Rishabh Sinha

Working Papers

In situ structural transformation and women's autonomy: Evidence from rural Bangladesh

This paper investigates how in situ structural transformation, where men transition out of agriculture but remain in their communities, affects norm-based restrictions on women's autonomy. Using panel data from rural Bangladesh and a Double/Debiased Machine Learning approach, I isolate the effect of male occupational transitions on six indicators of female seclusion. I develop a theoretical framework modeling these restrictions as an equilibrium between the income-driven demand for patriarchal status and the rising efficiency costs of enforcement. Empirically, I find modest and statistically insignificant aggregate relaxations across most restrictions. However, these averages mask substantial heterogeneity: restrictions tighten among small landholding households but relax significantly for wealthier households, with effects also varying by household gender composition. Ultimately, male-biased economic shocks do not uniformly liberalize or tighten gender norms. Instead, changes depend on the specific enforcement costs and status-generating properties of individual restrictions.

[Paper]

The effects of simultaneous versus sequential multiple health behavior change on subjective well-being in Mexico

Public health strategies often advocate for the simultaneous adoption of multiple health behaviors. However, this “bundled” approach may overwhelm the self-regulatory capacity of individuals facing high socioeconomic strain. This study aimed to causally estimate the impact of simultaneous versus sequential cessation of smoking and alcohol consumption on subjective well-being (SWB), and to evaluate how socioeconomic factors moderate these effects and underlying mechanisms in a developing-world context. This study used data from the Mexican Family Life Survey across three waves (2002–2012). A sample-restricted difference-in-differences strategy was employed to eliminate biases inherent in standard two-way fixed effects models under staggered adoption. From an inverse probability-weighted complete-case sample, 222 individuals were isolated into mutually exclusive cohorts: controls (N=153), simultaneous adopters (N=33), and sequential adopters (N=36). The analysis estimated dynamic treatment effects on 20 self-reported indices of psychological distress, evaluated physical activity to test for behavioral substitution, and stratified models by urban/rural residency, income, and gender. Simultaneous adoption generated a reduction in psychological distress at wave 2. For example, simultaneous adopters were 36 percentage points (pp) (p<0.05) less likely to feel angry or irritated, and 26 pp (p<0.05) less likely to fear something serious happening. However, these effects dissipated by wave 3. Mechanism evaluation showed no evidence that offsetting negative behavioral substitution (lower physical activity) was required to affect this change. Stratified analyses revealed the transient benefits of simultaneous adoption were driven almost entirely by high-income rural individuals. Conversely, simultaneous adoption yielded null effects for urban and low-income populations, and exacted an ego-depletion penalty among rural women (e.g., 86 pp (p<0.01) more likely to feel pessimistic). For these vulnerable groups, sequential adoption was the only pathway to achieving delayed SWB dividends at wave 3.

[Paper] [Supplementary Appendix]

Borrowing to build: Foreign technology licensing and domestic R&D in a developing economy

While foreign technology licensing is widely promoted to accelerate growth in developing economies, theoretical ambiguity persists regarding whether it substitutes for or complements domestic R&D. This study addresses this tension by examining the impact of licensing on internal innovation efforts. Utilizing establishment-level data from Vietnam, I employ propensity score matching with machine learning estimators to mitigate selection bias and isolate the effect of external technology licensing. The analysis reveals that licensing foreign technology increases the probability of domestic R&D activity by approximately 15 percentage points. To explain this complementarity, I build on the absorptive capacity framework, demonstrating that firms must actively adapt imported technologies to overcome local "resource mismatches" (e.g., capital and infrastructure constraints) and "product mismatches" (divergent consumer preferences). There exists strong empirical support for the product mismatch mechanism, as licensing establishments serving local markets exhibit significantly higher R&D engagement than those operating in international markets. These findings challenge the traditional view of licensing as a passive shortcut to technological advancement, revealing it instead as a catalyst for active innovation. Policymakers in developing economies, as such, should pivot away from protectionist restrictions and instead provide targeted incentives that support the adaptation of foreign technologies to local constraints.

[Paper]

Foreign direct investment and poverty: A review of theory and evidence (prepared for Palgrave Encyclopedia of Global Poverty)

Foreign Direct Investment (FDI) has emerged as a preeminent source of external finance for developing economies. Yet, its capacity to alleviate poverty remains the subject of intense theoretical debate and empirical contradiction. This review synthesizes the theoretical foundations of the FDI-poverty nexus, contrasting classical growth-mediated mechanisms and direct resource transfers against countervailing forces of labor displacement, skill-biased technological change, and exclusionary privatization. A comprehensive survey of the empirical literature reveals a profoundly heterogeneous landscape, demonstrating that the measured impact of FDI—ranging from poverty-reducing to immiserating—is acutely sensitive to the specific metrics utilized to operationalize poverty, as well as to geographical focus. To reconcile these divergent findings, the review examines the critical role of host-country contextual determinants, arguing that FDI functions less as an autonomous driver of poverty reduction and more as a catalyst whose efficacy is fundamentally constrained by a nation’s baseline economic development, absorptive capacity, financial depth, and institutional quality. Finally, it discusses actionable policy implications, concluding that maximizing the pro-poor potential of FDI requires abandoning monolithic liberalization paradigms in favor of targeted institutional frameworks designed to cultivate specific transmission channels and mitigate the inherent distributional trade-offs of global economic integration.

[Paper]

Evaluating predictors of medium-term job growth using machine learning

Using data from 116 countries spanning 2005-2023, this paper provides a global comparative assessment of the factors predicting medium-term job growth. I assemble 25 macroeconomic, demographic, health, and institutional variables and employ tuned machine-learning models to evaluate their predictive power. Gradient Boosting Regression delivers the strongest performance, and Shapley decompositions reveal three main results. First, GDP growth is not the dominant predictor of employment expansion, and its predictive strength has weakened over time, except during global recoveries. Second, past job growth is one of the most consistent predictors across income groups, indicating strong labor-market momentum, especially following global macroeconomic instability. Third, some structural forces, such as urbanization and quality of governance, contribute to job growth. But their effects shift with global macroeconomic conditions and a country's level of development. Overall, the evidence underscores the role of these contextual factors, suggesting that economic growth is often insufficient to sustain job creation.

[Paper]

When men leave agriculture: Structural transformation and domestic abuse in rural Bangladesh

Structural transformation (ST), the shift of workers from agriculture to non-agricultural occupations, is central to economic development, but its implications for women’s safety remain poorly understood. This study examines whether men’s exit from agriculture in rural Bangladesh alters the risk of physical and verbal abuse against women. Around one in every five households in the analytical sample exited agriculture. ST increased the probability of physical abuse by 2.94 percentage points (p < 0.10), a 52 percent rise relative to unexposed women. The effect on the incidence of verbal abuse was relatively modest and statistically insignificant. Subgroup analyses suggest women’s bargaining power, particularly in decisions over their own employment, may mediate the link between household occupational transitions and abuse. While ST is widely associated with socioeconomic gains, this study highlights risks to women’s safety that may emerge alongside these broader benefits. Economic progress and education may prove insufficient in reducing domestic violence. Policies must ensure women share in the benefits of transformation by expanding access to employment and strengthening their influence within households.

[Paper] [Supplementary Files]

Regulation and rivalry: How price controls reshape formal-informal competition in Central Asia

This paper examines how price regulation of formal firms affects competition from informal providers, an underexplored channel in the literature on regulation in developing economies. Using establishment-level data from five Central Asian countries and machine learning-augmented propensity score matching, I estimate the impact of price controls on perceived informal competition. Consistent with the segmented market view, regulation has no significant effects in Tajikistan, Turkmenistan, or Uzbekistan. In contrast, regulation significantly reduces informal competition in Kazakhstan and the Kyrgyz Republic, supporting the overlapping market hypothesis but with an effect that qualitatively contradicts standard predictions. Heterogeneity analyses show that these effects are most pronounced in industries with high substitutability and markets without state-owned enterprises. These findings contribute new evidence to the segmented-versus-overlapping markets debate and highlight the need to refine overlapping market frameworks to account for the variations in how price regulation influences informal–formal dynamics. From a policy perspective, the results underscore how regulatory design influences the relative competitiveness of formal firms in different contexts: an issue central to decent work, fiscal capacity, and long-term growth.

[Paper] [Executive Summary]

Partisanship turning counterintuitive: The role of selective enforcement

This paper examines the causal link between political position-taking and social distancing compliance during the COVID-19 pandemic, with India as the focal case study. While existing research has identified strong correlations between partisan affiliation and adherence to health directives, the potential for reverse causality—where political leaders align their stance with their base's preexisting behaviors—complicates these findings. India offers a unique context for analysis, as Prime Minister Narendra Modi's early advocacy for social distancing occurred before any observable partisan gap in compliance. Using an event study approach, I examine variations in social distancing practices across Indian states and union territories, grouped by political alignment with Modi. Contrary to expectations, regions aligned with Modi did not follow social distancing guidelines more rigorously than non-aligned regions, with non-aligned areas seeing a more pronounced increase in time spent at home. The paper proposes a theory of selective enforcement as a potential explanation for this counterintuitive outcome, wherein supporters of Modi in regions with lower perceived benefits from social distancing were less likely to comply due to partisan-driven leniency. These findings highlight the context-dependent nature of political influence on public health behaviors, challenging the generalizability of partisan effects observed in other settings.

[Paper]

What explains Latin America's low share of industrial employment?

This paper investigates the determinants of low industrial employment shares in Latin America. The analysis broadens the scope of the investigation beyond the established focus on comparative advantage and international trade by examining the role of labor market distortions, sectoral input-output linkages, and factor intensities in production technology. The cross-country differences in industrial employment are assessed using a transparent open-economy accounting framework followed by a multi-sector equilibrium model of structural transformation with trade. The findings from the former indicate that labor market distortions, rather than trade shares, are central to explaining why Brazil and Mexico allocate a smaller proportion of their workforce to industry. The equilibrium model corroborates these results, revealing that these countries experience further contraction in industrial employment shares when their productivity profiles are counterfactually changed to match those of more highly industrialized nations. This result arises because the two countries exhibit a comparative advantage in industrial production, contrary to the assumption implicit in the trade-based explanation. Further experiments suggest that, without substantial sectoral imbalances in productivity growth, these economies are unlikely to see significant expansion in their industrial bases as GDP per capita increases. These findings are likely to extend to other Latin American economies. Therefore, policymakers should prioritize resolving labor market distortions to promote industrialization.

[Paper] [Short Slides]

When political instability stifles growth in stable regimes: Insights from India's Emergency

Political economy theory posits that political instability can impede economic growth. Existing empirical investigations of this mechanism often treat regime instability and within-regime instability as independent dimensions, thereby overlooking their interrelated nature. This study addresses this identification problem by analyzing the economic repercussions of political instability within stable regimes. During India's 1975 Emergency, one dimension of instability was dormant, allowing for causal identification. Estimates derived from the synthetic control method indicate that the Emergency resulted in a substantial and sustained adverse impact on India's economy. Absent the Emergency, the GDP per capita ten years after its imposition would have been more than 70 percent higher. Falsification tests confirm the specificity of these effects on the Emergency. Damaging effects extended to other developmental metrics. For instance, life expectancy in 1985 would have been four months higher absent the Emergency. These findings challenge the prevailing notions that political instability within stable regimes yields negligible economic impact.

[Paper]

Distortions in intermediate markets and structural transformation in Latin America

The paper evaluates the concern of low industrial employment share in Latin America in favor of growth in the services share. A standard model of structural transformation with inter-sectoral linkages is extended to allow for the substitution of inputs sourced from different sectors. Barring a couple of instances, the elasticity of substitution is estimated to be less than one across sectors and countries. Sector-biased productivity growth and distortions in intermediate and labor markets determine the allocation of resources. Distortions in intermediate markets play a quantitatively vital role in determining the sectoral allocation of labor. Changes in distortions, particularly in the use of services inputs, can raise industrial shares and bring them closer to the peaks achieved by other countries in the past.

[Paper]

Closer, but no cigar: Intergenerational mobility across caste groups in India

This paper compares the intergenerational mobility of educational and occupational attainment of men from disadvantaged groups (Scheduled Castes (SC) and Scheduled Tribes (ST)) in India with the intergenerational mobility of men outside these groups during 1983-2009. Although there has been a modest convergence in mobility rates of non-SC/ST and SC/ST men in educational attainment, there has been no significant convergence in the mobility rates of occupational attainment. Upward mobility of SC/ST men remains much lower compared to non-SC/ST men. Additionally, the former are more susceptible than the latter to moving down the intergenerational ladder. The mobility gap varies over a large range across states, but the cross-state variation has declined, with convergence being higher in states with larger gaps initially. The paper finds no evidence of higher convergence in states with higher economic growth. As such, policies that focus on growth may not necessarily deliver convergence in outcomes across social groups.

[Paper]

Intergenerational occupational mobility and labor productivity

The paper documents that intergenerational occupational persistence is significantly higher in poor countries even after controlling for cross-country differences in occupational structures. I posit that high occupational persistence in poor countries is symptomatic of underlying talent misallocation. Constraints on education financing forces sons to choose fathers' occupations over the occupations of their comparative advantage. A version of Roy (1951) model of occupational choice is developed to quantify the impact of occupational misallocation on aggregate productivity. I find that output per worker drops by a factor of three relative to the benchmark US economy for the country with the highest level of occupational persistence.

[Paper]

Google Sites

Report abuse