Towards sustainable knowledge diasporas : The rationale for an appropriate technopolicy

Towards sustainable knowledge diasporas : The rationale for an appropriate technopolicy Jean-Baptiste Meyer Institute of Research for Development September 2006 Report for the UNESCO Diaspora Knowledge Networks Project (Preliminary version) Summary: This chapter puts the diaspora knowledge networks into perspective. The first part describes their historical development up to recent times, when precise questions were raised about their real consistency and efficiency. These questions are addressed and then answered, mobilising very recent evidence, presented in part 2. Part 3 draws on the lessons that may be drawn from both success and failure stories providing the material for a systematic analysis, through a sociological approach of networks. The last and conclusive part sketches out the policy orientations that may be infered from such a network analyis in which technologies are crucial. Because of the intricacy of technical and policy issues, we labelled this approach “technopolicy�.

During the mid-1990s, the diaspora option – i.e. the connection of the highly skilled expatriates with their country of origin in order to contribute to its development- emerged as a possible mitigation of the brain drain and of the shortage of adequate S&T human resources in the South. As a paradigmatic shift and revolutionnary policy option it has come under scrutiny and has naturally faced a number of critiques. These critiques questionned the magnitude of the phenomenon, the sustainability of diasporic initiatives and their impact on the countries developments.

Since the beginning of this decade, systematic studies have been conducted by different teams, bringing converging evidence. Today, the highly skilled diasporas can no longer be minimised and they even appear as a more consistent and powerful developmental trend than it had initially been thought. To highlight this phenomenon has required new investigative techniques and methods, due to the particular characteristics of the research object : virtual (in)visibility and worldwide dispersion. It is a fact, now, that the existence of numerous expatriate associations is fully established and verifiable.

1

It is from the Asian highly skilled diasporas that the most convincing evidence comes. Dozens of specialised associations from many countries have been identified. Some of these are being studied by various organisations and the processes through which they build themselves, link members and counterparts, and act towards collective goals are precisely described and analysed. At the same time, the results of some diaspora networks and highly skilled expatriate associations are disappointing. There are various examples of unsuccessfull attempts and many experiences with limited visible output and activities. Reports on such shortcomings are taken into account and their views thoroughly interpreted.

Searching for explanations of both successes and failures of diaspora knowledge network leads to a complex picture. There are no management general recipes since networks, countries, conditions and development processes are multiple and diverse. But there are lessons about effective ways of getting home and host countries as well as diaspora actors associated in productive manners. This requires a clear understanding of the networks dynamics and of the mediation instruments or institutions that connect heterogeneous entities together.

Sociological concepts may be used in order to understand these dynamics and mediation processes. The specialised litterature on social capital, socio-economics of innovation and scientific networking provides adequate keys for the interpretation of what happens in the diaspora knowledge networks. The logics at work show that a mix of technological as well as organisational devices may create the necessary mediators for actions to be realised. In fact, these associative techniques are action related, much distinct from passive artefacts or administrative procedures.

ICTs are opening avenues for DKN developments. They shape the links among distant partners within the expatriate associations and between these and other groups in host and home countries. However, like any cluster, pole of comptetitiveness, strategic alliance or pool of collective resources, DKN existence is tied to investments made by various actors, for which public and institutional policies are decisive. Such policies have a clear impact on 3 necessary and interdependent investment types: on ICTs (infrastructural), on institutions (organisational) and on projects (programatic or operational). No DKN may survive without at least one of these investments and those combining them do experience a high return. ***

2

Contents

Summary

1

I- The controversial development of the diaspora option

4

I-1- Initial case studies and preliminary findings

4

I-2- The time of doubts and critiques

7

II - Confirming the diaspora option with new evidence

9

II – 1- Worldwide growth of diaspora knowledge networks

9

II – 2- A closer look at booming Asian cases

12

III - Systematic analysis of DKN dynamics

17

III – 1- The common logics of DKN

17

III – 2- Network sociology

21

IV – Lessons for policy orientations

28

IV – 1- The art of mediation and the need for action

28

IV – 2- Investments for a win win solution

31

References

?

Appendix 1: searching the web with systematic methodology

?

Appendix 2: list and links of 155 diaspora knowledge networks

?

3

I- The controversial development of the diaspora option In the early 1990s the mobility of highly skilled people was neither in the academic nor in the policy agendas. After the 1970s and 1980s broad and multidisciplinary concerns about the brain drain, only few econometric studies were still dealing with some modeling exercises about losses of human capital experienced by some countries. However, the issue was pragmatically adressed on the field, by governmental and civil society’s actors. In a few countries, these actors started to develop the reconnection of highly qualified expatriates.

These original atttemps were soon discovered by scholars and thus followed, observed, analysed and commented by them. The first descriptions of these networks happened to coincide with the emergence of the so called new global economy and the significant acceleration of skills circulation worldwide that it spured, in the late 1990s. The findings of these studies came under scrutiny and, in the follwing years, a number of doubts and critiques appeared with regards to the materials collected, the interpretations given and the policy suggestions eventually proposed. This part briefly presents terms of this scientific debate, recalling the initial case studies and the arguments to which they gave birth.

I-1- Initial case studies and preliminary findings The diaspora option – using networks of highly skilled expatriates to foster development in their home countries – originally emerged in Latin America. It was in this part of the world that such experiments gained political credit most rapidly. A brief list of them is given below: -

AFUDEST (the Franco-Uruguayan association for scientific and technical development), which lasted from 1985 to 1994;

-

ALAS (the Latin American Association of Scientists), which was founded in 1987 and gave rise to DATALAC (a database of Latin American researchers resident in OECD countries) and IRNLAC (Inter-regional Network of Latin American and Caribbean scientists) which was behind programmes (such as ECOMED in the field of medical biology) directly supported by UNESCO;

-

Various experimental Argentinean networks (Cre@r, Prociar, Setcip) that were promoted by ALAS in the second half of the 1990s but suffered from the subsequent crisis in Argentina;

-

4

TALVEN (Talentos para Venezuela);

-

Caldas (Colombian Network of Researchers and Engineers abroad) from 1991 to the present day.

Of all the Latin-American networks, the Caldas network is probably the best known, having been the focus of various studies since its inception 15 years ago. It was the outcome of a combination of factors, namely a spontaneous initiative on the part of Colombian expatriates to form local associations (mainly in the United States, Spain, France, Britain, Switzerland, Belgium and Germany) and their decision to link up via the Internet, which was booming at the time, just as Colombia was launching its own proper science and technology system, backed up by a policy to promote international co-operation (Meyer et al. 1997). The Caldas network was a model of its kind owing to its rapid growth, but its subsequent development has been marked by numerous problems. These featured in a recent evaluation report, which looked back over its history and highlighted issues such as resource-allocation weaknesses, poor organisation and technical management on the communication front, a mismatch between administrative structures in Colombia and in the host countries, and the failure to develop the network and its skills in Colombia itself (Chaparro et al, 2004).

A broader historical review of all the Latin-American networks highlights a crucial lack of professionalism. Run by voluntary members with enthusiastic personal commitment but too little time to devote to such work, they are subject to the vagaries of organisations whose staff cannot ensure the regular continuity of projects and activities (Lema 2003). This lack of follow-up obviously undermines any scope for sustainable development. The problem has apparently been overcome by some networks whose expatriates are from a part of the world where government or business can take over from volunteers where necessary, namely Asia (see 3rd part, below) .

In Africa, the South African Network of Skills Abroad (SANSA) was created in 1998 and constituted a case study also thoroughly investigated. Before the end of the millenium it managed to gather more than 2000 members in 65 countries, with a very high level of qualification and senior positions in all sorts of professionnal fields (Brown 2003). As a highly successfull mobilisation process like the Colombian Red Caldas though in a different manner, it also faced problems in subsequent developments. A recent evaluation of its activities emphasises the fact that in spite of a sophisticated internet infrastructure, it is underutilised and dormant (National Research Foundation 2005).

5

In spite of their limitations, Red Caldas and SANSA proved the feasibility of a diaspora option, i.e. the real existence of off shore extensive human resources that could be mobilised by the country of origin. Some of the activities in which these contributions have been observed during these case studies as well as in other instances, are summarised in the list below: •

Exchange

of

scientific,

technical,

administrative

or

political

information

(contribution to the creation of the new Colombian National S&T system in the early 1990s, by prominent expatriates); •

Specialist knowledge transfer (waste management procedures from the Ecole Polytechnique Fédérale de Lausanne-Switzerland, with the Universidad del Valle, Cali-Colombia);

•

“Scientific or technological diplomacy” or promoting the home country in the R&D and business community of the host country (South African medical research in England, Indian IT entrepreneurs in Silicon Valley);

•

Joint projects, partly on a virtual basis (distant working, simulations);

•

Training: attending home-country sessions and meeting/mentoring students abroad (a feature shared by most networks of that sort);

•

Enterprise creation (including multinational subsidiaries) to assist the possible return of expatriates on a part-time or permanent basis (Chinese high tech firms with returnees in science parks);

•

Ad hoc consultations, for example on research/development projects (peer review, job recruitment, technology assessment).

The question of the theoretical challenges and policy changes that these case studies brought to traditional migrations’ approaches has been dealt with elsewhere (Meyer 2001; Meyer, Kaplan and Charum 2001, Barré et al. 2003). But the mixed results of these experiences as well as the lack of historical depth and comparative materials led to legitimate concerns about the real significance and feasibility of diaspora options.

6

I-2- The time of doubts and critiques Reviewing the evidence about the networks, various authors from different areas of studies or disciplines (migration, science policy, economics) have questionned the diaspora option (Gaillard and Gaillard 2003, Lowell and Gerova 2004, Lucas 2004). Their interrogations can be classified in the following categories: •

Those dealing with the magnitude of the phenomenon: how many networks are there in the world? Are those already identified and studied isolated events or representative sample? Do these examples of strong identification from expatriate talents feature a general dynamic or simply an exceptional form of international socialisation ?

•

Those puting into question the sustainability of the expatriate associations : are they technically equiped to last more than a few years? As temporary groupings of highly mobile individuals can they keep a membership stable enough to manage a necessary continuity? Can we actually check on a big enough sample of networks their average life expectancy ?

•

Those criticising the consistency and extrapolation of activities from internet visibility: are websites really picturing associations with a true social and productive life ? Aren’t they often short lived individual attempts left as remnants on the cyberspace, dead stars sending light of anterior existence ? Would the networks not be simply cosmetic communication realisations responding to the hunger of agencies desperately looking for attractive solutions to the brain drain?

•

Those expressing strong skepticism about the actual impact of the networks on the home countries S&T or economic developments: can they really contribute to local scientific communities when these are almost inexistent (in Africa for instance)? Can they compensate virtually the lack of physical resources in laboratories or enterprises? Can we attibute high tech or IT developments to these diasporas when the combination of local factors and conditions seems to potentially explain the determinants of change (in India for instance)?

7

When several of these arguments are taken into account together, the object of research and of policy - the very diasporic associations – is disolving in the air. If a group of highly skilled expatriates officially working to the development of their country of origin is an exception, with a very low probability of survival, with an internet window becoming a pretext for inflated international projections and expectations but whose real impact on the country of origin is simply an illusion, then there is nothing substantial left to the diaspora option.

In fact, many of the doubts mentioned in the former paragraphs got credibility from the limited evidence about the networks that existed until recent years. These networks are, indeed, new and elusive research objects whose systematic observation poses real challenges and requires innovative investigative methods. Parts of the critiques raised before may thus be adressed and answered through a deeper analysis. Today, with new evidence, the suspicions that highly skilled diasporic initiatives might have been a short lived post modern fantasy in social sciences and a fashionable tool for policy makers impatient of quick fix solutions may be discarded. Part II describes this evidence sustaining DKN importance while part III analytically combines it with the reports on several networks shortcomings, in order to draw lessons on the best ways to manage the diaspora option for brain gain.

8

II - Confirming the diaspora option with new evidence Today, the data collected about highly skilled diaspora networks overcomes most of the limitations pointed at by the authors previously mentionned. These data come partly from a systematic analysis of hundreds of websites on the one hand, and from specific studies of Asian diaspora networks on the other hand, to which we refer in the last section. In this section, the rigorous methodology used to search internet will be first described before showing the robust results that were thus produced.

II – 1- Worldwide growth of diaspora knowledge networks 5 census of diasporic highly skilled networks have been made at different times by various teams, aiming at grasping the magnitude of the phenomenon, beyond isolated case studies.

-

41 expatriate networks of ‘developed’ and ‘developing’ countries were identified at the occasion of the June 1999 Unesco World Science Conference (Meyer and Brown 1999)

-

106 networks referring exclusively at developing countries, in a state of the art about scientific diasporas, made by a panel of international experts in 2002 and published in September 2003 (Barré et al. 2003),

-

61 expatriate networks of ‘developed’ and ‘developing’ countries, in a report prepared for the World Bank, “Diasporas and Economic Development: State of Knowledge” (Lowell and Gerova) in September 20041;

-

158 networks referring exclusively at developing countries, in 2005, (Meyer and Wattiaux 2006) ;

-

191 networks of both developed and developing countries, in 2006 (Tobin and Sallee 2006)

There is an overwhelming evidence on the current existence and activity of DKN. This can be seen looking successively at 2 aspects: network census figures, current activity checked by effective contact.

1

A new and not yet published report by the same authors expands these findings to 97 “e-diasporas” organisations.

9

a) A high number of networks and always increasing The Lowell – Gerova report of September 2004 questionning the actual number of networks and minimising the size of the phenomenon refers to old figures from the 1999 1 st census (Meyer and Brown 1999) and ignores those produced in 2002, published and widely disseminated in 2003 (Barré et al. 2003). In the latter, the number passed from 41 identified networks refering to all kinds of countries of origin –‘developed’ as well as ‘developing’- to 106 exclusively selected for their link with and orientation to development of so-called third world countries. This increase was due to the systematic search techniques used and presented in section 2. The results were made available as an appendix to the state of the art on scientific diasporas with URL available for further enquiry (Barré et al. 2003). A new effort in 2005 increased the number to 158 identified diaspora knowledge networks (DKN, listed in appendix 2). If the Lowell – Gerova own search figures are added to these - droping the few duplicated networks - the total of identified DKN amounts to 173 and concerns 40 different developing nations plus 4 specific regional groupings. The fact that few duplications occured among several enquiries, made at different times by different teams, with different techniques, as well as the ever increasing list pleads in favour of a likelihood to find even more networks in the future. Considering that some areas with high diasporic potential remained so far rather in the shadow (Carribean, Middle East, China) it would not be surprising that the exhaustive number of DKN would be over 350 and covering a majority of non-OECD countries. The case of China presented below confirms this with striking evidence.

When looking exclusively at DKN associated to developing countries and actually involved in development strategies with these countries of origin, the increase of the networks identified through successive and partly cumulative studies is the following (figure 1): Figure 1: DKN identified through websearch at different times by different studies:

10

400 350 300 250 200 150 100 50 0 1999

2003

2004

2005

2006 (est.)

b) A fragile though durable activity The last census exercise, in 2005, has checked the activity of every single network. This was done by systematically collecting trace of very recent events (under 1 month) or, in absence of such a trace, calling the coordinators, administrators or webmasters. In case no sign at all was emited by the DKN and received by the research team, the network was not included in the confirmed active list. The result thus does not leave place for any doubt: at least 101 of the 158 networks are definitely active today, representing almost 2/3 of the total (corresponding letter ‘Y’, in appendix 2). Their life span varies considerably from one network to another but is roughly of 10 years in average, for what their reported history is. Some are old associations that decided to go on air (have a website) and others are just pure and fresh emanations from the internet.

More than half of the DKN identified in 1999 (and relevant in the following lists) are still active today even if and when they reported difficulties (see supra, first section, the evaluation processes of Caldas and SANSA). These percentages are not far from those of the Lowell – Gerova study but the interpretation naturally differs considering the modest size of their sample, compared to the one here. There is not a slow erosion of a small stock of DKN but rather a lively forest of sometimes precarious, sometimes durable organisms. In any case, this normal mortality rate should be considered in light of the corresponding high tech background. For instance, in the United States, on average, 3 out of 5 new business start-ups fail within their first 5 years. No one would, however, infer from this figure that investing in 11

technological innovation is a mistake... Nonetheless, the question of the networks viability should certainly be examined through a comparative analysis of failures and success stories, in order to understand their conditions of realisation. The list provided in appendix 2 offers a big sample for such sociological and historical enquiries.

As explained in section 2, all the 158 identified DKN have an internet window, since their existence has been pointed at essentially through this medium. For some of these networks (12 of them, classified “type A”, in appendix 1) their very existence is on the internet, as they are pools of human resources mainly tied by the website artefact. The more general networks (38, classified “type B”, in appendix 1) are large associations for which only a part of their activities are explicitly turned towards knowledge and development issues with the country of origin. The specialised networks (48, classified “type C”, in appendix 1) focus on such actions, often in a pretty similar fashion as the Caldas or SANSA networks.

II – 2- A closer look at booming Asian cases Some countries of Asia have had early attempts to mobilise their offshore human capital when noone thought of it. India, for instance, had inscribed the option of occasionally calling on the NRI (non-resident indians) experts to offset brain drain effects, as early as the 1960s and 1970s (UNCTAD-CSIR 1977). However, the expatriate associations concerning the region mainly become visible in the late 1990s. Since then, they have come to the forefront and become some sort of a model. This section describes the place of these diasporas as it appears through the internet study exposed before and it mobilises the recent findings of another study to give a comparative and more precise picture. a) Asia and highly skilled expatriate associations Of the major areas of the world, the Asian region is the one which counts the highest number of registered highly skilled expatriate associations and of home countries concerned by these.

57% of all the confirmed active networks refer to asian diasporas though non-asian countries may also have high numbers of networks (see table 1 and figure 2 and 3).

12

Table 1 : Breakdown of DKN by regions of the world (regions of origin) Diasporas Regions

Identified networks

Latin America Africa Asia Total

25 51 82 158

Networks active in 2005 15 27 56 98

Figure 2: identified networks per region of origin

AMERICA LATINA; 25 AFRICA; 51

ASIA; 82

Figure 4: Major countries of origin for identified and active DKN Total networks

Active 2005

16 14 12 10 8 6 4 2

13

Ar ge nt in a

Ch in a

Ko re a

Th ai la nd

Ne pa l

In di a

Et h

io pi a M or oc co Tu ni Ba sia ng la de sh

0

Number of home countries 9 11 18 38

There is a significant concentration of networks per country of origin. Especifically from Asia, India, China and Bangladesh represent half of the identified active networks.

Unsurprisingly, these countries with diasporic links traditionally, have the highest numbers of highly skilled expatriate associations.

If these figures give a general picture illustrative of the relative importance of Asia and of some countries within this region, they cannot, however, be taken as definite indicators. A recent study led by the Asian Development Bank shows indeed that many networks are not easily visible on the internet. For instance, many Chinese networks, with websites in Chinese languages, have escaped the scope of the study presented above. This confirms the hypothesis that the number of networks is much superior to the one extracted in 2005 and may specifically tend to show that China’s highly skilled expatriate associations have been underestimated. For 9 identified through websearch, the OCAO (Overseas Chinese Associations Office) pointed to 200 (Xiang Biao 2005). The case of China may be exceptional in size. However, it does point to the fact that the visible part of the DKN iceberg may reflect just a small part of the exact volume of such populations involved (Figure 5).

Figure 5: Numbers of Chinese DKN identified through websearch and by OCAO

9

200

OCAO

Websearch

If it is unlikely that the total number of actual DKN be 20 times higher than the one of the globally visible (identified through the web) networks, like in the Chines case, an estimate of twice this number is a reasonable hypothesis to work with (see figure 1). It obviously gives a

14

substantial set of potential users of DKN techno-organisational tools such as the ones described in the following sections. b) Some preliminary lessons from the case of China and India The case of China and India are among those most developed and a few lessons may already be extracted of their detailed description (Xiang Biao 2006, Leclerc and Meyer 2006).

However, this ADB study deserves more analysis and comments as it brings a wealth of interesting materials and perspectives, to be compared with other studies on other countries in Asia as well as in other parts of the world. Among the points more specific to the Chinese case which explain the magnitude and success of its highly skilled diaspora mobilisation, the following may be emphasised: -

estimate of the population of OCPs (Overseas Chinese Professionals) in the world is about 1 million with high concentration in North America;

-

there are more than 200 registered associations of these OCPs registered by the OCAO (Council for Overesas Chinese Abroad Office);

-

a definite policy (wei guo fuwu) has been set up in the late 1990s promoting linkages with talents in the diaspora;

-

5 ministries and a high number of provincial governments agencies as well as para statal entities are involved in programmes and activities with highly skilled expatriates;

-

short term visits, collaborative projects between OCPs and home academic communities, senior expatriate scientists lectures in China, occasional technical advice, local recruitments through big fairs or more selective encounters are part of the many activities displayed by the diaspora and counterparts in China.

The role of the associations of OCPs, whose number has increased during recent years, seems a key factor in the expansion of links between formal and informal networks. The internet happens to be the major media used by OCPs to stay connected with the country and among themselves.

Summarising the major conclusions of this study, it appears that the Chinese case strongly confirms the great number and current expansion of highly skilled expatriate associations, their real intensity of activity as well as their responsiveness to policy factors. The question that remains to be solved is more on the transformations it generates in the countries of origin, 15

on the developments of which it may -or not- be the impulse. Contradictory explanations have so far been given, on the case of India and China.

The role of the diaspora in the Indian IT industry’s developments has been particularly discussed during the past few years. Scholars unanimously think this role has been important, especially in lowering reputation barriers to trade, but divergences are recorded with regards to the priority given. Some authors (Lucas 2004, Kapur and Mac Hale 2005) subordinates the expatriates’ input to local factors in India (mainly cheap highly skilled labour), while others see the intervention of prominent NRI and associations in the United States as crucial and determining (Saxenian 2002, Khadria and Leclerc 2006). However, as in the case of the underestimated numbers of diaspora networks, the evidence tends to increasingly prove the key role of diaspora entities. Anecdotal evidence of NRI IT executives and entrepreneurs in America shows, indeed, their direct involvement and original contribution in high level activities development in Bangalore and other industrial places (Pandey et al. 2006). Strategic transformation in the field in India with the branch moving up the value chain from low to high skilled activities appears to be directly related to diaspora initiatives combined with the return of well trained IT employees after the slow down of activities in the USA in the early 2000s (Warrier 2006). Micro level and historical records of Indian IT firms reveal that a real transfer of technology has taken place, with India capturing knowledge intensive and competitive parts of the production process from expats (Leclerc and Meyer 2006).

The Indian high tech sector’s development is correlated to the impressive expansion and intensity of professionnal associations of NRI IT specialists and engineers in the United States and in California in particular. As mentionned above (see figure 4) India ranks first among the countries with Diaspora Knowledge Networks working for the developement of origin countries. Moreover, to these DKN with clear transfer and development purposes, should be added the many professional associations serving the careers, entrepreneurial, business endeavours and networking activities of Asian/Indian community members in the US. The profusion of actors and intermediaries makes this milieu extremely dense and fertile. This situation is matched in India where national and local governments as well as universities, technology institutes, professionnal associations, federations, commissions, and chambers of commerce are very present and active. Between India and America, continuous circulation of human and material agents feeds both poles with complementary tasks and objects.

Interactivity in this multiple and dispersed milieu is ensured by a systematic and creative use of computer mediated communication. It is such a basic existential attribute that the absence of personnal blog is, for instance, sacarstically described as a prehistoric situation.

16

III – Systematic analysis of DKN dynamics The increasing evidence about diaspora knowledge networks has gradually made possible more precise analysis of their results or shortcomings. A number of academic research or cooperation agencies’ works have brought a diverse set of case studies useful for assesssing the strengths and weaknesses of the brain gain diaspora option: on the Colombian Caldas network (Meyer et al. 1997, Charum and Meyer 1998 and 2000, Chaparro et al. 2004), on the South African Network of Skills Abroad as well as the South African Diaspora Network (Brown 2003, NRF 2005, Castro Sardi 2006, Marks 2006), on the Latin American early experiences (Lema 2003), on Argentina’s various attempts (Kuznetsov, Nemirovsky and Yoguel 2006), on the Indian numerous examples (Khadria 2003, Leclerc and Meyer 2006, Pandey et al. 2006), on the Philippines (Opiniano and Castro 2006) and Afghanistan (Hanifi 2006), on Armenia (Minoian and Freinkman 2006), on the huge Chinese diasporas (Yugui Guo 2003, Xiang Biao 2006) and on synthesis compiling various of these contributions and others (Barré et al., Institut de Recherche pour le Développement, 2003; Westcott, Asian Development Bank, 2005 and 2006; Kuznetsov, World Bank Institute 2006).

It is time to capitalise on this harvest of case studies, not only in an evaluation perspective through comparative analysis of individual experiences with successes and failures but also by drawing systematic lessons from these many stories and conceptualising the general dynamics at work in this diasporic scheme. With this overall understanding of particular situations we might be equiped to think about proper networks configurations and modalities.

III – 1- The common logics of DKN The DKN are very diverse. However, they are all built on the same basic objective: to take advantage of expatriate networks and human resources for the benefit of the origin country.

The logics to which they all respond is thus one of networking and connectivity. This is what makes the diaspora option very much distinct – though complementary- from the return option (figure 6). The former relies on permanent repatriation of individuals’ human capital to be physically reinserted in the local environment while the latter mobilises the expat’s networks through the single actor with whom direct contact is made. There is, thus, a multiplier effect.

17

Figure 6: Return and diaspora options Return option

Country of residence

Diaspora

Diaspora option

Country of residence

Diaspora

Country of origin

Country of origin

The networks of these expatriates are very diverse in nature : scientific, technical, institutional, professionnal, financial, etc. What make the expatriate particularly valuable, indeed, for its country of origin, are obviously the skills displayed in his/her cognitive activities and made available at home, but also the path to equipments and facilities unavailable locally, the status, credibility and peers recognition attached to formal employment in a firm or lab of international visibility and thus provided to a peripheral one without any cost, the access to major funding programmes (see US-NSF cases in Johnson 2003, EU multilateral projects in Granes and Morales 2000), the introduction to markets especially through the integration of procedures, standards and quality assessments (see Lucas 2004), etc.

It is often said that, through the diaspora option, the country of origin is able to access the social capital accumulated by the expatriates. However, this refers to an extensive version of social capital, much more than simply interpersonal relations. It includes obviously human capital (Becker 1962) but also intellectual capital (Stewart 1997), symbolic and cultural capital (Bourdieu 1979), institutional (Rey Valette 2004), technical, physical, financial capitals as well, since equipments, infrastructures and money are susceptible of being mobilised too. In the case of the South African network of Skills Abroad, some of these have been estimated. Human capital and social capital indicators have been defined, in a reductive though analytical manner2. The diploma and the socio-professionnal positions have been chosen as proxies, respectively for the former and the latter (see figure 7).

2

The notions of human and social capital are controversial and should therefore not be taken for granted. Their use here does take into account the interpretative flexibility allowing their utilisation though with critical distance. On the latter see Meyer 2001 and Meyer et al. 2001.

18

Figure 7 : Human capital

Social capital

Qualifications Post-Doctorate Post-Graduate Diploma Diploma Doctorate Honours Masters

Positions of SANSA members

4,3

40 35 30 25 20 15 10 5 0

10,2 15,1 28,8 31,8 45

Bachelors

76

S enior Acade mics

Junior Acade mics

Exe cutives

Cons ultants

S enior Pos itions (other)

Others

Pos tion Uns pec.

Both capitals are very high in the diaspora: the rate of Phds is twice as much as in the equivalent qualified population at home (figures of the year 2000); the proportion of higher positions in professionnal hierarchy among the expatriate professionnals is about 3/4.

While the former (human capital) requires the reconstruction of an adequate context in order to be intensively used if repatriated, the latter (extensive social capital) may be available in a remote manner.

The goal, for the country of origin, is to tap the expatriate professionnals and the resources connected to them. Due to the executive positions and power these profesionnals hold, their capacity to mobilise important resources is globally very high, making the potential multiplier effect quite effective. For such reasons, the involvement and commitment of even a small number of expatriates may be decisive, as shown by anecdotal evidence on other diasporas (for instance on the case of Chile, see Kuznetsov 2006, 5-6).

In terms of social capital approaches, DKN exhibit the 3 properties mentionned by some theoreticians (proposed by Woolcock 2000). Bonding comprises the internal relationships within the network tending to build up a community. Bridging refers to those associating distant and asymetrical partners of different countries. Linking happens among the many peers connected through the network. According to Brinkerhoff (2006a, 2006b), the 3 properties are very much interdependent in the networks dynamics: bridging may be efficient because bonding is strong. The confidence acquired among members reflects in them giving access to third actors, for example.

19

The evidence also demonstrates that in most cases, a significant part of the highly skilled expatriates are willing to help their country of origin. The individual motivations may be quite diverse: guilt feeling of having left and ‘made fortune’ away; activist commitment or sentimental rememberings; oportunities to keep in touch with relatives; expectations about professionnal developments; social or entrepreneurial expansion; occasion of international connections and cooperation agencies support; etc... Whatever the reasons, the diaspora members are, in principle, sensitive to the home country’s situation, open to its solicitation and available for cooperation. Eventhough only a fraction may actually respond to its call, the occasionnal survey of non respondents have shown that there was actually less profound disinterest than simple ignorance or unawareness of home country’s moves. The desire for networking, therefore, usually comes from both sides: the diaspora and the place of origin.

The motivation behind DKN creation, development and the commitment of their members deserves deeper explanations. The mutually reinforcing process between human and social capital has been pointed at (Helliwell and Putnam 1999, Denny 2003) and the highly skilled expats propensity to gather and build associations fits with this pattern (Banks and Tanner 1998, Gibson 2001). However, at the same time, the opportunity cost of professionals to get involved durably into non profit activities is comparatively high (Brown and Lankford 1992).

This also corresponds to what happens in many DKN: to keep the highly skilled expats in the network is difficult; more than to get them in, because their time is relatively less available than for other working or non active persons. Underproductive endeavors are thus quickly punished with an “exit” from those who might have embarked initially with enthusiasm.

However, beyond individual dispositions, there is an interesting collective process which is at work in the constitution of DKN. As mentionned by Brinkerhoff (2006b), the identity expression in such networks is very important. Our own results on Colombian and South African networks have shown that these identity expressions are very much constructed by the very actors when they decide to get involved in the network (Meyer et al. 1997, Meyer, Kaplan and Charum 2001). In such a constructivist approach to identity shaping, the role of communication technologies is crucial (see next section).

The identification of the professional expatriates to their country of origin is a product of networking through DKNs. However, it does correspond to a phenomenon rather classical in sociology. The labour market, traditionnaly regulated and bounded by national borders, is a very strong driver of national identities (Gellner 1983). It is therefore not surprising that the recognition –by origin countries- of professional abilities, crucial in the shaping of professionnal identities (Dubar 1991), generates in return an extension of the latter to the 20

national origin sphere. It is a re-identification through professional motives. This finding points to an important fact: identification processes are dynamic and action related. They are tied to current work activity more than to passive cultural remnants. This is especially so for cognitive professionnal practice. Knowledge production and development are indeed activities in which identitarian processes through collective projection into the future are very high (Meyer 2006). The new diaspora of scientists and engineers have contributed to highlight this phenomenon in an unexpected manner (Meyer 2004a). The increasing awareness of the importance of knowledge in development processes, stimulated by the emerging national systems of innovation in developing countries, as well as the opening to international winds by these countries in the 1990s - Argentina, Chile, China, Colombia, India, being good examples of such policies- have vivified an active re-integration process between the diaspora and the origin countries. This new integrative approach is coined as an actual policy process

by Gamlen (2005).

III – 2- Network sociology The dozens of networks mentionned in part II and exhibited in appendix 2 have almost all been created in the 1990s and 2000s, and are indissociable from the expansion of Arpanet, Bitnet, then Internet. The role of ICTs has been absolutely crucial in the spontaneous emergence of such DKNs in a humanly non coordinated though convergent manner. Even when traditional associations preexisted the constitution of actual DKN, the introduction of computer mediated communication boosted their activities. For all the networks registered in this database as well as for all the others known by other sources, the Internet is the major way, and by far, of exchanging information. In specific cases where the development of both internet and a single DKN coincided and could empirically be tracked and reported, the dynamics have shown that the electronic media have definitely been fundamental in the creation of the social entity (Granes and Meyer 2000). In the Caldas network, for instance, the email lists, news groups and institutional nodes have permited to identify diaspora members, to contact them for an initial and often successful ‘recruitment’, to mobilise their own networks, to provide a permanent common space shared by all, to build on complementarities with the Colombia based communities, etc...

At the same time, however, the tremendous input of the ICTs in the DKN process had an ambiguous effect. The proliferation of messages of all sorts, sometimes with few substantial matters, occasionally with controversial debates on sensitive issues with respect to the country 21

of origin, led to excessive noises. Like in the R-Caldas, it sometimes exhausted the founding fathers motivation, spoilt their energy by endless and sterile discussions, suffocated members potential initiatives in useless fora and generated bitter comments spurring further controversies and acrimony. In that sense, if the ICTs and internet have, without any doubt, been crucial in the development of DKNs, it may also be said that they caused many of these networks’ growth problems. These are no different from traditional problems in communication

studies

(noise,

conflicts,

misunderstandings,

accusations,

echoes,

amplification, rumours, etc.). But they precisely expanded because communication happened to be disconnected from action. In the Caldas network for instance, the coincidence between the proliferation of discussions within the R-Caldas list and the drying up of political initiative from Colciencias - initially the major stakeholder- in the network development after the mid 1990s, is of striking evidence.

The South African Network of Skills Abroad, though organised in a very different manner, exhibits similar deficiencies. Drawing lessons from the Colombian shortcomings in communication tools, it was thus provided with efficient instruments (Website, newsgroups, forum, email list, bulletin board,etc.) for its 2500 members located in 65 countries (Brown 2003). After 5 years, the agency managing the network concludes with a sub-utilisation of these resources (NRF 2005). The reason may be attributed to a lack of animation, human agency in these socio-technical networks being absolutely crucial (Turner et al. 2003). In this deficient human agency, more than administrative capacity and attention, the responsibility of weak political engagement may be underscored, like in the case of Colombia.

This is where an appropriate theory of action becomes useful, in order to understand what is needed to develop and sustain such new associations as the DKN ones. The sociology of science and technology or sociology of innovation has proposed a new conceptual framework with the actor-network theory developed during the last 20 years. It is extremely relevant for such objects of analysis as the Diaspora Knowledge Networks. The actor-network theory is based upon 4 concepts: problematisation, interessment, enrolment and mobilisation (Callon 1986, Latour 1987, Law 1999, Latour 2005). These concepts constitute the translation process through which socio-technical change may happen. Such a process may be described and made visible through dynamic mapping instruments (Callon M., Courtial J.-P., Turner W. 1991, for the theroretical and methodological approach and various chapters of the present volume for DKNs more specifically). In brief, for an innovation to succeed, those who create it (scientist, entrepreneur, R&D department, etc.) must reformulate a problem in adequate 22

terms, get the interest of entities susceptible of chosing the new path, enroll the allies (humans and non humans) who will define the new standard, mobilise and channel the resources and actors along these new lines, making these gradually irreversible. Thus, the innovation process is not the result of a self-imposing logic or natural diffusion of a superior artefact; it is a struggle between on the one hand, existing entities and networks and on the other hand, those trying to get relevance and momentum in order to make their own place below the sun.

Therefore, in many ways, innovating means a dissociation of a previous configuration, capturing actors and building a network in which these will hold together in a new manner.

Through problematisation, interessment, enrolment and mobilisation, there is thus diversion of previous elements and their associations to the benefit of new consortia.

This approach fits with the DKN situation. Let us examine more precisely what happens with the diaspora option presented in III-1. In this scheme, the expat is an actor-network, condensating his/her connections (social capital) in oneself. The DKN purpose is to capture this actor-network and divert him/her towards the home country, which is itself, at least potentially, a consortium of networks. In this sense, in order to realise the translation process, DKN must perform the problematisation, interessment, enrolment and mobilisation operations. In fact, today, most of the existing DKN may be considered as having succeeded in several of these operations, mainly : problematisation, mobilisation and enrolment. The experience shows indeed that the 3 have often been accomplished. a) Problematisation The problem of skills circulation has been successfully reformulated, with the shift from brain drain to brain gain during the last decades. With the latter has emerged a vision in which the possible reconnection has become a workable and promising option. Today, neither a single country nor an international organisation would reject the possibility of networking with expats in a systematic manner and this opens opportunities for many DKN. Many expatriates have implicitly or explicitly acknowledged their potential contribution to their country of origin, meaning that they were no longer unreachable and could be available for projects run by their fellows back home. This is an actual, pragmatic, individual brain gain stance, instead of the previous permanent expatriation scheme that used to prevail in these people’s mind. Personnal interviews show this historical reformulation of mobility as a life experience vs definite farewell to the place of birth. A crucial shift in this reformulation has been the introduction of internet communication, as a reconstruction of continuity between separate places and sequences of time. b) Mobilisation 23

All the cases where real attempts of setting up networks have been impulsed did manage to gather a significant number of expatriates. The extent of the mobilisation does vary considerably from one network to the other, from country to country, but the message of call from home has proven to be appealling to many (see above, III-1). This message has usually been spread through e-mailing lists made of personnal or institutionnal contacts. Interestingly, many expats confess that prior to such a call, they had been away from any national initiative.

The fact that it is based on professional and intellectual reasons and interests seems to have fueled the mobilisation process while political or patriotic channels appear to be more divisive or dissuasive. The new media, internet, being neutral on these aspects and in this particular context, has allowed reconnection on new grounds. c) Enrolment The DKN identified (part II and appendix 2) reveal a high degree of formal enrolment into associative structures. The expatriates agree to become members of a collective endeavour with an explicit purpose for which they play a definite role. The review of websites shows that a number of associations exhibit their affiliates personnal data and professional features. They are stored in data bases as individual components of the network, to which they agreed to serve the purpose, becoming by the same token accessible to anonymous actors that would reach the network for any reason or objective. The actors are thus “punctualised” in the network, as described by Callon et al. (1986b, 1991) The evidence clearly shows that adequate problematisation, significant mobilisation and actual enrolment have been achieved by the diaspora option in general and individual DKN in particular. However, there is one point on which the results have been most of the time much less satisfactory: interessment. And on interessment depends sustainability...

Interessment is what durably ties the actor to the network. Beyond simple incentives susceptible of generating psychological motivation, it covers all the intermediaries that stick the actor to a particular network. Inter-esse, in latin means to stand in between. The interessment devices (dispositifs d’intéressement) are those standing between the actors and nurturing collective action through which they link up. It takes for granted that the actor’s situation is in a competitive environment: interessment is what makes him/her select one network instead of others, among various possible connections.

In DKN, like in innovation networks in general, the interessment devices may be diverse: programmes, funding, invitations, meetings, rewards, contracts, information, etc. In a physically scattered group of actors, ICTs play an immense role among these devices. They 24

have the power of extracting the individual expatriate of its daily local networks to insert him/her within those of relevance in the origin country. However, this only makes sense for active purposes since the involvement in a DKN is not for communication per se but for action with regards to developments at home. This explains the weariness towards passive general mailing lists expressed by R-Caldas members for instance and the importance of interactive ad hoc devices such as those presented in the following chapters of this report.

ICTs have the power to attract expatriates on actions in origin country’s networks. But without action this attractive power vanishes into the air. This tends to plead for focused instruments, ones that would get along proper institutional or organisational forms on specific projects run jointly by DKN and home communities members.

However, there is a very specific feature of DKN in interessment procedures and one that may explain some of their shortcomings in that regard. At the opposite of conventional innovation networks as presented in the theoretical approach above, indeed, DKN do not aim at definitely and exclusively capturing the actor-networks represented by the expatriates. They, instead, try to only temporarily do so in such a way that these actors keep his/her multiple connections.

Taking apart the expatriate from his/her invaluable networks is the last thing that DKN would pursue, at the difference of traditional innovation processes. A DKN member is, by definition, the bridge between his/her own networks in the host country and the ones of the origin country. This collaborative (vs competitive, in the traditional innovation networks) scheme is theoretically possible and actually successfull in a number of cases but the tension does exist.

It may result in a withdrawal if the interessment procedure is not strong enough, as proved by a number of cases.

As mentionned above (III-1), the expatriate is generally very well inserted in his/her socioprofessionnal and other networks in the host country. These generate strong interessment devices locally which tend to hold the actor far from any non-competitive association proposed from a peripheral place. For instance, why would a prominent genetician leading international projects from the UK be interested in linking with peers in his native Bangladesh if no demand is relayed, no concrete perspective given, no support considered, no information provided ? Inspite of his awareness of this country’s problems (problematisation), personnal motivation (mobilisation) and initial involvement in tentative actions (enrolment) when called by his fellows at home, he may slowly or quickly drop out and even try to get protected from further useless disturbances in his daily work, when checking that his engagement is not met by investments on the other side.

25

In fact, many testimonies of DKN members point to the fact that passivity – a generic term for absence of (re)activity- in the country of origin led them to stand aside, sometimes with the feeling of having been first lured then deceived. A basic reason is time, as time translates the individual’s investment in networks, therefore the interessment exerted by these on the actor.

Eventhough the diaspora option is non exclusive since it combines existing and new associations, time is a non compressible variable which implies selection. Then, sometimes, the expatriate has to choose between the alternatives of time to be devoted to - in that particular sense competing- activities. At the opposite of the return option, in which competition with alternative activities from other associations are canceled or minimised by the expatriate’s physical reintegration, the diaspora option always has to provide attractive conditions for/by action since mobilisation is always partial. Moreover, in some reported instances, the expatriate’s investment into a DKN conflicted with his/her professional involvement, when expectations of associations of his/her networks in the host country have been unmet by corresponding ones in the origin country.

This clearly draws the line of conduct for DKN. They must aim at minimising the conflict of interest and be equiped accordingly. If the executives in the diaspora have very little time though a highly productive one because of their networks- then the DKN must provide timeproducing (vs time-consuming) intermediaries, in other words it must focus on productivity.

Technology is a key issue here but policy is not less important. The combination of both is necessary. Rapid access to the actors is only possible if information technology is available but also if these actors have been previously made aware of the option, mobilised, enroled and interested. This means a massive involvement of heterogenous actors, especially in the country of origin, to multiply mediations and possible interactions with the diaspora (see further, figure 8 and examples of China and India).

In many home countries this effort has not been made. The creation of institutions has recently been proposed as a possible answer and counterpart of the diaspora in the origin country (see Kuznetsov et al. 2006 and Chaparro et al. 2004). However, it is only a part of the answer. It unrealistically assumes that the diaspora groups will provide their skills and contacts to a governmental agency making interface with the local communities, as if both the former and the latter were homogeneous entities easily represented by a couple of single actors. In the case of the Colombian Caldas network, disengagement and criticisms from the expatriate members coincided with its institutionnalisation within Colciencias during the mid

26

and late 1990s. Instead of strenghtening the response to the diaspora supply it weakened its diffusion and ramification in the social fabric in Colombia.

Figure 8: Necessary mediations between actors within Diaspora Knowledge Networks

expat actors

DK N

mediations

ork netw

Host country

actors

ork netw

Home country

The institutionalist, structuro-functionalist approach (Merton 1973) in the sociology of science shows its limit when it turns to science policy at the meso level where networks operate. A macro actor like the home country State must certainly show the way and boost initiatives but should delegate and pass on decisions, measures and negociations to other characters. The Chinese and Indian cases leave no place for doubt on this point. Erroneously perceived as paramount of centralised intervention, these countries exhibit a tremendous involvement of heterogeneous actors, a multiple mediation process, in the development of their S&T diasporas (Xiang Biao 2006, Leclerc and Meyer 2007). Central state has had a clearly favorable policy indeed – wei guo fuwu in China, high level committee on diasporas in India- but the initiatives of provincial states, distinct administrative bodies, single institutions, para-statal organisations, ONGs, academias, firms, individuals, and so on, have had a decisive impact. The fact, that many different DKN and programmes developed at the same time for the same country of origin did not have a negative effect. At the opposite, it helped multiply

27

the opportunities of cross fertilisation. Therefore, instead of -or beyond- institution building, the State’s role is rather to suggest, facilitate and coordinate multiple actors initiatives. This is where governance actually takes place and it would not have become possible without distributed knowledge schemes opened by ICTs.

IV – Lessons for policy orientations The present section aims to synthesise the elements of the DKNs social dynamics that have been pointed at in the previous paragraphs and to articulate them with organisational and technological devices that can support the dynamics and their sustainability. The solutions come from a good combination of both, with no deterministic approach and rather an emphasis on policy relevant decisions. This approach is thus refered to as “technopolicy”, as technological and public -as well as institutional- policy issues and options are completely intertwined. In no ways does this presentation pretend to close the reflection and discussion on the subject matter. But it suggests some perspectives and practices to be explored, experimented, tested and evaluated.

IV – 1- The art of mediation and the need for action Comparing the 2 pictures of success stories and failures, a striking difference may immediately be noticed. While the first show large numbers of actors of all sorts involved in the relationship between home and host countries and besides these, the second exhibit a scarce population of a few individuals and organisations with little interaction. The Indian and Chinese cases even reveal that there is no exclusive way of populating these networks: the former relies on market, business and essentially private links and the latter on institutional, academic and mainly public ones. What matters is not status or category but rather plurality and involvement. No fear should thus be kept about possible redundancies and on a supposed

28

obligation of links optimisation in order to be efficient, like some social network and capital approaches or interpretations would tend to indicate (Burt 2000).

In fact, the representative model to follow has more to do with the lessons from the recent field of innovation studies (sociology, anthropology and economic geography). The evidence shows that highly innovative milieux are made of heterogeneous networks interweaving increasingly in a reduced space where productive associations tend to concentrate. The intensity of exchange is thus correlative to dense though exclusive links where tacit knowledge can flow easily and quickly. Though the litterature has focused on the local nature of such spillovers and transfers, it is also shown that actually, the geographical proximity is not the important factor. It is only a reflection of social relationships made possible by neighbourhood as it appears when physical and relationnal factors are analytically distinguished (Breschi S., Lissoni F. 2003).

What is precisely at stake in the DKNs is the recreation/recomposition of such neighbourhood conditions in spite of the dis-location of actors. To reproduce the relationnal intensity and interactive proximity of highly innovative milieux, information and communication technologies are absolutely crucial. So far, the experience shows that these have been critical for almost all the identified networks, in initial messages to build participation, recruit members and disseminate basic information but their potential for interaction has been severely neglected. When compared to the use of interactive technologies in multinational firms, DKNs appear very limited and at embryonic stage. For instance, in MNCs, the frequent parralel use of audio-conference through telephone lines for unformal exchanges, along with intranet high speed data transfers, completed by satellite occasionnal visio-conferences have been intensively used for the past 2 decades. This explains their tremendous growth and their increasing transfer of R&D and knowledge producing capacities and core activities, in nonoriginal, decentralised places. The DKNs should observe the corporate world tactics and

29

techniques in this regard and translate some of these modalities into their own needs and practices.

The vitues of DKNs are often described as their ability to transmit or channel information in a very effective manner, accelerating transactions, making them more reliable and smoother.

Bridging capacity, gate keeper’s role and reputation enhancer are the main qualities generally attributed to these networks. In such a perspective, the diaspora facilitates, rather in a passive way, the transmission of resources whose generation and utilisation are determined elsewhere, in both the host and home countries. These qualities, though real, do not make justice to the actual participation of the diaspora in the constitution and appropriate allocation of resources.

The actor-network theory provides, here again, a more accurate conceptualisation. The diaspora plays, indeed, a role as mediator and not only intermediary. For Latour (Latour, 2005) an intermediary just transmit something without transformation, but a mediator modifies it and adds a specific value, which is the case in the successfull asian examples mentionned in the former section (Leclerc and Meyer 2006).

As Kapur and Mac Hale (2005) rightly point out, what would the diaspora advantage consist in if it was only an additional information provider 3? Today, the supply of information is not a problem any more; what is important is to get the right and appropriate information for specific purposes. The diaspora thus acts as a selector of information in relation to certain needs and activities. There is no doubt that the data mining and mapping techniques described in the following chapters of this volume are of high interest in this regard, for a better intervention of diasporic actors.

However, the mediation role of the diasporic actors goes much beyond this information selector’s effect. In fact, DKNs are not simply go-between facilitating contacts and

3

Chap 7 on the role of the diaspora : “The problem facing economic agents today arises not from a lack of information but from a deluge of it, as people are bombarded with more information than they are equipped to handle. In order to control the flood of knowledge, they have come to rely heavily on social networks...”, (Kapur & Mc Hale, 2005, 112)

30

cooperation between home and host countries entities. They are not passive agents expecting initiatives to be relayed through their associations and communication facilities, apart from some dormant networks whose members sporadic mobilisation may sometimes keep individually connected though collective dynamics have slowly fraid. The very existence of DKN is tied to action. It is what fuels the networks. What distinguishes DKNs from cooperation agencies and from any traditionnal international scientific or business network, is the collective identification and commitment towards national knowledge goals. What jusitifies the voluntary involvement of their members in favour of their countries of origin, is the purpose of contributing by their skills to these countries developpement. This contribution is concretely realised through projects. If the networks are not fed by projects, if action does not nurture their life, they gradually congeal and loose their members, as several examples show. This has a direct incidence on both technology and management.

Technology must be geared towards serving these actions and concrete projects. In fact, the term “ICTs� - Information and Communication Technologies – does not seem the most appropriate in that it alludes to intangible, abstract phenomenons while interactive devices would be those most fitted to the networks purposes. However, the former (static websites, email diffusion lists, etc.) have prevailed so far over the latter in DKNs (Turner et al. 2003).

Such distributed knowledge practices allowing dis-located production and use of skills are described in this volume and give a new field of expansion for DKNs as entrepreneurial (obviously not necessarily lucrative) hubs.

Management should strategically be oriented towards projects. The diaspora is not a stock of human resources that comes in addition to intellectual/professionnal communities at home. It is an extension of these and should be associated to programmes of collective interest. This is where the interesment devices must take place. To decide upon these, technologies are

31

necessary: they may, for instance, provide the virtual room where the convergence of actors goals may be built.

IV – 2- Investments for a win win solution The role of the State in interessment devices is crucial not only by its own input but also by the example its sets for other actors, public or private. Recent surveys of DKNs unanimously converge on the importance of nation states involvement (IRD, Barré et al 2003; WBI Kuznetsov et al. 2006; ADB 2006). Governmental agencies may systematically provide infrastructural elements (databases, portals, websites, information in general) easing the access of the expat’s networks to local (origin countries) ones like in the case of Colombia today (see Castro Sardi, this volume). It may also grant action-tied resources, be they material (fundings, equipments, technical support), organisational (institutional support, procedural guidance, administrative assessment) or symbolic (rewards, mediatisation, exposure). Above all, it may act as intermediary among multiple actors (for instance chamber of commerce, scientific councils, universities, multi or national companies, etc.) both abroad and at home through its various services at different levels. Then the network logics takes on without the initial actor being contiguous to newcomers, in an expansion dynamic of increasing returns.

To fuel interessment devices, self-sustained action oriented processes, strategic investments are crucial along with interactive technologies. Whatever enthusiastic it is to cooperate with the country of origin, the diaspora will not durably do so without continuous relationnal facilitation. Passive usage of technologies, simply relying on supposedly inherent properties of communication hold by the artefacts will be of no use. Human agency in ICTs support of DKN is crucial (Turner et al. 2003, chapters in this volume) and it may be part of governmental or cooperative bodies role to ensure its presence, at various places along the chains of actors associations. Such an investment must not be neglected: the same happens in any cluster, incubator or technopole. Mediators assume collective learning tasks, circulation of tacit knowledge, representation under various forms, risk sharing on uncertain projects.

Without their participation - a public good indeed – little externalities, accumulation and transfers would be generated.

It would seem natural that origin countries, as obvious possible beneficiaries, make such kinds of investments. However, it is also the role of host countries as well as intergovernmental organisations to grow and feed these chains of intermediation because they potentially benefit 32

to both poles of migration systems (co-development perspective) and to a multilateral cooperative regime (global redistributive prosperity perspective). Technopolicy options have been identified that would make these investments on both the host and home countries possible. Diaspora incubators in OECD receiving countries have been conceived as connective platforms with both equipments (hard and soft) and strategic data (access to social, economic and other resources) to expand the DKN capacities and to professionalise them (BarrĂŠ et al 2003). Local antennas in origin countries have been suggested as the interface between external actors and national programmes (Lema 2003). There too, is the idea of both infrastructure giving access to potential partners, organisationnal entities providing virtual spaces where reprentative actors could meet and decide upon projects of common interest in and outside the country and programmes for which investments would be made to boost joint activities between home communities and expatriate members .

In a world where the hypothesis of the diaspora model generalisation (Bordes-Benayoun, Schnapper 2006), the blurring of borders in migratory schemes (PĂŠcoud, Guchteneire 2005) and the mobility/sedentary patterns transformation through ICTs (Diminescu 2005), take every day more consistency, nation-states projection into the future does require a proactive technopolicy. Forging the instruments of an emerging cosmopolitanism (Beck 2006) has indeed become a crucial issue. This is what is at stake here.

33

References Barré R., Hernandez V., Meyer J-B., Vinck D. (2003) Diasporas scientifiques/Scientific Diasporas, Paris, IRD Editions Banks and Tanner (1997) The State of Donation : household gifts to charity, 1974-1996, Institute for Fiscal Studies, London Beck U. (2006) Qu’est-ce que le cosmopolitisme? Flammarion, Paris Becker G. (1964) Human Capital, University of Chicago Press Bordes-Benayoun C., Schnapper D. (2006) Diasporas et nations, Odile Jacob, Paris Bourdieu P. (1979) La distinction ; critique sociale du jugement, Minuit, Paris Breschi S., Lissoni F. (2003). “Mobility and Social Networks: Localised Knowledge Spillovers Revisited”, paper presented at R&D, Education and Productivity, Paris, august 2003 Brinkerhoff J. (2006a) “Diasporas, Skills Transfer, and Remittances: Evolving Perceptions and Potential”, in Asian Development Bank, Converting Migration Drains into Gains, Harnessing the Resources of Overseas Professionnals, Manila Brinkerhoff J. (2006b) “Diasporas, Mobilization Factors and Policy Options”, in Asian Development Bank, Converting Migration Drains into Gains, Harnessing the Resources of Overseas Professionnals, Manila Brown M. (2003) “The South African Network of Skills Abroad: the South African Experience of Scientific Diaspora Networks”, in Barré et al.

Brown E. and Lankford H. (1992) “Gifts of money and gifts of time: estimating the effects of tax prices and available time”, Journal of Public Economics, 47 : 321-341 Burt R. (2000)« Structural Holes versus Network Closure as Social Capital » in Lin, N., Cook, C.S., Burt, R.S., de Gruyter, A. (eds), Social capital: Theory and Research

34

Callon M. (1986a) “Some elements for a sociology of translation: domestication of the scallops and the fishermen of St Brieuc Bay”, in Law J. (ed.), Power, Action and Belief: a New Sociology of Knowledge?, London, Sociological Review Monograph: Routledge and Keagan Paul Callon M, Law J., Rip A. (1986b) Mapping the Dynamics of Science and Technology, Mac Millan, Londres Callon M., Courtial J.-P., Turner W. (1991), “La méthode Leximappe: un outil pour l’analyse stratégique du développement scientifique et technique”, in Vinck D. Gestion de la Recherche, nouveaux problèmes nouveaux outils, De Boeck, Bruxelles : 207-277 Chaparro F., Jaramillo H., Quintero V. (2004) Aprovechamiento de la Diaspora e Insercion en Redes Globales de Conocimiento: El Caso de la Red Caldas, World Bank, Bogota Charum J., Meyer J-B. (eds.) Hacer Ciencia en un Mundo Globalizado; la Diaspora Colombiana en Perspectiva, TM Editores, Bogota Denny K. (2003) The effects of Human Capital on Social Capital : a Cross-Country Analysis, WP 03/16, The Institute of Fiscal Analysis, London Diminescu D. (2005) “Le migrant connecté: pour un manifeste épistémologique”, Migrations/Société, vol.17, n°102, pp ; 275-292 Dubar C. (1991), La socialisation: construction des identitiés sociales et professionnelles, Armand Colin, Paris Gaillard J., Gaillard A-M (2003) “Can the Scientific Diaspora save African science?”, Scidev Website Gamlen A. (2005) The brain Drain is Dead: Long live the New Zealand Diaspora, COMPAS Working Paper n°10, Oxford University.

Gellner E. (1983), Nations and Nationalism, Basil Blackwell, Oxford Gibson J. (2001) “Unobservable family effects and the apparent external benefits of education”, Economics of Education Review, 20 : 225-233

35

Granes J., Morales A. (1998) “Las potencialidades y limitaciones de la Red Caldas de Investigadores Colombianos en el Exterior: los Proyectos Internacionales Conjuntos: un estudio de caso”, in Charum J., Meyer J-B. (eds.) Hacer Ciencia en un Mundo Globalizado; la Diaspora Colombiana en Perspectiva, TM Editores, Bogota Granes J., Meyer J-B (2000) “Globalization of the National Scientific Community Through Electronic List: Lessons and Prospects From a Case Study in Colombia”, in CHARUM J., MEYER J.-B. (éds.), International Scientific Migrations Today (CD-ROM), Paris, Editions IRD-COLCIENCIAS Guo Y. (2003) “How Can the Chinese Intellectual Diasporas Bridge their Host and Home Countries as well as Help their Home Country Integrate into the International Community”, in Barré et al.

Hanifi S.M. (2006) “Material and Social Remittances to Afghanistan”, in Asian Development Bank, Converting Migration Drains into Gains, Harnessing the Resources of Overseas Professionnals, Manila Helliwell J.F. and Putnam R. (1999) “Education and Social Capital”, NBER Working Paper n°7121, Cambridge, National Bureau of Economic Research Johnson J. (2003) “S&T resources and programs in the United States for networking with developing countries.” In Barré et al. (eds.)

Kapur D., Mac Hale J. (2005) Give us your Best and Brightest : the Global Hunt for Talent and its Impact on the Developing World, Center for Global Development, Washington DC Khadria B. (2003) “Case study of the Indian Scientific Diaspora”, in Barré et al.

Kuznetsov Y. (ed.) (2006) Diaspora Networks and The international Migration of Skills: How Countries Can Draw on Their Talents Abroad, World Bank Institute, Washington DC Kuznetsov Y., Nemirovsky A., Yoguel G. (2006) “Argentina: Burgeoning Networks Abroad; Weak Institutions at Home”, in Kuznetsov (ed.)

Law J., Hassard J. (eds) (1999) Actor Network Theory and after, Blackwell Publishing, London

36

Latour B. (1987) Science in Action; How to Follow Scientists and Engineers through Society, Open University Press, Milton Keynes Latour B. (2005) Reassembling the social: an introduction to actor-network theory, Oxford University Press, Oxford, New York Lema F. (2003) “Professional migration from Latin America and the Carribean; from NGO to multilateral organisation and government involvement: three cas studies”, in Barré et al.

Leclerc E., Meyer J-B (2007) “Asian miracle and diaspora: the indian booming IT industry”, Asian Population Studies, 3, 1 (forthcoming) Lowell L., Gerova S. (2004) Diasporas and Economic Development: State of Knowledge, World Bank, Washington D.C. Lucas R. (2004) International Migration Regimes and Economic Development, EGDI, Stockholm Marks J. (2006) “South Africa: Evolving Diaspora, Promising Initiatives”, in Kuznetsov Y. Meyer J-B et al. (1997) “Turning Brain Drain into Brain Gain: the Colombian Experience of the Diaspora Option”, Science, Technology and Society vol 2, n°2 : 285-317 Meyer J-B (2001) “Network Approach versus Brain Drain: Lessons from the Diaspora”.

International Migration Volume 39: Issue 5, Special Issue 2 : 91-110 Meyer, J-B., Kaplan, D., Charum, J. (2001) “Scientific nomadism and the new geopolitics of knowledge”. International Social Sciences Journal/Revue Internationale des Sciences Sociales, 168 : 341-54 Meyer J-B (2004a) “Savoirs, diasporas et identités projectives ”. Hermès, 40 : 350-54 Meyer J-B (2004b) “Les diasporas de la connaissance : atout inédit de la compétitivité du sud”. La Revue Internationale et Stratégique, 55 : 69-76 Meyer J-B (2006) “Connaissance et développement: un lien à actualiser”, in Carton M. et Meyer J.-B., La société des savoirs : trompe l’œil ou perspectives ?, L’Harmattan, Paris Merton R. (1973) The Sociology of Science, Chicago, University Press of Chicago Minoian A.V., Freinkman L. (2006) “Armenia: What Drives First Movers and how Can their Efforts be Scaled up?”, in Kuznetsov Y. (ed.)

37

National Research Foundation (2005), SANSA Suvey Results, Icognition, NRF, Pretoria Opiniano J. M., Castro A.T. (2006) “Promoting Knowledge Transfer Activities Through Diaspora Networks: a Pilot Study on the Philippines”, in Asian Development Bank, Converting Migration Drains into Gains, Harnessing the Resources of Overseas Professionnals, Manila Pandey A., Aggarwal A., Devane R., Kuznetsov Y. (2006) “The Indian Diaspora: a Unique Case?”, in Kuznetsov Y. (ed.)

Pécoud A., de Guchteneire P. (2005) “Migration without borders: an investigation into the free movement of people” Global Migration Perspectives, n° 27, GCIM Rey Valette H. (2006) “Collective learning and social capital: the new stakes for governance”, in Carton M., Meyer J-B, The Knowledge Society “Trompe l’oeil or Accurate Perspetives?, L’Harmattan, Paris Saxenian A. L., Motoyama Y., Quan X., Wittenborn D. R. (2002), Local and Global Networks of Immigrant Professionals in Silicon Valley, San Francisco (Ca.), Public Policy Institute of California Stewart J. (1997) Intellectual Capital, London, Nicholas Bradley Turner W., Henry C., Gueye M. (2003) “Diasporas, développement et TICs”, in Barré R., Hernandez V., Meyer J-B., Vinck D. (2003) Diasporas scientifiques/Scientific Diasporas, Paris, IRD Editions UNCTAD - CSIR (1977) Case Study in Reverse Transfer of Technology; a Survey of Problems and Policy in India. Doc TD/B/C/.6/AC.4/6 Wescott C., Asian Development Bank (2005) Promoting Knowledge Exchange through Diasporas, Workshop on ‘Demographic Challenges and Migration’, Sydney, August 2005 Woolcock 2000 Xiang Biao (2006) Promoting Knowledge Exchange through Diasporas (the Case of the People’s Republic of China), University of Oxford, February 2006

38

Appendix 1: New methods of web investigation for networks identification; the limits of cruising in cyberspace with search engines Jean-Paul Wattiaux and Jean-Baptiste Meyer Today, there are more than 8 billion internet pages. Finding relevant networks websites among such floods of information is like navigating in an ocean for the discovery of new islands, without much indication of proximity and localisation.

For the initial systematic search, in 1999, the exploration relied upon repeated queries with a number of keywords deducted from the case studies (Caldas and SANSA) that had given some clues. Such terms as “expatriate”, “knowledge”, “abroad” and so on were then used extensively and applied to various countries names. After weeks of enquiries with pre-google search engines, the harvest of 41 networks was collected. Though this result materialised a number of intuitions that the isolated case studies had provided before, the ‘natural’ method used so far also exhibited its limits. The analysis indeed pointed at two specific difficulties.

1.1- The tremendous vocabulary diversity of networks’ names.

There is not a unique common vocabulary and language for such associations to name and describe themselves. Therefore, querying web search engines with a few generic keywords – for example: “association, professional, abroad” – can only retrieve a minor share of all DKN. As the purpose was to get as complete a list as possible, search engine queries by keywords would supposedly require using an exhaustive list of relevant keywords. And this list could obviously not be elaborated with certainty unless we already knew the complete list of websites. This tautological loop is easily demonstrated by studying the names of the 101 DKN cited in Appendix 1.

The index of all the meaningful words used in all the networks’ names shows a list of 207.

Only three words “Association, Network, American” occur in more than 10% of DKN names.

Only twelve more, “America, Scientists, Indian, Medical, Professionals, Abroad, Ethiopian, Red, Society, Chinese, Engineers, Physician” occur in more than 5% of DKN names.

Respectively, the remaining 192 words have very low occurrence. Of them, 153 have unique occurrence. Those “rare” words may be very effective keywords to find a particular DKN, but guessing these idiosyncratic keywords would be a task as difficult as to directly guess the possible names of all the unknown DKN. On the opposite side, if the fifteen previously cited words which have larger occurrence are used, thousands of answers surge, overflowing the queries, because they are not selective enough. The reason is clearly that all of them (but 39

“abroad”, which has only 7 occurrences) give the search engine no clue about the DKN topic.

To summarise, it could be said that as long as DKN are not explicitly stating by specific words that they are DKN, no keyword set would really allow one to approximate exhaustivity in the DKN website list.

1.2- The search engines technical limits and biases.

The principle of first generation search engines was that of the pioneer AltaVista: to mechanically surf the web, collecting web address of every visited webpage with every meaningful keyword for that webpage, and storing these data in a simple giant database. The power of such search engine is directly related to the number of webpages that have been visited and indexed in its database. Search engine databases have thus grown up to the point where several billions of webpages are referenced. A limitation in efficiency then appeared: the information desired was too often drowned by a flood of unwanted answers. Such billions of webpages database and the correlated answers’ overabundance are still the rule for today’s most popular search engines like Google, though offering a remedy. The latter uses a more sophisticated answering algorithm which gives an ordered list of answers, putting on top of the most frequently requested information by the majority of the people. Technically speaking, the algorithm gives prominence to the most popular websites. They are such tagged if they are recorded as being the most often targets of hyperlinks from pages in other websites.

By trying to take into account what could be the expectations of the majority of its users, the Google algorithm had obviously so greatly enhanced the global satisfaction rate that it became the leading web search engine. But there is an obvious drawback: when individual objective differs from that of most people, the most possibly relevant websites are simply but those best matching the selective keywords. Unfortunately, these specifically targeted websites may well be missed, since they can be drowned far in the depth of the popularity ordered Google websites listing, if they are not often target of web hyperlinks.

The paucity of DKN websites’ referencing may be due to social and technical reasons, such as tightness or closeness of concerned communities, limited creation and use of websites by these communities in their daily activity, or the website manager not having the proactive behaviour necessary to get these referencing hyperlinks from other website managers.

Nevertheless it is as important to know of the existence of such website as of that of the Google popularity contest winner. The most exhaustive list of relevant websites may therefore be more difficult to get by means of Google queries than by querying a first generation search engine.

40

2-A method based on socio-technical assumptions How to find these websites that are not to be found by querying web search engine individually?

A partial technical answer was to use a more powerful tool, that is a meta search engine, which incorporates the answers from simultaneous queries on multiple search engines, from heavy weight such as Google, AltaVista, Lycos and the like, up to less known specialized search engines. Copernic Agent Professional (www.copernic.com) was the one used in this study. Nevertheless, although it has proven itself a real time-saver, and powerful like a search engine based on the largest websites database, the meta search engine does not really bring a remedy to the specific difficulties of our task, because it lays on the answer of the very same ordinary search engine. In the 2002 study, it expanded the previous list of 1999 by mobilising more “search power” but fell short of solving the hasardous character of these enquiries.

The real answer consisted in defining an alternative strategy which does rely primarily on human competence and knowledge, instead of search engines.

First the assumption was made that relevant websites addresses were available somewhere in the Internet, except for when the authors wanted these to be unknown, running as a private Bulletin Broadcasting System (BBS).

Second, an “archaeological” method was used, simply remembering how was searched the Web in the times when there still was no Yahoo, the first large Web search engine in Internet history. In these times, information about websites was disseminated throughout all the Internet, and it had to be found either in discussion forums, either in ones of the many “Links” webpages which just incorporated a large selection of hyperlinks to websites.

Here are what the experience has shown to be the three most effective sources of “Links” webpages, for the 2002 study about diaspora networks: 1. Websites whose topic is related to a curriculum, such as professional federation websites or university alumni websites, 2. Website which addresses to the community of a country, such as embassy or consulate website, diaspora journal website, general diaspora website, 3. Diasporic association websites, like those we were looking for, themselves, since they often have “Links” page pointing to similar websites, as well as pointing to both 1. (curriculum related) and 2. (country community related) websites, 4. Search engines by means of their “Advanced Search” function: “Find pages that link to the page”, since “Links” pages are over present in answer of such requests.

41

The tactics was to surf the Web in an alternating movement, visiting diasporic association websites, from which were found “Links” webpages, where new diasporic association websites addresses were discovered, and so on. Several new diasporic association websites addresses were discovered in short periods of time, thanks to such tactics. Then, the productivity was regularly decreasing, as more and more websites listed in “links” pages had already been passed by. Ultimately, all relevant websites listed in “Links» pages were visited.

Thus, this web surf tactics shows the existence of a set of websites which are tied by hyperlinks. This set of websites is isolated, in the sense that it cannot give links to other new websites. Surfing in such a way permitted the discovery of archipelagos instead of simple islands, to recall the introductory metaphor to this section. But, the search for other groupings started from scratch everytime.

Other tactics were then used, in order to get a point of entry to a new set of hyperlinked websites. Three sources were mobilised: -

websites which were cited in the literature, or given by the experts panel of the 20022003 state of the art,

-

Google or Copernic queries with keywords which were selected not for possibly giving quantities of answer but for picking up at least one diasporic association, for example those four keywords “physician”, “association”, host country, origin country could give us the name of the association of physicians from such origin and working in such country – since it appeared that physicians were more likely than other professionals to constitute diasporic professional associations, that keyword was often used to find just a point of entry,

-

some Usenet forums named after the scheme “country.soc.culture” are classically point of aggregation of expatriates trying to speak with compatriots, and the relation of the diaspora with the origin country are not so rarely matter of discussion. Google offers the archives of these forums, where a few websites addresses have been discovered, which by any other way would have stayed unknown.

In other words, sociological assumptions were made and effective links used (professionnal corporation, ethnic communities and the like) to serve as structural axis in a heuristic process.

In an incremental manner, as the list of websites discovered expanded and the vocabulary enriched, and while geographical holes appeared more clearly, the search focus could improve and every single exploration become more efficient, especially for the 2005 enquiries of which the results are presented in the next section..

42

As a conclusion to this methodological chapter on tracking transnational networks by the web, the mix of automatic information tools and human knowledge should be emphasised. Last generation search engines are very effective in daily use and they permit to dig into an enormous amount of documentation with a tremendous speed. But how sensitive they may be for common enquiries, when turning to specific targets, they need to be fed, oriented, guided and completed by expert knowledge. For DKN studies, such a particular mix of human and technical mediations provides abundant evidence and gives very precise answers to doubts and critiques about the existence and stability of this new research object.

43