Research

Research Interests: Development Economics; Applied Microeconomics

Job Market Paper

Abstract: Microfinance has been shown to have limited average impacts on incomes and profits in developing countries. In addition to credit constraints, the poor also face labor market frictions which may lead them to make constrained occupational choices and investments in businesses. Some entrepreneurs start a business due to lack of alternatives (“involuntary entrepreneurs”) while others do so by choice (“voluntary entrepreneurs”). These two groups of entrepreneurs may respond very differently to microfinance but are difficult to distinguish empirically. We structurally estimate a heterogeneous agent model of occupational choices based on Karaivanov and Yindok (2022). We use data from two randomized experiments in microfinance to test the predictions of the model and differentiate voluntary from involuntary entrepreneurs. We find that the majority of entrepreneurs are involuntary who earn significantly lower income than voluntary ones. This misallocation at the extensive margin translates into significant misallocation in capital and labor investments in businesses. Microfinance leads to a positive impact on incomes for both types of entrepreneurs and wage workers by increasing credit and labor investments, but reduces the overall entrepreneurship rate. Income gains are driven by voluntary entrepreneurs and those with lower wealth levels. Simulating policy counterfactuals provides evidence that relaxing credit constraints lead to lower impacts than relaxing labor market constraints and thus, labor market policies may deserve relatively more consideration.
PEDL Grant Description, Research Note

Publications

Texts Don’t Nudge: An Adaptive Trial to Prevent the Spread of COVID-19 in India, with Sebastian Bauhoff, Dev Patel and James Potter
Journal of Development Economics, November 2021, Volume 153 [link]
Other publications: CGD Blog Article

Abstract: We conduct an adaptive randomized controlled trial to evaluate the impact of a SMS-based information campaign on the adoption of social distancing and handwashing in rural Bihar, India, six months into the COVID-19 pandemic. We test 10 arms that vary in delivery timing and message framing, changing content to highlight gains or losses for either one’s own family or community. We identify the optimal treatment separately for each targeted behavior by adaptively allocating shares across arms over 10 experimental rounds using exploration sampling. Based on phone surveys with nearly 4,000 households and using several elicitation methods, we do not find evidence of impact on knowledge or adoption of preventive health behavior, and our confidence intervals cannot rule out positive effects as large as 5.5 percentage points, or 16%. Our results suggest that SMS-based information campaigns may have limited efficacy after the initial phase of a pandemic.

Evaluating the effectiveness of Community Health Worker home visits on infant health: a quasi-experimental evaluation of Home Based Newborn Care Plus in India, with Tom Newton-Lewis
Journal of Global Health, October 2021, Issue 1 [link]

Working Papers/Research in Progress

Mobile Money and Savings in Tanzania

Abstract: This paper combines difference-in-differences methodology with matching techniques on repeated cross-sectional data to empirically investigate the impact of interest income distribution by Mobile Network Operators (MNOs) in Tanzania on savings and consumption behavior of the households using LSMS-ISA data. Starting September 2014, Tigo Pesa distributed interest income to its mobile money customers based on their mobile money wallet balance. The treatment households here are the households who use Tigo Pesa as their MNO while control households use other MNOs for mobile money services. I use local linear regression matching to match treatment observations with control observations separately for baseline and endline and then use one-to-one neighbor matching with no replacement to match treatment observations in baseline with endline. I find a statistically significant increase in the proportion of treatment households using mobile money to store or save for daily and large expenses relative to control households and crowding in effects on other traditional sources of savings such as bank accounts and membership in a savings or credit group. While I see an increase in the household expenditures, these results are not statistically significant.

Monitoring in the Context of Limited Budgets: Phone Calls to Primary Schools, with Jenny C. Aker and Juan C. Taborda

Abstract: Teacher absenteeism is an important constraint to learning gains in many low-income countries, especially in remote areas that are costly to monitor. In this randomized evaluation, we randomly assign primary schools to either a “mobile phone” intervention or none. In the mobile intervention, weekly phone calls were made to the school director, two teachers, the village chief and two randomly selected parents. We find that the mobile phone intervention improved teacher and school director attendance, with an order of magnitude of an additional three hours of instruction per week. The intervention also increased parents’ engagement in the school and aspirations for their children, and made teachers feel more appreciated. Yet, we do not find any statistically significant effects on math or reading scores in the first year. This suggests presence of other binding constraints that must be addressed to improve student learning.

Community Currencies & Enterprise Development, with Castro Gichuki & Shotaro N. Nakamura

Abstract: A community currency, as an alternative to the legal tender, circulated within specific communities could overcome the liquidity and credit constraints faced by the poor households specifically for enterprise development. Currently, there is no causal empirical evidence, to date and to our knowledge, on the efficacy of community currencies in inducing “social” transactions and spurring further business activities in developing markets. We are conducting primary data collection in these communities to understand how enterprises leverage social networks to grow, what types of goods and services are transacted through these community currencies and what other mechanisms are induced by them to spur enterprise growth.