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. Economic and Econometric Methods Used in the Tech Sector
« Back to Results

Economic and Econometric Methods Used in the Tech Sector

Paper Session

Friday, Jan. 7, 2022 3:45 PM - 5:45 PM (EST)

Hosted By: Econometric Society
  • Chairs:
    Elizabeth Mishkin, Uber Technologies, Inc.
  • Jonathan V. Hall, Uber Technologies, Inc.

The Paper of How: Estimating Treatment Effects with the Front-Door Criterion

Marc Bellemare
,
University of Minnesota

Abstract

We present the first application of Pearl's (1995) front-door criterion to observational data wherein the assumptions for point identification plausibly hold. For identification, the front-door criterion exploits exogenous mediator variables on the causal path. We estimate the effect of authorizing a shared Uber or Lyft ride on tipping by exploiting the plausibly exogenous variation in whether one actually shares a ride with a stranger conditional on authorizing sharing, on the full fare paid, and on a battery of fixed effects. We find that most of the observed negative effect on tipping is driven by selection. We then explore the consequences of violating the identification assumptions.

Reducing Interference Bias in Online Marketplace Pricing Experiments

David Holtz
,
University of California-Berkeley
Ruben Lobel
,
Airbnb
Inessa Liskovich
,
Airbnb
Sinan Aral
,
Massachusetts Institute of Technology

Abstract

Online marketplace designers frequently run A/B tests to measure the impact of proposed product changes. However, given that marketplaces are inherently connected, total average treatment effect estimates obtained through Bernoulli randomized experiments are often biased due to violations of the stable unit treatment value assumption. This can be particularly problematic for experiments that impact sellers’ strategic choices, affect buyers’ preferences over items in their consideration set, or change buyers’ consideration sets altogether. In this work, we measure and reduce bias due to interference in online marketplace experiments by using observational data to create clusters of similar listings, and then using those clusters to conduct cluster-randomized field experiments. We provide a lower bound on the magnitude of bias due to interference by conducting a meta-experiment that randomizes over two experiment designs: one Bernoulli randomized, one cluster randomized. In both meta-experiment arms, treatment sellers are subject to a different platform fee policy than control sellers, resulting in different prices for buyers. By conducting a joint analysis of the two meta-experiment arms, we find a large and statistically significant difference between the total average treatment effect estimates obtained with the two designs, and estimate that 32.60% of the Bernoulli-randomized treatment effect estimate is due to interference bias. We also find weak evidence that the magnitude and/or direction of interference bias depends on extent to which a marketplace is supply- or demand-constrained, and analyze a second meta-experiment to highlight the difficulty of detecting interference bias when treatment interventions require intention-to-treat analysis.

The Limits of Centralized Pricing in Online Marketplaces and the Value of User Control

Apostolos Filippas
,
New York University
Srikanth Jagabathula
,
New York University
Arun Sundararajan
,
New York University

Abstract

We report experimental and quasi-experimental evidence from a randomized controlled trial and its aftermath. The setting is of a peer-to-peer platform that transitioned from decentralized to centralized pricing. Centralized pricing increased the utilization of providers’ assets, resulting in higher revenues but also higher transaction costs. Barred from accessing the price system, providers made non-price adjustments such as reducing the availability of their assets, canceling booked transactions, and exiting the market. Providers allowed to retain partial pricing control reacted substantially less negatively while enjoying similar revenue increases. We highlight the challenges of implementing centralized pricing in platform settings and assessing its welfare implications. We show that partial control can mitigate challenges by allowing providers to express their private and heterogeneous preferences while retaining some benefits of centralization.
JEL Classifications
  • C1 - Econometric and Statistical Methods and Methodology: General