production environment and better disease prevention. So we can conclude that independent farmers have misperceptions about DCs’ effect of disease prevention.

5. Summary and conclusions This study has explored the connection between acceptance of DCs by independent farmers and specific milk safety, quality and other economic factors, examined the behavioral and production environmental change for farmers in DCs before and after they join, and also compared the safety and quality measurements across the two groups. We find factors contributing to farmers’ willingness to enter DCs, and the discrepancies between the actual effect and some of farmers’ misperceptions about these factors. Frequency of refusal for milk from independent farms, herd scale and farmers’ age contribute to farmers’ decision in moving into DCs, while milk price and disease prevention do not have significant influence. Independent farmers’ perception and the actual effects of DCs in improving the safety of raw milk and price are consistent, but there is misperception of DCs’ effect on disease prevention among independent farmers. Therefore, it is very important for DC promoters to educate independent farmers from disease prevention point of view.

The empirical results are based on a limited survey sample. However, the survey was conducted during the special window right after the 2008 severe milk safety scandal and the government reacted with promoting the DCs as a policy. Because DCs play a limited role in increasing price received by farmers, an obvious obstacle of moving independent farms into DCs, how to increase sale price of raw milk and further increase farmers’ profit is the critical issue of effectiveness of the DC policy.

References Buckley, C., S. Hynes and S. Mechan. 2012. Supply of an ecosystem service – farmers’ willingness to adopt Riparian Buffer Zones in agricultural catchments. Environmental Science and Policy 24: 101-109.

Calvin, L., F. Gale, D. Hu and B. Lohmar. 2006. Food safety improvements underway in China. Amber Wave 4: 16-21.

Jia, X., J. Huang, H. Luan, S. Rozelle, and J. Swinnen. 2012. China’s milk scandal, government policy and production decisions of dairy farmers: the case of greater Beijing. Food Policy 37: 390-400.

Leftley, R. and S. Mapfumo. 2006. Effective micro-insurance programs to reduce vulnerability. Opportunity International Network, Chicago, IL, USA. Available at: http://tinyurl.com/gwzd8lb.

Li, S. 2014. White paper on China dairy (2014). Available at: http://edepot.wur.nl/334381.

Mo, D., J. Huang, X. Jia, H.Luan, S. Rozelle J. and Swinnen. 2012. Checking into China’s cow hotels: have policies following the milk scandal changed the structure of the dairy sector? Journal of Dairy Sciences 95: 2282-2298.

Ortega, D.L., H.H. Wang, N. Olynk, J. Bai and L. Wu. 2012. Chinese consumers demand for food safety attributes: a push for government and industrial regulations. American Journal of Agricultural Economics 94: 489-495.

Ortega, D.L., H.H. Wang, L. Wu and N. J. Olynk. 2011. Modeling heterogeneity in consumer preferences for select food safety attributes in China. Food Policy 36: 318-324.

Patt, A., P. Suarez and U. Hess. 2010. How do small-holder farmers understand insurance, and how much do they want it? Evidence from Africa. Global Environmental Change 20:153-161.

International Food and Agribusiness Management Review

612

Holly Wang et al.

Volume 20, Issue 4, 2017

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0160 - Tuesday, August 22, 2017 12:37:03 PM - IP Address:24.21.169.207

Vanslembrouck, I., G.V. Huylenbroeck and W. Verbeke. 2002. Determinants of the willingness of Belgian farmers to participate in agri-environmental measures. Journal of Agricultural Economics 53: 489-511.

Wossink, G.A. and J.H. Wenum. 2003. Biodiversity Conservation by farmers: analysis of actual and contingent participation. European review of agricultural economics 30: 461-485.

Zhong, Z., S. Cheng, S. Kong and M. Tracy. 2014. Why improving agrifood quality is difficult in China: evidence from dairy industry. China Economic Review 31: 74-83.

International Food and Agribusiness Management Review

613

http://www.wageningenacademic.com/doi/pdf/10.22434/IFAMR2016.0160 - Tuesday, August 22, 2017 12:37:03 PM - IP Address:24.21.169.207