Horticultural Farming and Development Outcomes: Examining Human and Social Capital Investment Among Horticultural Households in Rural Indonesia

Abstract

Increasing demand for fruits and vegetables among the Indonesian population has transformed the agri-food industry of Indonesia to move beyond traditional staple crop farming. Relative to traditional staple food crops, horticultural crop production usually generates higher returns to land, creates more on-farm and off-farm employment and leads to higher real wages in local economies. This thesis aims to examine human capital and social capital investment among households focusing on horticultural crop production in rural Indonesia. The first analytical chapter aims to examine to what extent households cultivating horticultural crops as their main crops have better child education outcomes than households with traditional staple crops as their main crops. This study assesses the socioeconomic impacts of horticultural farming on household education investment among agricultural households in rural Indonesia. The study utilised a comprehensive household survey from Eastern Indonesia (IFLS EAST 2012) and applied econometric modelling that controls for the possible endogeneity of households’ horticultural farming participation with instrumental variable estimation. The sample in the current study consists of 1,246 children from 791 households. It is found that horticultural crop farming has a positive association with child education spending of the household, especially for boys and primary school children. It is also associated with increased amount of time spent in school for certain subgroups of children, namely girls and junior high school students. The second analytical chapter explores the risk preference among households having horticultural crops as their main crops and how it is associated with diversification toward nonfarm income and education spending. The study conducted an empirical investigation using three-stages least squares regression (3SLS) to estimate simultaneous equations and further employed GMM 3SLS which extends the 3SLS estimator by allowing for heteroskedasticity. Moreover, the study also incorporates gender perspectives as it analyses both husbands' and wives’ risk preference which have heterogeneous effects. The sample in the current study consists of 392 children from 284 horticultural farming households in East Java, Indonesia. The survey and framed risk experiment were conducted in September 2017 among 500 households of citrus cultivators located in Malang, Jembre, and Banyuwangi districts. It is found that wives’ lower risk aversion leads to higher non-farm income which is positively associated with child education spending. Moreover, non-farm income is found to have a larger positive association with education spending than income from citrus farming and other crops. These results, therefore, imply that the risks associated with horticultural farming could influence education spending through increasing income diversification toward non-farm source. The third analytical chapter aims to examine social capital investment among horticultural farmers through understanding the relationship between different types of agricultural social networks and farm productivity through technical efficiency analysis. Social capital accumulation is an important mechanism to overcome production and marketing constraints commonly faced by farmers. The study employed stochastic production frontier (SPF) analysis which accounts for productivity shifts due to induced changes in efficiency. The sample of the current study is 408 small-scale citrus farmers in rural Indonesia from a recent survey in September 2017. The survey covers one of the main citrus growing areas in Indonesia, including Malang, Jembre, and Banyuwangi districts. The findings are consistent with the previous research documenting positive effects of cooperative membership and farmer group membership on technical efficiency among smallholder farmers. However, there is no effect from having direct access to government authority for production-related information. The probit estimation suggests that, unlike cooperatives, farmer groups appeal to farmers regardless of their education and citrus farming experience. These results can, therefore, assist policy and program design to further promote agricultural social networks among rural households that help achieve higher agricultural productivity and rural development outcomes.