Volume 21 Issue 2 - Special Issue by IFAMA

International Food and Agribusiness Management Review

Special Issue: IFAMA 2017 Symposium Official Journal of the International Food and Agribusiness Management Association

Volume 21 Issue 2 2018

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.

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

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 21 Issue 2, 2018

Special issue: IFAMA 2017 symposium

TABLE OF CONTENTS 1.

IFAMA best paper competition: introduction to the special issue

179

Sector blending: evidence from the German Food Bank

181

Determinants of corporate cash holdings in times of crisis: insights from Brazilian sugarcane industry private firms

201

Job preferences of agricultural students in Germany â&#x20AC;&#x201C; A choice-based conjoint analysis for both genders

219

Allocating production risks through credit cum insurance contracts: the design and implementation of a fund for small cotton growers to access market finance

237

Governance of market-oriented fresh food value chains: export chains from New Zealand

249

Social sustainability in the ready-made-garment sector in Bangladesh: an institutional approach to supply chains

269

Jacques Trienekens

Meike Rombach and Vera Bitsch

Aviner Augusto Silva Manoel, Marcelo Botelho da Costa Moraes, David Ferreira Lopes Santos, and Marcos Fava Neves

Stephan G.H. Meyerding

Mauro Alem and Julio Jorge Elias

Jacques Trienekens, Mariska van Velzen, Nic Lees, Caroline Saunders, and Stefano Pascucci

Laura A. Carlson and Vera Bitsch

International Food and Agribusiness Management Review

OPEN ACCESS International Food and Agribusiness Management Review Volume 21 Issue 2, 2018; DOI: 10.22434/IFAMR2018.x002

IFAMA best paper competition: introduction to the special issue Special issue: IFAMA 2017 symposium

EDITORIAL Jacques Trienekens Professor, Department of Social Sciences, Management Studies, Wageningen University and Research, Hollandseweg 1, 6706 KN, Wageningen, the Netherlands

Corresponding author: jacques.trienekens@wur.nl

This special issue presents 6 selected papers that participated in the 2017 IFAMA best paper competition, among which the three finalists. The papers cover different subjects representing the broad interests of our agribusiness and food management community. All papers were double blind reviewed, in several review rounds, before being selected for publication: ■■ Sector blending: evidence from the German food bank, by: Meike Rombach and Vera Bitsch (IFAMA 2017 Best Paper competition winner). ■■ Determinants of corporate cash holdings in times of crisis: insights from Brazilian sugarcane industrial companies, by: Aviner Augusto Silva Manoel, Marcelo Botelho da Costa Moraes, David Ferreira Lopes Santos and Marcos Fava Neves (IFAMA 2017 Best Paper competition finalist). ■■ Job preferences of agricultural students in Germany – A choice-based conjoint analysis for both genders, by: Stephan G.H. Meyerding (IFAMA 2017 Best Paper competition finalist). ■■ Allocating production risks through credit cum insurance contracts: the design and implementation of a fund for small cotton growers to access market finance, by: Mauro Alem and Julio Jorge Elias. ■■ Governance of market-oriented fresh food value chains; export chains from New Zealand, by: Jacques Trienekens, Mariska van Velzen, Nic Lees, Caroline Saunders and Stefano Pascucci. ■■ Social sustainability in the ready-made garment sector in Bangladesh: an institutional approach to supply chains, by: Laura A. Carlson and Vera Bitsch.

179

Trienekens

Volume 21, Issue 2, 2018

We thank all the reviewers that have put (a lot of) effort in reviewing all the papers that participated in the competition, in time!: ■■ Ammar Abdul Aziz, The University of Queensland, Australia ■■ Birgit Schulze-Ehlers, Lincoln University, New Zealand ■■ Caroline Saunders, AERU-Lincoln University, New-Zealand ■■ Christopher Peterson, Michigan State University, USA ■■ Gerhard Schiefer, University of Bonn, Germany ■■ Gregory Baker, Santa Clara University, USA ■■ Hammad Badar, University of Agriculture, Faisalabad, Pakistan ■■ Hernan Palau, Universidad de Buenos Aires, Argentina ■■ Jeevika Weerahewa, University of Peradeniya, Sri Lanka ■■ Johann van Rooyen, Stellenbosch University, South Africa ■■ Lilly Camacho, The Commonwealth Scientific and Industrial Research Organisation (CSIRO), Australia ■■ Nic Lees, Lincoln University, New Zealand ■■ Nicola Shadbolt, Massey University, New Zealand ■■ Obi Ajuruchukwu, University of Fort Hare, South Africa ■■ Paul Dalziel, AERU-Lincoln University, New Zealand ■■ Péter Balogh, University of Debrecen, Hungary ■■ Vera Bitsch, University of Munich, Germany. ■■ Verena Otter, Georg-August Universität Göttingen, Germany ■■ Vincent Amanor Boadu, Kansas State University, USA We wish you a lot of reading pleasure! Jacques Trienekens, Wageningen University, The Netherlands Special issue editor

International Food and Agribusiness Management Review

180

OPEN ACCESS International Food and Agribusiness Management Review Volume 21 Issue 2, 2018; DOI: 10.22434/IFAMR2017.0064 Received: 12 July 2017 / Accepted: 7 September 2017

Sector blending: evidence from the German Food Bank Special issue: IFAMA 2017 symposium

RESEARCH ARTICLE Meike Rombach

and Vera Bitschb

aResearch

associate, and bProfessor, Technical University of Munich, Chair Economics of Horticulture and Landscaping, Alte Akademie 16, 85354 Freising, Germany

Abstract This study investigates forms of sector blending practiced by the Federal Association of German Food Banks and respective benefits and drawbacks emerging for the organization. The study builds on a prior sector blending categorization system. Twenty in-depth interviews with food bank members were recorded, transcribed, and analyzed using qualitative content analysis. In addition, webpage materials from German food retailers and press releases from the German Food Bank were analyzed. Imitation, interaction and industry creation were forms of sector blending found. The main drawback found was that the food bank fails to serve those outside the German social security system, such as the homeless. Benefits include a more professional image of the organization, and the enhancement of social capital. Overall, the additional generation of funds and resources were the most important aspects of sector blending efforts in the German food bank, as it helps food pantries adjust to an increasing number of people in need. Keywords: charitable food assistance, industry creation, in-depth interviews, qualitative content analysis JEL code: D60, D71, I38, I39 Corresponding author: meike.rombach@tum.de

ÂŠ 2017 Rombach and Bitsch

181

Rombach and Bitsch

Volume 21, Issue 2, 2018

1. Introduction The Federal Association of German Food Banks (abbreviated in the following as German Food Bank) is a charitable nonprofit organization that redistributes surplus food items to people in need (Lorenz, 2012). Founded in Berlin in 1993 as a social movement, the German Food Bank developed into a federal association with an umbrella structure since then (Federal Association of German Food Banks, 2017). The organization consists of 900 local food pantries spread across Germany staffed with about 50,000 volunteers (Von Normann, 2011). The local pantries elect a representative for each of the 16 German states (Selke, 2011a; Federal Association of German Food Banks, 2017). These state representatives form an executive board, which is supported by a head office and an advisory body (Federal Association of German Food Banks, 2017; Selke, 2011a). According to the homepage of the Federal Association of German Food Banks (2017), all representatives meet once a year in a general assembly where the executive board is elected. Local pantries differ in size, financial resources and operational structure (Von Norman, 2011). The webpage of the Federal Association of German Food Banks (2017) distinguishes between free and associated food pantries. While free pantries are independent in terms of their operating policies and procedures, they maintain close contact with the umbrella organization. Associated pantries, in contrast, are required to follow the umbrella organization’s policies and procedures. Food pantries in Germany served 1.5 million users in 2016 (Federal Association of German Food Banks, 2017). Pantry users are socially disadvantaged people suffering from food insecurity (Selke, 2011a). The main goal of the German Food Bank is to form a bridge between socially and economically disadvantaged people and food retailers who provide surplus food items which cannot be sold in retail outlets. The items could be mislabeled, out of season or have defects in packaging, but must be safe for consumption (Federal Association of German Food Banks, 2017). In addition to this bridging function, the German Food Bank actively fights food waste (Lorenz, 2012). Both goals are explicitly stated in the organizational motto ‘Food, where it belongs’ (Federal Association of German Food Banks, 2017). The German Food Bank’s activities can be seen as efforts to avoid food items being discarded. Further services provided by the German Food Bank are handing out meals in soup kitchens, providing breakfast to school children, distributing second-hand clothing and assisting users with medical appointments, banking tasks, dealing with government authorities and similar tasks (Lorenz, 2012; Reiniger, 2011; Von Normann, 2011). The German Food Bank and its activities have been well researched. Existing studies focus on the sociological and political aspects of poverty and food insecurity (Selke, 2011a), the food bank user perspective in the context of dignity (Hoffmann and Hendel-Kramer, 2011) and vulnerability (Sedelmeier, 2011). Other studies address public perceptions of the organization and its services (Häuser, 2011; Witt, 2011). Several studies (Selke, 2011b; Werth, 2011; Witt, 2011) indicate that the German Food Bank’s operations are not purely nonprofit, but in fact overlap with the for-profit and public sectors. This overlap in operations can be framed as sector blending. Sector blending occurs when an organization operating in a defined sector uses various approaches, activities, and relationships that blur the distinctions between nonprofit, public and for-profit organizations. The organization operates either in a manner similar to organizations in other sectors or operates in the same realms or both (Dees and Battle Anderson, 2003; Park, 2008). Dees and Battle Anderson (2003) emphasize the importance of sector blending as a means of organizational survival, as a way to establish legitimacy in the operating sector and to gain resources. At present, the number of food bank users is steadily increasing, of which mismatch between the quantity of food available in local food pantries and the number of users arises (Federal Association of German Food Banks, 2017; Hoffmann and Hendel-Kramer, 2011; Von Normann, 2011). Therefore receiving donations and funds is particular important for the organization, and sector blending can contribute to receiving both. However, because its operational principles strictly emphasize the non-profit nature of the organization, adopting for profit-business practices also could jeopardize the German Food Bank’s reputation and existence. As the Food Bank has become an important factor in the German social welfare system, the demise of the organization could lead to serious International Food and Agribusiness Management Review

182

Rombach and Bitsch

Volume 21, Issue 2, 2018

consequences for people in need. In this context, the present study adds an organizational perspective to the existing body of literature on food pantries in Germany, in which the German Food Bank’s operations are examined for evidence of sector blending. The objectives are to explore which forms of sector blending are employed by the German Food Bank, and which potential benefits and drawbacks emerge from sector blending for the organization.

2. Literature review Sector blending occurs due to global and cultural shifts (Wachhaus, 2013), as a consequence of organizations striving for market power and functionality (Bromley and Meyer, 2014), or due to resource dependencies and political pressure (Bromley and Meyer, 2014). The decline of governmental control in markets leads to market pressures that drive nonprofit organizations to adopt strategies similar to that of organizations operating in the for-profit and public sectors (Bromley and Meyer, 2014). Prior studies provide various examples of sector blending between the for-profit, nonprofit and the public sectors (Alcock, 2010; Billis, 1993; Brandsen et al., 2005; Park, 2008; Sagwa and Segal, 2000; Tuckman, 1998). The creation of ‘hybrid organizations’ (Evers, 2005; Hoffmann et al., 2012; Ménard, 2004; Pache and Santos, 2013), collaboration (Brandsen et al., 2005) and contracting (Dees and Battle Anderson, 2003) are commonly identified practices. The housing, health and educational sectors in the U.S. serve as examples where sector blending has taken place (Dees and Battle Anderson, 2003). Dees and Battle Anderson (2003) grouped various forms of sector blending into four categories: imitation and conversion, interaction, intermingling and industry creation (Table 1). These four categories are closely interrelated and the different forms of sector blending also occur in association with one another. Further, Dees and Battle Anderson (2003) explored benefits and drawbacks of sector blending. As benefits, they emphasized effective and appropriate resource allocation, sustainable solutions to social problems for social entrepreneurs, increased accountability for nonprofit organizations and increased financial strength. In terms of drawbacks, they enumerated threats to social performance (including mission drift), decrease in quality of service and decline in the level of advocacy against social injustice. Another critical aspect they addressed is the potential of undermining indirect social benefits. For example, Dees and Battle Anderson (2003) mentioned reduction in the creation of social capital as well as loss of the charitable character of the nonprofit sector. Another negative aspect of sector blending Dees and Battle Anderson (2003) emphasized is the creation of classes among service recipients. Social capital has been extensively studied in various disciplines (e.g. in sociology, Coleman, 1988; in philosophy, Bourdieu and Wacquant, 1992; in political science Putnam, 1995; in economics, Portes, 1998, refined concepts described by Bourdieu and Coleman), each of which has its own definition. Dees and Battle Anderson (2003) do not explicitly state which line of theory or school they follow. Instead, they provide examples of how social capital is created through members of organizations, within the organization and externally with other stakeholders and organizations (Dees and Battle Anderson, 2003: 23). These examples cover a broad enough spectrum to encompass the main themes included in each of the different schools mentioned above. Schneider (2009: 2) defined social capital as ‘relationships based in patterns of reciprocal, enforceable trust that enable people and institutions to gain access to resources like social services, volunteers, or funding.’ In developing this definition, Schneider identified the key elements common to all of the various schools of social capital theory – namely, networks, trust, norms and culture – and differentiated between different forms of social capital. The examples provided by Dees and Battle Anderson (2003) seem to relate well to Schneider’s definition. Accordingly, Schneider’s definition (2009) is used in the context of the present study, because it also fits well with the German Food Bank’s operations. To better understand the evidence of sector blending in the activities of the German Food Bank, the organizational structures of the German Food Bank are revisited, as these structures determine its operations. The term operations is used here to refer to the day-to-day activities of the food bank that serve both to create

International Food and Agribusiness Management Review

183

Rombach and Bitsch

Volume 21, Issue 2, 2018

Table 1. Sector blending classification system (adapted from Dees and Battle Anderson, 2003). Category

Explanation

Imitation and conversion

Adoption of strategies, concepts and practices of the business world; identifying target markets and consumer segments; application of management strategies and tools; charging fees for services provided; changing from non-profit to for-profit status. Example: Habitat for Humanity in the U.S. – organization building houses for those who otherwise could not afford them; in exchange, the new owners must repay a modest mortgage (generating fees for services).

Interaction Competing Contracting

Collaboration

Nonprofit organizations entering profit or public sectors and competing for market share, resources or consumers. Example: hospice care in the US For-profit organizations contracting with nonprofit organizations for both ‘nonprofitlike’ goods and services and goods and services traditionally provided by other for-profit businesses. Example: universities contracting with technology companies to transfer curriculum to media suitable for distance learning. Nonprofit organizations and for-profit organizations entering strategic partnerships and joint ventures that provide mutual benefits. Example: an anti-hunger organization in the U.S. entering into an agreement with American Express to market and raise money via cause-related marketing campaigns.

Intermingling

Creation of ‘hybrid organizations’; hybrid organizations are organizations, networks or umbrella groups with for-profit as well as nonprofit components; hybrid organizations allow nonprofit organizations to conduct activities not fitting neatly into the nonprofit structure. Example: Girl Scouts in the U.S. and the charity organization ‘United Way’ establishing for-profit subsidiaries to generate revenues to support their nonprofit programs by selling equipment and merchandise to local organizations and licensing the organizations’ names and logos.

Industry creation

Establishment of a new market within an existing market. Example: charter schools in the US are independent public schools that are often run and managed by parent-teacher partnerships, nonprofits, universities, and for-profit companies; some charter schools are new schools, others have been converted from traditional public schools; in return for demonstrated results, these schools are allowed to operate outside of traditional rules and regulations of the public school system.

unique value and achieve core objectives. Resources, capacity and output usually influence operations. As shown by Dees and Battle Anderson (2003), these three aspects are expected to improve due to sector blending. Based on the international literature on food bank management (González-Torre and Coque, 2016; McIntyre, 2015; Vitiello et al., 2015), particularly studies describing food bank structures and operations (Baglioni et al., 2016; Mohan et al., 2013; Tarasuk and Eakin, 2003, 2005), German food pantries can be classified as ‘front line systems’. In contrast to ‘logistical systems’ where food bank operations are focused on food bank donors and distribution to users is carried out by other organizations, front line system are dedicated to serve food bank users (Baglioni et al., 2016). The pantries operate either as soup kitchens, where prepared food is served (Lorenz, 2012); as distribution centers, where available food items are apportioned to food bank users by food bank volunteers (Selke, 2011a); or as supermarket-like pantries, where food bank users can choose from the available options within a monetary limit that is based on their level of need (Werth, 2011).

International Food and Agribusiness Management Review

184

Rombach and Bitsch

Volume 21, Issue 2, 2018

To maintain the pantries and guarantee service to food bank users, collaborations with sponsors and donors are essential. Selke (2011b) and Witt (2011) explained the benefits of collaboration for food donors and the German Food Bank using the example of event weeks carried out by a German food retail chain that invite consumers to buy products and place them in bags for donation to food pantries. The retailer donates the purchased food items to the food bank, and adds a predetermined quantity of products to the donation. The practice creates a win-win situation for the retailer and the food pantries, as it confers a positive image to the retailer while providing a large quantity of food donations to the food pantries (Witt, 2011). The Food Bank advertises collaboration by retailers and social welfare organizations its collaboration partners to food pantry users as well as on their internet site. Usually collaborations work well, because both parties benefit; but operational conflicts can occur in case of dependency on individual donors or sponsors (Hiss, 2010; Selke, 2011b). Selke (2011b) addressed another operational conflict related to brand protection. The Federal Association of German Food Banks protected the German name for food pantries ‘Tafel’ (dining table) as a brand name. Accordingly, only food pantries belonging to or associated with the organization are allowed to use the name, even though other nonprofit organizations provide the same service (Reiniger, 2011; Selke, 2011b). According to Selke (2011b), this creates conflict and hinders collaboration. Studies by Lorenz (2010b) and Witt (2011) researched the consequences of branding and the monopoly that resulted from it. They focused on the brand concept and emphasized the logo (a plate with cutlery and the slogan ‘Food where it belongs’). The brand concept can be associated with wealth, while food pantries are associated with poverty. Since logo and slogan should depict the reality behind the brand, both authors considered the branding questionable. German and international studies have highlighted equity and dignity as key operational principles for food pantry volunteers when serving food pantry users (Hoffmann and Hendel-Kramer, 2011; Lorenz, 2010a; McIntyre et al. 2015; Tarasuk and Eakin, 2003; Van der Horst et al. 2014). Volunteers are expected to distribute the available food fairly amongst the users, to treat them respectfully and avoid any form of shaming (Tarasuk and Eakin, 2003). In practice, these operational principles are not always followed. Structural asymmetry between users and volunteers can cause conflicts (McIntrye et al. 2015; Reiniger, 2011; Selke, 2011b, 2016; Van der Horst et al., 2014). With respect to dignity, Hoffmann and Hendel-Kramer (2011) found that in German food pantries, managers decide whether the food is provided free of charge or for a symbolic payment. Such payments are usually 10-30% of the prices charged at a local retailer. Payments are adjusted to the benefits users receive according to the Social Security Code (Hoffmann and Hendel-Kramer, 2011; Lorenz, 2010a). Von Normann (2011) pointed out that having to provide an official proof of neediness from the welfare agency to receive food from a food pantry was perceived as humiliating by users. However, the German Food Bank decided that the eligibility assessment was necessary, because deception had been documented (Von Normann, 2011). Lorenz (2010b) and Hoffmann and Hendel-Kramer (2011) criticized the German Food Bank on three main points. First, they posited that the German Food Bank has strayed from its foundational idea by bringing regular food items into the food pantry system, referring to the acquisition of non-surplus food items (e.g. through event weeks) in response to insufficient supply. Second, they confirmed prior work by Becker (2010) showing that the German Food Bank had also started addressing other aspects of poverty, such as education and youth unemployment. They postulated that food bank volunteers charged with these tasks are not necessarily qualified and suggested that these tasks were more appropriately carried out by what they see as better qualified personnel in other recognized German welfare organizations. Finally, they suggested that the German food bank aims to reach too many different target groups by opening food pantries for pets and children, and portrayed such operations as self-sustaining measures and attempts to create a social market within an existing social market (see also Selke, 2016). Judging from the studies discussed above, the operations of the German Food Bank include various forms of sector blending (Dees and Battle Anderson, 2003). The establishment of different food pantry types (Lorenz, 2012; Selke, 2011a; Werth, 2011), such as soup kitchens (Lorenz, 2012) and pantries with a supermarket International Food and Agribusiness Management Review

185

Rombach and Bitsch

Volume 21, Issue 2, 2018

character (Werth, 2011), can be summarized as imitation, as these structures mimic German food retail or food service venues on a smaller scale. Another example of imitation (in this case of the public sector, namely social security provision) is the eligibility assessment required of food bank users (Hoffmann and Hendel-Kramer, 2011; Von Normann, 2011). Collaborations with donors and sponsors (Witt, 2011) as well as with food pantry users (Lorenz, 2012) are forms of interaction. Similarly, brand creation can be interpreted as a form of interaction, more precisely, avoiding competition (Reiniger, 2011; Selke, 2011b). By the same token, brand creation and monopoly building represent a form of imitation of the for-profit sector. Whether sector blending is strictly beneficial to the German Food Bank or whether drawbacks also occur has not yet been studied.

3. Methods The study uses a qualitative research approach, because a qualitative approach is appropriate when a theory is to be developed, an unknown research topic to be explored, or a new perspective to be added to a previously well-investigated topic (Bitsch, 2005). The exploration of operations of the German Food Bank in relation to sector blending and its potential benefits and drawbacks takes a new perspective on researching food pantry operations. According to Bitsch (2005), a qualitative research approach is necessary when a study focuses on the perspectives and experiences of actors in their everyday lives. Further, it allows the identification of cultural framings and social realities (Bitsch and Yakura, 2007; Perera et al., 2016). Exploring food bank operations and forms of sector blending requires examination of the perspectives of multiple actors involved in these operations (Darbyshire et al., 2005). Also, food pantry users are considered a vulnerable population, which are more effectively reached using qualitative in-depth interviews (Hsieh and Shannon, 2005); thus, this approach was used here. Qualitative in-depth interviews allow the researcher to obtain answers to questions of a sensitive nature. This was particularly important in this study, as sector blending is not necessarily perceived as entirely positive. Potential negative effects on other parties might lead food pantry managers to avoid revealing operations that might be construed as sector blending (Hoffmann and Hendel-Kramer, 2011; Von Normann, 2011). Twenty in-depth interviews were conducted in August and September 2015 and from January to March 2016. Following Suri (2011) and Noy (2008), emphasizing purposeful sampling as particularly suitable for qualitative research, the interviewees’ roles at the German Food Bank, respectively their relationship with it, as well as their potential knowledge of forms of sector blending and their potential control over or benefits from sector blending activities were considered (Table 2). Interviewees included five food pantry managers, four food pantry users, five volunteers, one spokesperson from the head office of the Federal Association of the German Food Banks and five retail food donors, and represented both free and associated food pantries. Each interview lasted 60 to 90 minutes. Interviews were carried out either face-to-face or by phone, according to the interviewees’ preferences. A semi-structured interview guide outlined the main topics of the interviews. Topics were addressed through open-ended questions that were asked according to the interview flow. The topics discussed included food waste, food redistribution as well as other food pantry operations. The initial interview guide focused on food redistribution and volunteering at the food pantries. During the research process, modifications to the research guide were necessary in order to better reflect each interviewee’s duties as well as the terminology they used (see also Perera et al., 2016). Eighteen of the twenty interviews were audio recorded and transcribed verbatim before data analysis. On two occasions, the interviewees did not agree to audio recording and, therefore, field notes were taken. In addition to the interviews, information from webpages of the German Food Bank and food retailers were analyzed. The internet site of the Federal Association of German Food Banks served as the starting point. Press releases and other materials from the German Food Bank published between 2015 and 2016 were collected and analyzed. As the site also identified partnerships with food donors and sponsors, information from collaboration partners’ webpages were also included in the analysis. Field notes, press releases and interview transcripts were analyzed using qualitative content analysis. During the analysis process, raw text International Food and Agribusiness Management Review

186

Rombach and Bitsch

Volume 21, Issue 2, 2018

Table 2. Purposeful sampling approach. Group of interviewees

Role in/for the organization

Information sought from interviewee1

Assumed level of control over or benefits from sector blending activities1

Managers of local food pantry

Manage food pantry operations

Managers are knowledgeable about all types of food bank operations within their food pantry, within the umbrella organization and about external collaboration partners. They are knowledgeable about all forms of sector blending practiced by the German Food Bank.

Managers have control over sector blending activities. If the organization would not benefit from sector blending, these efforts would not be undertaken.

Food pantry users

Receive food

Food bank users are knowledgeable about the forms of sector blending they perceive when coming to the food bank to receive food and interact with personnel and other users. They are likely knowledgeable about forms of imitation and possibly interaction, if collaboration partners are advertised by the food pantries personnel.

Users benefit from sector blending activities but have no control over it.

Volunteers at a Serve food bank food pantry users, sort and pack food, collect food from retail chains

Due to volunteersâ&#x20AC;&#x2122; diverse tasks, they are likely to be knowledgeable about imitation and interaction. Since they are not involved in strategic decisions of the umbrella organization, they are not likely knowledgeable about industry creation.

Volunteers mostly follow the instructions of mangers. They do not have direct control over sector blending activities, and also do not benefit.

Spokesperson Official of the German communications Food Banks

The spokesperson communicates on behalf of the umbrella organization. Therefore, she is expected to know all forms of sector blending practiced by the German Food Bank.

The spokesperson has no control over sector blending activities, and also does not benefit.

Managers in food retail chains

Managers in ford retail chains are expected to be knowledgeable about forms of interaction, due to the collaborations between food retail chains and the German Food Bank.

Retail chains are likely to benefit from sector blending activities; otherwise, the collaboration would not exist. Extent and frequency of collaborations provide indirect control to retail chains.

Cooperation with local food bank, food donations

Adapted from Dees and Battle Anderson, 2000; Dart, 2004; Hoffmann and Hendel-Kramer, 2011; Lorenz, 2012; Selke, 2011b; Witt, 2011.

was systematically broken down and common themes were extracted. This was achieved through constant comparison and contrasting of the material. Ultimately, various forms of sector blending carried out by the German Food Bank were identified. The analysis process consisted of two main steps: open coding and the establishment of categories. During open coding, labels were assigned to text fragments. These labels reflected the key thought behind each text fragment. During the coding process, field notes, transcripts and other materials were carefully read several International Food and Agribusiness Management Review

187

Rombach and Bitsch

Volume 21, Issue 2, 2018

times. Throughout the analysis process, codes were conceptualized and relabeled. The coding process linked all relevant interview excerpts with codes and their corresponding definitions (Table 3). In the second step, categories were established by grouping codes according to their meanings and associations. Each category was named using content-characteristic words and defined. Category definitions were created by combining all related codes and their definitions. This process of qualitative content analysis was carried out using the software package Atlas.ti (ATLAS.ti Scientific Software Development GmbH, Berlin, Germany), which allows for systematic analysis of documents. Atlas.ti includes tools for annotating and coding the data. Prior to the analysis of the empirical data, the categorization of Dees and Battle Anderson was used to identify forms of sector blending in the existing body of literature on operations of the German Food Bank (Table 4). Table 3. Codes for the category ‘Industry creation’ with excerpts from interviews and web pages. Code

Excerpt

Receiving regular food items Operations of the food bank reflecting industry creation; the German Food Bank uses different strategies to receive regular food items from German food retailers directly or indirectly through collaborations.

‘From Monday (7 November 2016) on: [German food retailer] and [German food retailer] are asking you to buy a bag with seven products! These products will be donated to our local food banks. The bags are filled with products that are rarely donated due to their long shelf life. Spaghetti, ravioli, ketchup, milk rice, jam, tea and biscuits. The actual cost is currently 5.09 €, but the products are made available for 5.00 € in our markets. [Name of German food retail] will add 40,000 extra bags to the customer donations’ (Announcement on the webpage of a German food retailer).

Focus on other aspects of poverty Operations of the food bank reflecting industry creation; the German Food Bank offers services and products dedicated to other aspects of poverty that are not related to food; services can overlap with those of other welfare organizations.

‘The Food Bank Academy is a 100%-owned subsidiary of the Federal Association of German Food Banks and a nonprofit limited liability company. The purpose of the Food Bank Academy is to promote education. We would like to educate people that come actively to the food bank. We want to ensure the professional work of our boards and support volunteers in their commitment’ (Description on the webpage of the Food Bank Academy). ‘The food bank is helping greatly and is the solution for so many of us. I know that the food bank is supporting four schools in Munich’ (User of an associated food pantry, female, 50-60 years old).

Supply and demand Demand for food is higher than the amount of food that can be supplied in the food pantries; impetus for the German food bank to focus on industry creation.

‘What I am also doing is speaking with the big supermarket chains and the big donors. For example, I really go to the logistics center of [Name of a supermarket in German food retail] or [Name of a discounter in German food retail] to make sure that we have enough supply for the next year. The level of supply has dropped over the last year, but the number of people who want to receive food from the food bank has increased. So now you have an imbalance’ (Manager of an associated food pantry and country representative, male, 60-70 years old).

International Food and Agribusiness Management Review

188

Rombach and Bitsch

Volume 21, Issue 2, 2018

Table 4. Forms of sector blending activities carried out by the German food bank. Form of sector blending

In the literature

Imitation of the for-profit sector Retail structures, food Selke (2011a); service structures Werth (2011); Lorenz (2012)

Choice provided to food bank users

Selke (2011a); Werth (2011); Lorenz (2012)

Home delivery service No for users

In this Excerpt study Yes

Yes

Payment for food

Hoffmann and Hendel- Yes Kramer (2011)

Branding Motto

Witt (2011) Witt (2011)

No No

Imitation of public sector Proving eligibility Lorenz (2010a); Yes Hoffmann and HendelKramer (2011); Von Normann (2011); Lorenz (2012) Rating standards Normann (2011) Yes

Interaction Resource competition

Yes

‘Every station gets basic food, number one, potatoes, carrots, onions. In addition, bread is given out. Yes, yogurt, cheese, sausages. Rarely meat. Meat is not often available. We emphasize a healthy diet for these people [food bank users]. Fruit and vegetables’ (Manager and founder of a free pantry, female, 5060 years old). ‘They give me a small choice. However, in fact, I am not the only one. There more than 100 people and if I would say, I want 10 or 20 bananas, and then they would say, are you crazy? I mean I cannot’. (User of an associated food pantry, female, 20-30 years old). ‘Some guests get the food delivered to their home. We also pack for new families, a box, which is packed beforehand at the distribution center, which we pick up and deliver to them. We do this for the ones that cannot come here, because they are disabled.’ (Volunteer at an associated food pantry, female, 60-70 years old). ‘So they have the feeling of buying something with very little money. Nevertheless, they have the feeling of buying something. And for us it’s about 300 € income per week’ (Manager of an associated food pantry, male, 50-60 years old). Not applicable Not applicable

‘What helps us, is that, at the entrance of the food bank, the customers have to show their certificate’. (Volunteer at an associated food pantry, female, 2030 years old). ‘And so we separate them [food bank users]...the ones on Wednesday have green passes, and for Thursday they have white passes’ (Manager of an associated food pantry, male, 50-60 years old).

‘Lunch table! Our food bank in [Name of a North German town] recently got competition from the neighboring town. Another organization in [Name of the neighboring town] is collecting surplus food items for their purposes’ (Announcement on the webpage of an associated food pantry in North Germany).

International Food and Agribusiness Management Review

189

Rombach and Bitsch

Volume 21, Issue 2, 2018

Table 4. Continued. Form of sector blending Interaction (continued) Competition with other organizations providing same service related to food Competition with other welfare organizations Collaboration with sponsors and donors

Collaboration with other welfare organization

Contracting Intermingling Industry creation Supply and demand

In the literature

In this Excerpt study

Selke (2011b); Reiniger (2011)

Not applicable

Selke (2011b)

Not applicable

Selke (2011b); Witt (2011); Lorenz (2012); Von Normann (2011) Selke (2011b)

Yes

‘So companies donate their stuff, big companies like [name of an online retailer], they put their stuff in these containers to help’ (Manager and founder of a free pantry, female, 50-60 years old). ‘For many people it is quite normal to have a warm meal every day. Moreover, to eat something healthy. For many people this is not normal. For example, they do not have enough money. That is why there is the food bank’ (Announcement on the webpage of a welfare organization dedicated to helping people with disabilities). Not applicable

Not applicable

Yes

Lorenz (2010b); Yes Hoffmann and HendelKramer (2011)

Focus on other aspects Becker (2010); Lorenz Yes of poverty (2010b); Hoffmann and Hendel-Kramer (2011); Selke 2016 Bring regular food items into the food bank system in addition to surplus items

Lorenz (2010b); Yes Hoffmann and HendelKramer (2011)

‘Yes 200 people on Wednesday and 120 people on Thursday. On Thursday there are almost only refugees...90% of the refugees are Muslims. They do not eat pork and mostly we get pork’ (Manager of an associated food pantry, male, 50-60 years old). ‘After our lunch, at around 2 p.m., we support the children doing their homework. The ones that would like to can sit down in our food bank café and do their homework. If needed, we are there to help’ (Offer on the webpage of a food pantry for children in Eastern Germany). ‘Since 2009, [Name of a German food retailer] has been involving its customers in the engagement with the food bank. Every year, for the period of a few weeks, our customers and we are ‘filling plates together.’ In 2014, a total of 318,000 food bags were given to the local food banks’ (Announcement on the webpage of a German food retailer).

International Food and Agribusiness Management Review

190

Rombach and Bitsch

Volume 21, Issue 2, 2018

4. Results and discussion Forms of sector blending identified in the recent body of literature reviewed above focus predominantly on operations taking place within food pantries, partnerships with sponsors and donors, and the process of serving food pantry users. Therefore, the following sections are structured accordingly. The final section describes the benefits and drawbacks of sector blending for the German Food Bank. The discussion includes both, findings from this study as well as forms of sector blending identified in prior studies. 4.1 Forms of sector blending related to operations within German food pantries Given that different forms of food pantries exist in Germany (see also Lorenz, 2012; Werth, 2011), operations and duties described by the food bank volunteers interviewed are specific to the pantry in which they serve. Volunteers emphasized driving and reloading food trucks, inspecting, sorting and preparing food as their main duties in food pantries. Even though the operations volunteers are involved in seem diverse, they are related to food pantry users and how they receive food in the pantries. While some volunteers reported serving users coming to the pantry, others emphasized operations such as home delivery. Yeah. You can choose, but not everything. We give them prepared options they can choose. It is broccoli or cabbage, what do you like? (Volunteer at an associated food pantry, female, 20-30 years old). Some guests get the food delivered to their home. We also pack for new families, a box, which is packed beforehand at the distribution center, which we pick up and deliver to them. We do this for the ones that cannot come here, because they are disabled. The ones that cannot carry the food by themselves get it delivered to their home. (Volunteer at an associated food pantry, female, 60-70 years old). As shown in the statements on the food pantry operations in which volunteers are involved, forms of sector blending are present. The volunteersâ&#x20AC;&#x2122; statements are focused on the imitation of practices found in food retail, for instance, providing a variety of products or home delivery service. These findings correspond to the findings of Hoffmann and Hendel-Kramer (2011), Von Norman (2011) and Werth (2011). Von Normann (2011) discussed the existence of various types of food pantries which operate differently. Adding to this information, Werth (2011) more explicitly outlined the existence of supermarket-like food pantries where users are able to choose relatively freely from the available food items. As these prior studies were not dedicated to sector blending, these activities were not identified as forms of imitation. The sociological background of the prior studies likely led the authors to interpret their observations as forms of dignity and respect towards food pantry users (Hoffman and Hendel Kramer, 2011). When discussing their duties, managers focused on challenges, for instance, how to prevent cheating by food pantry users. They reported further challenges, such as the acquisition of funds and food donations. Managers emphasized food distribution schemes, quantities of food and partnerships with sponsors and donors. They mentioned their personal efforts as well as the importance of the Federal Association of German Food Banks as an umbrella organization with regard to the generation of funds and donations. We have different colors, yellow, pink and green. Then you know immediately. A person can come only once to each station. Through the color system they cannot try again through the back door. (Manager of an associated food pantry and country representative, male, 60-70 years old). We [the Federal Association of German Food Banks] have an agreement with them [a German wholesaler for food and tobacco] to keep 40 pallets. Therefore, we receive a pizza delivery from [a company producing frozen pizza]. And biannually, we receive a delivery of pizza on pallets. So, we can keep for example 30 pallets in our pantry in [a city in Bavaria]. The rest is distributed in other

International Food and Agribusiness Management Review

191

Rombach and Bitsch

Volume 21, Issue 2, 2018

pantries following the plans of the Federal Association. (Manager of an associated food pantry, male, 60-70 years old). Eleven million euros – We thank all [a discounter in German food retail] customers. Eleven million euros this incredible amount of money has been donated by [a discounter in German food retail] customers at the bottle deposit machine. (Headline on the webpage of a discounter in German food retail). The first manager statement above emphasizes the use of colors as a measure to ensure fair food distribution. This also reflects a form of imitation. In this case, the food pantry has borrowed measures from the public sector. The colors reflect the extent of neediness, evidence of which is required before becoming a food bank user. Similarly, before receiving any basic benefits, users must prove to the social security office the extent of their neediness. Standard rates of basic security benefits are payed accordingly. The managers interviewed and the webpages of German food retail discounters and supermarkets all reported on the amounts of food and funds being donated. In addition, webpages in particular, emphasized the partnerships between various food retail chains and the Federal Association of German Food Banks. According to Dees and Battle Anderson (2003), partnership presents a form of interaction, more precisely, collaboration between a non-profit organization and a for-profit organization. 4.2 Forms of sector blending within operations with respect to sponsors and donors German food retailers are important collaboration partners of the German Food Bank and are deeply involved in food bank operations. Within the partnership, both partners need to be able to trust each other. Accordingly, it is important to understand whether retailers evaluate the collaboration in the same positive light as it is presented by the Federal Association of German Food Banks. Some of the retailers interviewed confirmed the positive nature of the collaboration, while others expressed concern and shared examples of conflicts between themselves and the food bank. Other retailers addressed the issue of resource competition amongst various organizations involved with collecting surplus food items. We cannot stand them anymore; this is not correctly put. Well, those people came again and again. They always came back and said: ‘this is not alright. We will take this with us, but not that.’ But it depends on the person [the food bank volunteer who is collecting the food items]. Now we just donate it all to the...youth center. This does not hurt me, because I think, now the food is with someone who is very active in youth work, and the kids are very happy, even though the chocolate is expired. (Manager of a German food retail chain, female, previously donated to the German Food Bank once a week). Because after all it is the acceptance. It is not easy to explain to someone, clearly the food bank, they already do it, and now how should I best explain that? They are giving it to the needy...But again, on the other side, the Food Sharers, what they do is simple, they distribute it among themselves. There are certainly many who are not needy. And there it would, I think, diverge a bit. (Manager of a German food retail chain, male, donates once a week to Food Sharing). Some of the retail managers interviewed stated that they preferred to collaborate with organizations other than the German Food Bank, some stated that they wanted to end their current collaborations with the food bank, and some stated they had never wanted to collaborate with the German Food Bank to begin with. As these interviewees are giving their surplus items to other nonprofit organizations, competition was identified as another form of interaction. In contrast to Witt (2011) who highlighted the German Food Bank’s efforts to avoid competition with other organizations providing the same service through branding and monopoly creation, the present study found that resource competition could not be fully avoided. The contrasting findings can be explained by retailers’ evaluation of the partnership. Witt (2011) stated that both parties – food pantries and retailers – benefited from the collaboration, and that retailers are positively perceived by

International Food and Agribusiness Management Review

192

Rombach and Bitsch

Volume 21, Issue 2, 2018

society due to the collaboration. However, in the present study, some retailers interviewed feared reputational damage caused by publicity surrounding donations of surplus food items. 4.3 Forms of sector blending within operations related to food pantry users When focusing on the food pantry user as one of the central actors in the operations of the German Food Bank, managers, volunteers, donating retailers and the users themselves enumerated problems related to the current user situation and users’ demand for food as well as supply problems. Interviewees highlighted an increase in the number of food pantry users and a decline in the amount of available food. The interviewees mentioned increasing poverty among the elderly, provision of food to refugees and efforts to increase the amount of food available for users. The reason for the decrease in goods is that retailers have changed the way they do their logistics. Before they calculated more days and then received the goods from their logistics centers for the days they calculated. And if not enough was distributed then they had to give them out...This acquisition is now done daily, yeah. So now, they do not work with such a huge amount of goods but with fewer. (Volunteer at an associated food pantry, male, 60-70 years old). Well, the first reason is that we have many more asylum seekers. That is the first reason. The second reason is that the income of the women whose husbands have died is 400 €. That means that is the problem. And it is just getting worse. Yeah. Unfortunately, that is how it is. The women or retirees who come to the food bank have never worked before or do not work anymore because the husband worked and they looked after the children. Therefore, they have no income anymore. That is the problem today. (Manager of an associated food pantry and country representative, male, 60-70 years old). I did not know about it. It was a little bit a gift for me. There were some presents for my children on Christmas. It was a big event at the food bank...I feel thankful for it. It is very nice. Every child whose parents get standard social benefits gets a present. Lovingly done by people. Gifts, yeah! Socks or hats...and books, gloves. Very nice. (User of an associated food pantry, female, 20-30 years old). If you are a customer and see, uh, I can help the food bank, you buy the bags. And then the bag gets a sticker and goes to the food bank. And with this we also bring this idea to the public. (Manager of a German food retail chain, male, donates to the German Food Bank once a week). The statements presented indicate industry creation as another form of sector blending. Interviewees described features of a market system, alluded to the concepts of supply and demand and the inclusion of food items in the German Food Bank system which are not surplus items. They also elaborated on aspects of poverty that are not related to food but are being addressed by food banks. These findings corroborate findings by Hoffmann and Hendel-Kramer (2011). The criticism that the German Food Bank is not adhering to its foundational idea by carrying out activities not related to food and food insecurity may be appropriate from Hoffmann and Hendel Kramers’ point of view. However, from an economic point of view, the food bank is simply adapting to market conditions, which have changed between 1993 and 2017. The concern over not only supplying surplus food items is also questionable. Since 1993, the number of food pantries in Germany has increased to 900. The number of food pantry users has increased as well during this period (Federal Association of German Food Banks, 2017; Von Norman, 2011). In order for the services provided by the German Food Bank to be more than purely symbolic (i.e. to actually come close to meeting demand), it has become necessary to find ways to bring regular food items into the system. The inclusion of other market segments such as children, pets, and refugees is a consequence of cultural changes (Bromley and Meyer, 2014) and represents the German Food Bank’s adjustment to these changing conditions.

International Food and Agribusiness Management Review

193

Rombach and Bitsch

Volume 21, Issue 2, 2018

The various operations of the German Food Bank identified in the present study correspond with the sector blending categorization system described by Dees and Battle Anderson (2003). Aside from contracting (which is a form of interaction) and intermingling, all forms of sector blending described by Dees and Battle Anderson were identified in the present study (Table 3). Whether contracting and intermingling are not used by the German Food Bank or whether they could be identified in other materials than the ones analyzed remains to be determined in future research. Because the German Food Bank is advertising its nonprofit and charitable character, these forms are most likely not used, as they could harm the organizationâ&#x20AC;&#x2122;s image. Donors and sponsors might not want to support any form of for-profit business. The forms of imitation identified here were mainly related to concepts and practices common in German retail operations and furthermore in the German public sector, most notably the requirement for provision of proof of neediness in order to receive social security benefits. In contrast to findings from prior studies, aspects of industry creation identified at the German Food Bank were not found to be negative, but rather present necessary adaptations to cultural and market developments. 4.4 Benefits and drawbacks of sector blending The sector blending activities identified here with regard to the German Food Bank have both benefits and drawbacks for the organization. All forms of sector blending found in the present study as well as operations described in prior studies and identified as forms of sector blending here (Table 3) are evaluated below, following the classification of Dees and Battle Anderson (2003) (Figure 1).

Collaboration to receive public attention

Collaboration to receive resources

Interaction

Quasimonopoly leads to competitive advantage Quasimonopoly hinders exchange

Branding shows food bank is professional Payment leads to accountability Structure enhances social capital

Imitation of the forprofit sector

Eglibility check helps to control demand

Exclusion is an act against the mission

Imitation of the public sector

Federal Association of German Food Banks

Payment leads to additional income

Self sustaining

Competitive advantage

Industry creation

Drawing attention to societal problems

Figure 1. Overview of the benefits and drawbacks of sector blending. Benefits are depicted in green and drawbacks in red. International Food and Agribusiness Management Review

194

Rombach and Bitsch

Volume 21, Issue 2, 2018

Imitation of food retail and food service structures in German food pantries can be evaluated as beneficial, as these structures allow the creation of social capital. According to Dees and Battle Anderson (2003), social capital can develop at sites where many volunteers are involved. Soup kitchens and supermarket-like food pantries (Lorenz, 2012; Selke, 2011a; Werth, 2011) allow food pantry users to receive food and interact with one another as well as with volunteers. Following Schneider (2009), volunteers can be seen as the connection between the organization, the users and society, because the frequent interaction between volunteers and the users leads to reciprocal trust-based relationships that also benefit the German Food Bank as an organization. Another form of imitation – the establishment of an organizational logo and slogan (Witt, 2011) – is beneficial for the German Food Bank, as branding signals professionalism to current and potential users. In addition to imitation of retail practices and elements of branding, symbolic payments by users (Lorenz, 2010; Hoffmann and Hendel-Kramer, 2011) also have beneficial consequences for the German Food Bank. As explained by Dees and Battle Anderson (2011), shifting from a completely charitable relationship (where food is provided free of charge) to a customer relationship (where users are asked to provide a symbolic payment) improves accountability of the organization. The payment requirement increases the likelihood that food pantry users will hold the organization accountable if the food or services received are not of acceptable quality. Dees and Battle Anderson (2003) posited that paying customers usually provide feedback to the organization, make their complaints public, or seek other options for receiving goods or services in cases where the quality of the product or service is insufficient. Furthermore, charging users a symbolic payment results in a small amount of income for the food pantry, which can be used for other purposes, for instance, to buy equipment. Dees and Battle Anderson (2003) pointed out that earned income can lead to greater financial strength of nonprofit organizations. The German Food Bank’s imitation of the public sector (Hoffmann and Hendel-Kramer, 2011; Lorenz, 2012; Von Normann, 2011) also has both benefits and drawbacks. The eligibility assessment and distribution according to the extent of users’ neediness allows German food pantries to control supply and demand. Moreover, the eligibility assessment avoids cheating and free rider effects. However, the eligibility assessment also causes the exclusion of some people suffering from food scarcity, e.g. homeless people. Since people outside the social security system are excluded, this can be seen as a threat to the social performance of the German Food Bank, reflecting mission drift. According to Dees and Battle Anderson (2003), business structures or operations can steer nonprofit organizations away from their original mission. In the case of the German Food Bank, the original mission was to reduce food waste and provide food to socially and economically disadvantaged people. This mission would include homeless people. Dees and Battle Anderson (2003) further explained that many social services originally intended to serve the extremely poor focus instead on less disadvantaged people, rather than finding ways to serve both populations. Competing with other organizations that provide food to the needy as well as other welfare organizations with different foci also has both benefits and drawbacks for the German Food Bank. Providing education and other services, e.g. the Food Bank Academy (Federal Association of German Food Banks, 2017) can lead to improved service quality and distinction in the market. In addition, the legal protection of the organization’s name and the branding concept create a quasi-monopoly, which provides advantages over other organizations with similar objectives. Consequently, some of these organizations might be unwilling to collaborate and exchange knowledge. The relationships with sponsors, donors and other welfare organizations that consider the German Food Bank as a favorable collaboration partner are beneficial. Sponsors provide resources in the form of food items and funds that are needed by the food pantries (Lorenz, 2012; Selke, 2011b; Witt, 2011). In particular, event weeks dedicated to the German Food Bank have positive effects, because the food bank and its collaboration partners receive positive public attention. During event weeks, the food bank also receives food items rarely donated due to their long shelf life, e.g. canned food and cookies. These extra items received through collaborations are beneficial, since they help adjust the food supply to demand. Further, cooperation with

International Food and Agribusiness Management Review

195

Rombach and Bitsch

Volume 21, Issue 2, 2018

other welfare organizations not deterred by the German Food Bankâ&#x20AC;&#x2122;s quasi-monopoly are beneficial, as resources and knowledge can be exchanged. The present study as well as prior literature (Hoffmann and Hendel-Kramer, 2011; Lorenz, 2010) show aspects of industry creation. Industry creation can be evaluated as beneficial for the German Food Bank, as this form of sector blending has a self-sustaining effect that, from an economic point of view, is considered desirable even for a nonprofit organization. Including other aspects of poverty and different target groups is particularly beneficial, because these forms of diversification can lead to a competitive advantage. Also from a social point of view, such diversification is beneficial, because it helps both to draw attention to societal problems, such as poverty among children and the elderly, and to identify gaps in the German social security system. When evaluating the German Food Bankâ&#x20AC;&#x2122;s operations on the background of sector blending as a whole, social capital is enhanced through the adjustment of structures. But sector blending also leads to the exclusion of potential users from the system through eligibility control. With respect to economic benefits, the German Food Bankâ&#x20AC;&#x2122;s efforts of sector blending are aimed at addressing user demands and adjusting to cultural changes. A resulting drawback is interference with potential collaborations. Since the German Food Bank has remained focused on its mission and kept its charitable character, sector blending can be evaluated as positive overall.

5. Conclusions The study presented identified sector-blending activities in the German Food Bank, and showed that these are largely beneficial in both economic and social terms. Sector blending increases the generation of funds and resources, supports adjustment to increased demand by users, contributes to the professional appearance of the organization and enhances social capital. Drawbacks are mostly of a social nature, because the German Food Bank is not serving people outside the social security system. However, since the organization has neither lost its charitable character nor significantly jeopardized its mission, its activities benefit the German society overall. Sector blending efforts of the German Food Bank are of interest to marketing managers of food retail chains as well as agricultural producers considering collaborations. On the one hand, sector blending efforts emphasize the focus and professionalism of the German Food Bank as a potential collaboration partner. On the other hand, there are drawbacks, e.g. the exclusion of certain groups from the system, which are of importance for potential collaborators wanting to support the extreme poor. Since the present study emphasizes both benefits and drawbacks, it contributes to informed decision making as to whether the German Food Bank is a suitable collaboration partner for sponsors and donors. Furthermore, the study is valuable to organizations providing services similar to those provided by the German Food Bank. Other organizations could fill the niche that it is not serving. This strategy would lead to more effective and appropriate resource allocation in the market dedicated to supporting the poor, by allowing those not participating in the German social security system to be served. This includes not only the homeless, but also others who choose not to use the system due to embarrassment. In addition, the study serves to bring further attention to the fact that food insecurity is not only an issue in developing countries. In many developed countries such as Germany, food insecurity is a societal problem. In Germany, stigmatization of people, e.g. children and retirees, suffering from food insecurity still occurs (Selke, 2011a). Accordingly, governmental campaigns, social media as well as other forms of media reaching a large share of the German population could contribute to increased awareness. The German Food Bank may be willing to support these efforts with accumulated experience and knowledge. A limitation of the present study results from it being carried out in Bavaria, one of the wealthier German states. Inclusion of less wealthy states with higher unemployment, for instance in eastern Germany, may lead to somewhat different results, including other forms of imitation and interaction. However, conversion International Food and Agribusiness Management Review

196

Rombach and Bitsch

Volume 21, Issue 2, 2018

and intermingling are not likely to exist, because their for-profit nature contradicts the German Food Bank’s mission. Furthermore, an important actor was not included in the sampling for the present study, state and local government. Government representatives were not included, because their viewpoints were intensively discussed in prior sociological studies (Selke, 2011a,b; Von Norman, 2011). Still, these actors might have provided insights into forms of interaction between the German Food Bank and the official welfare system, e.g. collaboration and competition. With respect to sector blending, food bank structures and operations in countries with a similar economic situation should be investigated to identify solutions for current problems and further potential to improve food banks’ services and functioning. For example, the U.S. Department of Agriculture (USDA) can be seen as a collaboration partner supporting food pantries with food for emergency programs (USDA, 2017), while in the German context there is no such support from the government directly to the German Food Bank. Instead, there is an underlying tension, because the German Food Bank assumes duties of the German welfare system, and cutbacks in support to the poor have resulted in increased user dependence on the German Food Bank (Selke, 2016). Future research could also investigate the German Food Bank from a management perspective, since prior studies were mainly focused on users and on sociological issues. Accordingly, an investigation from a managerial perspective could benefit the German Food Bank’s operations and effectiveness and may allow comparison with prior U.S. (Vitiello et al., 2015) and European studies (González-Torre and Coque, 2016). For example, it would be beneficial to investigate human resource management in food pantries. As shown in Canadian studies (McIntrye et al., 2015; Tarasuk and Eakin, 2003), European studies (Van der Horst et al., 2014; Von Normann, 2011) and the present study, some of the operations and the sector blending efforts of the German and other food banks resulted in problems for food pantry users. Examples of eligibility assessments leading to exclusion or user shaming by volunteers do not necessarily imply mission drift, but might result from lack of management skills.

Acknowledgements This work was supported by the German Research Foundation (DFG) and the Technical University of Munich within the funding program Open Access Publishing. The authors like to thank E. Kang and D. Hermsdorf for their contributions to data collection and transcription as well as L.A. Carlson for technical editing.

References Alcock, P. 2010. A strategic unity: defining the third sector in the UK. Voluntary Sector Review 1(1): 5-24. Baglioni, S., B. De Pieri and T. Tallarico. 2016. Surplus food recovery and food aid: the pivotal role of non-profit organisations. Insights from Italy and Germany. International Journal of Voluntary and Nonprofit Organizations 27(3): 1-21. Becker, M. 2010. What is the impact of food bank offers on children? [In German: Welche Bedeutung haben Tafelangebote für Kinder?] In: Food Bank Society [In German: Tafelgesellschaft], edited by L. Lorenz. Transcript Verlag, Bielefeld, Germany, pp. 81-90. Billis, D. 1993. Sector blurring and nonprofit centers: the case of the United Kingdom. Nonprofit and Voluntary Sector Quarterly 22(3): 241-257. Bitsch, V. 2005. Qualitative Research: a grounded theory example and evaluation criteria. Journal of Agribusiness 23(1): 75-91. Bitsch, V. and E.K. Yakura. 2007. Middle management in Agriculture: roles, functions, and practices. International Food and Agribusiness Management Review 10(2): 1-28. Bourdieu, P. and J.L. Wacquant. 1992. An invitation to reflexive sociology. University of Chicago Press, Chicago, IL, USA.

International Food and Agribusiness Management Review

197

Rombach and Bitsch

Volume 21, Issue 2, 2018

Brandsen, T., W. Van de Donk and K. Putters. 2005. Griffins or chameleons? Hybridity as a permanent and inevitable characteristic of the third sector. International Journal of Public Administration, 28(9): 749-765. Bromley, P. and J.W. Meyer. 2014. They are all organizations. The cultural roots of blurring between the nonprofit, business, and government sectors. Administration and Society 49(7): 939-966. Coleman, J. 1988. Social capital in the creation of human capital. American Journal of Sociology 94(1): 95-120. Darbyshire, P., C. MacDougall and W. Schiller. 2005. Multiple methods in qualitative research with children: more insight or just more? Qualitative Research 5(4): 417-436. Dart, R. 2004. Being ‘business-like’ in a nonprofit organization: a grounded and inductive typology. Nonprofit and voluntary sector quarterly 33(2): 290-310. Dees, J.G. and B. Battle Anderson. 2003. Sector-bending: blurring lines between nonprofit and for-profit. Society 40(4): 16-27. Evers, A. 2005. Mixed welfare systems and hybrid organizations. International Journal of Public Administration 28(9): 737-748. Federal Association of German Food Banks. 2017. The German Food Bank. [In German: Die Tafel]. Available at: http://www.tafel.de/der-bundesverband.html. González-Torre, P.L. and J. Coque. 2016. How is a food bank managed? Different profiles in Spain. Agriculture and Human Values 33(1): 89-100. Häuser, G. 2011. The functioning of the food bank from the perspective of the federal association. [Die Wirkungen der Tafel aus der Sicht des Bundesverbandes] In: Transformation of Food Banks in Germany [In German: Transformation der Tafeln in Deutschland], edited by S. Selke and K. VS Verlag für Sozialwissenschaften, Wiesbaden, Germany, pp. 111-117. Hiss, S. 2010. Do business have social responsibility when they support the food banks? [In German: Übernehmen Unternehmen mit ihrer Unterstützung der Tafeln gesellschaftliche Verantwortung?] In: Food Bank Society [In German: Tafelgesellschaft], edited by L. Lorenz. Transcript Verlag, Bielefeld, Germany, pp. 69-80. Hoffman, A.J., K.K. Badiane and N. Haigh. 2012. Hybrid organizations as agents of positive social change: bridging the for-profit and non-profit divide. In: Using a positive lens to explore social change and organizations. building a theoretical research foundation, edited by K. Golden-Biddle and J.E. Dutton. Routledge, New York, NY, USA, pp. 131-152. Hoffmann, H. and A. Hendel-Kramer. 2011. A dignified offer? Survey of food bank users on impacts of food banks [In German: Angebot zur Würde? Befragung von NutzerInnen zu den Wirkungen der Tafeln]. In: Transformation of food banks in germany [In German: Transformation der Tafeln in Deutschland], edited by S. Selke and K. Maar. VS Verlag für Sozialwissenschaften, Wiesbaden, Germany, pp. 123-136. Hsieh, H.F. and S.E. Shannon. 2005. Three approaches to qualitative content analysis. Qualitative Health Research 15(9): 1277-1288. Lorenz, S. 2010a. Are food bank users customers, and should therefore pay at the food bank? [In German: Sind Tafelnutzer Kunden und sollen sie deshalb bei der Tafel zahlen?] In: Food Bank Society [In German: Tafelgesellschaft], edited by L. Lorenz. Transcript Verlag, Bielefeld, Germany, pp. 91-102. Lorenz, S. 2010b. New tasks for the food banks? Socio ecological means and purposes of food bank work [In German: Neue Aufgaben für die Tafeln? Zu sozioökologischen Mitteln und Zwecken der Tafelarbeit]. In: Food Bank Society [In German: Tafelgesellschaft], edited by L. Lorenz. Transcript Verlag, Bielefeld, Germany, pp. 175-184. Lorenz, S. 2012. Socio-ecological consequences of charitable food assistance in the affluent society: the German Tafel. International Journal of Sociology and Social Policy 32(7): 386-400. McIntyre, L., D. Tougas, K. Rondeau and C.L. Mah. 2015. Insights about food banks from a critical interpretive synthesis of the academic literature. Agriculture and Human Values 33(4): 1-17. Ménard, C. 2004. The economics of hybrid organizations. Journal of Institutional and Theoretical Economics 160(3): 345-376.

International Food and Agribusiness Management Review

198

Rombach and Bitsch

Volume 21, Issue 2, 2018

Mohan, S., M. Gopalakrishnan and P.J. Mizzi. 2013. Improving the efficiency of a non-profit supply chain for the food insecure. International Journal of Production Economics 143(2): 248-255. Noy, C. 2008. Sampling knowledge: the hermeneutics of snowball sampling in qualitative research. International Journal of social research methodology 11(4): 327-344. Pache, A. and F. Santos. 2013. Inside the hybrid organization: selective coupling as a response to conflicting institutional logics. Academy of Management Journal 56(4): 972-1001. Park, C. 2008. Do the boundaries between the nonprofit, public and business sectors blur? Comparing ‘within the nonprofit sector collaboration networks’ and ‘inter-sector collaboration networks’ in the social service field in Pittsburgh, Pennsylvania. International Review of Public Administration 13(2): 81-95. Perera, C., P. Auger and J. Klein. 2016. Green consumption practices among young environmentalists: a practice theory perspective. Journal of Business Ethics 138(4): 1-22. Portes, A. 1998. Social capital: its origins and applications in modern sociology. Annual Review of Sociology 24(1): 1-24. Putnam, R. 1995. Bowling alone: America’s declining social capital. Journal of Democracy 6(1): 65-78. Reiniger, W. 2011. Of bread baskets and other forms of food distribution [In German: Von Brotkörben und anderen Lebensmittelausgaben.]. In: Critique of Food Banks in Germany. [Kritik der Tafeln in Deutschland.], edited by S. Selke. VS Verlag für Sozialwissenschaften, Wiesbaden, Germany, pp. 259-264. Sagawa, S. and E. Segal. 2000. Common interest, common good: creating value through business and social sector partnerships. California Management Review 42(2): 105-122. Schneider, J.A. 2009. Organizational social capital and nonprofits. Nonprofit and Voluntary Sector Quarterly 38(4): 643-662. Sedelmeier, T. 2011. Food banks and vulnerability [In Geman: Tafeln und Verwundbarkeit]. In Transformation of Food Banks in Germany [In German: Transformation der Tafeln in Deutschland], edited by S. Selke and K. Maar. VS Verlag für Sozialwissenschaften, Wiesbaden, Germany, pp. 67-80. Selke, S. 2011a. Food banks and society. A sociological analysis of a polymorphic phenomenon [In German: Tafeln und Gesellschaft. Soziologische Analyse eines polymorphen Phänomens]. In: Food Banks in Germany [In German: Tafeln in Deutschland], edited by S. Selke. VS Verlag für Sozialwissenschaften, Wiesbaden, Germany, pp. 15-46. Selke, S. 2011b. The sorrows of others. The role of the food bank in between poverty construction and poverty eradication [In German: Das Leiden der Anderen – Die Rolle der Tafel zwischen Armutskonstruktion und Armutsbekämpfung]. In: Food Banks in Germany [In German: Tafeln in Deutschland], edited by S. Selke. VS Verlag für Sozialwissenschaften, Wiesbaden, Germany, pp. 279-302. Selke, S. 2016. Food banks as a moral organization. Principles and profit of the new poverty economy [In German: Tafeln als moralisches Unternehmen, Prinzipien und Profile der neuen Armutsökonomie]. In: Soup, Consultancy, Politics [In German: Suppe, Beratung, Politik], edited by S. Gilich and R. Keichler. VS Verlag für Sozialwissenschaften, Wiesbaden, Germany, pp. 217-244. Suri, H. 2011. Purposeful sampling in qualitative research synthesis. Qualitative Research Journal 11(2): 63-75. Tarasuk, V. and J.M. Eakin. 2003. Charitable food assistance as symbolic gesture: an ethnographic study of food banks in Ontario. Social Science and Medicine 56(7): 1505-1515. Tarasuk, V. and J.M. Eakin. 2005. Food assistance through ‘surplus’ food: insights from an ethnographic study of food bank work. Agriculture and Human Values 22(2): 177-186. Tuckman, H.P. 1998. Competition, commercialization, and the evolution of nonprofit organizational structures. Journal of Policy Analysis and Management 17(2): 175-194. USDA. 2017. The emergency food assistance program: about TEFAP. Available at: http://tinyurl.com/y75wfeoe. Van der Horst, H., S. Pascucci and W. Bol. 2014. The ‘dark side’ of food banks? Exploring emotional responses of food bank receivers in the Netherlands. British Food Journal 116(9): 1506-1520. Vitiello, D., J.K. Grisso, K.L. Whiteside and R. Fischman. 2015. From commodity surplus to food justice: food banks and local agriculture in the United States. Agriculture and Human Values 32(3): 419-430.

International Food and Agribusiness Management Review

199

Rombach and Bitsch

Volume 21, Issue 2, 2018

Von Normann, K. 2011. Food insecurity and ‘food bank work’ in Germany. Distribution political backgrounds and nonprofit based solution strategies [In German: Ernährungsarmut und ‘Tafelarbeit’ in Deutschland. Distributionspolitische Hintergründe und nonprofit-basierte Lösungsstrategien]. In: Food Banks in Germany [In German: Tafeln in Deutschland.], edited by S. Selke. VS Verlag für Sozialwissenschaften, Wiesbaden, Germany, pp. 91-112. Wachhaus, A. 2013. Governance beyond government. Administration and Society 46(5): 573-593. Werth, S. 2011. It should be different. From routine arises variety [In German: Es geht auch anders nach der Routine kommt die Vielfalt]. In: Food Banks in Germany [In German: Tafeln in Deutschland], edited by S. Selke. VS Verlag für Sozialwissenschaften, Wiesbaden, Germany, pp. 257-264. Witt, R. 2011. Service of the food bank as a premium brand [In German: Die Dienstleistung der Tafeln als Premiummarke]. In: Transformation of Food Banks in Germany [In German: Transformation der Tafeln in Deutschland], edited by S. Selke and K. Maar. VS Verlag für Sozialwissenschaften, Wiesbaden, Germany, pp. 85-102.

International Food and Agribusiness Management Review

200

OPEN ACCESS International Food and Agribusiness Management Review Volume 21 Issue 2, 2018; DOI: 10.22434/IFAMR2017.0062 Received: 11 July 2017 / Accepted: 7 September 2017

Determinants of corporate cash holdings in times of crisis: insights from Brazilian sugarcane industry private firms Special issue: IFAMA 2017 symposium

RESEARCH ARTICLE Aviner Augusto Silva Manoel a, Marcelo Botelho da Costa Moraesb, David Ferreira Lopes Santosc, and Marcos Fava Nevesd aPhD

Candidate in Controllership and Accounting, bProfessor of Financial Accounting at the Department of Accounting, d Professor at the Department of Business Administration, School of Economics, Business Administration and Accounting in Ribeirão Preto (FEA-RP), University of São Paulo (USP), Av. dos Bandeirantes 3900 FEA-RP, ZIP 14040-905, Ribeirão Preto, São Paulo, Brazil cAssistant

Professor, Department of Economy, Administration and Education, São Paulo State University (UNESP), School of Agricultural and Veterinarian Sciences, Jaboticabal. Rod. Prof. Paulo Donato Castellane S/N, ZIP 14.884-900 Jaboticabal, São Paulo, Brazil

Abstract The decision-making process regarding cash allocation, especially in times of recession, is somewhat challenging. Thus, the aim of this study is to evaluate the effects of the financial crisis in cash holdings of privately-held firms in the Brazilian sugarcane industry. For this purpose, a unique, hand-collected panel data of 31 firms for the period from 1998 to 2015 was used. The results show that the private firms of the industry maintain higher levels of cash than the average found by studies with Brazilian public companies. Furthermore, between the two dummy variables used, representing the subprime crisis and the most recent economic crisis in Brazil, only the second one was significant. Finally, in order to overcome the adverse conditions of the current Brazilian economic crisis, it was observed an increase in cash holdings, which, in turn, denotes the relevance of precautionary reasons in the understanding of cash management. Keywords: agribusiness, agri-food, cash management, treasury, financial crisis JEL code: G32, Q13, Q14 Corresponding author: aviner@usp.br

201

Manoel et al.

Volume 21, Issue 2, 2018

1. Introduction The initial point of discussion of this work is the argument of irrelevance introduced by Keynes (1936) regarding the issues of investment and financing of an organization in a perfect market. In this context, when in need of funding, firms would only go to the market to raise funds at an affordable rate. Therefore, there would be no need for them to allocate part of their assets to the cash and cash equivalents account, since they would have access to third party capital at a fair price to finance their investment activities (Keynes, 1936). Nevertheless, the theoretical assumptions proposed by Keynes (1936) would be valid only for organizations with access to perfect capital markets, which is not the case in the Brazilian market. Bearing this in mind, theories begin to emerge in order to understand the behavior of managers in relation to the whole process of the organizations’ cash policies decisions. Consequently, studies related to cash management are recurrent in academic literature, especially due to the fact that companies maintain significant percentage of their assets in cash. These problems are related to the existence of market imperfections, such as informational asymmetry, agency problems, transaction costs and financial constraints (Ferreira and Vilela, 2004; Jensen, 1986; Jensen and Meckling, 1976; MartínezSola et al., 2013). However, most of the efforts on the subject are concentrated on the analysis of large public (Companies listed in stock Exchange) firms (Bigelli and Sánchez-Vidal, 2012; Hall et al., 2014; Steijvers and Niskanen, 2013). These firms, in most cases, have a strong relationship with the capital market, so the demand for cash reserves, for instance, for precautionary reasons are lower than for privately-held firms (Hall et al., 2014). In this sense, Gogineni et al. (2012) point out that the fact that private firms do not have access to the capital market makes it more difficult to obtain sources of financing. Another fundamental difference between them is the ownership structure. While public (listed) companies usually have thousands of shareholders (diffused ownership), privately-held firms have only one or a few shareholders, and they often still belong to a single family, in which the degree of control can be valued by the owners (Bigelli and Sánchez-Vidal, 2012; Gogineni et al., 2012). Another point highlighted by Gogineni et al. (2012) among public and private firms is the level of informational asymmetry between insiders and outsiders. According to the authors, privately-held firms are less transparent to investors, which to a certain extent may aggravate the problems related to adverse selection. Therefore, the results presented so far in the literature, based mostly on samples of public firms, such as in countries considered more developed and in diversified financial environments cannot be generalized to privately-held firms without contextualization, since they differ in several aspects (Gogineni et al., 2012; Tahir and Alifiah, 2015). Dahrouge and Saito (2013) also indicate that the amount to be kept in cash is a controversial issue, but of interest to managers. Thus, the existence of market imperfections can result in potential agency problems, coupled with the fact that managers may have the desire to maintain control while avoiding external financing. Therefore, all the decision-making process related to cash management plays a key role in the firms’ financial policies (Steijvers and Niskanen, 2013). In financial crisis scenarios, where access to credit is lower or even the demand for resources is reduced (Carvalhal and Leal, 2013), there is an increase in interest in the effects of such factors on issues related to the financial liquidity of companies. The current recession of the Brazilian economy, started in 2014, affected by not only economic factors, but also political scandals may have greatly influenced the decision-making process related to cash management.

International Food and Agribusiness Management Review

202

Manoel et al.

Volume 21, Issue 2, 2018

Together with the economic crisis, it is important to bring the context of the sugar cane chain, that suffered a strong crisis from 2008 until the end of 2015, based on the level of debt and the low national and international prices of sugar and ethanol, the major products. Given all the above, the objective of this article was to evaluate the effects of the financial crisis, as well as the crisis of the sugarcane industry, especially influenced by the control of the gasoline price by the Brazilian Federal Government, on the cash holdings of sugarcane plants and distilleries. 1.1 Aimed contributions Studying the effects of recessionary periods under huge uncertainties on cash management is important both for corporate finance issues and organization’s managerial practices. The context of crisis may suggest, for example, that some factors, which used to be imponderable, can provide opportunities for further studies in order to understand the management practices. Consequently, this whole context offers a valuable opportunity to analyze its effects on cash holdings. Besides the fact that most of the efforts on this subject focus on public (listed) companies, Tahir and Alifiah (2015) also mention the direction of these studies only at company level. Thus, the authors point out that it is increasingly difficult to ignore the importance of each sector, as well as their specificities in studies on the topic since they can decisively influence the percentages kept in cash by each sector. Therefore, there is an opportunity to contribute with current literature in understanding the business environment in each sector, especially in the case concerning cash management due to each sector´s characteristics (specificities), which can affect their cash management. Consequently, the results obtained in research in specific sectors can help define the customized cash policies for each sector (Tahir and Alifiah, 2015). Given the theoretical assumptions listed, and the way in which privately-held firms may be affected differently from listed companies by the aforementioned factors in their cash holdings, this study initially aimed to contribute to the literature on cash management providing evidence on a market which is still obscure in Brazil: the study of privately-held firms. The importance of private firms is also a relevant factor, given that the number of non-listed companies is much higher, playing an indispensable role in any economy and still employing a higher proportion of the workforce (Gogineni et al., 2012). In addition, the Brazilian market, characterized by Dahrouge and Saito (2013), may further affect the cash management of Brazilian companies, especially those who do not have access to the capital market, due to the high fundraising, higher and inflation rates and costs of capital. In situations of financial constraints, for example, organizations can allocate a higher percentage of their assets to cash in order to anticipate any adverse situation, as well as to continue with their investment policies. This occurs because financial constraints often make it impossible for Brazilian organizations to raise funds at a rate compatible with their investment projects (Ferreira and Leal, 2011). Evidence shows that when the 2008 economic crisis hit the sugarcane business in Brazil, investments decreased dramatically. The crisis environment can also influence the organization’s cash management, since there are fewer investment opportunities and higher market uncertainties (Baum et al., 2006; Nason and Patel, 2016). Despite unfortunate, the crisis scenario allows to deepen the studies on the subject and to shed light on the behavior of organizations and their effects on cash management (Campello et al., 2011; Van Der Stede, 2011). Manoel et al. (2016) mention a peculiar feature of the industry, in addition to the effects caused by the crisis, which enables further exploration of its effects on cash management: the high ownership concentration. In this sense, Steijvers and Niskanen (2013) provide evidence that the ownership structure actually moderates the relationship between the company’s management and its cash holdings. International Food and Agribusiness Management Review

203

Manoel et al.

Volume 21, Issue 2, 2018

Moreover, according to Neves et al. (2014) the Brazilian sugarcane industry demonstrates its strength for the Brazilian economy by producing various forms of sustainable and renewable energy, such as sugar, ethanol and electricity. Consequently, the sector is capable of meeting today’s global demand for cleaner and greener energy without compromising the environment and the lives of future generations. This industry is very sensitive in financing issues, since cash flows are dramatic. Agricultural activities in this industry have a huge initial investment and the first harvest will be taken almost 2 years after planting, and industrial investments due to licenses, complexity and other events may take three years to start running after initial investments were performed. An industry with a crushing capacity of 3 million tons of sugar cane per year will demand an initial investment of around US$ 400 million and cash generation will start only at least two years after the investment (capital expenditure). Finally, following the evidence obtained by Ozkan and Ozkan (2004), this research also used the Generalized Method of Moments (GMM) for dynamic panel data in order to consider the problems of endogeneity in the empirical analysis of the company’s cash management. For the authors, the issue of endogeneity is a relevant factor in studies on the topic since shocks that affect cash reserves may also affect some of the regressors.

2. Hypothesis development In recent years, the sugarcane industry has been facing one of its worst crises, as mentioned by Neves et al. (2014) and Santos et al. (2015). This crisis was caused not only by climatic factors, such as the absence of rainfall, especially in 2014 and early 2015, but also the control of gasoline prices by the Brazilian federal government. The policy of controlling the price of gasoline, according to Sant’Anna et al. (2016), occurred in order to control inflation. This policy, however, has resulted in negative consequences for the industry. For Sant’Anna et al. (2016), this meant that the price of ethanol needed to be kept low while the production costs were rising 11.5% a year. Moreover, Neves et al. (2014) mention the increase in the indebtedness of the mills and distilleries of the sector, the reduction of the investment opportunities and the narrow or even negative margins of the current investments. In this respect, 62 economic groups either closed down or filed for bankruptcy protection just in the period from January to October, 2014, considerably increasing the number of unemployed workers, especially in the cities of Sertãozinho and Piracicaba, in the State of São Paulo, characterized by its longtime tradition in the sugarcane agro-industry equipment manufacturing. Consequently, it was expected that the industry recession scenario also affected the firms’ cash reserves, since the literature on the theme indicates that in such situations the precautionary reasons are greater, making, therefore, that companies keep higher cash holdings in order to avoid adverse market conditions. Furthermore, Carvalhal and Leal (2013) point out that, during the subprime crisis (2008), the emerging markets, in which Brazil is included, were also affected by lack of credit, confidence and decline in economic activity. Thus, the aim of this study was to evaluate the effects of the financial crisis, as well as the crisis of the sector, influenced especially by the gasoline price control by the Brazilian federal government and lower international sugar prices, on cash holdings of mills and distilleries of sugarcane industry. Given all the above, the research hypothesis is: H1: there is a positive association between cash holdings in times of economic recession and sector (industry) crisis.

International Food and Agribusiness Management Review

204

Manoel et al.

Volume 21, Issue 2, 2018

3. Research methodology In this topic the methodology that conducts all this research is described, as well as the sample used, the proposed model and the operational definition of each variable. This research used the annual financial statements through a unique, hand-collected panel data set of 31 privately-held firms of the Brazilian sugarcane industry. The data used are nominal and were hand-collected from the websites of the organizations (for the ones that published) and their publications in the Official Journal of the Brazilian Government. The period analyzed corresponds to the interval between the years of 1998 and 2015, according to the availability of the data of each firm throughout the period. This time interval includes the new technological and market paradigm of the sector that involves the production of sugar, ethanol and energy. In addition, there was a complete deregulation of the sector in 1996/1997, which makes companies less protected to the market fluctuations. Therefore the decision was to start the sampling period from the year 1998. Finally, the use of an 18-year panel allowed to analyze the evolution of the cash levels during that time, as well as to observe their variations and the possible impact factors. At first the regression model with panel data through a balanced panel was thought as the best method. However, due to the difficulty in accessing information of some variables throughout the period, an unbalanced panel was chosen so that the number of observation would not be reduced. The panel data regression model, composed of a cross section in a time series, allows, according to Wooldridge (2013), to study phenomena that can undergo changes over time and the individual analyzed. Moreover, this model also allows analyzing a greater amount of information, as well as a greater number of degrees of freedom and, consequently, generating better estimation efficiency of the parameters. Finally Wooldridge (2013) points out that the regression model with panel data allows a greater variability of the data, which in a certain way can contribute to reduce the problems of multicollinearity among the variables used. Thus, due to the use of a sample composed of 31 companies over 18 years, initially 558 observations were obtained. Given the unavailability of some data for the variables used in the model, 102 observations were removed from the sample. Those with negative Equity were removed, following Bigelli and Sรกnchez-Vidal (2012) recommendations. Therefore, a final sample of 434 observations for the same 31 companies was used, according to Table 1. The GMM model, which uses a time lag fitted model for the dependent variable, excluded 76 observations; a final sample with 358 observations was available. The variables used in the model are presented according to a previous suggestion given by the reviewed the literature. Table 1. Research sample.1 1998-2015

Sample

Observations

Total (-) Companies with missing values for the variables used (-) Companies with Negative Equity (=) Final sample used to GMM model (-) Time lag (GMM) (=) Observations in regression

558 (102) (22) 434 (76) 358

GMM = Generalized Method of Moments.

International Food and Agribusiness Management Review

205

Manoel et al.

Volume 21, Issue 2, 2018

3.1 Dependent variable The dependent variable used in this work was the same one suggested and used by Dittmar, Mahrt-Smith and Servaes (2003) and Opler et al. (1999). Obtained by the natural logarithm of the value resulting from the sum of the cash and equivalents divided by the total net assets (total assets minus cash and equivalents of cash). 3.2 Independent variables ■■ Subprime and Brazilian market crisis Precautionary reasons cause organizations to reserve part of their assets in cash to avoid any adverse situation. Thus, since financial crises represent a trivial factor in the financial decisions of companies and lead firms to a credit crunch in Brazil, according to Carvalhal and Leal (2013), it is expected, therefore, that in such context, companies maintain higher levels of cash. In addition to the problems related to climate factors and the entire context of the recession of the Brazilian economy, the sugarcane industry has faced one of its worst crises in recent years due to the control of gasoline prices by the Brazilian federal government, according to Santos et al. (2015) and Sant’Anna et al. (2016) and lower international sugar prices, mainly caused by Brazil, that due to the low prices of gasoline and the difficulties for ethanol to compete, converted more cane to sugar, flooding the sugar market. Baum et al. (2006) also point out that in times of increased uncertainty about the macroeconomic environment, organizations will increase their cash holdings to offset the negative effects of the economic downturn on their cash flows. Consequently, under a low credit access scenario, the marginal benefit of cash reserves may be higher, especially in financially restricted firms (Bliss et al., 2015). Therefore, a positive association between financial crisis contexts and companies’ cash reserves is expected. In order to capture the effect of the financial crisis, two dummy variables were used, with value 1 for the years of 2008 and 2009 representing the subprime crisis, and 0 otherwise. The other one represents the crisis of the sugarcane industry and the Brazilian market, with the value 1 for the years 2014 and 2015, and 0 otherwise. 3.3 Control variables ■■ Liquidity Al-Najjar (2013) argues that the cost to convert more liquid assets into cash is lower than other assets. Consequently, companies with a larger number of liquid assets do not have to maintain large cash reserves, since they can, for example, convert these funds into cash when they are needed. Thus, following Al-Najjar (2013) and Ferreira and Vilela (2004), it is expected a negative association between liquidity and cash reserves. The variable used as a proxy for liquidity was the Current Liquidity, obtained through the division of Current Assets by Current Liabilities. ■■ Size In relation to the Size, obtained by the natural logarithm of the Net Assets, a negative association with the cash levels is expected. This expected result is in accordance to the study by Kim et al. (1998), Hall et al. (2014) and Steijvers and Niskanen (2013), who argue that larger organizations maintain smaller cash reserves. This fact is consistent with the argument that smaller firms face higher costs to raise funds, and in order to avoid significant transaction costs, they maintain higher cash holdings. Moreover, these companies have a

International Food and Agribusiness Management Review

206

Manoel et al.

Volume 21, Issue 2, 2018

higher probability of financial difficulty due to the higher degree of informational asymmetry (Ferreira and Vilela, 2004). ■■ Profitability For Al-Najjar and Clark (2017), the most profitable institutions are expected to maintain higher cash reserves since, according to the pecking order theory, they are more likely to invest in value-for-money investment projects Net Present Value and can therefore use the resources generated in order to maintain their investment activities. In addition, Hall et al. (2014) mention that the most profitable companies probably have easier access to external financing with a lower cost of funding. Therefore, these companies tend to hold higher cash reserves, to take advantage of these benefits. In this research as return variable, the Return on Equity (ROE), obtained by the division of Net Income by Equity is used. ■■ Cash generation At first, the use of variables related to Cash Flow is considered. However, Brazilian firms were not required to present the Cash Flow Statement prior to 2008. Therefore, the use of a sample period prior to this date made it impossible to use data derived from this financial statement (Brazil, 2007). Thus, two financial indicators were chosen to be used: the Asset Turnover and the Operating Margin, respectively obtained by dividing the Total Revenue by Net Assets and Operating Income by Total Revenue. The basis for the use of the two variables follows the positioning of the Dupont Model, in which both are used in the financial analysis as indicators of generation and income retention (Assaf Neto, 2014; Ross et al., 2013). ■■ Debt The Debt variable was also used in this study, obtained by dividing Liabilities by the sum of Liabilities and Equity. Bigelli and Sánchez-Vidal (2012) still mention that the debt of organizations, provided by bank loans, should lower the financial constraints of privately-held firms in order to serve as a cash substitute. Therefore, a negative association between the level of indebtedness and the cash balance was expected. ■■ Leverage Ozkan and Ozkan (2004) argue, in their study, that highly leveraged organizations are more likely to incur financial difficulties. Han and Qiu (2007) complement Ozkan and Ozkan (2004) by mentioning that more leveraged firms may need to accumulate a larger percentage of their assets in cash in order to pay off their debt in the future. Thus, both authors suggest a positive relationship between leverage and cash holdings. The variable used as proxy for Leverage was obtained by dividing Total Debt by Net Asset, following the operational definition suggested by Assaf Neto (2014). ■■ Age Al-amri et al. (2015) argue that organizations substantially reduce their cash levels throughout their life cycle. The results obtained by the authors corroborate this fact, signaling that mature companies maintain lower cash holdings, since they manage their cash flows better. Thus, in order to capture the effect of the life cycle on cash management, the natural logarithm of age was used. International Food and Agribusiness Management Review

207

Manoel et al.

Volume 21, Issue 2, 2018

■■ International Financial Reporting Standards Another topic that has been widely discussed in the Financial and Accounting literature is the adoption of the International Financial Reporting Standards (IFRS). The purpose of establishing a set of international standards was to not only improve the comparability, quality and reliability of information provided to users, but also reduce the cost of preparing the financial statements and provide more timely and accounting relevance (Antunes et al., 2012; Armstrong et al., 2010). The Brazilian convergence to International Standards, unlike what occurred in most European countries with mandatory adoption in 2005, was held in two stages. The first one started in the beginning of 2008, through Law 11.638/2007, when partial adoption took place. The mandatory adoption, in turn, was in 2010 with the adoption of the full set of IFRS standards for the Consolidated Financial Statements, following the standards issued by the Accounting Pronouncements Committee in Brazil (Antunes et al., 2012). Due to the aforementioned peculiarities, as well as the discretion in the adoption of IFRS by private firms, a dummy variable was adopted to control the effects of the adoption of the standards, in which the value 1 was assigned from the individual declaration of each company in the preparation of its financial statements, and 0 before this period. Finally, it is important to mention that among the sample used, the year of adoption varied from 2008 to 2011. ■■ Net working capital The Current Assets can be easily converted into cash and, thus, are considered as cash substitutes. Therefore, the Net Working Capital (NWC) was also used as a control variable, as suggested by Opler et al. (1999) and Bigelli and Sánchez-Vidal (2012). A negative association between NWC and cash reserves was expected. The variable used to control the effects of cash substitutes was obtained, according to Opler et al. (1999) and Gao et al. (2013), through the NWC, in which Current Assets were discounted from Current Liabilities and from Cash and Equivalents, and the resulting value was, then, divided by Net Assets. 3.4 Econometric model Following Ozkan and Ozkan (2004), the GMM estimate for dynamic panel data was used to address the issue of endogeneity. The theoretical assumptions for using this model are that it implicitly assumes that firms can instantly adjust their target cash levels after changes in the firm’s specific characteristics or due to random shocks. Thus, a model that assumes that organizations seek a cash target level was used. Therefore, the percentages of the assets held in cash in a period will also be explained by their decisions taken in previous years. According to García-Teruel and Martínez-Solano (2008), dynamic panel regressions are characterized by the existence of autocorrelation due to the consideration of the lagged dependent variable as an explanatory variable. The consistency of the estimates of the GMM is subject to an ideal choice of instrument, in which its validity depends on the absence of serial correlation of higher order in the idiosyncratic component of the error term. Like Ozkan and Ozkan (2004), García-Teruel and Martínez-Solano (2008), the consistency of the estimates for the absence of second-order serial correlation was tested as proposed by Arellano and Bond (1991). In addition to this test, the Sargan Test of excessive identification restrictions, which tests the lack of correlation between the instruments and the residues was used.

International Food and Agribusiness Management Review

208

Manoel et al.

Volume 21, Issue 2, 2018

Therefore, Equation 1 with the proposed econometric model follows: CASHit = β0 + β1CASHi,t–1 + β2Subprime Crisisi,t + yβ3 Brazilian Crisisi,t + β4 CLi,t + β5 Sizei,t + β6 ROEi,t + β7 ATi,t + β8 OMi,t + β9 Debti,t + β10 Levi,t + β11 Agei,t + β12 IFRSi,t + β13 NWCi,t + ui,t (1) In which: ■■ Cash = Cash and Equivalents, obtained by the natural logarithm of the sum of the cash and cash equivalents, divided by the Net Assets (total assets minus cash and cash equivalents). ■■ Subprime Crisis = Subprime Crisis, dummy variable that assumes the value 1 for the years 2008 and 2009 and 0 for the remaining sample period. ■■ Brazilian Crisis = Brazilian Crisis, dummy variable that assumes the value 1 for the years 2014 and 2015 and 0 for the remaining sample period. ■■ CL = Current Liquidity, obtained by the division of Current Assets by Current Liabilities. ■■ Size = Size, obtained by the natural logarithm of Net Assets. ■■ ROE = Return on Equity, obtained by the division of Net Income by Equity. ■■ AT = Asset Turnover, obtained by dividing Total Revenue by Net Assets. ■■ OM = Operating Margin, obtained by the division of Operating Income by Total Revenue. ■■ Debt = Debt, obtained by the division of Total Liabilities by the sum of Total Liabilities and Equity. ■■ Lev = Leverage, obtained by the sum of Total Short-Term Debts and Total Long-Term Debt, and the resulting value was divided by Net Assets. ■■ Age = Age, obtained by the natural logarithm of the age. ■■ IFRS = IFRS adoption dummy, in which the value 1 was attributed from the individual declaration of each firm from the adoption of IFRS in the preparation of its financial statements, varying between years 2008 and 2011, and 0 previously. ■■ NWC = Net Working Capital, obtained by subtracting Current Assets by Current Liabilities, and the resulted value divided by Net Assets. ■■ u = Error term.

4. Results Table 2 shows the Descriptive Statistics of the research sample. All variables were winsorized at 5% level. Table 2. Descriptive statistics of variables. Variable

Mean

1st quartile

Median

3rd quartile

Std. dev.

Cash Subprime Crisis Brazilian Crisis Current Liquidity Size ROE Asset Turnover Operating Margin Debt Leverage Age IFRS NWC

10.38 0.106 0.092 1.786 12.515 0.377 0.760 0.141 0.644 0.068 3.675 0.369 -0.014

1.614 0.000 0.000 0.774 11.672 -0.025 0.486 0.005 0.522 0.030 3.466 0.000 -0.136

5.592 0.000 0.000 1.253 12.580 0.084 0.663 0.096 0.652 0.055 3.807 0.000 -0.021

13.015 0.000 0.000 2.046 13.325 0.191 0.908 0.171 0.750 0.091 4.094 1.000 0.080

0.132 0.308 0.290 2.263 1.119 8.830 0.498 0.850 0.156 0.059 0.660 0.483 0.193

International Food and Agribusiness Management Review

209

Manoel et al.

Volume 21, Issue 2, 2018

In sequence, Figure 1 shows the change in cash levels throughout the period. In addition to the numbers indicated in this Figure, it was found that, on average, organizations of the sugarcane industry maintain 10.38% of their Total Assets allocated in Cash and Equivalents. The value found, therefore, is higher than the average obtained in the works with Brazilian listed companies, such as shown by Dahroug and Saito (2013), which obtained an average of 6.5%. The higher average for the sample of privately-held firms in this study was as expected, given that private companies may face greater barriers to obtain external funding than those listed on stock exchanges. Thus, due to precautionary reasons, the mills and distilleries maintain higher cash levels. Moreover, the achievement of a higher level of cash reserves is in line with the results obtained by Lee and Powell (2011) because, according to the authors, capital-intensive firms, as observed in the studied sector, maintain higher cash holdings. Hall et al. (2014) still attribute the fact that listed companies maintain lower cash levels given their greater access to the capital market in relation to privately-held firms. Therefore, the results obtained by the authors â&#x20AC;&#x201C; that private firms maintain higher levels of cash than those listed â&#x20AC;&#x201C; were also obtained in this study. The mean value found here, however, is misaligned to that obtained by Gao et al. (2013) for their sample, since the authors verified that private firms in the US market hold 17.17% of their assets in cash; representing, therefore, a percentage 50% higher than public firms (9.39%). The authors attributed this difference to the major agency problems of listed companies, as well as to the more exacerbated financial constraint of the group of privately-held firms. Thus, the specific characteristics of each market can explain this difference. These include, for example, the existence of greater investment opportunities for privately-held firms in the US market, the countryâ&#x20AC;&#x2122;s own institutional characteristics, and the existence of mechanisms to control the managers opportunistic actions in relation to the problems of agency resulting from the maintenance of higher cash holdings. Moreover, the results found by Kusnadi and Wei (2011) corroborate this fact, since the authors observed that there is a great variation of the cash levels between companies of developed and developing countries. After discussion of the descriptive statistics, it was verified, at first, whether the variables of the model presented potential problems of collinearity. When observing Table 3, it was verified that there are no high correlations between the variables used, except for the variables Net Current Capital and Current Liquidity. In addition, Table 4 denotes the Variance Inflation Factor. Through this test, it was verified that no value was above 10, including the two aforementioned variables, which could denote, according to Wooldridge (2013), multicollinearity problems. Consequently, multicollinearity was not a problem in this model. Average cash level (%) 13.90

11.73

12.13 8.60

14.04

13.54

8.77

12.79

9.78

9.55

12.15 8.22

8.19

7.06

8.34

7.98

10.49

8.30

1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015

Figure 1. Variation of cash holdings.

International Food and Agribusiness Management Review

210

Manoel et al.

Volume 21, Issue 2, 2018

The normality of the variables was verified. The results indicated, through the Shapiro-Wilk test, the nonnormality of all variables. The normality of the residues did not present normal distribution either. Considering all these points, Table 5 shows the results obtained by the GMM with standard asymptotic errors, given the problem of heteroscedasticity of the sample. The results of the Sargan test indicate that the instrument variable (Cash t-1) used in the estimation model has no correlation with the error term. The first result of the model confirms the dynamic behavior of the cash decisions, since the coefficient of the lagged variable of cash is positive and significant at 1%. Consequently, the decision of the cash levels of a certain period is influenced by the percentages of cash held in previous dates. The results are consistent with those obtained by Ozkan and Ozkan (2004) and GarcĂa-Teruel and MartĂnez-Solano (2008) indicating that the organizations of the industry, in general, have a target level of cash to be maintained. Table 3. Correlation matrix of variables. 1 Ln Cash 2 Subprime Crisis 3 Brazilian Crisis 4 CL 5 Size 6 ROE 7 AT 8 OM 9 Debt 10 Lev 11 Age 12 IFRS 13 NWC

1 -0.06 0.05 0.54 -0.12 0.30 0.48 0.09 -0.32 0.05 -0.12 -0.06 0.28

1.00 -0.11 -0.09 0.18 -0.06 -0.12 -0.03 0.13 0.05 -0.02 0.16 -0.08

1.00 -0.06 0.34 -0.04 -0.13 0.07 0.05 0.08 0.16 0.42 -0.14

1.00 -0.21 1.00 0.23 0.04 1.00 0.38 -0.59 0.29 1.00 0.18 0.27 0.56 -0.07 1.00 -0.47 0.11 -0.39 -0.30 -0.27 1.00 -0.08 -0.17 -0.25 0.12 -0.16 0.16 1.00 -0.06 0.14 0.01 -0.06 0.10 -0.05 -0.05 1.00 -0.13 0.66 -0.03 -0.31 0.17 0.08 0.05 0.17 1.00 0.72 -0.21 0.25 0.32 0.09 -0.45 -0.31 -0.12 -0.29

Table 4. Variance Inflation Factor (VIF).1 Variables

VIF

Subprime Crisis Brazilian Crisis Current Liquidity Size ROE Asset Turnover Operating Margin Debt Leverage Age IFRS NWC

1.113 1.313 2.578 2.970 1.979 2.146 1.738 1.517 1.352 1.071 2.172 2.807

Values above 10 indicate potential multicollinearity problems.

International Food and Agribusiness Management Review

211

1.00

Manoel et al.

Volume 21, Issue 2, 2018

Table 5. Generalized Method of Moments cash holdings regression.1 Variables

Coefficients

Constant Cash t-1 (instrumental var.) Subprime Crisis Brazilian Crisis Current Liquidity Size ROE Asset Turnover Operating Margin Debt Leverage Age IFRS NWC Observations (n) Corr (y,y) Sargan test (df)

-0.048 0.047 0.220 0.056 -0.235 0.155 0.406 0.187 0.711 0.094 -0.118 0.242 0.868 0.279 0.903 0.296 -0.273 0.391 0.118 0.730 0.267 1.407 1.618 1.362 0.043 0.205 -2.642 0.665 358 0.6346 167.058 (135) with P-value (0.0318)

Standard error

P-value

-1.037 3.919 -1.513 2.173 7.566 -0.487 3.110 3.049 -0.698 0.161 0.190 1.188 0.210 -3.975

0.300 8.91e-05*** 0.130 0.029** 3.84e-014*** 0.626 0.001*** 0.002*** 0.485 0.872 0.850 0.235 0.833 7.02e-05***

= statistically significant at 10%; ** = statistically significant at 5%; *** = statistically significant at 1%.

Concerning the dummy variable representing the Subprime Crisis, there is a negative association, without obtaining statistical significance. Therefore, the coefficient, contrary to the one suggested by the literature, indicates that the effects of this crisis do not seem to have a decisive impact on the cash levels maintained by the mills and distilleries as much as the current crisis of the sector, which Santos et al. (2015) characterize as one of the worst which ever took place in Brazil. Thus, the effects of the subprime crisis, ceteris paribus, seems to be lower in Brazil, at least in relation to cash management. In this sense, corroborating the arguments of Santos et al. (2015), it was possible to observe a positive association at the level of 5% for the dummy variable representing the current Brazilian and Industry Crisis. Factors such as the absence of rainfall and the control of the price of gasoline by the federal government to control inflation seem to have influenced the cash management of the companies in the industry. For Santâ&#x20AC;&#x2122;Anna et al. (2016), gasoline price controls adversely affected the sugarcane industry, as gasoline prices were artificially kept below international prices, and ethanol prices, on the other hand, were determined by market. As a result, gasoline has become more efficient than ethanol, since ethanol presents a higher consumption per mile than gasoline, causing the price of ethanol to become too expensive. According to Santâ&#x20AC;&#x2122;Anna et al. (2016), all the context mentioned above was responsible for the crisis in the sector, so that the results obtained here still indicate that the crisis in the sector positively affected, ceteris paribus, the cash levels maintained by the companies analyzed. A possibility to obtain a positive coefficient would be due to an apparent reduction of the credit supply during the years 2014 and 2015. For Bliss et al. (2015), this occurs because, in financial crisis situations, there may be a strong reduction in the supply of credit, or even an increase in the cost of funding. However, Bliss et al. (2015) emphasize, in their work, that periods of crisis are not only associated with credit shortage. Therefore, there may also be a shock in the demand for financing due to the whole context of market uncertainties.

International Food and Agribusiness Management Review

212

Manoel et al.

Volume 21, Issue 2, 2018

As a result, the whole environment of uncertainty and crisis in Brazil caused managers to increase the average level of cash, aiming to overcome the problems caused by the credit crunch and reduced investment opportunities. Consequently, the precautionary reason proved to be of great value in understanding the cash management of privately-held firms of the Brazilian sugarcane industry, which in turn, was also obtained by Trejo-Pech et al. (2015). In their study, Trejo-Pech et al. (2015), made use of a sample of agribusiness companies listed on the US stock market during the period from 1970 to 2012. The results obtained by them showed that precautionary reasons are essential for the management of the cash of agribusiness companies, since cash is needed to quickly execute their growth opportunities, as well as limit the transaction costs related to obtaining resources from third parties. Despite not only the positive and significant association with this variable, but also the average cash level higher than that found by Dahroug and Saito (2013) for the Brazilian listed companies, the maintenance of an average cash level of 10.38% as a concern was mentioned. This is due to the fact that the cases of companies that went bankrupt or went into bankruptcy process are high in the sector, and the maintenance of lower cash balances compared, for example, to that found by Gao et al. (2013) of 17.17% for their sample of privately-held firms, may be contributing decisively to the crisis in the sector. Thus, given the maintenance of lower percentages of the assets in cash, the organizations may face greater difficulties in obtaining resources from third parties. Moreover, companies with low cash levels during times of crisis may be more susceptible to failure. In this context, the market may interpret lower levels as precursors of bankruptcy, and maintaining larger cash holdings may signal the availability of resources needed to pursue the organization’s investment activities. Consequently, companies could maintain higher cash holdings so that, even in periods considered ‘normal’, they could take advantage of possible investment opportunities that, without the maintenance of higher reserves, would be impossible, due to the difficulty of access to resources at an affordable cost that did not prevent the realization of their projects. Regarding smaller and younger companies, factors that previous studies indicate that would play a relevant role in determining cash levels, especially for privately-held firms, had not significant results. It was expected that both Size and Age were negatively associated with cash, however, only the first one got a negative, but nonetheless, insignificant association. Specifically for Size, Trejo-Pech et al. (2015), in a study with agribusiness firms, obtained the same negative association, with statistical significance, though. As noted by the regression results, there is a substitution effect between NWC and the dependent variable at the 1% level. This negative association had already been expected because of the assumptions of Opler et al. (1999), Ozkan and Ozkan (2004), which signal the lower cost of converting non-cash net assets into cash in relation to other assets. Therefore, the presence of more liquid assets allows companies to maintain lower percentages of their assets in cash. The Leverage, according to Steijvers and Niskanen (2013) can also act as a substitute for cash reserves. Thus, the authors suggest a possible negative relationship between leverage and cash, counteracting the positive association suggested by Han and Qiu (2007) and Ozkan and Ozkan (2004). However, the results signaled a positive association and yet without statistical significance. Like Leverage, the Debt variable, according to Bigelli and Sánchez-Vidal (2012) could also serve as a substitute for cash. However, it also presented a positive association without the achievement of significance, contradicting, therefore, the assumptions signaled by the authors.

International Food and Agribusiness Management Review

213

Manoel et al.

Volume 21, Issue 2, 2018

The profitability variable used (ROE), on the other hand, confirms the theoretical assumptions of Al-Najjar and Clark (2017) and Hall et al. (2014), in which the most profitable companies are expected to maintain larger cash balances. Given that, it was obtained a positive association between the ROE and cash levels at the level of 1%. Furthermore, the positive and significant influences of the Asset Turnover (AT), Return on Equity (ROE) and Current Liquidity (CL), at 1%, demonstrate the importance of financial management for companies in this sector, whose results may suggest an integrated analysis of these dimensions. It was noted that the greater availability of resources in the millâ&#x20AC;&#x2122;s cash flow depends on the ability of these companies to generate income from their assets (AT) and to manage resources and leverage on a competitive basis to shareholders (ROE). Moreover, the positive and significant effect of liquidity also includes the importance of financial freedom in managing working capital and demonstrates that there is no tradeoff between liquidity and profitability for the sector, bearing in mind that the greater availability of cash is a function of higher levels of profitability and liquidity. Nevertheless, it is possible to associate to this analysis the negative and significant impact of the indebtedness in the cash, which shows that companies with higher cash levels use this internal resource primarily to use third-party capital to meet capital demands, which brings the behavior of these companies closer to Pecking Order Theory. Finally, Operating Margin and IFRS were not significant determinants in the cash holdings. In order to make the results, which were obtained through regression of the model of the GMM, more robust, the database was also analyzed by means of the Panel Data Regression Model. This procedure was carried out due to the fact that one criticism of the GMM (Johnston and Dinardo, 1997) is that it removes part of the sample used which, in turn, is already limited in this study. Therefore, the non-significance obtained for some of the model variables, especially for the variable representing the subprime crisis, which had a P-value of 0.13, could be due to the limited sample size used. Thus, it was carried out the following test model specification: Welch test (4.5e-028), Breusch-Pagan (3.1e-063) and Hausman (3.0e-019) so that their results indicated the best fit of the Fixed Effects model to the database used. Considering all these points, the Fixed Effect Regression Model with Robust Standard Errors (due to the heteroscedasticity problem) was also used. The results obtained, as well as the signs, in turn, were similar. The only items that diverged were the significance levels of the ROE and Asset Turnover, which were previously significant at the 1% level, and were now at 5% and 10% respectively. In specifically addressing the variable representing the subprime crisis, it continued to be non-significant and with the same coefficient, so that the new p value obtained was 0.49. Consequently, the results reached in this model corroborate the previous ones. Finally, another factor that could influence the results of this study is inflation. Curtis et al. (2017) demonstrated that inflation is an important determinant of the companyâ&#x20AC;&#x2122;s cash holdings. The Brazilian economy is denoted by high inflationary levels (Dahrouge and Saito, 2013), but the results obtained, through the National Extended Consumer Price Index, by including this variable as a control, was significant in neither model. Therefore, it was decided not to consider inflation in the model.

5. Conclusions Many of the efforts made on the cash literature so far are focused on public companies, and only at company level. Thus, the specificities of privately-held firms (like ownership structure, level of information asymmetry and access to sources of financing), as well as the peculiar characteristics of each sector, Manoel et al. (2016) mention, in particular, the high shareholder concentration of the Brazilian sugarcane industry. It means that the studies carried out so far cannot be generalized to other companies without proper contextualization. The decision-making process in relation to the total levels of assets held in cash by firms is challenging, especially in times of crisis. Thus, this study analyzes the effects of the Subprime and the Brazilian market International Food and Agribusiness Management Review

214

Manoel et al.

Volume 21, Issue 2, 2018

crises on the cash levels of the sugarcane industry, through a unique, hand-collected sample of privatelyheld firms. Studies that analyze the cash management of firms in this sector may be decisive for understanding such an important item of financial management. Furthermore, the lack of an appropriate cash management may have contributed incisively, ceteris paribus, in the cases of bankruptcy and bankruptcy processes of these firms in Brazil. This research can contribute for decision makers to adopt more specific management policies, according to their peculiarities, especially when forecasting periods of economic recession, aiming to emerge stronger in troubled times like the ones analyzed. Likewise, the mills and distilleries that are able to strengthen their cash holdings at a time prior to recessions may take advantage of future investment opportunities, which in turn, would not be possible in cases of lack of internal resources in a timely manner. This fact takes place because of the characteristics of the Brazilian market â&#x20AC;&#x201C; see Lozano and Caltabiano (2014), such as low access to credit and dependence on the banking system with one of the highest real interest rates in the world, meaning that resources obtained from financing are not considered as ideal substitutes compared to those generated internally. Companies that survive the crisis periods may also suffer in subsequent years since the funds inflow of in cash can take a long time, depending on the financial receipt cycles. As a result, the delay in rebuilding cash reserves at pre-recession levels may cause firms to increase their debt levels, especially short-term debt at high interest rates. Consequently, it further compromises the future performance of the industry, as well as worsens the quality of the debt structure. Aiming to make the results obtained more robust, this research also controlled for endogeneity problems, because of its possible existence in studies on cash management. For better understanding, see Ozkan and Ozkan (2004). Thus, the empirical results obtained in this study corroborate the previous studies regarding the fact that private firms maintain higher percentages of their assets in cash than public ones. This result can be attributed to easier access by listed companies to sources of financing, even in a context of financial constraint. Thus, as a precaution, the companies of the industry maintain larger cash holdings. In relation to the two dummies representing the Subprime and Brazilian Crises, only the second one was statistically significant with a positive coefficient. The first one, on the other hand, was not significant and did not show the expected positive sign. Evidence obtained also suggests that companies with higher NWC have lower cash levels, given their greater convertibility in cash. Finally, the Current Liquidity, Asset Turnover and ROE variables were also significant in the understanding of the cash management, in which all three had a positive association at the 1% level. Regarding the limitations of this study, in addition to the restricted number of the sample, it was also mentioned the quality of the financial statements, especially for the period prior to the adoption of IFRS, since, not infrequently, the disclosed information complied with current legislation, but without providing additional information. Therefore, the non-significance results obtained by some of the control variables of the model, as well as the P-value (0.13) of the variable representing the subprime crisis may be due to these limitations. However, the results obtained using the Model of Fixed Effect Regression with Robust Standard Errors were also similar to those obtained through GMM. Consequently, one of the limitations of GMM, specifically regarding the reduction of the already limited sample number, was not significant in the results presented previously. Furthermore, on account of the difficulty of accessing the explanatory notes, the research was unable to include more variables in order to expand the understanding on the theme. Thus, the inclusion of variables on a possible relationship of the company with a single bank, the number of board members, mechanisms of corporate governance, the existence of an external chief executive officer, as well as proxies related to the investment opportunities could not be performed either. International Food and Agribusiness Management Review

215

Manoel et al.

Volume 21, Issue 2, 2018

Despite all these limitations, this research is still relevant, since the results obtained here in a single context of the Brazilian economy can serve as a basis for future research. Thus, expanding the scope of this study to public companies of the sugarcane industry in Brazil is a natural way. Consequently, with the development of research in this scope, new studies can compare their results with those obtained here, in order to expand the understanding on cash holdings. Studies comparing cash management in companies from the world’s leading sugarcane producers such as China, India, Australia and Thailand can also be of great value to the cash management literature. The research also offers some insights for financial managers, being this an important managerial implication.

References Al-Amri, K., M. Al-Busaidi and S. Akguc. 2015. Conservatism and corporate cash holdings: a risk prospective. Investment Management and Financial Innovations 12(1): 101-113. Al-Najjar, B. 2013. The financial determinants of corporate cash holdings: evidence from some emerging markets. International Business Review 22(1): 77-88. Al-Najjar, B. and E. Clark. 2017. Corporate governance and cash holdings in MENA: evidence from internal and external governance practices. Research in International Business and Finance 39(1): 1-12. Antunes, M.T.P., M.C.P. Grecco and H. Formigoni. 2012. Mendonça Neto. A adoção no Brasil das normas internacionais de contabilidade IFRS: o processo e seus impactos na qualidade da informação contábil. Revista de Economia e Relações Internacionais 10(20): 5-19. Arellano, M. and S. Bond. 1991. Some tests of specification for panel data: Monte Carlo evidence and an application to employment equations. The Review of Economics Studies 58(2): 277-297. Armstrong, C.S., M.E. Barth, A.D. Jagolinzer and E.J. Riedl. 2010. Market reaction to the adoption of IFRS in Europe. The Accounting Review 85(1): 31-61. Assaf Neto, A. 2014. Finanças Corporativas e Valor. 7th ed. Atlas, São Paulo, Brazil. Baum, C.F., M. Caglayan, N. Ozkan and O. Talavera. 2006. The impact of macroeconomic uncertainty on non-financial firms’ demand for liquidity. Review of Financial Economics 15(4): 289-304. Bigelli, M. and J. Sánchez-Vidal. 2012. Cash holdings in private firms. Journal of Banking and Finance 36(1): 26-35. Bliss, B.A., Y. Cheng and D.J. Denis. 2015. Corporate payout, cash retention, and the supply of credit: evidence from the 2008-2009 credit crisis. Journal of Financial Economics 115(3): 521-540. Brazil. 2007. Lei n° 11.638, de 28 de dezembro de 2007. Altera e revoga dispositivos da Lei no 6.404, de 15 de dezembro de 1976, e da Lei no 6.385, de 7 de dezembro de 1976, e estende às sociedades de grande porte disposições relativas à elaboração e divulgação de demonstrações financeiras. Brasília, 2007. Avalaible at: http://tinyurl.com/o3fgkfv. Campello, M., E. Giambona, J.R. Graham and C.R. Harvey. 2011. Liquidity management and corporate investment during a financial crisis. Review of Financial Studies 24(6): 1944-1979. Carvalhal, A. and R.P.C. Leal. 2013. The world financial crisis and the international financing of Brazilian companies. Brazilian Administration Review 10(1): 18-39. Curtis, C.C., J. Garín, and M.S. Mehkari. 2017. Inflation and the evolution of firm-level liquid assets. Journal of Banking and Finance 81: 24-35. Dahrouge, F.M. and R. Saito. 2013. Políticas de cash holdings: uma abordagem dinâmica das Empresas Brasileiras. Revista Brasileira de Finanças 11(3): 343-373. Dittmar, A., J. Mahrt-Smith and H. Servaes. 2003. International corporate governance and corporate cash holdings. The Journal of Financial and Quantitative Analysis 38(1): 111-133. Ferreira, E.J. and R.P.C. Leal. 2011. Cash holdings of Brazilian and US firms: size and industry effects. Journal of International Finance and Economics 11(1): 75-84. Ferreira, M.A. and A.S. Vilela. 2004. Why do firms hold cash? Evidence from EMU countries. European Financial Management 10(2): 295-319. Gao, H., J. Harford and K. Li. 2013. Determinants of corporate cash policy: insights from private firms. Journal of Financial Economics 109(3): 623-639.

International Food and Agribusiness Management Review

216

Manoel et al.

Volume 21, Issue 2, 2018

García-Teruel, P.J. and P. Martínez-Solano. 2008. On the determinants of SME cash holdings: evidence from Spain. Journal of Business Finance and Accounting 35(1-2): 127-149. Gogineni, S., S.C. Linn and P.K. Yadav. 2012. Evidence on the determinants of cash holdings by private and public companies. Working Paper. University of Oklahoma, Norman, OK, USA. Hall, T., C. Mateus and I.B. Mateus. 2014. What determines cash holdings at privately held and publicly traded firms? Evidence from 20 emerging markets. International Review of Financial Analysis 33(3): 104-116. Han, S. and J. Qiu. 2007. Corporate precautionary cash holdings. Journal of Corporate Finance 13(1): 43-57. Jensen, M. 1986. Agency costs of free cash flow, corporate finance, and takeovers. American Economic Review 76(2): 323-329. Jensen, M. and W. Meckling. 1976. Theory of the firm: managerial behavior, agency costs and ownership structure. Journal of Financial Economics 3(4): 305-360. Johnston, J. and J. Dinardo. 1997. Econometric methods. McGraw Hill Higher Education, California, CA, USA. Keynes, J.M. 1936. The general theory of employment, interest and money. Harcourt Brace, London, UK. Kim, C., D.C. Mauer and A.E. Sherman. 1998. The determinants of corporate liquidity: theory and evidence. Journal of Financial and Quantitative Analysis 33(3): 335-359. Kusnadi, Y. and K.C.J. Wei. 2011. The determinants of corporate cash management policies: evidence from around the world. Journal of Corporate Finance 17(3): 725-740. Lee, E. and R. Powell. 2011. Excess cash holdings and shareholder value. Accounting and Finance 51(2): 549-574. Lozano, M.B., and S. Caltabiano. 2014. Cross institutional cash and dividend policies: focusing on Brazilian firms. Applied Economics 47(3): 239-254. Manoel, A.A.S., J.P.A. Eça, and M.B.C. Moraes. 2016. Custo do capital próprio em Empresas Brasileiras do setor sucroenergético: um estudo considerando a adoção das normas internacionais de contabilidade. Revista Universo Contábil 12(2): 117-137. Martínez-Sola, C., P.J. García-Teruel and P. Martínez-Solano. 2013. Corporate cash holding and firm value. Applied Economics 45(2): 161-170. Nason, R.S. and P.C. Patel. 2016. Is cash king? Market performance and cash during a recession. Journal of Business Research 69(10): 4242-4248. Neves, M.F., R.B. Kalaki, V.G. Trombin and J.M. Rodrigues. 2014. Sugarcane industry development analysis from the perspective of agro-industrial system quantification. Research in Agriculture 1(2): 42-57. Opler, T., L. Pinkowitz, R. Stulz and R. Williamson. 1999. The determinants and implications of corporate cash holdings. Journal of Financial Economics 52(1): 3-46. Ozkan, A. and N. Ozkan. 2004. Corporate cash holdings: an empirical investigation of UK companies. Journal of Banking and Finance 28(9): 2103-2134. Ross, S.A., R.W. Westerfield, and J.F. Jaffe. 2013. Corporate finance. 10th ed. McGraw-Hill, New York, NY, USA. Sant’anna, A.C., A. Shanoyan, J.S. Bergtold, M.M. Caldas, and G. Granco. 2016. Ethanol and sugarcane expansion in Brazil: what is fueling the ethanol industry? International Food and Agribusiness Management Review 19(4): 163-182. Santos, G.R., E.A. Garcia and P.F.A. Shikida. 2015. A crise na produção do etanol e as interfaces com as políticas públicas. Radar: Tecnologia, Produção e Comércio Exterior 1(39): 27-38. Steijvers, T. and M. Niskanen. 2013. The determinants of cash holdings in private family firms. Accounting and Finance 53(2): 537-560. Tahir, M.S. and M.N. Alifiah. 2015. Corporate cash holding behavior and financial environment: a critical review. International Journal of Economics and Financial Issues 5(1): 277-280. Trejo-Pech, C.O., M.A. Gunderson, T.G. Baker, A.W. Gray and M.D. Boehlje. 2015. Assessing cash holdings in agribusiness. International Food and Agribusiness Management Review 18(4): 85-104. Van Der Stede, W.A. 2011. Management accounting research in the wake of the crisis: some reflections. European Accounting Review 20(4): 605-623. Wooldridge, J.M. 2013. Introductory econometrics: a modern approach, 5th ed. South Western, Nashville, TN, USA.

International Food and Agribusiness Management Review

217

OPEN ACCESS International Food and Agribusiness Management Review Volume 21 Issue 2, 2018; DOI: 10.22434/IFAMR2017.0060 Received: 4 July 2017 / Accepted: 7 September 2017

Job preferences of agricultural students in Germany – A choice-based conjoint analysis for both genders Special issue: IFAMA 2017 symposium

RESEARCH ARTICLE Stephan G.H. Meyerding Postdoctoral Researcher, Department of Agricultural Economics and Rural Development, Marketing for Food and Agricultural Products, Platz der Göttinger Sieben 5, 37073 Göttingen, Germany

Abstract Knowledge of the job preferences of agricultural students is essential for the competitiveness of the industry. Yet no study is available in Germany. Other studies mostly used direct questions for the preference measurement, which raises possible concerns about a socially desirable response bias. For this reason, the present study combines a quasi-experiment (choice-based conjoint analysis) and direct questions (Likert scales) to measure the job preferences of 568 agricultural students in Germany and compares the results. In addition to finding gender differences, the study found that ‘income’ and ‘future perspective’ are the most important job characteristics for the job choice of agricultural students, and that they also prefer an increasing ‘work-lifebalance’ as well as a ‘rural location’ for their future employer. Insights about job characteristics’ attractiveness lead to a more transparent environment in which employers and (potential) employees make better-informed decisions, resulting in increased job satisfaction, performance and career sustainability. Keywords: war-for-talents, social desirability, work-home-conflict, gender, job choice JEL code: J28, J24, J43, J81, J71, J64 Corresponding author: stephan.meyerding@uni-goettingen.de

219

Meyerding

Volume 21, Issue 2, 2018

1. Introduction Each year over 9,000 new agricultural graduates enter the German job market (Dudek et al., 2015: 32). Upon graduation, these new agricultural professionals choose positions based on their skills and experience, goals, position availability, and job characteristic preferences (e.g. ‘opportunity for advancement’, ‘location’, ‘work-life-balance’, ‘salary’). Both employers and (potential) employees can benefit from job preference studies. Insights about job characteristics’ attractiveness as well as accurate information about applicants’ characteristics being available to employers leads to a more transparent situation in which candidates and employers can make better-informed decisions, resulting in increased job satisfaction, performance and career sustainability. The present study seeks to identify job characteristics influencing job attractiveness and choice among students of agricultural science in Germany. To the best of the author’s knowledge there is no study available which evaluates the job preferences of agricultural students in Germany. Even in the US, most studies use primary sources which are around 20 years old (Cheney, 2000; McGraw et al., 2012; Schneider, 1985). Much of the existing information on agricultural students’ and professionals’ job preferences in the US was obtained from topics addressed in salary studies (Barkley et al., 1999; Broder and Deprey, 1985; Popp et al., 2010). Other existing studies have examined working agricultural professionals (Marchant and Zepeda, 1995; Thilmany, 2000), but analyses have been descriptive as opposed to modeling choice behavior (McGraw et al., 2012). Furthermore, many studies of agricultural economics professionals in the US have analyzed respondents from the academic field with a focus on the relationship between gender and salary (Arbaiya, 2008; Thilmany, 2000). Although Hine and Cheney (2000) analyzed job preferences of agricultural economics professionals and examined gender and ethnic differences in preferences, no factors influencing the job choice were presented. Therefore, the current study seeks to address gaps in knowledge about job preferences among agricultural students in Germany. The study is novel for two reasons. First, it quantifies job preferences by using a choice-based conjoint analysis (CBCA) in a quasi-experiment, which reduces the effect of socially desirable responses, which might occur when using stated preferences alone (Meyerding, 2016c); it compares the results of the CBCA with the results of a measurement with Likert scales (stated preference approach). Second, it focuses on agricultural students in Germany, where no comparative study is available, and compares the results by gender, which previous German studies have not done (Lehberger and Hirschauer, 2016). In Germany, the results of the present study are of particular interest, since there is a shortage of skilled labor in the agricultural sector due to demographic changes (Meyerding, 2016a,b). Additionally, the present study builds on the existing foundation of literature on job choice, job preferences, and the relationship between occupation and gender. The results of the study provide employers with information on how to attract applicants who match well with respective work and company cultures and are likely to become satisfied and productive employees (Meyerding, 2015, 2016d). Prospective applicants can gain insights into personal choice decisions based on their own subjective preferences and goals (McGraw et al., 2012). 1.1 Previous job preference studies Beginning in the 1970s, early job preference studies in agriculture have mainly focused on the US, and have become more abundant over the last two decades. Past studies tended to examine samples from a specific profession or academic major, including agricultural economists (Hine and Cheney, 2000), agricultural college graduates (Barkley et al., 1999), farm operators (Stallmann and Nelson, 1995), academic sports management faculty (Mahony et al., 2006), accounting students (Bundy and Norris, 2011; Omar et al., 2015; Trump et al., 1970), and education PHD students (University of Iowa, 2011). In the majority of studies, the participants were college students making or simulating an initial professional position decision (Bundy and Norris, 2011; Butler et al., 2000; University of Iowa, 2011). Other research has investigated factors that have influenced current working professionals’ decision to take up their current positions (Hine and Cheney, 2000; Mahony et al., 2006; McGraw et al., 2012; Stallmann and Nelson, 1995). One recent German study evaluated the job preferences of horticultural students as well as vocational scholars and master craftsmen in horticulture (Meyerding, 2016d). Another study evaluated the job preferences of employees of horticultural International Food and Agribusiness Management Review

220

Meyerding

Volume 21, Issue 2, 2018

companies in Germany (Meyerding, 2015). However these two available studies for Germany only focused on the horticulture subsector, which might be not transferable to agriculture as a whole (Meyerding, 2016e). 1.2 Review of methodologies in previous studies Most job preference studies have relied on surveys to collect the necessary data. The response rate of the present study (32%) is comparable to those of past studies (e.g. 27%) (Barkley, et al., 1999), 55% (Hine and Cheney, 2000), and 68% (Bundy and Norris, 2011)). Most of the reviewed study results have been descriptive rather than analytic (McGraw et al., 2012). The University of Iowa (2011), Hine and Cheney (2000), and Trump et al. (1970) used survey instruments and reported preference rankings and/or descriptive statistic results. A few studies have also reported analytical results. Bundy and Norris (2011) had participants rate their preferences on Likert scales and conducted a chi-square analysis. In addition to descriptive analyses, Mahony (2006) also conducted a multiple regression analysis of factors associated with ‘participants’ willingness to leave their current job’. A probit model estimating the probability of off-farm employment for farm operators was used by Stallman and Nelson (1995); however, they collected only demographic and human capital data and not the participants’ preferences. Barkley, Stock, and Sylvius (1999) published multiple regression models for starting and current salaries of agriculture college graduates. In addition, possible factors affecting their job choice were identified from independent job preference variables. Butler et al. (2000) identified the most important job characteristics for accounting students by comparing their self-reported job preferences with human resources experts’ opinions of students’ preferences. McGraw et al. (2012) evaluated determinants that affect agricultural economics professionals’ job choice between government and academic positions using a binominal probit model. Also the studies in German horticulture, identifying the job characteristics preferences of students and vocational and master craftsmen scholars, as well as employees (Meyerding, 2015, 2016d,e), used direct questions and mainly reported descriptive statistics. All reviewed studies used direct questions for measuring preferences. These studies therefore might contain social desirability biases (Meyerding, 2016c). Social desirability is the tendency to answer questions in a way that the participant believes is desired by society or the interviewer. Consequently the answers to direct questions sometimes do not reflect actual behavior when the participant is not observed. For this reason, in addition to direct questions, the present study used a choice-experiment using CBCA to quantify the part-worth utilities of the job characteristics under investigation. 1.3 Studies sampling students In other studies, four types of job characteristics that are important to students have been identified: ‘advancement opportunities’, ‘compensation’ (including not only salary, but also ‘health benefits’ and ‘future earning potential’), ‘job security’, and ‘work environment’ (including social as well as professional relationships with others at the workplace) (Bundy and Norris, 2011; Butler et al., 2000; McGraw et al., 2012; Trump et al., 1970; University of Iowa, 2011). Although Bundy and Norris (2011) found that ‘starting salary’ was not a highly important factor in job choice, ‘compensation’, broadly defined, was an important characteristic. Although accounting students ranked ‘salary’ only 22nd out of 35 characteristics, ‘job security’, ‘health benefits’, and ‘expected future income’ were first, forth, and ninth, respectively. Other studies with student samples have probably underestimated the influence of long-term ‘compensation’ as a result of the ambiguous nature of the response options of ‘salary’ or ‘compensation’. These two response possibilities may have been interpreted as ‘starting salary’, but for students or recent graduates, potential or expected ‘future compensation/salary’ is possibly a better indicator of the importance of ‘compensation’ or ‘salary’ for their job choice (McGraw et al., 2012). Future ‘compensation’ and/or ‘salary’ might also be only one aspect of the broader job characteristic ‘future prospects’, which was highly ranked by the sample of horticultural students in Germany (Meyerding, 2016d,e), vocational and master craftsmen (Meyerding, 2016d) as well as employees of horticultural companies (Meyerding, 2016e). ‘Good future prospects’ include not only the ‘salary’, but also other kinds of career development, which could include taking another role such as teaching or leadership activities. It is defined as development in the direction of the individual’s goals (Meyerding, 2016d). The job characteristic ‘good future prospects’ has been found to be the most influential International Food and Agribusiness Management Review

221

Meyerding

Volume 21, Issue 2, 2018

for job satisfaction in German horticulture for different subgroups (Meyerding, 2015, 2016d,e). The most important job characteristics in a study identifying job preferences of horticultural students in Germany were, in descending order: ‘fair treatment of employees’, ‘work-home conflict’, ‘emotional dissonance’, ‘considerate supervisor’, and ‘supportive supervisor’ (Meyerding, 2016d). 1.4 Studies sampling working professionals For employees, the three most important types of attributes have been job ‘location’, ‘working environment’ (including social and professional relationships at the workplace), and ‘compensation’ (including salary and health benefits) (Barkley et al., 1999; Hine and Cheney, 2000; Mahony et al., 2006; McGraw et al., 2012). The major difference among current working employees and students is the relative importance of ‘future advancement opportunities’ to students and of ‘job location’ to employees (McGraw et al., 2012). In a study by Barkley et al. (1999), job characteristics preferences measured for initial job choice were quite similar to those for current job choice. ‘Job location’ and ‘benefits’ were ranked first and second, respectively, for both ‘starting salary’ and ‘current salaries’ of agricultural professionals who had graduated with agricultural degrees (e.g. in animal science, agribusiness, food science, natural resources) in the US (McGraw et al., 2012). However, differences may exist across fields as a result of different participant preferences or survey techniques. There may be a difference between results for direct and indirect questions, due to the influence of social desirability bias on the former (Meyerding, 2016c). Other differences might also occur. Barkley et al. (1999) did not ask about the importance of ‘opportunities for advancement’, but results from Mahony et al. (2006) indicate ‘opportunities for advancement’ (rank/tenure) were important to current professionals. Furthermore, for agricultural economists, ‘income’ was ranked only 17th out of 24 job characteristics, but ‘being a good match to career objectives’ was second after ‘work environment’ (Hine and Cheney, 2000), a concept not assumed to be relevant or not considered by other studies (Barkley et al., 1999; Mahony et al., 2006; McGraw et al., 2012). In a recent study by Meyerding (2015, 2016e,) employees of horticultural companies in Germany preferred the job characteristic ‘fair treatment of employees’ the most, followed by ‘new learning’, ‘considerate supervisor’, ‘skill use’, and absence of ‘emotional dissonance’. One limitation of Meyerding’s study was that he used only direct questions for the preference measurement. As mentioned earlier the results might therefore be subject to social desirability bias.

2. Materials and methods Job offers with seven job characteristics were used in the quasi-experiment. The current investigation involved a survey with (in the following order) 16 choice sets, to perform a CBCA and the evaluation of the importance of 28 different job characteristics (including ‘salary’, ‘future prospects’, ‘image of the company’, and ‘work-life-balance’, among others) on a Likert scale. Additionally, participants were requested to give some sociodemographic data, like their age, semester, gender, desire to run their own business, and study focus. 2.1 Description of the sample To assess agricultural students’ preferences for job characteristics, a questionnaire with 54 queries was conducted in a web-based format. The survey was spread through lectures, several e-mail lists, personal contacts and social media. The students were not compensated for contributing to this study. The study was done in November 2016 in Germany. Agricultural students from different universities participated in the survey. In total, 568 agricultural students contributed to the investigation. As the CBCA can be executed only on comprehensive data sets, all shown results (including the Likert scale parts and sample description) are grounded on the group of 568 choice set completers. No significant variances, in terms of socio-demographic features, were found among non-completers and completers of the CBCA (16 choice sets). For the group of completers, the socio-demographic attributes and answers to the questions (regarding whether participants could ‘imagine running their own business’, as well as regarding their ‘academic major’) shown in Table 1 apply. Amongst the choice set completers, the bulk was male. The mean age of the choice set completers International Food and Agribusiness Management Review

222

Meyerding

Volume 21, Issue 2, 2018

Table 1. Summary statistics of the 568 choice set completers.1 Characteristic

Age Semester Imagines running own business Study focus business administration Study focus plant science Study focus resource management Study focus animal science Study focus agribusiness and social science 1 ***

All (N=568)

Men (N=326)

Women (N=242)

Mean % 6 or 7

MannWhitney-U test P-value

23.00 5.11 4.76 4.39 3.21 2.67 2.89 4.29

23.46 5.27 5.39 4.67 3.53 2.65 2.59 4.50

22.24 4.88 3.91 4.03 2.78 2.70 3.30 4.04

0.000*** 0.120 0.000*** 0.006*** 0.000*** 0.742 0.001*** 0.040**

N/A N/A 47.20 40.10 25.90 9.80 18.40 39.90

N/A N/A 62.00 42.00 25.90 9.60 14.40 42.60

N/A N/A 27.20 37.90 14.00 10.10 23.60 36.30

significant at the 0.01 level (two-tailed); ** significant at the 0.05 level (two-tailed); * significant at the 0.10 level (two-tailed).

was 23 years; male participants were about one year older than female participants. The participants came from all semesters, with no significant differences between the genders. The sample is comparable to the population of agricultural students at universities in Germany as a whole (Statistisches Bundesamt, 2016). In Table 1, the first two columns show the data for all participants. Columns three and four present the data for the male subsample, and columns five and six, the data for the female subsample. Column seven presents the P-value of the Mann-Whitney-U test, which indicates differences between the genders. The questions regarding ‘imagining running their own business’ and ‘study focus’ were asked using a seven point Likert scale. In every second column for each group, the percentage is given of the share of the group who answered six or seven to the different questions. As can be seen in Table 1, 62% of the male subsample plans to run their own business in the future, whereas only 27% of the female subsample rated this question on the two highest values (strongly agree and agree). Significant differences could also be observed in the study focuses of the two sexes. The majority of men were business administration, plant science, agribusiness, or social science majors, while women were significantly more likely to be animal science majors than their male peers. 2.2 Choice-based conjoint analysis The two usual approaches, traditional conjoint analysis (TCA) or preference-based conjoint analysis, and CBCA, can be differentiated. TCA asks (directly) about preferences for the characteristics under investigation, whereas CBCA evaluates preferences more naturalistically by monitoring an amount of choice decisions (Green and Srinivasan, 1978, 1990; Green et al., 2001). For the application of CBCA in this study, seven phases had to be completed (Meyerding, 2016c): design of the stimuli; design of the selection situation; specification of a utility model; specification of a selection model; estimation of the utilities; interpretation, implementation; and disaggregation of the utilities. The mixture of the characteristics and their specifications presented in Table 2 finally results in 96 diverse stimuli (characteristics profiles). The characteristics used in the CBCA are those that are mentioned in a series of studies investigating job preferences (Esters, 2008; Esters and Bowen, 2005; Gore, 2006; Jones and Larke Jr, 2001; Scofield, 1994). The first characteristic is the ‘image’ of the company, with a focus on sustainability and the specifications reflecting all three dimensions of sustainability (ecologic, social, and economic). The sustainability ‘image’ of the company is assumed to be important, especially for younger generations, when it comes to their job International Food and Agribusiness Management Review

223

Meyerding

Volume 21, Issue 2, 2018

Table 2. Characteristics and their specifications of jobs analyzed in the choice-based conjoint analysis. Characteristics

Characteristic specifications

Image

No information Ecologically sustainable Socially sustainable Economically sustainable 30,000 € p.a. 40,000 € p.a. 50,000 € p.a. 60,000 € p.a. No information Career opportunities in the job Stepping stone for better job Below average Average Above average No information Good for the CV Employees are considered elite Good reputation with business partners Village Town City Metropolis Part-time 40 hours 40-50 hours More than 50 hours

Income

Future

Work-life-balance

Prestige

Place

Working hours

preferences (Meyerding, 2016d). The second characteristic is ‘income’, with specifications from ‘30,000 € p.a.’ to ‘60,000 € p.a.’, which reflects the possible span of starting incomes for agricultural graduates in Germany and has been shown to influence job preferences and job choice in a variety of studies (McGraw et al., 2012). Another characteristic which has been shown to be especially important for job satisfaction (Meyerding, 2015) and job preference (Hine and Cheney, 2000; Mahony et al., 2006) is ‘future prospects’, with the specifications of career ‘opportunities for professional growth in the job under consideration’ or that the ‘job under consideration is a stepping stone for another, better job’. Other characteristics are ‘work-life-balance’ (Lehberger and Hirschauer, 2016; Meyerding, 2016d,e: 30-37) and ‘prestige’ (Jurgensen, 1978; Warr, 1990), the later with the specifications, ‘good for the CV’, ‘employees are considered elite’, and ‘good reputation of the company with business partners’. One characteristic that has not been frequently investigated in job preference studies, but which is especially important for agricultural companies, is the ‘place’ or location of the company (Meyerding, 2016b). For example, horticultural companies in Germany named urbanization/ rural depopulation among the top causes for fundamental changes in their company in the next decade, which is one of the reasons for the labor shortage in the sector (Meyerding, 2016b). ‘Working hours’ are also an important characteristic that should be mentioned, because high workload is one of the distinguishing characteristics of jobs in agriculture compared to other industries in Germany (Lehberger and Hirschauer, 2016). Farmers in Germany work 1,713 hours a year, making them the group with the greatest working time (industry/production 1,400 hours, construction 1,517 hours; Deutscher Bauernverband e.V., 2013). The

International Food and Agribusiness Management Review

224

Meyerding

Volume 21, Issue 2, 2018

number of ‘working hours’ is also assumed to have an important impact on job preferences of agricultural students (Esters and Bowen, 2005; Meyerding, 2016d; Scofield, 1994). The selection (phase two) was done by generating an orthogonal reduced factorial design (using SPSS (IBM, Armonk, NY, USA), orthoplan), which lead to 32 cards. To construct the final choice design, the factorial design (after shifting – cyclic variation with three alternatives) was imported into SPSS and the 32 choice sets were generated using the plancards function. One of these choice sets is presented in Figure 1. Because evaluating 32 choice sets would have been too demanding for the participants, each participant needed to answer only 16 choice sets. Every time the questionnaire was started, a random variable was initiated with the value 0 or 1. If the random variable was 0 the participant saw the first 16 choice sets and if the random variable was 1, the participant saw the last 16 choice sets. After a welcome message, the student was requested to do a quasi-experiment (16 choice sets), where the participant had to select one out of three different jobs (stimuli) or select the none option (Figure 1). In the third phase (design of a utility model), the part-worth model was chosen since the characteristics ‘image’, ‘future’, ‘work-life-balance’, ‘prestige’, ‘place’, and ‘working hours’ have different relevance for every participant which cannot be projected. As in the CBCA, selections done by participants were monitored. Additionally to the utility model, a choice model was required (phase four). This was to define and explain the decision-making procedure of an individual. As is common for the CBCA, the multidimensional logit choice model was used. In the logit model the choice likelihood is defined only by the variances of the utilities. The estimation of the utilities (phase five) was complete by maximizing a log-likelihood function to estimate the parameters of a probability function for resolving non-linear minimization difficulties (Meyerding, 2016c). Phase six was the explanation and application. The total quantity of estimated part-worth utilities and overall utility amounts are not important when using the current value model; just the differences matter. The partworth utilities are an indicator of the utility of a characteristic specification compared to a basic category (Meyerding, 2016c). Afterwards the survey was conducted; part-worth utilities were estimated due a Cox regression. To prepare the survey data for the Cox regression a range of different jobs had to be completed (Backhaus et al., 2015: 227). In this process, a structure with the identical model is implemented for the logit choice model, which maximizes the identical likelihood function of the layered Cox regression, as it arises also in the CBCA (Meyerding, 2016c).

Job advertisement Image of employer

Work-life balance Prestigious

particularly ecologically sustainable 40,000 EUR no advancement opport./ important experiences below average no information

Size of location Working hours

city 40-50 hours per week

Gross salary p.a. Future orientation

Job advertisement Image of employer

Work-life balance Prestigious

particularly socially sustainable 50,000 EUR career opportunities with the employer average good reference for CV

Size of location Working hours

metropolis >50 hours per week

Size of location Working hours

Gross salary p.a. Future orientation

Gross salary p.a. Future orientation Work-life balance Prestigious

Figure 1. One of the different 32 choice sets.

International Food and Agribusiness Management Review

225

particularly economically sustainable 60,000 EUR job is a springboard for better work above average employees are considered elite village/small town part time

None of them

Meyerding

Volume 21, Issue 2, 2018

The part-worth utilities of all characteristics specifications were estimated for three groups: (1) all participants; (2) male subsample; and (3) female subsample. For the group of all participants, the value for the likelihood ratio statistic, which is chi-square distributed, is 129,768.694 (df (20)). The image ‘no information’, the income ‘30,000 € p.a.’, the future ‘no information’, the work-life-balance ‘average’, the prestige ‘no information’, the place ‘village’, and the working hours ‘part-time’ were selected as the base specifications in the Cox regression to estimate the part-worth utilities in the CBCA. Their part-worth utilities were consequently set to zero. Based on these basis specifications the utilities of the additional characteristic specifications can be understood. With a P-value of 0.00 the present model is highly significant for the group of all participants. Each part-worth utility is also highly significant (P<0.01), except for ‘40 hours’ (characteristic working hours) with P=0.94. For the group of male participants, the value for the likelihood ratio statistic, which is chi-square distributed, is 65,353.658 (df (20)). The same characteristics specifications as for the group of all participants were selected as the basic specification in the Cox regression. With a P-value of 0.00, the present model is also highly significant for the male subsample. Most part-worth utilities are also highly significant (P<0.01), except for ‘ecologically sustainable’ (P=0.04), ‘employees are considered elite’ (P=0.02), ‘city’ (P=0.62), ‘40 hours’ (P=0.73) and ‘40-50 hours’ (P=0.21). For the female subsample, the value for the likelihood ratio statistic is 46,536.043 (df (20)). Again the same characteristics specifications as for the other groups were selected as the basic specifications. With a P-value of 0.00, the present model is again significant for the female subsample. The majority of part-worth utilities are significant (P<0.01). However, some part-worth utilities show lower P-values or are not significant: ‘ecological sustainability’ (P=0.50), ‘socially sustainability’ (P=0.06), ‘economic sustainability’ (P=0.27), ‘employees are considered elite’ (P=0.86), ‘good reputation with business partners’ (P=0.22), ‘town’ (P=0.14), ‘city’ (P=0.69), and ‘40 working hours’ (P=0.24). 2.3 Assessment of job preferences using Likert scales The most frequently used method to evaluate the preferences for job characteristics is using Likert scales (Esters and Bowen, 2005; Jurgensen, 1978; McGraw et al., 2012; Meyerding, 2015, 2016d,e). A variation of characteristics, together with characteristics also included in the choice-based-conjoint analysis, had to be rated on a scale, with six options ranging from unimportant to essential. As earlier stated, the characteristics to be rated using Likert scales contain similar ones to those included in the CBCA. These are images which focus on sustainability – but not differentiated into the three dimensions – ‘income’, ‘future prospects’, and ‘work-life-balance’ (work-home conflict). The characteristics ‘prestige’, ‘place’, and ‘working hours’ are not used as questions for the Likert scale ratings. The differences between the items used for the two methods are because the Likert scale part was initially done to complement the quasi-experiment and not to compare the results of the different methods. The 28 job characteristics used for the preference measurement using Likert scales are based on the vitamin model by Warr (2007), which was developed for the measurement of job satisfaction and was also tested and used in recent studies by Meyerding (2015, 2016d) in German horticulture. Some of these characteristics were also found to be important in studies among agricultural students in the US (Esters, 2008; Esters and Bowen, 2005; McGraw et al., 2012; Scofield, 1994). The MannWhitney-U test was used to identify differences between the means of different subgroups.

3. Results In this section, the results for the conjoint analysis for all groups are presented first, followed by the results for the Likert scale.

International Food and Agribusiness Management Review

226

Meyerding

Volume 21, Issue 2, 2018

3.1 Part-worth utilities of job characteristic specifications as a result of the choice based conjoint analysis for the group of all participants The resulting part-worth utilities for the group of all participants are presented in Figure 2. The highest income has the highest part-worth utility, 1.76, followed by the second highest income (‘50,000 € p.a.’) with 1.20. After the income, the future perspective ‘stepping stone’ earns the third highest part-worth utility (0.80), followed by the future perspective ‘career opportunities in the job’ (0.76) and the prestige attributes ‘employees are considered elite’ (0.23) and ‘good for CV’ (0.23). Only after income, future, working hours, place, work-life-balance, and prestige, an attribute of image show a part-worth utility of 0.13. As stated before, just the variances are important, which means the part-worth utilities need to be understood compared to the basic specification. That also means that the place attribute ‘village’ yields a part-worth utility of 0.52 compared to the attribute ‘metropolis’, and is therefore more important than prestige. The same applies to the working-hours attributes ‘part-time’ and ‘40 hours’ compared to the attribute ‘more than 50 hours’. From four image attributes, the study participants prefer ‘social’ and ‘economic sustainability’ (both 0.13), but the part-worth utilities are very low compared to the other characteristics. Not surprisingly, the participants prefer the highest income of ‘60,000 € p.a.’. When it comes to future prospects, ‘stepping stone for a better job’ is ranked highest, followed by ‘career opportunities in the job’. In terms of work-life-balance, ‘above average’ is preferred (0.12), but the attribute ‘below average’ has an impressive negative part-worth utility of -0.43. For prestige, the part-worth utilities for the positive attributes differ little, but are higher than those of the characteristic image. Surprisingly, the place attribute ‘village’ is preferred by the agricultural students. For the characteristic working hours, the attribute ‘40 hours’ is preferred, working ‘part-time’ does not show a higher utility, ‘40-50 hours’ show a negative part-worth utility of -0.20, and working ‘more than 50 hours’ increases the negative impact to -0.60 compared to ‘part-time’ or ‘40 hours’.

1.76

1.50 1.20

1.00

0.89 0.76 0.59

0.50 0.23 0.23

0.11 0.13 0.13

-1.50

Income

Future

Work-lifebalance

Prestige

40-50 hours

40 hours

-0.20

None of these

-0.52

-0.60

More than 50 hours

-0.47

Part-time

-0.26

Metropolis

City

0.00 0.00

Town

Village

Employees are considered elite

0.00

Good for the CV

-0.43

No information

0.00

Above average

Average

Below average

Stepping stone for better job

Career opportunities in the job

No information

60.000 €

50.000 €

40.000 €

30.000 €

0.00

Good reputation with business partners

Image

Economically sustainable

-1.00

0.28

0.12 0.00

Socially sustainable

-0.50

0.00

Ecologically sustainable

0.00

No information

Parth-worth utilities (dimensionless)

0.80

The part-worth utilities expressing the utilitiy of a characteristic value in comparison to a reference value (the reference is marked in gray for each characteristic)

Place

Working hours None

Figure 2. Results of the choice-based conjoint analysis for the group of all participants (N=594). Characteristics’ importance from left to right: 2.75, 37.21, 18.82, 11.63, 5.92, 10.99 and 12.68%.

International Food and Agribusiness Management Review

227

Meyerding

Volume 21, Issue 2, 2018

3.2 Part-worth utilities of job characteristic specifications as a result of the choice-based conjoint analysis for the male subgroup The part-worth utilities for the male subgroup are presented in Figure 3. The highest income again has the highest part-worth utility, 0.93, followed by the second highest income (‘50,000 € p.a.’) with 0.61. After income, the future perspective ‘stepping stone for a better job’ earns the third highest part-worth utility (0.47), followed by the work-life-balance attribute ‘above average’ (0.12), the prestige attribute ‘good for the CV’ (0.35), and the image ‘socially sustainable’ (0.33). For the characteristic image, the specification ‘socially sustainable’ (0.33) shows the highest part-worth utility, which is about twice that of ‘economically sustainable.’ Unsurprisingly, the male participants prefer the highest income. When it comes to future prospects, ‘stepping stone for a better job’ (0.47) is again ranked the highest. In terms of work-life-balance, ‘above average’ is preferred (0.37), but ‘below average’ has an impressive negative part-worth utility of -0.62. For the characteristic prestige, ‘good for the CV’ (0.35) showed the highest part-worth utility. For the location (place) of the company the ‘village’ is again preferred, but this time the ‘city’ comes second (-0.03). For working hours, again the attributes of ‘part-time’ and ‘40 hours’ are ranked the highest. 3.3 Part-worth utilities of job characteristic specifications as a result of the choice-based conjoint analysis for the female subgroup The part-worth utilities for the female subsample are shown in Figure 4. The highest income receives again the highest part-worth utility of 0.97, followed by the second highest income with 0.63. After income, the future perspective ‘career opportunities in the job’ earns the third highest part-worth utility (0.54), followed by the third highest income and the future perspective attribute ‘stepping stone for a better job’ (0.54 and 0.50, respectively). After this come the work-life-balance attribute ‘above average’ (0.31) and the prestige attribute ‘good for the CV’.

1.50

0.93

0.51

0.35

Future

Work-lifebalance

Prestige

-0.06

None of these

40-50 hours

40 hours

-0.15

More than 50 hours

-0.37

Part-time

-0.03

Metropolis

-0.20

City

0.00 -0.02

Town

No information

Good for the CV

Employees are considered elite

-0.62

Village

0.00

Above average

Average

Below average

Stepping stone for better job

No information

0.00

Good reputation with business partners

Income

60.000 €

50.000 €

0.00

40.000 €

30.000 €

Socially sustainable

Image -1.50

0.37

0.14 0.18 0.00

Economically sustainable

-1.00

0.47

0.17

0.00

Ecologically sustainable

-0.50

0.36

0.33 0.14

0.00

0.61

Career opportunities in the job

0.50

No information

Parth-worth utilities (dimensionless)

1.00

-0.30

The part-worth utilities expressing the utilitiy of a characteristic value in comparison to a reference value (the reference is marked in gray for each characteristic)

Place

Working hours None

Figure 3. Results of the choice-based conjoint analysis for the male subgroup (N=594). Characteristics’ importance from left to right: 10.19, 28.70, 14.51, 19.14, 11.42, 11.42 and 4.63%.

International Food and Agribusiness Management Review

228

Meyerding

Volume 21, Issue 2, 2018

1.50

0.97

Image -1.50

Future

Work-lifebalance

Prestige

-0.20

None of these

40-50 hours

-0.35

More than 50 hours

-0.37

40 hours

-0.03

Metropolis

-0.16

Part-time

0.00 -0.07

0.00

City

0.08

Town

Good for the CV

-0.55

Employees are considered elite

-0.01

No information

0.00

Above average

Average

Below average

Stepping stone for better job

0.00

Career opportunities in the job

60.000 €

50.000 €

40.000 €

Income

No information

0.00

-0.08 0.00

30.000 €

0.05

Economically sustainable

-1.00

0.00

0.29

0.14

Socially sustainable

-0.50

0.31

Ecologically sustainable

0.00

0.54 0.50

Village

0.50

0.63

Good reputation with business partners

0.54

No information

Parth-worth utilities (dimensionless)

1.00

-0.04

The part-worth utilities expressing the utilitiy of a characteristic value in comparison to a reference value (the reference is marked in gray for each characteristic)

Place

Working hours None

Figure 4. Results of the choice-based conjoint analysis for the female subgroup (N=242). Characteristics’ importance from left to right: 6.67, 29.40, 16.36, 16.67, 9.09, 11.21 and 10.61%. For the characteristic image, the specification ‘socially sustainable’ shows the highest part-worth utility (0.14). However ‘economically sustainable’ receives a negative outcome (-0.08) compared to ‘no information’. Also the female subsample preferred the highest income (0.97), with a huge gap between this and the next highest income (0.63). For women, the ‘career opportunities in the job’ (0.54) are more important than the other future perspective ‘stepping stone for a better job’ (0.50). For work-life-balance, female students value the attribute ‘above average’ (0.31), and ‘below average’ shows a substantial negative utility (-0.55) compared to an average work-life-balance. For prestige, women valued ‘good for the CV’ (0.29) the most, and, surprisingly, the attribute ‘employees are considered elite’ earns a small negative part-worth utility of -0.08 compared to ‘no information’. Women again preferred the ‘village’ as a location for their future work, followed by, with much less utility (-0.03), the ‘city’. When it comes to working hours women tend to prefer ‘part-time’ and the drop-off of part-worth utility for more work hours is more substantial as for the other groups. 3.4 Results of the Likert scale evaluation for all three groups The second method used to determinate job preferences of agricultural students in Germany was the traditional Likert scale (1=unimportant to 6=essential) – the most frequently used technique. Table 3 presents the results (mean values) of the assessment of the job characteristics favored by agricultural students. As mentioned above, the characteristics are based on earlier research. A brief explanation of the items can be found in the article by Meyerding (2016d). The second column shows the mean for all participants, the third for the male subsample and the fourth for the female subsample. The P-values of the Mann-Whitney-U test are presented in the far right column of Table 3, indicating significant differences between the subgroups. The order starts with the highest value for the group of all participants and ends with the lowest. For all participants, the ‘fair treatment of the employees’ is the most important job characteristic, when using a direct question for preference measurement, followed by ‘work-home-conflict’, ‘new learning’, ‘supervisor behaves considerately’, and ‘good future prospects’. The characteristic ‘pay level’ attains only a middle ranking, placed 17 out of 28, with no significant difference between the genders. This result for ‘pay level’ (income) differs a lot from the results of the quasi-experiment for all three groups (Figures 2-4), where the highest income received the highest part-worth utility of all attributes for all characteristics under investigation. International Food and Agribusiness Management Review

229

Meyerding

Volume 21, Issue 2, 2018

Table 3. Results of the Likert scale part for all three groups.1 Job characteristic

All

Men

Women

MannWhitney-U test P-value

Fair treatment of employees Work-home-conflict New learning Considerate behavior of supervisors Good future prospects Skill use Job security Emotional dissonance Adequate equipment Quality of social contact Significance to self Task discretion Task coherence Pleasant environment Organizations morality in society Clear role requirements Pay level Availability of feedback Supervision is supportive Range of different tasks Safe work practices Amount of social contact Difficulty of job demands Value to society Conflict between job demands Future predictability Influence over the wider orga. Number of job demands

4.89 4.66 4.58 4.51 4.47 4.38 4.30 4.23 4.17 4.10 4.09 4.00 3.98 3.96 3.92 3.90 3.86 3.83 3.80 3.74 3.69 3.62 3.56 3.48 3.42 3.32 3.25 3.24

4.60 4.48 4.55 4.30 4.54 4.43 4.21 4.12 4.27 4.08 4.15 4.04 4.06 3.78 3.75 3.82 3.93 3.87 3.72 3.84 3.52 3.59 3.63 3.39 3.26 3.36 3.30 3.27

5.23 4.91 4.60 4.74 4.44 4.30 4.42 4.34 4.02 4.12 4.02 3.95 3.87 4.19 4.12 4.04 3.81 3.78 3.91 3.63 3.91 3.68 3.48 3.60 3.64 3.28 3.20 3.22

0.000*** 0.000*** 0.732 0.000*** 0.124 0.059* 0.042** 0.061* 0.012** 0.568 0.131 0.309 0.121 0.000*** 0.000*** 0.003*** 0.104 0.308 0.108 0.038** 0.000*** 0.288 0.105 0.052* 0.000*** 0.373 0.290 0.418

Standard deviation between 1.0 and 1.3; *** significant at the 0.01 level (two-tailed); ** significant at the 0.05 level (two-tailed); * significant at the 0.10 level (two-tailed).

Significant differences between the results of the Likert scale evaluation of men and women could be found in the present study. For women, the characteristics ‘fair treatment of employees’, ‘work-home-conflict’, ‘supervisor behave considerate’, ‘job security’, ‘emotional dissonance’, ‘present environment’, ‘organizations’ morality in society’, ‘clear role requirements’, ‘safe work practices’, ‘value to society’, and ‘conflict between job demands’ are more important than for men. For the male subsample, only the job characteristics ‘skill use’, ‘adequate equipment’, and ‘variation of tasks’ are more important than for the female subsample. The most important job characteristics for men are ‘fair treatment of employees’, ‘learning opportunities’, ‘good future prospects’, ‘work-home-conflict’, and ‘skill use’. For women the same order applies with the exception of ‘work-home-conflict’ and ‘good future prospects’, which are in third and fourth place, respectively. This indicates that men value ‘good future prospects’ more than ‘work-home-conflict’, and women prefer ‘work-home-conflict’ (or rather, less of that conflict) more than ‘future prospects’.

International Food and Agribusiness Management Review

230

Meyerding

Volume 21, Issue 2, 2018

4. Discussion The results of the conjoint analysis for the group of all participants indicate that ‘income’ has the highest impact on job choice for agricultural students in Germany for both female and male subsamples. This result is different from the results of the direct questions about job characteristics preferences, where ‘income’ is ranked 17th out of 28 job characteristics. Similar results using direct questions were observed by Meyerding (2016d) for horticultural science students in Germany using the same items as in the present study. Horticultural students ranked income in 18th place out of 28 (Meyerding, 2016d). The middle place rank of ‘income’ when using direct questions was also observed in a series of other studies in the US, Germany, and the United Kingdom (Bundy and Norris, 2011; Butler et al., 2000; Cheney, 2000; Esters and Bowen, 2005; Jones and Larke Jr, 2001; Meyerding, 2016e; Scofield, 1994; Jurgensen, 1978; Trump et al., 1970). As mentioned above, this gap between the results could be due to the effect of social desirability bias. Social desirability is the trend of participants to act or reply in a way they assume society considers appropriate (Crowne and Marlowe, 1960). This effect leads to the difficulty that the possible employee states he/she prefers, for example, an employer whose business is ‘ecologically sustainable’, but does not show this behavior to the same extent in a real job-choice situation when not being observed (Meyerding, 2016c). The findings of research analyzing job preferences on the basis of direct questions are, then, problematic. The results of this study suggest that the use of a mixture of approaches to get an understanding of the real behaviors of job seekers and employees is very important. For examples of other approaches and a discussion of choiceexperiments see Beckley et al. (2012). The second most important job characteristic according to the conjoint analysis is ‘future prospective’, a job characteristic which is rarely mentioned by other research in this arena (McGraw et al., 2012). This feature is not only important for job choice and preferences of agricultural students, but has also been found to be the most important job characteristic influencing the job satisfaction of horticultural employees (Meyerding, 2015), as well as of vocational and master craftsmen scholars in German horticulture (Meyerding, 2016d). The result of the present study suggest that employers in agriculture should pay increased attention to communicating the ‘future prospects’ of employees and job applicants, and try to paint a positive but realistic picture of the future. This helps increase the attractiveness of employers, as well as increasing the job satisfaction of the employees already in business. The female subsample favors the future prospect ‘career opportunities in the job,’ whereas the male subsample prefers the attribute ‘stepping stone for a better job.’ This difference can probably be interpreted as the tendency of women to focus on stability and of men to focus on development. Another important job characteristic is ‘work-life-balance’, which is also named as a megatrend in the agricultural subsector horticulture (Meyerding, 2016b). A below average ‘work-life-balance’ seems to have a very high negative impact on job choice in agriculture for both men and women. This needs to be mentioned because work in agriculture is more time consuming than in other sectors (Lehberger and Hirschauer, 2016). The need for an improved ‘work-life-balance’ should lead to changing processes in the structural and procedural organization of agricultural companies, in order to be an attractive employer in the future (Meyerding, 2016a). An unexpected result was present for the job characteristic ‘location’. Agricultural companies report struggling with urbanization (Meyerding, 2016b) as one reason for their labor shortage. The results of the conjoint analysis suggest that agricultural students prefer to work in rural areas and the part-worth utilities show an inverse relation to the population size of the ‘location’ of the company. This result is positive for agricultural companies looking for agricultural graduates, but it is not clear if this preference is also present for other groups of potential employees of agricultural companies, so companies in rural areas might still have problems finding workers. The unexpected result might be due to self-selection in the sample of the present study. People who show a preference for rural areas might choose to study agriculture more often than others. In addition, for the overall German population the preference for urban locations is increasing (Kurz, 2004).

International Food and Agribusiness Management Review

231

Meyerding

Volume 21, Issue 2, 2018

For ‘working hours’ the part-worth utilities decrease with increasing the duration. As discussed above, this job characteristic is of special importance in agriculture, as the work is more time consuming than in other industries (Lehberger and Hirschauer, 2016). One interesting result that should be mentioned is that ‘part-time’ showed almost the same part-worth utility as work duration of ‘40 hours’ a week. ’40 hours’ seems to be the preferred duration of work a week for agricultural students in Germany. Only for the female subsample is there a slightly higher preference for ‘part-time’ work. This preference might be linked to a wish to reduce ‘work-family-conflict’ in the future. ‘Part-time’ work, as well as flexible working hours, significantly reduces ‘work-family conflict’ (Meyerding, 2015, 2016e). Another unexpected result is the relatively low importance of the ‘image’ of the employer in terms of sustainability. For the group of all participants, the part-worth utilities of all three sustainability dimensions are at about the same low level compared to the attribute ‘no information’. This is in line with the results of the direct measurement using a Likert scale, where ‘organizations’ morality in society’ is ranked 15th out of 28. On the other hand, the job characteristic ‘fair treatment of employees’ is in first place. It is not assumed that social desirability is the reason for this gap. It is more likely that ‘social sustainability’ is not associated with the ‘fair treatment of employees’ by the study participants. In the conjoint analysis all sustainability images showed a positive part-worth utility for both women and men, with the exception that for the female subsample the attribute ‘economically sustainable’ showed a small negative part-worth utility compared to ‘no information’ (about the company’s image in terms of sustainability issues). To be economically sustainable (or in other words, successful) seems to have a negative connotation for the female subsample. The reason for this is unclear. The order of sustainability issues in the result of the conjoint analysis as well as the Likert scale measurement differ to those of horticultural students in Germany, where, for example, the ‘organizations’ morality in society’ was ranked relatively high, 10th out of 28 job characteristics (Meyerding, 2016d). Maybe agricultural and horticultural students differ in their values regarding sustainability. The high value to students of a positive sustainability image that is assumed in other studies (Meyerding, 2016d) cannot be supported with the results of the quasi-experiment in the present study. This is only true for the job characteristic ‘image’ and may be due to wording and to the participants‘ lack of knowledge. The results of the Likert scale measurement indicate that ‘fair treatment of employees’, ‘work-home-conflict’, ‘new learning’, ‘considerate behavior of supervisors’, ‘good future prospects’, and ‘skill utilization’ are the preferred job characteristics of agricultural students in Germany. The results show that agricultural companies need to focus on these job characteristics to be an attractive employer today and in the future. The differences between the subgroups suggest that for men the job characteristics relating to work content are of special interest, whereas women focus more on the work environment. 4.1 Limitations There are some broadly documented weaknesses of conjoint analysis in general. One case is that participants occasionally use simplification tactics to reply to challenging full-profile tasks. Respondents might consider just the most significant characteristics; this could end up in overstated differences in significance among the most and the least significant aspects. This could have led to an overestimation of the most important factor, ‘income’, in this study (though, allowing for such an overestimation, it nonetheless emerges as the most important job characteristic). Respondents employ more energy making real-world choices than they do making judgements in a CBCA, particularly in high-involvement decision-making situations such as can be assumed for job choice (Meyerding, 2016c). Other explanations for the variances among the outcomes of the two approaches, beyond the possible social desirability effects already proposed as the main explanation, has to be stated. For instance, the statement words in the Likert scale portion can affect the evaluation (Meyerding, 2016c). Also, the items in the Likert scale task do not perfectly represent the concepts used as job characteristics in the conjoint analysis. For a better comparison of the two methods the items should be identical.

International Food and Agribusiness Management Review

232

Meyerding

Volume 21, Issue 2, 2018

One more restriction for the choice-experiment was the comparatively high amount of choice sets (16), which was problematic for the study subjects and probably led to non-completers; nevertheless, the non-completers do not vary in terms of sociodemographic characteristics from the set of completers (Meyerding, 2016c). 4.2 Future research A supplement investigation may be recommended using adaptive conjoint analysis, which is assumed to be more suitable for high-involvement tasks (Bruns, 2015) such as job choice and job characteristics’ preference measurement. The items in the Likert scale part should, in a follow-up study, be identical to those in the conjoint analysis of the present study. In the present study, only two subgroups for both sexes were analyzed, and it is clear that students or employees have many other personal characteristics that influence job preferences (Warr, 2013). Conjoint analysis is restricted to only a few characteristics that can be included in the analysis. For this reason future research should also evaluate more personal characteristics, such as ‘personality characteristics’ or ‘political views’, and more job characteristics, to get a more detailed picture about the job preferences of agricultural students, by using, for example, a structural equation model. It would also be of interest to investigate the job preferences of groups other than students of agriculture.

5. Conclusions It can be summarized that main variances exist in findings produced from direct questions, as with Likert scales, and quasi-experimental results. One potential cause for these variances in the identical sample might be the social desirability effect. Consequently findings of research examining job preferences that primarily apply direct questions may be biased to the preference for socially-valued job features. For this reason, a mixture of methods to get a knowledge of the real manners of job seekers, students, and employees is important, as demonstrated by the present study. Future research should cluster the data to create student segments that are more homogeneous and compare the preferences of these segments to draw a more detailed picture. The results show a rather different picture of the preferences relating to the ‘sustainability image’ of the employer, and of the importance of ‘income’ for job preferences and job choice than do various investigations in this area. The outcome that ‘sustainability image’ and ‘prestige’ are associated with relatively low partworth utilities indicates that, confronted with other, associated choices including ‘income’, ‘future prospects’, and so on, students do not see much value in such characteristics. This is different from findings reported by other studies in this field (Bundy and Norris, 2011; Hine and Cheney, 2000; Mahony, 2006; McGraw et al., 2012; Meyerding, 2016d; Trump et al., 1970; University of Iowa, 2011). It also implies that there is more emphasis on ‘income’ and ‘future prospects’ when choosing a job in agriculture than is claimed by activists and media figures who promote sustainability management in agriculture as a tool for attracting potential employees.

References Arbaiya, A. 2008. Factors affecting the salary of agricultural economics professionals. Master’s thesis, University of Arkansas, Fayetteville, AR, USA. Backhaus, K., B. Erichson and R. Weiber. 2015. Fortgeschrittene Multivariate Analysemethoden: Eine anwendungsorientierte Einführung. [Advanced multivariate analysis methods: an application-oriented introduction.] 3. überarbeitete und aktualisierte Auflage. Springer Gabler, Berlin, Germany. Barkley, A.P., W.A. Stock and C.K. Sylvius. 1999. Agricultural graduate earnings: the impacts of college, career, and gender. American Journal of Agricultural Economics 81(4): 785-800. Beckley, J.H., D. Paredes and K. Lopetcharat. 2012. Product innovation toolbox: a field guide to consumer understanding and research. John Wiley and Sons, Hoboken, NJ, USA. Broder, J.M. and R.P. Deprey. 1985. Monetary returns to bachelors and masters degrees in agricultural economics. American Journal of Agricultural Economics 67(3): 666-673.

International Food and Agribusiness Management Review

233

Meyerding

Volume 21, Issue 2, 2018

Bruns, F. 2015. Franchise-Attraktivität: eine präferenzanalytische Studie am Beispiel von Investoren im markengebundenen Vertragshandel der Automobilwirtschaft in der Volksrepublik China. [Franchise attractiveness: a preference analysis study based on the example of investors in the traded automotive trade agreement in the people’s Republic of China.] 1. Auflage. GRIN Verlag, München, Germany. Bundy, P., and D. Norris. 2011. What accounting students consider important in the job selection process. Journal of Applied Business Research 8(2): 1-6. Butler, S.A., D.E. Sanders and S.M. Whitecotton. 2000. Student and recruiter insights on the importance of job attributes. Journal of Managerial Issues 12(3): 337-351. Cheney, L.M. 2000. Tracking agricultural economics professionals. Review of Agricultural Economics 22(1): 17-22. Crowne, D.P. and D. Marlowe. 1960. A new scale of social desirability independent of psychopathology. Journal of Consulting Psychology 24(4): 349-354. Deutscher Bauernverband e.V. 2013. Branchenvergleich: Arbeitszeiten in der Landwirtschaft unterscheiden sich kaum noch. [Industry comparison: working hours in agriculture are hardly different.] Available at: http://tinyurl.com/y9b9fqzn. Dudek, K., B. Glässner and C. Tauch. 2015. Hochschulrektorenkonferenz 2015 – Statistiken zur Hochschulpolitik 1/2015. [Hochschulrektorenkonferenz 2015 – Statistics on higher education policy 1/2015.] Available at: https://tinyurl.com/ydxqzhra. Esters, L. 2008. Influence of career exploration process behaviors on agricultural students’ level of career certainty. Journal of Agricultural Education 49(3): 23-33. Esters, L.T. and B.E. Bowen. 2005. Factors influencing career choices of urban agricultural education students. Journal of Agricultural Education 46(2): 26-35. Gore, P.A. 2006. Computer-based career exploration: usage patterns and a typology of users. Journal of Career Assessment 14(4): 421-436. Green, P.E., A.M. Krieger and Y. Wind. 2001. Thirty years of conjoint analysis: reflections and prospects. Interfaces 31(3): 56-73. Green, P.E. and V. Srinivasan. 1978. Conjoint analysis in consumer research: issues and outlook. The Journal of Consumer Research 5(2): 103. Green, P.E. and V. Srinivasan. 1990. Conjoint analysis in marketing: new developments with implications for research and practice. Journal of Marketing 54(4): 3-19. Hine, S. and L.M. Cheney. 2000. Career choices and challenges among agricultural economists. Review of Agricultural Economics 22(1): 34-41. Jones, W.A. and A. Larke Jr. 2001. Factors influencing career choice of African American and Hispanic graduates of a land-grant college of agriculture. Journal of Agricultural Education 42(1): 39-49. Jurgensen, C.E. 1978. Job preferences (what makes a job good or bad)? Journal of Applied Psychology 63(3): 267-276. Kurz, K. 2004. Labour market position, intergenerational transfers and home-ownership: a longitudinal analysis for west German birth cohorts. European Sociological Review 20(2): 141-159. Lehberger, M. and N. Hirschauer. 2016. Recruitment problems and the shortage of junior corporate farm managers in Germany: the role of gender-specific assessments and life aspirations. Agriculture and Human Values 33(3): 611-624. Mahony, D.F., M. Mondello, M.A. Hums and M. Judd. 2006. Recruiting and retaining sport management faculty: factors affecting job choice. Journal of Sport Management 20(3): 414-430. Marchant, M.A. and L. Zepeda. 1995. The agricultural economics profession at the crossroads: Survey results of faculty salary, employment, and hiring prospects. American Journal of Agricultural Economics 77(5): 1322-1328. McGraw, K., J.S.H. Popp, B.L. Dixon and D.J. Newton. 2012. Factors influencing job choice among agricultural economics professionals. Journal of Agricultural and Applied Economics 44(2): 251-265. Meyerding, S.G.H. 2015. Job characteristics and job satisfaction: a test of Warr’s vitamin model in German horticulture. The Psychologist-Manager Journal 18(2): 86-107.

International Food and Agribusiness Management Review

234

Meyerding

Volume 21, Issue 2, 2018

Meyerding, S.G.H. 2016a. Organization 2020 – An empirical study of the current and future importance of organizational issues for horticultural companies and their perceived competence in Germany. Russian Journal of Agricultural and Socio-Economic Sciences 51(3): 58-69. Meyerding, S.G.H. 2016b. Change management study of horticulture 2015 – Conditions and success factors. Journal of Organisational Transformation and Social Change 13(2): 123-146. Meyerding, S.G.H. 2016c. Consumer preferences for food labels on tomatoes in Germany – A comparison of a quasi-experiment and two stated preference approaches. Appetite 103: 105-112. Meyerding, S.G.H. 2016d. Job satisfaction and preferences regarding job characteristics of vocational and master craftsman scholars and horticulture students in Germany. Review of Agricultural and Applied Economics 19(1): 30-49. Meyerding, S.G.H. 2016e. Mitarbeiterzufriedenheit in KMU: mit einer empirischen Erhebung auf Grundlage des Vitamin-Modells und Potenzialanalyse der Mitarbeiterzufriedenheitsmessung für die soziale und ökonomische Dimension der Nachhaltigkeit am Beispiel von Gartenbauunternehmen.[Employee satisfaction in SMEs: an empirical survey based on the vitamin model and the potential analysis of employee satisfaction measurement for the social and economic dimension of sustainability, as exemplified by horticultural companies] Berichte aus der Betriebswirtschaft. Shaker Verlag, Aachen, Germany. Omar, M.K., A. Zakaria, S. Ismail, J.S.L. Sin and V. Selvakumar. 2015. Job selection preferences of accounting students in Malaysian private universities. Procedia Economics and Finance 31: 91-100. Popp, J.S.H., D.J. Newton, D. Pittman, D.M. Danforth and A. Abdula. 2010. Comparison of factors influencing salaries of agricultural economics professionals in academic and federal employment. Available at: http://tinyurl.com/yage5px9. Schneider, V.E. 1985. The market for agricultural economists: discussion. American Journal of Agricultural Economics 67(5): 1223. Scofield, G.G. 1994. An Iowa study: factors affecting agriculture students career choice. North American Colleges and Teachers of Agriculture Journal 38(4): 28-30. Stallmann, J.I. and J.H. Nelson. 1995. Employment history and off-farm employment of farm operators. Journal of Agricultural and Applied Economics 27(2): 475-487. Statistisches Bundesamt. 2016. Bildung und Kultur – Studierende an Hochschulen: Wintersemester 2015/2016. [Education and culture – students at higher education: wintersemester 2015/2016.] 2016. Available at: https://tinyurl.com/ybv6fony. Thilmany, D. 2000. Gender based differences of performance and pay among agricultural economics faculty. Review of Agricultural Economics 22(1): 23-33. Trump, G.W., H.S. Hendrickson, W.L. Compfield, D.H. Chamer, T.F. Keller, S.W. Lyverse, H.J. Mottice, J.M. Saada, W.P. Tenney, M.F. Usry and T.H. Williams. 1970. Job selection preferences of accounting students. Journal of Accountancy (129): 84-86. University of Iowa. 2011. College of Education. Factors influencing job choice: 2010-11. Available at: https://tinyurl.com/yb4q6wrt. Warr, P. 1990. Decision latitude, job demands, and employee well-being. Work and Stress 4(4): 285-294. Warr, P. 2013. Jobs and job-holders: two sources of happiness and unhappiness. In: The Oxford handbook of happiness, edited by S.A. David. Oxford University Press, Oxford, UK, pp. 733-750. Warr, P.B. 2007. Work, happiness, and unhappiness. Lawrence Erlbaum Associates, Mahwah, NJ, USA.

International Food and Agribusiness Management Review

235

OPEN ACCESS International Food and Agribusiness Management Review Volume 21 Issue 2, 2018; DOI: 10.22434/IFAMR2017.0116 Received: 4 July 2017 / Accepted 16 November 2017

Allocating production risks through credit cum insurance contracts: the design and implementation of a fund for small cotton growers to access market finance Special issue: IFAMA 2017 symposium

RESEARCH ARTICLE Mauro Alema and Julio Jorge Elias

aSenior

Infrastructure Specialist, Infrastructure and Environment Sector, Inter-American Development Bank, 1300 New York Avenue, N.W. Washington, D.C. 20577, USA bProfessor,

Department of Economics and Business School, Universidad del CEMA, Av. Cordoba 374, Buenos Aires (1054), Argentina

Abstract Collateral requirements that lenders place on small cotton producers can lead to risk rationing and to farmers’ dependence on downstream parties. This paper presents a cotton fund that consists of a set of contracts – credit, insurance, warrant and forward – that enables producers to tackle specific agricultural risks and gain access to market finance. These financial contracts proved to be successful at guaranteeing the fund as issuer of state-contingent debt securities in the capital markets. The fund, as an intermediary, lent to cooperatives to help finance small cotton producers in northern Argentina. The paper explains the experimental design of this innovative fund and presents a potential alternative to government intervention by moving away from ex post subsidies for small producers and, instead, facilitating ex ante private credit. The paper contributes to the literature on rural financial intermediation by designing a new mechanism to raise funds coming from relatively uninformed investors and creating collateral substitutes through delegated monitoring to overcome asymmetric information and limited commitment. Keywords: cotton, agribusiness, rural development, financial intermediary, credit, insurance, risk sharing JEL code: O13, Q14, Q13 Corresponding author: je49@ucema.edu.ar

237

Alem and Elias

Volume 21, Issue 2, 2018

1. Introduction Rural economies are characterized by fragmented and incomplete markets, where the ability of small farmers to make long-term investments, take calculated risks and ultimately smooth consumption streams is limited by available financial services. Even though the physical proximity of small, tight-knit groups could facilitate risk pooling, local markets cannot offer sufficient diversification opportunities and the costs of funds can be high. Hence, farmers and cooperatives can ultimately face considerable residual risk. The lack of diversified outside financial intermediaries, such as rural banks and insurance companies, can be attributed to asymmetric information and limited enforcement. These problems reduce the credible promises that a would-be issuer of state-contingent promises can ensure. Making new financial services and contract forms available creates new trading opportunities and allows farmers to specialize in high value income generating activities. This article addresses the design of a fund as a market-based solution to allocate production and distribution risks of small cotton producers and cooperatives. It also presents an experimental implementation of the fund in the province of Chaco, Argentina. In recent years, despite exceptional natural conditions and a long tradition in cotton production in northern Argentina, small producers and cooperatives have been rationed from private market finance. Government intervention through public banks and subsidies crowded out the development of private markets to insure and finance production. Like Guirkinger and Boucherâ&#x20AC;&#x2122;s (2008) model of private information and the literature on financial intermediation since Stiglitz and Weiss (1981), financial innovation can be understood as efforts to find new mechanisms to create collateral substitutes. One prominent strategy, followed in this paper, is for the lender to use intermediaries or delegated monitors to reduce information asymmetries. In a prototypical mechanism design moral hazard problem, standard credit contracts can be improved by incorporating contingencies for bankruptcy,1 and some interim, potentially random monitoring can be predicted. â&#x20AC;&#x2DC;The contingencies, though costly to verify, kick in when the borrower most needs them, at very low project realizationsâ&#x20AC;&#x2122; (Townsend, 2003). If small producers lack access to insurance, then collateral has an additional adverse effect on loan demand as producers may not be willing to risk losing their assets. The design of inter-linked contracts introducing new parties to monitor and insure the production risk (surety bond) and the distribution risk (warrant) of cotton fiber is a key contribution of this paper to the literature. By embedding the proposed contractual arrangement in a cotton trust fund (CTF), a state-contingent contract can be drawn up to transfer risk from the borrower to third parties: insurance and warrant companies. The CTF was designed as a financial fund to issue market debt securities and intermediate finance coming from outside, uninformed investors for small producers in cotton cooperatives. Stiglitz and Weiss (1981) showed that credit rationing may arise as an imperfect mechanism to attenuate problems of asymmetrical information and limited liability. The CTF is a tool that attempts to solve these problems by providing a substitute for collateral and a monitoring mechanism. The paper shows how credit and insurance contracts can be combined to facilitate access to market finance. Conning and Udry (2007) provide an excellent review of the vast theoretical literature on credit constraints applied to the agricultural sector in developing countries. They highlight the essential role played by intermediaries in the expansion of exchange possibilities for the economy by creating new instruments and contractual forms to overcome the trading gaps and missing markets that information and enforcement problems create. Our paper contributes to the literature of credit rationing and rural financial intermediation by introducing a new mechanism that creates collateral substitutes and monitoring mechanisms to overcome these obstacles to trade.

The theory is generalized to encompass aggregate, observable shocks in Phelan (1994), and to allow interim communication, or unsolicited announcements of unobserved idiosyncratic shocks (Prescott, 2003).

International Food and Agribusiness Management Review

238

Alem and Elias

Volume 21, Issue 2, 2018

Some related financial tools have been implemented with mixed results in other countries. One example is the development of warehouses in rural areas that enable smallholder farmers to access finance from credit cooperatives by formalizing their crops as collateral (Coulter and Onumah, 2002). Towo and Kimaro (2014) showed that the warehouses have played an important role in facilitating agriculture related activities in Tanzania. Another related system is an inventory based credit system, where groups of farmers place their product in a warehouse and a lending institution uses the inventory as collateral to extend loans to farmers. The group of borrowers has a collective responsibility over the management of the inventory. The main difference to the warehouse is the focus on groups to build trust. This system has been tried in Ghana. It generated major immediate benefits to participating farmers but has not proven economically sustainable because of the small volumes of grain involved (Coulter and Onumah, 2002). The remainder of the paper proceeds as follows. Section 2 provides some background on Chacoâ&#x20AC;&#x2122;s cotton industry and on the traditional financing scheme of small cotton producers. Section 3 discusses the link between the CTF and the theory of credit rationing and rural financial intermediation. Sections 4 and 5 analyze the critical issues involved in designing and implementing the set of financial contracts embedded in the fund to finance cotton producers. Section 6 discusses market outcomes and Section 7 concludes.

2. Background 2.1 The cotton industry Cotton crop production in the province of Chaco represented 69% of the total cotton crops sown in Argentina in 2009-2010 (MAGPN, 2010). Cotton fiber production is the first link in the textile manufacturing process. Figure 1 shows the evolution of the price of cotton for the period 2001-2011, which shows the substantial risk faced by producers. Throughout the years 2002-2005, the series exhibited high volatility with wide fluctuations, a sharp drop in 2009 and a large increase in 2010.

3.50 3.00 2.50

$/kg

2.00 1.50 1.00 0.50

pJa 01 n M -02 ay Se -02 p Ja -02 n M -03 ay Se -03 p Ja -03 n M -04 ay Se -04 p Ja -04 n M -05 ay Se -05 p Ja -05 n M -06 ay Se -06 p Ja -06 n M -07 ay Se -07 p Ja -07 n M -08 ay Se -08 p Ja -08 n M -09 ay Se -09 p Ja -09 n M -10 ay Se -10 p Ja -10 n M -11 ay Se -11 p11

0.00

Figure 1. Evolution of the price of cotton (cotton free on board price in $/kg; type D Long 27,0 mm (1 1/16â&#x20AC;?)), Argentina, September 2001-October 2011 (Ministerio de Agroindustria de Argentina, 2017).

International Food and Agribusiness Management Review

239

Alem and Elias

Volume 21, Issue 2, 2018

2.2 The traditional financing scheme Small cotton farmers have historically been organized in cooperatives to finance production inputs, to share processing facilities, and to market the cotton fiber. Cooperatives accessed public bank loans to buy seeds and fertilizers that were then distributed to member farmers to produce raw cotton. Chaco counted with 15 active cooperatives in 2010 that were associated with a second-degree cooperative called Unión de Cooperativas Algodoneras Limited (UCAL) which was composed of 1,866 small-scale cotton producing members. These cooperatives produced and processed approximately 36% of total raw cotton, a figure that was diminishing. Active members decreased as cooperatives were unable to access credit to finance their crop inputs. By the end of 2009, cotton cooperatives had accumulated debt with public banks which left them with no choice but to access informal in-kind credit from input suppliers at implicit interest rates reaching 70% in many cases. Input suppliers act as monitoring lenders who must lend primarily out of their own equity and charge exceedingly high interest rates to recover monitoring costs (Aleem, 1990). The opportunity to improve led the government of Chaco, through a specialized agricultural finance unit (UCEF), to work together with two financial intermediaries – a bank and an insurance company –to review the traditional financing scheme and design an innovative fund to provide market finance to cotton cooperatives.

3. The link between the cotton trust fund and the theory of credit rationing and rural financial intermediation Before the CTF was developed, small cotton producers had access to informal in-kind credit and governmentassisted loans but not to insurance. In most cases, access to formal private finance was limited as lenders requested land collateral which made it difficult for small producers to demand credit. In their seminal paper on credit rationing, Stiglitz and Weiss (1981) studied the impact of limited liability on feasible equilibrium contracts. By placing an upper bound on an agent’s exposure to consequences when a project fails, limited liability may end up encouraging agents to choose excessively risky projects. They showed that this could lead to credit rationing in equilibrium. Banerjee (2003) shows how this could even result in the collapse of the market in some cases. There are three main but not necessarily exclusive tools that are used in practice to overcome the problem of limited liability: using collateral guarantees, legal enforcement strategies and monitoring (Conning and Udry, 2007). The traditional informal scheme of financing for cotton producers in Chaco by input suppliers via in-kind loans (seed, fertilizers) can be interpreted as a form of monitoring (Burkart and Ellingsen 2004; Watts 1994) since these practices make it difficult for borrowers to divert credit for other private uses. The CTF consists of a set of contracts – credit, insurance, warrant and forward – which addresses the limited liability problem by combining the three tools. First, it allowed small cotton producers to generate financial collateral at market prices using the only collateral that these producers had: their uncertain future production. To create the collateral, the producer essentially ceded two rights to the fund: their rights to harvested raw cotton under a contract that specifies the date of delivery and price, and their rights to receive insurance indemnities against climate and other agriculture risks. The limited liability problem in this case is solved by a surety issued by an insurance company to guarantee the delivery of the future production at a given location. Second, the surety insurance also operated as a monitoring mechanism since the insurance company visited small producers and cotton cooperatives throughout the production process. Third, the use of the warrant and the trust contract reduced the costs of legal enforcement since in the case of default, the surety indemnized the fund so that outside investors could recover their investment in the trust fund and the insurance company would become a residual claimant to the cooperative.

International Food and Agribusiness Management Review

240

Alem and Elias

Volume 21, Issue 2, 2018

With the support of this set of contracts, the CTF could obtain a high credit rating from a market rating agency that allowed the CTF to obtain funds from the capital market at competitive market rates. These funds were lent to cooperatives to buy seeds and other crop inputs. Cooperatives paid for the insurance premiums (both climate and surety bond) and on-lend (in-kind) to small cotton producers with a small transaction cost. The private insurer carried out a market due diligence process which was key to avoiding an oversupply of funds that would end up financing projects whose return would be much lower than the market return.

4. Experimental design of the financial product (pilot phase) Despite the good conditions of the processing facilities, a review of the financial statements of cotton cooperatives revealed significant losses and negative net worth in recent years, with only two exceptions. These two cooperatives were selected to undertake the experimental design of a new financial contract to cover the production and distribution risks of cotton fiber by a third party – a specialized insurance company acting as delegated monitor – thus making it possible to inter-link credit with a surety contract, as Townsend (2003) suggested. The structured finance scheme allowed a public investment bank, Banco de Inversión y Comercio Exterior (BICE), to lend to these two cooperatives. In 2010, UCEF, Sancor Seguros (a leading insurance company), and BICE designed an insurance cum credit contract as a new financial product to finance cotton cooperatives to buy raw cotton and pay back with the proceeds of the sale of cotton fiber. An experimental approach was followed to implement the new contract at a pilot scale before launching the financial product to raise market financing. The new financial product consisted of a surety insurance policy to cover cotton fiber delivery risk linked to a standard credit contract and implied the following contractual arrangement: ■■ At the time of harvest, the two cooperatives received a short-term loan (90 days) from BICE to buy raw cotton from member producers. ■■ The cooperatives processed the raw cotton in their ginning equipment and obtained cotton fiber to market at spot prices. ■■ The loan was guaranteed by a surety insurance policy issued by Sancor Seguros to cover the nondelivery of a warrant of cotton fiber in the quantity stipulated by the loan contract. ■■ The surety insurance policy was ceded to BICE as a guarantee for the cancellation of the loan until a warrant was constituted. The surety insurance policy covered all events that could prevent the placement of the warrant of cotton fiber deposited in an authorized storage or warrant company (Control Union). The credit contractual agreement was then structured on the basis that the warrant is accepted as liquid and preferred collateral by banking regulation. The insurer performed a due diligence analysis (delegated monitoring) of the two cooperatives before issuing the surety. By doing so, it covered the risk that the cooperative might not have the amount of export-quality cotton fiber to constitute a warrant. The surety was ceded to the bank as collateral for the loan repayment. The surety insurance was also introduced to work as a monitoring mechanism and control activity, a complementary tool that is used in practice to overcome the problem of limited liability (Conning and Udry, 2007). The surety covered the ‘proper use’ of the money lent to the cooperative and acted as indirect collateral given that the indemnity covered the amount of the bank loan plus interest.

International Food and Agribusiness Management Review

241

Alem and Elias

Volume 21, Issue 2, 2018

The surety intended to cover the cooperative’s management risk and its premium was set at 2% by the insurer, which compared favorably to the standard 6.5% costs of land mortgage contracts. After the disbursement of the loans to the cooperatives, the new financial scheme operated as follows: ■■ The cooperative bought raw cotton from small-scale producing members, which was then processed to obtain export-quality cotton fiber. ■■ The cotton fiber was marketed at spot prices and the proceeds were used to buy more raw cotton to process, allowing the cooperative to roll over the proceeds of the sales several times and increase the margin in return. ■■ Before reaching the 90th day of the loan disbursement, the cooperative: i was to pay back the loan plus interest and the surety claim would not come into effect; ii might not pay back the loan but would deliver a warrant (endorsed to the bank) to the amount stipulated in the insurance contract and the surety would not come into effect; iii might not pay back the loan and the surety would come into effect so that the insurance company would pay the bank the indemnity. ■■ In case of (ii) the bank would liquidate the warrant and recover the amount lent. A significant part of real-world financial innovation can be understood as efforts to find ways to relax incentive constraints so as to reduce the size of the limited liability rents that restrict the range of feasible contracting (Conning and Udry, 2007). Delegated monitoring is one strategy used to decrease information asymmetries aimed directly at lowering the agent’s return from moral hazard (Tirole, 2006). Sancor Seguros acted as delegated monitor as its experts carried out the due diligence and supervision process to assess the two cooperatives’ management and production capabilities to use the proceeds of a loan to buy raw cotton, produce, and sell cotton fiber in the spot markets. The experts for this independent technical review would set the amount of cotton fiber that each cooperative would be able to produce in a 90-day window during the harvest months (typically between March and July) of the 2009-2010 campaign. The loan amount would be limited to 70% of the dollar value that was to be obtained after multiplying the insured tons of cotton fiber by the spot market price of US$0.80 per kilogram. At the harvest season in 2010 the bank approved two loans for approximately US$500,000 to Roque Saénz Peña and San Bernardo cooperatives. The operation was successful and served as an experimental phase before launching an intermediated financial fund to issue short-term debt securities. The fund could now raise finance for crop inputs and sowing, in addition to harvesting and ginning cotton. The fund also included the possibility of incorporating other agricultural products. 4.2 The financial product framework innovation: contract-based collateral The financial product consisted initially of two related contracts: a standard credit contract and an insurance contract. The insurance contract guaranteed the bank the constitution of a warrant for a specific amount of cotton fiber by the end of the loan term. The financial product provided incentives for the cooperatives to maximize cotton fiber production and its sale in the market, thus avoiding moral hazard problems. In turn, small cotton producers could sell their raw cotton to the cooperative at market prices as a way to cancel their obligations, obtain a return and be able to repeat the scheme in the next campaign. Figure 2 shows the key innovation in the financial product. The product consisted of the surety bond issued by a third party (a specialized agricultural insurance company) that covered the management risk of the cooperative by assuring the bank that a minimum amount of export-quality cotton fiber (a commodity with an international price) would be deposited as a warrant in a third-party company. Given that the value of that deposit certificate was considered a liquid asset by banking regulations, the bank did not require additional collateral for the operation.

International Food and Agribusiness Management Review

242

Alem and Elias

Volume 21, Issue 2, 2018

90 days

Warrant t

Warrant t+1

Delivery risk insurance

Figure 2. ‘Warrant-Backed’ loan. 4.3 Implementation challenges The provincial government was a key driver in the implementation. Through UCEF, it acted as a specialized unit transmitting the benefits of the financial contracts (credit cum insurance) to cotton cooperatives as an alternative to the traditional approach of mortgage-backed bank loans. In general, it was important to see the big picture but, as Duflo (2017) points out in ‘The Economist as Plumber’, the details matter a great deal. And because those details matter a great deal, a joint team composed of UCEF, BICE and Sancor Seguros visited the two cooperatives to explain the implications of these contracts and to gain the trust of small producers. A second major obstacle was to assure the efficient functioning of local and export markets for cotton fiber as the financial contract critically depended on the marketing of the fiber to pay off the loan and terminate the surety contract. Any impediment to trade would have a negative effect on the financial contract designed.

5. Financial product design: the cotton trust fund 5.1 A design based on market principles State involvement in banking has been one of the causes of the relatively poorer effective intermediation in developing countries (Adams et al., 1984). The approach to the experimental design was based on the need to recover market principles to allocate production risks to market actors and replace the traditional financing scheme (mortgage-backed bank loans) and ex post government intervention in providing production subsidies and debt and tax relief programs to small producers in cotton cooperatives. The role of the government was reduced to facilitate market function so that cooperatives would be able to sell cotton fiber at spot prices in the future. The financial structure of the trust fund consisted of replicating the two credit operations implemented for the processing and sale of the cotton, but now including the finance of cotton production to small farmers associated with the cooperatives. The fund now extended the term of the loan to cover the entire production cycle (9 months). At this early stage, on-lending to members also included multi-risk insurance to cover climate events. Second-order cooperatives that grouped the commercialization of cotton fiber were constituted as purchaser trustors in the forward sale contract at a specified price. Cooperatives located in the province of Chaco were established as trustor producers who would cede their rights under the forward sale contracts of cotton fiber to the fund’s trustee, BICE. Producer trustors would also cede their rights to the indemnities of climate risk insurance policies granted by Sancor Seguros to the CTF. On the other hand, the federation of cooperatives (UCAL) and a local buyer of cotton fiber (Buyatti) were established as purchaser trustors, who would cede their rights to a surety bond provided by Sancor Seguros that was to cover the cooperatives’ risk of cotton fiber delivery and the constitution of a warrant in the designated depository company. BICE would act as the trustee of the CTF with the mandate of issuing

International Food and Agribusiness Management Review

243

Alem and Elias

Volume 21, Issue 2, 2018

debt securities in the capital market to obtain resources that would be intermediated to cooperatives and small producers to finance the production of cotton that would be harvested in 2011. The forward sales contracts regulated the price, the delivery conditions, and the rights of each of the parties. Under these contracts, producers’ cooperatives undertook the delivery of a specified quantity of tons of cotton fiber of a specified quality on fixed dates. The purchaser trustor was to be instructed by the trustor producers to pay the price to BICE, as CTF’s trustee, as was stipulated in the forward sale contracts. The forward sales contract foresaw that the producer trustor would cancel its obligation of delivery of the cotton fiber by paying a sum of money equivalent to the price agreed in the contract. It was anticipated that if the price of cotton fiber in the market surpassed the price fixed in the forward sales contracts, the producer trustor would be able to use that option and cancel its obligations by making a later cash payment. Figure 3 and 4 show the flows of cotton, cash and contractual arrangements of the CTF and the evolution of the price of cotton during the implementation of the first series (CTF I) with the chronological detail of the operation, respectively.

Cotton market

Merchandise flow

7a Unión Cooperativas AgrícolasAlgodoneras --------------------------------4a Trustors purchasers

Cash flow Other

Sancor Cooperativa de Seguros Limitada --------------------------------Insurer

Trust account

2 Purchase contract 3 Direct buyer

Producers/cooperatives --------------------------------4b Trustors

3 2

Warrant --------------------------------Certificate of deposit

Algodón I – Financial trust

1 Investors 7

1 Cooperativa Sáenz Peña --------------------------------Deposit taker

Control Union Argentina SA --------------------------------Deposit company

Banco de Inversión y Comercio Exterior --------------------------------Issuer/trustee

Figure 3. Financial structure of the cotton trust fund.

International Food and Agribusiness Management Review

244

8 BDO Becher --------------------------------Auditor

Alem and Elias

Volume 21, Issue 2, 2018 3,50 3,30 3,10

Date of issue of VRD and CP

2,90

Deadline for repayment to the trust Execution and auctioning of the warrant (in case of non-repayment)

$/kg

2,70 2,50 2,30 2,10 1,90 1,70

Disbursement to producers´ cooperatives

Due date of VRD to capital market investors

l-1 0 ug -1 0 Se p10 O ct -1 N 0 ov -1 0 D ec -1 0 Ja n11 Fe b11 M ar -1 1 A pr -1 M 1 ay -1 1 Ju n11 Ju l-1 A 1 ug -1 1 Se p11 O ct -1 1

1,50

Figure 4. Evolution of the price of cotton during the cotton trust I, with the chronological detail of the operation (cotton free on board price in $/kg; type D Long 27,0 mm (1 1/16”)), Argentina, July 2010-October 2011. CP = certificate of participation; VRD = Cotton trust fund debt securities (own elaboration using data from Ministerio de Agroindustria de Argentina, 2017). 5.2 Separation and mitigation of risks The CTF transaction was based on the identification of every event affecting production, processing, and sale of cotton fiber in the market and the mechanism designed to mitigate each one independently. The mechanism used market principles to align the participants’ incentives: ■■ Risk of cotton fiber delivery: the risk that the trustor producer cooperative would not deliver the cotton in accordance with the terms of the forward sale contract was mitigated by the surety insurance policy issued by Sancor Seguros (rated B2/A1.ar by Moody´s (2010)). ■■ Multi-risk agricultural insurance policy: in addition to the surety policy, the trustor producer cooperative’s production was covered by a multi-risk insurance policy. The insurance company would pay a specific amount to the CTF’s trustee (BICE) in case of flood, drought, and specific agricultural diseases. ■■ Volatility risk of cotton prices: the transaction was open to the price risk of cotton should the following three events concurrently occur: (1) if the market price of the cotton on the date of delivery of the merchandise were to fall below the price fixed in the sale contract of US$0.97 per kilogram of cotton; (2) if the trustor producer cooperatives were to constitute the warrants and therefore the surety delivered by Sancor Seguros would cease to be relevant; and (3) if the designated trustor purchasers failed to comply with their obligations in the forward sales contract. To eliminate this risk, the historical behavior of cotton prices in the Argentine market was studied and concluded that the transaction supported a reduction in the prevailing cotton price of US$1.67 up to 42%. The maximum historical negative deviation for the period analyzed was 41%. ■■ Counterparty risk of the purchase contract: if any of the trustor purchasers defaulted on their obligations under the forward sales contracts, the CTF’s trustee would take possession of the merchandise deposited and would have to carry out an extrajudicial auction of the cotton fiber. The costs of this operation were included in the transaction costs and a reserve fund was set up for this purpose. International Food and Agribusiness Management Review

245

Alem and Elias

Volume 21, Issue 2, 2018

6. Results of the experimental design The CTF debt securities (VRD) obtained a local A2.ar credit rating of up to US$64 million and had a Certificate of Participation of 5% of said amount retained by trustors producers. These securities would have a variable yield with a margin of 2% over the market reference rate (BADLAR index) with a maximum nominal interest rate of 18% and a minimum of 13%. BICE issued a first tranch (CTF I) of debt securities for a period of nine months for US$8.5 million that was successfully placed in the local capital market. The final cost of the funds for producers was 23% (13% of interest rate of the VRD +2% of the surety insurance policy +7% of the multi-risk insurance policy +1% of the cost of the trust). The resources raised by the trust allowed 11 cooperatives to access finance at prime market rates thanks to the high credit rating of the transaction, which was the same rating as the company Sancor Seguros. Once the trustors (purchasers and producers) had ceded their right to the trust, they became part of contractual arrangements to support the transaction, together with the forward sale contracts. This complex structure of contracts, based on market design principles as in Townsend (2003), was made possible by identifying and separating the risks, as was discussed in Section 5. An important part of these risks was transferred to the private sector, Sancor Seguros, the warrant company and ultimately market investors, with clear comparative advantages to absorb them. Another key aspect was the lack of government subsidies in this operation. All transactions and contracts involved in the CTF implied fees and prices at market levels. This feature implied a radical change in the government’s traditional interventional approach in the sector. The operation of the CTF faced important challenges during its implementation which calls for further studies to replicate the fund in other contexts. One particular challenge was an external event that significantly affected the functioning of the country’s cotton fiber export market. Because of climate events in other cotton-producing regions in the world, the price of fiber in the international market soared to over US$3.50 per kilogram. Thus, given the potential impact of this situation on local prices, the market expected the government to intervene and close the export market. In this context, cotton cooperatives faced difficulties to close sales contracts that posed high risks on the successful completion of the transaction. Second, the need to attract outside investors demanded that strict timing and procedures in the case of default would have to be included in the contract. The advantage of avoiding lengthy court litigation procedures was balanced by the need to collect loan repayments on time to repay investors. In practice, the actual collection of repayments from cooperatives resulted in the need to restructure a few loans as cooperatives delayed repayments to the fund for a few weeks.

7. Conclusions In environments with the peculiarities of the cotton sector in the province of Chaco, conventional credit mechanisms may be affected by the producers’ lack of patrimonial and collateral support, as well as by the high volatility of the price of commodities. This context requires a field research approach to understand and measure diverse risks. The theory of risk allocation and mechanism design provide alternative hedging mechanisms to cope with specific risks that need to be identified before designing contract arrangements. The fund was able to provide Chaco cotton cooperatives and cotton growers – who were unable to access conventional loans – with access to market financing. These actions fostered the formalization of cotton production activities that market actors (insurance and warrant companies, as well as the trustor bank) could monitor. This paper’s main contribution to the literature is the design and implementation of an interlinked loaninsurance contract to overcome information asymmetries and limit liability problems, as Stiglitz and Weiss discuss (1981). The new financial contract presented in Section 4 allowed credit cooperatives to partially overcome limited liability and increase access to market finance by leveraging the credit rating level of the International Food and Agribusiness Management Review

246

Alem and Elias

Volume 21, Issue 2, 2018

insurance provider. The innovation in the contract is simple and consisted of issuing a warrant backed by a surety bond. This innovation allowed a bank to lend to cooperatives despite their weak financial situation. The key result is that the bank delegated monitoring to the insurance company, which then invested resources in visiting cooperatives and screening and monitoring their productive cycle throughout the life of the contract, as was predicted in the mechanism design literature (Townsend, 2003). This device worked as substitute collateral which effectively ameliorated credit constraints to rural farmers. The simple financial contract was later embedded in a financial fund, which consisted of a complex structure of various hedging contractual arrangements to eliminate risks. The fundâ&#x20AC;&#x2122;s implementation faced increased challenges as more cooperatives and new actors (cotton purchasers) were needed to hedge price changes with forward contracts. Despite the challenges, the fund was successfully implemented during a first stage and provided small producers with proper incentives to increase their production. Hence, the most suitable option was to produce and sell in the market. As more farmers increase production, companies providing insurance will also benefit by being able to better diversify risks as described in the theory of optimal allocation (Townsend, 2003). For the investor, the debt securities issued are relatively safe and offer market levels of profitability. The challenge lies in implementing these types of sophisticated ex ante contracts in agricultural environments plagued by informal contractual and financial relationships. The pilot, experimental design approach was successful at introducing small cotton growers to innovative hedging mechanisms employed in modern finance. The support and professional stance of the government unit was also a key factor in the success. The role of the intermediary institution in the mechanism design for risk allocation deserves further research and experimentation, and this paper attempts to contribute to the research agenda presented by Conning and Udry (2007) on the role of financial intermediaries in rural markets. The key innovation in this paper was the interlinked loan-insurance contract, which was used as the basis to design a financial fund. We believe that the role of the fiduciary agent as a new financial intermediary is a promising attempt to overcome information asymmetries in rural markets, as well as facilitating capital market products at low transaction costs. It is reasonable to consider that other private actors can play the role of the trustee. The case may be that private actors will still wait and see if these types of intervention can be replicated before investing in this new technology. This experience showed an innovative form of public-private partnership in which each actor contributes know-how. By so doing, it allows market instruments to channel public policy and actors to increase production that is sustainable and reduces poverty. This financial tool also confirms that proper risk management is one of the pillars of agricultural development policies. This product enables agricultural producers to improve both their production and marketing and increase their incomes.

Disclosure The ideas and opinions expressed in this article are those of the authors and do not necessarily represent the official position of the Inter-American Development Bank.

Acknowledgment We thank three anonymous referees and the editor Jacques Trienekens for very useful comments. This paper is based on a chapter of Alem and Elias (2016) that was awarded by the Latin American Association of Development Financing Institutions (ALIDE) with the Rommel Acevedo Latin American Award on Development Banking in 2016. We thank Lucia Alem and Laura Fernandez for assistance in the proofreading of the article.

International Food and Agribusiness Management Review

247

Alem and Elias

Volume 21, Issue 2, 2018

References Adams, D.W., H.G. Douglas and J.D. Von Pischke. 1984. Undermining rural development with cheap credit. In: Undermining rural development with cheap credit, edited by D.W. Adams, H.G. Douglas and J.D. Von Pischke. Westview special studies in social, political and economic development series, Boulder, CO, USA, pp. 1-7. Aleem, I. 1990. Imperfect information, screening, and the costs of informal lending: a study of a rural credit market in Pakistan. World Bank Economic Review 4(3): 329-349. Alem, M and J.J. Elias. 2016. Vigencia y Futuro de la Banca de Desarrollo: Un Enfoque desde la Asignación de Riesgos y Recursos. Unpublished Manuscript. [Available upon request with corresponding author]. Banerjee, A.V. 2003. Contracting constraints, credit markets, and economic development. In: Advances in economics and econometrics: theory and applications, edited by M. Dewatripont, L.P. Hansen and S. Turnovsky. Eighth World Congress, Volume III, Cambridge University Press, Cambridge, UK. Burkart, M. and T. Ellingsen. 2004. In-kind finance: a theory of trade credit. American Economic Review 94(3): 569-590. Conning, J. and C. Udry. 2007. Rural financial markets in developing countries. Ch. 56. In: Handbook of agricultural economics, edited by R. Evenson and P. Pingali. Vol. 3. Elsevier, New York, NY, USA, pp. 2857-2908. Coulter, J and G. Onumah. 2002. The role of warehouse receipt systems in enhanced commodity marketing and rural livelihoods in Africa. Food Policy. 27: 319-337. Duflo, E. 2017. The Economist as plumber. American Economic Review 107(5): 1-26. Guirkinger, C. and S.R. Boucher. 2008. Credit constraints and productivity in Peruvian agriculture. Agricultural Economics 39: 295-308. Ministerio de Agroindustria de Argentina. 2017. Subsecretaria de mercados agropecuarios. cotizaciones semanales de la cámara algodonera: infomación histórica. Available at: http://tinyurl.com/y8fz9uqt. Ministerio de Agricultura Ganaderia y Pesca de la Nacion (MAGPN). 2010. Informe Argentina: sobre situacion actual y perspectivas de la industria alodonera. 69 Reunion del Comite Consultivo Internacional del Algodon, Lubock, Texas, USA. Available at: http://tinyurl.com/ybx7ybnz. Moody’s. 2010. Informe de calificacion – fideicomiso financiero algodón I. Moody’s, September 2, Buenos Aires, Argentina. Phelan, C. 1994. Incentives and aggregate shocks. Review of Economic Studies 61: 681-700. Prescott, E.S. 2003. Communication in private-information models: theory and computation. The Geneva Papers on Risk and Insurance Theory 28(2): 105-30. Stiglitz, J.E. and A. Weiss. 1981. Credit rationing in markets with imperfect information. American Economic Review 71(3): 393-410. Tirole, J. 2006. The theory of corporate finance. Princenton University Press, Princeton, NJ, USA. Townsend, R.M. 2003. Microcredit and mechanism design. Journal of the European Economic Association 1(2-3): 468-477. Towo, N.N. and P.J. Kimaro. 2014. Warehouse receipt system: a solution towards smallholder farmers´ financial constraints? International Journal of Economics, Commerce and Management 2(7). Available at: http://ijecm.co.uk/wp-content/uploads/2014/07/2719.pdf. Watts, M.J. 1994. Life under contract: contract farming, agrarian restructuring, and flexible accumulation. In: Living under contract: contract farming and agrarian transformation in sub Saharan Africa, edited by P.D. Little and M.J. Watts. University of Wisconsin Press, Madison, WI, USA, pp. 21-77.

International Food and Agribusiness Management Review

248

OPEN ACCESS International Food and Agribusiness Management Review Volume 21 Issue 2, 2018; DOI: 10.22434/IFAMR2017.0063 Received: 18 August 2017 / Accepted: 7 September 2017

Governance of market-oriented fresh food value chains: export chains from New Zealand Special issue: IFAMA 2017 symposium

RESEARCH ARTICLE Jacques Trienekens a, Mariska van Velzenb, Nic Leesc, Caroline Saundersd, and Stefano Pascuccie aProfessor

and bResearcher, Wageningen University and Research, Hollandseweg 1, 6706 KN Wageningen, the Netherlands

cSenior

lecturer and dProfessor and Director, Lincoln University, P.O. Box 85084, Lincoln 7647, New Zealand

eProfessor,

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

Abstract The competition in international food markets is increasingly moving towards products with higher levels of added value and higher degrees of differentiation, requiring companies to become more market-oriented. Market orientation is ‘the extent to which an actor in the marketplace uses knowledge about the market, especially about customers, as a basis for decision-making on what to produce, how to produce it, and how to market it’. Market orientation comprises three constructs: market intelligence generation, dissemination and responsiveness. Value chain governance can facilitate market orientation requirements. Value chain governance includes network governance, contracting and informal relationships. Knowledge about how governance can facilitate a value chain’s market orientation is limited. Therefore, the aim of this study is to explore how the governance of a global food value chain can facilitate the value chain’s market orientation. The study applies a multiple case study design. Four in-depth case studies were conducted on global food value chains from New Zealand to Western Europe dealing with the products apples, kiwis, venison and lamb. Interviews were conducted with actors from these four value chains in the Netherlands as well as in New Zealand. In each value chain actors with similar functions were interviewed in order to make the results comparable. Analysis of the case studies shows that network governance (i.e. leadership, shared governance and facilitation), contractual agreements (i.e. type and content: price, volume, quality) and informal relationships (i.e. trust and commitment) can contribute to the market orientation of a value chain. Leaderships and shared governance, in combination with good informal relationships in the chain, as well as contractual incentives, are main contributors to market orientation in global fresh food value chains. The paper adds to the still very scarce literature on governance of value chains and market orientation of value chains. Keywords: market orientation, value chain governance, fresh global food chains JEL code: Q13 Corresponding author: jacques.trienekens@wur.nl © 2017 Trienekens et al.

249

Trienekens et al.

Volume 21, Issue 2, 2018

1. Introduction The competition in international food markets is increasingly moving toward products with higher levels of added value and higher degrees of differentiation, requiring companies to become more market-oriented (Grunert et al., 1996). A market-oriented company aims to continuously deliver superior value to its customers (Han et al., 1998) ‘by identifying and satisfying customer needs more effectively than its competitors’ (Kirca et al., 2005). In this way, the company tries to create competitive advantage (Grunert et al., 2005). Market requirements, i.e. requirements for the final customer, should be coordinated in the chain by a suitable governance structure (Elg, 2008; Grunert et al., 2005; Trienekens and Wognum, 2013). Governance mechanisms may include formal arrangements (contracts: Wever, 2012) and informal arrangements such as trust and reputation (Lu, 2012; Trienekens et al., 2012). The study of these mechanisms has so far focused primarily on the dyad. In studying value chains as entities we need to extend this view to mechanisms that cover the chain end-to-end. The objective of the paper is to explore the relationships between market orientation and governance of international fresh food chains, based on an in-depth analysis of four fresh food export chains from New Zealand to Europe. The paper adds to the still very scarce literature on market orientation of value chains and (related) value chain governance.

2. Market orientation and governance 2.1 Market orientation Two main definitions of market orientation can be distinguished: namely, one from a behavioural perspective and one from a cultural perspective (Homburg and Pflesser, 2000; Kirca et al., 2005; Vieira, 2010). Kohli and Jaworski (1990) defined market orientation in terms of organisational behaviour and stated that market orientation comprises three types of activities, namely market intelligence generation, dissemination and responsiveness. Narver and Slater (1990) defined market orientation in terms of organisational culture and stated that market orientation comprises customer orientation, competitor orientation and inter-functional coordination. These are organisational norms and values which encourage behaviour that is in line with market orientation (Kirca et al., 2005). Recently, there has been an increasing interest in market orientation from a value chain perspective (Elg, 2008; Grunert et al., 2005, 2010). This interest is relevant, since the degree of market orientation of one company in the value chain can be influenced by the degree of market orientation of other actors in the value chain (Grunert et al., 2005). Additionally, in the global business environment, competition is transiting from company level to value chain level (Green Jr et al., 2006; Jain and Benyoucef, 2008). Therefore, in order to offer superior value to the market, companies need to collaborate with other value chain actors and develop an inter-firm strategy for market-orientation (Elg, 2008). Grunert et al. (2005) used the definition of Kohli and Jaworski (1990) to define market orientation from a value chain perspective as ‘chain members’ generation of intelligence pertaining to current and future end-user needs, dissemination of this intelligence across chain members, and chain wide responsiveness to it’. In this definition, intelligence generation refers to all activities conducted by value chain members aimed at gathering information on end-users, typically consumers, customers and competitors (Grunert et al., 2005). This gathered information is also called market information. Dissemination or communication of intelligence refers to all exchanges of information on consumers, customers and competitors between value chain members (Grunert et al., 2005). Responsiveness includes the actions of value chain members in response to the gathered market information to create superior value for consumers and customers (Grunert et al., 2005). Effective communication or dissemination is a prerequisite for responsiveness in the chain,

International Food and Agribusiness Management Review

250

Trienekens et al.

Volume 21, Issue 2, 2018

since responsiveness is only possible if market information is effectively shared in the value chain (Kähkönen and Tenkanen, 2010). 2.2 Value chain governance Value chain governance may facilitate market orientation (Elg, 2008; Grunert et al., 2005). However, knowledge about this subject is limited (Elg, 2008; Karami et al., 2015). So far most of the literature on the governance of inter-business relationships has focused on bilateral buyer-supplier relationships (Menard, 2004; Wever, 2012; Williamson, 1991). One stream in the value chain literature that goes beyond the dyad is the global value chain approach, focusing on the governance of end-to-end value chains (Dolan and Humphrey, 2004; Gereffi et al., 2005; Gibbon et al., 2008; Kaplinsky, 2000). Gereffi et al. (2005) defined the complexity of information and knowledge transfer required to sustain a particular transaction as explanatory factors for governance choices in global value chains, thereby underlining the important role of lead parties and power asymmetry in these chains. If we consider the current practice, power normally lies with large organisations downstream of the value chain, such as retail or marketing and distribution organisations, or with large manufacturing firms. Suppliers, such as farmers, rarely have a powerful position in value chains (Ponte and Sturgeon, 2014). Lead firms are the main decision takers as regards requirements on quality, quantity and timing of delivery and pricing. Power can result from different sources, distinguishing between sources of coercive and non-coercive power (Hunt and Nevin, 1974; Leonidou et al., 2008). The main difference between the two is that for sources of non-coercive power, individuals willingly yield power to another individual (non-aggressive), while for coercive power there are potential punishments which makes individuals yield power to another individual (aggressive) (Hunt and Nevin, 1974; Leonidou et al., 2008). However, the literature on the governance of end-to-end value chains is scarce. Therefore, and as we consider value chains as netchains (Lazzarini, 2001), including horizontal and vertical relationships covering the chain end-to-end, we include network governance in our theoretical approach. Network governance can be defined as ‘the set of mechanisms that supports and sustains cooperation among participating organisations to enhance the likelihood of achieving network-level goals’ (Alvarez et al., 2010). Networks focus on improving collaboration between organisations in the network, which can lead to better alignment of, among other things, supply and demand in the value chain (Barratt, 2004). Therefore, it is likely that network governance can facilitate market orientation of a value chain. Provan and Kenis (2008) explored different forms of network governance and propose that these can be categorised along two dimensions. First, network governance may or may not be brokered. At one extreme, in case of highly brokered network governance, there are few direct organisation-to-organisation interactions. In this case a single organisation is responsible for the governance of the network and thereby takes a leadership position (Provan and Kenis, 2008). At the other extreme, within lowly brokered network governance, the network can be governed by all the organisations within the network. In this case, every organisation interacts with all other organisations in order to govern the network (Provan and Kenis, 2008). The second dimension along which network governance forms can be categorised is whether the network is externally governed or participant governed. In the case of externally governed networks, an organisation that is not involved in the network activities is responsible for the governance of the network. Within the dimension of participantgoverned networks, the network might be governed by a single (lead) organisation or by all organisations that comprise the network (shared governance) (Provan and Kenis, 2008). The global value chain literature underlines the importance of lead parties, information exchange and supplier capabilities in the explanation of value chain governance structures. In the network literature; lead parties, shared governance (and information exchange) and network facilitation are key elements to explain network governance. For our current research we select three network governance constructs, namely leadership, shared governance and facilitation. Leadership reflects one actor in the value chain, being the leader or ‘channel captain’. There is a power asymmetry as the actor taking the leadership role has more power than International Food and Agribusiness Management Review

251

Trienekens et al.

Volume 21, Issue 2, 2018

the other chain actors. The leader is responsible for coordinating major network-level activities and key decisions (Gereffi et al., 2005; Provan and Kenis, 2008). Shared governance reflects actors being responsible for governing (part of) the network together (Provan and Kenis, 2008) and communication and information exchange between these actors (Gereffi et al., 2005). In this paper, facilitation reflects the provision of actors in the value chain with all kinds of support, such as access to particular information, assistance or provision of infrastructure (Leonidou et al., 2008). The need for facilitation also reflects the capabilities of the chain actors (Gereffi et al., 2005). ■■ Contracting Different types of contracts can be distinguished, such as spot-market contracts, verbal agreements, formal contracts, equity-based contracts and vertical integration (Raynaud et al., 2005; Wever et al., 2010). Transaction Cost Economics (Williamson, 1991) has often been applied to study bilateral (i.e. buyer-supplier) governance mechanisms (Grunert et al., 2011; Ménard, 2004; Raynaud et al., 2009). In spot markets, coordination takes place via price and competition (Martinez and Zering, 2004; Williamson, 1991). At the other end of the continuum, hierarchies reflect vertical integration of two companies, in which there is coordination via administrative control with associated monitoring rights (Raynaud et al., 2005; Wever et al., 2010; Williamson, 1991). Hybrids, such as partnerships and alliances, are in between these two extremes (Ménard, 2004). The degree of coordination or control increases from spot markets to hierarchies (Trienekens and Wognum, 2013). Market orientation requires intelligence generation, intelligence dissemination and responsiveness of the actors in the value chain (Grunert et al., 2005). Therefore, certain contract attributes may be more suitable for facilitating market orientation than others, for example contracts that enable firms to flexibly respond to market requirement, i.e. contracts that can be responsive in price, quality and/or delivery requirements. Specific investments, for example, in company shares, can also be explanatory major reason for the choice of governance structures (Williamson, 1991; Wever et al., 2012); in many cases these often lead to more integrated structures. In addition, behavioural and demand/supply uncertainty are transaction attributes impacting on the governance choice (Williamson, 1991; Wever et al., 2012). Behavioural uncertainty (e.g. opportunistic behaviour of actors) can be decreased through stricter coordination of transactions, i.e. through more hybrid-type or hierarchical types of bilateral agreements (i.e. formal written contracts, equity-based contracts, vertical integration) (Wever, 2012). In contrast, demand or supply uncertainty would require more flexibility and less strict contracts. ■■ Informal coordination mechanisms Besides formal relationships (e.g. through contracts) there are also informal relationships between actors in a value chain (Alvarez et al., 2010; Lu, 2012; Trienekens et al., 2012). Trust and commitment can be identified as relevant aspects influencing the degree of market orientation. Informal relationships, including trust and commitment, can be beneficial for the exchange of information and thereby facilitate a value chain’s market orientation (Grunert et al., 2010; Jeong and Hong, 2007). Trust can be defined as ‘the degree to which a firm believes that its trading partner is honest and/or benevolent’ (Bigne and Blesa, 2003). Grunert et al. (2010) studied the market orientation of decision makers in two cross-border value chains, namely Norwegian salmon exported to Japan and Danish pork exported to Japan. The research showed that trust creates openness and thereby enhances information exchange between actors in the value chain, contributing to intelligence communication (Grunert et al., 2010). Micheels and Gow (2011) studied the relationship between market orientation and performance by conducting surveys of beef producers in the United States. It was found that trust between actors in the value chain directly contributed to market orientation and indirectly contributed to the performance of a company. More specifically, trust contributed to the exchange of market information (Micheels and Gow, 2011). Commitment can be defined as ‘an exchange partner believing that an ongoing relationship with another is so important as to warrant maximum efforts at maintaining it; that is, the committed party believes the International Food and Agribusiness Management Review

252

Trienekens et al.

Volume 21, Issue 2, 2018

relationship endures indefinitely’ (Morgan and Hunt, 1994). In addition to trust, commitment can be an important antecedent of the degree of market orientation of a value chain, since commitment can stimulate effective market information exchange in a value chain (Grunert et al., 2005; Micheels and Gow, 2011). The reason for this is that commitment creates openness in the value chain (Grunert et al., 2010). Kähkönen and Tenkanen (2010) studied the impact of power on information sharing in the Finnish food industry. It was concluded that a lack of commitment or one-sided commitment decreases the willingness to share information in the value chain (Kähkönen and Tenkanen, 2010). Kwon and Suh (2005) studied the relationships between trust, commitment and various aspects of value chain management. It was observed that commitment is needed between value chain actors in order to achieve common goals (Kwon and Suh, 2005); in this case, a particular degree of market orientation. Trust and commitment are interrelated, since trust contributes to commitment (Karami et al., 2015; Kwon and Suh, 2005; Micheels and Gow, 2011; Ryu et al., 2009).

3. Theoretical framework In our research we measured market orientation through three constructs based on (Grunert et al., 2005; Kohli and Jaworski, 1990): market intelligence, market responsiveness and market communication. Three basic elements of value chain governance are distinguished: (bilateral) contracts throughout the chain, informal coordination mechanisms and network governance. Based on Wever (2012) we define three components of contracts: price, volume and quality. Furthermore, we include trust and commitment as informal coordination mechanisms (Bigne and Blesa, 2003; Karami et al., 2015; Kwon and Su, 2005; Micheels and Gow, 2011). Network governance includes leadership, shared governance and facilitation. This brought us to the following conceptual model of our research. Figure 1 shows the main concepts and the main relationships to be expected based on theoretical considerations. We expect that network governance may impact on the collection of market information (market intelligence), market communication (upstream of the value chain) and market responsiveness (throughout the chain). We expect that network governance (all three components) may have a (positive) effect on trust between the value chain actors and commitment of these actors. Trust and commitment, for their part, will have a positive effect on the communication of market information throughout the chain. Facilitation (one of the elements of network governance) may impact on the type of arrangements and support arrangements on price, quality and volume (e.g. storage of products). Contractual arrangements will affect the responsiveness to the market of actors throughout the chain.

4. Methodology The research is exploratory, as very little research has been done so far on the relationship between market orientation and value chain governance. We analyse four export chains from New Zealand to Western Europe on market orientation, value chain governance and the relationship between these concepts. Three selection criteria were established for the choice of case studies. First, the value chain needs to be global, i.e. the production takes place in a different country from where the produced products are being sold. In this way the research addresses the increasing globalization of food value chains (Ren et al., 2014). Second, the selected value chains should probably be governed in different ways, in order to create a meaningful comparison between the different value chains and to more easily observe contrasting patterns in the data (Eisenhardt and Graebner, 2007). Third, for the same reason as above, the selected value chains should probably have different degrees of international market penetration. We investigated two fruit chains and two meat chains, which all have primary production stages in New Zealand and marketing and distribution stages in North Western Europe. For confidentiality reasons the chains are named A, B (two fruit chains) and C, D (two meat chains). Chain A produces and exports kiwi. It consists of around 2,500 producers, 50 post-harvest operators (pack houses and cool store facilities), the global marketing and distribution organisation, and distribution companies International Food and Agribusiness Management Review

253

Trienekens et al.

Volume 21, Issue 2, 2018 Contracts (bilateral) Type of arrangement Price Volume Quality

Network governance

Facilitation

Market responsiveness

Leadership

Market intelligence

Shared governance

Market communication Market orientation

- Trust - Commitment Informal coordination

Figure 1. Conceptual model. around the world. The marketing and distribution company is owned by producers and facilitated by legislation, as kiwi producers in New Zealand may only export through this company (with the exception of Australia). The apple chain B consists of independent or contract growers as well as orchards owned by a large packing and distribution company that sells and distributes apples around the world. Growers supply apples to 37 post-harvest operators (pack houses and cool storage facilities). Some of these pack houses are also owned by the marketing and distribution organisation. The marketing and distribution organisation is an independent company and a clear leader in this chain. The organisation also owns the plant variety rights of two internationally well-known apple varieties. These varieties can only be exported through the marketing and distribution organisation. Other varieties can also be exported through other companies. The venison chain C that we studied consists of 21 deer farmers (both breeders and finishers) which form a venison producer group aiming to produce high-quality and sustainable venison for international markets. The farmers supply their deer to one independent slaughterhouse. A facilitating company organises processing, logistics and marketing (including export). The farmers and this company each own 50% of the value chain company that again hires the facilitating company. Distribution and sales in Western Europe is done by a high-quality food wholesaler. The lamb chain D consists of around 25 sheep farmers, including breeders and finishers, who together form an association. The farmers collaborate with a veterinarian, who is one of the initiators of the project. The farmers aim for sustainable production in terms of animal welfare, social responsibility and environmental responsibility. The practices must lead to a consistent and high-quality meat brand for the national as well as the international market. The farmers supply their lambs to a large New Zealand meat company, which is responsible for processing and the export to the Netherlands. Distribution in the Netherlands is performed by a Dutch importer and distribution company. This company tries to distribute the products to high-end butcher shops and wholesalers, supplying high-end restaurants, at premium prices. International Food and Agribusiness Management Review

254

Trienekens et al.

Volume 21, Issue 2, 2018

Table 1. List of interviewees. Apple value chain 1. Manager of orchard owned by distribution organisation 2. CEO of orchards, pack house and cool store facilities recently acquired by distribution organisation 3. Value chain manager distribution organisation New Zealand 4. Marketing manager distribution organisation Benelux Kiwi value chain 1. Kiwi grower 2. Director of pack house and cool store 3. Value chain manager distribution organisation New Zealand 4. Marketing manager distribution organisation New Zealand Venison value chain 1. Deer farmer 2. CEO venison exporter and chain coordinator 3. Marketing manager Deer Industry New Zealand 4. Manager of game meat at wholesaler in the Netherlands Lamb value chain 1. Sheep farmer 2. Marketing manager lamb processor and exporter 3. Importer and distributor in the Netherlands

Table 2. Operationalisation of market orientation concepts. Market orientation concepts

Indicators

Literature

Intelligence generation

Market research on customer and consumer wishes and needs Measurement of customer and consumer satisfaction Meetings with customers and consumers Meetings with industry partners Research on competitors’ actions Frequency of performing these activities Wishes of customers and consumers Customer and consumer satisfaction Trends in consumer wishes Competitors’ actions Complaints Sales volumes Frequency of sharing and receiving information New product development driven by market information Periodic review of whether products are in line with consumers’ wishes Adapting products to consumers’ or customers’ wishes and complaints Adapting products or prices in response to competitors’ actions Timing of response

(Bigne and Blesa, 2003; Deshpandé and Farley, 1998; Deshpandé et al., 1993; Grunert et al., 2005; Jaworski and Kohli, 1993; Matsuno et al., 2005; Narver and Slater, 1990; Tomášková, 2009)

Intelligence communication

Responsiveness

(Bigne and Blesa, 2003; Grunert et al., 2005; Jaworski and Kohli, 1993; Matsuno et al., 2005)

(Bigne and Blesa, 2003; Deshpandé et al., 1993; Grunert et al., 2005; Jaworski and Kohli, 1993; Matsuno et al., 2005; Tomášková, 2009)

International Food and Agribusiness Management Review

255

Trienekens et al.

Volume 21, Issue 2, 2018

The concepts in Figure 1 have been applied in a questionnaire that has been used for in-depth interviews along all four value chains, in New Zealand as well as in Western Europe. Next to the analysis of ample secondary material on the four sectors, 20 interviews were performed along these four chains. Interviews were in-depth and lasted between 1.5 and 3.5 hours. The aim was to get precise information about governance mechanisms and market orientation from different perspectives (different functions) along the chains. In Table 1, an overview is provided of the interviewees. The interviewees were selected based on their role in the value chain and their ability to oversee (different stages of) the value chain. The interviews were performed in the last quarter of 2015 (September-November). For the lamb value chain only three interviews were conducted, since for this value chain extensive information was already available from Van Vlerken (2015). Table 3. Operationalisation of value chain governance concepts. Value chain governance concepts

Indicators

Literature

Network governance Leadership

Decision taking Initiate decision taking Organisation of meetings Shared governance Meetings between several members Main agenda of meetings Frequency of meetings Facilitator Facilitation with knowledge, resources, financial rewards Contracts Type of agreement Spot-market contract Verbal agreement Formal written contract Equity-based contract Vertical integration Price agreements Specifications of agreement Which actor sets price? Duration of price agreements Criteria used for price setting Circumstances under which agreement can be changed Volume agreements Specifications of agreement Duration of volume agreements Circumstances under which agreement can be changed Quality agreements Requests from customers to change quality Response to requests to change quality Informal relationships Trust Role of trust Confidence about accuracy and correctness of information Compliance with agreements Other actors take your interests into account when taking decisions Commitment Role of commitment Dedication of time, effort and resources to support others Willingness to make long-term investments when other actors wish to

International Food and Agribusiness Management Review

256

(Provan and Kenis, 2008)

(Leonidou et al., 2008) (Raynaud et al., 2005; Wever, 2012)

(Wever, 2012)

(Kwon and Suh, 2005; Micheels and Gow, 2011)

Trienekens et al.

Volume 21, Issue 2, 2018

Interviewees were contacted via e-mail or by telephone to invite them to participate and to give them an explanation of the research. If the interviewee was willing to cooperate, an appointment for the interview was planned. Before the actual interview was conducted the questionnaire was sent to the interviewee for the purposes of preparation. Interviews were semi-structured, which means that in addition to questions prepared beforehand, there was room for additional questions based on answers from the interviewee. At the start of each interview, permission was requested for recording. Additionally, a concise description of the research was provided and an indication was given about topics to be discussed during the interview. The topics to be discussed in the interviews were derived from the theoretical framework described in section 3 of this paper. Tables 2 and 3 show operationalisation of the concepts from the theoretical model. Indicators and related questions were based on the literature as much as possible.

5. Results 5.1 Market orientation ■■ Intelligence generation In the kiwi chain (A) the marketing and distribution company applies intensive intelligence generation: different types of research, namely consumer usage and consumer attitudes research, annual tracking of markets (consumption trends for kiwis, consumer brand awareness, loyalty and engagement), pre-campaign testing (consumer perception and purchase intention), post-campaign analysis (consumer recap of campaign and buying behaviour), conduction of category information (sales volumes), in-store sampling (consumer liking, purchase intention and comparison to competitors), sensory testing with consumers (at the launch of a new cultivar), monitoring quality and price of kiwis of competitors (firmness and dry matter), research into specific areas (convenience, taste and brand), feedback from customers about markets, performance of kiwis, complaints, and interactions with consumers via social media. In the apple chain (B) the marketing and distribution organisation conducts market research approximately once every five years, when a new variety is developed (sensory research, product name, etc.). Furthermore, there are regular meetings between customers and a regional manager in Europe, to discuss market developments and customer satisfaction. At events and fairs consumers can taste apples. Venison chain C conducted market research with restaurant chefs in Europe before the launch of their products. They have also appointed ambassador chefs who provide market information (quality of venison, positioning, consumer satisfaction) and encourage consumers to give feedback. Venison workshops for chefs and promotions in restaurants are also organised. The industry body, which supports chain C, gathers full sales data from the distribution company in the Netherlands. Moreover, research is conducted (by the industry body) on customer preferences and restaurant chefs’ and consumer usage and attitudes towards venison. In addition, consumer comments are tracked in the press and social media. Feedback from customers about satisfaction, complaints, trends in the market and competitor prices are also researched. Interactions with consumers via social media have so far had less success. The lamb chain (D) has not performed market research on opportunities in the Netherlands and customer required product specifications before the launch of the meat product. Feedback from customers on product quality comes from the importer. There is, however, no feedback from customers’ customers (i.e. restaurants). Prices of other New Zealand lamb products are monitored. Limited research is conducted on customer satisfaction with chilled lamb. In addition, market research is done by a marketing consultant on customer and consumer opinions about the lamb production. So far interactions with consumers via social media have not been successful. International Food and Agribusiness Management Review

257

Trienekens et al.

Volume 21, Issue 2, 2018

■■ Intelligence communication The marketing and distribution company of the kiwi chain (A) receives information from customers and their European subsidiary: outcomes of market research, complaints, feedback on performance, prices and quality of competitors. It communicates this information to the breeding company, growers and post-harvest operators via industry consultation groups (growers, post-harvest operators, marketing organisation), company website, online seminars, magazines, and during annual tours (‘road shows’) and meetings. The European subsidiary of the apple chain (B) updates the head office in New Zealand about the market situation on a weekly basis, via e-mail, on prices and stock levels. Twice a year both parties meet in New Zealand. Growers are kept up-to-date on situations in the market via e-mail and via the company’s supply representatives, who regularly visit the growers to advise them and receive crop estimates. Occasionally customers visit growers and pack houses, where market information is discussed. The chain facilitator of the venison chain (C) receives sales data and restaurant feedback from the industry body and market information from the European distribution organisation (such as sales of particular cuts). Market information is shared during meetings, visits by stakeholders (farmers, facilitator and customers) and via daily e-mail contact with customers. The facilitator shares information with the slaughterhouse when there is a problem or change in product specifications. Market information with farmers is communicated through an industry magazine. In the lamb chain (D) the importer shares feedback from customers (e.g. on product specifications, such as cuts, size, labelling), prices of competitors, market development and possible aspects for improvement with the processing company via e-mail (and in some cases, also with farmers). The processing company shares relevant market information with farmers via e-mail or telephone and occasionally in face-to-face meetings. Within the group of farmers, information is also shared via an online forum. Farmers and the processing company would like to receive more market information. ■■ Responsiveness In the kiwi chain (A) a lot of attention is paid to breeding of new kiwi varieties, in line with consumer wishes. Setting standards (e.g. size, dry matter, firmness) and incentive systems for growers and post-harvest operators is also based on consumer wishes and competitor actions. The establishment of plans for optimal ripening takes place for storage in New Zealand, transportation and in the market, in response to customer feedback on the firmness of kiwis. The company responds to wishes for year-round supply by having growers in the Northern and Southern hemisphere. In the apple chain (B) price changes or new competition are always discussed with the head office before action is undertaken (e.g. lower prices). Responsiveness to consumer requirements takes place through the development of new apple varieties (which take at least three years). Moreover, the company tries to have orchards/contracts in the Northern as well as the Southern hemisphere in order to guarantee year-round supply. The industry body for the venison chain (C) supported the idea of telling the product story in a trial in 2016 based on restaurant feedback. In addition, the industry body is developing a national QA programme in response to customer feedback. Changes in packaging and storytelling took place after a consumer survey in England. Change in product specifications and volumes are based on customer feedback. There is a response to complaints, and development of cuts takes place in collaboration with customers. Supply and demand are matched in terms of volumes. Prices might be adapted in response to competitor actions. In the lamb chain (D) the aim is to base changes in product specifications on customer feedback and complaints. There was a change from frozen to chilled lamb when New Zealand restaurants were satisfied with the quality. However, some Dutch customers were not satisfied with this switch. Although there are International Food and Agribusiness Management Review

258

Trienekens et al.

Volume 21, Issue 2, 2018

customer requests for a larger variety of cuts, the number of cuts has been reduced since it was too costly. Supply does not always meet demand. Farmers are not able to respond to customer demand for year-round supply. Lamb prices can be adapted in response to competitor prices or customer feedback. Table 4 gives an overview of the strength of market orientation in the four cases. The kiwi marketing and distribution organisation is a well-established global company that is able to generate extensive market intelligence and communicate the market intelligence throughout the chain, where this chain includes the farmer and packhouse stage in the interaction. In particular, the kiwi chain has developed many activities to increase market responsiveness. The apple chain does most market research in relation to the launch of new varieties and also pays less attention to chain wide communication and responsiveness compared to the kiwi chain. The venison chain has a close collaboration with restaurant chefs in the Western European market (including the appointment of ‘ambassador chefs’) to obtain information on the most valued cuts. The lamb chain more or less ‘ends’ with the importer in the European market, where the importer hasn’t yet built close contacts with parties further downstream of the chain. Table 4. Market orientation in 4 value chains.1 Case A: kiwi Case B: apple Case C: venison Case D: lamb 1

Market intelligence

Intelligence communication

Responsiveness

++ ++ ++ +/-

++ + ++ +/-

++ + + -

++: strong attention; +: moderate attention; +/-: limited attention; -: lack of attention.

5.2 Value chain governance ■■ Contracts Kiwi value chain (A) has formal written contracts throughout the chain. There is also vertical integration in this chain as growers are shareholders of post-harvest operators and of the marketing and distribution organisation. Volumes of growers are specified in supply agreements. Allocation plans are made to match demand and supply in terms of volume, size and pack type. Based on these allocation plans, the company allocates pack types and delivery schedules to grower entities and post-harvest operators. Prices are based on the pool of market returns for growers. Payments are also based on size and taste/dry matter. There are incentives for growers for early harvest and long cold storage. Post-harvest operators have incentives in the packing process. In the apple value chain (B), there are formal written contracts throughout the chain. Agreements on volumes of growers are based on production forecasts. Prices are based on a pool of returns for growers. Sometimes premiums are payed to growers that aim at markets with tougher requirements (taste, size). With customers there are no price agreements; market developments and demand decide the price. To grow the apple varieties owned by the marketing and distribution organisation, growers have to pay royalties. The company uses several incentives such as a penalty system for lower quality products and a premium for long-term storage at growers (to meet demand fluctuations). Post-harvest operators have incentives in the final packaging process (timing of packaging). Venison value chain (C) is organised through verbal agreements and equity-based contracts between farmers and the facilitating company. There are formal written contracts between the slaughterhouse and the facilitator (on behalf of farmers). Verbal agreements govern the relationship between the facilitator and the high-end distributer in Europe. Volumes of farmers are specified in a supply agreement. During the season, International Food and Agribusiness Management Review

259

Trienekens et al.

Volume 21, Issue 2, 2018

the facilitator checks with customers about sales of venison and with farmers about the supply of deer. In the event of supply shortages other farmers make up the shortfall, as they are committed to helping each other. The farmer base price is based on the returns in the market. Fluctuations in farmer base prices can take place during the year to influence the supply of deer by farmers. Lamb value chain (D) is characterised by verbal agreements. Volumes of farmers are specified in a supply agreement. There is a fixed volume agreement between the processing company and the importer/distributor in Belgium (1000 lambs per week). Farmers have difficulties living up to these volume agreements. The processor tries to push the farmers for a consistent supply. However, there are no penalties when farmers donâ&#x20AC;&#x2122;t comply. Farmers get the market price and a premium per kilogram of live weight of lamb from the processor. The processor aims to have a fixed price for the importer, who aims to set monthly prices with its customers. Shareholding of farmers is required in order to be part of this lamb chain. Table 5 depicts the contract elements of the four cases. The most vertically integrated chain is the kiwi value chain (A), followed by the venison chain (C) and the apple chain (B). Case D is the least integrated value chain. Table 5. Type of contract in the four cases. Case A: kiwi Case B: apple

Type of agreement

Price

Volume

Quality

Written contracts + equity based Written contracts

Based on pool of returns Based on pool of returns Price/kg based on returns in market Market price

Based on supply agreement Based on production forecasts Based on supply agreement Based on supply agreement

Based on size, taste/dry matter Related to market, taste/ size According to quality standard According to quality standard

Case C: venison Verbal and written contracts + equity based Case D: lamb Verbal and written contracts

â&#x2013; â&#x2013; Network governance In the kiwi chain the market and distribution organisation is the lead company, which is responsible for intelligence generation activities, communication of market intelligence, facilitating, setting and checking of production standards and incentive systems and management of the product flow. Its basic source of power is that it is the single point of entry for New Zealand producers to international markets (with the exception of Australia). Shared governance is strongly supported by intensive communication and consultation between chain actors. There are monthly meetings of industry consultation groups where outcomes of market research are discussed and fortnightly conferences on quality between market representatives and post-harvest representatives. In addition, there is an annual tours programme during which retailers, wholesalers and authorities visit New Zealand. Moreover, growers and post-harvest operators visit markets (organised by the lead company as well as by post-harvest operators). Grower roadshows are attended by growers, post-harvest operators and the lead company three times a year. The lead company facilitates chain collaboration by giving assistance to growers for the implementation of Global GAP, knowledge transfer on growing techniques via grower education systems, workshops and online seminars, and intensive production support.

International Food and Agribusiness Management Review

260

Trienekens et al.

Volume 21, Issue 2, 2018

In the apple chain also, the market and distribution company is the lead party. The company is responsible for intelligence generation activities and optimises the product flow (harvest, packaging, storage, transport) through planning and a system of incentives. It sets standards for growing and packing the own apple varieties. Its source of power is vertical integration, through ownership of several orchards and packing houses, and plant variety rights for two major varieties. It supports shared governance through weekly meetings between managers of owned orchards (about technical issues, packing of apples and developments in the market) and monthly meetings between contract growers, managers of owned orchards, post-harvest operators and the leading company (about sales, complaints and prices). Visits of representatives of the company to post-harvest operators takes place on a weekly or twoweekly basis (about new packaging and label requirements, changes in market demand, timing of supply). Meetings between managers of the companyâ&#x20AC;&#x2122;s subsidiaries in Europe and the head office in New Zealand take place twice a year. Grower meetings take place four to six times a year. Annually there are grower visits to markets and customer visits to New Zealand. Decisions are taken by the lead company. The lead company facilitates chain collaboration through knowledge transfer on growing techniques to growers on field days and visits by third party technical people. In addition, the company sends newsletters to growers with technical and market requirements. In the venison chain the facilitator takes the leadership role. It is responsible for some intelligence generation activities, connects actors, optimises the product flow and influences the behaviour of other actors (incentive systems). These activities are supported by Deer New Zealand which undertakes market research for the sector and provides marketing funds to the venison companies. The source of power of the facilitator is the trust of farmers in its competences to connect both ends of the value chain, and to develop and maintain all relationships in the value chain. Shared governance is an important issue in this value chain. In the farmer group there is intense communication and farmers try to help each other to improve. If a farmer wants to become part of the producer group, other farmers can use their veto if they believe that the farmer has a different philosophy or is not able to supply according to specifications. Meetings of the producer group and the facilitator take place four times a year (discussions about the market, farm management practices, industry-related topics). The Annual General Meeting is attended by farmers, the facilitator and one or two customers. Moreover, there is a board meeting of farmers and facilitator four times a year (discussions about sales volumes). There are customer visits to New Zealand and grower visits to the market (discussions about demand, supply, wishes of consumers, complaints, trends in the market, prices, sales volumes, and aspects for improvement). The lead company facilitates some market research. In the lamb chain leadership is less present. The processor tries to optimise the product flow and has developed some incentive systems. The main leader in this chain is a farmer who was one of the initiators of the chain. However, connections with the market are organised by the slaughterhouse. Shared governance is supported by communication and consultation between the board of the farmers group and the processor several times a year. The Annual General Meeting is attended by farmers and the processor. Until now the importer has not taken part in these meetings. Two farmers visited the importer. The processor visits the importer annually (but not only for this value chain). There is communication via e-mail, but actors would like to have more face-to-face meetings. A â&#x20AC;&#x2DC;Farm Quality Groupâ&#x20AC;&#x2122; provides advice to farmers on farm management practices. Table 6 shows clearly the differences in network governance structures. The chains with a clear leader and support for shared governance are the most integrated chains. It is interesting to note that the most integrated chains (A-kiwi and C-venison) use (partly) non-coercive power to organise the product flow, and also have more intense communication and shared governance.

International Food and Agribusiness Management Review

261

Trienekens et al.

Volume 21, Issue 2, 2018

Table 6. Network governance aspects for the four cases. Leaderships Case A: kiwi

Case B: apple Case C: venison Case D: lamb

Type of power

Lead party is marketing organisation

Power based on sole entry point (coercive power) Non-coercive power (based on competence trust) Lead party Power partly based on is marketing vertical integration and organisation contracts (coercive power) Lead party is Power based on facilitator competence trust (noncoercive power) No clear lead Slaughterhouse provides party market access

Shared governance

Facilitation

Intense communication and Marketing, knowledge consultation throughout the transfer and training chain Decisions taken by lead party; meetings to optimize product flow Intense communication and consultation throughout the chain Limited communication between end-to-end chain partners

Marketing, knowledge transfer and training Marketing and customer contacts (supported by sector organization New Zealand) Some training organized by the farmer group

â&#x2013; â&#x2013; Informal relationships In the kiwi value chain permanent and good relationships exist between growers, post-harvest operators and the lead company as well as between the lead company and customers in which there is openness, trust, twoway dialogue, integrity and respect. Transparency, incentive systems and belief in the single desk structure contribute to the relationships between the New Zealand actors. Relationships between growers and postharvest operators are also generally good. The role of trust is considered huge as it underpins every stage in the value chain, namely the trust that every actor in the value chain does a good job and acts to the best of its abilities. The commitment of growers is supported by consistent and safe returns, as well as by the shareholding system. Commitment to the chain is positively related to trust between the actors. In the apple value chain relationships between lead company and the growers are good according to most growers. However, some growers do not want to commit their volumes to the lead company and therefore the lead company has partly vertically integrated by establishing its own orchards. Relationships between the lead company and post-harvest operators can be awkward as the lead company can be very stringent in what is accepted in terms of quality specifications of the apples. There are good relationships between the lead company and customers. In general, the focus of the value chain is more on vertical integration than on relationships. Trust, supported by communication, does support transactions in the value chain. There is trust in the accuracy and correctness of information received. Open communication and transparency (as stimulated by trust) contribute to commitment. The venison value chain is characterised by tight and good relationships between farmers, slaughterhouse and the facilitator due to open communication. Moreover, there are good relationships between farmers and customers in the market due to face-to-face meetings and the emotional attachment of farmers to customers. Relationships between the facilitator and the European distributor are crucial for market orientation. In this regard, face-to-face meetings are crucial for building relationships. The role of trust is considered huge as it influences the willingness to share information and communicate and it replaces the need to constantly control and check the behaviour of other actors. There is trust in compliance with agreements, and the competences of other actors. The actors are also highly committed to the value chain, supported by the system of shareholding. Commitment results in actors spending more time, effort and resources on the value chain. In the lamb value chain there are good and collaborative relationships between farmers and the processor due to a commitment to grow and develop the product in the market. There are also good business relationships between the processor and the importer (which are not limited to this chain). The importer has relationships International Food and Agribusiness Management Review

262

Trienekens et al.

Volume 21, Issue 2, 2018

with distributors however there is little connection to the final customer (restaurants and specialist butchers). Relationships between farmers and the importer are not optimal due to misunderstandings, possibly caused by a lack of face-to-face meetings (difficult due to geographical distance and lack of resources). The challenges related to volumes and prices put pressure on the trust present in the value chain. Many farmers are quite committed to the value chain due to their shareholding in the Farm Quality Group, shared passion and belief in the programme. However, not all farmers seem highly committed to delivering the volumes required in the market. From Table 7 we see that trust and commitment are closely related to shared governance (Table 6) and (noncoercive) leadership. In the following section we will further analyse the relationships between market orientation and governance. Table 7. Levels of trust and commitment in the four case chains.1 Case A: kiwi Case B: apple Case C: venison Case D: lamb 1

Trust

Commitment

++ + ++ +

++ + ++ +/-

++: strong presence; +: moderate presence; +/-: limited presence of trust/commitment.

6. Discussion and conclusions Table 8 gives insights into the main relationships between the concepts that we applied in our research. In three of the four value chains, there is a clear actor who has taken the leadership role. The absence of a leader in the lamb value chain could explain why this value chain is less market-oriented than the other three. There are typical similarities in the activities that reflect the leadership role: ■■ All three leaders have a strong focus on markets and customers or, in other words, aim to create a market-oriented chain. ■■ All three leaders are responsible for conducting intelligence-generating activities. The kiwi and apple lead parties conduct a large variety of intelligence-generating activities in their value chain and the venison chain conducts some of these activities. ■■ All three leaders oversee the whole value chain from the farmers to the market. This enables the leader to facilitate intelligence communication from one end of the value chain to the other. ■■ All three leaders are involved in optimising the product flow as they all communicate demand and supply with the different actors in the value chain and try to match these. ■■ All three leaders influence the behaviour of other actors in the value chain and thereby support market responsiveness of the chain. For example, both fruit chains established detailed standards for growers and post-harvest operators. The lead party in the venison chain established an incentive system for farmers related to product specifications of the venison. The points show that leadership in a value chain may contribute to market intelligence generation, communication and responsiveness, which is in line with the findings of Elg (2008) and Kahkonen and Tenkanen (2010). The use of coercive power by the lead organisation can create tension in the relationship, because the lead organisation can force the other party to do things that this party would otherwise not have done (Leonidou et al., 2008). In kiwi value chain A, the impact of the use of coercive power on relationships between marketing organisation and farmers is moderate, as the shareholding structure in this chain gives growers the power to International Food and Agribusiness Management Review

263

Trienekens et al.

Volume 21, Issue 2, 2018

Table 8. Overview of relationships between governance elements and market orientation of four case value chains.1 Chain leader Shared governance Facilitation Strictness contract Equity based Trust Commitment Market intelligence Communication Responsiveness 1

Case A:kiwi

Case B: apple

Case C: venison

Case D: lamb

++ ++ ++ ++ ++ ++ ++ ++ ++ ++

++ + ++ ++ +/+ + ++ + +

++ ++ + +/++ ++ ++ ++ ++ +

+/+/+/+/+/+ +/+/+ -

++: strong presence/attention; +: moderate presence/attention; +/-: limited presence/attention; -: lack of attention.

take a stand if they do not agree with decisions or actions taken by the lead company. Moreover, alignment of actors in this chain is supported by the good relationships between the growers, post-harvest operators and the lead company. In addition, this chain has a special shared governance structure with its industry consultation groups, in which growers, post-harvest operators and the lead company are represented. These industry consultation groups meet once a month to discuss major industry issues, such as taste reviews. The meetings of the consultation groups contribute to intelligence communication through the value chain, since market information can be shared. In addition, the meetings contribute to responsiveness of the value chain as decisions can be taken about how to respond to market information, such as outcomes of sensory research. In the apple value chain (B), the leader uses much more coercive power to align activities with its own objectives. Shared governance in this chain is at a lower level than the kiwi chain, as are trust and commitment. Another example of shared governance, in which part of the chain was governed by several actors, was found within the lamb and venison value chains, in which the farmers formed a farmersâ&#x20AC;&#x2122; group. Within these farmer groups, there is communication on how farm management practices can be improved. The venison producer group in particular is characterised by strong collaboration between the farmers and also with other actors in the chain, as well as by social pressure between the farmers to apply good farming practices and on the other actors to perform in line with the common objectives. The existence of the venison producer group, and its strong collaboration with the facilitator, slaughterhouse and distributor in Europe, contributes to the responsiveness of the value chain. In venison value chain C the lead party applies non-coercive power based on competence trust. There is a high level of shared governance in this chain, as well as high levels of trust and commitment between the different actors. In lamb chain D no clear lead party is present, and, although there is close collaboration within the farmer group, shared governance and trust throughout the chain are moderate while commitment is lower than in the other chains. Leadership can be supported by facilitation of market research, market intelligence communication throughout the chain and farmer support. Additionally, findings indicate that the facilitating role can be taken by a value chain actor as well as by an external organisation, that is not, or only to a very limited extent, involved in the product flow of the value chain. This is the case for the venison value chain. Regarding contractual arrangements, in three of the four studied value chains, not just one but multiple types of bilateral agreements are being used. In the kiwi value chain there are written as well as equity-based agreements, while in the venison value chain we see (long-term) verbal agreements next to equity-based agreements. In the apple chain the focus is on written agreements and in the lamb chain on (long-term) verbal agreements. A clear finding in this research is that spot-market agreements are not used in any of the four value chains. Peterson et al. (2001) described spot-market contracts as being focused on the short International Food and Agribusiness Management Review

264

Trienekens et al.

Volume 21, Issue 2, 2018

term and associated with limited information exchange. Therefore, a certain level of integration might be conditional to being market-oriented. 6.1 Conclusions Leadership was found to be an important characteristic of value chain governance as it could contribute to all three constructs of market orientation. The actor taking a leadership role has the ability to oversee the whole value chain and connect actors in the value chain, which contributes to intelligence communication. Additionally, the lead organisation is able to optimise the product flow and influence the behaviour of other value chain actors, both of which contribute to value chain responsiveness. The leader also performs activities aimed at gathering market information, which contributes to intelligence generation. Moreover, the leader can create and stimulate one common vision in the value chain, namely the focus on customers and consumers. By doing so, the leader can stimulate market orientation to be the common goal of the value chain. Leadership was found to be closely related to facilitation, since the actor taking a leadership role can also act as a facilitator. The facilitator can provide other actors with knowledge and/or assistance, and thereby contribute to their responsiveness. Furthermore, facilitation can contribute to intelligence generation by conducting intelligence-generating activities. Leadership can be the result of coercive and non-coercive sources of power. Non-coercive sources of power were found to have a positive effect on the informal relationships in the value chain. In contrast, the coercive sources of power had a negative impact on these informal relationships, as the lead organisation may force other actors to do things that these actors would otherwise not have done (Grunert et al., 2005; Kähkönen and Tenkanen, 2010; Leonidou et al., 2008). This can create tensions in the informal relationships. However, trust and commitment, as constructs of informal relationships, can contribute to all three constructs of market orientation, namely to intelligence communication (by increasing the willingness of actors to share information), responsiveness (by increasing the willingness of actors to dedicate time, effort and resources to the value chain) and intelligence generation (via their contribution to relationship quality and related information exchange). It was found that by involving different actors in the decision-making via shared governance, the negative impact of leadership on informal relationships could be decreased. Shared governance, characterised by the presence of many face-to-face meetings, also contributes to intelligence communication and responsiveness, as market information can be exchanged and decisions regarding responsiveness can be taken during these meetings. In addition to shared governance, shareholding of actors can also lower the negative impact of leadership on informal relationships. When actors have a shareholding in the value chain, they have the power to stand up if they do not agree with decisions or actions taken by the leader. 6.2 Management implications, limitations and further research Based on the insights obtained in this research, recommendations can be made to actors of global food value chains, which aim to be market-oriented. The insights from this research can be used by value chain actors to analyse their value chain and find points for improvement in order to increase their market orientation. In addition, all of these aspects of governance are interrelated and can therefore positively or negatively influence each other. For example, leadership in particular was found to contribute to market orientation in several ways. However, when using coercive power, leadership can also have a negative impact on the informal relationships in the value chain, while these informal relationships are important contributors to market orientation. Therefore, special attention must be paid to these informal relationships. ■■ Further research This exploratory research into market orientation and value chain governance has delivered new insights for further research. First, only a little research has been performed on market orientation of value chains (e.g. Grunert et al., 2005). Even less research has been performed on the governance of value chains. This study International Food and Agribusiness Management Review

265

Trienekens et al.

Volume 21, Issue 2, 2018

contributes strongly to the literature in the field of market orientation as well as value chain governance. For the latter in particular, we have designed an innovative theoretical framework, combining insights from contract theory (e.g. Williamson, 1991), social network theory (e.g. Lu et al., 2012) and network governance (e.g. Provan and Kenis, 2008). The framework and approach can be used by other scholars and add to our insights in market orientation and value chain governance. This research has several limitations. The research assessed the contribution of governance to market orientation. The effect of the level of market orientation on the performance of the value chain (for example, profitability) was not investigated. Moreover, governance costs of the value chain were also not included in the study. Further research may include these elements. Four chains were addressed in this exploratory research to arrive at new insights into the relationships between market orientation and value chain governance. As the number of chains and the qualitative research performed in each of these chains must be considered rather limited, further research should focus on more and more differentiated chains from different sectors, with more quantitative based in-depth research.

References Alvarez, G., C. Pilbeam and R. Wilding. 2010. NestlĂŠ Nespresso AAA sustainable quality program: an investigation into the governance dynamics in a multi-stakeholder supply chain network. Supply Chain Management 15(2): 165-182. Barratt, M. 2004. Understanding the meaning of collaboration in the supply chain. Supply Chain Management 9(1): 30-42. Bigne, E. and A. Blesa. 2003. Market orientation, trust and satisfaction in dyadic relationships:a manufacturerretailer analysis. International Journal of Retail and Distribution Management 31(11): 574-590. DeshpandĂŠ, R. and J.U. Farley. 1998. Measuring market orientation: generalization and synthesis. Journal of Market Focused Management 2(3): 213-232. DeshpandĂŠ, R., J.U. Farley and F.E. Webster. 1993. Corporate culture, customer and innovativeness orientation, in a quadrad japanese firms: analysis. Journal of Marketing 57(1): 23-37. Dolan, C. and J. Humphrey. 2004. Changing governance patterns in the trade in fresh vegetables between Africa and the United Kingdom. Environment and Planning 36: 491-509. Eisenhardt, K.M. and M.E. Graebner. 2007. Theory building from cases: opportunities and challenges. The Academy of Management Journal 50(1): 25-32. Elg, U. 2008. Inter-firm market orientation and the influence of network and relational factors. Scandinavian Journal of Management 24(1): 55-68. Gereffi, G., J. Humphrey and T. Sturgeon. 2005. The governance of global value chains. Review of International Political Economy 12(1): 78-104. Gibbon, P., J. Bair and S. Ponte. 2008. Governing global value chains: an introduction. Economy and Society 37(3): 315-338. Green Jr, K.W., R. McGaughey and K.M. Casey. 2006. Does supply chain management strategy mediate the association between market orientation and organizational performance? Supply Chain Management 11(5): 407-414. Grunert, K.G., L.F. Jeppesen, K.R. Jespersen, A.M. Sonne, K. Hansen and T. Trondsen. 1996. Market orientation in food and agriculture. Kluwer, Boston, MA, USA. Grunert, K.G., L.F. Jeppesen, K.R. Jespersen, A.-M. Sonne, K. Hansen, T. Trondsen and J.A. Young. 2005. Market orientation of value chains: a conceptual framework based on four case studies from the food industry. European Journal of Marketing 39: 428-455. Grunert, K.G., T. Trondsen, E.G. Campos and J.A. Young. 2010. Market orientation in the mental models of decision makers: two cross-border value chains. International Marketing Review 27(1): 7-27.

International Food and Agribusiness Management Review

266

Trienekens et al.

Volume 21, Issue 2, 2018

Grunert, K.G., N. Wognum, J. Trienekens, M. Wever, N.V. Olsen, and J. Scholderer. 2011. Consumer demand and quality assurance: segmentation basis and implications for chain governance in the pork sector. Journal on Chain and Network Science 11(2): 89-97. Han, J.K., N. Kim and R.K. Srivastava. 1998. Market orientation performance :organizational is innovation a missing link? Journal of Marketing 62(4): 30-45. Homburg, C. and C. Pflesser. 2000. A multiple-layer model of market-oriented organizational culture: measurement issues and performance outcomes. Journal of Marketing Research 37(4): 449-462. Hunt, S.D. and J.R. Nevin. 1974. Power in a channel of distribution: sources and consequences. Journal of Marketing Research 11: 186-193. Jain, V. and L. Benyoucef. 2008. Managing long supply chain networks: some emerging issues and challenges. Journal of Manufacturing Technology Management 19: 469-496. Jaworski, B.J. and A.K. Kohli. 1993. Market orientation: antecedents and consequences. Journal of Marketing 57(3): 53-70. Jeong, J.S. and P. Hong. 2007. Customer orientation and performance outcomes in supply chain management. Journal of Enterprise Information Management 20(5): 578-594. Kähkönen, A.-K. and M. Tenkanen. 2010. The impact of power on information sharing in the Finnish food industry. British Food Journal 112(8): 821-835. Kaplinsky, R. 2000. Globalisation and unequalisation: what can be learned from value chain analysis. Journal of Development Studies 73(2):117-146. Karami, M., S. Malekifar, A.B. Nasiri, M.B. Nasiri, H. Feili and S.U.R. Khan. 2015. A conceptual model of the relationship between market orientation and supply chain performance. Global Business and Organizational Excellence 34(2): 75-85. Kirca, A.H., S. Jayachandran and W.O. Bearden. 2005. Market orientation: a meta-analytic review and assessment of its antecedents and impact on performance. Journal of Marketing 69(2): 24-41. Kohli, A.K., and B.J. Jaworski. 1990. Market orientation: the construct, research propositions, and managerial implications. The Journal of Marketing 54(2): 1-18. Kwon, I.-W.G. and T. Suh. 2005. Trust, commitment and relationships in supply chain management: a path analysis. Supply Chain Management 10(1): 26-33. Lazzarini, S.L., F.R. Chaddad and M.L. Cook. 2001. Integrating supply chain and network. Journal on Chain and Network Science 1(1): 7-22. Leonidou, L.C., M.A. Talias and C.N. Leonidou. 2008. Exercised power as a driver of trust and commitment in cross-border industrial buyer-seller relationships. Industrial Marketing Management 37(1): 92-103. Lu, H., S. Feng, J.H. Trienekens and S.W.F. Omta. 2012. Network strength, transaction-specific investments, inter-personal trust, and relationship satisfaction in Chinese agri-food SMEs. China Agricultural Economic Review 4 (3): 363-378. Martinez, S.W. and K. Zering. 2004. Pork quality and the role of market organization. Agricultural Economic Report 835. Available at: http://tinyurl.com/ydybhs63. Matsuno, K., J.T. Mentzer and J.O. Rentz. 2005. A conceptual and empirical comparison of three market orientation scales. Journal of Business Research 58(1 SPEC.ISS): 1-8. Ménard, C. 2004. The economics of hybrid organizations. Journal of Institutional and Theoretical Economics 160(3): 345-376. Micheels, E.T. and H.R. Gow. 2011. The moderating effects of trust and commitment on market orientation, value discipline clarity, and firm performance. Agribusiness 27(3): 360-378. Morgan, R.M., and S.D. Hunt. 1994. The commitment-trust theory of relationship marketing. Journal of Marketing 58(July): 20-38. Narver, J.C. and S.F. Slater. 1990. The effect of market orientation on business profitability. Journal of Marketing 54(4): 20-35. Peterson, H.C., A.F. Wysocki and S.B. Harsh. 2001. Strategic choice along the vertical coordination continuum. International Food and Agribusiness Management Review 4: 149-166. Ponte, S. and T. Sturgeon. 2014. Explaining governance in global value chains: a modular theory-building effort. Review of International Political Economy 21(1): 195-223

International Food and Agribusiness Management Review

267

Trienekens et al.

Volume 21, Issue 2, 2018

Provan, K.G. and P. Kenis. 2008. Modes of network governance: structure, management, and effectiveness. Journal of Public Administration Research and Theory 18(2): 229-252. 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. Raynaud, E., L. Sauvée and E. Valceschini. 2009. Aligning branding strategies and governance of vertical transactions in agri-food chains. Industrial and Corporate Change 18(5): 835-868. Ren, Z., A. Saengsathien and D. Zhang. 2014. Modeling and optimization of inventory and sourcing decisions with risk assessment in perishable food supply chains. IEEE International Conference on Industrial Engineering and Engineering Management (pp. 934-939). Available at: http://tinyurl.com/y9l94u5y. Ryu, I., S. So and C. Koo. 2009. The role of partnership in supply chain performance. Industrial Management and Data Systems 109(4): 496-514. Tomášková, E. 2009. The current methods of measurement of market orientation. European Research Studies 12(3): 135-150. Trienekens, J. and N. Wognum. 2013. Requirements of supply chain management in differentiating European pork chains. Meat Science 95(3): 719-726. Trienekens, J.H., P.M. Wognum, A.J.M. Beulens and J.G.A.J. van der Vorst. 2012. Transparency in complex dynamic food supply chains. Advanced Engineering Informatics 26(1): 55-65. Van Vlerken, P. 2015. Opportunities for capturing a consumer premium for high-value food products – a case on the Kumanu lamb supply chain. Wageningen University, Wageningen, the Netherlands. Vieira, V.A. 2010. Antecedents and consequences of market orientation: a Brazilian meta-analysis and an international mega-analysis. Brazilian Administration Review 7(1): 40-58. Wever, M. 2012. Chain-wide consequences of transaction risks and their contractual solutions – managing interdependencies in differentiated agri-food supply chains. Wageningen University, Wageningen, the Netherlands. 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(2): 228-237. Williamson, O.E. 1991. Comparative economic organization: the analysis of discrete structural alternatives. Administrative Science Quarterly 36(2): 269-296.

International Food and Agribusiness Management Review

268

OPEN ACCESS International Food and Agribusiness Management Review Volume 21 Issue 2, 2018; DOI: 10.22434/IFAMR2017.0114 Received: 31 March 2017 / Accepted: 16 November 2017

Social sustainability in the ready-made-garment sector in Bangladesh: an institutional approach to supply chains Special issue: IFAMA 2017 symposium

RESEARCH ARTICLE Laura A. Carlson

and Vera Bitschb

aResearch

Associate, and bProfessor, Chair of Economics of Horticulture and Landscaping, Technical University Munich, Alte Akademie 16, 85354 Freising, Germany

Abstract Ready-made-garment (RMG) production for sale in the EU and USA is a key source of economic development for Bangladesh. The 2013 collapse of Rana Plaza revealed worker safety and other social issues in RMG factories in Bangladesh, showing that formal, top-down approaches to these problems, including corporate codes-of-conduct and reforms in Bangladeshi labor laws, have little effect. Supply chain sustainability is a key issue for business, government and civil society. Satisfactory theoretical approaches to promoting social sustainability in supply chains are lacking. A case study using qualitative document analysis identifies the key institutional factors related to social sustainability in the Bangladeshi RMG industry, with a modified version of the Institutional Analysis and Development (IAD) framework as an analytical frame. Key elements of other frameworks for social sustainability are discussed in terms of how well the IAD framework captures those concepts, and how employing the IAD could enhance supply chain analysis. Keywords: action situation, IAD framework, new institutional economics, rules-in-use, social and cultural context JEL code: B25, B27, D02, D46, D62, D63, D64, D70, D80, D91, E14, F23, F6, J5, J8, K31, L67, O1 Corresponding author: laura.carlson@tum.de

ÂŠ 2018 Carlson and Bitsch

269

Carlson and Bitsch

Volume 21, Issue 2, 2018

1. Introduction The collapse of Rana Plaza in Bangladesh in 2013 left 1,134 workers from the ready-made garment (RMG) industry dead, bringing the issue of worker safety in the supply chains of major garment retailers in Bangladesh into world-wide public attention (Kenney, 2013). Bangladesh is a country of approximately 150 million people, nearly a third of whom are classified as poor (World Bank, 2013). In the short period of time since Bangladesh first became an independent nation in 1971, the country has managed significant decreases in such key negative social indicators as poverty, infant mortality and population growth, particularly since 2000 (World Bank, 2013: 49). The World Bank identifies several factors that have contributed to this decline: Real per-capita income increased an average of 19% between 2000 and 2005, and 54% between 2005 and 2010 across all income levels, with increases among the poorest 10% of the population significantly lower at 29% (2000-2005) and 7% (2005-2010) (World Bank, 2013: 44). Demographic changes, increased education levels and increases in foreign remittances have all contributed to the decrease in poverty, but the single most important contribution has come from an increase in labor income (World Bank, 2013: 49). The majority of the population of Bangladesh (nearly 110 million people) still resides in rural areas (Bangladesh Bureau of Statistics, 2013: 24). Accordingly, in 2013, the largest share of all employed (45.1%) was in agriculture, with services encompassing another 34.1%, and only 20.8% employed in industry (Bangladesh Bureau of Statistics, 2013: xvi). By far, the largest single industrial employer in Bangladesh is the labor-intensive RMG industry, with an official estimate of 2,997,000 employees in 2013, accounting for approximately 5% of the total employed population of 58,073,000 (Bangladesh Bureau of Statistics, 2013: 137-139). Despite the small percentage of the total population active in the sector, the 20 billion USD earnings per year the RMG industry generates (World Bank, 2013: 2) constitute 80% of the total export value of Bangladesh (Kathuria and Malouche, 2016: 9). As a result, RMG producers enjoy a powerful position in Bangladeshi society. A New York Times article published a few days after the Rana Plaza collapse stated, critics have argued that the outsize importance of the industry has made the government reluctant to take steps that could increase costs or alienate foreign brands...Meanwhile, some factory owners say they cannot raise wages or invest in upgrading facilities because of the low prices paid by Western brands (Yardley and Manik, 2013). Labor activists interviewed following the collapse claimed that neither the Bangladeshi government nor Western buyers are taking their share of the responsibility for protecting workers in this industry (Yardley and Manik, 2013). Bangladeshi RMG production represents a broader trend in the globalization of production processes into complex global supply chains that has led many to focus on the concept of supply chain sustainability, which emphasizes ‘‘interrelationships among society, the environment, and economic/industrial development’’ (Hutchins and Sutherland, 2008: 1688). This focus has generated a great deal of collective action involving social sustainability issues (Bartley and Child, 2011) that has contributed to companies in many labor-intensive production sectors establishing ‘‘extensive supplier labor codes’’ (Maloni and Brown, 2006: 35). However, ‘‘implementation of such codes of conduct has been considerably more successful in some sectors than others’’ (Roberts, 2003). The business case for such codes is that ‘‘poor environmental and social conditions in corporate supply networks can pose significant reputational risks to big name brands’’ (Roberts, 2003). However, an investigation of the effectiveness of codes of conduct for improving supply chain sustainability found, ‘‘Given the limited overall improvements that codes lead to, it is surprising that companies continue their multi-billions of investments in codes and auditing’’ (Egels-Zandén and Lindholm, 2015: 38). This failure is in part due to the fact that, as Moon (2007: 298) pointed out, sustainability issues (particularly social ones) are ‘‘highly contextual’’, and are ‘‘subject to issue attention cycles in which events or findings give them urgency’’. This is characteristic of ‘‘wicked problems’’ which are by nature unsolvable, as they International Food and Agribusiness Management Review

270

Carlson and Bitsch

Volume 21, Issue 2, 2018

change in relationship to new events and cannot be ‘‘fully defined and separated from their context’’ (Bitsch, 2016: 17). This changeability is exacerbated by the array of actors involved who often have ‘‘contradictory perspectives, beliefs, values and goals’’ with respect to a given situation (Bitsch, 2016: 17). In discussing efforts to design rule systems that support sustainable management of natural resource systems, Ostrom (2005: 255) points out the impracticality of producing one-size-fits-all rule-based solutions (policies or institutional structures) to complex natural resource problems. This claim is backed up by the results of a number of case studies of long-enduring natural resource governance systems in a variety of legal, cultural and environmental settings that indicate that the rules from which each successful system is constructed vary ‘‘dramatically from one system to the next’’ (Ostrom, 2005: 258). To help make sense of some of this variety, Ostrom and her colleagues made use of the empirical data derived from these case studies to develop and apply the Institutional Analysis and Development (IAD) framework (Ostrom, 2005: 9). Based on this theoretical framework, a set of eight design principles for enduring sustainable natural resource management systems was developed (Ostrom, 2005: 259). This framework can be used to categorize rules according to who makes them and what aspect of the resource system they affect. Several authors have tried to create similar conceptual frameworks to help guide the search for the key elements necessary for promoting social sustainability (Beske et al., 2014; Boström, 2012; Cuthill, 2010; Henderson et al., 2002; Murphy, 2012; Seuring and Müller 2008). In the literature review that follows, first some background on social sustainability issues in supply chains is presented. This leads into a discussion of recently published theoretical frameworks for assessing social sustainability. The key concepts that form those frameworks are identified and compared with the key elements of the IAD framework and the eight design principles mentioned above. The remainder of the manuscript is structured as follows: the method section describes the data collection and analysis methods used to generate an empirical dataset describing the Bangladeshi RMG industry as a resource system. The results and discussion section that follows first presents some of the key features of the Bangladesh RMG industry obtained from the empirical data using the concepts that make up the IAD framework as organizing elements. Based on the framework, emphasis is on the institutional arrangements that define the cultural and social issues and relationships between actors that have unique impacts on this particular part of this particular supply chain in this particular location. The ultimate goal, however, is to assess the application of the IAD framework to help organize and support theoretical approaches to sustainability in supply chain settings.

2. Literature Review Most products today move through a progression of individual business transaction steps on the way to the final consumer, a structure commonly referred to as the supply chain (Hutchins and Sutherland, 2008: 1689). At each of these steps, the possibility exists for negative externalities (defined as costs of doing business that are born by those other than the producer of the good) (Lipschutz, 2004: 197). Externalities are commonly thought of as negative environmental effects, but social impacts occur as well. Among the negative social externalities that have been identified in complex international supply chains in labor-intensive industries such as Bangladeshi RMG are ‘‘low wages, substandard working conditions, forced overtime, child labour, and lack of a right of free association (union organizing)’’ (Lipschutz, 2004: 197). As a result of the increasingly global nature and complexity of supply chains and an ever-growing interest in sustainability, new actors traditionally seen as outside the supply chains of firms have taken action to ensure that the full social consequences of this form of economic organization are considered. These groups engage in ‘‘naming and shaming’’ campaigns that go beyond traditional labor organizing to bring attention to poor working conditions and other issues related to labor (Bartley and Child, 2011: 425). The development of corporate codes of conduct and sustainability initiatives is often attributed to the collective action efforts of these stakeholders (Bartley and Child, 2011: 427). These new rule systems represent a ‘‘move from factoryInternational Food and Agribusiness Management Review

271

Carlson and Bitsch

Volume 21, Issue 2, 2018

centered, state regulation focusing on individual sites of production to supply-chain and ’’brand’’ regulation focusing on multiple actors in a production chain’’ (O’Rourke, 2003: 6). The aspects of most supplier codes of conduct today that apply to social sustainability issues are based on principals for fair labor practices developed and promoted by the United Nations labor agency, the International Labor Organization (ILO) (Dae-oup Chang and Wong, 2005: 141-142; O’Rourke, 2003: 6-7). In some cases, codes of conduct are drafted and policed by the companies alone, but in other cases unions, Non-governmental organizations (NGOs) and other stakeholders have developed their own rules and monitoring systems that are then adopted by companies (O’Rourke, 2003: 10-11). Examples of the latter are often developed and propagated by multi-stakeholder initiatives such as the Fair Labor Association and the Clean Clothes Campaign (Dae-oup Chang and Wong, 2005: 142) that otherwise are largely engaged in monitoring efforts. The variety and number of external monitoring efforts has clearly introduced even more complexity into global RMG supply chains, which can create additional problems for implementation as, ‘‘Stakeholders can compete for legitimacy, influence and recognition from companies and public opinion’’ (Albareda, 2009; Arenas et al., 2009: 176). This complexity of actors makes managing a supply chain particularly difficult. ‘‘The term supply chain management has been used to explain the planning and control of materials and information flows as well as the logistics activities not only internally within a company but also externally between companies’’ (Chen and Paulraj, 2004: 119). This concept has been expanded to explicitly include the concept of sustainability, and is referred to as Sustainable Supply Chain Management (SSCM) (Turker and Altuntas, 2014: 837). Scholars in the field of SSCM tend to discuss sustainability in the supply chain in terms of ‘‘outcomes’’ that result from ‘‘managerial decisions and/or behaviors’’ (Pagell and Wu, 2009: 38). The emphasis of much of the SSCM literature is on the relationship between environmental and economic outcomes with little attention paid to social sustainability impacts (Pagell and Wu, 2009: 38). particularly in relationship to the ‘‘cultural or ethical ramifications of decisions’’ (Hutchins and Sutherland, 2008: 1688). Recognizing this lack, several authors have made efforts towards developing theoretical frameworks that capture the key features necessary for promoting social sustainability in supply chains (Beske, et al., 2014; Boström, 2012; Coe et al., 2008; Cuthill, 2010; Henderson et al., 2002; Hess, 2008; Murphy, 2012; Seuring and Müller, 2008). In early work, Henderson et al. (2002) presented a framework based on what the termed global production networks as an alternative to the commonly used term supply chain. In their view, the ‘‘chain metaphor’’ places too much emphasis on the sequential nature of the production process, and that looking at the production process in terms of networks of actors allows analysis of the ‘‘multi-dimensional and multi-layered lattices of economic activity’’ (Henderson et al., 2002: 442). Their analysis moved sustainable supply chain theory forward in that it focused on ‘‘social processes’’ between actors in the supply chain, including labor (Henderson et al., 2002: 444). Recognition of the importance of the diversity of actors that make up a global production networks is one of the key characteristics of this theoretical perspective, whose adherents also stress the importance of the embeddedness of parts of the network (Coe et al., 2008: 279). This latter distinction is key, as it begins to take into account both the importance of cultural differences in different countries that may affect relationships within the chain of production, and potential avenues of political pressure for various actors that may be outside the immediate cultural setting of a segment of the supply chain, but are still within the ‘‘network’’. Seuring and Müller (2008: 1704) conducted a structured review of the supply chain management literature to move towards a theoretical framework for SSCM. They identified ‘‘barriers and supporting factors’’ to SSCM that focus largely on issues of communication, information exchange, monitoring and costs associated with the increasing complexities of supply chains. Similarly, Boström (2012: 7-10) identified ‘‘six challenges for operationalizing and integrating social sustainability’’, including unrealistically high expectations of win-win-win scenarios that evenly consider all three sustainability pillars. Boström feels such expectations inherently fail to account for necessary tradeoffs. He points out that the some of these high hopes are created by often ‘‘vague, subjective, and ideological framing’’, and argues that the roots of this framing lie International Food and Agribusiness Management Review

272

Carlson and Bitsch

Volume 21, Issue 2, 2018

in the historical development of the sustainability concept that began with a focus on environmental issues. For Boström, the perpetuation of the separate pillars concept in sustainability framing reflects a lack of institutional links between environmental and social issues. And further, that all of the above contribute to the continued promotion of ‘‘global capitalism for sustainable development’’; which ignores the fact that, under current institutional arrangements, the distribution of the benefits of capitalism (as well as the costs) is inherently inequitable and unsustainable. This leads to Boström’s final point, that in order to overcome the challenges already delineated, special attention must be paid to who is involved in the decision making process and how their input is integrated into rule making and enforcement, rather than simply examining the outcomes of those processes. Some of the high expectations and vague, subjective and ideological framing to which Boström refers can be seen in the framework for analyzing sustainable development put forth by Murphy (2012: 21). Murphy lists the following as the most important aspects that contribute to social sustainability: social cohesion, equity, participation, and awareness for sustainability. He goes on to describe policy areas that fit under each of these aspects, and ways to analyze them, which are almost exclusively based on ‘‘commitment’’ to undertake broad, general actions. Similarly vague are the defining concepts Cuthill (2010: 366) put forth based on action research related to sustainable urban development in Australia. These include social capital, social justice and equity, engaged governance and social infrastructure. Beske et al. (2014: 133-139) presented an integration of the theories from the Dynamic Capabilities literature and those from the field of SSCM. They present ‘‘categories in which SSCM practices can be structured’’ that include strategic orientation, continuity, collaboration, risk management, and pro-activity in relation to other stakeholders. The main elements they identified from Dynamic Capabilities are knowledge management, partner development, supply chain reconceptualization, co-evolving and reflexive control. It seems that both approaches focus on practices firms can take to promote sustainability. As evidenced by many of the concepts included in the various frameworks described above, such as participation, engaged governance, partner development and co-evolving, many supply chain actors are expected to engage in collective and collaborative efforts with one another that require cooperation beyond simple business transactions. The continued devolution of government control over market processes only increases this need, and more and more collective efforts are evolving to create complex interactions. To better understand these interactions, it is useful to look at existing approaches to studying cooperation in collective action situations. Evidence from extensive experimental economics work has shown that the intrinsic preferences of a significant percentage of research subjects actually create a willingness to cooperate with others, even in situations where not cooperating is certain to bring a higher personal monetary payoff. These preferences, however, become weaker after repeated ‘‘bad experiences’’ of betrayal by others (noncooperation) (Ostrom, 2005: 129). Additionally, the intrinsic willingness to cooperate may also be affected by externally composed constraints that tend to ‘‘crowd out’’ that willingness if participants see them as controlling (Ostrom, 2005: 113). Such externally imposed constraints may be formal or informal, but all can be thought of as institutions (North, 1989). Institutions in this context are ‘‘human-constructed constraints or opportunities within which individual choices take place and which shape the consequences of their choices’’ (McGinnis, 2011). In order to fully understand the ways in which individual actors interact with one another through implementing particular practices, it is necessary to understand the institutions that govern those interactions and support those practices. This includes both formal institutions such as written rules, laws and contracts, and informal ones such as norms, cultural constraints and gender roles. In describing institutional analysis, Ostrom (1990, 55) states, ‘‘The basic strategy is to identify those aspects of the physical, cultural, and institutional setting that are likely to affect the determination of who is to be involved in a situation, the actions they can take and the costs of those actions, the outcomes that can be achieved, how actions are linked to outcomes, what information is to be available, how much control individuals can exercise, and what payoffs are to be assigned to particular combinations of actions and outcomes’’. The IAD framework is a tool developed to organize each of these elements according to their relationships with one another. International Food and Agribusiness Management Review

273

Carlson and Bitsch

Volume 21, Issue 2, 2018

The IAD framework ‘‘identifies, categorizes, and organizes those factors’’ which are most important in understanding ‘‘the ways in which institutions operate and change over time’’ (McGinnis, 2011: 169). The arena of analysis to which the IAD framework is applied is referred to as the action situation, or ‘‘the social spaces where individuals interact, exchange goods and services, solve problems, dominate one another, or ﬁght’’ (Ostrom, 2011). An actor can be either an individual or a group of individuals acting collectively, and is defined based on the combination of particular attributes that include resources, knowledge, decision making and valuation processes (Ostrom, 2011: 11). All of these attributes are assigned based on the unique set of rules to which each actor is subject, which include both rules-in-form (rules that exist on paper) (Ostrom, 2005: 138) and ‘‘working rules’’, or ‘‘rules-in-use’’ (Ostrom, 2005: 19-20). The distinction between rules-in-form and rules-in-use is particularly important in a system in which ‘‘there may be central laws and considerable efforts made to enforce them, but individuals generally attempt to evade rather than obey the law’’ (Ostrom, 2005: 20). In-depth analysis of these sets of rules in various resource management systems using the IAD framework has yielded a set of eight design principles common to long-standing resource systems. (Ostrom, 2005). According to these principles, ‘‘clearly defined boundaries’’ of the resource system are recognizable to users, including who has rights ‘‘to harvest resource units’’ (design principle 1). Users perceive there to be ‘‘proportional equivalence between benefits and costs,’’ meaning rules that limit extraction of resource units ‘‘are related to local conditions and to rules requiring labor, material, and/or money inputs’’ (design principle 2). Functional ‘‘collective-choice arrangements’’ exist that allow users of the resource to modify ‘‘harvesting and protection rules’’ that affect their potential actions (design principle 3). ‘‘Monitoring’’ is in place to ‘‘actively audit biophysical conditions and user behavior’’, and monitors ‘‘are at least partially accountable to the users and/or are the users themselves’’ (design principle 4). Further, ‘‘graduated sanctions’’ on ‘‘users who violate rules-in-use’’ are likely to come ‘‘from other users, from officials accountable to these users, or from both’’ (design principle 5). There are ‘‘conflict-resolution mechanisms’’ in place in which ‘‘low-cost, local arenas to resolve conflict among users or between users and officials’’ can be addressed in a timely way (design principle 6). There is at least ‘‘minimal recognition of rights to organize,’’ in that users that ‘‘devise their own institutions are not challenged by external governmental authorities, and users have long-term tenure rights to the resource’’ (design principle 7). Finally, ‘‘appropriation, provision, monitoring, enforcement, conflict resolution, and governance activities’’ that take place within the resource system ‘‘are organized in multiple layers of nested enterprises’’ (design principle 8) (Ostrom, 2005: 259). The fact that the IAD framework was developed by scholars primarily interested in the governance of public goods and common pool resources is reflected in its structure. McGinnis (2011) provides a description of the key elements of the IAD framework and the action situation it illustrates. In brief, each actor in an action situation is assigned to one or more positions based on attributes they possess which are determined by a combination of intrinsic qualities of the actor and various types of rules. Based on their assigned positions, participants must make choices among an array of possible actions. The set of possible actions they consider, including those they choose to take (and the outcomes of these actions) are affected by (and in turn affect) the actions chosen by other participants (and the outcomes of these actions) in an iterative dynamic process. This process is acted upon by a wider set of factors that include the rules-in-use within the resource management system, specific key attributes of the resource and the attributes of the community with primary rights to the resource. The latter aspect includes, ‘‘all relevant aspects of the social and cultural context within which an action situation is located’’ (McGinnis, 2011: 175). The environmental pillar of sustainability is easily and often connected to the management of common pool resources such as forests, fisheries and biodiversity. We argue that social sustainability issues in supply chains have characteristics that are also similar to common pool resources. The added value associated with the good produced as it moves along the chain is the resource being appropriated. The users of this resource system are those actors who manage to appropriate some amount of that added value along the chain. In that sense, each individual or group of individuals is attempting to appropriate as much of the amount of the added value as they consider to be possible and necessary given the amount of resources they themselves have International Food and Agribusiness Management Review

274

Carlson and Bitsch

Volume 21, Issue 2, 2018

and must contribute. The rules that apply to the appropriation of that value are different for different user groups and even for some individuals within groups. We can state that the resource is subtractable (another characteristic of common pool resources) in that there is a finite amount of value at the end of the chain that can be finally captured in the price of the good. This total value can, however, include both intrinsic and extrinsic value that apply to users at any point along the chain. Thus, it is not always measured in economic outcomes. Social sustainability issues are related to the relative appropriation of that value among users of the resource – workers, RMG factory owners, western retailers and consumers. Similar to appropriation problems in common pool resources, there is a tendency towards under-provision of social ‘‘goods’’, free-riding and cheating which in combination may lead to unequal value appropriation and other negative externalities. The IAD has been applied in a wide variety of locations to assess appropriation problems resulting from various institutional structures governing common pool resources (Ostrom, 2005: 9). Thus, we propose a modified version of the IAD framework as a potential theoretical tool for moving towards a better understanding of the issues that contribute to the wickedness of problems with social sustainability in global supply chains. The eight design principles and the main concepts stressed by the other frameworks discussed above will be help guide the analysis The assumption is that the IAD framework offers something additional through its focus on the relationships between institutions and the actors they govern, rather than solely on the actions of actors or the outcomes of those actions. Based on this assumption, we have constructed a modified IAD framework for use in analyzing supply chain social sustainability issues using data from the empirical case study example of the RMG industry in Bangladesh (Figure 1). The only modification we have made to the original IAD is to the first of the key elements in the framework: while the IAD framework includes the characteristics of the (natural) resource being appropriated, we replace this element with (1) the unique characteristics of the Bangladeshi segment of the RMG supply chain, particularly, the ‘‘size, productivity, and predictability of the resource system’’ and ‘‘the extent of mobility of the resource units’’ (Ostrom, 2011: 23). We maintain the remaining elements of the original IAD that focus on; (2) social and cultural attributes of the community in which the resource (in this case, the supply chain segment) is based; and (3) the rules-in-use in that particular segment of the supply chain, particularly those that relate to ‘‘collective-choice rules that the users may adopt authoritatively in order to change their own operational rules’’ (Ostrom, 2011: 23). It is important to note here that the rules-in-use may differ both from the rules-in-form for that segment as well as from the rules-in-use (or form) in other segments of the supply chain. These first three elements combine to affect the environment

Characteristics of the supply chain segment Participants

Attributes of the community (social and cultural context)

(Inter)actions

Outcomes Rules-in-use

Action situation

Figure 1. Framework for institutional analysis of supply chains (adapted from Ostrom, 2005). International Food and Agribusiness Management Review

275

Carlson and Bitsch

Volume 21, Issue 2, 2018

in which the participants interact; (4) the action situation, which includes the participants, the actions they take, and the outcomes these actions generate. It is important to recognize the iterative nature of this process (represented by the feedback arrows within the action situation), as participants react to the outcomes of their own actions and the actions and outcomes of other participants. The modified framework is applied to a set of empirical data about the Bangladeshi RMG industry that relate to social sustainability issues. This empirical data is derived from qualitative document analysis of existing peer-reviewed scientific studies of the RMG industry in Bangladesh and a variety of other source types. Thus, existing data about the case of Bangladeshi RMG production is used to examine the effectiveness of applying a modified version of the IAD framework to identify the key institutional relationships that are important in analyzing issues of social sustainability in supply chain situations. The approach used to collect and analyze that empirical dataset is described in the following section.

3. Methods A case study approach is used to describe the operational situation in the RMG sector in Bangladesh, to identify the external and internal motivations facing the appropriators of units of value from that resource system, as well as the outcomes of the combination of those factors in terms of social sustainability. According to the IAD framework, operational situations are ‘‘All action situations where individuals engage in the provision, production, distribution, appropriation, assignment or consumption of goods and services’’ (Ostrom, 2005: 56). A case study approach is seen as appropriate here because ‘‘the study investigates a contemporary phenomenon’’ and ‘‘seeks an in-depth understanding of the phenomenon within the framework of the actors involved’’ (Mugera and Bitsch, 2005: 82), as well as ‘‘the dynamics present within single settings’’ (Eisenhardt, 1989: 534). Based on Stake (2005) the work presented here can be considered an instrumental case study, as it has been undertaken ‘‘mainly to provide insight into an issue or to redraw a generalization’’ with regard to the problems surrounding social sustainability in supply chains. The Bangladesh RMG industry was selected as an appropriate case because it combines many of the key issues of supply chain social sustainability in developing countries and, largely due to the Rana Plaza incident, has received a substantial amount of attention from retailers, activists, scholars and governments. Thus, a rich array of existing data is available for this case. All data used to develop the case study presented here are derived from secondary sources. This includes seventeen peer-reviewed scientific articles, fifteen documents from international and Bangladeshi government agencies, five reports from private policy agencies, five from labor organizations, seven from other nongovernmental activist organizations, two from multi-stakeholder initiatives, eighteen from Bangladeshi or western print and radio news sources, web pages from two retailers and three industry associations, and five books.1 The data collection process began with qualitative coding of an initial document that provided an overview of the the RMG sector in Bangladesh (Bhattacharya et al., 2002). Based on that coding, subsequent document searches were conducted to locate additional sources to more fully describe the themes identified in the first document. New sources were coded, compared to one another and to previous sources, new themes identified and existing themes made more concrete. In addition to seeking additional information on conceptual issues, important actors and institutional arrangements identified in the coding process were further pursued through addition of new source material. In cases where multiple sources were identified with information pertinent to a given concept, sources based on interviews or surveys of actual participants in the RMG industry within Bangladesh were prioritized. Source selection and conceptual coding continued in this way until sufficient detail was reached to fill out the structure of the IAD framework. The approach to data collection and analysis described above is similar to that used in grounded theory, in that the individual sources used were examined and compared and assigned conceptual codes developed based on a process of constant comparison (Strauss and Corbin, 1998). Throughout the process, memos 1

A complete list of the sources used is available from the authors upon request.

International Food and Agribusiness Management Review

276

Carlson and Bitsch

Volume 21, Issue 2, 2018

were used to make note of interrelationships between concepts and actors and to document missing areas of information to guide further sampling. Coding of the data and memo writing were both accomplished using the qualitative data analysis software package ATLAS-ti (Version 7.5.18 , ATLAS.ti GmbH, Berlin, Germany). The benefits of using these types of software go beyond efficiency and convenience. Qualitative data analysis can serve as a way to keep track of the research process, thus helping to lend it credibility (Hwang, 2008). In addition, it helps ensure a systematic process of analyzing qualitative data by serving as both a memory aid and as a device to identify key concepts and apply meaningful codes to those concepts (Konopásek, 2008). Grounded theory as a method has been used by other authors in supply chain analysis (for a review see Denk et al., 2012). Randall and Mello (2011: 864) advocate for the use of grounded theory in the movement towards theory in supply chain management because it provides ‘‘deep insights into social phenomena’’ particularly, ‘‘the underlying meanings in human experiences, interactions, and relationships that form company strategies and actions in a supply chain context’’ and the ‘‘behavioral dimensions at the individual, organizational, and inter-organizational level’’. The results and discussion section that follows presents some of the key features of the Bangladesh RMG industry derived from the qualitative empirical analysis described above. The modified IAD framework (Figure 1) is used as a structuring device. The eight design principles will be referred to (as design principle 1, 2, etc.) throughout as well as the main concepts from the other frameworks described above where they apply.

4. Results and discussion 4.1 Characteristics of the supply chain segment The story of the modern Bangladesh RMG industry is a rather short and intense one. A key factor in the development of the RMG industry in Bangladesh was the Multi-Fiber Agreement (MFA) – a now-defunct global trade agreement (formal rule) establishing quotas on garment exports designed to protect US and EU textile and RMG industries from threats from foreign competition (Ahmed et al., 2014). This led companies from restricted countries to seek out manufacturing partners in other countries not affected by quotas. Desh Garments, one of the first entrants to the Bangladeshi industry, was thus supported by foreign direct investment from the South Korean firm Daewoo in an effort to avoid MFA restrictions (Ahmed et al., 2014). Well-connected Bangladeshis secured favorable policies from Bangladeshi government allies that facilitated initiation of RMG production with little capital investment (Yunus and Yamagata, 2014). Duty-free status of goods imported to bonded warehouses and back-to-back loans that eliminated the need to have the cash up front to buy cloth and accessories made it both cheaper and easier for Bangladeshi export-oriented RMG producers to begin to fill export orders (Yunus and Yamagata, 2014; Zohir, 2001). Thus, Bangladeshi RMG entrepreneurs, created at least partial boundaries (design principal 1) around the resource (the value generated by RMG production in Bangladesh) by effectively limiting competition from outside. In fact, with the help of formal policies originating both from outside and within Bangladesh they essentially created their own resource. These policy advantages helped the Bangladeshi RMG export industry to grow rapidly from a 1980 total of a mere 30 factories to more than 5,100 in 2011 (Yunus and Yamagata, 2014). The Bangladeshi RMG industry demonstrated an average annual growth of 12.3% over the period from 2002 to 2011 (Yunus and Yamagata, 2014). By 2008, Bangladesh was the fourth largest global exporter of RMG, with most of those exports going to the European Union (EU – nearly 60%) and the United States (US – nearly 30%) (Yunus and Yamagata, 2014). Bangladesh’s market share continued to grow, and in 2009 they became second only to China in terms of single-country RMG exports (Kathuria and Malouche, 2016). The 2012 Survey of Manufacturing Industries reported a total of 6,984 RMG manufacturers (Bangladesh Bureau of Statistics,

International Food and Agribusiness Management Review

277

Carlson and Bitsch

Volume 21, Issue 2, 2018

2013). Thus, the boundaries of the resource (the total value available for appropriation) continued to expand, as did the pool of actors appropriating value from it. Export Protection Zones (EPZs) with special rules such as duty-free imports of raw material and tax-free exports of finished products to encourage foreign investment are a common strategy used by less developed countries like Bangladesh to encourage foreign investment (Dicken, 2011: 201). As a result of such policies, Dicken (2011: 459) states, ‘‘employment tends to be geographically concentrated in the large, burgeoning cities and in the EPZs’. In the case of Bangladesh, however, while the RMG industry is concentrated largely in two major Bangladeshi cities (Dhaka and Chittagong) the majority of textile factories in Bangladesh are not located in EPZs (Yunus and Yamagata, 2014: 81) They are also overwhelmingly wholly or majority-owned by Bangladeshis, a fact that sets the industry in Bangladesh apart from that in other major RMG-exporting countries such as Cambodia and Taiwan (Rahman et al., 2007: 5). This is indicative of Bangladesh as a whole. According to the Bangladesh Bureau of Statistics, of an estimated 42,792 individual manufacturing establishments, only 263 are considered to be foreign-owned and only 160 are identified as being joint ventures between local and foreign owners (Bangladesh Bureau of Statistics, 2013). The regional concentration of largely locally-owned RMG operations should, theoretically contribute to the satisfaction of many of the design principles (design principles 1, 3, 4, 5 and 6). A geographically concentrated resource is more easily understood by appropriators, and communication among them should be less costly. Thus, self-organization to accomplish monitoring sanctioning and conflict resolution should involve fewer transaction costs. Domestic actors involved in the Bangladeshi RMG resource system appear, at least, to have a great deal of potential control of that resource. Who those actors are and what factors complicate collective action amongst them is explored in more detail in the discussion of the social and cultural attributes of community and the participants in the action situation that follows. One way in which Bangladesh does not differ from other major RMG producers is that women workers have played a key role in the development of the industry. Some argue that this is a phenomenon inherent to RMG, ‘‘the industry’s growth strategy depends on the social discrimination of women and as a consequence of this, on their major willingness to accept low qualified, low paid and, often, informal employment opportunities’’ (Ferenschild and Wick, 2004: 9). In order to evaluate the validity of this claim in the context of Bangladesh, some of the key social and cultural characteristics of the country or ‘‘community’’ in IAD terms, as well as key attributes of women as participants must be examined. 4.2 Social and Cultural Attributes of the Community The nature of the political system in Bangladesh is not conducive to good governance. Historical factors contributed to difficulties that have faced political leaders in Bangladesh from its founding. Centuries of different colonial rulers contributed to the formation of ‘‘local lords...in charge of law and order’’ (Van Schendel, 2009: 51). This system resulted in the development of ‘‘a ruling elite ‘‘separated from the masses of the population by their military and political power’’ (Van Schendel, 2009: 51). This system, which was ‘‘open to the successful adventurer’’(Van Schendel, 2009: 52) was and remains a great resource for the RMG entrepreneurs, as evidenced by the favorable policies mentioned above. Another legacy of the colonial period, is a tradition of identity politics that originated under British rulers who treated Muslims in colonial India as a unified entity – a concept which was largely false in the Bengal region given the wide variation in the religious and social practices and beliefs of the inhabitants (Van Schendel, 2009: 81-82). In fact, an important group of cultural and intellectual leaders in eastern Bengal promoted much more secular and less conservative values that were key to moving Bangladesh towards independence from Pakistan, itself a nation whose boundaries were created wholly based on religious divisions. Still, the conservative element remains strong in Bangladesh which is overall, still largely rural. The result is political competition between two ruling elites – an urban elite, which is largely represented by the Awami League (AL) party – and a rural elite, which tends to be more supportive of the Bangladesh Nationalist Party (BNP) (Rahman and Langford, 2012; Van Schendel, 2009). ‘‘The broad spectrum of conservative Muslim and right-wing religious groups

International Food and Agribusiness Management Review

278

Carlson and Bitsch

Volume 21, Issue 2, 2018

believes that Bangladesh should derive its basic national identity from its Muslim and Islamic heritage, a view not acceptable to secular parties like the AL’’ (Rashiduzzaman, 1997: 267). The power struggle between the AL and the BNP is manifested most clearly in the practice of hartal – politically-motivated country – or sector-wide strikes that have been a part of the political process in Bangladesh since before statehood (Van Schendel, 2009). Hartals continue to be a one of the main strategies used by the two parties to undermine the power of the other, as evidenced by the following excerpt from the May 16 2016 Daily Star, a newspaper published in the Bangladeshi capitol of Dhaka, ‘‘Both the major political parties – whether the ruling party Awami league (AL) or the opposition BNP – have used hartal as a major political weapon to create mass unrest in the country since 1991. If BNP is in power, the AL goes on to call hartal after hartal, and if AL is in power, the BNP gets busy calling hartals.’’ Thus, violent protest is routinely used in Bangladesh as a legitimate means to enforce or resist more formal institutional arrangements. The question of whether this can be considered to represent the presence of adequate conflict mechanisms (design principle 6) perhaps depends on whether one looks at outcomes or process. As will be discussed below, similar violent means have been the most effective tool for RMG workers to secure promised wage increases. It is difficult to see it as movement towards the social cohesion and trust-building seen as key to moving towards more socially sustainable supply chains (Beske et al., 2014). Election results in Bangladesh are also strongly affected by the sort of activity described above. ‘‘It is well known that voter intimidation and vote-buying are widespread in both rural and urban settings’’ in Bangladesh (Stiles, 2002: 839). In order to reward the often violent and oppressive acts perpetrated on behalf of the party, both parties are known to follow a system of patronage politics. ‘‘No prominent leader or party in Bangladesh has been free from the taint of favoring activists who had committed criminal or violent offenses’’ (Rashiduzzaman, 1997: 261). ‘‘Rent seeking is an equilibrium outcome in Bangladesh and acknowledged by all economic actors in the garment industry: owners, labor, and foreign buyers. According to these stakeholders, corruption is a predictable and manageable cost of doing business’’ which is only offset through the maintenance of low labor costs (Ahmed et al., 2014: 265). Thus, violence, corruption and intimidation are among the most important resources used to obtain and maintain political power in Bangladesh. This extends to so-called civil society. As most Bangladeshi unions are seen as connected to one political party or another, (Ahmed et al., 2014; Fair Wear Foundation, 2013), patronage extends to top union positions as well. The resulting weakness of the government and perceived illegitimacy of unions translates into weak enforcement of labor laws, building and fire safety codes, and a general distrust of unions among workers, RMG entrepreneurs and the general public (Ahmed et al., 2014). This distrust in unions has a significant impact on the effectiveness of collective action efforts organized by international unions and NGOs who tend to insist on the importance of formal rules in support of union formation and recognition. Thus, effective collective choice mechanisms (design principle 3) do not exist for addressing key social sustainability issues in Bangladesh, and the ones that enjoy the strongest advocacy do not appear to fit well to the local context. This does not support Cuthill’s (2010) notion of engaged governance or Beske et al. (2014) concept of collaboration among actors. Another key cultural component of Bangladesh is its strongly patriarchal society in which women were traditionally expected to follow the will of a male protector (Kabeer and Mahmud, 2004). This normative structure had already begun to devolve when the RMG industry began to take hold in Bangladesh. Large-scale death and imprisonment of men during the Bangladeshi war of independence from Pakistan and subsequent labor migration of men to other countries has increasingly left women to fend for themselves or replace the missing men as contributors to family income (Kabeer and Mahmud, 2004; Van Schendel 2009). This highlights the dynamic nature of institutions in any given setting. The robustness of an institution refers to its ability to recover to some new equilibrium after external shocks. In response to the dual threats of poverty and lack of male protection and new opportunities for wage employment, young rural women migrated in large numbers to urban centers where RMG production is concentrated. Traditional norms requiring women to stay in their fathers’ or husbands’ homes were relaxed due to external economic pressures. The informal International Food and Agribusiness Management Review

279

Carlson and Bitsch

Volume 21, Issue 2, 2018

institutions governing women’s behavior began to change. The result was in an average annual increase in the female labor force of 10% in the period from 1974 to 1999/2000, (the total increase in labor participation was only 3% during this same period) (Murayama, 2006: 65). This dynamic has served to create more social capital for women and dramatically altered the set of informal, but still monitored and enforced rules and norms that govern the possible actions women are able to take in Bangladesh. Still, as wage earners women have become legitimate appropriators of value from the RMG resource system, with rights and responsibilities associated with their position. A closer look at some of the rules-in-use in that system will shed more light on how they affect female workers and other actors differently. 4.3 Rules-in-use Despite the proliferation of supplier codes of conduct mandated by western brands and retailers that purport to guarantee the rights of workers to collective bargaining, legal minimum wages, a limited number of working hours per week and guaranteed days off (Huq et al., 2014), workers report excessively long shifts, lack of employment security, threats of layoffs and factory closings due to increased expenses from code compliance, and verbal and physical and abuse at the workplace (International Labor Rights Forum, 2015). Thus, the rules-in-form appear to have little legitimacy or explanatory power in Bangladesh. One study confirmed the ‘‘rent-seeking equilibrium outcome’’ extends to RMG factories in that RMG factory owners and managers admit to widespread cheating with regard to codes of conduct, and representatives of western buyers acknowledge their complicity, stating that it is part of their strategy of nurturing mutually trusting relationships with suppliers (Huq et al., 2014). Various elements related to trust and long-term-relationship building are identified in all of the frameworks described above. Between at least some participants in the Bangladeshi segment of the RMG supply chain, building trust is, in part, based on mutually agreed upon rules-in-use that are directly oppositional to externally-imposed rules-in-form – collaborative cheating. Among the arguments given for this is that the rules do not suit the needs of the workers themselves. Decreased working hours due to externally imposed codes of conduct reportedly serve only to increase work loads and production speed expectations (Action Aid, 2015). This violates the proportional equivalency principle (design principle 2) and thus detracts from the legitimacy in the eyes of those subjected to the rule. The appropriators themselves break it as it only serves to increase the share of the value they get for their effort. More important avenues for positive change for workers are good behavior of management and on-time payment of wages (Awaj Foundation, 2013). That the principles of equity and social justice identified in most frameworks are clearly not present in Bangladeshi RMG factories is clearly illustrated by the position of women there. While women are a key component of the RMG workforce, female workers are paid less than their male counterparts, are seldom promoted to positions of power over other workers, and are subject to sexual harassment both on the way to work and on the job (Kabeer and Mahmud, 2004). As participants in the action situation (and appropriators of the resource), then, they are still subject to different rules than their male counterparts. Some report having full control over their wages, while others report handing them over to a husband or father. While familial rules like these are assigned, monitored and enforced outside the operational situation being analyzed, they affect it in that they limit the opportunities women have elsewhere, and thus, the set of choices women would even conceive of within the action situation itself. This hints at the nested sets of action situations and associated institutional configurations to which appropriators are subjected (design principle 8). Still, the large concentration of women working and living together in close conditions could be expected to generate new forms of social capital amongst those women that could be leveraged to appropriate more value. In general, though, collective choice arrangements (design principle 3) involving RMG workers in Bangladesh are absent. For example, while the legal basis for trade unions has always been present in Bangladeshi law, overall trade union participation of the labor force in Bangladesh remains small (Al Faruque, 2009). ‘‘The highly fractionalized nature of the union movement in Bangladesh along with government neglect of workers’ organizations has stunted the growth of a significant organized labor movement. Moreover, the nature of the political competition between the AL and the BNP does not require the two parties to seek the support International Food and Agribusiness Management Review

280

Carlson and Bitsch

Volume 21, Issue 2, 2018

of organized labor in order to form a winning coalition’’ (Ahmed et al., 2014: 266). Political appointments to union leadership positions further capture and dilute the potential power of unions as agents of collective action in Bangladesh. Left no formal means to register their grievances (design principle 6), the most effective tool of RMG workers in Bangladesh appears to have been labor unrest. It was due to labor unrest in the Bangladesh RMG industry in 1994, that a committee including three representatives each from RMG workers, RMG employers and the Bangladeshi Government was formed to try to negotiate (Al Faruque, 2009: 23). Though workers’ representatives on this committee reportedly made numerous efforts to use the labor committee structure to improve things for workers, the industry continued to violate each agreement, resulting in sporadic, factorylevel protests, some of which were negotiated by factory-level unions and some mediated by local NGOs (Rahman and Langford, 2012: 97). Ultimately, these protests contributed to changes in the formal rules applicable to RMG working conditions embodied in the Bangladesh Labour Law of 2006. The Bangladesh Labour Law of 2006 consolidated ‘‘as many as 25 separate legislations’’ pertaining to labor and working conditions that had previously been in place in Bangladesh (Rahman, 2015: 66). Key features of the 2006 legislation include specific provisions that prohibit children under age 14 from working in most settings and restrict the number of hours that both adolescents (aged 14-18) and women can work, mandate the provision of worker identity cards and letters of appointment, provide death and maternity benefits for workers, and compensation to family members in case of death or disability (Rahman, 2015: 66). Thus, they assign rights and prohibit certain activities for particular subsets of potential resource appropriators. Although collective bargaining rights for workers are included in the law, for a factory union to be eligible for registration with the Bangladeshi government and legally represent workers to the Ministry of Labour and Employment in Bangladesh, 30% of all workers in the establishment must be members (Rahman, 2015: 68). In, addition, the law stipulates that a worker can only be a member of the union as long as he or she remains employed by the establishment at which the union is housed; thus, fired workers are immediately out of the union even if they hold office in it (Rahman, 2015: 68). A common tactic of employers is to dismiss a worker for misconduct, thereby making that worker ineligible to remain a trade union officer (Rahman, 2015). While a worker can complain to the Labour Court, they face problems of underlying corruption at the courts and serious backlog of cases which, in some instances, can stretch back many years (International Trade Union Confederation, 2008). Thus, low-cost conflict resolution venues (design principle 6) are not present for violations of formal labor laws in Bangladesh. The effective absence of these formal rules helps support the development and perpetuation of informal rules that actually govern the treatment that RMG workers, union members and organizers in RMG factories receive both from management within the factory (intimidation, threat of firing, and increased workloads), as well as in the community at large (visits at home from hired thugs and threats to family members) (Action Aid, 2015). 4.4 Action situation, participants, interactions and outcomes This section presents some of the key participants in the RMG action situation in Bangladesh and identifies some of the most important institutional factors affecting their actions. The goal is to help illustrate both the different sets of incentives facing individual actors based on the formal and informal rules that govern their actions, and how those incentives are continually being altered by the iterative process of interactions with other participants and outcomes. ■■ Participants: multi-stakeholder initiatives unique to Bangladesh Largely as a result of the sense of urgency brought about by the Rana Plaza incident, the Bangladesh Accord on Fire and Building Safety (the Accord) was signed in May 2013 between the Industrial global trade union, four NGOs, and over 200 (largely European) garment brands and retailers, led by Hennes & Mauritz (H&M) (Reinecke and Donaghey, 2015). The Accord is considered unique, due both to its contractual nature that ostensibly holds western corporations legally accountable for fire and building safety issues in factories International Food and Agribusiness Management Review

281

Carlson and Bitsch

Volume 21, Issue 2, 2018

owned and managed by independent RMG suppliers, and because unions were involved in its creation. This ostensibly new, but still largely top-down approach is a 5-year commitment that includes detailed inspections for building and fire code violations in the RMG factories in Bangladesh from which its members source, online publication of factory locations and results of inspections, and mandatory remediation of the problems identified in those inspections (Bangladesh Accord, 2013). Due to the poor governance structure of Bangladesh described above and the power of the RMG industry in Bangladesh, implementation of the Accord has been slow, and few factories have received failing marks. Repairs that have occurred have been financed by the local RMG suppliers, who have largely eschewed financial help from the western brands who signed the Accord agreement (Daily Star, 2016a). More significantly, in several cases, the Bangladeshi government has reportedly refused to force RMG entrepreneurs to comply with Accord inspector recommendations to immediately close factories found to have serious safety issues (Greenhouse and Manik, 2014). Still, the Accord has been referred to as a ‘‘significant new departure in global supply chain labour governance’’ (Reinecke and Donaghey, 2015). This, however, is based on the perception of increased accountability through existing formal institutions in Bangladesh, whose already inadequate resources have been only partially upgraded to deal with the increased work (International Labor Rights Forum, 2015). ■■ Participants: industry organizations The Alliance for Bangladesh Building and Safety (Alliance) is a Wal-Mart-led initiative with goals and procedures that, at least on the surface, appear to be nearly identical to those of the Bangladesh Accord (Alliance for Bangladesh Worker Safety, 2016). The big difference is that the Alliance is purely an industry group that includes 26 largely US-based brands and retailers who declined to sign the ostensibly legally binding Accord (Reinecke and Donaghey, 2015). Thus, these participants have chosen to exclude themselves from the potential additional monitoring and sanctions and associated conflict resolution mechanisms the international legal level (design principles 4, 5, 6) for conditions in their producer factories in Bangladesh. Some have reported Accord member factories to be more compliant and Alliance members more tolerant of old abuses (International Labor Rights Forum, 2015). As many Accord factories are also Alliance factories, Alliance members have the potential to free-ride on the efforts of the Accord. For their part, Accord signees cannot exclude Alliance members (or others) from the benefits of any improvements they are able to make. Thus, Accord signatories could be acting as resource providers for appropriators with whom they are in competition. This does not seem like a sustainable solution. Bangladeshi RMG suppliers are were the initial providers of the resource and continue to play a major role in its maintenance. Also, as these agreements were worked out largely amongst actors from outside of Bangladesh, the quality of involvement of local actors (as referred to by Beske et al., 2014; Boström, 2012; Cuthill, 2010) is low, undermining legitimacy still further. RMG workers (except through indirect representation by the global union and the government inspectors) are not able to participate in rule-making, monitoring and holding monitors accountable, or sanctioning (design principles 3, 4, 5 and 6 are violated). The importance of RMG as a driver of development, employer of unskilled workers and contributor to labor income has yielded RMG producers a great deal of power in Bangladesh. This power is institutionalized in the Bangladesh Garment Manufacturers and Exporters Association (BGMEA), which quickly assumed a quasi-governmental role, for example, in assigning Bangladesh’s MFA quotas to producers (Yunus and Yamagata, 2014). In this way, the BGMEA has a great deal of control over the boundary of the RMG resource system in terms of who is allowed to extract value from it. Knitwear producers in Bangladesh formed their own, related organization in the Bangladesh Knitwear Manufacturers and Exporters Association (BKMEA), largely in response to the threat of changes in the EU General system of preferences (GSP) country of origin (rules-of-origin restrictions) related to knitwear garments that threatened quota-free status of Bangladeshi knit goods (Yunus and Yamagata, 2014: 81). Thus, even a threatened formal rule change can generate collective efforts to deal with that change and result in new coalitions of resource users with their own unique agendas for influencing rules.

International Food and Agribusiness Management Review

282

Carlson and Bitsch

Volume 21, Issue 2, 2018

The power of the BGMEA and BKMEA is augmented by the fact that a large amount of research into the garment industry in Bangladesh is accomplished based on lists of companies and other data supplied by one or both of these two industry groups (Action Aid, 2015; Yunus and Yamagata, 2014; Zohir, 2001). Thus, they control a large amount of information as it flows both up and down the supply chain. Up to now, the BGMEA and its member entrepreneurs have been able to resist efforts from western brands, unions, workers and activists to significantly change the rules-in-use with regard to labor conditions and wages. This can be attributed to their control of important social capital and social infrastructure (Beske et al., 2014; Cuthill, 2010; Murphy, 2012). ■■ Participants: large retail buyers Buyers from the Bangladesh RMG industry include the world’s largest volume retailers, including major US buyers like Wal-Mart, Gap, K-mart, Sears, J.C. Penney, and Levi Strauss; and the main RMG buyers from the EU, including C&A, H&M, and Marks and Spencer (Kabeer and Mahmud, 2004: 143). By extension, then, their customers (the consumers of their goods) are also appropriators of the value in the RMG supply chain. Due to the continued social costs that arise from this relationship, foreign appropriation of the value provided by the Bangladeshi RMG resource system is considered by some to be significantly more than their contribution to the costs, violating many of the key elements of social sustainability (design principle 2; equity: Murphy, 2012; social justice and equity: Cuthill, 2010). The largest single buyer of RMG from Bangladesh, with annual purchases reaching five billion dollars in 2016 (Daily Star, 2016b), is H&M who continue to make attempts to try to better incorporate these costs into their own rule systems for governing supply chains. Having established a supplier code of conduct in 1997, H&M opted in February of 2016 to replace the code with a ‘‘Sustainability Commitment’’ which states, ‘‘We want to make sustainability an integrated part of all our global business relationships and work closely with our suppliers and business partners in order to achieve long lasting impacts across our entire value chain’’ (H&M, 2017). Given the large amounts of product H&M sources from Bangladesh, events like Rana Plaza represent a massive threat to such a carefully cultivated reputation. Thus, despite the fact that H&M was not named as one of the companies sourcing from any of the factories at Rana Plaza, it was the first official signatory to the Bangladesh Accord on Fire and Building Safety (Reinecke and Donaghey, 2015). H&M seized the opportunity provided by the Rana Plaza collapse to try to gain more control over its supply chain partners in Bangladesh, both through stakeholder pressure and financial means, and in so doing, increase brand reputation. This increase in social capital can be considered additional value that H&M hopes to extract from the Bangladeshi RMG resource system, offset by the costs of additional monitoring and sanctions. Despite its proactivity, the focus of H&M’s efforts still seems to be on enforcing the formal rules at work in Bangladesh, rather than trying to better understand and affect changes in the more important informal rules at work there. Another large retail actor in the Bangladeshi RMG industry is Wal-Mart – the single largest US-based buyer of clothing from Bangladesh (Human Rights Watch, 2015; Ullah Mirdha, 2016). Wal-Mart has been widely criticized for unfair labor practices in its own retail and distribution systems, as well as in its supply chains (Bloomberg, 2003; Kolben, 2007). In connection with its US operations, Wal-Mart has been accused of many of the same violations deemed problematic in the Bangladesh RMG industry, including wage and hour violations, employment discrimination and locking employees in stores overnight (Greenhouse, 2004; Kolben, 2007). One of the biggest criticisms of Wal-Mart, however, is its animosity towards unions. ‘‘Wal-Mart’s anti-union animus is evident at the beginning of the employee’s tenure when she receives the employee handbook, which provides that we are not anti-union; we are pro-Associate. It is our position that every Associate can speak for himself or herself without having to pay hard-earned money to a union in order to be listened to and have issues resolved (Kolben, 2007: 286).’’

International Food and Agribusiness Management Review

283

Carlson and Bitsch

Volume 21, Issue 2, 2018

As a retailer of their own branded merchandise, Wal-Mart, like H&M, does not possess any manufacturing systems of its own. However, due to its size and buying power, Wal-Mart’s activities have impacts across a large number of related industries in its supply chain and those of competitors. By forming its own coalition of buyers in response to Rana Plaza, Wal-Mart is attempting to capture some of the value of the social capital H&M is reaping from their proactivity (Beske et al., 2014), while significantly limiting the costs of capturing that value by avoiding sanctions, monitoring and conflict resolution mechanisms associated with the Accord agreement (design principles 4, 5, 6). ■■ Participants: Bangladeshi ready-made garment entrepreneurs The global shift in production of goods to less-developed countries has already been touched on above. What makes the RMG industry unique is its focus on fast fashion – in which the strategy for profit-making is based on the concept of convincing consumers to buy cheaper clothing more often. ‘‘The fast fashion industry owes its special features and structure to its requirements for shortened lead-times, faster inventory turnovers and high order fulfillment rates for customer demand at its peak points’’ (Turker and Altuntas, 2014: 838). This places RMG suppliers in the position that raising their prices even slightly or not meeting production agreements threatens the loss of production contracts. Buyers report being more concerned with things other than workers’ rights and working conditions that could have negative effects on the economic efficiency of sourcing from Bangladesh, such as unreliable electricity supply and corruption (Berik and Rodgers, 2010: 68). Bangladesh’s competitive advantage over other potential RMG producers is mainly considered to be the low labor cost, with costs of the poor governance system (corruption and inefficiency of infrastructure) eating up some of the value that could potentially otherwise be appropriated by workers (McKinsey and Company, 2011). Provision of the resource in terms of infrastructure, both physical and social (Cuthill, 2010) is not being accomplished well under the current rule system. This provision problem is evidenced in some of the ‘‘new’’ arrangements that have been made since Rana Plaza. The financial arrangements offered through the Bangladesh Accord and Alliance, require not only a commitment from buyers to RMG producers to make up some of the costs of providing the resource, but also a commitment from the RMG producers to give up some of their autonomy. By taking the financial assistance offered through these agreements, RMG entrepreneurs might potentially be giving up a significant amount of control that, up to now, they have not wanted or needed to cede due to the acceptance of the rules-in-use. By forming a coalition with organized labor and creating a new set of rules that is binding under both legal and financial contracts, Accord-signatory buyers have attempted to change the rules-in-form to give them more control over facilities where value is being created and appropriated. However, as in the case of the Accord, none of the entrepreneurs offered financial help from the Alliance partners have chosen to take it (Butler, 2013). Cooperation from RMG entrepreneurs in terms of making necessary repairs has also not been promising. According to the Third Annual Report from the Alliance published in October of 2016, ‘‘To date, the Alliance has suspended 97 factories for failure to make progress on repairs that address safety concerns’’ (Alliance for Bangladesh Worker Safety, 2016: 9). It seems this collective effort has failed to create sufficient social capital or engaged governance (Cuthill, 2010) to create effective monitoring and sanctioning mechanisms (design principles 4 and 5) in RMG factories in Bangladesh. Though collaboration is clearly present (Beske et al., 2014), the effects of this collaboration are negated by other factors. A look at how RMG workers are included in these and other collective action agreements within the RMG supply chain may shed light on why. ■■ Participants: ready-made garment workers in Bangladesh While most accounts cite figures of 4 million people employed in the industry, 80% of whom are women (Fair Wear Foundation, 2013; International Labor Rights Forum, 2015; Reinecke and Donaghey, 2015), the BGMEA is consistently the source of this information. According to the 2012 Bangladesh Census of Manufacturing Industries, however, the industry employs 2,762,335 workers, 1,257,464 (ca. 45%) of whom are men and 1,504,871 (ca. 55%) of whom are women (Bangladesh Bureau of Statistics, 2013).

International Food and Agribusiness Management Review

284

Carlson and Bitsch

Volume 21, Issue 2, 2018

Several studies provide information as to the characteristics of RMG workers in Bangladesh that are helpful in understanding the way the rules-in-use affect the attributes they possess with regard to appropriating value from RMG production. Two recent surveys of RMG workers were conducted by the Awaj foundation, a Bangladeshi NGO whose stated purpose is ‘‘giving a voice to Bangladesh’s Women Garment Workers’’ (Awaj Foundation, 2017). The first survey, conducted in 2013, assessed 1215 garment workers from more than 250 factories, 69% of whom were women. The second survey in 2015-16 covered 1007 garment workers, 85% of whom were women. In contrast to the frequent characterization of RMG workers as largely ‘‘illiterate’’, slightly over 17% of respondents in the 2013 survey had not attended school at all, while nearly a quarter of those surveyed had attended school for 9 years or more. The 2016 report states, ‘‘The percentage of workers without even one year of formal schooling decreased from 17% in 2013 to 6% in 2016, which is very low compared with the literacy rate of the general population of Bangladesh (61.5%)’’ (Awaj Foundation, 2016: 19). In a study by Rahman et al. (2007:6), only 20% of the workers surveyed were considered ‘‘unskilled’’, while 44% were assigned to the ‘‘skilled’’ category. Most had at least primary education, with men being generally more educated than women. Skill level does not seem to have a high impact on wages, as one study showed that 80% of those surveyed earned between 3,000-4,999 Taka (38-62 USD) per month, independent of education level (Awaj Foundation, 2013: 41). It seems from these figures that the social capital available to RMG workers (at least in terms of skills and education) is growing. Increased literacy will also bring increased access to information, another key element mentioned in some way by the authors all of the supply chain social sustainability frameworks described above. Rather than focusing on long-term relationships with their employers, Bangladeshi RMG workers choose to move frequently among factories (Awaj Foundation, 2013: 38-39; Murayama, 2006: 86). This can be attributed to the potential gains workers have by practicing this strategy of mobility, ‘‘the more different factories the participant has worked in, the bigger are her/his chances to earn a higher salary (Awaj Foundation, 2013: 44). Thus, despite the seeming insecurity of jobs with no formal contracts or identity cards, it seems that RMG workers have devised a strategy to make the relative fluidity of the employment arrangements work in their favor by moving about frequently in search of better pay. This strategy, however, does not appear to be conducive to fulfilling long-term relationships, continuity or social capital in efforts to build trust. In questioning workers about what they like best and dislike most about their current work place, the number one positive response (from 27.5% of those surveyed) was a ‘‘good factory environment’’ and the number one dislike (26.6% of respondents) was ‘‘bad behavior of management and supervisors’’ (Awaj Foundation, 2013: 50-52). The next important positive responses were ‘‘timely salary payment’’ (23.9%) and good behavior of management and supervisors’’ (10.4%). Among the most common dislikes were ‘‘violation of leave facility’’ (12.9%) and delayed salary and overtime payment (9.8%) (Awaj Foundation, 2013: 50-52). These responses seem to indicate that workers are indeed concerned about how they are being treated, and that not only the economic value of the resource is important to them. The report goes on to state, ‘‘Interestingly, aspects that are usually high on the agenda of the media, NGOs and other pressure groups, are not significantly reflected in the interviewees’’ expressions of what they dislike. None of the workers mentioned ‘‘issues of worker participation, e.g. in form of worker representation committees and health and safety committees or by organizing in trade unions’’ (Awaj Foundation, 2013: 54-55). This may be attributable to the general distrust of unions in Bangladesh described above, and points to a need to develop other, more legitimate, means of supporting worker participation in creating, monitoring and sanctioning behavior in Bangladeshi RMG factories. ■■ Participants: Bangladeshi women in their role as ready-made garment workers Women have been able to increase their social capital (what Cuthill (2010: 366) calls ‘‘a theoretical starting point for social sustainability’’) through wage work in the RMG industry in Bangladesh, but are still subject to constraints from the male-dominated society of Bangladesh. Thus, the aspects of social justice and equity (Cuthill, 2010; Murphy, 2012; design principle 2) are not present within the operational situation in this instance. Still, in interviews, women in the RMG industry in Bangladesh reported having greater power of International Food and Agribusiness Management Review

285

Carlson and Bitsch

Volume 21, Issue 2, 2018

choice about life decisions, such as marriage age and partner, than they would have had before beginning garment work (Kabeer and Mahmud, 2004). Many stated that they chose to seek work in the garment industry despite objections from their male family members. ‘‘Almost all the female workers in the 1990 garment industry survey reported that the decision to work in the garment industry was their own and in many cases it was done in opposition to the wishes of the male household head’’ (Zohir, 2001: 50). Thus, their new position as appropriators of the RMG resource is contributing to continued change in the informal rules and norms that women face in Bangladesh, once again highlighting the iterative and dynamic nature of the relationships between actors and institutions. ■■ Interactions and outcomes In 2006, beginning with independent protests at two RMG factories in Dhaka over unpaid wages, violent suppression by police generated enough anger in the RMG worker community to fuel a mass violent protest movement over a period of days between May 10 and May 23 (Rahman and Langford, 2012). The entirety of Dhaka City and its adjacent areas became a battlefield because the workers were extremely militant and enraged, and as a result normal life ground to a standstill. On 23 May approximately 16 factories were burnt, 50 factories were vandalized, 200 vehicles were ransacked and a worker was killed (Rahman and Langford, 2012: 90). These actions resulted in the Tripartite Memorandum of Understanding in which the labor committee sat down with the workers to negotiate an agreement to settle the dispute (Rahman and Langford, 2012: 90). According to Rahman and Langford (2012: 90-91), this seemingly spontaneous show of solidarity by mass numbers of RMG workers had three major impacts, (1) RMG entrepreneurs saw a real physical threat to their establishments; (2) Bangladeshi government officials sat down to negotiate in earnest with RMG worker representatives for the first time; and (3) a minimum wage board was formed that resulted in an increase in wages by 75%. Thus, changes in formal rules were made due to workers’ violent protests. Other issues were also agreed upon in these meetings, many of which were eventually incorporated into the Bangladesh Labour Law of 2006. However, despite these ‘‘achievements’’, many RMG entrepreneurs did not increase wages, and in response to further inaction from the Bangladesh government, violent worker protests began again in 2010. ‘‘From January 1 to June 30, 2010, there were an estimated 72 incidents of labor unrest, leaving at least 988 workers injured in clashes with police’’ regarding key issues related to worker mistreatment by management, non-payment of wages, factory closings without notice or payment of wages due, and not allowing leaves and holidays as required (Claeson, 2010: 20). Economic forces external to the RMG sector contributed to the unrest. ‘‘During the past two years, as workers have seen their meager earnings eroded by double-digit inflation, protests and violent clashes with the police have become increasingly common’’ (Yardley, 2012). Perhaps not surprisingly, given the rather violent nature of the collective action undertaken by RMG workers, government authorities stepped in to stop it (design principle 7). In fact, the Bangladeshi government seems to have done little to support workers but talk, ‘‘the Minister of Labor urged garment factory owners to pay all wages and payments owed to workers’ (Claeson, 2010: 20). However, as shown by the worker survey results outlined above, many workers (particularly women) were still not receiving the minimum wage agreed on by 2013 when the Rana Plaza collapse occurred. Again, formal rule changes brought about by violent protest were not sufficient to alter the rules-in-use. In the wake of Rana Plaza, the Bangladeshi government entered into an agreement known as the Sustainability Compact with the EU, US, ILO and several other national government actors. As part of this agreement, the Bangladesh Labour Act of 2006 was amended to increase worker’s collective bargaining rights and workplace safety. Key changes were the elimination of a provision that had mandated the Bangladesh Government’s Registrar of Trade unions to provide the names of the officials of trade unions that have applied for registration to employers. (Human Rights Watch, 2015) However, NGOs, workers and labor activists report that this is still common practice. Once again, changes in rules-in-form did little to alter rules-in-use.

International Food and Agribusiness Management Review

286

Carlson and Bitsch

Volume 21, Issue 2, 2018

Unions have, to date, not been able to successfully overcome the constraints on organizing RMG workers created by the strategies of powerful political and social elites in Bangladesh and the BGMEA. The Accord agreement created a new formal alliance between global unions and Western brands that may prove to be an effective enough strategy to alter some aspects of these power relationships that are based on informal rules. However, the RMG entrepreneurs’ strategy of refusing to comply with or benefit from Accord policies seems to indicate that they intend to continue their past successful strategies of relying on their political clout within Bangladesh and resistance to relinquishing power to western brands and local and global unions. The Bangladeshi government seems to be backing them up, as evidenced by their refusal to mandate the closing of factories deemed as unsafe by Accord and Alliance inspectors. The distrust of unions by nearly all parties in Bangladesh may also be a hindrance that requires the creation of new types of worker representation that can more effectively negotiate between workers and RMG entrepreneurs.

5. Conclusions The IAD framework, developed by institutional economists based on extensive research into public good provision and common pool resource management, provides a structure for examining institutional incentives facing individual decision makers about appropriation of value from a common pool resource. The present study has used of a modified IAD framework to apply existing theoretical concepts to a case study in supply chain social sustainability, and therefore, has shown the IAD framework to be an effective tool to organize the key variables in these dynamics at the operational level. Using information gained from document analysis to identify the key institutions governing economic and social processes in the RMG industry in Bangladesh, provides a greater understanding of the reasons behind the continued lack of progress towards social sustainability in this segment of the RMG supply chain. The failure of attempts to promote social sustainability through corporate codes of conduct and reforms in Bangladeshi government policy indicates that other, less formal, institutional arrangements, such as norms and rules-in-use, are more important in determining the actions of particular actors in the Bangladeshi RMG supply chain. Based on the eight design principles associated with the IAD framework, the existing institutional arrangements have been analyzed in the context of the local community and related to the specific circumstances of the supply chain segment. The analysis has shown that Ostrom’s design principles incorporate most of the main elements identified in the social sustainability frameworks proposed by other authors. While environmental issues associated with RMG production in Bangladesh have not been touched on here, the original framework is specifically designed to make the institutional linkages between environmental, social and economic pillars of sustainability explicit, thus addressing one of the key critiques put forth by Boström (2012). One main element that the IAD approach adds to the other frameworks discussed here is that the institutional arrangements serve themselves as indicators of the potential for long-term sustainable resource management. This addresses what Hutchins and Sutherland (2008) see as a key challenge in efforts to promote sustainability, namely operationalization of the key concepts it entails. Another is the way that the IAD-based institutional approach incorporates the sociocultural and political environment in which the supply chain segment operates directly into the analysis. Boström’s (2012) assertion that special attention must be paid to who is allowed to be involved in rulemaking in supply chains is also specifically addressed by the IAD framework and associated design principles. Application of the IAD framework to other locations and other industries has the potential to identify areas with similar constraints on the action situation in which participants have been able to successfully change the dynamic internal to the action situation to create actual improvements in social sustainability. To that end, the results of the analysis of RMG production in Bangladesh show that particular attention should be paid to the existence of monitoring and sanctioning that is visible to all resource appropriators. To better understand how information is transmitted through the supply chain, the rules-in-use should be examined rather than focusing on the rules-in-form, as is common in supply chain sustainability research and policymaking. Policies that support collective action at the local level, rather than ones that prescribe top-down solutions where monitors are not directly accountable to appropriators (particularly workers, International Food and Agribusiness Management Review

287

Carlson and Bitsch

Volume 21, Issue 2, 2018

but also factory owners, managers and consumers) seem more appropriate than those where monitors are accountable primarily to RMG buyers and consumers, at least in the Bangladeshi context. Other sectors in similar settings should be examined, in which labor issues are also a sustainability topic to see if this applies in those contexts as well. As Maloni and Brown (2006: 43) pointed out, there are many parallels between the food industry and RMG, ‘‘labor and human rights also present a complicated issue in the food industry, potentially exposing the supply chain to the same reactions and protests experienced by the apparel industry’’. Thus, food and agricultural supply chains with important segments located in countries with similar histories, cultures and political arrangements could be expected to behave in similar ways to the RMG industry in Bangladesh. Comparisons among supply chains for different products in the same setting using an institutional perspective could provide additional insights into the unique characteristics of that resource, the community in which it is based and the rules-in-use in that context, and help move existing supply chain sustainability theory forward. Studies designed to capture the effects of similar social, economic and political dynamics between key actors in fruit and vegetable export supply chains could be especially useful in this regard, particularly in other developing country settings where research into supply chain sustainability has been surprisingly scarce (Seuring and Gold, 2013). The role of factory managers and line supervisors in perpetuating some of the less tangible problems in the Bangladeshi RMG industry, such as verbal and physical abuse and sexual harassment has been alluded to in many worker interviews as well as by factory owners. (International Labor Rights Forum, 2015). Still, little is known about these actors, who they are, where they come from, and what incentives they face to maintain the system of ill-treatment of factory workers documented above. Such mid-level managers are an unknown quantity in research into supply chain problems, because they are not accessible to researchers (Chan and Siu, 2010), or because they are simply not included in analyses (Bitsch and Yakura, 2007). Despite extensive efforts during the theoretical sampling process, no suitable sources were found to help shed more light on these particular participants in the Bangladeshi RMG action situation. With regard to Bangladeshi RMG workers, finding ways to help them self-organize that do not rely on unions may be more effective than efforts up to now that assume traditional unions are the most appropriate. The analysis has shown that the rules-in-use and the characteristics of the community combine to render union negotiations ineffective as either a conflict resolution method or a means through which RMG workers can participate in rule-making and enforcement. The characteristics of the resource, such as its geographic concentration and the relative homogeneity of its ownership should be an advantage to workers and those trying to provide them with tools and information that would support that effort. Development of innovation collective choice mechanisms that take advantage of those characteristics and take local conditions into account is needed. Given the importance of rules-in-use (rather than rules-in-form) as well as their relative invisibility, in-depth analysis of these rules requires field research. This presents new challenges, however. As Ostrom (2011) points out, rules-in-use are often implicit understandings that change almost imperceptibly. The limited (but expanding) set of choices available to women in Bangladesh provides an example of rulesin-use that continue to change and evolve in relationship to changes in other elements of an action situation. The tacit nature of this sort of local knowledge makes it especially difficult for outside observers to identify and understand, in part because those who do understand it consider it common knowledge and thus, do not verbalize it. Thus, further research into the Bangladeshi RMG supply chain (or other labor-intensive chains) should be based on prolonged engagement using in-depth qualitative methods such as participant observation and unstructured interviews (preferably conducted in the native language of the participants). The major theoretical contribution of this effort was to show that looking at social sustainability issues as a resource appropriation problem using an institutional framework to help organize theoretical concepts from more traditional supply chain and development approaches can yield additional insights into the difficulties behind efforts to promote social responsibility. The points of leverage that are likely to be effective in each supply chain situation are different and based on the complex, dynamic relationships between the participants International Food and Agribusiness Management Review

288

Carlson and Bitsch

Volume 21, Issue 2, 2018

(individuals and groups) in particular segments of that supply chain. As all relationships are based on an agreed-upon set of rules, analyzing the actually functioning rule set (rules-in-use), rather than simply trying to understand outcomes can help identify potential points for improvement. The IAD framework helps organize those elements and the eight design principles serve as a guide to identify institutional arrangements that help or hinder movement towards social sustainability.

Acknowledgement This work was supported by the German Research Foundation (DFG) and the Technical University of Munich within the Open Access Publishing Funding Programme. Background research was conducted at the Fraunhofer-Institute for Systems and Innovations Research as part of the project ‘‘Transition in globalen Wertschöpfungsketten: Förderung der sozialen Nachhaltigkeit (SoNa-WSK)’’, funded by the German Federal Ministry for Research and Education (BMBF). The authors wish to thank Miriam Bodenheimer of the Fraunhofer-Institute for her guidance and helpful comments throughout the research process.

References Action Aid. 2015. Diversify and conquer: transforming Bangladesh into an industrialised country. Available at: http://tinyurl.com/zayugsc. Ahmed, F.Z., A. Greenleaf and A. Sacks. 2014. The paradox of export growth in areas of weak governance: the case of the ready made garment sector in Bangladesh. World Development 56: 258-71. Al Faruque, D. 2009. Current status and evolution of industrial relations system in Bangladesh (Asia-Pacific Working Paper Series). ILO Regional Office for Asia and the Pacific, New Delhi, India. Alliance for Bangladesh Worker Safety. 2016. Protecting and empowering bangladesh garment workers. Available at: www.bangladeshworkersafety.org. Arenas, D., J.M. Lozano and L. Albareda. 2009. The role of NGOs in CSR: mutual perceptions among stakeholders. Journal of Business Ethics 88(1): 175-197. Awaj Foundation. 2013. Workers’ voice report 2013: insight into life and livelihood of Bangladesh’s RMG workers. Available at: Online:awaj.info/wpcontent/uploads/2013/10/Workers_Voice_Report_2013.pdf. Awaj Foundation. 2016. Workers’ voice report. 2016: the working conditions in Bangladesh’s RMG industry after rana plaza. Available at: http://tinyurl.com/y7gt8eef. Awaj Foundation. 2017. History and background. Available at: http://awaj.info/history. [link obsolete]. Bangladesh Accord. 2013. Accord on fire and building safety in Bangladesh. Available at: http:// bangladeshaccord.org. Bangladesh Bureau of Statistics. 2013. Survey of Manufacturing Industries 2012 (SMI 2012). Statistics and Informatics Division (SID) Bangladesh Bureau of Statistics, Dhaka, Bangladesh. Bartley, T., and C. Child. 2011. Movements, markets and fields: the effects of anti-sweatshop campaigns on U.S. Firms, 1993-2000. Social Forces 90(2): 425-451. Berik, G. and Y.V.D.M. Rodgers. 2010. Options for enforcing labour standards: lessons from Bangladesh and Cambodia. Journal of International Development 22(1): 56-85. 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 152: 131-143. Bhattacharya, D., M. Rahman and A. Raihan. 2002. Contribution of the RMG sector to the Bangladesh economy. CPD Occasional Paper Series, 50(6). Center for Policy Dialogue, Dhaka, Bangladesh. Bitsch, V. 2016. Sustainability as innovation: challenges and perspectives in measurement and implementation. In: Diffusion and transfer of knowledge in agriculture, Matière à débattre et decider, edited by C. Huyghe, P. Bergeret and U. Svedin. Available at: http://tinyurl.com/ybxlyq48. Bitsch, V., and E.K. Yakura. 2007. Middle management in Agriculture: roles, functions, and practices. International Food and Agribusiness Management Review 10(2): 1-28. Bloomberg. 2003. Is Wal-Mart too powerful? Available at: http://tinyurl.com/y92vexb4.

International Food and Agribusiness Management Review

289

Carlson and Bitsch

Volume 21, Issue 2, 2018

Boström, M. 2012. A missing pillar? Challenges in theorizing and practicing social sustainability: introduction to the special issue. Sustainability 8(1): 3-14. Butler, S. 2013. One in six Walmart factories in Bangladesh fail safety review. The Guardian. Available at: http://tinyurl.com/y7se3xpq. Chen, I.J. and A. Paulraj. 2004. Towards a theory of supply chain management: the constructs and measurements. Journal of Operations Management 22(2): 119-150. Claeson, B. 2010. Enemies of the nation or defenders of human rights? Fighting poverty wages in Bangladesh. SweatFree Communities: International Labor Rights Forum. Available at: http://tinyurl.com/y98c9n73. Coe, N.M., P. Dicken and M. Hess. 2008. Global production networks: realizing the potential. Journal of Economic Geography 8(3): 271-295. Cuthill, M. 2010. Strengthening the ‘‘social’’ in sustainable development: developing a conceptual framework for social sustainability in a rapid urban growth region in Australia. Sustainable Development 18: 362-373. Dae-oup, C. and M. Wong. 2005. After the consumer movement: toward a new international labour activism in the global garment industry. Labour, Capital and Society 38(1/2): 126-155. Daily Star. 2013. A country of hartals. Available at: http://tinyurl.com/y7z57g4e. Daily Star. 2016a. Garment factory reforms halfway through. Available at: http://tinyurl.com/yao5qb4n. Daily Star. 2016b. H&M to buy more garment items from Bangladesh: official. Available at: http://tinyurl. com/y9k9sg7e. Denk, N., L. Kaufmann and C. Carter. 2012. Increasing the rigor of grounded theory research – a review of the SCM literature. International Journal of Physical Distribution and Logistics Management 42(8/9): 742-763. Dicken, P. 2011. Global shift: mapping the changing contours of the global economy. Sage, London, UK. Egels-Zandén, N. and H. Lindholm. 2015. Do codes of conduct improve worker rights in supply chains? A study of fair wear foundation. Journal of Cleaner Production 107: 31-40. Eisenhardt, K.M. 1989. Building theories from case study research. Academy of Management Review 14(4): 532-550. Fair Wear Foundation. 2013. Bangladesh country study 2013. Available at: http://tinyurl.com/ybopqef4. Ferenschild, S. and I. Wick. 2004. Global game for cuffs and collars: the phase-out of the WTO agreement on textiles and clothing aggravates social divisions. Südwind e.V, Siegburg, Germany. Greenhouse, S. 2004. Workers assail night lock-ins by Wal-Mart. New York Times, January 18. Greenhouse, S. and J.A. Manik. 2014. Stalemate over garment factory safety in Bangladesh. New York Times, June 25. H&M. 2017. Sustainable fashion. Available at: http://tinyurl.com/ycycblht. Henderson, J., P. Dicken, M. Hess, N. Coe and H.W.C. Yeung. 2002. Global production networks and the analysis of economic development. Review of International Political Economy 9(3): 436-464. Human Rights Watch. 2015. ‘‘Whoever raises their head suffers the most’’ workers’ rights in Bangladesh’s garment factories. Available at: http://tinyurl.com/y8n7xjkq. Huq, F.A., M. Stevenson and M. Zorzini. 2014. Social sustainability in developing country suppliers: an exploratory study in the ready made garments industry of Bangladesh. International Journal of Operations and Production Management 34(5): 610-638. Hutchins, M. and J.W. Sutherland. 2008. An exploration of measures of social sustainability and their application to supply chain decisions. Journal of Cleaner Production 16(15): 1688-1698. Hwang, S. 2008. Utilizing qualitative data analysis software: a review of Atlas.TI. Social Science Computer Review 26(4): 519-527. International Labor Rights Forum. 2015. Our voices, our safety: Bangladeshi garment workers speak out. Available at: http://tinyurl.com/ya68ohlh. International Trade Union Confederation. 2008. Annual survey of violations of trade union rights – Bangladesh. Available at: http://tinyurl.com/ycmz2n7j. Kabeer, N. and S. Mahmud. 2004. Rags, riches and women workers: export-oriented garment manufacturing in Bangladesh. Chains of fortune: linking women producers and workers with global markets, edited by M. Carr. Commonwealth Secretariat, London, UK, pp. 133-164. International Food and Agribusiness Management Review

290

Carlson and Bitsch

Volume 21, Issue 2, 2018

Kathuria, S. and M. Malouche. 2016. Toward new sources of competitiveness in Bangladesh: key findings of the diagnostic trade integration study. Directions in Development. World Bank, Washington, DC, USA. Available at: http://tinyurl.com/yc83v8gp. Kenney, C. 2013. The tragic number that got us all talking about our clothing. Planet Money, National Public Radio, December 26. Kolben, K. 2007. Wal-Mart is coming, but it’s not all bad: Wal-Mart and labor rights in its international subsidiaries. UCLA Journal of International Law and Foreign Affairs 12: 275-332. Konopásek, Z. 2008. Making thinking visible with Atlas.ti: computer assisted qualitative analysis as textual practices. Forum Qualitative Social Research 9(2). Available at: http://tinyurl.com/yb4uog3f. Lipschutz, R.D. 2004. Sweating it out: NGO campaigns and trade union empowerment. Development in Practice 14(1/2): 197-209 Maloni, M.J and M.E. Brown. 2006. Corporate social responsibility in the supply chain: an application in the food industry. Journal of Business Ethics 68(1): 35-52. McGinnis, M.D. 2011. An introduction to IAD and the language of the Ostrom workshop: a simple guide to a complex framework. Policy Studies Journal 39(1): 169-183. McKinsey and Company. 2011. Bangladesh’s ready-made garments landscape: the challenge to growth. Available at: http://tinyurl.com/y9958aqc. Moon, J. 2007. The contribution of corporate social responsibility to sustainable development. Sustainable Development 15(5): 296-306. Mugera, A.W. and V. Bitsch. 2005. Managing labor on dairy farms: a resource-based perspective with evidence from case studies. International Food and Agribusiness Management Review 8(3): 79-98. Murayama, M. 2006. Globalisation and female employment in Bangladesh. In: Employment in readymade garment industry in Post-MFA era: the case of India, Bangladesh and Sri Lanka, edited by M. Murayama. Institute of Developing Economies, Japan, pp. 53-101. Murphy, K. 2012. The social pillar of sustainable development. Sustainability 8(1): 15-31. North, D. 1989. Institutional change and economic history. Journal of Institutional and Theoretical Economics 145(1): 238-245 O’Rourke, D. 2003. Outsourcing regulation: analyzing nongovernmental systems of labor standards and monitoring. Policy Studies Journal 31(1): 1-29. Ostrom, E. 1990. Governing the commons: the evolution of institutions for collective action (political economy of institutions and decisions). Cambridge University Press, Cambridge, UK. Ostrom, E. 2005. Understanding institutional diversity. Princeton University Press, Princeton, NJ, USA. Ostrom, E. 2011. Background on the institutional analysis and development framework. Policy Studies Journal 39(1): 7-27. Pagell, M. and Z. Wu. 2009. Building a more complete theory of sustainable supply chain management using case studies of 10 exemplars. Journal of Supply Chain Management, 45(2): 37-56. Rahman, S. 2015. The compliance of labour regulations by TNCs in Bangladesh-a host state’s perspective. European Journal of Law and Political Sciences 3(1): 66-75. Rahman, M., D. Bhattacharya and K.G. Moazzem. 2007. Bangladesh’s apparel sector in post-MFA period: a benchmarking study on the ongoing restructuring process (No. 23152). Center for Policy Dialogue, Dhaka, Bangladesh. Rahman, Z. and T. Langford. 2012. Why labour unions have failed Bangladesh’s garment workers. In: Labour in the global south, edited by S. Mosoetsa and M. Williams. International Labour Organization, Geneva, Zwitserland, pp. 213. Randall, W. and J. Mello. 2012 Grounded theory: an inductive method for supply chain research, International Journal of Physical Distribution and Logistics Management 42 (8/9): 863-880. Rashiduzzaman, M. 1997. Political unrest and democracy in Bangladesh. Asian Survey 37(3): 254-268. Reinecke, J. and J. Donaghey. 2015. After Rana Plaza: building coalitional power for labour rights between unions and (consumption-based) social movement organisations. Organization 22(5): 720-40. Roberts, S. 2003. Supply chain specific? Understanding the patchy success of ethical sourcing initiatives. Journal of business ethics 44(2): 159-170.

International Food and Agribusiness Management Review

291

Carlson and Bitsch

Volume 21, Issue 2, 2018

Seuring, S. and S. Gold. 2013. Sustainability management beyond corporate boundaries: from stakeholders to performance. Journal of Cleaner Production 56: 1-6. Seuring, S. and M. MĂźller. 2008. From a literature review to a conceptual framework for sustainable supply chain management. Journal of Cleaner Production 16(15): 1699-1710. Stake, R. 2005. Qualitative case studies. In: The Sage handbook of qualitative research. (third edition), edited by N. Denzin, and Y. Lincoln. Sage, Thousand Oaks, CA, USA. Stiles, K. 2002. International support for NGOs in Bangladesh: some unintended consequences. World Development 30(5): 835-846. Strauss, A. and J. Corbin. 1998. Basics of qualitative research: techniques and procedures for developing grounded theory (second edition). Sage, Thousand Oaks, CA, USA. Turker, D. and C. Altuntas. 2014. Sustainable supply chain management in the fast fashion industry: an analysis of corporate reports. European Management Journal 32(5): 837-849. Ullah Mirdha, R. 2016. Bangladesh remains second largest garments exporter, against all odds. Daily Star (Dhaka). Available at: http://tinyurl.com/y7mk78cr. Van Schendel, W. 2009. A history of bangladesh. Cambridge University Press, Cambridge, MA, USA. World Bank. 2013. Bangladesh â&#x20AC;&#x201C; Poverty assessment: assessing a decade of progress in reducing poverty, 2000-2010 (Bangladesh Development Series No. 31). World Bank, Washington, D.C., USA. Available at: http://tinyurl.com/ycjth5yk. Yardley, J. 2012. Export powerhouse feels pangs of labor strife New York Times. Available at: http://tinyurl. com/c9d5fng. Yardley, J. and J.A. Manik. 2013. Building collapse in bangladesh leaves scores dead. New York Times, April 24. Available at: http://tinyurl.com/y877q6hx. Yunus, M. and T. Yamagata. 2014. Bangladesh: market force supersedes control. In: The garment industry in low-income countries: an entry point of industrialization, edited by T. Fukunishi and T. Yamagata. Palgrave MacMillan, Basingstoke, UK, pp. 77-104. Zohir, S.C. 2001. Social impact of the growth of garment industry in Bangladesh. The Bangladesh Development Studies 26(4): 41-80.

International Food and Agribusiness Management Review

292