Foreign Aid, Domestic Politics, and Time-to-Development: An Analysis of Subnational Aid Efficacy

Focusing on the understudied causes of variation in domestic aid efficacy, this research studies the politics of sub-national development. While allocation and implementation of aid projects is affected by donor governments, multilateral aid agencies, NGOs, and a wide variety of private actors, the domestic political leaders of recipient countries play a particularly important role. Based on work by Dreher et al. (2016), Bueno de Mesquita et al. (2003), and others, this research develops a theoretical claim for how electoral incentives make office-seeking leaders engineer higher levels of aid effectiveness for the specific constituencies they rely on for political survival. Empirically, this argument is tested with the help of a mixedmethods approach that combines statistical analyses of geocoded sub-national development data with evidence from a comparative case study of Indian states.

Foreign Aid, Domestic Politics, and Time-to-Development: An Analysis of Sub-national Aid Efficacy Minta Siripong, Department of Biology Tobias Hofmann, Department of Political Science This research has been supported by: Abstract Research Question Focusing on the understudied causes of variation in domestic aid efficacy, this research studies the politics of sub-national development. "The West spent $2.3 trillion on foreign aid over the last five decades and still had not managed to get twelve-cent medicines to children to prevent half of all malaria deaths." - William Easterly, The White Man's Burden, 2006 While allocation and implementation of aid projects is affected by donor governments, multilateral aid agencies, NGOs, and a wide variety of private actors, the domestic political leaders of recipient countries play a particularly important role. Based on work by Dreher et al. (2016), Bueno de Mesquita et al. (2003), and others, this research develops a theoretical claim for how electoral incentives make office-seeking leaders engineer higher levels of aid effectiveness for the specific constituencies they rely on for political survival. Empirically, this argument is tested with the help of a mixedmethods approach that combines statistical analyses of geocoded sub-national development data with evidence from a comparative case study of Indian states. Measuring Development Motivation: Why is it that in spite of the large amounts of money, time, and effort spent on improving conditions in developing countries, most studies on foreign aid have only found modest effects in addition to significant variation? Research questions: • Does foreign aid promote development? 2. % of children aged 6-8 attending school 3. male - female % of children attending school 4. infant & child mortality rates; % of immunized children 5. fertility rate of adolescent women 6. n.a. 7. % of households with piped water & flushed toilet 8. % of households with phone, cellphone, & internet access Argument Research Design Literature: • "Aid is not just a one-size-fits-all form of assistance" (Pritchett & Woolcock, 2004) Time period: 25 years, 1991-2016 Units of analysis: 36 administrative states and union territories • "When given the freedom to freely allocate aid, political leaders of recipient countries may partake in favoritism of certain regions […] or use aid to buy votes or reward government supporters" (Dreher et al., 2016) • Geocoded disbursements for all IBRD and IDA World Bank projects (AidData, 2017) • Does partisan alignment between the national government and state-level representatives improve the effectiveness of regional World Bank projects and attainment of UN Millennium Development Goals (MDG)? Hypothesis: Controlling for the initial level of development, aid leads to faster development in districts aligned with the national government. In other words, regional aid efficacy is mediated by partisan alignment. Comparative Case Study • "If Kerala sends BJP MPs to Lok Sabha and if the NDA retained power, we will make Kerala the number one state in India" (Times of India, 2019) Mixed-methods research design that combines statistical analyses with a comparative case study of Indian states. • "Aid capture […] make development difficult in recipients of poor governance" (Dalgaard & Erikson, 2009) • Do the electoral incentives of democratic political leaders interfere with development and affect aid efficacy? • Why do some regions develop faster than others? • "Jharkhand is blessed with natural resources, this land has abundant coal and iron ore, but still the people here are poor […] Development of Jharkhand is only possible when there is a BJP government in the state" (Asian News International, 2014) 1. % of underweight children; % households with IWI < 35 d Political leaders can not only influence aid allocation, but have incentives to affect sub-national aid effectiveness via administrative project support, matching funds, etc., to ‘buy' the ongoing support of their electoral winning coalition. Amit Shah, President of the Bharatiya Janata Party (BJP): Dependent variables are inspired by the UN MDG: Minta Siripong University of Utah Department of Biology [e-mail] Data and sources: • Sub-national development indicators (Global Data Lab, 2019) • National and state-level results of 10th-16th Lok Sabha elections (Election Commission of India, 2018) Methods: • Quantitative: various multiple regression techniques, incl. fixed-effects instrumental variable regression • Qualitative: comparative case study of Kerala and Jharkhand Quantitative Analysis Aid, partisan alignment, and MDG attainment: MDG: Aid Alignment Aid * Alignment MDGt-1 Controls Fixed-effects Constant Observations R2 #1 #2 #4 #5 -0.001* 0.002** -0.005* 0.000 (0.001) (0.001) (0.002) (0.001) -1.607** 2.323** -4.515** 0.205 (0.696) (1.185) (1.873) (1.262) -0.001** 0.001** -0.002* 0.000 (0.000) (0.001) (0.002) (0.001) 0.965*** 0.946*** 0.872*** 0.968*** (0.020) (0.020) (0.055) (0.023) yes yes yes yes yes yes yes yes 0.246 -1.250 1.626 -5.283*** (0.873) (0.986) (2.309) (0.916) 525 525 525 525 0.952 0.926 0.947 0.980 Robust standard errors in parentheses. *p < .1. **p < .05. ***p < .01. Summary: The combined evidence from the statistical analysis and comparative case study highlight the domestic politics of aid efficacy (except for fertility) as the effect of aid on development is conditioned by the partisan alignment of national government and local members of the Lok Sabha. Outlook Next steps: Collection and analysis of more and better data, incl. data for other countries and additional sub-national development indicators. Follow-up UROP project: Closer look at the dynamics between voters and political leaders with respect to local aid projects and regional development (Findley et al., 2016): • Lab experiment at the U: Analysis of the incentives for differentiated development policies under different electoral institutions and responses to informational treatments about government behavior. • Online survey experiment via MTurk: Analyzes electoral responsiveness of Indian voters to information about government support for local aid projects. References: • Bueno de Mesquita et al. 2003. The Logic of Political Survival. Cambridge, MA: MIT Press. • Dalgaard, C. J., and L. Erickson. 2009. Reasonable Expectations and the First Millennium Development Goal. World Development 37: 1170-1181. • Dreher et al. 2016. Aid on Demand. AidData Working Paper 3: 1-66. • Findley et al. 2016. Who Controls Foreign Aid? International Organization 71: 633-663. • Pritchett, L., and M. Woolcock. 2004. Solutions when the Solution is the Problem. World Development 32: 191-212.