Research

Publications

  1. Hossain, M. A., Mahajan, K., & Sekhri, S. (2022). Access to toilets and violence against women. Journal of Environmental Economics and Management, Available online 11 June 2022, 102695.

  2. Sekhri, S, Hossain M.A., & Khosla, Pooja (2022). Access to Colleges, Human Capital, and Empowerment of Women. Journal of Human Resources 0620-10948R4; published ahead of print March 9, 2022, doi:10.3368/jhr.0620-10948R4

  3. Hossain, M., Malek, M. A., Hossain, M. A., Reza, M. H., & Ahmed, M. S. (2019). Agricultural microcredit for tenant farmers: Evidence from a field experiment in Bangladesh. American Journal of Agricultural Economics, 101(3), 692-709.

Revise & Resubmit

1. Unintended Consequences of a Well-Intentioned Policy: Impact of Credit on Child Labor in Bangladesh (R&R at Journal of Human Resources)

Canonical models of credit suggest that relaxing credit constraints can increase human capital investment in children. However, when credit brings new business opportunities within reach in economies with many labor market frictions, increased access to credit might increase the use of child labor by increasing the opportunity cost of a child's time. In this study, I use data from a randomized controlled experiment to examine the effect of an agricultural credit expansion program in Bangladesh and find an increase in child labor. I present evidence that this increase in child labor is due to a rise in new opportunities for children to work in household self-employment activities. I also find that treated households with fewer working adults use more child labor and spend less on education. While I do not see any effect on schooling outcomes, the time budget survey reveals that children from treated areas spend significantly less time studying. Overall, these findings raise concerns about the unintended inter-generational consequences of easing credit constraints to increase self-employment. (Link to the full paper)

2. Water in Scarcity, Women in Peril (with Sheetal Sekhri) (R&R at The Journal of the Association of Environmental and Resource Economists )

We establish causal effects of groundwater scarcity on sexual violence against women using nationally representative data from India. Negative shocks, measured as the year-to-year variation in subsurface water from the long-term district-specific mean, result in an increase in reported rapes. We bolster our identification by providing insights from geology that explain why annual fluctuations in subsurface water are largely geogenic, and hence exogenous. Our empirical findings reveal that these negative shocks increase the time required for women to collect water from outside the house, thus exposing them to a heightened risk of violent attacks. We show evidence that makes alternative explanations untenable. (Link to the full paper)

Working Papers

1. Performance Gains from Gender Match in Higher Education: Evidence from a Setting with Entrenched Gender Stereotypes

Worldwide, women are severely underrepresented in science, technology, engineering, and mathematics (STEM) education, developing countries in particular. This study sheds light on whether female college students in a male-dominated STEM course like economics benefit academically from being taught by female instructors in a setting where gender stereotypes are an entrenched social phenomenon. I use a novel and confidential administrative dataset from a renowned economics program in Bangladesh to show that when matched with female teachers, female students gain in terms of grade performance, as well as longer-term outcomes such as degree completion time and the likelihood of enrolling into an economics master’s program. The quasi-random allocation of students to mandatory courses, with no scope for students to select the courses or instructors, addresses the self-selection concerns. Comparing the test scores of the blindly and non-blindly graded exams for the same course, I rule out the explanation that the gain from gender-matching is driven by gender preference in teachers' assessment. I find that the benefit of matching increases with female teachers' rank, experience, and academic qualifications. I show suggestive evidence that female teachers' effectiveness in teaching female students is an important channel, aside from the oft-cited role model effect, through which same-sex teacher assignments improve female students' academic achievements. (Link to the full paper)

2. Political Affiliation and Conflicts: Evidence from Nigeria (with Marup Hossain) (Under Review)

Does the political affiliation of local politicians affect organized violence? Applying a regression discontinuity design to closely contested constituencies in the 2011 Nigerian House of Representatives election, we show that number of violence and resulting deaths or human displacements are lower in areas where the ruling-party candidates win rather than losing a close election. The effect of political affiliation on violence is magnified when the local constituencies do not have onshore petroleum deposits, are further away from the nearest big city, or the state governor is from the opposition parties. We provide suggestive evidence that political affiliation affects violence through the increased allocation of security-related funds from the federal government to the local governments or improvement in local economic opportunities. (Link to the full paper)

3. Reinforcing Inequality: Consequences of Elevated Fluoride Exposure and Inequitable Mitigation (with Sheetal Sekhri, Emily Gonzalez, Rajiv Gupta, and Eric Robertson) (Under Review)

We establish causal links between elevated fluoride exposure in drinking water, cognition, and health of children. Chronic elevated exposure is deleterious for children, and severely affects poor children's human capital. At the same time, high exposure affects poor individuals' time spent on economic activities, household chores, and water collection. State institutions fail to implement adequate mitigation measures, and there is little evidence of mitigation by households. We observe minimal adaptation overall but none by the poor or the sick households. Environmental exposure leads to a self-reinforcing cycle of poverty: exposure affects cognition and health of children leading to adverse inter-generational consequences, and the poor are not able to offset the risk, depressing economic mobility and perpetuating inequality. (Link to the full paper)

4. COVID-19 and Gender Gap in Labor Market: Evidence from Nigeria (with Marup Hossain) (Under Review)

Abstract: The COVID-19 pandemic-driven economic downturn can have substantial implications for the gender gap in the labor market in developing countries, where women are already worse-off in job participation and earnings than men. Using multiple rounds of individual-level survey data before and after the pandemic and incorporating a difference-in-differences design, we show that overall employment has reduced more for women than men in Nigeria. Women also experienced a larger shift from business employment to farm-based employment, which may further aggravate women's economic condition to the extent the labor market returns in farming activities are lower than that of business activities. (Link to the full paper)

Works in Progress

  1. Labor Market Effects of Public Sector Salary Increase: Evidence from a Natural Experiment (Presentation Slides)

  2. Credit Access and Women's Bargaining Power within the Household: Evidence from a Randomized Controlled Trial (with Mohammad Abdul Malek)

Abstract: Using a randomized field experiment conducted in Bangladesh, we show that microfinance enables the women from the treated areas to participate more in self-employment activities and earn more compared to the women from the control households. While we do not find any effect of microcredit on average women’s bargaining power within the household, we detect significant heterogeneity. The effect of microcredit is significantly higher for households with more female labor force participation, or with more female self-employment income, or with a more educated female family member. We develop a simple model of intra-household bargaining power that supports our empirical findings. Finally, we show that women’s higher bargaining power leads to better child developmental outcomes, reinforcing earlier findings from the development literature that, the bargaining process often leads to better outcomes for the children, such as education, and health, when interventions are targeted towards women.(Link to the full draft)

  1. Exposure to Elevated Fluoride, Cognition, and Behavior (with Sheetal Sekhri, Emily Gonzalez, Thomas Kishore, Rajiv Gupta, and Eric Robertson)

Abstract: Excessive fluoride in groundwater (exceeding the WHO’s safe limit of 1.5 mg/L) is endemic in 25 countries across the globe. The neurotoxicity of fluoride at higher levels of consumption is not well-established. In this study, we use a quasi-experimental approach to examine the consequences of exposure to elevated levels of fluoride on cognition, domains of intelligence, and behavioral development of children. We collected an array of water and household characteristics data on individuals from 815 households in 275 villages across eight districts of Rajasthan, India. We used the natural variation of geogenic origin in fluoride levels in drinking water and a regression discontinuity design around the safe limit to examine the causal effects of elevated fluoride exposure. We find that children exposed to elevated levels of fluoride higher than 1.5 mg/L had worse outcomes than children exposed to lower levels. Intelligence, as measured by the normed average of three sub-scales of Malin’s Intelligence Scale for Indian children (MISIC), was 4.5 points or 5.9 % lower. Right-handed grip strength and dexterity (measured by assembly score of Purdue pegboard) were inversely related to elevated fluoride exposure. Finally, we also discerned a statistically significant relationship between fluoride and conduct problems. (Draft available on request)

  1. Effects of Physician Supply on Health Outcomes: Evidence from a Natural Experiment in Rural Bangladesh (with Nirman Saha)

Abstract: While there has been a marked improvement in the health outcomes of the general population in developing countries in recent years, meeting people's health needs in remote and rural areas remains a key challenge. The shortage of qualified physicians in rural areas, a common feature in most low- and middle-income countries, has often been touted as one of the most important underlying factors. In this paper, we assess whether an increased supply of physicians in rural areas improves health outcomes. Under a unique intervention commenced in Bangladesh in 2013, the government recruited around 6,000 physicians through a highly qualified exam-- an increase of more than 25 percent of physicians' existing stocks. We use the variation generated by this program, one of the most massive such interventions globally, to estimate the effects of increased physician's supply on local general health outcomes and health care access. Since the increase in the number of physicians in an area is endogenous, we employ an instrumental variable approach to estimate the causal effects. We find an improvement in several health outcomes and other health care access proxies due to the intervention. However, the outcomes could be much better had there been a more equitable distribution of the newly recruited physicians.

  1. Impact of Police Emergency Response System on Crimes and Accidental Deaths: Evidence from the `Dial 100' Project in India (with Sandip Sukhtankar)

Abstract: In 2015, the state of Madhya Pradesh (MP), India, launched a statewide police emergency response system under the name “Dial 100” to provide quick, efficient, and quality service to people requiring any police help in emergency conditions. We employ a difference-in-difference approach to identify the causal impact of this intervention on different crimes, by comparing the incidence of crimes in the boundary districts of MP with the districts of the neighboring states that share a boundary with MP before and after the intervention. After controlling district and year fixed effect, we find that the intervention significantly reduced different types of crimes such as murder, robbery, and burglary. We, however, do not find any significant of the intervention on number of births and deaths registered.

  1. Shocks and Brave Farmers: Evidence from a Randomized Agricultural Microfinance Experiment in Bangladesh (with Yasuyuki Sawada, Mohammad Abdul Malek, & Minhaj Mahmood)

Abstract: We examine the role of an agricultural microcredit program in facilitating farmers’ risk coping strategies when experiencing various shocks. We construct a simple two-period model of technology adoption to show consumption credit as an insurance mechanism. The standard balancing tests show that the treatments and shocks are both exogenously given and, thus, our regression results can be interpreted as causal relationships. Shocks considerably increase credit uptake among the treatment groups. We observe a significant impact of microcredit on outcome indicators, such as, credit use for farming, land use under share cropping and leasing condition, technology adoption, shifting day laboring to self-farming, and increasing farming income. (Draft available on request)

  1. Credit Crunch and Human Capital: Evidence from Micro-Finance Regulation in India (with Yutong Chen and Sheetal Sekhri)

In 2010, an ordinance by the state of Andhra Pradesh, India, brought to a complete halt of operations of all the private microfinance organizations in the state, adversely affecting their loan recovery and liquidity. We exploit the variation in the credit supply generated by this natural experiment to explore the effect of a reduction in credit supply on children's human capital attainment and labor supply. We use a synthetic Difference-in-difference strategy to show that a reduction in microcredit access lowered children's school enrollment by six percentage points. We also observe significant heterogeneity in the treatment effect -- Primary, Middle, or Secondary schooling are negatively affected, whereas Higher Secondary Schooling is significantly positively affected. The likelihood that children participate in self-employment activities also falls significantly. While the overall likelihood of children's participation in wage-earning activities is not affected, there is heterogeneity in the treatment effect -- Consistent with the results on school enrollment, younger children's likelihood of participating in wage-earning activities increase, while that of adult children fall. All these results highlight the role that sunk cost on children's education and parental expectation about their children's education play in determining parental human capital investment in children.