Research

Papers

Anticipation and Consumption

(with Linh Tô)

Abstract

This paper introduces a model of how the timing of information affects consumption decisions and tests its predictions in both developed and developing contexts. In our model, consumers form intertemporal plans and experience utility from anticipating future consumption. The model predicts excess sensitivity of spending to receiving a windfall, with smaller spending responses when there is more time to anticipate receiving the payment. The prediction that waiting leads to more patient decisions does not depend on whether consumers are liquidity constrained. Using Nielsen Consumer Panel data, we find higher marginal propensities to spend for households scheduled to receive the 2008 Economic Stimulus Payments sooner. Using data from randomized experiments in Kenya and Malawi, we find higher savings and assets among households scheduled to wait longer before receiving lump-sum unconditional cash transfers. Finally, we discuss existing evidence on how consumption responds to gains, losses, and news in light of our model.

Download

Daily Labor Supply and Adaptive Reference Points

(with Linh Tô, Forthcoming, American Economic Review)

Abstract

This paper provides field evidence on how reference points adjust, a degree of freedom in reference-dependence models. To examine this in the context of cabdrivers' daily labor-supply behavior, we ask how the within-day timing of earnings affects decisions. Drivers work less in response to higher accumulated income, with a strong effect for recent earnings that gradually diminishes for earlier earnings. We estimate a structural model in which drivers work towards a reference point that adjusts to deviations from expected earnings with a lag. This dynamic view of reference dependence reconciles conflicting "neoclassical" and "behavioral" interpretations of evidence on daily labor-supply decisions.

Download

 

The Public-Housing Allocation Problem: Theory and Evidence from Pittsburgh

Abstract

We present a model of public-housing allocation, in which objects that arrive stochastically over time must be matched with applicants on a waiting list. Our framework delivers a strategy-proof, ex-ante efficient and envy-free mechanism which allows applicants to trade off their preferences for different units and waiting times by choosing among a set of waiting lists. A counterfactual change from existing mechanisms, in which applicants receive offers after units become available and can refuse only a limited number of times, to the proposed mechanism improves welfare by a lower bound of $6,429 per applicant in a sample of Pittsburgh households.

Download

 

Other Publications

The Design of Public-Housing Policies

in preparation for More Equal by Design: Economic Design Responses to Inequality, eds. Scott Duke Kominers and Alexander Teytelboym, Oxford University Press.

Abstract

Social programs are a central means for addressing inequality. This chapter proposes to view social programs through the lens of market design, focusing on the case of public-housing assistance. One of the goals of this chapter is to deliver lessons for housing policy that can be gleaned from market design. After presenting relevant features of the market for affordable housing, we highlight various dimensions of inequality. We explore design considerations that pertain to inequality among those who do receive assistance as well as inequality in terms of which households receive housing assistance. We describe recent research that attempts to address these issues and suggest directions where further research can make progress. The chapter aims to show more broadly that the market for public housing also exemplifies general lessons for market deign.

Download

 

Matching with Stochastic Arrival

AEA Papers and Proceedings, 109: 209–12. May 2019.

Abstract

This paper examines efficiency and fairness properties in a dynamic allocation problem. The model applies to situations in which objects of different types arrive stochastically over time and must be assigned to agents in a queue, such as the allocation of public housing units. The main result demonstrates the impossibility of designing an allocation mechanism in an environment with stochastic arrival that can guarantee resulting assignments that are efficient or fair ex post.

Download