Work in Progress
Updated January 2018
Abstract: The textbook policy response to traffic externalities is congestion pricing. However, quantifying the welfare consequences of pricing policies requires detailed knowledge of commuter preferences and of the road technology. I study the peak-hour traffic congestion equilibrium using rich travel behavior data and a field experiment grounded in theory. Using a newly developed smartphone app, I collected a panel data set with precise GPS coordinates for over 100,000 commuter trips in Bangalore, India. To identify the key preference parameters in my model – the value of time spent driving and schedule flexibility – I designed and implemented a randomized experiment with two realistic congestion charge policies. The policies penalize peak-hour departure times and driving through a small charged area, respectively. Structural estimates based on the experiment show that commuters exhibit moderate schedule flexibility and high value of time. In a separate analysis of the road technology, I find a moderate and linear effect of traffic volume on travel time. I combine the preference parameters and road technology using policy simulations of the equilibrium optimal congestion charge, which reveal notable travel time benefits, yet negligible welfare gains. Intuitively, the social value of the travel time saved by removing commuters from the peak-hour is not significantly larger than the costs to those commuters of traveling at different, inconvenient times.
with Yuhei Miyauchi, updated November 2017
Abstract: We show how urban commuting flows extracted from cell phone transaction data can be used to measure the spatial distribution of income and economic activity within cities. We use data from Dhaka and Colombo to construct commuting flow matrices for several million users in each metropolis, with fine spatial coverage and daily frequency. We relate commuting to labor productivity using a model of workplace choice that predicts a gravity equation. We recover workplace productivity values that rationalize observed commuting patterns; this procedure essentially assigns higher productivity to locations with higher in-commuting, ceteris paribus. Empirically, we show that commuting flows from cell phone data correlate strongly with flows from a transportation survey in Dhaka. We then show that model-predicted income is a robust predictor of self-reported survey income. We apply our method to measure spatial and temporal variation in economic activity. First, we compare the urban economic structures of Dhaka and Colombo. Secondly, we calculate the economic costs of hartals (a form of strike) and find that people travel less on hartal days, an effect concentrated on routes with high predicted income.
Abstract. This paper examines two related hypotheses: the ability of urban drivers to effectively bypass policies that restrict road traffic, and whether these behavioural responses render such policies ineffective. I study an unexpected, large scale driving restriction policy experiment in Delhi in 2016. In the short run, around half of the affected drivers are able to lawfully bypass it by switching to existing unrestricted private travel modes. However, the policy also led to a precisely estimated decrease in average driving travel time excess delay. Both effects are broadly similar during a second, anticipated round of the policy. Methodologically, this paper makes two contributions: traffic congestion is quantified using rich data from Google Maps, and short-term driving substitution patterns are identified using panel daily driver data and the essentially random assignment of odd and even license plates.
Citywide effects of high-occupancy vehicle restrictions: Evidence from “three-in-one” in Jakarta, (with Rema Hanna and Ben Olken), Science, Vol. 357 (6346), 2017.
Abstract: Widespread use of single-occupancy cars often leads to traffic congestion. Using anonymized traffic speed data from Android phones collected through Google Maps, we investigated whether high-occupancy vehicle (HOV) policies can combat congestion. We studied Jakarta’s “three-in-one” policy, which required all private cars on two major roads to carry at least three passengers during peak hours. After the policy was abruptly abandoned in April 2016, delays rose from 2.1 to 3.1 minutes per kilometer (min/km) in the morning peak and from 2.8 to 5.3 min/km in the evening peak. The lifting of the policy led to worse traffic throughout the city, even on roads that had never been restricted or at times when restrictions had never been in place. In short, we find that HOV policies can greatly improve traffic conditions.
Debunking the Stereotype of the Lazy Welfare Recipient: Evidence from Cash Transfer Programs Worldwide, (with Abhijit Banerjee, Rema Hanna, and Ben Olken), World Bank Research Observer, Vol. 32 (2), 2017.
Abstract. Targeted transfer programs for poor citizens have become increasingly common in the developing world. Yet, a common concern among policy-makers and citizens is that such programs tend to discourage work. We re-analyze the data from seven randomized controlled trials of government-run cash transfer programs in six developing countries throughout the world, and find no systematic evidence that cash transfer programs discourage work.
Rapid Innovation Diffusion in Social Networks, (with Peyton Young), Proceedings of the National Academy of Sciences, Vol. 111 (3), 2014.
Abstract. Social and technological innovations often spread through social networks as people respond to what their neighbors are doing. Previous research has identified specific network structures, such as local clustering, that promote rapid diffusion. Here we derive bounds that are independent of network structure and size, such that diffusion is fast whenever the payoff gain from the innovation is sufficiently high and the agents’ responses are sufficiently noisy. We also provide a simple method for computing an upper bound on the expected time it takes for the innovation to become established in any finite network. For example, if agents choose log-linear responses to what their neighbors are doing, it takes on average less than 80 revision periods for the innovation to diffuse widely in any network, provided that the error rate is at least 5% and the payoff gain (relative to the status quo) is at least 150%. Qualitatively similar results hold for other smoothed best-response functions and populations that experience heterogeneous payoff shocks.
Fast Convergence in Evolutionary Equilibrium Selection, (with Peyton Young), Games and Economic Behavior, Vol. 80, 2013.
Abstract. Stochastic best response models provide sharp predictions about equilibrium selection when the noise level is arbitrarily small. The difficulty is that, when the noise is extremely small, it can take an extremely long time for a large population to reach the stochastically stable equilibrium. An important exception arises when players interact locally in small close-knit groups; in this case convergence can be rapid for small noise and an arbitrarily large population. We show that a similar result holds when the population is fully mixed and there is no local interaction. Moreover, the expected waiting times are comparable to those in local interaction models.