|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
1 SOCIAL DYNAMICS AND ACCESS TO SOCIAL CAPITAL OF GGAVATT PARTICIPANTS IN VERACRUZ, MEXICO By SEBASTIAN GALINDO GONZALEZ A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2009
2 2009 Sebastian Galindo Gonzalez
3 To my wife, Paula ; and to our children Sergio, Sebastian, and Ricardo.
4 ACKNOWLED GMENTS First, I would like to thank my supervisory committee. I am very thankful for all your valuable contributions to my formation as a scholar. I thank Dr. Roger Natzke who introduced me to the Extension science and assisted me in pursuing my dreams. I thank Dr. Mirka Koro Ljungberg for opening new dimensions of knowledge for me and constantly challenging my cognitive capabilities. I thank Dr. Mark Brennan for always having his door open for me and guiding me through the realms of community developme nt. I thank Dr. Tracy Irani for teaching me how to conduct a research study. I thank Dr. Nick Place for instilling the passion for Extension in me. Last, but not least, I thank Dr. Glenn Israel for helping me grow, both as a person and as an intellectua l by constantly motivating me to excel and teaching me to learn from my failures and successes. Thanks Glenn for being such a great role model! I am also grateful for the huge support that I have received from the Department of Agricultural Education. I thank Rachel Harris and Holly OFerrell for the wonderful job that they do keeping our Department running. I thank Jodi Modica for always telling me what I needed to do; I would be lost without your advice. I thank my fellow graduate students for their friendship. I thank all the faculty members, staff, and students with whom I took classes, worked, or simply shared a conversation in the hallway, for everything that I have learned from them. I thank Dr. Ed Osborne for his financial support and for prov iding me with challenging opportunities to teach and collaborate in research and extension activities with other graduate students and faculty members. My research could not have been possible without the support of other institutions. I wish to thank t he Mexican Council on Science and Technology (CONACyT) for the invaluable financial support that they provided. I thank the Center for Latin American Studies of the University of Florida and the Tinker Foundation for the grant that allowed me to conduct m y
5 fieldwork. I thank the Universidad Veracruzana, particularly Dr. Raul Arias Lovillo for the support that he provided during my initial years in graduate school. I also thank Dr. Genaro Ruiz Arriaga, Secretar y of Agricultural and Rural Development, Fore stry and Fisheries in Veracruz, and his team for all their help. Thanks go to all the people that have shared so many moments with me during this journey. Thanks to Janice and Alfredo for being such good friends. I thank my parents for all their sacrif ices and hard work to raise such a big and wonderful family. I thank my brothers, Juan and Manuel, for supporting my studies after our father died. I thank Aurora for all her help. Finally, I thank Paula, Sergio, Sebastian and Ricardo for giving me the love and motivation that kept me going during all these years. It is great to be a Florida Gator!!!
6 TABLE OF CONTENTS page ACKNOWLEDGMENTS .................................................................................................................... 4 LIST OF TAB LES ................................................................................................................................ 9 LIST OF FIGURES ............................................................................................................................ 11 ABSTRACT ........................................................................................................................................ 12 CHAPTER 1 INTRODUCTION ....................................................................................................................... 14 Background .................................................................................................................................. 14 Ongoing Extension Reform ................................................................................................. 14 Power of Groups .................................................................................................................. 16 A program for livestock producers ............................................................................. 17 Evaluating g roup -extension p rograms ........................................................................ 19 Social Capital and Agricultural Development ................................................................... 20 Problem Statement ...................................................................................................................... 21 Research Question ............................................................................................................... 22 Purpose of the Study ............................................................................................................ 22 Significance of the Study ............................................................................................................ 22 Overview of Methodology .......................................................................................................... 23 Delimitation of the Study ............................................................................................................ 23 Conclusion ................................................................................................................................... 24 2 THEORETICAL AND CONCEPTUAL FRAMEWO RKS .................................................... 25 Agricultural Extension ................................................................................................................ 25 Forces of Change ................................................................................................................. 25 Extension S ervices for Groups ............................................................................................ 29 Advantages of Groups ......................................................................................................... 30 Assessing the Results of Group -extension Programs ........................................................ 32 Social Capital ............................................................................................................................... 33 Evolution of the Concept ..................................................................................................... 35 The Forms of Social Capital ............................................................................................... 44 Exploring Social Capital ..................................................................................................... 46 Agricultural Extension and Social Capital ................................................................................ 48 What is the GGAVATT? ............................................................................................................ 49 Evaluability Assessment ..................................................................................................... 52 Conceptualization of Social Capital Theory in the GGAVATT ....................................... 53 Summary ...................................................................................................................................... 56
7 3 METHODOLOGY ...................................................................................................................... 63 General Perspective ..................................................................................................................... 63 Context of the Study ................................................................................................................... 65 Population ............................................................................................................................. 66 Quantitative Segment .................................................................................................................. 67 Census ................................................................................................................................... 68 Do Respondents Differ from Non respondents? ............................................................... 68 Instrumentation .................................................................................................................... 69 Procedures Used .................................................................................................................. 71 Recruitment of participants ......................................................................................... 71 Cognitive evaluation of the group parti cipants instrument ...................................... 72 Group administered survey training ........................................................................... 72 Data Collection Process ....................................................................................................... 73 Data Processing .................................................................................................................... 73 Validity and Reliability ....................................................................................................... 75 Data Analysis ....................................................................................................................... 75 Qualitative Segment .................................................................................................................... 76 Participants ........................................................................................................................... 77 Instrumentation .................................................................................................................... 78 Recruitment of Participants ................................................................................................. 79 Data Collection Process ....................................................................................................... 79 Data Organization ................................................................................................................ 81 Credibility of the Study ....................................................................................................... 83 Data Analysis ....................................................................................................................... 83 Summary of Methodology .......................................................................................................... 85 4 RESULTS .................................................................................................................................... 94 Social Capital of Groups and Participants ................................................................................. 94 Distribution of Bonding, Bridg ing, and Linking Social Capital ....................................... 95 Relationships Between Social Capital and Independent Variables .................................. 96 Multivariate Relationships .................................................................................................. 99 Multilevel Regression Results .......................................................................................... 101 Social Capital in Participants Lives ........................................................................................ 104 Bonding Social Capital of GGAVATT Participants ....................................................... 104 Collective story on bonding social capital ................................................................ 104 Structure of th e collective story on bonding social capital ...................................... 107 Insights from Labovian analysis of the collective story on bonding social capital ...................................................................................................................... 108 Bridging Social Capital of GGAVATT Participants ....................................................... 109 Collective narrative on bridging social capital ......................................................... 109 Structure of the collective story on bridging social capital ..................................... 111 Insights from Labovian analysis of the collective story on bridging social capital ...................................................................................................................... 112
8 Linking Social Capital of GGAVATT Participants ........................................................ 113 Collective story on linking social capital .................................................................. 113 Structure of the coll ective story on linking social capital ........................................ 115 Insights from Labovian analysis of collective story on linking social capital ....... 115 Com paring Narrative Structures of the Three Collective Stories ................................... 116 Numbers and Words Come Together ....................................................................................... 118 Summary .................................................................................................................................... 119 5 SUMMARY AND DISCUSSION ........................................................................................... 140 Introduction ............................................................................................................................... 140 Statement of the Problem .......................................................................................................... 140 Review of the Methodology ..................................................................................................... 141 Summary and Discussion of Results ........................................................................................ 142 Recommendations ..................................................................................................................... 148 APPENDIX A IRB APPROVAL ...................................................................................................................... 151 B SURVEY INSTRUMENT FOR GROUP ADVISORS ......................................................... 152 C SURVEY INSTRUMENT FOR GROUP PARTICIPANTS ................................................. 154 D INFORMED CONSENT FOR GROUP ADVISORS ............................................................ 159 E INFORMED C ONSENT FOR GROUP PARTICIPANTS ................................................... 161 F INFORMED CONSENT FOR SOCIAL INTERVIEWS ...................................................... 163 G SUBJECTIVITY STATEMENT ............................................................................................. 165 LIST OF REFERENCES ................................................................................................................. 167 BIOGRAPHICAL SKETCH ........................................................................................................... 178
9 LIST OF TABLES Table page 3 1 Demographics of group advisors in Veracruz, Mexico ...................................................... 87 3 2 Demographics of group advisors of groups responding the survey ................................... 88 3 3 Chi -square statistics from comparing categorical variables of respondents and nonrespondents. ............................................................................................................................ 89 3 4 Indexes created for data reduction using factor a nalysis. ................................................... 90 3 6 Composite multiplicative variables. ..................................................................................... 92 3 7 Interview groups and participants pseudonyms. ................................................................ 93 4 1 Demographic attributes of advisors and respondents. ...................................................... 120 4 2 Descriptive statistics for bonding, bridging, and linking social capital. .......................... 121 4 3 Mean and standard deviation of relevant continuous independent variables. ................. 122 4 4 Percentile distribution of responses for reflexive controls. .............................................. 123 4 5 Mean comparisons of bonding, bridging, and linking social capital across levels of the categorical independent variables. ................................................................................ 124 4 6 Correlations between bonding, bridging, and linking social capital and scale independent variables. .......................................................................................................... 126 4 7 Multiple regression model for bonding social capita l. ...................................................... 127 4 8 Multiple regression model for bridging social capital. ..................................................... 128 4 9 Multiple regression model for linking social capital ....................................................... 129 4 10 Multilevel multiple regression model for bonding social capital. .................................... 130 4 11 Multilevel multiple regression model for bridging social capital. ................................... 131 4 12 Multilevel multiple regression model for linking social capital. ..................................... 132 4 13 Example of co -cons truction in the collective story on bonding social capital. ............... 133 4 14 Narrative structures from the collective story on bonding social capital and their interpreted contribution to social capi tal. ........................................................................... 135
10 4 15 Narrative structures from the collective story on bridging social capital and their interpreted contribution to social capital access. ................................................................ 137 4 16 Narrative structures from the collective story on linking social capital and their interpreted contribution to social capital access. ................................................................ 139
11 LIST OF FIGURES Figure page 2 1 Structure of the GGAVATT model (SAGARPA & INIFAP, 2002). ................................. 57 2 2 Impact model for the GGAVATT program. ......................................................................... 58 2 3 Process theory model of the GGAVATT program. ............................................................. 59 2 4 Theory -based modified impact model for the GGAVATT program. ................................. 60 2 5 Conceptual model for social capital in the GGAVATT. ..................................................... 61 2 6 The three forms of social capital in the GGAVATT program. ........................................... 62 3 1 Schematic representation of the concurrent mixed method used in this study. ................. 86 4 1 Structure of the collective story on bonding social capital. ............................................... 134 4 2 Structure of the collective story on bridging social capital. .............................................. 136 4 3 Structure of the collective story on linking social capital. ................................................. 138
12 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy SOCIAL DYNAMICS AND ACCESS TO SOCIAL CAPITAL OF GGAVATT PARTICIPANTS IN VERACRUZ, MEXICO By Sebastian Galindo -Gonzalez August 2009 Chair: Glenn D. Israel Major: Agricultural Education and Communication Considering that contemporary e xtension systems tend to become decentralized, market driven, plu ralistic, and characterized by farmers taking a leadership role, it is evident th at the coordinated and active participation of farmers in modern e xtension systems is required Extension scholars around the world have proposed the implementation of group-e xtension approaches as an alternative to increase the access to social capital at the local level and help smallholders meet the challenges of development. The need arises then to investigate the mechanisms through which this type of intervention may aff ect social interaction of its beneficiaries. This dissertation is a report of an evaluation study of a Mexican groupextension program The GGAVATT (Livestock Groups for Technology Validation and Transference) is a technology transfer program for livesto ck producers developed in Veracruz, Mexico in the early 1980s. In this e xtension program, neighboring farmers get organized into groups of a minimum of 10 participants to receive technical advice from a research institution through a group advisor. This study makes use of a quantitative and a qualitative segment to explore the suggested association between farmers participation in the GGAVATT program and access to social capital. The broad purpose of this study is to create a profile of the social dynam ics of Veracruzan farmers in terms of the forms and types of social capital that are accessible to them
13 as participants in the GGAVATT program T he results from both segments of this study demonstrate that GGAVATT participants have access to different for ms of social capital which translates in their lives into collective action and empowerment. The quantitative results provide evidence of the positive effect that time and motivation have on access to the different forms of social capital. A gap exists b etween the levels of social capital of men and women participating in this program; stakeholders need to work actively in reducing this gap. The qualitative results show social capital in action and its importance to obtain different benefits. The benefi ts of connectedness are discussed and recommendations are made to teach participants, advisors, and administrators to understand and value them. Finally, t he need for further research is discussed
14 CHAPTER 1 INTRODUCTION Extension scholars around th e world have propos ed the implementation of group-e xtension approaches as an alternative to increase the access to social capital at the local level and help smallholders meet the challenges of development. The need arises then to investigate the mechanism s through which this type of intervention may affect social interaction of its beneficiaries. This dissertation is a report of an evaluation study of a group -extension program found in Mexico. M embership in th is program and access to three different form s of social capital, namely, Bonding, Bridging, and Linking are explored. In addition, t he effects and implications of participant demographics, group characteristics, and community uniqueness on access to social capital are discussed. The background of t he study, the research problem and its significance, and a brief description of the methodology are presented in this chapter. Background Ongoing Extension Reform Multiple authors have documented the changes that both agricultural and rural extension have experienced in modern times. According to Swanson (2008) the structure and scope of extension and agricultural advisory services around the world have been transformed by the effect of social, political, and economic factors during the last few decades. This reform has been a constant process and, as Rivera and Alex (2004a, p. 347) indicate, is inevitable and will continue into the future. In the same way, Mercoiret et al. (2007) write that those same forces of change have affected the production conditions of farmers around the world. Swanson (2008) illustrates this
15 relationship by pointing out how as farmers become more commercialized extension services tend to become privatized. In the Global Review of Good Agricultural Extension and Advisory Service Practices Swanson adds that contemporary extension systems following good practices tend to be decentralized, market driven, pluralistic, and characterized by a leading role of the farmers. Swansons description differs from what was the norm in most developing countries during the 20th century when most extension systems were heavily centralized (particularly Ministry -based systems), driven by technological de velopment and not the market, and highly hierarchical relying on topdown approaches with limited if any, farmer input. From Swansons words we can understand that farmers now have, and will increasingly have, new roles and greater responsibilities in the development process. This is the reason why Qamar (2000) calls extension programs to have the development of the problem solving and decision -making skills of farm ers as their top priorities. Within this context of change, Mercoiret et al. (2007) allege that some farmers have been able to benefit from the opportunities of market openness while a very large number of smallholders have been left behind due to their lack of infrastructure and access to support services. This point is further e mphasized by Rivera and Alex (2004b) who discuss how the poor have less access to training courses and extension activities due to their lesser levels of social capital. In other words, the poor are often limited in extension activities by their lack of social connections that can inform and encourage them to attend these types of events. Similarly, Rogers (2003, p. 460) offers the following generalization in his diffusionist theory: The consequences of the diffusion of innovation usually widen the socioeconomic gap between the audience segments previously high and low i n socioeconomic status.
16 This situation turns even more critical when we consider the large number of smallholders in the world. Swanson (2008, p. 1) observes that there are nearly one billion small -scale farm households in developing countries, with the vast majority facing problems of hunger, malnutrition and poverty. According to Rogers (2003) a good strategy to narrowing the gaps originated by an unequal access to innovations is the formation of groups of smallholders which will endow them with the required economic, political, and so cial strength to promote change. The inference follows that by getting organized into farmers or producers groups, smallholders might actually be able to respond effectively to the current challenges of development. In this same direction, Abaru, Nyakun i, and Shone (2006) argue that farmers organizations can provide farmers with opportun ities to become beneficiaries of the market economy. This means that individual farmers are empowered by getting organized into groups, and this newly found collective identity and voice allow them to participate from a more advantaged position in the deve lopment process. Power of Groups According to Abaru et al. (2006) farmers organizat ions can be valuable allies in development efforts and play a communication role between farmers, extension, and research. This situation can be observed in many developed countries, where farmers organizations play key roles in production, marketing, and representation of producers interests before diverse institutions (Mercoiret, Pesche, & Bos c, 2007) These developed countries are characterized for having supportive policies for organizations with farmers in charge (Abaru, Nyakuni, & Shone, 2006) In s ome countries these farmers organizations have evolved into more complex and powerful specialized commodity groups. The formation of cooperatives or organized groups of farmers has been seen by governments and international donors as a feasible mechanism to make services available to
17 rural populations. For instance, multiple producer organizations were established in Africa, Asia, and Latin America to serve as instruments for the implementatio n of top-down developmental strategies (Mercoiret, Pesche, & Bosc, 2007) Unfortunately, many of these cooperatives have collapsed, particularly those without strong grassroots organization, leaving farmers in poverty and powerless (Abaru, Nyakuni, & Shone, 2006) Another explanation for the disintegration of some of these groups can be found in the words of Swanson (2008, p. xi) who warns of the fact that In most developing countries, public extension systems have been discouraged from organizing farmers, farm women and rura l youth, because these groups could place political demands on the national government. The creation of farmers organizations has given birth also to the development of extension programs tailored to suit the needs of these groups; this is known as groupextension. Cuellar et al. (2006) say that the formation of groups to work towards common objectives in agriculture is a cost -effective way of ensuring sustainable development. For instance, in France, Argentina and Me xico, group -extension has been successfully conducted through the Centers for the Study of Agricultural Techniques (CETA), the Regional Consortiums for Agricultural Experimentation (CREA), and the Livestock Groups for Technology Validation and Transference (GGAVATT), respectively (AACREA, 2006; Roman Ponce, 2001) ; similar groups exist in Chile, Uruguay, Brazil, and other countri es. A Program for livestock p roducers The GGAVATT is a model for livestock technology transfer developed in Veracruz, Mexico, by researchers from the National Institute of Forestry, Agriculture and Livestock Research (INIFAP) in the 1980s. The GGAVATT is defined as a mechanism for technology transfer and validation in which 10 to 20 farmers, from a given geographical region and with similar production systems, get organized into a group with the purpose of increasing animal
18 production and productivity through the adoption of technologies generated in research centers (SAGARPA & INIFAP, 2002) Currently, almost every State in Mexico, and some Central American countries, has adopted the GGAVATT model (Mora Alfaro, 2002) The GGAVATT has certain characteristics that make it an exceptional model of study. Some of these are: (a) farmers have a leadership role in this program because it was originated at the grassroots level as a result of the interaction of farmers and researchers; (b) is a pluralistic model because it has a mixed funding mechanism in which government and group members share the advisors salary; (c) it has survived multiple governmental transitions, both at the state and federal levels; and (d) the structure of the program permits the exploration of inner (within the group), horizontal (between groups), and vertical (between groups and institutions) social interactions. Each group has an advisor that visits the farms of all group members at least once a month. At the beginning of the year, group members and the advisor decide which technologies will be adopted and collectively establish goals for each individual farm. A monthly meeting is held in one of the farms to repor t on the progress of the group, with an emphasis on the results of the farm hosting the meeting. At the end of the year the program is evaluated based on the stated goals; these results are presented in the following monthly meeting and are used to decide in the goals of the future year. Representatives of research and governmental institutions maintain regular contact with the groups through the advisor and may be present during the monthly meetings of each group. Multiple studies have shown the effective ness of the GGAVATT program in improving both economic indicators (Chagoya Fuentes, Gonzlez, Gonzlez, & Gonzlez, 2002; Espinosa
19 Garcia, Wiggins, Gonzalez Orozco, & Aguilar Barradas, 2004) and technological ado ption rates (Galindo Gonzlez, 2001; INIFAP & SAGARPA, 2001; Roman Ponce, 2001) Evaluating group -extension programs In a direct critiq ue of the approaches used for the evaluation of some extension programs, Cabero and van Immerzeel (2007) remark that important achievements of these projects are rarely documented in evaluation reports. They explain that this is due to the narrowness of the evaluation f oci that are concerned only with technical and economic issues ignoring other important matters such as social capital. This position is shared by Mercoiret et al. (2007) who emphasize that rural producer organizations operate on at least three dimensions other than the one comprising the technical and economic activities. These t hree dimensions are: reduction of vulnerability and food security, natural resource management, and last, but not least, human and social capital enhancement. The creation of social capital is not an exclusive property of group -extension programs. In fact most social programs, such as those fostered through extension services, often have as one of their underlying goals the development of social capital, even if their objectives are not articulated in such language (Productivity Commission, 2003, p. XIII) However, the mechanisms through which social capital is accrued or accessed may differ from program to program. It has been proposed that those programs that encourage democratic input and par ticipation from its participants are likely to generate social capital and that it is the responsibility of policymakers to ensure that if social capital is not being created, at least it is not been damaged (Product ivity Commission, 2003) This failure to identify in an explicit way the development of social capital as an intended objective of social programs has marginalized it from being explored through regular program evaluations. Generally, the evaluations of social interventions are focused only in the explicit
20 intended objectives of the program (Rossi, Lipsey, & Freeman, 2004) The risk exists however, that some of these social programs may be, inadvertently, damaging the existent stock of social capital present in the communities where they are operating (Productivity Commission, 2003) Social Capital and Agricultural Development It has been proposed that the alleviation of a wide variety of social ills requires access to sufficient stocks of social capital (The Saguaro Group, 2000) Swanson (2008) believes that to achieve the development goals of contemporary extension systems, many farmers need improved access to markets and a more effective communication with other stakeholders (i.e. policymakers, researchers, and extension providers). He s tates further that these can be achieved by developing social capital in the form of different types of farmer and producer groups. Bourdieu (1986) used the term social capital in reference to the advantages and opportunities available to the individuals as result of their membership in a given network. The advantages and opportunities that Bourdieu is talking about r efer to the acquisition of economic capital or other privileges by virtue of belonging to a certain group. James Coleman (1988) defined the term i n a similar way, as the resources that individuals may derive from their access to social ties. In Bourdieus definition (1986) it is possible to identify three main elements of social capital: networks, norms, and reciprocity. Although Coleman is credited for the introduction of trust as a constituent of social capital, Bourdieu describes something analogous to it w hich he names faith and bad faith to identify its positive and negative directions. Both authors also believe that the different forms of capital can be transformed into each other. For example, social capital can be transformed into economic and human capital. Putnam (2000) identifies two forms of social capital. He calls bonding social capital that which exists within homogenous groups and bridging social capital that which connects
21 different, or heterogeneous, groups. The bonding social c apital of the GGAVATT groups is represented by the strong ties that connect participants within the same group. The GGAVATTs tend to be homogenous groups formed by neighbors of the same community, with a shared ethnicity, and frequently connected by kinshi p ties. Although this characteristic is a strength of the group that is likely to foster trust and cooperation among group members, the participants need to establish horizontal ties connecting with different groups (that may, or may not, have shared demog raphics with them) in order to access more opportunities for development (Beugelsdijk & Smulders, 2003; Narayan & Cassidy, 2001; Uphoff, 1996) The importance of having both forms of social capital can be better understood considering the words of Granovetter (1973; 1983) who says that those ties that connect people with their close friends and family are called strong (like the ones found in bonding social capital), while the ones connec ting with acquaintances are called weak (like the ones present in bridging social capital), and that those weak links are critical to keep individuals updated with information that is unlikely to reach them through their strong ties. Michael Woolcock (2001) describes a vertical form of social capital that connects people with governmental, political, or private institutions, and calls it linking social capital. Probl em Statement If we reconsider the description of a contemporary extension systems offered by Swanson (2008) (i.e. decentralized, market -driven, pluralistic, and characterized by farmers taking a leadership role), the need for coordinated and active participation of farmers in modern extension systems is evident It is important then to explore the sug gested association between farmers participation in group -extension programs and access to social capital. Special attention must be placed in the identification of the activities fostered by this type of program that situate the
2 2 participant farmers as p rotagonists of the social capital process and active co -constructors of a better collective future. Research Question What are the attributes and activities of GGAVATTs that are associated with greater levels of social capital? Purpose of the Study The b road purpose of this study is to create a profile of the social dynamics of Veracruzan farmers in terms of the forms and types of social capital that are accessible to them as participants in the GGAVATT program. The specific objectives of this study are : 1 To provide a description of the social capital of GGAVATT groups and participants by: a ) Describing the d istribution of dependent (i.e. bonding, bridging, and l inking social capital) and independent variables among respondents. b ) Exploring the bivariate relat ionships present between dependent and independent variables. c ) Investigating the multivariate relationships affecting bonding, bridging, and l inking social capital. d ) Determining the distribution of the variation explained by the multivariate models at the in dividual and group levels. 2 To contextualize how social capital takes form in the lives of the participants by looking for evidence of access to social capital, represented through transformations in roles and cooperation, in the stories that GGAVATT parti cipants co construct around their participation in the program 3 To connect the description and context obtained in specific objectives 1 and 2 to represent the complexity and multidimensionality embedded in the interactions fostered by the GGAVATT program and discuss its plausible consequences. Significance of the S tudy The results from this study are valuable at multiple levels. First, GGAVATT stakeholders can use the results to gain a deeper understanding of the program and its possible consequences.
23 Se cond, governmental and non-governmental institutions dedicated to rural and agricultural development can use these findings to improve the design of current and future extension programs. The ideal would be to foster the establishment of programs that prov ide farmers with more opportunities to create, accumulate, and access social capital. Next, policy makers can use the findings to legislate in a way that creates a more favorable environment for the blossoming and operation of farmer led organizations. Nex t, stakeholders of diverse social programs can broaden the scope of their evaluations to explore for possible effects on social capital. They might also want to combine different methods to obtain more relevant information. Last, the results of this study contribute to the body of knowledge on social capital, particularly as it relates with farmers organizations and rural development. Overview of Methodology This responsive evaluation (Guba & Lincoln, 1986) was planned within the overarching notion that social phenomena are highly complex and should be explored using a combination of methodologies. The study follows a goal -free, theory -driven evaluation approach (Chen & Rossi, 1980) to explore access to different forms and types of social capital as an unintended, unanticipated, direct consequence of the participation in the GGAVATT program (Rogers, 2003) The connections between program participat ion and social capital access are investigated with quantitative and qualitative methodologies arranged in a parallel or concurrent mixedmethod design (Tashakkori & Teddlie, 2003) Delimitation of the Study The population for this stud y included only those GGAVATT participants whose groups were receiving governmental monetary support from the Veracruzan Secretariat of Agricultural and Rural Development, Forestry and Fisheries (SEDARPA) at the time of collecting the data.
24 This means tha t the results of the quantitative segment of this study can only be generalized to that population. For the quantitative segment of this study it was decided to conduct a census of the population. However, the number of groups funded by SEDARPA was temp orarily reduced during the data collection period because many groups did not meet the deadline s to submit required paperwork originating a lower response rate than expected. Even though this situation was addressed as nonresponse error it may still be a source of error affecting the generalizability of the results. The survey administration process was particularly challenging because the population was numerous the groups were distributed across a large geographical region the data needed to be colle cted from multiple groups on the same date (because their monthly meetings were already scheduled), and the literacy levels of group participants were very diverse. The group advisors were trained to collect the data from their groups to overcome these ch allenges. This strategy was effective but it is also a source of error capable of affecting the validity of the data. Probably due to the large proportion of program participants that are males and the criteria designed for selecting participants for the qualitative segment of the study, only males were part of the interview groups This is likely to affect the richness of the co -constructed data because it l acks the womens perspective. Conclusion This first chapter has provided an introduction for the present study. The following chapters will progressively present a more complete picture of the theoretical and conceptual frameworks that formed and informed this research, the methodologies that were utilized, and the obtained results. In the final ch apter, the main results are summarized and discussed.
25 CHAPTER 2 THEORETICAL AND CONC EPTUAL FRAMEWORKS Agricultural Extension For centuries, rural inhabitants around the world have relied on the structured exchange of information fostered by different ty pes of extension services to obtain knowledge and develop the skills required to solve their immediate problems and improve their lives (Ensminger & Sanders, 1945; Jones & Garforth, 1997; Masefield, 1950, 1972; Nagel, 1997; White, 1970, 1977) As a consequence of the urbanization process the scope of extension programs has widened and now they are serving a large proportion of urban p opulations. Extension interventions attempt to change human behavior and help individuals, families and communities ac hieve better living conditions using nonformal educational methods to communicate relevant information (Ban & Hawkins, 1988; Rling, 1988) Forces of Change The development of t he diffusion research tradition that started in the 1920s became the leading paradigm in extension. This was particularly true within the discipline of rural sociology where extension evolved and acquired the status of a science cultivating t hree main schools of thought or perspectives namely social interaction, research, develo pment and diffusion, and problem -solving persp ective s (Havelock, 1971; Rogers, 2003; Rling, 1988) Extension has transformed under the influence of both moderni zation, which advocates for the replacement of old practices with new ones (Ensminger & Sanders, 1945; Gwynne, 2003) and dependency, which alleges that underdevelopment could no t be overcame by means of the topdown approaches proposed by modernization (Gwynne, 2003) Elements of these two opposing development theories are present to greater or lesser extent in all extension systems. In addition to the effects of philosophical viewpoints, the structure and scope of extension and agricultural
26 advisory services have been continuously and unavoidably transformed by social, political, and economic factors during the last few decades (Rivera & Alex, 2004a; Swanson, 2008) Mercoiret et al. (2007) mention that those same forces of change have affected the production conditions of farmers around the world. They explain that the intensification and modernization of agriculture, as wel l as the increased integration of rural peoples with markets, have changed the ways in which farmers work. From this we can understand that forces of change affect the extension syste m as a whole and also each of its components individually, which in tur n promotes further internal transformation of the system. Swanson (2008) illustrates these relat ionships by pointing out how as farmers become more commercialized extension services tend to become privatized. In the Global Review of Good Agricultural Extension and Advisory Service Practices Swanson (2008) points out that contemporary extension s ystems tend to be decentralized, market driven, pluralistic, and characterized by farmers taking a leadership role. Swanson indicates that this change results from the inappropriateness of traditional structures to promote effective rural transformation i n the emerging global agricultural economy. From Swansons (2008) words we can understand that farmers have now, and will increasingly have, new roles and greater responsibilities in the development process. This idea is emphasized by Qamar (2000) who notes that extension has moved beyond the passive, top down, technology transfer and currently focuses on the d evelopment of the problem solving and decision -making skills of farmers. This is achieved by encouraging more active and vital functions of farmers within the system mainly through the use of participatory approaches, decentralization, and the promotion of pluralism in the planning, funding, and delivery of extension services. For Rivera (2008) developing an appropriate set of skill s is crucial for
27 capacity building. In other words, Extension interventions should not only focus on the technical training of its beneficiaries but also in a more comprehensive effort of capacity building. As Rivera (2008) claims, training focuses basically on the development of human competence while capacity building encompasse s a broader plan to enable institutions and participants to meet their development goals. Efforts aimed towards capacity building in agriculture would usually require the active and coordinated involvement not only of extension and its participants but al so of other institutions and actors within the Agricultural Knowledge and Information Systems (AKIS); these include markets, government, and research, among others. Mercoiret et al. (2007) allege that within this context of change some farmers have been able to benefit from the opportunities of market openness while a very large number of smallholders have been left behind due to their lack of infrastructure and access to support services. They observe that the market openness resulting from economic and institutional reforms cannot be capitalized by many small and medium size farm ers. The implications of these facts are that smallholders are less likely to take advantage of developmental opportunities and this marginalization will separate them even more from large size farmers. Most smallholders are primarily concerned with ach ieving food security rather than producing for a market and profiting from their produce (Abaru, Nyakuni & Shone, 2006) This point is further emphasized by Rivera and Alex (2004b) who discuss how the poor have less access to traini ng courses and extension activities due to their lower levels of social capital. In other words, the poor are often limited in extension activities by their lack of social connections that can inform and encourage them to attend these types of events. Th is means that poor farmers are likely to become poorer in comparison with those farmers that can access extension services. Rogers (2003, p. xi) offers the following generalization in his diff usionist theory to
28 explain this process: The consequences of the diffusion of innovation usually widen the socioeconomic gap between the audience segments previously high and low in socioeconomic status. A large number of smallholders are restricted fro m accessing markets due both to their lack of capacity to obtain affordable production inputs and their unsatisfied need for adequate technical knowledge that allow them the optimal use of the limited inputs they can afford (Abaru, Nyakuni, & Shone, 2006; Cuellar, Hedlund, Mbai, & Mwangi, 2006) This situation perpetuates poverty and marginalization and for many cases resu lts from not offering smallholders technological solutions appropriate to their realities. This is evidence of the weakness, or inexistency, of linkages between farmers, Extension, and research which limits technological agricultural innovation (Cuellar, Hedlund, Mbai, & Mwangi, 2006) Swanson (2008) explains that with almost o ne billion small -scale farm hous eholds in developing countries governments and donors are under pressure to find solutions suitable for the improvement of rural livelihoods of smallholders. This point is emphasized by the Food and Agriculture Organization of the United Nations (FAO & World Bank, 2001) which holds that currently, and in the foreseeable future, alleviating poverty and hunger in the world would mean to deal with the problems of smallhol ders and their families. Rogers (2003) proposes the formation of groups of smallholders as a good strategy to narrowing the gaps originated by an unequal access to innovations The rationale behind this idea is that this type of organization will endow smallholders with the required economic, political, and social strength to promote change. The inference follows that by getting organized into farmers or producers groups, smallholders are empowered and this might actually allow them to respond effectively to the current challenges of development.
29 Similarly Abaru, et al. (2006) argue that farmers organizations can provide farmers with opportunities to become beneficiaries of the market economy. This means indeed that individual farmers are empowered by getting organized into groups, and this newly found collective personality and voice allow them to participate in a more advantaged position of the development process. However, the formation of groups is not a magical solution or something that should be taken lightly. When groups are form ed in a hurry and without a careful consideration of the local social, economic, and cultural milieu, they are condemned to fail; the identification and promotion of authentic farmers organizations is a big challenge and a process that should be carefully studied (Abaru, Nyakuni, & Shone, 2006) Beyond scale, there are other elements that need to be taken into consideration when the formation of groups is proposed as a development strategy in a given region; among these are: policies and laws that give farmers the control of their organizations, identification and connections with markets, required training and skills that extensionists will need to meet the needs of sustainable farmers organizations, and the detection and improvement of links, or connections, that might be required for growing the group further than grassroots level (Abaru, Nyakuni, & Shone, 2006) Extension Services for Groups Kubo (1979) defines group farming system as a production group composed of more than two farm households who assemble and mutually agree to collaborate in part or fully in the process of agricultural production. Groups of farm ers are formed to share resources (e.g. information, labor, land, technology, etc.) and to work towards common goals. Regardless of the many ways in which groups may be established, one characteristic that remains constant across all types of voluntary as sociation is that participants must perceive a benefit, or an advantage, from their involvement.
30 Advantages of Groups According to Abaru et al. (2006) farmers organizations can be valuable allies in development efforts and play a communication role between farmers, Extension and research. They believe that these organizations play a strategic role strengthening the farmer Extension research linkages and making farmers central characters in the developmental dialogue. Kenyas National Agriculture and Livestock Extension Program insists that groups formed around a common purpose is the most cost -efficient manner to form sustainable development in the agriculture sector (Cuellar, Hedlund, Mbai, & Mwangi, 2006) Local groups and associations are considered valuable assets for sustainable environmental and economic development due to the social connectedness that they promote (Pretty, 2003) In many developed countries farmer s organizations play key roles in pr oduction, marketing, and representation of producers interests in front of diverse institutions (Mercoiret, Pesche, & Bosc, 2007) In those cases, their linkage effect goes beyond research and Extension, connecting farmers with other economic, institutional, and political actors. Well known examples of these organizations are the powerful commo dity groups in the United States. Developed countries are characterized for having supportive policies for organizations with farmers in charge and this has been found to be a factor of success for this type of groups For example, Abaru et al. (2006) identify decentralization, access to micro -credit, available extension services, and libe ralization as some positive policies present in Uganda that favored the blossoming of farmers organizations. Pretty (2003) elaborates on the benefits of a favorable environment as a condition for the creation and sustainability of local groups recommending, among others, the establishment of new incentives, protective regulations, and the removal of destructive subsidies as part of the policy reform required for this purpose.
31 In some regions of the world, the formation of cooperatives or organized groups of farmers has been seen by governments and international donors as a feasible mechanism to make services available to rural populati ons. For instance, multiple producer organizations were established in Africa, Asia, and Latin America to serve as instruments for the implementation of top -down developmental strategies (Mercoiret, Pesche, & Bosc, 2007) For example, the government of Taiwan established the Joint Farming Operations for Rice Cultivation in 1964 in response to a decrease in food production mainly due to migration to cities, non -farm employment, and reduction in size of the farms; this initiative was so successful that it later expanded to other crops and livestock (P i Feng, 1979) The Uganda Land Management Program (ULAMP) has successfully adopted this approach building representative farmers organization at the grassroots level (Abaru, Nyakuni, & Shone, 2006) These associations have also worked well in developed countries like The Netherlands where farmers see cooperatives as business organizations. However, the y have failed in some developing countries due to both external and internal constraints (Braverman, Guasch, Huppi, & Pohlmeier, 1991) Many of the cooperatives that have collapsed lacked strong grassroots organization and their disintegration has left farmers in poverty and powerless (Abaru, Nyakuni, & Shone, 2006) Another explanation for the disintegration of some of these groups can be found in the words of Swanson (2008, p. xi) who warns us of the fact that i n most developing countries, public extension systems have been discouraged from or ganizing farmers, farm women and rural youth, because these groups could place political demands on the national government. It is important to keep in mind that the formation of groups and other participatory strategies have caused, as a direct consequence the activation of segments of the population that may have been historically marginalized.
32 The formation of cooperatives and other organized groups of farmers produced the proliferation of Extension interventions designed to serve the needs of the se organizations. This situation has fostered the development of a new approach known as groupExtension In France, Argentina, and Mexico group-Extension has been successfully conducted through the Centers for the Study of Agricultural Techniques (CETA) the Regional Consortiums for Agricultural Experimentation (CREA), and the Livestock Groups for Technology Validation and Transference (GGAVATT), respectively (AACREA, 2006; Roman Ponce, 2001) Similar groups and experiences exist in many countries of the world. Assessing the Results of G roup -e xtension P rograms In a direct critique of the approaches used for the evaluation of some Extension programs, Cabero and van Immerzeel (2007) remark that important achievements of these projects are rarely documented in evaluation reports. They explain that this is due to the narrowness of the evaluation foci that are concerned only with technical and economic issues ignoring other important matters such as social capital. This is a common practice because evaluations of social interventions tend to be focused only in the explicit intended objectives of the program (Rossi, Lipsey, & Freeman, 2004) Cabero and van Immerzeel continue presenting the cas e of projects using the Raymi methodology, which are highly participatory and aimed to the diffusion of indigenous knowledge, alleging that social capital, self -esteem, and cultural energy are traditionally neglected in evaluation efforts even though they have shown to be important capacity building elements of these programs. This position is shared by Mercoiret et al. (2007) who emphasize that rural producer organizations operate on at least three dimensions other than the one comprising the technical and economic activities. These three dimensions are: reduction of vulnerabili ty and food security, natural resource management, and last, but not least, human and social capital enhancement. They accept that agricultural production is probably the greatest
33 concern for group participants due to its central role in satisfying basic needs of their households, however, they hold that no group could be sustainable if it focuses exclusively on just one of these areas. The creation of social capital is not an exclusive attribute of extension programs i n fact, most social programs ofte n have as one of their underlying goals the development of social capital, even if their objectives are not articulated in such language (Productivity Commission, 2003, p. XIII) It has been alr eady established that t his failure to identify in an explicit way the development of social capital as an intended objective of social programs has excluded it from being explored through regular program evaluations. T he mechanisms through which social ca pital is accrued or accessed may differ from program to program. It has been proposed that those programs that encourage democratic input and collaboration from its participants are likely to generate social capital. As Pretty (2003) notes, this is particularly true for those initiatives directed to the management of natural resources in which collective action has been institutionalize d through different local organizations (e.g. water users groups, youth clubs, church groups, etc.). He goes on to say that the destructive impact that agriculture has had on the environment is in great measure the result of the insistence of governments and developmental institutions of changing the behavior of individuals rather than that of groups or communities. What is clear is that it should be the responsibility of policymakers to ensure tha t if social capital is not been created, at least it is n ot been inadvertently damaged (Productivity Commission, 2003) Some social programs may harm social capital by creating a strong separation and poor communication between their participants and others in the community. Social Capital Enough has been said about social capital and its relationship with groupExtension approaches; however, the term has not been properly introduced. T he views of Pierre Bourdieu
34 presented in his iconic text on the topic The forms of capital (1986) are used to organize our thoughts. For Bourdieu, capital is a tangible or intangible, embodied or institutionalized, intrinsic power suitable of being accumulated, that exists in different forms and that represents an immanent social law. He believes that without conceptualizing capital in all its forms it is impossible to make sense of the structure and functioning of the social world. The three forms of capital that Bourdieu describes are economic capital, cultural capital, and social capital. Social capital has been defined as a community attribute, and is one of the resources that organizations and communities can invest in (Flora & Flora, 2003, p. 214) Social capital is the collection of supportive interpersonal interactions that exist among individuals (Israel & Beaulieu, 2002) Social capital remains in some ways a blurred concept, maybe in part because scholars from multiple social disciplines h ave focused on the same phenomenon but had tried to understand and explain it using the discourses of their particular disciplines. This has originated an overlapping of ideas and an impressive amount of publications talking about trust, networks, norms, and other elements as essential factors for collective action, and therefore, for the preservation and development of society. Portes (1998) maintains that scholars are reaching the consensus that s ocial capital is represented by the capacity of social actors to secure benefits accessible to them by virtue of their membership in social networks. Jules Pretty (2003) concurs and adds that social capital facilitates cooperation because it lowers the costs of social transactions because people have the required confidence to invest in collective activities; this is also consistent with Colemans propositions (1988) Pretty (2003) lists four features of social capital: (a) relations of trust; (b) reciprocity and exchanges; (c) common rules, norms, and sanctions; and (d) connectedness, networks, and groups.
35 The notion that local involvement and collective acti on can be beneficial both for individuals and communities is not new. According to Portes (1998, p. 2) even though the origins of this concept can be traced back to the writings of social theorists like Durkheim and Marx, the attractiveness of social capital is two -fold: First, the concept focuses attention on the positive consequences of sociability while putting aside its less attractive features. Second, it places those positive consequences in the framework of a broader discussion of capital and calls attention to how such nonmonetary forms can be important sources of power and influence, like the size of one s stock holdings or bank account. Evolution of the Concept The term social capital was coined in the early 1900s by the rural educationalist Lyda Judson Hanifan, referring to goodwill, fellowship, mutual sympathy and social intercourse among a group of individuals and families who make up a social unit, the rural communi ty (Hanifan, 1916, p. 130) The term was used a gain almost fifty years later by Jane Jacobs (1961) to describe the i mportance of social networks for the life of urban settings. There are two main perspectives originated in the 1980s from which current definitions of social capital have been derived, one from Bourdieu and the other from James Coleman. Bourdieu (1986) used the term in reference to the advantages and opportunities available to individuals as result of their membership in a given network. These advantages and opportunities are related with the acquisition of economic capital or other privileges by virtue of belonging to a certain group. For Bourdieu (p. 248) social capital is the aggregate of the actual or potential resources which are linked to possession of a durable network of more or less institutionalized relationships of mutual acquaintance and recognition or in other words, to membership in a group. He explains further that membership is not enough to access its social capital; members need to engage in symbolic and/or material exchan ges conducted under the
36 guidance of a whole set of instituting acts. In other words, participants in any organization should be active and act according to the norms of the group in order to be recognized as a true member and access the benefits of the as sociation. These benefits that may be derived from a given network are diverse and depend upon the nature of the organization. However, if we retake Bourdieus notion of three forms of capital needed to understand the social world, we can assume that in general terms these benefits would be manifested as economic, cultural, or social capital (Portes, 1998) For example, participants in a group -extension program can have economic benefits (e.g. technical informati on to increase production, access to markets), increase their embodied cultural capital with what they learn from peers and technical advisors, and they might even receive a certificate from an educational institution certifying their expertise on a partic ular practice, such as artificial insemination, which would represent an institutionalized form of cultural capital. For Bourdieu (1986, p. 252) individuals in a network also realize a long -term investment in sociability for which the ideal return will come from other individuals in the network (or foreign to the network but as a consequence of his /her membership) in the form of a nonspecific indebtedness which is called gratitude. Portes (1998) supports this notion indicating that individuals deliberately invest economic and cultural capit al to obtain social capital. Bourdieu (1986) explains that the diverse benefits that build up from socialization in the network is what constitutes the basis for solidarity in the group and he adds that these profits are not necessarily pursued in a conscious way, rather they arise from a more or less continuous series of exchanges through which recognition is affirmed and reaffirmed. With this comment Bourdieu indicates that recognition, or reciprocity (Gouldner, 1960) is at the same time an important norm and a product of the socializati on that mediates the activity within the group generating social
37 cohesion. At this point it is very important to remember that social capital is the capability of obtaining certain benefits by virtue of participating in a given network and should not be c onfused with the benefits themselves (Portes, 1998) It is possible to identify three main elements in Bourdieus (1986) social capital definition: a social network, norms, and recognition. The common rules and norms, including recognition, are designed to moderate individual behavior plac ing the group interests above those of the individuals ; this produces the environment of confidence that motivates members to invest their different resources in collective activities (Pretty 2003) Due to their central role as the settings where social activity occurs social networks are basic elements for socialization. Social networks are the result of purposeful investments of individuals or groups to establish or reproduce relationsh ips that might be useful in the short, medium, or long term (Bourdieu, 1986) ; these investments may be done consciously or unconsciou sly. Social ties, or invisible bonds, are created by the interaction between individuals. The ties that connect people with their closer friends and families are called strong tieswhile the ones connecting them with acquaintances are called weak ties (Granovetter, 1973; 1983) Most of the ties present in a given community are weak ties, or in other words, low -density networks. It has been proposed that an increase of the width and density of ties might be considered an accumulation of social capital or societal thickness (Fox, 1996) This is consistent with Bourdieu (1986) who claimed that an increase in the size of the network would be reflected in an increase on the stock of social capital. However, the network is only one element in social capital, and even though it would be expected to believe that modifications in its structure would affect the potential social capital that might be accessed through it (Blau, 1977) there are also cognitive elements (e.g. norms and recognition) that definitely affect how
38 much social capital may ultimately be available. Both formal (e.g. school) and informal (e.g. family) netwo rks are forms of social capital (Putnam, 2001). Granovetter emphasizes the importance of weak links in keeping the individuals updated with information that is unlikely to reach them through their strong ties. From this we can understand that the greater the numbers of weak ties that people have the greater their opportunities to receive information that might be beneficial for them. Access to information is considered a key factor affecting rural poor contact with extension (Rivera & Alex, 2004b) and has been used as a proxy indicator of social capital (Grootaert, Narayan, Nyhan Jones, & Woolcock, 2004) A study from Indonesia exploring the impact of social capital on household welfare and poverty suggests that social capital increases the sha ring of information, reduces exploitive behavior, and improves the collective decision making process (Grootaert, 1999) Godquin and Quisumbing (2005) conducted a longitudinal study of group membership and social networks of rural households in the Philippines. They found that households that are richer, better educated, and living closer to urban areas are more likely to participate in groups and to have access to greater social and economic networks. James S. Coleman (1988) also developed a concept of social capital to explain social action. He did so by combining elements of rational action and characteristics of the social context, or in other words, integrating elements from sociological and economical traditions. For Coleman, as it is for Bourdie u, social capital is productive because it allows people who posses it to obtain certain benefits that would be inaccessible to them otherwise. For him it is the function of social capital which ultimately defines it. Coleman (1988, p. 98) goes on to say that social capital is not a single entity but a variety of different entities, with two elements in common: they all consist o f some aspect of social structures, and they facilitate certain actions of actors whether persons or corporate actors within the structure. This idea of looking at social capital
39 both in terms of structural attributes of the network and the processes that take place within this structure has inspired other researchers to study the topic. For example, Israel et al. (2001) looked at the ro le of process and structural attributes of family and community social capital on educational achievement of high -school students. They found strong evidence of a positive effect of both process and structure attributes on educational achievement, particu larly in the case of family social capital, although the effect was also present for community social capital. In subsequent studies, Israel and Beaulieu (2002, 2004) explored the structura l and process attributes of family, school, and community social capital on students performance on standardized tests, where they found further evidence of the important role that these three forms of social capital play in the development of human capit al. Bourdieu believes that all three forms of capital (i.e. economic, cultural, and social) are fungible, meaning that it is possible to change units of one type for units of another; but there is always a cost associated in those transformations. For Co leman, the fungibility of social capital (and the other forms of capital that he calls physical and human) might be specific to certain situations. In other words, even though people posses certain stock of social capital it is possible that they are unable to use that asset for securing a determined benefit. For example, membership in a farmers organization could be very productive to access nonformal training on farming practices but may be totally useless for be ing accepted in to a universitys graduat e program. Coleman (1988) conceptualizes social capital as an attribute of society. He explains that social capital exists in the relations bet ween and among members of a given social network. This is consistent with Bourdieus (1986) notion that membership w ithout socialization is not enough. In other words, if social capital resides in the relations between and among members of
40 the network, the only way of accessing it is by actively engaging in social transactions within that network. However, it has been proposed that social capital is affected by characteristics of individuals and groups (Alesina & La Ferrara, 2000; Glaeser, Laibson, Scheinkman, & Soutter, 1999) Coleman (1988) intr oduced the element of trust in the definition of social capital, while retaining networks, norms, and activity as key parts of the concept. Trust acts like a lubricant that enhances cooperation, reduces transaction costs, and helps create social obligatio ns (Alesina & La Ferrara, 2000; Pretty, 2003) The role of trust in sociology can be trace d back to social exchange theory (Blau, 1977; Homans, 1958, 1974) that explains socialization as an exchange process involving certain costs and rewards for the individuals. In a review of trust and d i s trust in organizations, Kramer (1999) describes five types of trust: (a) disp ositional trust relates to the fact that individuals naturally differ significantly in their predisposition to trust others; (b) history based trust is associated with the tendency of individuals to make a decision to trust, or distrust, someone based on t he results of cumulative previous experiences; (c) category -based trust is that in which people are more likely to trust, or distrust, someone based on group membership; (d) role -based trust describes a situation in which people are trusted, or distrusted, based on the role that they have and not on who they are, for example, some people are trusted because certain degree of credibility or expertise is associated with the function that they perform; (e) rule -based trust explains how the set and enforcement of formal and informal rules affect the trust that people place in other within an organization. When people trust they are willing to commit resources to a given activity even though the results of that activity depend on the decision of others to collaborate (Glaeser, Laibson, Scheinkman, & Soutter, 1999) The more people trust, the more likely th ey are to engage in social activities; however, people are more
41 inclined to trust others that are similar to them (Alesina & La Ferrara, 2000) Trustworthines s of a network is increased when people behave in ways that increase the trust of others in the network that trusted them (Glaeser, Laibson, Scheinkman, & Soutter, 1999) According to Coleman (1988) trustworthiness of the network is highly d ependent of the actual structure of the network. For him, the closure of the social structure is directly proportional to the existence of effective norms and trustworthiness Closure means the existence of sufficient ties between a certain number of pe ople to guarantee the observance of norms (Portes, 1998, p. 6) Coleman particularly refers to the higher levels of apparent trust present in homogenous closed communities that allow social and eco nomic activities to occur with ease. He provides some anecdotal examples to explain how closure of the network affects these apparent levels of trust. The structural effects of the network on peoples behaviors have been described previously (Blau, 1960, 1977) and some empirical studies have relied on this characteristic of the network to produce recommendations directed to i mprove different outcomes related with community development (Israel & Beaulieu, 2002; Israel & Beaulieu, 2004; Israel, Beaulieu, & Hartless, 2001) Eve n thou gh Coleman tried to keep the concept of reciprocity out of his definition of social capital, it seems that his notion of trust and Bourdieus recognition might work well together. In fact, some scholars believe that trustworthiness is the same as reciproc ity (Ashraf, Bohnet, & Piankov, 2006) Fo r example, people may trust people that they know because they share membership in one or more network and people that they do not know because they have a general level of confidence resulting from previous experiences with strangers from a determin ed social context (Pretty, 2003) In both of the previous cases a certain degree of recognition or reciprocity are present and affect the trust that a given individual is willing to place in a social
42 exchange. From this we can understand that trust is affected by both the networks structure and past social processes and changes on it are capable of affecting future process. The term social capital was popularized by Robert Putnam in his famous book Bowling alone: the collapse and revival of American community (2000, p. 19) where he defines social capital as the set of connections among individuals social networks and the norms of reciprocity and trustworthiness that arise from them and contends that social contacts are as valuable to the productivity of individuals and groups as either physical or human capital. He had used the term previously to describe some local cultural traits affecting community agency (Putnam, 1993) The term community agency refers to the capacity for collective action possessed and shared by the members of the community (Luloff & Bridger, 2003) Agency is a state of independence, a right of the people, the empowerment to use their capacity to create, to change, and to live according to their own meaning systems (Bhattacharyya, 1995) For Putnam low levels of social capital are the cause of numerous social problems affecti ng both developed and developing countries and he identifies civic engagement as a key indicator of community social capital. Community development requires that communities have the capacity to mobilize the resources needed to identify and respond to its own needs (Kelly & Caputo, 2006) Social participation, through civic engagement and community action, is the engine that generates the emergence of the community field and its development (Brennan, 20 04) Civic engagement is the catalyst to transform the interaction between residents into significant focused efforts to develop community (Wilkinson, 1991) Local part icipation in development efforts provides insider knowledge and promotes empowerment and interaction of residents (Dinham, 2005) People can be empowered and obtai n a voice by engaging in civic associations. Unfortunately, the people in most need of a voice often lack the time or the resources to get involved (Powell &
43 Geoghegan, 2006), or the opportunity to volunteer in civic service may be accessible only for a few privileged individuals (McBride, Brav, Menon, & Sherraden, 2006) For Robert Putnam (2000) social capital moves in a continuum between the micro and the macro level and can be considered both a public and a private good. The connections available to individuals help them achieve their individual goals, and at the same time, these well connected individuals represent an asset for the extended group, and their accumulation of social capital is part of the communitys stock. Putnam (2001) agrees with Coleman in that some forms of social capital are good only for specific purpos es, and concurs with Granovetters claim on the importance of weak ties. He also warns that not all consequences of social capital are positive for the community in general, in fact, like in the case of organized criminal groups these effects may be highl y damaging. Social capital may have other negative consequences such as discrimination when homogenous groups marginalize others who are different from access to certain resources (Portes, 1998) Putnam (2000) situates the concept of reciprocity as part of the mutual obligations that individuals acquire from membership in a given network. A very interesting conceptualization in Putnams understanding of the norm of reciprocity is when he observes that r eciprocity is not always specific. In other words, people sometimes do things for others expecting something in return from the same persons; in this case reciprocity is specific. However, Putnam explains that people may also help others with the expecta tion that some one will help them if they ever need it; he calls this generalized reciprocity and it has also been called diffuse reciprocity by Pretty (2003) Is possible again to observe how the norm of reciprocity, the level of trust, and the activity in the network are intrinsically linked. According to Pretty (2003) reciprocity and social exchanges are likely to increase trust; that is as long as they are in line with what is normally
44 expected in the particular social context where they occur. Reciprocity remains a very controversial element in social capital theory because as every norm is a social construction highly dependent on context and situatedness (Moody, 2008) The network is considered to be a structural element of social capital, while trust and norms are cognitiv e elements of the construct (Krishna & Shrader, 1999) The Forms of Social Capital Putnam (2000) distinguishes two forms of social capital. He calls bonding social capital the one that exists within homogenous individuals. In other words, it is the capital that exists in the relationships among people with similar backgrounds, demographic attributes, place of re sidence, interests, etc, and is visible by the presence of different local organizations (Pretty, 2003) For example, the bonding social capital that exists in a given neighborhood contributes to the likelihood that neighbors would come together to address common problems (Larsen et al., 2004) The shared culture and a common identity influence the confidence of rural communities to come together and address their problems (Brennan, 2005) Community development is in the interest of the whole local residents, and the input and participation of each one is required. The second form of social capital that Putnam (2000) describes is bridging social capital This form represents another type of horizontal connection that link individuals with others that may differ f rom them in multiple dimensions. This social capital is the one that exist in the relationships among heterogeneous individuals, principally connecting people across communities organizations or groups (Pretty, 2003) Reconsidering Bourdieus (1986) proposition con cerning the relationship between size of the network and stock of social capital and Granovetters (1973, 1983) theory of the strength of weak ties, it is evident that peopl e with access to bridging social capital would have greater chances of procuring benefits than those limited to socializ ing only with their peers.
45 Michael Woolcock (1998) integrates a new perspective where he reconsiders the effects of bridging and bonding social capital at the micro and macro levels. He talks about integration and synergy when bonding s ocial capital acts at the micro and macro levels, respectively. And he names linkage and organizational integrity the effects of bridging social capital at the micro and macro levels, respectively. Woolcock situates this novel approach in a context where development is occurring simultaneously in bottom up and top-down directions across the micro and macro levels and characterized by participation of civil society in formal institutions. However, he claims that besides the already identified connecti ons within groups (bonding) and between groups (bridging) there are other connections needed. The connections that are missing are the ones between civil society and ma cro level institutions of the public and corporate sector s This third form of social capital is known as linking social capital (World Bank, 2001) and is formed by the vertical connections that connect individuals and groups with institutions. Through linking social capital groups are capable of interacting with different type of institutions to modify their policies or download resources (Pretty, 2003) It has been proposed that it is no t possible to have bridging social capital without having bonding social capital first (Putnam, 1993; Schuller, 2007) and that enough evidence exist that both forms of social capital are different constructs (Larsen et al., 2004) However, Schuller (2007) claims that the distinction between bonding and bridging social capital is completely context dependent and in most occasions is very difficult to determine when one, or another, is acting. In many ways the same can be said about the distinction between bridging and linking social capital, which may suggest that the three forms, although different constructs, share common elements and are part of a continuum. Schuller adds that there is not one group completely homogenous and with members lacking any type of connection to others in different
46 networks and that, therefore, the interest should be placed in exploring how all forms of social capital work together connecting both homogeneous and heterogeneous groups across the community field. This synergy is key for community development to occur which requires local residents to come together and act (Wilkinson, 1991) Explori ng Social Capital Multiple attempts have been made to explore the presence of social capital in diverse settings. Due to the inherent intangibility of the construct, researchers have used their ingenuity to develop different approaches to its study. Some of these approaches have included the use not only of the already described constituents of social capital but also some of its proximal and distal indicators (Stone, 2001) For example, Christian Grootaert, Lead Economist in the Social Development Department at the World Bank, and his colleagues developed in 2004 an Integrated Questionnaire for the Measurement of Social Capital designed to be used on developing countries for the collecti on of quantitative data at the household level (Grootaert, Narayan, Nyhan Jones, & Woolcock, 2004) Their questionnaire includes questions on six dimensions of social capital: (1) groups and networks, (2) trust and solidarity, (3) collective action and cooperation, (4) information and communication, (5) social cohesion and inclusion, and (6) empowermen t and political action. The first one is concerned with structural social capital, the second deals with cognitive social capital, the third one focuses on the processes of social capital, and the final three explore proximal indicators that have been lin ked to social capital. A companion of this instrument was published in 2006 and focuses on the use of qualitative methodologies to collect contextual data (Dudwick, Kuehnast, Nyhan Jones, & Woolcock, 2006) Another well known study was conducted by Onyx and Bullen (2000) who surveyed 1,200 adults from five communities in Australia exploring cognitive social capital (i.e. trust and
47 norms), structure and process of social capital (i.e. participation in networks), and social agency as a proximal indicator of social capital. Using factor analysis they were able to identify how these elements grouped into seven unidimensional factors in their data. These factors are: (a) participation in the local community, (b) social agency, (c) f eelings of trust and safety, (d) neighborhood connections, (e) family and friends connections, (f) tolerance of diversity, (g) value of life, and (h) work connections. Other studies have focused only on incorporating the main elements and forms of social capital into their approaches and conduct a detailed exploration of each one of them. A very good example comes also from Australia where Edwards (2004) looked at the quality of the networks (looking at levels of trust and reciprocity) the structure of the networks (e.g. size, openness, density, etc.), the transactions or processes that occur within the network, and finally at the different forms of social capital (i.e. bonding, bridging, and linking). There are many more studies on social capital using a variety of indicators to study the concept product of the multiple definitions and interpretations of the term. Schuller (2007) warns that in a case like this it is even more important to assess the validity of any attempt to study social capital based on the internal consistency of the particular study between the theoretical and conceptual frameworks presented and the methodology followed. Researching cooperatives and other groups of farmers multiple studies have found an important rol e of family and kinship in the success of these organizations (Bennett, 1983; Hu, Huang, Hendrikse, & Xu, 2007; Huppi & Gershon, 1990; Kubo, 1979; Reed, 1979; Wong, 1979) It has already been discussed that this type of connections are mainly represented by dense networks with strong connections among fairly homogeneous individuals; in other words, this would represent a fundamental form of bonding social capital. Therefore, it would be valid
48 to believe that the presence, or absence, of this type of conne ctions may have an effect on access to social capital making it a relevant proximal indicator to explore. Agricultural Extension and Social Capital It has been proposed that the alleviation of a wide variety of social ills requires access to sufficient s tocks of social capital (The Saguaro Group, 2000) Swanson (2008) believes that to achieve the development goals of contemporary extension systems, many farmers need improved access to markets and a more effective communication with other stakeholders (i.e. policymakers, researchers, and extension providers). He states further that these can be achieved by developing social capital in the form of different types of farmer and producer groups. These local organizations represent a structural form of social capital that are likely to provide farmer members with opportunities to obtain different benefits that could be useful to improve their productivity, obtain more skills, enhance the quality of their products, or any other progress that they need to improve their lives. But how ex actly does that process work? Extension was defined at the beginning of this chapter as a human activity striving to improve the lives of individuals through nonformal education that will provide them with the knowledge and skills to solve their immediate problems. Therefore it would be valid to claim that extensions mission is, essentially, the timely development of relevant human capital. According to Rivera (2008) the development of human resources is the ultimate goal of development and, besides school systems and the private and public workforce organizations, a large number of people rely upon nonformal educational programs, like the ones fostered by extension, to obtain this education. However, as Coleman and other scholars have demonstrated, there is a clear role of social capital in the formation of human capital. Accor ding to Balatti and Falk (2002, p. 296) the effects of learning are only spilled -over on the society through social capital because: Social capital is implicated in effective adult learning
49 in three most important ways: 1. The processes of drawing on and building social capital are intrinsic to the learning process. 2. The processes of drawing on and building social capital are directly or indirectly implicated in realizing socioeconomic outcomes from adult learning. 3. The realization of socioeconomic benefits through learning is broug ht about as much through the learners identity formation and reformation as through knowledge and skills. The stock of social capital present in a given community allows for the integration of learned skills and knowledge into the everyday life of the c ommunity (Kessels & Poell, 2004) and, according to the OECD (2001), learning and economic success cannot be achieved when social capital is lacking. Schuller (2007) explains that both human and social capitals are harder to obtain and to value when they are not linked together. Rivera (2008) identifies supporting commercial interests, the promotion of a knowledge economy, and the advancement of the worth of human capital as the three main challenges that extension is currently facing; to achieve these tasks sufficient social capital would be required. Extension programs can successfully promote social capital development (Pretty, 2003) if they recognize, and act accordingly: (a) that the knowledge and skills generated from the active engagement of farmers and extensionists repr esent a communal property; (b) that social interventions which include processes mediated by trust, reciprocity, and cooperation are more effective to disseminate information both across participants and nonparticipants; (c) that the promotion of social le arning should be the core of their programs; and (d) that social processes are very important for the development and diffusion of technologies. What is the GGAVATT? In Mexico, many institutions of the federal government, NGOs, and private sector have ha d a role in extension activities over the years. With different levels of success, Mexico has used several extension models and approaches for promoting rural development at the national,
50 regional, and local levels. One of these extension models is the G GAVATT. The acronym GGAVATT stands for Grupos Ganaderos de Validacin y Transferencia de Tecnologa (Livestock Groups for Technology Validation and Transference). The GGAVATT is a model for livestock technology transfer developed in Veracruz, Mexico, b y researchers from the National Institute of Forestry, Agriculture and Livestock Research (INIFAP) and farmers from the northern region of the state between the late 1970s and the early 1980s. The idea for the program originated when one researcher, wh o have just recently returned to Veracruz after obtaining a doctoral degree in the University of Florida, noted that every weekend when he visited his familys farm, where he was implementing new technologies, neighboring farmers would come to see what he was doing and to ask questions about ways in which they could also try some of these innovations. He shared this experience with his colleagues at INIFAP and they decided to replicate the experience with another group of farmers closer to their research c enter. These early experiences were the basis to develop a methodology to work with the farmers. The informal groups became formal when the program was institutionalized by INIFAP and the rules of membership in these organizations were established in con junction with the farmers. Other institutions became involved with the program to attend the needs of the many groups proliferating across the state. Today, the stakeholders of the GGAVATT program in Veracruz include: (a) participating farmers; (b) Farme rs Unions (i.e. North, Central, and South); (c) local farmers associations; (d) local, state, and federal governments; (e) educational and research institutions (e.g. INIFAP, University of Veracruz, and UNAM); and (f) Produce Foundation, among others. I n other words, the program was born because of the willingness of neighboring farmers to get organized to learn and collaborate
51 (bonding social capital) and has evolved based on the capacity of these farmers to effectively link with other organizations and institutions (bridging and linking social capital). The GGAVATT is defined as a program for technology transfer and validation in which 10 to 20 farmers, from a given geographical region and with similar production systems, are organized into groups wit h the purpose of increasing animal production and productivity through the adoption of technologies generated in research centers (SAGARPA & INIFAP, 2002) The general objective of the GGAVATT program is to increase the production and profitability of the participant farmers through technology transfer; the model also aims to improve the quality of life of the participants and their families as well as promoting the conservation and rational use of the natural resources (SAGARPA & INIFAP, 2002) In its beginnings, the scope was limited to those farmers producing dual -purpose cattle, today, there are groups working with dairy, beef, swine, ovine, goats, and bees (Roman Ponce, 2001) and the program is accessible to any livestock farmer interested in adopting new technologies. The GGAVATT program recognizes three components in its organization: (a) the group participa nts are known as the livestock component; (b) the group advisor, or technical component of the program, is a person with a background in agricultural sciences appointed by the group participants; and (c) the institutional component that is responsibl e for the development and validation of technology. Figure 2 1 shows a representation of the GGAVATT model. The different organizational activities that each GGAVATT group needs to carry out, like electing officials, appointi ng a technical advisor, linking with a research institution, among others, promote the activation of participants through empowerment and democratic practices. The monthly and annual meetings are key processes in the GGAVATT program and it is during
52 these activities that a big part of the socialization takes place. Each month one of the members of the group is responsible for the organization of the meeting in his or her house; the whole family of the participants is traditionally involved in the preparat ions and the meeting adopts a festive character. During these monthly meetings the groups officials and advisor report to the other participants on the status of the different scheduled goals and activities. At the end of the year a meeting is held to p resent the results of the annual evaluation and to establish the goals for the next year. All the official activities required during these meetings are conducted in a format to which the farmers are habituated; these are the same procedures followed regu larly during the meetings at the local or regional Farmers Unions. In addition, GGAVATT participants may attend regional and national meetings where institutions, farmers, and advisors have a greater opportunity for interaction, networking, and sharing e xperiences. In addition to the group meetings, the program conducts monthly regional meetings where group advisors and representatives from research and governmental institutions get together to exchange information and plan different activities. During the first five years of existence of a group, the salary of the advisor is paid jointly by the group participants (all members contributing the same amount) and the government (state and federal funds in equal proportion). The federal funds come from the Mexican Secretariat of Agriculture, Livestock, Rural Development, Fisheries and Food (SAGARPA) and the state funds are provided by the Veracruzan Secretariat of Agricultural and Rural Development, Forestry and Fisheries (SEDARPA). It is the latter institu tion who manages the GGAVATT program in Veracruz. Evaluability Assessment During the summer of 2005 an evaluability assessment (Rossi, Lipsey, & Freeman, 2004) of the GGAVATT program in Veracruz was conducted by the principal investigator. This procedure made it possible to establish that all stakehold ers agreed on the need to explore the
53 possible social effects of the program. They had realized that policymakers in the state were requir ing access to this information to make better informed decisions concerning the funding of social programs There al so was anecdotal evidence supporting the notion of conducting an inquiry on that topic. St akeholders also expressed their intentions to use the results and recommendations of an evaluat ion for improving the program. The logic models of the GGAVATT progra m (i.e. impact and process theory) developed as part of the evaluability assessment method were useful to explore how the programs logic fit with different social theories. Figure 2 2 shows a simplified impact model for the GGAVATT program where it is possible to observe how program activities lead to direct, intended, and anticipated productive outcomes. The process theory model of the GGAVATT program is presented in Figure 2 3 showing the diff erent activities that occur within the structure of the program. Finally, Figure 2 4 is a modified version of an impact model for the GGAVATT. It shows how the different types of social interaction fostered by program activi ties might be associated with positive social impacts. After a careful review of different social theories, social capital theory seems to be a good fit to explore these possible associations between program processes and social impacts. It is important to be aware that logic models are just representations of educated guesses about how the world works in the context of a particular situation and these representations are based on a certain number of assumptions and may be affected by factors that are out of the control of the program. Conceptualization of Social Capital Theory in the GGAVATT Figure 2 5 shows a conceptual model in which the GGAVATT group is represented as a social network, or structure, in which different pr ocesses take place through the interaction of the levels of trust, reciprocity, and activity of the participants. These processes would lead to accessing different forms of social capital (maybe more than one at the same time) which will in
54 turn translate into different types of benefits for the group and maybe even spill -over to the extended community. Some of these benefits are the ones present in the GGAVATTs logic model (see Figure 2 2) that show how active participation in the program (investment in socialization) leads to an increase in knowledge (development of human capital) which leads to a change in behavior (social and human capitals acting together) that will result into obtaining economic benefits (economic capital). However, these processes occurring inside the network are affected by attributes of the group participants and of the group. Examples of the former include: gender, age, education, years in the program, additional access to social capital (previous and current), etc. Among the latter, is needed to consider attributes of the advisor (e.g. gender, age, education, experience, etc.), region of the state, years of existence, homogeneity, etc. The inclusion of demographic attributes of advisors and group participants (e.g. gender, a ge, education occupation ) as predictor variables in the statistical analyses is necessary because it is expected that differences in these characteristics affect how individuals socialize (Grootaert, Narayan, Nyhan Jones, & Woolcock, 2004; Spellerberg, 2001) For example, younger GGAVATT participants are likely to be active in different social networks than older participants and, ther efore, to have different access to social capital. The presence of family members or very close friends of participants or advisors in the groups as well as the marital status is likely to affect not only the networks to which they have access but also their position within these networks and the density and quality of bonds connecting them with others (Edwards, 2004; Grootaert, Narayan, Nyhan Jones, & Woolcock, 2004; Spellerberg, 2001; Stone, 2001) ; all these factors together may contribute to differences in access to social capital.
55 It is also expected that the number of years that a participant has been in the program is positively associated with greater levels of social capital because of the time required to develop trust and reciprocity. Region of the state may have an effect on social capital levels of GGAVATT participants due both to sociocultural differences across these regi ons and variation in the number and type of local or regional networks available to them (Edwards, 2004; Spellerberg, 2001; Stone, 2001) The level of tolerance within groups and communities can also affect the levels of social capital (Edwards, 2004; Grootaert, Narayan, Nyhan Jones, & Woolcock, 2004; Spellerberg, 2001) In other words, if people within the groups and the communities have positive attitudes towards ea ch other and accept diversity in a respectful way, the possibilities of creating conflicts and violence are reduced and the opportunities for fostering a productive socialization are enhanced. Finally, the level of h omogeneity of the group (e.g. how simil ar participants are in terms of religion, age, education, SES, etc.) is likely to be positively associated with bonding social capital and negatively associated with both bridging and linking social capital (Grootaert, Narayan, Nyhan Jones, & Woolcock, 2004) Some of the program processes in which GGAVATT participants take part are not constrained to socialize among group members and advisors. Group participants in the GGAVATT program interact, among others, with participants of other GGAVATT groups, with non -participant farmers, with other members of their communities, and with diverse institutions. In other words, membership in the GGAVATT program allows these participants to bond, bridge, and link with others and, therefore, different types of trust, reciprocity, and activity are required for the different processes. For example, activities taking place within the boundaries of the group will involve the trust that is present among group members and, for most of the cases, the specific reciprocity that exists in the group. However, when participants are engaged in
56 bridging or linking activities re ciprocity would tend to be less specific and more generalized, or universal, while trust will be specific to the context and based on previous experiences. A visual representation of this idea is presented in Figure 2 6 Not e that in this figure bridging and linking social capital have the same elements with the exception of trust. Having the same activities at both levels responds to the fact that the activities fostered by the GGAVATT that are likely to lead participants t o accessing bridging social capital are the same that may lead to linking social capital. Summary The theoretical and conceptual frameworks of this study have been described in Chapter 2. The important role that farmers organizations can play in curren t agricultural knowledge and information systems was described and then the focus was placed on the GGAVATT program and how these concepts apply in this setting.
57 Participant farm Participant farm Participant farm Participant farm Participant farm Participant farm Participant farm Participant farm Technology Validation Module Advisor Research institution Participant farm Participant farm Participant farm Participant farm Participant farm Participant farm Participant farm Participant farm Participant farm Participant farm Participant farm Participant farm Participant farm Participant farm Participant farm Participant farm Technology Validation Module Advisor Research institution Figure 2 1 Structure of the GGAVATT m odel (SAGARPA & INIFAP, 2002) (Back to text )
58 Figure 2 2 Impact model for the GGAVATT pr ogram. (Back to text )
59 Figure 2 3. Process theory model of the GGAVATT pro gram (Back to text )
60 Figure 2 4. Theory-based modified impact model for the GGAVATT p rogram. (Back to text )
61 Figure 2 5. Conceptual model for social capital in the GGAVATT. (Back to text )
62 Figure 2 6. The three forms of social capital in the GGAVATT progr am. (Back to text )
63 CHAPTER 3 METHODOLOGY General Perspective The evaluation research reported here possesses three main methodological characteristics. First, the evaluation design was co -constructed by the evaluator and stakeholders in direct response to the needs of the program through negotiation and collaboration. This is what Guba and Lincoln (1986) identified as responsive evaluation, or fourth generation evaluation. Second, the focus of this evaluation moves beyond the explicit intended economic and technical goals of the program and looks into its unanticipated social consequences becoming a form of goal -free evaluation (Worthen, 1990) Last, but not least, an in depth review of the programs theory and processes was conducted to identify plausible paths through which the program might be affecting the lives of its benefic iaries. This procedure was used to construct the alternative theory -driven impact model (Chen & Rossi, 1980) used to explore the social effects that were suspected to exist either from theoretical or anecdotal evidence, or both. The compl exity of the phenomenon under study called also for a combination of methods. The advantages of using mixed -methods for the evaluation of social program have been widely discussed in the literature. For example, Greene et al. (2001, p. 27) state that the fundamental uncertainty of scientific knowledge especially about complex, multiply determined, dynamic social phenomena can be better addressed through the multiple perspectives of diverse methods than through the limited lens of just one. In a similar way, Lucke et al. (2001) propose that quantitative approaches alone may lack the required sensitivity to capture a faithful picture of the effects of a program in a community. Another advantage of combining methods in an evaluation study, such as this, is that the results can be presented in multiple forms (e.g. reports, factsheets, success stories, etc.) which make them valuable for more st akeholders; this is also likely to
64 increase their likelihood of utilization. For these reasons, this research followed a parallel mixed method, also known as concurrent mixed method (Tashakkori & Teddlie, 2003) This method is charact erized for having two relatively independent segments, one addressing qualitative questions and the other quantitative questions. As Figure 3 1 shows, each segment uses its own appropriate data collection and analysis methods From the results of each segment, it will be possible to draw conclusions that will be then pulled together to form meta inferences at the end. It is needed at this point to address the philosophical distinctions between the quantitative and the quali tative segments of the study. The former assumes that a truth exists, this truth is embedded in the object of study, and it is suitable of being determined through the application of the scientific method. In other words, the quantitative segment has its ontological roots in realism, identifies epistemologically with objectivism, and adheres to the theoretical perspective of positivism (Crotty, 1998) The latter, the qualitative segment of this study, is founded on social constructionism. This theoretical perspective sustains the concept of th e social construction of reality (Crotty, 1998) Ontologically, social constructionism identifies with idealism due to its acceptance of the existence of numerous concurrent truths, or as Hatch (2002) states, multiple realities are co nstructed in the constructivist research paradigm. In social constructionism, the knowledge is constructed through negotiated transformative dialogue (Gergen & Gergen, 2003) and this knowledge is not considered to faithfully represent a g iven external reality but to be a situated representation of what has been socially agreed to be real. Within this context, meaning is co -constructed by researchers and participants (Crotty, 1998; Hatch, 2002) It is very important that each segment of the study remains faithful to its corresponding theoretical perspective because that contributes to the internal consistency to the research. Each theoretical perspective (i.e. positivism and social constructionism) would yield
65 different types of information, such as statistics and narra tives, which will help to create a more comprehensive profile of the social dynamics of GGAVATT participants. The results from any given segment should not be seen as a confirmation or contradiction of the other. As it has already been established, the ph ilosophical assumptions about truth and knowledge on which they are founded are diverse and, therefore, represent different dimensions of the same phenomenon that may or may not share similarities. Therefore, the results from each segment must be seen as complementary because they are reflecting different perspectives of reality. This study was approved by the Institutional Review Board of the University of Florida on February 21, 2007 (See Appendix A ). Context of the St udy Mexico borders to the north with the United States of America, to the east with the Gulf of Mexico and the Caribbean Sea, to the south with Belize and Guatemala, and to the west with the North Pacific Ocean. Mexico is a transition zone between North A merica and Central America and, therefore, a country of contrasting realities. On the one hand, in 2002 half of Mexicos population was living in poverty and one fifth in extreme poverty (World Bank, 2009) ; 75% of those in extreme poverty residing in rural areas. On the other hand, it is not uncommon to fin d the names of Mexican business men listed among the richest people in the world. The federal r epublic of Mexico is organized into 31 states and one federal district. The Mexican state of Veracruz is located in the east central part of the country. Veracruz shares borders with the states of Tamaulipas to the north, Puebla, Hidalgo, and San Luis Pot osi to the west, Oaxaca, Chiapas, and Tabasco to the south, and the Gulf of Mexico to the east. With its region divided into 210 municipios, Veracruz is the third most populous state in the country, with almost 7 million inhabitants (Gobierno del Estado de Veracruz, 2006)
66 With its characteristic geographical diversity and highly fertile soils Veracr uz is an ideal place for agricultural activities. This is particularly relevant in a country like Mexico where 15.1% of the labor force works in agriculture (C.I.A., 2009) In the case of Veracruz, 31.7% of the economically active population (i.e. 2.7 million) work in an agricultural activity (INEGI, 2004) Veracruz is considered the Americas gateway for bovine and equine production; the Spaniards introduced both species through Veracruz. There are 3.7 million hectares currently dedicated for livestock production in the Verac ruz. The state is the leading producer of cattle and is one of the most important producers of poultry and honey (Gobierno del Estado de Veracruz, 2006) Forty percent of Veracruzans live in rural communities with less than 2,500 inhabitants. Migration, predominantly of males to the United States, has affected the population growth of the state. Po pulation The study was conducted between February and May of 2007 with the 83 GGAVATT group s funded by the Veracruzan Secretariat of Agricultural and Rural Development, Forestry and Fisheries ( SEDARPA) in Veracruz at that time; these groups had 980 partici pants. These are not the only GGAVATT groups in Veracruz; in fact, there might be about 180 groups working across the state. However, it was decided to work with the 83 groups currently funded by SEDARPA because of the accessibility to collect data from them. SEDARPA funds each group (using a mixture of federal and state funds) for a total of five years. The participant farmers are also responsible for paying part of the advisors salary; the farmers share increases proportionally during the five years of funding through SEDARPA. At the end of this period, the groups can continue with the farmers paying 100% of the advisors salary; some groups have continued and some have disbanded. All of the groups in the state (~180) have been at some point in tim e funded by SEDARPA. The GGAVATT is a partial coverage program which means
67 that only a proportion of all the farmers in the state of Veracruz are participating in it. This proportion is very low if considering that the state has approximately 7 million i nhabitants of which a little more than 2 million are involved in agricultural activities (Gobierno del Estado de Veracruz, 2006) The typical group had been working for two years and had an average of 11.8 participants. Most of these groups (69%) focused on the production of dual -purpose cattle (i.e. bovines dedicated to the production of both milk and meat). The other 31% of the groups worked with diverse livestock including: dairy, beef, swine, ovine, and even bees. The average group advisor in Veracruz had three years of experience and almost 38 years of age. As the results in Table 3 1 show, almost 90% of advisors were males and 70% of them had another job to complement their incomes. A large number (65%) were veterinarians and for most of them (61%) this was their first experience as group advisors. Quantitative and qualitative data were collected from group advisors and participants receiving economic support from SEDARPA. The research plan was presented to the states Secretary of Agriculture in February 27, 2007 in the facilities of SEDARPA in Xalapa, Veracru z. As a result of this meeting, SEDARPA granted access to the groups and logistic support for the data collection activities of the study. Quantitative Segment The quantitative segment of the study followed a cross -sectional design to explore the associat ion of program participation and access to social capital. A post then pre test model (Colosi & Dunifon, 2006) was used to introduce reflexive controls in the survey instruments to explore if group participants perceived that their access to social capital has changed since they started participating in the GGAVATT program. For the quantitative segment of the study, individual participants are considered the unit of analysis.
68 Census It was decided to conduct a census of the total population (i.e. 980 participants in 83 groups) as part of the design of this study. However, data was collected from the 83 advisors in the state and only from 406 group participants in 38 GGAVATTs. There are three different possible causes for not receiving resp onses from the rest of the groups, some of these are: the group refused to answer the survey, the advisor did not collect the data during the monthly meetings, or the groups questionnaire package was not returned to the investigator. It was not possible to obtain further clarification on this matter. The typical group that participated in the study has been operating for 2.4 years and had 12 participants, however, not all the participants in these groups (n=456) responded to the survey. This may be the result of some participants that missed to attend their monthly meetings when data were collected. The 406 participants that responded the survey represent 89% of the total members of these 38 GGAVATT groups. A large proportion of these groups (47%) belong to the central region while 29% were from the north and 24% from the south. Seventy four percent of these groups focused on dual -purpose cattle. Advisors from these 38 groups had a n average of 35 years of age and had been working for 2.6 years in the program. Table 3 2 presents some demographic attributes of the 38 GGAVATT advisors. Do Respondents Differ from Non -r espondents? Although the total number of respondents was greater than the originally required sample size fo r this population, it was required to determine if those who did not answer the survey were any different than those who did. This is a common procedure in survey research to address the issue of nonresponse error (Dill man, 2000) Israel (1992b) propose d several strategies to deal with nonresponse error and argu ed that a poor response rate to a survey or questionnaire can make the results virtually useless. The data collected from the 83 group advisors was very useful
69 to compare respondents against nonrespondents. Table 3 3 summar izes the Chi -square values and their associated p -values resulted from the comparison of categorical variables of respondents and non respondents. A significant association was found to exist between having a first time advisor and a nswering or not answering the survey. The region of the state was also significantly associated with the likelihood of being a respondent or nonrespondent. In other words, groups from the Central region of the state were less likely to respond to the survey and groups with first time advisors were more likely to respond to the survey. For this reason, it was decided to control for both of these variables during quantitative data analysis process. Non significant differences were found to exist between r espondents and nonrespondent when comparing the means of continuous variables (e.g. technological adoption, productivity, collaboration, economic performance, and communication with advisor) ; these indicators were reported by the group advisors. The comm unities of both groups were not different either in terms of the availability of local organizations such as: credit unions, farmers unions, NGOs, church, etc.). In other words, respondents and nonrespondents had access to similar formal networks in the ir communities Respondents had been in the GGAVATT for more years (p=.01) than nonrespondents (2.43 vs. 1.67 for respondents and non respondents, respectively). It was observed that more males tended to belong in the group of respondents and more female s tended to be non respondents (p=.0 8 ). Instrumentation Two instruments were developed, one for group advisors and the other for GAVATT participants. The design of both questionnaires followed suggested formats of the Tailored Design Method (Dillman, 2000) Some of the items in these instruments were developed by the principal investigator while others were adapted or, in very few cases, used verbatim from other
70 researchers (Australian Bureau of Statistics, 2004; Grootaert, Narayan, Nyhan Jones, & Woolcock, 2004; Onyx & Bullen, 2000) The first questionnaire (see Appendix B) is a 2 -page survey instrument consisting of three sections. The first section (i.e. question Z1 to Z19) asks group advisors to rate the performance of their groups on different dimensions relevant for the program (e.g. technology adoption, productivity, economic performance, etc.). In the second section (i.e. questions Z20 to Z27), advisors provide information about the presence of forma l networks in the communities where their groups exist. The final section (i.e. questions Z28 to Z45) ask s about the groups (e.g. size, type of production, etc.) and the demographic attributes of the advisor The items included in the second section were adapted from other sources (Grootaert, Narayan, Nyhan Jones, & Woolcock, 2004; Van Schaik, 2002) while the rest of the items were developed by the principal investigator. The second survey instrument, consisting of five pages, includes questions on the following: elements of bonding social capital, elements of bridging social capital, elements of linking social capital, reflexive controls, and demographic attributes of the group participants organized into eight sections (see Appendix C ). The first section (i.e. questions A1 to A13) focuses on general, interpersonal, and institutional trust, exploring them at two main levels intra group and extra -group. The second section (questions A14 to A22) contains a series o f questions that measure the homogeneity and equality within the groups. The following section is designed to serve as a reflexive control to gauge changes in trust. The fourth section in the survey (questions A23 to A30) explores the members participat ion level in activities in other formal networks, the relative importance of the GGAVATT in the lives of these participants, and it also includes a reflexive control. In the following section (B1 to B23), participants respond questions
71 related with their participation in program activities and how much importance do they place on these activities. The sixth section (questions B24 to B31) measures the number of institutional contacts that participants currently have and the ones that they had before the GG AVATT. The next section (questions C1 to C34) contains questions related with reciprocity and other norms, democracy, collaboration, solidarity, and tolerance. The eight section (questions D1 to D14) explores how satisfied or dissatisfied the participant s are with different elements of the program. The final section asks group members to provide demographical information (questions E1 to E11) All of the sections were mostly designed with items adapted from multiple sources (Grootaert, Narayan, Nyhan Jones, & Woolcock, 2004; Narayan & Cassidy, 2001; Van Schaik, 2002) ; item A1 was used verbatim from other studies because it is a question commonly used in studies exploring generalized trust. Pro cedures Used This section describes activities that were conducted to collect the data and to prepare the collected data for analysis. Recruitment of participants The 83 group advisors were invited to participate in the study during six regional meetings. These meetings took place between March 1, 2007 and March 12, 2007 in the following locations: Veracruz, Tierra Blanca, San Andres, Xalapa, Coatzacoalcos, and Tuxpan. The principal investigator explained the purpose of the study and read aloud the Spanis h version of the informed consent form (see Appendix D ) to the advisors. Group participants were recruited during the monthly meetings of their groups. The group advisors had a key role in the recruitment of participants f or the study and the survey administration. During the groups monthly meeting the advisor told the participants about the study, invited them to participate, and read aloud the informed consent for participants (see
72 Appendix E ). The advisors proceeded to administer the survey in those groups where GGAVATT members decided to participate in the study. Cognitive evaluation of the group participant s instrument A cognitive evaluation (American Statistical Association, 1997; Willis, 1999) of the survey for group participants was conducted during the six regional meetings with group advisors In these assemblies t he principal investigator read aloud the whole instrument and asked a dvisors for their understanding of the meaning of each of the questions. This procedure allowed to pilot test the face validity of the instrument and to familiarize the advisors with an i nstrument that would be their responsibility to administer. Through this method, no significant problems with the wording of the questions in the survey for group participants were identified. Group -administered survey t raining The low level of literacy among rural population in the state (World Bank, 2002b) and the unfeasibility to have just one person (the investigator) collecting data across the state in so many groups made necessary the design of a particular protocol for survey administration. Group advisors were trained t o collect the data from group participants. The training followed a modified protocol combining Dillmans recommendations for group administered surveys (Dillman, 2000) with techniques to use in groups where literacy mi ght be a barrier. The principal investigator was responsible for the development of the protocol and training the advisors. The training consisted in teaching advisors the procedure that they should follow to administer the survey to group participants a nd covered the following: (a) explaining the purpose of the study to participants; (b) informed consent process; (c) distributing the materials; (d) explaining the instructions for each section of the survey; (e) use of specially designed visual support to demonstrate participants the correct way to answer without influencing their responses; (f) reading each question allowing for a reasonable time for answering before
73 proceeding to read the next question; (g) appropriate procedure to collect answered surve ys; (h) thanking participants for their responses and inviting them to contact the principal investigator if more information was needed; and (i) returning the surveys to the principal investigator. Each advisor left the training with a package containing all the material required to administer the survey to the ir group and detailed instructions for its use. Data Collection Process The GGAVATT advisors were surveyed during their regional meetings following Dillmans recommendations for group administered s urveys. Each advisor received two copies of the informed consent and the instrument. A signed copy of the informed consent and the answered instrument were collected from them after they finished answering. The data collection process took approximately 15 minutes in each of the meetings. To collect data from group participants, group advisor s used approximately 40 to 45 minutes during the monthly meetings of their groups. Each GGAVATT member received two copies of the informed consent and the survey. The advisor followed the protocol for survey administration that they learned during their regional meetings. At the end, the members enclosed the completed surveys in an envelope and delivered it to the group advisor. The group advisor s returned the sur veys to the principal investigator during the ir next regional meeting. Data Processing Once the survey instruments were collected, each possible answer was assigned a particular code. Answers to Likert type questions were numbered progressively using a sc ale from 1 to 5, where five represents to the most positive answer. Questions with only two possible answers were coded and where the latter represents either the most positive (i.e. the yes from a Yes/No answer) or the most common answer ( e.g. the male from a Male/Female answer). These codes were later reversed for running the multiple regressions in
74 order to situate the most common combination of categorical variables as the comparison group. The codes were captured in a database creat ed in Microsoft Access. To avoid problems with missing data, values were imputed using the Missing Value Analysis function of the Statistical Package for the Social Sciences (SPSS). Replacing the missing values in a dataset reduces conceptual and computa tional difficulties during the analysis (Schafer & Graham, 2002). The method used for estimating replacement values was Expectation Maximization (EM); each iteration in this method consist of an E step and an M step (Hill, 1997) In the first step, the missing data are substituted with functions of conditional expectations found c onsidering observed values and current parameter estimates. The maximum likelihood estimates of the parameters are calculated in the second step as if the missing values were already replaced. The database was then reduced by consolidating individual var iables into indexes following procedures recommended by Carmines and Zeller (1979) ; all the indexes met criteria for unidimensionality (a single factor was extracted from principle components analysis). Table 3 4 contains the indexes that were created from items measured on five Point Likert type scales, their corresponding questions in the survey and their respec tive factor loadings, eigenvalues, explained variances, and Chronbachs alphas. Additionally, other indexes were created by adding values obtained from different series of Yes/No questions (Yes = 1; No = 0). These indexes are presented in Table 3 5 Three composite variables were created to represent the Bonding, Bridging, and Linking forms of social capital using indexes presented in Tables 3 4 and 3 5., and based on Figure 26 presented in Chapter 2. These multiplicative variables are presented in Table 3 6 The rationale for multiplying the components instead of just adding them is that the mere membership in a given network is not enough to access the available social capital. In other word s, farmers may
75 belong to the GGAVATT but if they are not active and act in congruency with the group expectations they would likely be limited in fully access ing the resources of the network. The advantages of using a multiplicative index instead of an ad ditive one have been established in the literature on human development (Herrero, Martinez, & Villar, 2008) Finally, all of the c ontinuous variables were centered around the mean before incorporating them into a regression model. Validity and Reliability The four types of validity described by Cronbach (1971) were addressed in the study. A panel of experts from the University of Florida revised the instrumentation to determine its content, construct, and face validity. This panel of experts was composed by faculty members of the Departments of Agricultural Education and Communication, Educational Psychology, and Family, Youth and Community Sciences. In addition to that, the cognitive evaluation conducted with group advisors also helped to ensure the face validity of the survey instrument for group participants. Construct validity was further assessed through factor analysis and Cronbachs alpha to determine internal consistency. Data Analysis Recommendations from Isra el (1992a) were followed. After determining th at there were no coding errors present in the dataset, t he first step in the data analysis process was to explore the distribution of responses across dependent and independent variables. For categorical variables the analysis consisted of frequency distr ibution across categories and descriptive statistics were computed for continuous variables. The next step in the analytical process was to investigate bivariate relationships between the three dependent variables (i.e. bonding, bridging, and linking soci al capital) and the independent variables. For categorical independent variables t test statistics (Cohen, 1988) were
76 computed to determine if the means for the different levels were significantly different (p tendency for significance; p also applied to all the other test -statistics in th e study). Correlation (Cohen, 1988) analysis was used for continuous independent variables. Multivariate regression was used to explore the effects of independent variables on the three dependent variables (Tabachnick & Fidell, 1996) This process was taken one step further by running multi level regression analysis for mixed models (Peugh & Enders, 2005) or hierarchical linear models, to explore how the variation explaine d by the multiple regression models was distributed between two levels (i.e. at the participant level and at the group level). The multilevel analysis was an appropriate methodology for this study because the participants were naturally nested within grou ps and this type of analysis takes into consideration this characteristic to estimate the effect of explanatory variables at the participants and groups levels. The criterion followed for the incorporation of new variables into the regression model was that for each degree of freedom that was introduced into the model the change on the 2 Log Likelihood value should be significant at the .05 level of alpha in a Chi -square distribution. Collinearity analyses (Belsley, 1991) were conducted for each model and variables were eliminated when strong collinearity was present. Qualitative Segment Social constructionism is an appropriate theoretical perspective to study how GGAVATT participants conceptualize change in their lives. This change can be understood in terms of how the shared meanings that they have created as a group provide the basis for acti ons where they access the different forms of social capital (Gergen, 2001; Gergen & Gergen, 2003) In this process, is important to appreciate the central role of dialogue as the link betwee n the past, present, and future of the participants (Gergen, 1999; Gergen & Gergen, 2003) because it is
77 through this dialogue that the constant construction of realities takes place by the action of words. For Gergen and Gergen (2003, p. 2) the pivotal assumption around which the constructionist dialogues revolve is that what we take to be knowledge of the world and self finds its origins in communal interchange. For social constructionism, it is important t o interpret how people come to explain their worlds. The GGAVATT is an extension model which has characteristics that are likely to increase the opportunities for significant interaction to occur among its members. The nature of the contact promoted by the program activities makes this an excellent setting to foster transformative dialogue among its participants at the same time that empowerment of farmers is achieved through democratic participatory practices. In a social constructivist study, it is no t relevant if the participant is telling the truth or not what matters is why this participant is expressing that idea. In other words, what is really relevant is why this person is telling that particular story and what does that mean? For the qualitative segment, narrative gathered from interview groups was considered to represent the shared voice of program participants and, therefore, th e whole set of individuals that participated in th ose interviews r epresent our collective unit of analysis. Partici pants A total of 17 GGAVATT members were selected to participate in this study from the groups coordinated by the States Secretary of Agriculture. The participant selection was done using criterion sampling (Patton, 2002) and the criteri a to participate included: being an active GGAVATT member, be considered a leader by the members of the group, and being democratically nominated by the majority in the group. All selected participants were males with ages ranging between their early30s to their early 70s. They all have farming as their main economic activity. Most of the participants (11) came from groups in the central region of the
78 state while the north and the south were represented by three participants each one. There was a huge variation across literacy levels of these participants; one of them was illiterate and a couple had professional degrees. They represent a sub -set of the participants described in Table 4 1 of Chapter 4. Table 3 7 presents th e interview groups and the pseudonyms of these participants. Instrumentation The data collection method used for this study was the interview because it naturally yields socially constructed knowledge. The interviews were conducted in Spanish by the res earcher who is a native Spanish speaker. The subsequent preparation and analysis of the data was also performed in Spanish to preserve the original meaning. However, pieces of data were translated to English during each step of the process to document the credibility and confirmability of results through peer review. The social constructionist interview presents specific challenges for the researcher, and shapes completely the final result of the study (Crotty, 1998) The instruments were designed to allow researcher and participants to co cons truct the knowledge together (Gubrium & Koro Ljungberg, 2005; H olstein & Gubrium, 2003) and to allow flexibility in the dialogue (Berg, 2004). Interview schedules that included the main conversational topics of interest for the study and allowed for flexibility in the dialogue were used and researcher and interviewe es negotiated the construction of knowledge around them. As in most interviews, the presence of the researcher shaped the data through the interaction with participants. However, in this case that was the goal because researcher and participants were co -constructors of knowledge in the constructivist paradigm (Crotty, 1998; Hatch, 2002) This social constructionist endeavor was constant across the different interview groups and created the foundations for a common story to emerge. From his active participation and repeated readings o f the transcriptions, the researcher brought with him the experiences and
79 thoughts from the first interview group to the next one and the next one successively; and by doing this helped to connect the voices of all participants. Recruitment of Participan ts During the monthly group meetings, advisors told group participants that the study required volunteers for social interviews. After explaining the purpose of the study, the advisor read aloud the informed consent (see Appe ndix F ) and provided the criteria for selecting a participant for the interviews (i.e. being an active GGAVATT member, be considered a leader by the members of the group, and being democratically nominated by the majority in the group ). The group partici pants that were interested in participating and met the eligibility criteria were later contacted by the principal investigator and received information about the time and place for their interviews. Data Collection Process There are two basic types of int erviews namely individual, or one on-one, interviews and group interviews and they mainly differ on the number of active participants. An active participant is that which is actively engaged in the social construction of knowledge during the interview process (Holstein & Gubrium, 2003) Individual interviews are characterized by having just one active participant, or intervi ewee, and one interviewer. The level of involvement of the interviewer would depend on the epistemological orientation of the study (Holstein & Gubrium, 2003) ; this applies for all types of interviews. The group interview is characterized for the presence of two or more active interviewees. Probably the best known group interview approach is the focus group. Focus groups a re dialogues among a group of people on a specific topic that ideally have between six and eight participants, one moderator, and one observer (Israel & Galindo Gonzalez, 2008) When the number of inter viewees is too small to be considered a focus group (four or less) then it can be called a social interview. Another
80 difference between focus groups and social interviews is that there is no need to have an observer; however, this might be changed if answ ering the research questions requires access to that type of data. The social interview was selected as the qualitative data collection strategy for this study. During these interviews, participants were asked to talk about what a good GGAVATT is and how would they describe a successful GGAVATT group. Five interview groups were formed with the 17 participants; nine participants were grouped in three interview groups of three participants each, while the remaining eight participants were divided in two int erview groups of four participants each. Each interview group decided to participate in only one interview session. The three geographical regions of the state of Veracruz were represented by the five interview groups; three represented the central regio n of Veracruz, while the remaining two interview groups represented the northern and southern regions of the state, respectively. The dialogue topics to be included in the interviews, as well as the number of interview sessions to be held, were negotiate d with the participants before starting the interview. Once the dialogue topics were decided, the format proposed by Gergen (1999) for a transformative dialogue was followed during the actual interviews. This technique allowed for activat ing the particip ants and achiev ing a richer co -construction of meanings. The advantages of using interviews for data collection were many. The data collected this way were a firsthand account of the realities of participants and the format allowed for clarification and expansion of the participants answers. From a social constructionist perspective, the dialogue promoted a negotiation of power that activated participants. The main limitation of the interviews is that they were time consuming, particularly during transcription of the data.
81 The audio from every interview session was recorded using two digital audio recorders ; the rationale behind having two recorders was to reduce the risk of missing part of the interview in case of malfunctioning of one of them Int erviewees were informed of these recordings at the beginning of the interview session, through an informed consent. According to McMillan and Schumacher (2006), mechanical recording of data is an appropriate strategy to enhance the validity of the study. The role of the interviewer was that of an active participant in the dialogue. All of the participants appeared to be comfortable sharing their thoughts during the sessions and enjoyed learning from the experiences in other groups. When it was noticed t hat one participant was dominating the conversation, the researcher directed probes to other interviewees to activate them and bring their perspectives to the dialogue. Each interview session included a 15 minute break and it was very interesting to obser ve how during this time participants continued talking about the topics of the session and some of them exchanged information to plan a future visit with their respective groups. Data Organization This section describes several methodological steps taken after collecting the data and that were required to prepare it for analysis. Verbatim transcriptions were generated from the digital audio files. The identity of the participants was kept confidential through the use of pseudonyms for the transcriptions. These transcriptions were done by the researcher following appropriate guidelines (Bogdan & Biklen, 1998; Gee, 2005) At least five hours of work were required to transcribe each hour of interview. The transcripts from each interview were open coded (S trauss, 1987; Strauss & Corbin, 1998) The coding was done sentence by sentence and in -vivo codes were used as much as possible. The coding made it possible to identify five categories or motifs (Gee, 2005) that were present in the data It was observe d that the same categories were present in all interview
82 group s. These categories were: characteristics of a good GGAVATT, social transformation, technological transformation; identity of GGAVATT participants, and interaction with non participants. This characteristic of the data together with the already described connecting role of the researcher in the co -construction of knowledge across interview groups were the basis to make the decision of merging all stories into one collective story. This collect ive story represents the voices of all the interviewees. The next required step was to reduce the amount of data. A social capital filter was used as a data reduction technique. Through multiple readings of the data it was possible to identify the segm ents of text that were about elements or proximal indicators of social capital and use them to create a new, and reduced, dataset. Only those segments of data related with social capital concepts (e.g. networks, trust, collaboration, etc.) were kept in th e new dataset. The selected segments were organized into three categories according to which one of the three main forms of social capital (i.e. bonding, bridging, and linking) they were more related to. According to Hatch (2002) narratives are one of the expected products of constructivist research paradigms. It was decided to organize the segments into the three forms of social capital because, as was described in Chapter 2, the program fo sters intra and extra group processes that are likely to be associated with access to these three forms of social capital and the data showed evidence of that (See Figures 2 5 and 2 6). For example, the segments included in the collective story on bonding social capital talk about collaboration inside the group and the importance of trust and shared norms for success of the group. The segments included in the collective stories on bridging and linking social capital were those that exemplify the interact ions of group participants with their extended communities and institutions, respectively.
83 Credibility of the Study The most important indicator of validity in qualitative research is probably the presence of the internal consistency that is achieved whe n there is a clear and logical connection between the purpose of the study, the research questions, the philosophical orientation, the theoretical perspective, the methods used to carry out the study, and the representation of results (Gubrium & Koro Ljungberg, 2005) Besides this general consideration, there are many techniques that can be used to address specific issues related with the methodological rigor of the study. In this study, t o assess the credibility of the findings, some of the participants were contacted again and asked to review the transcriptions or the initial results of the analyses to determine if these reflect the reality of what was disc ussed during the sessions. This is a common practice known as member checking (McMillan & Schumacher, 2006) Credibility addresses the issue of fit between respondents views and the repres entation made by the researcher (Lincoln & Guba, 1985) To determine the dependability of the study (Lincoln & Guba, 1985), the whole research process was systematically documented using an audit trail. Peer review was also used throughout the process because this is likely to enhance both the credibility and the confirmability of the findings (Lincoln & Guba, 1985; McMillan & Schumacher, 2006.) The researchers subjectivity statem ent (See Appendix G ) is also an indicator of validity. Data Analysis The content, structure, and narrative elements of each story were analyzed. The analysis of the content consisted of identifying the main topics of the story and elaborating a summary of it. The structure of the stories was explored and framed following an example provided by Gee (2005; p.163). Studying the structure this way allowed for identifying how the stories were created or co -constructed. The data c ollected through social interviews were composed of
84 multiple narratives or stories and these narratives are parts of the realities that were socially constructed during the interview process. Through these stories, the participants constantly created and recreated their worlds, and themselves, in an ever -changing form that responded to contextual variations and allowed for the simultaneous existence of multiple dimensions of reality (Sands, 2004) Narrative analysis is an analytical method emerged from the literary tradition. The interpretive nature of narrative analysis allows for a holistic appraisal of peoples lives (Daiute & Lightfoo t, 2004) For a social constructionist study, narrative analysis is the ideal choice because narratives are at the core of the social construction of reality and the making sense of the world (Gergen, 1994; Gergen & Gerge n, 2003) The data collected through social interviews is composed of multiple narratives or stories. These narratives are parts of the reality that was socially constructed during the interview process. Through these stories participants constantly c reate and re -create their worlds and themselves (Sands, 2004) T he three stories were analyzed using Labovian narrative analysis (Labov, 1972) Labovian analysis is characterized for deconstructing the narrative into six interrelated components (Murray, 2003) These parts are: (a) the abstract, which presents the gist of the narr ative; (b) the orientation, which provides the contextual framework of the narrative, in terms of the who, where, when, and how; (c) the complicating action that describes the central event of the narrative; (d) the evaluation of the action which c an take three different approaches, external evaluation, embedded evaluation, and evaluative action; (e) the result that is a relation of what finally happened; and (f) the coda, which may be in the form of a reflection and a summary of the events. This m ethod made it possible to free the narrative from any material additional to the central story. In other words, it allowed the core narrative to emerge (Murray, 2003). In
85 these core narratives it was possible to situate the transformations that particip ants have undergone as members of a GGAVATT. Narrative analysis is a rigorous and auditable technique that allows for structural comparisons and provides a valuable insight from the participants perspective. Narrative analysis generates narratives. As t he analysis progressed, these stories unfolded revealing new layers of meaning. Summary of Methodology This responsive evaluation (Guba & Lincoln, 1986) was planned within the overarching notion that social phenomena are highly complex and should be explored using a combination of methodologies. The study follows a goal -free, theory -driven evaluation approach (Chen & Rossi, 1980) to explore access to different forms and type s of social capital as an unintended, unanticipated, direct consequence of the participation in the GGAVATT program (Rogers, 2003). The connections between program participation and social capital access are investigated with quantitative and qualitative methodologies arranged in a parallel or concurrent mixedmethod design (Tashakkori & Teddlie, 2003).
86 Figure 3 1. Schematic representation of the concurrent mixed method used in this study (Back to text )
87 Table 3 1. Demographics of group advisors in Veracruz, Mexico Frequency (% ; n=83 ) Advisors gender 13 Female 87 Male Advisor lives in community 71 No 29 Yes Advisor has a second job 30 No 70 Yes First time advisor 39 No 61 Ye s Advisors Education 35 Other a 65 Veterinarian a Includes: a gronomists, b iologists, and s pecialists, among others (Back to text )
88 Table 3 2. Demographics of group advisors of groups responding the survey Frequenc y (% ; n=38 ) Advisors gender 10 Female 90 Male Advisor lives in community 63 No 37 Yes Advisor has a second job 32 No 68 Yes First time advisor 26 No 74 Yes Advisors Education 34 Other a 66 Veterinarian a Includes: agronomists, biologists, and sp ecialists, among others (Back to text )
89 Table 3 3. Chi -square statistics from comparing categorical variables of respondents and non respondents Chi square p value Survey item Gender of advisor .45 .50 Z39 Adviso r lives in community 2.14 .14 Z40 Advisor has a second job .07 .79 Z41 First time advisor 4.43 .03** Z42 Advisor has family in group .15 .70 Z43 Compadres in group .99 .32 Z44 Education of Advisor .02 .89 Z45 Type of production 7.48 .19 Z29 R egion of the state 5.68 .06* Z30 *Tendency to significance ** Significant at the .05 level of alpha (Back to text )
90 Table 3 4. I ndexes c reated for d ata r eduction using factor analysis Index Components Survey Item F actor loading Eigen value Explained variance (%) Cronbachs alpha Bonding social capital Groups trust Group mates Officials Advisor Institutional contact A8 A12 A10 A11 .72 .80 .81 .81 2.48 62.00 .77 Groups reciprocity Reciprocity 1 Reciprocity 2 Reciprocity 3 C1 C2 C19 .96 .91 .87 2.49 83.09 .89 Bridging social capital Interpersonal trust Neighbors Non participants Other groups A6 A13 A9 .89 .84 .94 2.38 79.47 .87 Linking social capital Institutional trust Police Political parties Army A3 A4 A5 .86 .85 .82 2.14 71.28 .80 Bridging and Linking social capital Generalized reciprocity Reciprocity 4 Reciprocity 5 Reciprocity 6 Reciprocity 7 C3 C4 C8 C34 .81 .75 .72 .71 2.24 55.90 .74 Other Group tolerance Tolerance 1 Tole rance 2 Tolerance 3 C21 C22 C23 .84 .91 .85 2.26 75.24 .83 Community tolerance Tolerance 5 Tolerance 6 Tolerance 7 C25 C26 C27 .89 .92 .81 2.31 76.97 .85 Importance of intra group activities Monthly meeting Annual evaluation Elections Planning B16b B1 7b B20b B23b .70 .79 .66 .77 2.76 55.24 .79 Importance of extra group activities Regional meeting National meeting Group visiting B18b B19b B21b .81 .90 .75 2.03 67.78 .76 Institutional trust change Local government State government Research institutio n A23 A24 A27 .79 .85 .76 1.93 64.41 .72 (Back to text )
91 Table 3 5. Additive indexes created for data reduction. Survey Index values Index Component item Minimum Maximum Bonding social capital Intra group activi ty Monthly meeting Annual evaluation Elections Planning B16a B17a B20a B23a 0 4 Bridging and linking social capital Extra group activity Regional meeting National meeting Group visiting B18a B19a B21a 0 3 Other Homogeneity Same religion Same ge nder Same SES Same education Same occupation Same residence A14 A15 A16 A17 A18 A22 0 6 Formal networks in community Farmers Union Local farmers association Cooperative Church Political party Credit union Sports group NGO Z20 Z21 Z22 Z23 Z24 Z25 Z26 Z 27 0 8 Previous related institutional contacts Local government State government Research institution Farmers Union B24b B25b B27b B28b 0 4 Current related institutional contacts Local government State government Research institution Farmers Union B2 4a B25a B27a B28a 0 4 Previous unrelated institutional contacts Political party Church NGO Communications media B29b B30b B31b B26b 0 4 Current unrelated institutional contacts Political party Church NGO Communications media B29a B30a B31a B26a 0 4 (Back to text )
92 Table 3 6. Composite m ultiplicative variables Composite variable Components Bonding social capital Group s trust Intra -group activity Group s reciprocity Bridging social capital Interpersonal trust Extra -group activity Generalized reciprocity Linking social capital Institutional trust Extra -group activity Generalized reciprocity (Back to text )
93 Table 3 7. Interview groups and participants pseudonyms. Interview Group Parti cipants Pseudonym s San Juan Len Diego Ramiro Sierra Antonio Rogelio Jess Hctor Jutla Santiago Gerardo Juan Retiro Cristbal Pedro Luis Calatan Pablo Jacinto Roberto Eduard o (Back to text )
94 CHAPTER 4 RESULTS Chapter 3 describes the different methodological procedures that were followed. T his chapter presents the results from the analysis of the data organized according to the specific obje ctives of the study Social Capital of Groups and Participants The material presented in this section responds to the first specific objective of this study that was to provide a description of the social capital of GGAVATT groups and participants. Ninet y four percent of the group participants responding to the survey consider the GGAVATT as one of the three most important organizations that they belong to. Most of the respondents (74%) reside in the community where the group operates and for an even gre ater proportion (76%) the farm is their main source of income. Among the 406 group participants, almost 60% had participated in the program between one and five years while 35% had done so for less than a year. The majority of the group participants reported to have their families and best friends in the GGAVATT group (59% and 63% for family and friends in group, respectively); only 31% of the group advisors had family members in their g roups. Ninety-one percent of respondents were males. The most rele vant demographics are presented in Table 4 1 ; in it is possible to observe that most group advisors and participants are males (90% and 91% for advisors and participants, respectively). A large proportion of advisors are veter inarians (66%) and while the majority of participants live in the communities where their groups operate and have farming as their main economic activity most advisors (63%) live outside these communities.
95 Distribution of Bonding, Bridging, and Linking Social Capital Table 4 2 summarizes the descriptive statistics for the three forms of social capital and the distribution of these variables across regions of the state. The mean value of bonding social capital of group partici pants is 44.95 in a scale from zero to a hundred. Looking at the mean values for different regions in the state is possible to observe that group participants in the south and north seem to have greater levels of bonding social capital than those in the c entral regions. In a scale from zero to seventy -five, t he overall means for bridging and l inking social capital are 18.67 and 15.09, respectively. It is important to notice that participants in the north region seem to have levels above the means for the three different forms of social capital. The means and standard deviations of relevant continuous independent variables are presented in Table 4 3 With the exception of homogeneity and the presence of formal networks in com munities all other variables were measured using a scale from one to five. Group participants generally perceive that group activities are very important; the means for i mportance of group activity and i mportance of groups bridging/linking activity are 4.35 and 4.07, respectively. Even though the descriptive analyses of the data show that GGAVATT participants are a fairly homogeneous group (i.e. 90% are males, 84% considered their socioeconomic status (SES) to be medium, for 75% the farm is their main e conomic activity, and 73% are local residents in a scale from zero to six t he mean value of perception of homogeneity is 3.11. This is very interesting because looking at the descriptive statistics of participants presented at the beginning of this chapter it would seem that the respondents are a fairly homogeneous group. However, if we look for example at SES, which is a complex cognitive construct per se it is possible to notice that even though almost 85% of respondents self -rated their SES as mediu m, only 45% of respondents perceive that most members of their groups have the same SES. This suggests the existence of other socio-cultural variables affecting how participants compare
96 themselves against their peers. O n average, a lmost five formal organizations are available to GGAVATT participants in each of their communities The use of reflexive controls in the instruments made it possible to determine that group participants perceive that the trust that they have on some institutions and individual s has changed since they started their participation in the program. This is particularly true for the trust that they have in the institutions that are directly related with the program. As the content of Table 4 4 shows, 67 % of the participants reported that their trust in these institutions either increased or significantly increased during this period. The same applies to the trust they place in their neighbors to which almost 47% of group participants reported a similar positive change. This finding suggests that the participants involvement in program processes have somehow affected positive perceptions about the trustworthiness of other people in their communities and program stakeholders. In addition to the informa tion presented in Table 4 4, direct mean comparisons allowed to find significant differences (p activity (2.46 vs. 3.46, respectively), previous related institutional contacts and current related institutional contacts (1.73 vs. 2.54, respectively), and previous and current unrelated institu tional contacts (1.27 vs. 1.56, respectively). Relationships B etween Social Capital and Independent Variables Comparing the means of the dependent variables, bonding, bridging, and linking social capital, across the levels of independent categorical vari ables it was possible to determine the presence of significant associations; these comparisons are shown in Table 4 5 There is a highly significant (p of both bridging and linking social capital. Gender of the group participant is also associated with bridging social capital. Men are si gnificantly (p have greater levels of
97 bridging social capital and tend (p This might still be a reflection of traditional views in Mexican culture that fostered a passive, and to a certain degree shy, role of women in society (Paz, 1962; Ramrez, 1977) This would be likely to affect wom ens effectiveness in socializing with others, particularly those who are different or have a position of power. There is a highly significant (p second job and greater levels of both bridging and linking social capital. It is possible that, depending on the nature of their second job, advisors with a second job have more opportunities to act as bridges connecting their groups with other networks in which they are active as a result of their additi onal activities. Greater bonding and bridging social capital levels are associated in a highly significant way (p 0.05) for linking social capital. When the advisor has family in t he group, the levels of both bridging and linking social capital are significantly (p participants have family members in their groups, these participants are very likely (p have greater levels of bridging social capital and tend (p social capital too. Group participants education is positively associated (p bonding social capital. When the advisor is a veterinarian, participants are very likely (p to have greater levels of bridging social capital and tend (p linking social capital. The more years that group participants have been in the program, the more likely (p ater levels of all three forms of social capital. The same is true for the years that the group has existed, with the exception that the association with bonding social capital is only significant at the .05 level of alpha. Group participants between 41 and 50-
98 years old have significantly (p tend (p than do younger participants There is a highly significant (p region of the state and bonding social capital (39.10, 47.06, and 55.33 for Central, North, and South regions, respectively). Group participants from the North tend (p Producers of dual -purpose catt le (DPC) are more likely (p bridging social capital. No significant associations were found between any of the forms of social capital and having an advisor that is a local resident, SES of the group participant, the grou p participant being a local resident, or when farming is the main economic activity for the group participant. The associations between the thr ee forms of social capital and continuous independent variables were also tested; T able 4 6 presents the obtained correlation coefficients and their statistical significance. A positive moderate correlation (Davis, 1971) of practical and statistical significance (p activities and bonding social c apital; statistically significant (p ) low correlations exist between the same variable and the other two forms of social capital. This finding suggests that when participants consider program activities to be relevant they are more likely to be act ive in the network and access the benefits of organization. Group tolerance has a low positive correlation with bonding social capital ( p 1 ). Another positive low correlation exists between trust change in neighbors and bridging social capital ( p 0.01) ; and the same is the case between the presence of formal networks in the communities and bonding social capital ( p 0.01) The importance that participants place on bridging and linking activities is positively correlated with all three forms of s ocial capital. Positive moderate correlations of statistical and
99 practical significance (p number of previous related institutional contacts of group participants and the levels of both bridging and linking social capital; t he same variable has a low correlation with bonding social capital. The number of current related institutional contacts, as well as the number of unrelated previous institutional contacts, is positively correlated with all three forms of social capital. Bridging and linking social capital have positive low correlations with both the advisors age and the number of unrelated current institutional contacts. Previous network activity has a low positive correlation with bonding and bridging social capital ( p ). Multivariate Relationships Table 4 7 presents the multivariate regression model for bonding social capital. The model is highly significant (p -square = .19) of the varia tion in the dependent variable The model shows that the importance that group participants place on group activities is significantly (p same way, as years in the program increase, the level of bond ing social capital increases significantly (p tolerance are also positively associated (p levels of social capital. Group participants from the South region of the state are more likely (p bonding social capital. The multivariate regression model for bridging social capital is presented in Table 4 8 This model is highly significant (p and explains 30% of the variation (Adjusted R -square = .30). Again, the importance that group participants place on the bridging and linking activities of the program has a highly significant (p capita l. The longer the time that group participants have been in the program, the more likely (p
100 of groups existence which has a significant (p ive effect on bridging social capital. The levels of bridging social capital increase in a significant way (p trust more in their neighbors. Group participants from the North and South have significantly (p els of bridging social capital than those in the Central region of the state. The largest effect on bridging social capital is produced by the gender of the advisor. Male advisors are significantly (p cial capital. Even though the effect of group participants gender is not as impressive as that from the advisor, male participants are also positively associated (p First time advisors and advisor s with a veterinary degree are significantly associated with greater levels of bridging social capital (p The numbers of previous and current related institutional contacts also have a posi tive effect on bridging social capital (p unrelated institutional contacts of program participants is also positively associated (p with an increase in bridging social capital levels Table 4 9 presents the multivariate regression model for linking social capital. This is also a highly significant model (p (Adjusted R -square = .22). As was the case for bridging social capital, the importance placed by participants on bridging and linking activities of the GGAVATT program has a highly significant (p years that part icipants have been in the program, as well as the number of previous and current related institutional contacts, the greater the levels of linking social capital (p homogeneity of the group increases, participants are less likely (p have greater levels of linking social capital. Male advisors are again associated with a large and significant (p
101 0.01) increase on the levels of linking social capital. First time advisors and advisors with a second job tend (p ed with greater levels of linking social capital. Multilevel Regression Results The multilevel multivariate regression model for bonding social capital is presented in Table 4 10. The 2 Restricted Log Likelihood value for the interc ept only is 3, 646.26. When we introduce all the variables from the model presented in Table 4 7, this value decreases to 3,567.59 (see explanation in Chapter 3 concerning the criteria for introducing variables in the multiple regression models). The covariance parameters show that in the intercept only model almost 80% of the variation in the data exists at the individual or residual level (415.06 vs. 109.37 for residual and group levels, respectively). These parameters are reduced in the model to 398.74 and 24.71 for residual and group levels, respectively. This result shows that 79% of the variation at the group level is explained by the model; and the remaining variation is not significant anymore. However, the model is capable of explaining only 4% of the variation at the individual level and the residual is still highly significant (p and significance of the components of the model are almost the same as the ones described for Table 4 7 corresponding to t he multivariate regression model. Table 4 11 summarizes the results from the multilevel multivariate regression model for bridging social capital. In this case, the 2 Restricted Log likelihood value for the intercept only model is 3,290.47 and it is reduced to 3,175.35 after the introduction of all the variables from the model shown in Table 4 8. The covariance parameters for the intercept only model indicate that almost 40% of the variation is located at the group level, while the remaining 60% exists at the individual level; these parameters are reduced 55% and 8% in the full model, respectively. Both covariance parameters are still significant in the full model (p residual level, respect ively). The behaviors of some components in this multilevel model differ
102 than the one observed in Table 4 8. For example, the importance placed on bridging and linking activities and the trust in neighbors continue to have a positive effect but now this effect is just significant at the .05 level of alpha. Similarly, the effects of advisors gender and the number of unrelated institutional contacts, whose positive effects ceased to be significant and now represent only a tendency (p same happened to the negative effect of living in the South region of the state. Totally opposite is the case of the number of current institutional contacts that moved from being just a tendency (see Table 4 8) into significance at the .05 level of alpha. Finally, three components that had significant effects are not significant in the multilevel model, these are: living in the North, having a first time advisor, and having a veterinarian for advisor. The third multilevel model is presented in Table 4 12 and it corresponds to the model for linking social capital. The 2 Restricted Log Likelihood value for the intercept only model is 3,100. 22 and when the variables from the multivariate model presented in Table 4 9 are introd uced this value decreases to 3,018.69. The covariance parameters show that in the intercept only model a little more than one third of the variation in the data exists at the group level and the other two thirds are at the individual level (102.05 vs. 56. 88 for residual and group levels, respectively). These parameters are reduced in the model to 94.26 and 34.88 for residual and group levels, respectively. This result shows that while the full model explains almost 40% of the variation at the group level it only explains less than 8% of the variation at the individual, or residual, level; both covariance parameters are still significant in the full model. This multilevel model presents some differences compared to the one in T able 4 -9 These differences are: (1) the importance placed on group bridging and linking activities is not a significant factor for linking social capital anymore; (2) an increase in group homogeneity still has a negative
103 effect on linking social capital however, this effect is now smaller and not as significant as before; and (3) the same is true for the effect of advisors gender. The results from the hierarchical linear models for bonding, bridging, and linking had some similarities and differences. For example, the time that participants have spent in the GGAVATT program was directly associated with greater levels of all forms of social capital. In other words, the larger the numbers of years that group participants invest in socializing in the pr ogram the more likely they would have access to the different forms of social capital. This makes sense because trust and reciprocity require time to develop. Gender of both participants and advisors has an important effect for bridging and linking socia l capital. For instance, groups with female advisors are less likely to have greater levels of bridging and linking social capital than those with male advisors (p female group participants are also associated with lesser levels of bridging social capital. (p 0.05). The number of previous related institutional contacts that a participant has affects positively the levels of all three forms of social capital for that particular participant (p 0.05 for bonding and p contacts has a similar effect for bridging and linking forms of social capital (p 0.01 for bridging and linking, respectively). While participants from the South region have greater (p of bridging social capital. The importance that participants place on the different activities fostered by the GGA VATT program is positively associated with greater levels of bonding (p 0.01) and bridging (p to a particular form of social capital; an increase in the value of this variabl e is associated (p 0.05) with greater levels of bridging social capital. An opposite relationship exists between
104 homogeneity of the group and the levels of linking social capital; those groups that are more homogeneous are less likely (p greater levels of that form of social capital. Social Capital in Participants Lives The second specific objective of this study was to contextualize how social capital takes form in the lives of the participants by looking for evidence of access to soc ial capital, represented through transformations in roles and cooperation, in the stories that GGAVATT participants co -construct around their participation in the program The material presented in this section contributes to that goal. Normal overlapping exist among the different social spheres in which GGAVATT participants interact (e.g. group, community, institutional, etc.) making it very difficult to find stories that are examples of just one form of social capital. Therefore, those three stories sho uld not be taken as the only way to represent the voices of those participants. On the contrary, it is recommended to think of them as one single story, the story of the social capital of GGAVATT participants in Veracruz. Tab le 4 13 has an example of the co -construction present in the collective stories. It can be observed how the stories from various participants belonging to different groups intertwine forming a collective story. Bonding Social Capital of GGAVATT Particip ants The first collective narrative is about bonding social capital and is titled What a good GGAVATT is Th is story is about how participants interact within the group establishing and enforcing rules to meet their goals : Collective story on bonding s ocial capital A good GGAVATT is that where all the members trust each other and are willing to pursue a common goal. There should be respect among the members, they should coexist harmoniously, and they should see each other as family. That is a good GGA VATT.
105 We all collaborate in the GGAVATT and we have obtained many things as group. At the beginning we did not have a grass chopper, much less a tractor, all the work was manual. We got together to buy electric fences. As individual farmers we were unab le to buy an electric fence; we did not even know about them. Therefore, we decided to buy them as a group and raffle off the turns. We were 12 or 13 gavateros. We got together to make a raffle for the turns. Once the order of the turns was established we calculated the amount of money that each one needed to provide to cover the cost. Every 15 days each one of us made that money available to buy a new electric fence. That is how we acquired them. For example, we got an animal feed dealership. It al lows us to pay a smaller price for the fee d than others do. A representative of the fee d company came to our monthly meeting to give a talk. After that, he asked if we were interested in selling his products in this community. All the members voted and decided to become his deal ers. Since then, we buy the fee d that we need for our cows at a special price. The group also sells fee d to other farmers that are not part of the GGAVATT at a higher price. That same way we got grass choppers, medicines, and ma ny other things. We have used that same procedure to build silos; we became organized to visit each weekend one farm and work together. The group makes its own decisions. When someone has an idea we all listen to it. If someone has an idea and shares it with the group, we consider it and, according to the opinion of the majority, it is accepted or rejected. The group is very democratic. We all have a voice. If someone proposes something, we all listen to it and vote. We all participate. In addition, the vote of the president has the same weight t hat any other and it does not matter how many cows a member has. The majority of us, the people integrating the group, we have the tendency of helping just for the sake of helping, as long as it does not cost us. I also feel that, as a person, one is always hoping that if I open myself to help someone perhaps that person or another people will help me also; we always expect something in return unconsciously. When we help someone we always expect to get someth ing in return, but it can be just a smile of gratitude or any other thing. For example, when somebodys cow is having problems at calving, the members of the group will give him a hand disinterestedly, even if that person is not a GGAVATT member. However if we think about lending someone outside the group a tractor for example, then that is very difficult. For that type of situation, like lending expensive machinery outside the group, there are different rules; but for the rest I feel that there are not In the group, we have the mandate of helping each other in return for labor, and we take turns for it. Every Monday we go to a different farm and work there the whole day. It seems to be working well for us. Neighbors that do not belong to the group may ask for our help occasionally, and we would help without expecting them to give anything in return. There is a person who abandoned the group because he did not want to keep track of the records required by the program. However, he is still welcome t o visit our farms to see what we are doing and ask questions. The most important thing is that the person that gets help is grateful, because that is what is valued for this case. If you help someone and that one does not value your
106 effort, because it is an effort what you are doing, then you have no desire left to help that person again. I believe that unconsciously we are used to help. If someone does not follow the rules inside the group, there may be consequences. That has already been established, the one who is not present at the monthly meeting has to pay a fine. There is a penalization. The most important thing is the attendance to the meetings. Everybody is aware of this commitment; they know the date and the place for the meeting, and the maj ority of members are usually there. We are still working in establishing all the bylaws. We know the rules already in our minds but we have not written them yet. They are not signed yet; many of them are still informal. There are always second chances. Nobody is perfect and sometimes the rules are broken. However, you can have a second, and probably a third, chance. However, the third is your last chance, after that you are out. It has not happened yet in our group. We also understand that tolerance is needed; we are all together in this group and should help others to follow the rules. Everyone must be committed. They need to feel that dedication to the group and be aware that today is the day of the meeting and we all should wear the shirt. In my g roup, we all wear a uniform to our meetings. We made those shirts for our annual evaluation and, after that, decided to wear them for the monthly meetings too. Today is our day, the GGAVATTs day, and we all are going to the meeting and will receive our guests. The reality is that the rules here are very flexible. From the beginning, the group was constituted by people willing to be there, nobody forced them. The moment that somebody wants to leave, or does not want to follow the rules, the doors are wi de open. To continue or leave is a personal decision for each one. In the GGAVATT we have learned a lot and have made improvements on many things. I feel that the program is working well. We were doing everything the traditional way, as our parents taught us. And I do not mean that this was wrong, no. But it needs, or needed, some technical guidance to do it better. We are also teaching our children the little things that we learn here. I have two sons; they go to the school in the morning and they help a little bit in the farm in the afternoons. I talk to them about what we are learning in the group so they can figure out what the GGAVATT is, and how it works. They get informed and up to date now. And I believe they enjoy it too. Because before w e suffered to feed and support the cattle during the dry season, and today we even have the luxury of giving away some pasture to the people who need it. Sharing this information with the children has increased their involvement with the farm. The childr en are more enthusiastic now. They have always helped in the farms because that is the custom; I helped my father and they help me. However, they are more motivated now that they are observing the results of what we learn. We have big expectations for the future. I hope to keep on improving, excelling myself for my own benefit. We have little time of being in the GGAVATT and we are already observing positive results, with more time it will be possible to have a better standard of living. There are some things, maybe too many, on which we still are inefficient, but by using the right technologies we can change that and become more efficient. That will help us, and our children, to have a better quality of life.
107 There is a saying that goes Just do good without selecting to whom. As human beings, I do believe that we must always think and practice this philosophy without expecting something in return. By being organized, so as we are, we are focusing more into this cause. We want to live better! Struct ure of the collective story on bonding social capital The structure of this story is represented in Figure 4 1 The story is formed by three sub stories, each one representing a quality that a good GGAVATT must possess. Sub -story 1 revolves around the concept of collaboration among group members and the tangible outcomes that can be achieved through it. A couple of examples of those outcomes are provided in sub sub -stories 1.1 and 1.2 (the acquisition of electric fences and the establishment of an animal feeds dealership, respectively). Sub -story 2 emphasizes the need for clear rules inside the group and the enforcement of these rules. The attendance at the monthly meeting is particularly relevant for the GGAVATT members, a nd this is described in sub-sub -story 2.2. Sub-sub -story 2.1 deals with the concept of reciprocity and makes a clear distinction between the limits and expectations that apply inside and outside of the group. The final sub-story, in a more personal tone, presents the GGAVATT participant as someone who envisions a better future and works actively for it. In this first story, GGAVATT participants provide examples describing what a good GGAVATT is. Trust, collaboration, democratic practice, respect for th e norms, and active participation are common elements in those examples and are highlighted as group values necessary for the common success. The story presents the GGAVATT group as a reliable, supportive, and motivating organization for its part icipants. In this organization each gavateros has a shared responsibility for the success of the group and they have institutionalized a set of norms to guide their interactions. The GGAVATT participants co -construct their futures within this structure and extend the t ransformative dialogue to their children.
108 Insights from Labovian analysis of the collective story on bonding social capital Table 4 14 contains examples of narrative structures in the first collective story that relate with the access to bonding social capital of GGAVATT participants. The c entral role of trust as a collective asset necessary for the success of the group is e vident in the abstract where it is established that A good GGAVATT is that where all the member s trust each other Trust provides the cohesiveness that the group needs to perform effectively. This idea is further emphasized in the story telling that There should be respect among the members, they should coexist harmoniously, and they should see each other as family. GGAVATT participants have certain expectations of the behavior that needs to be observed within the group to be recognized as a member. GGAVATT p articipants have some restrictions as individual farmers and they bring these limitations with them into the group. This situation is exemplified in the complication when they realize that A t the beginning we did not have grass chopper, much less a tractor, all the work was manual. However, the difference is that now the problem is shar ed among members and they have the capability of looking for a solution together. And they were able to overcome these limitations by working together We got organized to buy those things. This sentence summarizes the groups capacity to act collectively when they are challenged Through this newly found agency, GGAVATT participants acquire material goods and cooperate on everyday tasks; We have used that same procedure to build silos; we became organized to visit one farm each weekend and work togethe r. The evaluation and coda show that GGAVATT participants feel comfortable with their current achievements look to the future with confidence and perceive that membership in the group bri n gs more than economic benefits into their lives : They place a h igh value on being organized and on the possibility of helping themselves and other s through this organization.
109 Bridging Social Capital of GGAVATT Participants The second collective narrative is about b ridging social capital and is titled The extended c ommunity Th is story is about the struggle and success in the interactions between GAVATT participants and other people in their communities. : Collective narrative on bridging social capital I believe that the GGAVATT was the foundation for cattle pro duction to become an important economic activity in the region. We started doing things that nobody else was doing in the region. For example, that everybody used proven sires, that people started using improved grasses, electric fences, and rotational g razing, among others. Some practices are of difficult adoption, even to the interior of the group. However, I have seen many others practices diffuse and be adopted not only inside but outside the group. Much of what we have learned from the GGAVATT, the people around us started replicating almost immediately. One of the original commitments of our GGAVATT was that the information that we obtained had to be available for the whole community. This has happened not only with people from the community but also with people visiting from neighboring communities. It took time, but many people have imitated what we have done. From the beginning the people were copying us. They did not always use the technologies as were intended, but nevertheless they improv ed drastically. Although they did not want to accept it they were copying us. When they noticed that our cows produced more, looked prettier, and were healthier, they started to ask. Our work is in the fields and it is somehow a very open activity. In general, we know the majority of the people. Although the people were not participating with us in the GGAVATT, there was always contact with some member of the GGAVATT and they were exchanging information. Many of our ranches are just on the side of the road. When we are working with the cattle, people will walk -by and ask for information on what we are doing. Mariano, for example, has a brother and nephews that are not in the GGAVATT. His nephews helped him to prepare a silo. That silo was enough to f eed his cattle for two months during the dry season. The next sowing season his nephews decided to adopt this technology. The most interesting thing is that they had their silo ready sooner than everybody did, and always using technology from GGAVATT. O ther persons who were not inside the group, but that have a direct relation with the members, already started adopting different technologies. Claudio, who is a GGAVATT member, adopted good practices of hygiene for milking. His nephew decided to adopt the se same practices when his cows were having problems with mastitis. He did so because he has observed the good results that his uncle was having. Even though Claudios nephew was not a GGAVATT member, he was completely convinced of the benefits of adopti ng the right practices. He started to attend
110 our meetings as a listener, but since last month we decided to adopt him in the group and now is our newer member; now he is planning to do a corn silo. I think that we are having an informative impact in the c ommunity. Today, in the majority of the communities in the region there are, or have been, GGAVATTS. The advisor has been a key factor to involve people from the community with the group. From the beginning the doctor [the groups advisor] was very open to help any person that approached him. The doctor has been always willing to talk with the people and provide them with every type of information, although they were not members of the GGAVATT; if they had some problem he would give them the solution. Y es, some people would come saying that they had a sick cow and if the doctor was here he would go with them to treat her. He was consulted not only on dairy cattle, sometimes it was horses, beef cattle, anything that the people needed. The GGAVATT has aff ected the agricultural activity in the region. Cattle production was important before the GGAVATT, but not as it is today. It was mainly a backyard activity, something secondary. Producing corn was more important for the people and much more valuable; it was definitely more important. The people were always saying that cows were not a business, that they represented only a small savings account. Cows were considered unimportant for the economy. However, when they realized that by applying some methods it could became a more profitable activity, they changed their perceptions and started considering it as an important productive activity. Before the GGAVATT we were troubled. My compadre and I were ready to cross the river to go to the other side; we w ere going as wet backs to the United States. We thought about it seriously and were determined because the situation here was really harsh, however, we decided to try the GGAVATT and now we are better. Now we know that livestock production can be a good business with the correct advice, and we do not want to leave. The change was noticeable; I would say that after only three years the effect of the GGAVATTs work was evident in the region. It is really amazing the large number of farmers in the communi ty that are, or have been, members of the GGAVATT through the 17 years that the group has been active here. It seems like it belongs to the whole community. GGAVATT participants are members of larger communities and relate on a daily basis with nonparti cipant farmers, sometimes this relation is quite difficult. Many people criticize us saying "why should I pay for that [technical advice] if they are doing the same or worse than us? However, I consider that the reality is not like that. It is not the s ame to see things from the outside than being inside and knowing exactly how much it is costing us to produce a liter of milk. Even when our success is evident, they criticize us; usually they would say that what we are obtaining is because the government is giving it to us. We have been through different situations where people have signaled and accused us of many things. Suddenly, there were people saying that we were crazy and that the GGAVATT was useless. These types of comments can easily damage you r motivations and make you
111 abandon your goals. However, you must hold to a very well founded idea to avoid being moved by those comments. Those people criticizing us are deceiving themselves. Before the GGAVATT, we thought that we were doing things in a good way, but when we started to calculate costs, we realized that we were mistaken. I am convinced that in order to reach our goals we are required to go through multiple experiences, and most of all, live through them. If we had not put an effort in f orming the group, maybe no one would be talking about us. Structure of the collective story on bridging social capital As Figure 4 2 illustrates, this collective narrative consists of 4 sub-stories The first sub story descri bes how participants have helped to disseminate technological innovations to nonparticipants in their communities. The second sub-story highlights the importance of the role of the a dvisor in opening the group to the community. Sub -story 3 elaborates on the impacts of the interactions described in the two previous sub -stories, emphasizing the economic transformation and the community attachment. The final sub -story reflects on the fact that the interaction between participants and non participants has n ot always been easy. In fact, some GGAVATT participants have been subject of direct attacks from their neighbors: We have been through different situations where people have signaled and accused us of many things. Suddenly, there were people saying that we were crazy and that the GGAVATT was useless. The topic of this sub -story relates perfectly with Machiavellis words from The Prince (2003, p. 20) And it ought to be remembered that there is nothing more difficult to take in hand, more perilous to conduct, or more uncertain in its success, than to take the lead in the introduction of a new order of things. The collective story about bridging social capital situates the GGAVATT participants as members of larger communities and pre sents some examples of how group and extended community interact. In this constant negotiation between participants and non -participants, GGAVATT members identify themselves as agents of change and the presence of the groups as a valuable resource for the ir communities. The gavateros are role models for their peers and they
112 take this as a serious responsibility that they have with their communities even though their interaction with neighbors is not always cordial. The role of the advisor is described a s a key factor for a successful development of bridging linkages between group and extended community. The group provides the community with information that promotes transformation, and the community contributes to the sustainability of the groups by fos tering the formation of new groups and the incorporation of new members in established ones. Insights from Labovian analysis of the collective story on bridging social capital In Table 4 15, some examples of narrative struct ures in the second collective story are presented accompanied by their interpreted contribution to access to bridging social capital. T he abstract indicates that the GGAVATT has been an instrument of regional economic development. Cows were considered u nimportant for the economy. However, when they [nonparticipants] realized that by applying some methods it could become a more profitable activity, they changed their perceptions and started considering it as an important productive activity. The compl ication of the story shows that w hen non -participants observed the good results that GGAVATT members were achieving, they approached the participants and asked for information: Many of our ranches are just on the side of the road. When we are working wit h the cattle, people will walk -by and ask for information on what we are doing. GGAVATT participants needed to decide between sharing the information that they were obtaining or keeping it for themselves. They decided for the former because according to them One of the original commitments of our GGAVATT was that the information that we obtained had to be available for the whole community. The result of this decision was that in just a few years of work some regions of the state have been positively a ffected by the work of the GGAVATT groups presenting now a larger number of thick productive networks The evaluation is positive because this attitude of GGAVATT participants has translated into more opportunities for young
113 farmers that do not need to le ave family and land looking for progress: Before the GGAVATT we were troubled. My compadre and I were ready to cross the river to go to the other side; we were going as wet -backs to the United States. We thought about it seriously and were determi ned because the situation here was really harsh. However, we decided to try the GGAVATT and now we are better. Now we know that livestock production can be a good business with the correct advice, and we do not want to leave anymore. Group participants also benefit from this openness to the community because it has proven to be an effective way of recruiting new members which contributes to the sustainability of the groups and also because of the satisfaction that they obtain from making positive change s in their communities. Linking Social Capital of GGAVATT Participants The third collective narrative is about linking social capital and is titled Dealing with institutions Th is story shows how GGAVATT participants interact with institutions and the benefits that they derive from these relationships. Collective story on linking social capital There is more power in operating as a group than as individuals. As a group we have enough force. One of the important parts of acting like a group is our enh anced capacity to access resources from institutions. There are many programs out there, there are all kinds of supports, but they are never going to come to us if we do not do a little work. This is how we obtained the tractor. There was an agricultural exposition in Xalapa and we had the chance to talk with the states secretary of agriculture there. We asked for support from his office to buy a tractor and he committed to provide 50% of the cost. When we returned to our community we hold a meeting at the farmers association to invite all the farmers to participate. At the end, only the GGAVATT participants decided to get involved. We were 11 members and we were able to put the money to buy the tractor. We tried to get a motocultor that way too. Fi rst we attempted to obtain it with support from DesaRur but that did not turn out well. We requested money from them and it was approved, but they did not allow us to choose where to buy it. They made us buy it from a company that they had already selec ted. The problems with the motocultor started as soon
114 as it arrived. Its quality was really poor the pieces were falling apart. The engine was supposedly German but everything else was Chinese. When we complained about the quality of the product and sa id that we wanted to exchange or return it, they did not help at all. We finished losing not only their support but also a big part of the money that we had originally invested in the motocultor. It was a really bad experience and we felt like the childr en whose candies were taken. But now we have a good motocultor thanks to the doctor. He travelled to Chipilo to attend a cattle show and saw this motocultor there. When he returned he told us about it and we, as a group, agreed to cooperate to buy it. Later, the groups officials traveled to Chipilo and bought it, and there it is, working for everybody. We are lucky of having the doctor as our advisor. First, an Engineer came trying to form a group in the community and wanting to be the advisor of that group. He almost succeed, however, we were unable to reach the minimum number of cows required by the GGAVATT program. It was good though, because we later learned that the Engineer had a history of forming groups just to get paid but he did not work at all with them. Sometime later, a woman named Rosario came to the region saying that she was working for DesaRur and was here to form groups and help us. She asked for money to prepare the paperwork. She received money from some people and she even brou ght people from DesaRur to visit with us. She later asked us to sign some papers saying that she had been working with us and she promised that those papers were going to be useful for us to get support from DesaRur The truth is that nothing happened; s he disappeared with the papers and the money. Finally, there was a meeting at the Local Farmers Association and the doctor was there. After the meeting he explained what the GGAVATT program is, and what it is not. He promised to return with more inform ation and with the people from SAGARPA and SEDARPA; he did so and, even though we were still distrustful, we decided to work with him and we are very happy with the results. As a group it is easier to request, and obtain, money to buy a tractor, a milking machine, or a grass chopper that will be useful not only to one person but to many. We have always had the problem of having very low prices to sell our products. Four years ago, a candidate for representative visited us and we shared our concerns about t he price of the milk. After he won, he returned willing to solve our problem. He had realized that the low price of milk was a common problem in his whole district. He hired a private consultant firm to assess the problem and identify solutions. This f irm found that we could increase the price of our product by adding more value to it. It was proposed that the farmers in the district got organized to establish a milk processing plant. The representative talked with the municipal presidents in his dist rict, and invited all the farmers to get organized. As a result, nine Societies of Rural Production (SPR) were formed in seven municipalities; many of these farmers were already integrated into GGAVATT groups. These societies came together to form a big company and hired another firm to develop the project for the milk processing plant. The result is that we are currently in a quite advance phase of building a plant with an approximate cost of 1.6 million dollars, and the capacity to process 70,000 liter s per day. The 202 associates are contributing 25% of the total cost and the rest of the money has been obtained from
115 multiple governmental programs. We are sure that this project is not only a project for the dairy producers; it is a project for regiona l development. In dealing with all these institutions, I have learned a couple of things. First, it is a fact that there are many programs and resources available, but they are not sufficient to alleviate all the problems that rural people have. Second, I believe that the biggest problem of the system is its truly incredible levels of corruption. Every person wants a percentage of the money that you are going to receive. This way we are not going to solve the problems, indeed. Structure of the collectiv e story on linking social capital This collective narrative about linking social capital is formed by three sub -stories (see Figure 4 3 ). The first sub -story provides examples of how the groups have been able to obtain valuab le support for equipment for them (and their communities) through negotiation with governmental and parastatal institutions. The second sub-story is about the people that ultimately represent the different institutions ; a series of examples of good and ba d experiences with these individuals are presented to illustrate these interactions The last sub -story describes how a request for assistance originated in a GGAVATT group turns into a 1.6 million dollar project with 202 farmers involved. The story finishes with a reflection on what the GGAVATT participants have learned about dealing with institutions. E xamples of the use of the collective presence of GGAVATT participants in their dealings with diverse institutions and the people representing them are present in this collective narrative In the story the gavateros reflect on the benefits that can be derived from interacting with institutions, on the roles that individuals play representing a given institution and connecting it with the people, and fi nally, on the power that they have to transform their regions through their own organization. Insights from Labovian analysis of collective story on linking social capital As illustrated in Table 4 16, the gist of the story is that participants feel empowered and employ this power for action W hen the y realize d their need for help to face the market
116 challenges they use d their collective voice to request assistance from a political candidate: The candidate promised to help t hem as a result of their dialogue After he won, he returned to solve our problem. He had realized that the low price of milk was a common problem in his whole district so he hired a private consultant firm to assess the problem and identify solutions. The consulting firm identified both the need to add more value to the product and the lack of capacity of individual farmers, or even small groups of farmers, to address that task effectively. They came up with the idea of organizing the farmers of the region into Societies for Rural Production that came together to estab lish a milk processing plant. The result is that we are currently in a quite advance phase of building a plant with an approximate cost of 1.6 million dollars, and the capacity to proc ess 70,000 liters per day. Dealing with institutions, GGAVATT participants have realized that as a group it is easier for them to access governmental funds However, they have also understood that these funds are insufficient to solve the problems of ru ral people and, therefore, it is unrealistic to wait for paternalistic policies and programs to achieve real development. This is probably the best example of empowerment found in the story and represents an evolution in the conceptualization of agricultu ral extension in Mexico that is moving from solving the problems of the people to helping people solve their problems They also recognize the big problem present in many institutions of the country : I believe that the biggest problem of the system i s its truly incredible levels of corruption. Every person wants a percentage of the money that you are going to receive. This way we are not going to solve the problems, indeed. Comparing Narrative Structures of the Three Collective Stories The Labovi an analysis of the three collective stories have shown the role of different narrative elements in situating the meaning that GGAVATT participants make of their access to different forms of social capital. There are two constant motifs present across the storylines
117 The first one of them is evident in the narrative element result of the three stories. This motif is the capacity for collective action that GGAVATT participants exhibit. This collective action transforms into new capacities in their group s, communities, and regions and results from placing enough amounts of trust in group members (bonding), neighbors (bridging), and institutions (linking). The final product of this capacity for action is represented by a feeling of empowerment ; this is th e second constant motif P articipants share this feeling in their stories in the form of the narrative element evaluation where they reflect on the final results of what they have done as GGAVATT members. Both collective action and empowerment have bee n previously used as proxy indicators of social capital valuable for its study (Grootaert, Narayan, Nyhan Jones, & Woolcock, 2004; Onyx & Bullen, 2000) The capacity for c ollective action of GGAVATT participants is evident in the results of the stories at three different levels. First, farmers are capable of overcoming different types of limitations by institutionalizing collaboration within their groups. Next the openness of the groups to the communities is associated with a proliferat ion and thickening of productive networks that is evident in the incorporation of new members into established groups and the formation of new ones. Finally, at a regional level, the size of the networks is increased and its structures modified as a resul t of the merging of multiple GGAVATT groups and individual farmers forming new and different organizations more adequate to face greater challenges of the access to markets. Diverse forms of e mpowerment are also evident in the evaluations of the three co llective stories. The gavateros feel more secure and confident which translates into a positive attitude looking into the future. Additionally, they have the power to stay in their communities and work what belongs to them; they do not need to leave to l ook for opportunities because they are
118 capable of creating their own. But overall, they have realized that the power to solve their problems resides within themselves. They know how to get help and they know that they should ultimately be responsible of solving their own problems. Numbers and Words Come Together The third, and final, objective of this study was to connect the description and context obtained in specific objectives 1 and 2 to represent the complexity and multidimensionality embedded in t he interactions fostered by the GGAVATT program and discuss its plausible consequences. This section presents a general account on how these two products interconnect; however, the discussion of these interconnections is part of Chapter 5 In c omparing th e results from both segments it is possible to observe that even though some similarities and differences are present the overall feeling is of complementarity The quantitative segment illustrates factors that distinguish and affect an individuals socia l capital between groups, while the qualitative segment presents a collective representation of how GGAVATT participants struggle and success accessing social capital in their daily activities. Looking at the similarities between both segments it is evident that GGAVATT participants have access to all three forms of social capital (i.e. bonding, bridging, and linking) through different programs activities. Both segments also show that this access, and its resulting capacity for collective action, can b e positively affected by factors such as time, current institutional connectedness, and the importance that participants place on the programs activities. T he quantitative segment shows that gender, region of the state, homogeneity of the group, the num ber of previous related and unrelated contacts, as well as the current number of unrelated contacts, are significantly associated with variations in access to social capital. However, the role of these factors is not evident in the qualitative narratives.
119 Finally, while the quantitative segment shows how some attributes of participants and groups are associated with access to greater or lesser levels of social capital for individuals the qualitative segment exemplifies how this access materializes into participants lives as collective action and empowerment allowing them to overcome certain limitations. The qualitative narratives represent how GGAVATT participants actively socialize with others (inside and outside the group; horizontally and vertically) in synergy with the trust that they place in the network and the expectations that they have. This is consistent with the rationale behind the creation of the multiplicative indexes for bonding, bridging, and linking social capital used in the quantitat ive segment. Summary The results from both segments of the study, organized by specific objectives, have been presented in Chapter 4. A description of the social capital of GGAVATT groups and participants was provided, it was possible to put in perspec tive how social capital takes form in the lives of program participants, and an effort was made to depict how description and context come together to offer us a glimpse of the social capital of GGAVATT participants. The discussion and implications of the se results are presented in Chapter 5.
120 Table 4 1. Demographic attributes of advisors and respondents Attribute Distribution Advisors gender Female 10 % Male 90 % Advisor lives in community No 63 % Yes 37 % Advisor has a second job No 32 % Yes 68 % First time advisor No 2 % Yes 74 % Advisors Education Other a 34 % Veterinarian 66 % Group participant is local r esident No 26 % Yes 74 % Farm is main activity No 24 % Yes 76 % Group participant h as family in group No 41 % Yes 59 % Group participant s gender Female 9 % Male 91 % Marital status of group participant Other b 21 % Married 79 % Groups t ype of production Other c 27 % Dual purpose cattle 73 % a Includes: Agronomists, biologists, and agricultural specialists. b Includes: Single, divorced, and widowers. c Includes: Dairy cattle, beef cattle, sheep, and bees. (Back to text )
121 Table 4 2. Descriptive statistics for bonding, bridging, and linking social capital. Statistic Bonding social capital Bridging social capital Linking social capital Mean 44.95 18.67 15.09 Standard deviation 22.82 16.22 12.59 Median 42 .00 15.28 13.28 Minimum 0 .00 0 .00 0 .00 Maximum 100 .00 75 .00 65 .00 Central region (n=201) 39.10 18.96 14.25 North region (n= 115) 47.06 20.19 17.18 South region (n=90) 55.33 16.05 14.26 (Back to text )
122 Table 4 3. Mean and standard deviation of relevant continuous independent variables Variable Mean SD Importance of group activity 4.3 5 .65 Group tolerance 1.9 3 1. 12 Homogeneity 3.11 1.71 Organizations available in community 4.89 1.97 Importance of group bridg ing / linking activity 4.07 .90 Community tolerance 2.28 1.22 (Back to text )
123 Table 4 4 Percentile distribution of responses for reflexive controls Decreased significantly Decreased Neutral Increased Increased significantly Trust in related institutions 2 3 29 59 8 Trust in neighbors .5 5 48 32.5 14 Trust in strangers 9 11 64 14 1 Trust in political parties 9 22 50 13 6 Trust in communication media 12 17 54 14 3 Local involvement .7 9 60 21 9 (Back to text )
124 Table 4 5 Mean comparisons of bonding, bridging, and linking social capital across levels of the categorical independent variables Categorical variable Bonding Bridging Linking Advisors gender Female 49.10 .9.12*** 6.95*** Male 44.50 19.71*** 15.97*** Advisor is local resident No 45.53 17.96 14.65 Yes 43.84 20.02 15.91 Advisor has a second job No 42.57 13. 85*** 12.75*** Yes 46.21 21.20*** 16.32*** First time as advisor No 37.53*** 14.67*** 12.92** Yes 47.88*** 20 24*** 15.94** Advisor has family No 44.43 17.53** 14.26** Yes 46.11 21.18** 16.92** Advisors education Other a 43.20 15.48*** 13 .59* Veterinarian 46.00 20.57*** 15.98* Member is local resident No 47.19 19.83 16.42 Yes 44.15 18.25 14.61 Farm is main activity No 46.51 18.03 14.83 Yes 44.45 18.87 15.17 Has family in group No 44.33 15.68*** 13.82* Yes 45.39 20.75** 15.97* Has best friends in group No 44.35 17.04 13.78 Yes 45.30 19.60 15.84 SES Low 40.80 18.37 16.00 Medium 45.81 18.70 14.97 High 38.44 19.37 13.84 a Includes: Agronomists, biologists, and agricultural specialists. b Includes: Dairy c attle, beef cattle, sheep, and bees. *Tendency to significance. ** Significant at the .05 level of alpha *** Significant at the .01 level of alpha
125 Table 4 5. Continued. Categorical variable Bonding Bridging Linking Gender of participant Female 44.61 13.24** 11.74* Male 44.99 19.22** 15.43* Participants education Elementary 40.51*** 16.81 14.23 Secondary 45.67*** 20.15 16.59 Higher 51.24*** 19.88 14.67 Age of participant 40 or less 40.92** 18.03** 15.40* 41 50 48.34** 22.31** 17.1 4* More than 50 45.36** 16.78** 13.60* Years in the program Less than 1 43.43*** 13.37*** 11.48*** 1 3 43.02*** 19.70*** 16.36*** More than 3 53.08*** 26.63*** 19.01*** Region of the state Central 39.10*** 18.96 14.25* North 47.06*** 20 .19 17.19* South 55.33*** 16.05 14.26* Type of production Other b 45.42 14.34*** 14.23 Dual purpose cattle 44.78 20.27*** 15.40 Years the group has existed Less than 1 40.24** 15.70*** 11.53*** 1 3 46.88** 18.57*** 15.89*** More than 3 4 6.29** 25.42*** 19.04*** a Includes: Agronomists, biologists, and agricultural specialists. b Includes: Dairy cattle, beef cattle, sheep, and bees. *Tendency to significance. ** Significant at the .05 level of alpha *** Significant at the .01 level of al pha (Back to text )
126 Table 4 6 Correlations between bonding, bridging, and linking social capital and scale independent variables. Bonding social capital Bridging social capital Linking social capital Importance of group activity .35 ** .17 ** .18** Group tolerance .14 ** .08 .07 Homogeneity .00 .05 .12 Trust change in neighbors .08 .16** .05 Trust change in strangers .01 .01 .01 Formal networks in community .15 ** .0 4 .08 Previous network activity .16** .15 ** .09 Current network activity .00 .0 9 .08 Importance of bridging/linking activities .24** .17** .19 ** Community tolerance .0 6 .13 ** .02 Direct institutional trust change .14 ** .04 .07 Trust change in p olitical parties .02 .01 .04 Trust change in communication media .02 .04 .12* Previous related institutional contacts 19 ** .33** .33** Current related institutional contacts .23** .24** .24** Related institutional contact change .03 10 .10 Unrelated previous institutional contacts .12 .11 .12* Unrelated current institutional contacts .08 .14** .16** Unrelated Institutional Contact Change .05 .62 .06 Advisor's Age .0 6 .18 ** .14** Correlation i s significant (p ** Correlation is highly significant (p 0.01). (Back to text )
127 Table 4 7 Multiple regression model for bonding social capital. Component Estimate P value Intercept 47.35 .00*** Importance of group activities 7.53 .00*** Years in the program 5 .04 .00*** Region is South 11.73 .00 *** Region is North 4.09 .11 Previous related institutional contacts 2.37 .02** Current related institutional contacts 1.50 .16 Group tole rance 1.83 .02** Advisor is female 5 .00 .17 Adjusted R square: .194 Significance of the model: .000 ** Significant effect. *** Highly significant effect (Back to text )
128 Table 4 8 Multiple regression model for bridging social capital Component Estimate P value Intercept .39 .93 Importance of bridging / linking activities 3.5 4 .00*** Years in the program 4.67 .00*** Trust change to neighbors 3.61 .00*** Region is not North 5.43 .00*** Region is not S outh 7.90 .00*** Length of groups existence 2.75 .02 ** Advisor is female 9.11 .00*** Group participant is female 5.07 .03** Is not a first time advisor 5.46 .00*** Advisor is not a veterinarian 3.60 .01 ** Previous related institutional con tacts 2.93 .00*** Current related institutional contacts 1.1 6 .10 Unrelated institutional contact change 2.37 .02** Advisor is not a local resident 1.21 .43 Farming is not the main activity 1.14 .48 Adjusted R square: .30 Significance of the model: .00 *Tendency to significance. **Significant effect. *** Highly significant effect. (Back to text )
129 Table 4 9 Multiple regression model for linking social capital Component Estimate P value Intercept 11.54 00*** Importance of bridging / linking activities 2.36 .00 *** Years in the program 2.69 .00 *** Previous related institutional contacts 1.98 .00 *** Current related institutional contacts 1.55 .01*** Homogeneity 1.03 .00 *** Region is not North .21 .8 9 Region is not South 1.2 1 .4 6 Advisor is female 8.21 .00 *** G roup participant is female 2.7 6 .15 Advisor does not have a second job 2.02 .10 Is not a f irst time as advisor 2.42 .10* Adjusted R square: .22 Significance of the model : .00 *Tendency to significance. ***Highly significant effect. (Back to text )
130 T able 4 10. Multilevel multiple regression model for bonding social capital Component Estimate P value Intercept 48.18 .00*** Importance of group activities 7.60 .00 *** Years in the program 4.47 .01*** Region is not South 11.48 .00 *** Region is not North 3.96 .22 Previous related institutional contacts 2.0 7 .04 ** Current related institutional contacts 1.41 .19 Gr oup tolerance 1.6 9 .03** Advisor is female 4.5 2 .32 2 Restricted Log Likelihood; Intercept only: 3,646.26 Model: 3,567.59 Covariance parameters. Intercept only: Group level = 109.37 Residual = 415.06 Model: Group level = 24.71 Residual = 398 .74*** **Significant effect. ***Highly significant effect. (Back to text )
131 Table 4 11. Multilevel multiple regression model for bridging social capital Component Estimate P value Intercept 6.46 42 Importance of bri dging / linking activities 2.20 .01 ** Years in the program 3.9 3 .00 *** Trust change to neighbors 2.01 .02 ** Region is not North 4.5 6 .19 Region is not South 6.35 .09* Length of groups existence 1.59 .49 Advisor is female 8.26 .08 G roup participan t is female 5.0 1 .02 ** Is not f irst time as advisor 4.59 .18 Advisor is not a veterinarian 3.7 1 .17 Previous related institutional contacts 2.10 .00 *** Current related institutional contacts 1.4 4 .04 ** Unrelated institutional contact change 1.6 3 .0 9* Advisor is not a local resident 1.32 .64 Farming is not the main activity 3.49 .23 2 Restricted Log Likelihood; Intercept only: 3,290.47 Model: 3,175.35 Covariance parameters; Intercept only: Group level = 101.12 Residual = 161.77 Model: Group level = 45.75** Residual = 148.54*** *Tendency to significance **Significant effect. ***Highly significant effect. (Back to text )
132 Table 4 12. Multilevel multiple regression model for linking social capital Component E stimate P value Intercept 12.87 00*** Importance of bridging / linking activities 1.01 .1 4 Years in the program 2.62 .01*** Previous related institutional contacts 1.62 .00 *** Current related institutional contacts 1.48 .01*** Homogeneity .94 .02** Region is not North .62 .8 3 Region is not South .17 .95 Advisor is female 7.72 .05 ** G roup participant is female 1.97 .26 Advisor does not have a second job 1.76 .46 Is not a f irst time advisor 2.32 .41 2 Restricted Log Likelihood; Intercept only:3,100.22 Model: 3,018.69 Covariance parameters. Intercept only: Group level = 56.88 Residual = 102.05 Model: Group level = 34.88** Residual = 94.26*** **Significant effect. ***Highly significant effect. (Back to text )
133 Table 4 13. Example of co -construction in the collective story on bonding social capital. Interview Group Participant Collective story San Juan Diego A good GGAVATT is that where all the members trust each other and are willing to pursue a co mmon goal. Jutla Santiago There should be respect among the members, they should coexist harmoniously, and they should see each other as family. That is a good GGAVATT. San Juan Le n We all collaborate in the GGAVATT and we have obtained many thi ngs as group. At first we did not have a grass chopper, much less a tractor, all the work was manual. Calat n Pablo We got together to buy electric fences. As individual farmers we were unable to buy an electric fence; we did not even know about them. We decided to buy them as a group and raffle off the turns. We were 12 or 13 Gavateros. We got together to make a raffle for the turns and once the order was established, we calculated the amount of money that each one needed to provide to cover the cost. Every 15 days each one of us made that money available to buy a new electric fence. That is how we acquired them. Jutla Gerardo We also got an animal feeds dealership. A representative of the company came to our monthly meeting to give a talk and aft er that he asked if we were interested in selling his products in this community. All the members voted and we decided to become his dea lers. Since then, we buy the fee d that we need for our cows at a special price. The group also sells f ee d to other farme rs that are not part of the GGAVATT at a higher price. San Juan Diego That same way we have obtained grass choppers, medicines, and many other things. We have used that same procedure to build silos and we became organized to visit one farm each weekend and work together. (Back to text )
134 Story 1 : What a good GGAVATT is Sub story 1: Everybody collaborates Sub sub story 1.1: Electric fences Sub sub story 1.2: The animal feeds dealership Sub story 2: The group has rules Sub sub story 2.1: Reciprocity Sub sub sub story 2.1.1: Inside and outside the group Sub sub story 2.2: The monthly meeting Sub sub sub story 2.2.1: Missing the meeting Sub story 3: Transforming the future Sub sub story 3.1: Teaching the children Sub sub story 3.2: The aspirations Figure 4 1. Structure of the collective story on bonding social capital. (Back to text )
135 Table 4 14. Narrative structures from the collect ive story on bonding social capital and their interpreted contribution to social capital. Excerpt from story and sub-stories Narrative structure Interpreted contribution to social capital access A good GGAVATT is that where the members trust each other. Abstract Central role of trust as a collective asset necessary for the success of the group. At the beginning we did not have grass chopper, much less a tractor, all the work was manual. Complication Their individual limitations are now a group challen ge and they addressed them as such We got organized to buy those things. Every 15 days every one of us collaborated equally with money to buy an electric fence. That is how we all have electric fences now. Result Group membership allows farmers to work collectively to t ransform their current realities GGAVATT participants embrace big expectations for the future. Evaluation Membership provides security that translates into positive attitudes. Just do good without selecting to whom Coda Membership p rovides participants with opportunities to improve other lives. (Back to text )
136 Story 2 : The extended community Sub story 1: Diffusion of technology Sub sub story 1.1: Others are copying us Sub sub story 1.2 : Preparing a silo Sub sub story 1.3: Milking hygiene Sub story 2: The Advisors role Sub story 3: Regional transformations Sub sub story 3.1: Economic activity Sub sub story 3.2: Community attachment Sub story 4: Social pressure S ub sub story 4.1: Why should I pay to do what you are doing? Sub sub story 4.2: Keep your focus Figure 4 2. Struct ure of the collective story on bridging social capital (Back to text )
137 Table 4 15. Narrative structu res from the collective story on bridging social capital and their interpreted contribution to social capital access Excerpt from story and sub-stories Narrative structure Interpreted contribution to social capital access The GGAVATT was the foundation for cattle production to become an important economic activity in the region. Abstract Farmers identified the GGAVATT as a viable structure to achieve economic development. When they noticed that our cows produced more, looked prettier, and were healthi er, they started to ask. Complication The GGAVATT, as a source of information, represents an asset both for participants and non participants. Today, there is a GGAVATT in the majority of the communities in the region. Result Productive social networks i n the region have increased and thickened. Now that we know that with the correct advice we can produce, we do not want to leave our communities. Evaluation Through the GGAVATT, community attachment has increased. We are having an informative impact in the community Coda Information and communication in the community have been affected by the GGAVATT. (Back to the text )
138 Story 3 : Dealing with institutions Sub story 1: Equipping the group Sub sub story 1.1:Buyi ng a tractor Sub sub story 1.2: Problems to get a motocultor Sub story 2 : People and Institutions Sub sub story 3.1: The Engineer Sub sub story 3.2: Rosario Sub sub story 3.3: The Doctor Sub story 3 : Union is strength Sub sub stor y 3 .1: What is the problem? Sub sub story 3 .2: Getting organized Sub sub story 3 .3: Future of the region Lessons learned Figure 4 3. Structure of the collective s tory on linking social capital. (Back to text )
139 Table 4 16. Narrative structures from the collective story on linking social capital and their interpreted contribution to social capital access. Excerpts from sub-stories Narrative structure Interpreted contribution to social capital access We have mor e power as a group than as individuals. Abstract Participants are empowered for political action. We have always had the problem of having to sell our products at very low prices, and we asked a politician for help. Complication The farmers have now a c ollective personality and voice that is used to establish dialogues with people in power positions. Nine Societies of Rural Production (SPR) were formed in seven municipalities. Many of these farmers were already integrated into GGAVATT groups. Result Ch anges in the size and structure of the network as a strategy to access greater levels of social capital There are many programs and resources available but they are not enough to alleviate all the problems of rural people. Evaluation Participants unders tand that they need to have an active role in development and, therefore, they are truly empowered. As a group it is easier to request, and obtain, support to buy equipment that will be useful to many people. Coda The benefits negotiated as a group may sp ill over to the extended community. (Back to text )
140 CHAPTER 5 SUMMARY AND DISCUSSI ON Introduction Extension scholars around the world have proposed the implementation of group-extension approaches as an alternative to increase the access to social capital at the local level and help smallholders meet the challenges of development. The need arises then to investigate the mechanisms through which this type of intervention may affect social interaction of its beneficiari es. This dissertation is a report of an evaluation study of a group -extension program found in Mexico. Membership in this program and access to three different forms of social capital, namely, bonding, bridging, and linking, are explored. In addition, the effects and implications of participant demographics, group characteristics, and community uniqueness on access to social capital are discussed. Statement of the Problem If we reconsider the description of good practices in contemporary extension sys tems offered by Swanson (2008) (i.e. decentralized, market -driven, pluralistic, and characterized by farmers taking a leadership role), the need for coordinated and active participation of farmers in modern extension systems is evident. It is important th en to explore the suggested association between farmers participation in group -extension programs and access to social capital. Special attention must be placed in the identification of the activities fostered by this type of program that situate the par ticipant farmers as protagonists of the social capital process and active co constructors of a better collective future. To determine the attributes and activities of GGAVATTs that are associated with greater levels of social capital, the purpose of this s tudy is to create a profile of the social dynamics of
141 Veracruzan farmers in terms of the forms and types of social capital that are accessible to them as participants in the GGAVATT program. The specific objectives of this study are: 1 To provide a descript ion of the social capital of GGAVATT groups and participants by: a ) Describing the distribution of dependent (i.e. bonding, bridging, and linking social capital) and independent variables among respondents. b ) Exploring the bivariate relationships present betwee n dependent and independent variables. c ) Investigating the multivariate relationships affecting bonding, bridging, and linking forms of social capital. d ) Determining the distribution of the variation explained by the multivariate models at the individual and group levels. 2 To contextualize how social capital takes form in the lives of the participants by looking for evidence of access to social capital, represented through transformations in roles and cooperation, in the stories that GGAVATT participants co con struct around their participation in the program 3 To connect the description and context obtained in specific objectives 1 and 2 to represent the complexity and multidimensionality embedded in the interactions fostered by the GGAVATT program and discuss i ts plausible consequences. Review of the Methodology This responsive evaluation (Guba & Lincoln, 1986) was planned within the overarching notion that social phenomena are highly complex and should be explored using a combination of methodologies. The s tudy follows a goal -free, theory -driven evaluation approach (Chen & Rossi, 1980) to explore access to different forms and types of social capital as an unintended, unanticipate d, direct consequence of the participation in the GGAVATT program (Rogers, 2003) The connections between program participation and social capital access are investigated with quantitative and qualitative methodologies arrang ed in a parallel or concurrent mixed method design (Tashakkori & Teddlie, 2003)
142 Summary and Discussion of Results A holistic review of the results from both segments of this study demonstrates that GGAVATT participants have access to different forms of social capital which translates in their lives into collective action and empowerment This is of the utmost importance because as the collective stories show, participants would not be able to generate and take advantage of opportuni ties for development if they dont have a more or less balanced stock of bonding, bridging, and linking social capital. Through their bridging and linking ties the groups are able to find opportunities present in larger networks that can be later capitali zed if they posse s s strong bonding ties. This is consistent with the findings that Dahal and Adhikari (2008) obtained studying collective management of resources at the community level. They found that a united community was capable of influencing state policies through an increased access to resources and information from an expanded network. For example, the narratives show that the GGAVATT provides participants with opportunities and a favorable environment to interact with diverse institutions. This form of connection is important because it enables the groups to communicate with people in influential positions, from whom they may be able to obtain valuable information, funding, or even the development of a more favorable political environment for th eir productive activities. This has led to situations in which the local GGAVATT group is taking the lead to download different kinds of resources to their communities such as equipment or marketing advice According to Mercoiret et al. (2007) this type of organizations can only play an effective role of intermediation between farmers and other actors when they operate in an institutional environment favorable for that activity. Additionally, the results from both segments have shown how important are the current institutional contacts for participants access to the different forms of social capital. The more people that participants know in positions of influence within the institutions related with
143 GGAVATT, the more likely they would be to have greater levels of social capital and, consequently, to become empowered through their capacity for collective action. The results from both segments also have provided evidence of the positive effect that time and motivation have on access to the different form s of social capital. The statistics show that those participants that had been GGAVATT members for a longer time are more likely to have greater levels of social capital; and this effect of time is also present in the collective stories. If we think abou t the elements forming the three dependent variables of this study (i.e. bonding, bridging, and linking social capital; see Figure 2 6) it would be expected for time to play such an important role. Trust is something that takes time to develop, even though there must be already a certain level of trust since the moment that participants decide to get together and form a group, and may result from multiple factors as was described in Chapter 2. In addition to that, the establishment of reciprocity, which s ome scholars have equated to trustworthiness, goes hand to hand with the development of trust and, therefore, is also dependent on time. Finally, as trust and trustworthiness develop over time participants would likely be more motivated to participate in the activities and this motivation would translate into action. In direct connection with the participants motivation, both segments have also shown that the importance that participants place on the different program activities is also positively associ ated with greater levels of social capital; this importance is evident for other participants as this excerpt from the collective story on bonding social capital reveals: Everyone must be committed. They need to feel that dedication to the group And G GAVATT participants continue this story exemplifying how important it is for them to enforce the norm of attending the monthly meetings and the celebratory nature of these meetings involving the participation of the host family. It is this commitment and interest in the group which would ultimately lead to what Bourdieu (1986)
144 referred to as recognition. The implic ation of this finding is that every day in the program counts. The advisor needs to be very careful to make sure that the pass age of time is having a positive effect on the stocks of trust, reciprocity, and motivation within the group. The quantitative segment has shown that there are other factors affecting the levels of social capital of GGAVATT participants. For example, when the gender of the advisor, the participants, or both, is female this has a negative effect on the levels of bridging and linki ng social capital. As was already discussed, this might be a reflection of traditional Mexican culture. According to Ramrez (1977) early in her life a Mexican woman learn s to keep to herself and assume a passive role. Additionally, many formal, and informal, Mexican institutions are still heavily domin ated by men and, for some of these men, machismo is still a barrier that imped es interac tion as peers with women (Paz, 1962) Fortunately, this feature of Mexican culture has been slowly decreasing during the last 45 years. Diaz -Guerrero (1975) reported progress on this area almost 35 years ago. However, it is not strange to find men and women that still bear remnants of this form of social construction of gender on their lives. This cultural characteristic is likely to be, at least partially, responsible for the gender differences observed in relation to bridging and linking social capital. However, the collective stories from the qualitative segment did not provide any indication that gender was an issue f or GGAVATT participants. This might be a consequence of the composition of the interview groups which were formed only by men. It is possible that these men either do not consider gender to be an issue in the social dynamics of the group or just simply d o not want to talk about it. Region of the state was also found to have a significant effect on the levels of different forms of social capital for GGAVATT participants in the results from the quantitative segment of the study This might also be due to cultural differences across regions. In terms of agricultural
145 production, people from the North and Central regions of the state are considered more progressive and open to change than people from the South, who are considered more traditionalists. Extr apolating this idea into the realm of this study, it might be valid to presume that these characteristics are likely to affect the levels of social capital of GGAVATT participants. For example, participants from the South would have greater levels of bonding social capital but lack bridging linkages which would be consistent with what the quantitative results have shown. However, the collective stories from the qualitative segment of the study did not show any differences across regions of the state on ho w the access that participants have to different forms of social capital is manifested into their lives. This might be the result of the selection process where leaders were identified by group members to participate in the study and, therefore, they are likely to be progressive as a whole. The quantitative segment also showed that the number of institutional contacts, both previous or current and related or unrelated, are positively associated with greater levels of social capital which shows that connec tedness is a key element in access to social capital. In general, the hierarchical linear models explained a good proportion of the variation at the group level; however, the variation at the individual, or residual, level was poorly explained. This fac t in conjunction with the previously described discrepancies in homogeneity (i.e. how participants self rate their SES level and how they compare their SES to those of the members of their groups) suggest that a completely different set of sociocultural v ariables might be needed to explain the variation at the individual level The qualitative segment of the study showed that participants access different forms of social capital which leads to collective action and empowerment. The GGAVATT groups have p roven to be valuable resources for the communities where they operate. H orizontal connections
146 that are present between GGAVATT groups and other groups in their communities represent an important source of bridging social capital. Local residents access t hese networks in search of information which they later use to transform their communities and, in return, the program has been invigorated through the creation of new groups and the incorporation of new members. This is consistent with Mercoiret et al. (2007) which claim that farmers participating in group extension programs are an important source of technology dissemination to non-participant peers. This has also helped program participants to broaden their social network which is likely to lead to the finding of new opportunities. The participants perceive that the GGAVATT h as endowed them with a sense of belonging and a suppor tive social network which contribute to the development of attachment to the community (Brehm, Eisenhauer, & Krannich, 2004; Lev-Wiesel, 2003) This is particularly relevant in a country like Mexico that has been suffering of a prolonged migration of young rural residents to cities and to the United States of America in search of economic opportun ities. GGAVATT participants identified the importance of having clear rules and norms to guide the participation in group processes. This is a key element of success because clear rules increase transparency and accountability and help reduce conflicts (Abaru, Nyakuni, and Shone 2006). The training that GGAVATT participants obtain within the group context represents another advantage for their communiti es. In the case of the GGAVATT many members and former members have later occupied positions of power at the local level. The leadership and record keeping skills developed in the program have translated into enhanced administrative transparency and accountability at the local level; this effect has been observed in similar programs (Abaru, Nyakuni, and Shone 2006) By m aking available to its participants opportunities to develop and access bonding, bridging, and linking forms of social capital, the
147 GGAVATT program seems to be a viable extension approach to meet the current challenges of rural and agricul tural development. Abaru et al. (2006) say that farmers organizations have the capacity to analyze and solve their own problems and promote local development and democracy. These characteristics are highly desirable when the goal is to have decentralize d pluralistic systems with farmers leading in response to changing market conditions. Based on empirical evidence (Grootaert, Narayan, Nyhan Jones, & Woolcock, 2004; Hu, Huang, Hendrik se, & Xu, 2007; Kubo, 1979; Spellerberg, 2001; Stone, 2001) kinship was expected to affect social capital of groups. However, neither the advisor having family in the group nor the group participants having family in their groups had a significant effec t on social capital levels. It is possible that these variables have a certain type of association (e.g. conditional, spurious, etc.) with other variables; maybe some of the unknown sociocultural variables previously mentioned. Lacking an identification of these sociocultural variables represents a limitation of this study. The findings from this study open the door for diverse reflections on social capital theory. First, t he creation of a m ultiplicative index for social capital combining the levels of t rust, reciprocity, and activity within the network seems to be a useful indicator for explor ing this phenomenon. However, further studies are needed to explore the usefulness of this index in different settings and to determine the specific weight that ea ch elements should have in the equation to increase its accuracy. Next, using the individual as unit of analysis but analyzing the distribution of variation at both the individual and group levels has shown that most of the variation affecting the levels of the three different forms of social capital is found at the individual level. This suggests that social capital is a group attribute (i.e. of the network) that presents an enormous variation at the individual level (e.g. individual attributes, position in
148 network, and density and quality of ties within the network) This is very important because even though the findings at the group level are more useful for policy development, exploring the differences at the individual level will be helpful for fine tuning the policies. Finally returning to the basic element in social capital theory was useful to provide a focus to the study and to reduce the noise of confounding variables. The use of trust, reciprocity, and level of activity within the network (because in this particular case membership in the network was already a given) as the three key components of social capital helped to separate and identify the concept in a clear way making it suitable for analysis. The results have also shown that the l evel of social capital differs across participants and groups In other words, some participants have greater levels of social capital than others. The collective stories illustrated some problems that had occurred with advisors or people representing di fferent institutions, resulting in groups disbanding or never forming. The qualitative data contains additional evidence showing that some participants have abandoned the program for a variety of reasons. These findings suggest that the GGAVATT program does not work the same for everybody and it might be necessary to explore the reasons behind this. Recommendations The GGAVATT needs to decrease that gap between men and female levels of bonding and bridging social capital (Diamond, 2004; Maarse, Wentholt, & Chibudu, 1998) This is very important because every day more farmers in the world are women but they only receive 5% of the extension services worldwide (F.A.O., 2009) The World Bank has found that with the incorporation o f a gender component into development programs the effectiveness is likely to increase in a significant way (World Bank, 2002a) Some strategies to address this issue are: (a) t ake group advisors both men and women to Xalapa to meet the people working inside the governmental institutions and to learn about the different benefits that can be derived from a
149 stronger communication with these individuals. This activity is likely to enhance communication between advisors and governmental institutions and to help female advisors to overcom e traditional cultural barriers; (b) organize fieldtrips for the people working in side governmental institutions in Xalapa to visit with advisors and clientele to have a firsthand observation of their daily work and needs This may allow institutional emp loyees to become more sensitive to the realities of the advisors and their clients which could translate into bringing the voices of these men and women in to the planning process at the institutional level ; (c) e ducate advisors clienteles, and institutio nal contacts about the advantages of connectedness and intercultural communication skills This strategy has the potential to increase the importance that these individuals place on connectedness and therefore increase their level of activity towards that purpose ; (d) in combination with the training in items a -c above, establish women mentoring groups Current female participants and advisors can act as mentors for new female group members and advisors. These mentoring groups can also increase awareness among non participant women about the advantages of participating in a program like this ; and ( e ) d evelop promotional program materials includ ing pictures showing women participating in the program to motivate more women to be involved either as advisors or as participants. All stakeholders need to understand the benefits of connectedness (Dahal & Adhikari, 2008; Diamond, 2004; Larsen et al., 2004; Pretty, 2003) Advisors should try to identify the most important local groups in the communities where their GGAVATTs operate and establish linkages with them to promote communication across different local institutions. T hese other local groups might also receive information from the group advisors or participants on how they can get li nked with governmental institutions.
150 The need for further research on different areas can be established from the results of this study. These areas are: (a) explore socio -cultural variables that may affect the access to social capital; (b) look not onl y at the presence of family, or kin, ties but also at the quality of these ties; (c) explore if group size has an effect on social capital levels; (d) test the multiplicative index used in this study on different setting and with different units of analysi s (e.g. community level and organizational level) ; (e) explore if the levels of social capital of a group are associated with its productivity and adoption of technology; and ( f ) look at groups that have disbanded. Not all GGAVATT groups have been success ful. A large proportion of groups that are formed lack the capacity to survive the initial five years. Other groups that survive this period disband after completing it. If the value of the GGAVATT to the communities has been established, it would be im portant to explore the reasons behind this disintegration of groups to l ook for factors affecting success and develop strategies to alleviate them. There are two general considerations that would be recommended to be applied to conduct research in all the areas mentioned in this paragraph First, the use of mixed -methods because of the richness and quality of information that it provides. And second, the conceptualization and operationalization of social capital should be limited to its essential compone nts (i.e structural and cognitive).
151 APPENDIX A IRB APPROVAL (Back to text )
152 APPENDIX B SURVEY INSTRUMENT FO R GROUP ADVISORS
153 (Back to text )
154 APPENDIX C SURVEY INSTRUMENT FO R GROUP PARTICIPANTS
158 (Back to text )
159 APPENDIX D INFORMED CONSENT FOR GROUP ADVISORS
160 (Back to text )
161 APPENDIX E INFORMED CONSENT FOR GROUP PARTICIPANTS
162 (Back to text )
163 APPENDIX F INFORMED CONSENT FOR SOCIAL INTERVIEWS
164 (Back to text )
165 APPENDIX G SUBJECTIVITY STATEME NT I was born in the largest city of the world, Mexico City. For the first 15 years of my life, contact with rural life was limited to what was possible for me to experience in my trips to the countryside Nevertheless, that minimal expos ure to the influence of the country life was enough to awake in me a deep interest for the quality of life of rural communities which has become a central topic in my study. My first real contact with agriculture and rural lif e took place during my days in veterinary school. When I decided to become a v eterinarian, my objective was to establish a practice for small animals in a city. However, two events influenced the outcome of my professional life after graduation. On the one hand, the curriculum of the v eterinary school in Veracruz, where I had moved during high school, was focused on the study of large animals, particularly bovines. On the other hand, I had experienced so many good things working in rural communities that I became convinced that, as James Herriot has wisely remarked, farmers are truly the salt of the earth. After worki ng many years among farmers, my interest in agriculture has increased. However, the focus of my interest is not anymore in the animal health or productivity but in the social context where agriculture takes place. Inside this framework, my main concern i s to understand how the life of rural communities can be improved in a form that is significant at various levels of evidence. In multiple times, and in different roles and settings, I have interacted with diverse GGAVATTs stakeholders. These experienc es have given me the opportunity to appreciate the program from numerous viewpoints. Most of the group advisors are veterinarians or agronomists, and I believe that they are convinced that the best way to help the farmers is by
166 increasing the productivity of the farms; the researchers share and strengthen this idea of progress based on diffusion of technologies. For them, the GGAVATT also represents an excellent setting to test new ideas under real conditions. I also consider that for the authorities and development agencies, the GGAVATT is an adequate structure, that is already present, to communicate information and channel resources to the farmers. For the farmers, the GGAVATT is a form of organization that belongs to them, an asset that empowers thro ugh the diffusion of information and the communication with peers, institutions, and authorities. It was through dialogue with these individuals that was possible for me to identify the need of studying the GGAVATT from diverse perspective s. (Back to text )
167 LIST OF REFERENCES AACREA. (2006). About the CREA movement. Retrieved 03/15, 2006, from http://www.aacrea.org.ar/genericos/inglesa.htm Abaru, M. B., Nyakuni, A., & Shone, G. (2006). Strengtheni ng farmers organizations: The experience of RELMA and ULAMP Nairobi, Kenya: World Agroforestry Centre Alesina, A. F., & La Ferrara, E. (2000). Who Trusts Others? Journal of Public Economics, 85(2002), 207234. American Statistical Association. (1997). H ow to conduct pretesting. Alexandria, VA: American Statistical Association. Ashraf, N., Bohnet, I., & Piankov, N. (2006). Decomposing trust and trustworthiness. Experimental Economics, 9 (3), 193208. Australian Bureau of Statistics. (2004). Measuring social capital: An Australian framework and indicators Canberra. Balatti, J., & Falk, I. (2002). Socioeconomic Contributions of Adult Learning to Community: A Social Capital Perspective. Adult Education Quarterly, 52(4), 281298. Ban, A. W. v. d., & Hawkins, H S. (1988). Agricultural extension Burnt Mill, Harlow, Essex, England, New York: Longman Scientific & Technical ; Wiley. Belsley, D. A. (1991). A Guide to using the collinearity diagnostics. Computational Economics, 4 (1), 33 50. Bennett, J. (1983). Agric ultural cooperatives in the development process: Perspectives from social science. Studies in Comparative International Development (SCID), 18(1), 3 68. Beugelsdijk, S., & Smulders, S. (2003). Bridging and Bonding Social Capital: which type is good for economic growth? European Regional Science Association. Bhattacharyya, J. (1995). Solidarity and agency: Rethinking community development. Human Organization, 54(1), 60. Blau, P. M. (1960). Structural effects. American Sociological Review, 25(2), 178193. Bla u, P. M. (1977). A macrosociological theory of social structure. The American Journal of Sociology, 83(1), 26 54. Bogdan, R., & Biklen, S. K. (1998). Qualitative research for education: an introduction to theory and methods (3rd ed.). Boston: Allyn and Bac on. Bourdieu, P. (1986). The forms of capital. In J. G. Richardson (Ed.), Handbook of theory and research for the sociology of education (pp. xxiv, 377 p.). New York: Greenwood Press.
168 Braverman, A., Guasch, J. L., Huppi, M., & Pohlmeier, L. (1991). Promoti ng rural cooperatives in developing countries: T he case of subSaharan Africa. Washington, D.C.: World Bank. Brehm, J., Eisenhauer, B., & Krannich, R. (2004). Dimensions of community attachment and their relationship to well being in the amenity rich rural west. Rural sociology, 69(3), 405429. Brennan, M. A. (2004). IFAS community development in search of a common understanding of community. from http://purl.fcla.edu/UF/lib/FY715pdf Brennan, M. A. (2005) The importance of incorporating local culture into community development. EDIS. Florida Cooperative Extension Service, University of Florida. from http://edis.ifas.ufl.edu/pdffiles/FY/FY77300 .pdf C.I.A. (2009). The world f actbook Retrieved March 1, 2009, from https://www.cia.gov/library/publications/the -world -factbook/geos/mx.html Cabero, J., & van Immer zeel, W. (2007). Building learning networks for small -scale farmers: Pachamama Raymi as an innovative knowledge management system. Knowledge Management for Development Journal, 3(2), 52 63. Carmines, E. G., & Zeller, R. A. (1979). Reliability and validity assessment Beverly Hills, Calif.: Sage Publications. Chagoya Fuentes, J. L., Gonzlez, O. O., Gonzlez, O. E., & Gonzlez, B. P. (2002). Evolucin de los parmetros productivos y reproductivos de los hatos de ganado bovino del GGAVATT Tepetzint l a, Veracru z. Paper presented at the XXVI Congreso Nacional de Buiatria, Acapulco, Mexico. Chen, H. T. & Rossi, P. H. (1980). The multi -goal, theory -driven approach to evaluation: A model linking basic and applied social science. Social Forces, 59(1), 106122. Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, N.J.: L. Erlbaum Associates. Coleman, J. S. (1988). Social capital in the creation of human capital. The American Journal of Sociology, 94, 95 120. Colosi, L., & Dun ifon, R. (2006). Whats the difference? post then pre & pre then post. Cornell Cooperative Extension. Cronbach, L. J. (1971). Test validation. In R. L. Thorndike, W. H. Angoff & E. F. Lindquist (Eds.), Educational measurement (2d ed., pp. xx, 768 p.). Washington: American Council on Education. Crotty, M. (1998). The foundations of social research: Meaning and perspective in the research process. London; thousand Oaks, Calif.: Sage Publications.
169 Cuellar, M., Hedlund, H., Mbai, J., & Mwangi, J. (2006). Ph ase I: Impact assessment Stockholm, Sweden: The National Agriculture and Livestock Extension Programme (NALEP). Dahal, G. R., & Adhikari, K. P. (2008). Bridging, linking, and bonding social capital in collective action: The case of Kalahan Forest Reserve in the Philippines (No. 79). Washington, D.C.: International Food Policy Research Institute (IFPRI). Daiute, C., & Lightfoot, C. (2004). Narrative analysis: S tudying the development of individuals in society Thousand Oaks, Calif.: Sage. Davis, J. A. (1971). Elementary survey analysis Englewood Cliffs, N.J.: Prentice Hall. Diamond, J. (2004). Local regeneration initiatives and capacity building: Whose 'capacity' and 'building' for what? Community Dev J, 39 (2), 177189. Daz Guerrero, R. (1975). Psychology of the Mexican: Culture and personality Austin: University of Texas Press. Dillman, D. A. (2000). Mail and internet surveys: The tailored design method (2nd ed.). New York: John Wiley & Sons. Dinham, A. (2005). Empowered or over -powered? The real experien ces of local participation in the UK's New Deal for Communities. Community Dev J, 40 (3), 301312. Dudwick, N., Kuehnast, K., Nyhan Jones, V., & Woolcock, M. (2006). Analyzing social capital in context: A guide to using qualitative methods and data. Washing ton, D.C.: World Bank Institute. Edwards, R. W. (2004). Measuring social capital: An Australian framework and indicators Canberra, Australia: Australian Bureau of Statistics. Ensminger, D., & Sanders, I. T. (1945). What extension is. In E. d. S. Brunner, I. T. Sanders & D. Ensminger (Eds.), Farmers of the world; the development of agricultural extension (pp. 1 7). New York,: Columbia University Press. Espinosa Garcia, J. A., Wiggins, S., Gonzalez Orozco, A. T., & Aguilar Barradas, U. (2004). Economic susta inability at the farm level: a study on family milk production enterprises in Mexico. Tecnica Pecuaria Mexico, 42 (1), 55 70. F.A.O. (2009). Research and extension: A gender perspective. SD dimensions from http://www.fao.org/SD/fsdirect/fbdirect/FSP004.htm F.A.O, & World Bank. (2001). Farming systems and poverty: Improving farmers' livelihoods in a changing world. Rome and Washington D.C.
170 Flora, C. B., & Flora, J. L. (2003). Social capital. In D. L. Brown, L. E. Swanson & A. W. Barton (Eds.), Challenges for rural America in the twenty first century (pp. 214227). University Park: Pennsylvania State University Press. Fox, J. (1996). How does civil society thicken? The political construction of social capital in Mexico. World Development, 24(6), 10891103. Gal indo Gonzlez, G. (2001). Uso de innovaciones en el grupo de ganaderos para la validacin y transferencia de t ecnologa "Joachin", Veracruz, Mexico. Terra, 19 (4), 385-392. Gee, J. P. (2005). An introduction to discourse analysis: theory and method (2nd ed. ). New York: Routledge. Gergen, K. J. (1994). Realities and relationships: S oundings in social construction. Cambridge, Mass.: Harvard University Press. Gergen, K. J. (1999). An invitation to social construction London: Sage. Gergen, K. J. (2001). Social construction in context London ; Thousand Oaks, Calif.: Sage. Gergen, K. J., & Gergen, M. M. (2003). Social construction : a reader London; Thousand Oaks, Calif.: SAGE. Glaeser, E. L., Laibson, D., Scheinkman, J. A., & Soutter, C. L. (1999). What is Soci al Capital? The determinants of trust and trustworthiness. National Bureau of Economic Research Working Paper Series, No. 7216. Gobierno del Estado de Veracruz. (2006). Ve racruz, gobierno electrnico. Retrieved December 10, 2006, from www.veracruz.gob.mx. Godquin, M., & Quisumbing, A. R. (2005). Groups, networks, and social capital in rural Philippine. Paper presented at the International Research Workshop on "Gender and Collective Action", Chiang Mai, Thailand. Gou ldner, A. W. (1960). The norm of reciprocity: A preliminary statement. American Sociological Review, 25, 161178. Granovetter, M. (1973). The strength of weak ties. The American Journal of Sociology, 78(6), 13601380. Granovetter, M. (1983). The strenght of weak ties: A network theory revisited. Sociological Theory, 1 201233. Greene, J., Benjamin, L., & Goodyear, L. (2001). The merits of mixing methods in evaluation. Evaluation, 7(1), 25 44. Grootaert, C. (1999). Social capital, household welfare, and pov erty in Indonesia: Washington: The World Bank.
171 Grootaert, C., Narayan, D., Nyhan Jones, V., & Woolcock, M. (2004). Measuring social capital : an integrated questionnaire Washington, D.C.: World Bank. Guba, E. G., & Lincoln, Y. S. (1986). The countenances of fourth generation evaluation: Description, judgment, and negotiation. In D. S. Cordray & M. W. Lipsey (Eds.), Evaluation studies review a nnual (Vol. 11, pp. 70 88). Newbury Park, CA: Sage Publications. Gubrium, E., & Koro -Ljungberg, M. (2005). Contending with border making in the social constructionist interview. Qualitative Inquiry, 11(5), 689715. Gwynne, M. A. (2003). Applied anthropology: A career -oriented approach. Boston: Allyn and Bacon. Hanifan, L. J. (1916). The rural school community center. An nals of the American Academy of Political and Social Science, 67, 130 138. Hatch, J. A. (2002). Doing qualitative research in education settings Albany: State University of New York Press. Havelock, R. G. (1971). Planning for innovation through disseminat ion and utilization of knowledge Ann Arbor: Center for Research on Utilization of Scientific Knowledge University of Michigan. Herrero, C., Martinez, R., & Villar, A. (2008). Multidimensional welfare: An application to the measurement of human development : BBVA Foundation working paper. Hill, M. A. (1997). SPSS missing value analysis 7.5. Chicago, Ill.: SPSS Inc. Holstein, J. A., & Gubrium, J. F. (2003). Inside interviewing: new lenses, new concerns Thousand Oaks [Calif.]: Sage Publications Homans, G. C. (1958). Social behavior as exchange. The American Journal of Sociology, 63(6), 597606. Homans, G. C. (1974). Social behavior; its elementary forms (Rev. ed.). New York,: Harcourt, Brace. Hu, Y., Huang, Z., Hendrikse, G., & Xu, X. (2007). Organization and strategy of farmer specialized cooperatives in China. In Economics and Management of Networks (pp. 437462). Huppi, M., & Gershon, F. (1990). The role of groups and credit cooperatives in rural lending. The World Bank Research Observer, 5 (2), 187204. INE GI. (2004). Encuesta nacional de empleo: precisiones estadsticas de los indicadores estratgicos. Instituto Nacional de Estadstica, Geografa e Informtica: Mxico, D.F.
172 INIFAP, & SAGARPA. (2001). GGAVATT de lechera familiar "La labor": primera evaluaci n. : INIFAP. Israel, G. D. (1992a). Elaborating program impacts through data analysis (No. PEOD 3). Gainesville, FL: IFAS, University of Florida. Israel, G. D. (1992b). Sampling issues: Nonresponse (No. PEOD 9). Gainesville, FL: IFAS, University of Florida Israel, G. D., & Beaulieu, L. J. (2002). The influence of social capital on test scores: How much do families, communities and schools matter? [Electronic Version]. AAAE Online from http://aaaeonline.ifas.ufl.edu/Regions/Southern/israel2002.htm Israel, G. D., & Beaulieu, L. J. (2004). Investing in communities: Social capital's role in keeping youth in school. Journal of the Community Development Society, 34(2), 35 57. Israel, G. D., Beaulieu, L. J., & Hartless, G. (2001). The influence of family and community social capital on educational achievement. Rural sociology, 66(1), 43 68. Israel, G. D., & Galindo Gonzalez, S. (2008). Using focus group interviews for planning or evaluat ing extension programs [Electronic Version] from http://edis.ifas.ufl.edu/pdffiles/PD/PD03600.pdf Jacobs, J. (1961). The death and life of great American cities New York: Random House. Jones, G. E., & Garforth, C. (1997). The history, development, and future of agricultural extension. In B. E. Swanson, R. P. Bentz & A. J. Sofranko (Eds.), Improving agricultural extension : a reference manual (pp. ix, 220). Rome: FAO. Kelly, K., & Caputo, T. (2 006). Case study of grassroots community development: Sustainable, flexible and cost -effective responses to local needs. Community Dev J, 41 (2), 234245. Kessels, J. W. M., & Poell, R. F. (2004). Andragogy and social capital theory: The implications for human resource development. Advances in Developing Human Resources, 6(2), 146157. Kramer, R. M. (1999). Trust and distrust in organizations: Emerging perspectives, enduring questions. Annual Review of Psychology, 50, 569598. Kr ishna, A., & Shrader, E. (1999). Social capital assessment tool. Paper presented at the Conference on social capital and poverty reduction, Washington, D.C., June 2224. Kubo, Y. (1979). The cooperative farming system in the mixed farming area of Hokkaido, Japan. In J. Wong (Ed.), Group farming in Asia: E xperiences and potentials (pp. 3 -13). Singapore: Singapore University Press. Labov, W. (1972). Language in the inner city : S tudies in the Black English vernacular Philadelphia,: University of Pennsylvania Press.
173 Larsen, L., Harlan, S. L., Bolin, B., Hackett, E. J., Hope, D., Kirby, A., et al. (2004). Bonding and bridging: Understanding the relationship between social capital and civic action. Journal of Planning Education and Research, 24(1), 64 77. Lev Wie sel, R. (2003). Indicators constituting the construct of 'perceived community cohesion'. Community Dev J, 38 (4), 332343. Lincoln, Y. S., & Guba, E. G. (1985). Naturalistic inquiry Newbury Park, CA: Sage Publications. Lucke, J. C., Donald, M., Dower, J., & Raphael, B. (2001). Considerations in the design of a mixed -method cluster evaluation of a community program for 'at risk' young people. Evaluation, 7(1), 110131. Luloff, A. E., & Bridger, J. C. (2003). Community agency and local development. In D. L. Brown & L. E. Swanson (Eds.), Challenges for rural America in the twenty first century (pp. 203213). University Park: Pennsylvania State University Press. Maarse, L., Wentholt, W., & Chibudu, A. (1998). Making change strategies work: G ender sensitive, clie nt oriented livestock extension in Coast Province, Kenya Amsterdam: Royal Tropical Institute. Machiavelli, N. (2003). The prince London; New York: Penguin Books. Masefield, G. B. (1950). A short history of agriculture in the British colonies Oxford,: Cl arendon Press. Masefield, G. B. (1972). A history of the colonial agricultural service Oxford: Clarendon Press. McBride, A. M., Brav, J., Menon, N., & Sherraden, M. (2006). Limitations of civic service: Critical perspectives. Community Dev J, 41 (3), 3073 20. McMillan, J. H., & Schumacher, S. (2006). Research in education: E vidence-based inquiry (6th ed.). Boston: Pearson/Allyn and Bacon. Mercoiret, M. R., Pesche, D., & Bosc, P. M. (2007). Rural Producer Organizations (RPOs) for pro -poor sustainable development (No. 41398). Washington: The World Bank. Moody, M. (2008). Serial reciprocity: A preliminary statement. Sociological Theory, 26(2), 130151. Mora Alfaro, J. (2002). Desarrollo rural, cambio institucional y e xtensin rural en Centroamrica y Mxico. Sa n Jos, C.R.: Instituto Interamericano de Cooperacin para la Agricultura Murray, M. (2003). Narrative psychology and narrative analysis. In P. Camic, J. Rhodes & L. Yardley (Eds.), Qualitative research in psychology (pp. 95 112). Washington, DC: American Psychological Association.
174 Nagel, U. J. (1997). Alternative approaches to organizing extension. In B. E. Swanson, R. P. Bentz & A. J. Sofranko (Eds.), Improving agricultural extension: A reference manual. (pp. ix, 220). Rome: FAO. Narayan, D., & Cassidy, M. F. (2001). A dimensional approach to measuring social capital: Development and validation of a social capital inventory. Current Sociology, 49(2), 59 102. Onyx, J., & Bullen, P. (2000). Measuring social capital in five communities. Journal of Applied Be havioral Science, 36(1), 23 42. Patton, M. Q. (2002). Qualitative research and evaluation methods (3 ed.). Thousand Oaks, Calif.: Sage Publications. Paz, O. (1962). The labyrinth of solitude: life and thought in Mexico. New York,: Grove Press. Peugh, J. L. & Enders, C. K. (2005). Using the SPSS mixed procedure to fit cross -sectional and longitudinal multilevel models. Educational and Psychological Measurement, 65(5), 717741. Pi Feng, H. (1979). An outline of group farming experience in Taiwan. In J. Wong (Ed.), Group farming in Asia: experiences and potentials (pp. 51 65). Singapore: Singapore University Press. Portes, A. (1998). Social capital: Its origins and applications in modern sociology. Annual Review of Sociology, 24(1), 1 24. Pretty, J. (2003). Social capital and connectedness: Issues and implications for agriculture, rural development and natural resource management in ACP countries (No. 8032): ACP EU Technical Centre for Agricultural and Rural Cooperation (CTA). Productivity Commission. (2003). S ocial capital: Reviewing the concept and its policy implications Canberra: AusInfo. Putnam, R. D. (1993). The prosperous community: social capital and public life. The American Prospect, 13, 35 42. Putnam, R. D. (2000). Bowling alone: the collapse and rev ival of American community New York: Simon & Schuster. Putnam, R. D. (2001). Social capital: Measurement and consequences. In OECD & HRDC (Eds.), The contribution of human and social capital to sustained economic growth and well -being, International Symposium Report Qamar, M. K. (2000). Agricultural e xtension at the turn of the millennium: trends and challenges Rome: FAO.
175 Ramrez, S. (1977). El mexicano, psicologa de sus motivaciones (4. ed.). Mxico, D.F.: Editorial Grijalbo. Reed, E. (1979). Two appro aches to cooperation in rice production in South Korea. In J. Wong (Ed.), Group farming in Asia: experiences and potentials (pp. 14 36). Singapore: Singapore University Press. Rivera, W. M. (2008). Agricultural extension system reform and the challenges ahead. Paper presented at the National Federation of Coffee Growers of Colombia 70th anniversary seminar on "Science and Tropical Agriculture for the 21st Century", Bogota, Colombia. 1012 September 2008. Rivera, W. M., & Alex, G. (2004a). The continuing rol e of government in pluralistic Extension systems. Paper presented at the AIAEE 20th Annual Conference, Dublin, Ireland. Rivera, W. M., & Alex, G. (Eds.). (2004b). Volume 1. Decentralized systems: Case studies of international initiatives Washington: The W orld Bank. Rogers, E. M. (2003). Diffusion of innovations (5th ed.). New York: Free Press. Rling, N. G. (1988). Extension science: Information systems in agricultural development Cambridge Cambridgeshire ; New York: Cambridge University Press. Roman Ponc e, H. (2001, 1618 August 2001). Impacto del modelo GGAVATT en la transferencia de tecnologa pecuaria. Paper presented at the AMMVEB, Veracruz, Mexico. Rossi, P. H., Lipsey, M. W., & Freeman, H. E. (2004). Evaluation: A systematic approach (Seventh ed.). Thousand Oaks, CA: Sage. SAGARPA, & INIFAP. (2002). Manual para la formacin de capacitadores: modelo GGAVATT INIFAP: Mexico. Sands, R. (2004). Narrative analysis: A feminist approach. In D. Padgett (Ed.), The qualitative research experience (pp. 48 75). Belmont, C A: Thomson Brooks/Cole. Schafer, J. L., & Graham, J. W. (2002). Missing data: Our view of the state of the art. Psychological Methods, 7(2), 147177. Schuller, T. (2007). Reflections on the use of social capital. Review of Social Economy, 65(1), 11 28. Spellerberg, A. (2001). Framework for the measurement of social capital in New Zealand. Wellington, New Zealand: Statistics New Zealand. Stone, W. (2001). Measuring social capital: Towards a theoretically informed measurements framework for researching social capital in family and community life (No. 24): Australian Institute of Family Studies.
176 Strauss, A. L. (1987). Qualitative analysis for social scientists Cambridge [Cambridgeshire] ; New York: Cambridge University Press. Strauss, A. L., & Corb in, J. M. (1998). Basics of qualitative research: techniques and procedures for developing grounded theory (2nd ed.). Thousand Oaks: Sage Publications. Swanson, B. E. (2008). Global review of good agricultural Extension and advisory service practices Retr ieved December 10, 2008, from http://www.fao.org/docrep/011/i0261e/i0261e00.htm Tabachnick, B. G., & Fidell, L. S. (1996). Using multivariate statistics (3rd ed.). New York, NY: HarperCollin s College Publishers. Tashakkori, A., & Teddlie, C. (2003). Handbook of mixed methods in social & behavioral research Thousand Oaks, Calif.: SAGE Publications. The Saguaro Group. (2000). Bettertogether. The report of the Saguaro seminar: Civic engagement in America. Retrieved January 30, 2007, from http://www.bettertogether.org/bt_report.pdf Uphoff, N. (1996). Collaborations as an alternative to projects: Cornell experience with university NGO g overnment networking. Agriculture and Human Values, 13(2), 42 51. White, K. D. (1970). Roman farming. Ithaca, N.Y.: Cornell University Press. White, K. D. (1977). Country life in classical times London: Paul Elek. Wilkinson, K. P. (1991). The community in rural America New York: Greenwood Press. Willis, G. B. (1999). Cognitive interviewing: A "how to" guide : Short course presented at the annual meeting of the American Statistical Association. Wong, J. (1979). Group farming in Asia: experiences and potenti als Singapore: Singapore University Press. Woolcock, M. (1998). Social capital and economic development: Toward a theoretical synthesis and policy framework. Theory and Society, 27(2), 151208. Woolcock, M. (2001). The place of social capital in understanding social and economic outcomes. ISUMA, 2 (1), 11 17. World Bank. (2001). World development report 20002001: Attacking poverty New York: Oxford University Press. World Bank. (2002a). Integrating gender into the World Bank's work: A strategy for action. Washington. World Bank. (2002b). Mexico. Veracruz -Llave poverty diagnostic (No. 25161ME). Washington: The World Bank.
177 World Bank. (2009). Poverty in Mexico fact sheet. Retrieved March 1, 2009, from http: //go.worldbank.org/MDXERW23U0 Worthen, B. (1990). Program evaluation. In H. Walberg & G. Haertel (Eds.), The international encyclopedia of educational evaluation (pp. 42 47). Toronto, ON: Pergammon Press.
178 BIOGRAPHICAL SKETCH Sebastian Galindo -Gonzal ez received his Ph.D. from the University of Florida (UF) in the summer of 2009. He started his doctorate in 2005, and his areas of research are international extension reform, social capital as an unanticipated consequence of technology diffusion, and the use of mixed -methods for program evaluation. He came to UF in 2002 after graduating in veterinary medicine from the University of Veracruz, in 2001, to obtain a masters degree from the Department of Animal Sciences. Sebastian has authored, or co autho red, multiple peer reviewed publications and even a book chapter on international extension. He has taught graduate courses on program evaluation and international extension systems in the Department of Agricultural Education and Communication of UF. Seb astian became the inaugural recipient of the prestigious Jimmy G. Cheek Graduate Medal of Excellence Award in the spring of 2009 in that same institution.