2025 Theses Doctoral

Statistical Inference in Competitive Equilibrium

Liao, Luofeng

This thesis studies statistical inference and A/B testing in settings with interference arising from competitive market effects.

We study these effects in two fundamental market equilibrium models: the linear Fisher market (LFM) equilibrium and first-price pacing equilibrium (FPPE). LFM arises from fair resource allocation systems (such as physical allocation of food to food banks) and more generally distribution systems (such as user attention to different types of social media notifications on Instagram, or jobseekers’ attention to job posts in LinkedIn). For LFM, we assume that the observed data is captured by the classical finite-dimensional Fisher market equilibrium, and its steady-state behavior is modeled by a continuous limit Fisher market.

The second type of equilibrium we study, FPPE, arises from internet advertising applications, where adver- tisers are constrained by budgets and advertising opportunities are sold via first-price auctions. Pacing is a prevalent approach for managing advertiser budgets, where the platform assigns each advertiser an autobidder that controls expenditure by adaptively shading bids. For platforms that use pacing-based methods to smooth out the spending of advertisers, FPPE provides a steady-state description of the outcome of pacing-based markets.

In Chapter 1 we develop a theory of statistical inference for LFMs and FPPE. In Chapter 2 we investigate theoretically sound bootstrap approaches for LFM and FPPE. In Chapter 3 we apply the theory to reduce interference bias in parallel A/B tests on ad auction platforms.