This website uses cookies.

By clicking the "Accept" button or continuing to browse our site, you agree to first-party and session-only cookies being stored on your device to enhance site navigation and analyze site performance and traffic. For more information on our use of cookies, please see our Privacy Policy.

  1. Home
  2. ASSA Annual Meeting
  3. 2022
  4. Preliminary Program
  5. From Data to Decisions
« Back to Results

From Data to Decisions

Paper Session

Saturday, Jan. 8, 2022 12:15 PM - 2:15 PM (EST)

Hosted By: Econometric Society
  • Chair: Sanjog Misra, University of Chicago

Incorporating Social Welfare in Program-Evaluation and Treatment Choice

Debopam Bhattacharya
,
University of Cambridge
Tatiana Komarova
,
London School of Economics

Abstract

The econometric literature on program-evaluation and optimal treatment-choice takes functionals of observable outcome-distributions as target welfare, and ignores program-impacts on unobserved utilities, including utilities of those whose outcomes are unaffected by the intervention. We show that in the practically important setting of discrete-choice, under general preference-heterogeneity and income-effects, the distribution of indirect-utility is nonparametrically identified from average demand. This enables project-evaluation and treatment-targeting based on aggregate consumer-utility and planners' distributional preferences while allowing for unrestricted consumer heterogeneity at the same time. We also demonstrate theoretical connections between aggregate indirect-utility and Hicksian compensation and Bayesian cost-benefit analysis for treatment allocation problems. Two empirical applications illustrate our results.

Estimating Nesting Structures

Ali Hortaçsu
,
University of Chicago
Jonas Lieber
,
University of Chicago
Julien Monardo
,
Telecom Paris
Aureo de Paula
,
University College London

Abstract

The nested logit model is commonly used to estimate demand in differentiated products
markets. However, it and its generalizations require an assumed nesting structure. In this
paper, we propose to estimate the nesting structure from the data. For this, we build on a
recent generalization of the nested logit model that allows any possible nesting structure
and is consistent with utility-maximization by heterogeneous consumers. In this setting,
estimating the nesting structure amounts to estimating a linear model with many endogenous variables, which is challenging. We show theoretically and in simulations that
non-negativity constraints coming from economic theory are sufficient to recover the nesting structure from data. In doing so, we explore the regularization properties of the nonnegative least squares estimator as demonstrated in the statistical literature and expanded
here to an instrumental variable context. This estimator may be of independent interest

The Value of Data for Price Targeting

Susan Athey
,
Stanford University
Rob Donnelly
,
Instacart
Ayush Kanodia
,
Stanford University
Aaron Kaye
,
University of Michigan
Mitchell Linegar
,
Stanford University

Abstract

This paper analyzes the value of information for targeting price discounts in shopping applications. It applies methods that combine standard consumer choice models from marketing and economics with matrix factorization techniques from machine learning, whereby users have latent preferences and products have latent attributes, and these are learned from data about consumer choice in a setting where prices vary over time. The paper applies the model to individual panel data from supermarket shopping for a large retailer. The paper analyzes the value of data for increasing profits through personalized price targeting, assessing the relative importance of enriching the model (adding more latent factors) versus more precisely estimating parameters of a fixed model, finding that enriching the model as data grows is an important contributor to improved performance. The paper shows that increasing the length of the history of data used for given set of individuals is substantially more valuable for targeting to those users than adding data about more products or additional users. The results have implications for privacy policy and competition policy.
JEL Classifications
  • C5 - Econometric Modeling
  • C3 - Multiple or Simultaneous Equation Models; Multiple Variables