Abstract: Can own-source revenues (OSR) enable local governments to manage public health when service provision is controlled by higher-level departments? In India, Gram Panchayats (GPs) govern public health, while delivery is undertaken by public healthcare institutions. This separation creates accountability frictions, as public healthcare institutions follow administrative mandates that may not reflect local preferences, making OSR a key discretionary instrument through which GPs can secure responsiveness. We examine whether greater OSR strengthens GP capacity to respond to adverse health shocks, focusing on the second COVID-19 wave in rural Tamil Nadu, when GPs were assigned time-bound mitigation responsibilities without predictable fiscal support. We combine a primary COVID survey of 389 GPs across six districts with administrative and secondary data on health infrastructure, electoral competition, and local public finances. To address endogeneity, we instrument a GP’s OSR using the leave-one-out mean OSR of fiscally and economically similar peer GPs. Higher OSR significantly increases discretionary COVID spending and reallocates expenditures toward public health and sanitation. Effects are larger in electorally competitive GPs, in those closer to Health Sub-Centres, and among pradhans more aware of assigned COVID functions, suggesting significant complementarities with administrative presence and political competition. The results operate partly through enhanced deliberative decision-making, indicating that fiscal autonomy strengthens democratic processes in public health management. Overall, the findings support policies that expand local revenue-raising capacity to improve crisis responsiveness in decentralized systems.
Abstract: This paper examines whether investments in healthcare generate electoral returns for local governments in settings characterised by fragmented service delivery and constrained fiscal autonomy. Using a candidate-term panel spanning five electoral cycles (1994–2015) from rural India, we study how own-source revenue (OSR) generation and sector-specific public expenditure jointly shape the re-contesting decisions of incumbent Gram Panchayat (GP) pradhans. We develop a theoretical model in which an incumbent allocates discretionary revenues across public goods that differ in voter attribution. The model yields three predictions: OSR extraction imposes an electoral penalty; public spending can mitigate this penalty, but only in proportion to its attributability to local action; and weakly-attributed health spending provides limited insulation against this penalty and may even amplify it at high levels of OSR. Empirical estimates from a linearised structural equation closely mirror these predictions. Higher OSR is associated with a significantly reduced probability of re-contesting. Health spending offsets this penalty only at low levels of revenue extraction; at higher OSR levels, the interaction between health expenditure and own revenues turns negative and statistically significant. Employment schemes, which voters more readily attribute to local incumbents, retain positive electoral returns across the OSR distribution. Household survey evidence on attribution corroborates the mechanism: citizens are substantially less likely to credit the Gram Panchayat for health outcomes than for sanitation, roads, or employment provision. These findings highlight how weak attribution and shared accountability undermine political incentives to invest in socially valuable but institutionally fragmented public services.
[Working paper with Pritha Dev and Hari K. Nagarajan] [Current Status: Revision submitted to the Journal of Economic Behavior and Organization]
Abstract: How does physical space influence social network formation? We examine the causal impact of neighborhood composition on caste-based homophily in rural Indian villages. Using a game-theoretic model of sequential household settlement and a new instrumental variable based on historical household arrival patterns, we show that households with more same-caste neighbors are substantially more likely to form caste-homophilous social ties. We develop a dynamic location choice model where households from two castes sequentially settle in a village. The households make their location decisions based on existing households and their expectations about the future entrants. The model shows that with or without households preferring to locate next to other households from their own caste, arrival patterns and group size critically shape settlement outcomes. Next, leveraging nationally representative data on 96,000 households, we find that caste-based homophily is increasing in the number of same caste immediate neighbors even though households rarely form links with neighbors. Our results are robust to alternative definitions of key variables, alternative estimation methodology, alternative sampling techniques and a placebo test. We document interesting heterogeneities of neighborhood composition effects across caste categories, landholding statuses and street-level public good densities. This study highlights how the physical organization of economic space critically shapes patterns of social capital, with implications for the design of inclusive rural policies.
Abstract: Does financial inclusion change the allocation of resources within households, or does it merely expand formal account ownership? We study this question using the exogenous expansion of doorstep banking agents in India, Business Correspondents (BC), following the 2010 Reserve Bank of India circular that relaxed their recruitment and deployment rules. Combining an expanding panel of Indian rural households with the estimates of individual resource shares, we show that BC expansion increased women’s control over household resources. Women in treated households experienced 10.50 percentage point increase in their resource shares relative to women in untreated households. The Shift-Share-DiD estimates show substantially larger local effects of compliers whose financial inclusion was shifted following the BC rollout. Placebo tests based on the national drive for universalization of financial inclusion (PMJDY) and checks against alternative channels indicate that the effect is not attributable to account opening, welfare transfers, or local political reservation. The evidence suggests that financial inclusion improves women’s welfare when policy design lowers the practical and informational costs of using accounts, thereby translating formal access into gains in intra-household agency.
[Current Status: Draft coming soon]
Abstract: Estimates suggest that over a 100 million women are “missing” from the economy due to systematic neglect and discrimination. This paper extends the concern to the intrahousehold domain by asking whether women’s share of household resources tends to decline over time, imposing on them an age penalty. While various initiatives aim to strengthen women’s agency through economic participation, it remains unclear whether these lead to sustained improvements in control over household resources. We examine whether financial inclusion can mitigate this age penalty for women in rural India. Leveraging quasi-exogenous variation in the expansion of banking agents (Business Correspondents) following an RBI directive, we estimate the effects of improved financial access on women’s resource shares. These agents deliver doorstep banking services and reduce gender-specific barriers to account usage. Using a three-wave panel across 30 villages in Uttar Pradesh, we find that financial inclusion raises women’s resource shares, particularly between ages 25 and 40. Early-life access flattens the age-penalty curve, suggesting cumulative benefits. These findings underscore financial inclusion as a policy that arrests the slow disappearance of women from within the household and the larger economy.
[Current Status: Drafting]
[Current Status: Drafting]
[Current Status: Drafting]
[Current Status: Drafting]
