Improving the social welfare of rural households in South Asia and Africa through effective rural development interventions

Abstract

United Nations Sustainable Development Goals recognise rural development in low and middle income countries as critical for reducing global poverty because most of the world’s poorest and marginalised people are largely concentrated in rural areas of South Asia and Africa. Several meta-analyses of evaluations of rural development strategies, investments and policies have concluded that the actual impact of rural development initiatives on the poorest subpopulations remains poorly understood. Further, development economists bemoan the lack of comprehensive and rigorous quantitative evaluations that can adequately inform the process of designing effective rural development policies. This thesis contributes to the literature on quantitative evaluation of various initiatives for improving social welfare outcomes in low and middle income countries in the world’s poorest rural regions. Specifically, this thesis describes four studies evaluating four distinct ex-post and ex-ante past and prospective rural development initiatives and considers a broad set of social welfare outcomes to contribute to effective investments and policies for improving the livelihood of the poorest subpopulation. A unique feature of this thesis is that it enables a comparison between two spatially and socio-culturally disparate contexts in South Asia and Africa, across a broad set of rural social welfare outcomes. The thesis utilised quantitative economic estimation approaches to evaluate various initiatives for improving social welfare outcomes, and was designed to address four key issues emerging from a review of evaluation literature. Specifically, this thesis: 1) compared the performance of large- and small-scale infrastructure investments; 2) assessed the importance of considering potential adverse effects of rural development programs; 3) compared complementary multi-objective program designs with single-objective programs; and 4) compared targeted interventions that consider family structures and gender dynamics with universal rural development initiatives. The first analytical chapter utilised a stochastic benefit cost analysis to estimate the net benefit of a 100 million-dollar (2017 USD) prospective irrigation-expansion investment to support irrigated agriculture in the Nam Ngum River Basin, a tributary of the Mekong, in Lao PDR between 2009 and 2030. The second analytical chapter evaluated the causal influence of microcredit loans on primary school enrolment using quasi-experimental treatment-effects methods based on 2010 Bangladesh census data. The third analytical chapter employed a survey-informed stochastic benefit-cost analysis to estimate the net benefit of incorporating climate-smart technologies to existing livestock donation programs in Rwanda’s Western and Eastern provinces. The fourth analytical chapter mixed-effects generalised linear panel regression models to evaluate the influence of mobile phone use on agricultural outcomes based on 2012 and 2015 surveys of rural households in Bangladesh’s seven major administrative divisions. The thesis study found that: 1) community-scale rural development programs may perform better than large-scale regional rural development schemes; 2) rural development investments and policies should consider impacts on a broad set of cross-sectoral rural social welfare outcomes; 3) complementary multi-objective policy packages that make provisions for foreseeable inadvertent adverse impacts of rural development interventions may perform better than single-objective intervention; and 4) targeted and tailored policy interventions that take into account heterogeneous household characteristics and gender dynamics may perform better than universal interventions.