Open Dissertation Final - Luis Sevilla.Pdf

Open Dissertation Final - Luis Sevilla.Pdf

The Pennsylvania State University The Graduate School College of Agricultural Sciences SOCIAL NETWORKS AND THE EXCHANGE ECONOMY IN RURAL MOZAMBIQUE: A STUDY OF OFF-FARM LABOR AND CROP MARKETING BEHAVIORS A Dissertation in Agricultural, Environmental and Regional Economics & Demography by Luis Sevilla © 2013 Luis Sevilla Submitted in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy December 2013 The dissertation of Luis Sevilla was reviewed and approved* by the following: Jill L. Findeis Distinguished Professor Emeritus of Agricultural, Environmental and Regional Economics & Demography Dissertation Advisor Chair of Committee David Abler Professor of Agricultural, Environmental and Regional Economics & Demography Stephan Goetz Professor of Agricultural, Environmental and Regional Economics & Demography Gary King Professor of Biobehavioral Health Rhonda BeLue Associate Professor of Health Policy and Administration Ann Tickamyer Professor and Head of Department of Agricultural Economics, Sociology, and Education *Signatures on file in the Graduate School. ii ABSTRACT Of the 3 billion living in rural areas in less developed regions of the world, approximately 1.2 billion people live in extreme poverty (The Economist, 2013; World Bank, 2013), and 70% of the 1.2 billion people have some dependency on agriculture (Cleaver, 2012). In sub-Saharan Africa, 47% of the population lives in extreme poverty (United Nations, 2012), 66% of the total population lives in rural areas, and more than 90% depend on agriculture for their livelihoods (Asfaw et al., 2010). Unfortunately, subsistence agriculture operates as a safety net for the poor population rather than as a driver of economic growth (World Bank, 2005). To combat extreme poverty, greater economic growth and income equality will be required (Chandy et al., 2013) and this may be achieved through poverty reduction strategies that target the productivity, profitability, and sustainability of poor farm households (Asfaw et al., 2010). By promoting rural economic growth, to include farm and off-farm opportunities, households can directly benefit from increased food security and incomes (Cord, 2002). Many economic activities in developing countries are influenced by non-market interactions with family, friends, and acquaintances. These mutually beneficial relationships require an investment of limited household resources such as time and money away from productive activities but in the long run may expand household resources (Bolin et al., 2003). Previous studies have found that social networks affect household incomes (di Falco and Bulte, 2011; Haddad and Maluccio, 2003; Narayan and Pritchett, 1999), agricultural technology adoption (Bandiera and Rasul, 2006; Isham, 2002), employment and credit (Munshi, 2011; Wahba and Zenou, 2005), productivity (Fafchamps and Minten, 2002), and risk sharing (Fafchamps, 2011). This dissertation investigates the influence of social networks on economic behavior of agricultural households in rural Mozambique. Specifically, the dissertation has two research objectives: 1) to better understand the effect of social networks on male and female iii labor allocation and off-farm work choices, and 2) to determine if social networks impact agricultural marketing behaviors of rural agricultural households. This dissertation uses an ex ante baseline socioeconomic questionnaire administered to Mozambican households by the Institute for Agricultural Research of Mozambique (IIAM) and Pennsylvania State University (PSU) as part of a multidisciplinary project funded by the McKnight Foundation. The project aims to improve food security and agro-ecosystem sustainability through the development and diffusion of common beans (Phaseolus vulgaris) bred to grow well in low phosphorous soils of Africa. Face-to-face interviews were conducted between August 2008 and August 2009 in eight villages throughout Central and Northern Mozambique. As a baseline study, the interviews provided an initial picture of household composition, labor allocation, agricultural production and technology adoption (including beans), and social networks. To achieve objective 1, a series of models are estimated to better understand the interrelationships between social networks and labor allocation and off-farm work decisions. A bivariate probit model is used to simultaneously estimate labor participation models for adult male and female respondents when both are present in the household. The bivariate probit model tests for jointness in decision-making between adult males and females in dual-headed households. Next, separate univariate probit models are used on the full sample to estimate models of labor participation decisions for adult male and female respondents. To test for the endogeneity of ego networks, a probit model with continuous endogenous regressors (IV probit) is included to model the off-farm labor participation decisions of adult males and females. Lastly, a multivariate probit is used to model off-farm work choices, particularly hired farm labor, non-farm wage labor, and non-farm self-employment. iv In the bivariate probit analysis, the statistical insignificance of the cross equation correlation and the failure to reject the Wald test statistic for the hypothesis that the two equations are independent implies non-jointness in decision-making (at least for this sample). Therefore, the labor participation decision for men and women is estimated with two separate univariate probit models. Results from the univariate probit analysis indicate that men and women rely on different ego network types to access off-farm labor. In Mozambique, having more friends (friendship ego network size) positively affects women’s off-farm labor participation; however, kin ego network size is not a statistically significant predictor of off-farm labor. For men, the opposite is true, friendship ego networks are not statistically significant predictors of off-farm labor but larger kin ego networks positively affect the off-farm labor participation decision. Results show that, regardless of network type, both men and women benefit from knowing more people (total ego network size). However, given a total network size, an increase in the proportion of contacts who are kin has the opposite effects on men and women. For women, a more kin homogenous total ego network reduces the probability of off-farm employment whereas for males, a more kin homogenous total ego network size increases the likelihood of off-farm employment. Additionally, results show that age is a statistically significant predictor of female off-farm labor participation. Also, education in Mozambique seems to result in different outcomes for men and women. More education among women reduces their likelihood of off-farm labor participation but increases men’s probability of working in off-farm labor. The multivariate probit analysis extends the off-farm labor participation model to look at the different off-farm work choices for men and women. Results from the multivariate probit indicate the use of different ego network types to access high and low wage off-farm labor. The v results show that women who are self-employed and/or have higher wage non-farm jobs also have smaller friendship ego networks. For men, friendship ego networks are not statistically significant predictors of the different off-farm work choices. However, having more relatives in the village (kin ego network) increases access to work on others’ farms. Knowing more people (total ego network size) does not result in favorable labor outcomes for men and women; women with larger total ego network are less likely to be self-employed and both men and women with larger total ego networks are more likely to be employed in low wage work on others’ farms. A greater proportion of kin in total ego networks reduces women’s employment in hired farm labor and increases the likelihood of employment in non-farm wage labor. For men, the opposite is true; more kin homogeneous total ego networks positively affects employment on others’ farms as well as non-farm self-employment. In addition to the ego network results, the multivariate probit indicates that increased education reduces women’s dependence on low wage hired farm labor but does not result in increased access to higher wage non-farm labor or self-employment. More than education, women benefit from improved health. Healthier women are less likely to be self-employed and more likely to participate in higher wage non-farm labor. For men, the opposite is true; improved health is not a statistically significant predictor of any of the off-farm work choices, however, increased education positively affects male participation non-farm self- employment. Other notable results include the negative effect of farm size on male participation in hired farm labor as well as the negative effect of distance to the nearest city/town on female participation in non-farm wage labor employment and self-employment. To achieve objective 2, a double hurdle model is used to determine the effect of social networks on the market participation decision and value of sales for agricultural households. Three separate double hurdle models are estimated to study marketing behavior of three crops: vi beans, maize, and other crops. To test for and correct for potential endogeneity, a control function method is used. Interpretation of the results includes marginal effects of the independent variables on the probability of participating in crop

View Full Text

Details

  • File Type
    pdf
  • Upload Time
    -
  • Content Languages
    English
  • Upload User
    Anonymous/Not logged-in
  • File Pages
    225 Page
  • File Size
    -

Download

Channel Download Status
Express Download Enable

Copyright

We respect the copyrights and intellectual property rights of all users. All uploaded documents are either original works of the uploader or authorized works of the rightful owners.

  • Not to be reproduced or distributed without explicit permission.
  • Not used for commercial purposes outside of approved use cases.
  • Not used to infringe on the rights of the original creators.
  • If you believe any content infringes your copyright, please contact us immediately.

Support

For help with questions, suggestions, or problems, please contact us