From 2000 to 2011, the number of mobile phone subscriptions in developing countries increased from 250 million to 4.5 billion.* My work explores the opportunities created by these new networks.

Working Papers
Accepted, Review of Economic Studies. [ Slides | Supplemental Appendix ]
This paper develops a method to estimate and simulate the adoption of a network good. I estimate demand for mobile phones as a function of individuals’ social networks, coverage, and prices, using transaction data from nearly the entire network of Rwandan mobile phone subscribers at the time, over 4.5 years. I estimate the utility of adopting a phone based on its eventual usage: subscribers pay on the margin, so calls reveal the value of communicating with each contact. I use this structural model to simulate the effects of two policies. A requirement to serve rural areas lowered operator profits but increased net social welfare. Developing countries heavily tax mobile phones, but standard metrics that neglect network effects grossly understate the true welfare cost in a growing network, which is up to 3.14 times the revenue raised. Shifting from handset to usage taxes would have increased the surplus of poorer users by at least 38%.
with Darrell Grissen
Many households in developing countries lack formal financial histories, making it difficult for banks to extend loans, and for potential borrowers to receive them. However, many of these households have mobile phones, which generate rich data about behavior. This paper shows that behavioral signatures in mobile phone data predict loan default. We evaluate our approach using call records matched to lending outcomes in a middle income South American country. Individuals in the highest quartile of risk by our measure are 7.4 times more likely to default than those in the lowest quartile. The method is predictive for both individuals with financial histories, and those without, who cannot be scored using traditional methods. We benchmark performance on our sample of individuals with (thin) financial histories: our method performs no worse than models using credit bureau information. The method can form the basis for new forms of lending that reach the unbanked.
with B. Douglas Bernheim, Jeffrey Naecker, and Antonio Rangel
A central task in microeconomics is to predict choices in as-yet-unobserved situations (e.g., after some policy intervention). Standard approaches can prove problematic when sufficiently similar changes have not been observed or do not have observable exogenous causes. We explore an alternative approach that generates predictions based on relationships across decision problems between actual choice frequencies and non-choice subjective evaluations of the available options. In a laboratory experiment, we find that this method yields accurate estimates of price sensitivities for a collection of products under conditions that render standard methods either inapplicable or highly inaccurate.

Although the spread of new technologies is vital for economic development, it is difficult to study with traditional sources of data. The mobile phone represents a new technology which automatically records every potential learning experience, and nearly every remote interaction with peers who could share their own learning experiences. In 2006, a Rwandan mobile phone operator introduced a new plan that represented substantial savings for over 85% of subscribers. This project uses operator data to investigate how individuals learned about this new plan, and aims to differentiate between learning by doing, from the experience of social network neighbors, and from official sources.
The Joint Network Structure of Spillovers and Policy: Evidence from a Mobile Phone Handout in Rwanda
with Darrell Grissen
Forthcoming, American Economic Association Papers and Proceedings
Many households in developing countries lack access to credit: physically providing small loans to poor and remote populations is costly. However, the digitization of developing countries enables a new model: digital credit delivered directly via mobile phones. Mobile money enables inexpensive financial transfers, and mobile phones capture behavior that can predict repayment when mined with machine learning. This paper evaluates the potential of digital credit to reach those excluded from current financial systems.
In Progress
Hidden Quality
Competition in Network Industries: Evidence from Mobile Telecommunications in Rwanda
Making Decisions with Manipulated Data
with Joshua Blumenstock
White Papers
When analyzing cellular network usage, data on cell tower locations may not be complete. This document outlines a simple procedure to estimate the geographical locations of unknown towers based on call handoffs with known towers.