Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Statistical Discrimination in Ratings-Guided Markets

Yeon-Koo Che, Kyungmin Kim, Weijie Zhong

TL;DR

In this work, the possible vulnerability of the ratings-based social learning to discriminatory inferences on social groups is identified and policy implications in terms of regulating trading relationships as well as algorithm design are explored.

Abstract

We study statistical discrimination of individuals based on payoff-irrelevant social identities in markets that utilize ratings and recommendations for social learning. Even though rating/recommendation algorithms can be designed to be fair and unbiased, ratings-based social learning can still lead to discriminatory outcomes. Our model demonstrates how users' attention choices can result in asymmetric data sampling across social groups, leading to discriminatory inferences and potential discrimination based on group identities.

Statistical Discrimination in Ratings-Guided Markets

TL;DR

In this work, the possible vulnerability of the ratings-based social learning to discriminatory inferences on social groups is identified and policy implications in terms of regulating trading relationships as well as algorithm design are explored.

Abstract

We study statistical discrimination of individuals based on payoff-irrelevant social identities in markets that utilize ratings and recommendations for social learning. Even though rating/recommendation algorithms can be designed to be fair and unbiased, ratings-based social learning can still lead to discriminatory outcomes. Our model demonstrates how users' attention choices can result in asymmetric data sampling across social groups, leading to discriminatory inferences and potential discrimination based on group identities.

Paper Structure

This paper contains 17 sections, 8 theorems, 9 equations, 2 figures.

Table of Contents

  1. Introduction
  2. Model
  3. Players.
  4. Matching.
  5. Trade.
  6. Ratings.
  7. Belief and Search Decision.
  8. Solution Concept.
  9. Equilibrium Characterization
  10. Steady state distributions and beliefs
  11. Trade vs. No Trade
  12. Non-discriminatory equilibrium
  13. Discriminatory Equilibrium
  14. Stability of Ratings-Based Discrimination
  15. Discussions
  16. ...and 2 more sections

Key Result

Lemma 1

Suppose $(u_{H}+u_{L})/2>p$. Then, $u_{B}(\lambda)$ is always decreasing, while $u_{G}(\lambda)$ is monotone decreasing if and only if If $k>\underline{k}$, then there exist $\overline{\lambda}_{G}>\underline{\lambda}_{G}>0$ such that $u_{G}(\lambda)$ is increasing if and only if $\lambda\in(\underline{\lambda}_{G},\overline{\lambda}_{G})$.

Figures (2)

  • Figure 1: The blue solid curves depict the buyer-indifference condition \ref{['eq:symmetric_equal_profit']}, while the red dashed curves depict the market clearing condition \ref{['eq:symmetric_market_clear']}. The common parameters used for this figure are $\delta=1$, $\alpha=0.1$, $u_{H}=2$, and $u_{L}=p=1$
  • Figure 2: The blue solid curve represents the buyer-indifference condition \ref{['eq:symmetric_equal_profit']}. The parameters used for this figure are $\delta=0.2$, $\alpha=0.5$, $u_{H}=3$, $u_{L}=1$, $w=1.5$ .

Theorems & Definitions (9)

  • Lemma 1
  • Proposition 1
  • Proposition 2
  • Theorem 1
  • Definition 1
  • Theorem 2
  • Proposition 3
  • Proposition 4
  • Theorem 3