Working Papers
1. Managing to Learn (with Sabrin Beg and Adrienne M. Lucas) Under Review
To improve public services, public sector managers must encourage reticent civil servants to enact effective reforms. We show through a randomized controlled trial that school principals, i.e., school mangers, can act as leaders to improve InstructionalManagement (0.3SD) and student learning (0.11SD) with existing systems and personnel. Additional management training improved People Management but not student test scores. Managerial enhancements and student test score gains persisted. Our findings resolve conicting results regarding the role of management in public sector productivity and demonstrate how public sector managers can signal reform effectiveness through personal commitment, acting as leaders.
Impact analysis available here
Link to NBER Working Paper #31757
IPA summary available here and learn more about Differentiated Learning here
2. Firm Location, Childcare, and the Gender Profit Gap (with Solène Delecourt, Anya Marchenko, and Layna Lowe) Under Review
There is a significant gender profit gap among small firms in developing countries. How do a firm’s initial conditions contribute to this disparity? If men and women start different businesses – operating in different locations or sectors – and those initial conditions are hard to change, it could limit overall profitability for women. In our representative sample of 3,077 businesses in Kenya, women’s profits are 47% lower than men’s. We identify two initial and sticky conditions contributing to this gap. First, women locate their businesses in less profitable locations, closer to home. Second, women operate in crowded sectors, whereas men are over five times more likely to be monopolists than women. These constraints are more binding for women with greater childcare responsibilities, emphasizing the importance of initial conditions in perpetuating the gender profit gap.
3. When Given Discretion, Teachers Did Not Shirk: Evidence from Remedial Education in Secondary Schools (with Sabrin Beg, Jason Kerwin, Adrienne M. Lucas, and Wahed Rahman) Under Review
Public-sector organizations face a tradeoff: allowing workers discretion at the point of service to adapt to local needs, versus rigid harmonization to ensure uniform service delivery. We examine this trade-off in the context of secondary schools in Odisha, India, where the centrally set curriculum is nearly 4 grades above the learning levels of the mean student. We conduct a randomized intervention that assigned schools to either a rigid or a flexible version of a remedial learning intervention that displaced the curriculum. We compare learning outcomes and teaching quality to the status quo. Both interventions increased learning by 0.11SD, about 60 percent of a year of learning, with gains throughout the learning distribution. We find no crowd-out of grade-level mastery, and no change in the likelihood of earning passing Board Marks one year later. Discretion did not lower the quality of implementation or induce shirking. Allowing teachers flexibility to adjust classroom content to student needs was beneficial and had limited downsides.
Paper available here (includes appendices)
4. Learning Beyond School: Another Chance for Out of School Adolescent Girls in Rural Pakistan (with Sabrin Beg and Adrienne M. Lucas)
Out of school adolescents have few opportunities to engage with formal or informal schooling even if the reason for their lack of engagement with schooling was temporary. Through a 134- village randomized controlled trial in rural Punjab, Pakistan, we find that an alternative, more intense and more culturally palatable out of school learning bootcamp increased literacy and numeracy among adolescent girls who had either never attended or dropped out of school. The program increased the amount of socializing that the girls did with friends but did not affect their, their household heads’, or their village leaders’ beliefs about female empowerment, gender equality, or the proper education amount or labor market roles for women.
5. Health Insurance and High Cost Borrowing: The Effect of Medicaid on Pawn Loans, Payday Loans, and other Non-Bank Financial Products (with Katie Fitzpatrick and Xiangchen Liu)- Under Review
Winner of the Richard L. D. Morse Applied Consumer Economics Award for a professional paper recognizes research that addresses practical and everyday problems that consumers face
Linking Current Population Survey supplements from 2009-2019, we evaluate the effect of the Medicaid expansion on insurance coverage and financial security, measured by fringe bank use (e.g., payday lenders, check cashers, etc.). With synthetic control, we estimate aggregate and state-level effects. We show that Medicaid expansions increased coverage by 6 percentage points (21%) and reduced demand for fringe bank credit by 2 percentage points (20%). We document substantial state variation in the effects of Medicaid expansion. We use machine learning to correlate treatment effects to state-level characteristics. Treatment effects are associated with the state’s Medicaid policies and economic environment.
7. Uninsured, Unbanked, and Unconnected: Health Insurance, Financial Exclusion, and the Digital Divide (with Katie Fitzpatrick)
We test whether financial exclusion and digital exclusion are potential barriers to health insurance coverage after the Affordable Care Act (ACA). We define financial exclusion as whether a household lacks a bank account, a credit card, or relies on “fringe banks” such as pawn shops, payday lenders, and other non-bank financial service providers. We define digital inclusion as those without home internet access. We link three years of individual responses from two supplements of the nationally representative Current Population Survey from the period after the ACA was enacted (2015, 2017, 2019). Using a sample of non-elderly households we assess the relationship between health insurance coverage and financial exclusion stratified by income (below 138% of the FPL or between 138-400% of the FPL), digital exclusion, and health status using a multivariate regression analysis. Individuals who remain without health insurance coverage after the ACA are those most likely to experience financial exclusion, and this financial exclusion may create additional difficulties to obtain coverage through online portals. Low-income, uninsured adults are 11.5 percentage points (34.8%) more likely to live in a household without a bank account and less likely to have a credit card and emergency savings than low-income, insured adults; they are 5.0 percentage points (9.9%) more reliant on fringe banks. Similar correlations exist between financial inclusion and health insurance coverage among moderate-income households. This work suggests that policies to improve internet, particularly in rural areas, as well as policies to increase bank account ownership may have important spillover effects.
8. The Sources of Researcher Variation in Economics (with many authors) - Under Review
We use a rigorous three-stage many-analysts design to assess how different researcher decisions—specifically data cleaning, research design, and the interpretation of a policy question—affect the variation in estimated treatment effects. A total of 146 research teams each completed the same causal inference task three times each: first with few constraints, then using a shared research design, and finally with pre-cleaned data in addition to a specified design. We find that even when analyzing the same data, teams reach different conclusions. In the first stage, the interquartile range (IQR) of the reported policy effect was 3.1 percentage points, with substantial outliers. Surprisingly, the second stage, which restricted research design choices, exhibited slightly higher IQR (4.0 percentage points), largely attributable to imperfect adherence to the prescribed protocol. By contrast, the final stage, featuring standardized data cleaning, narrowed variation in estimated effects, achieving an IQR of 2.4 percentage points. Reported sample sizes also displayed significant convergence under more restrictive conditions, with the IQR dropping from 295,187 in the first stage to 29,144 in the second, and effectively zero by the third. Our findings underscore the critical importance of data cleaning in shaping applied microeconomic results and highlight avenues for future replication efforts.
Link to Paper (NBER Working Paper 33729)
Works in Progress
Barriers to Enrolling in Health Insurance: Estimates on the Value of Time (with Rebecca Thornton) Manuscript preparation
Differentiated Learning Experimentation (with Adrienne M. Lucas, Sabrin Beg, and Stephanie Bonds) Fieldwork completed
Improving Childcare Quality Through Social Franchising (with Emily Beam and Maira Reimão) Fieldwork ongoing
Systems-Based Approaches to Improving Teaching Effectiveness (with Sabrin Beg, Christina Brown, James Berry, and Adrienne Lucas) Fieldwork ongoing
Medicycles: Sustainable, Integrated Outreach Clinics in Remote Communities of Uganda (with Maria Dieci, Mahesh Karra, Xufeng Liu, and Aggrey Semeere) Fieldwork planning