Paper Type

Short

Paper Number

PACIS2026-1450

Description

Generative AI (GenAI) is reshaping platform-mediated online freelancing markets, yet its heterogeneous income distribution effects across task types and the moderating role of platform mechanisms remain underexplored. This study develops a task-heterogeneity framework to examine GenAI’s potential impacts on within-category income dispersion across AI-substitutable, AI-complementary, and AI-native freelance tasks, as well as the income quantile heterogeneity of these effects. A 23-month panel dataset from Upwork is planned for future empirical implementation, with a Difference-in-Differences (DiD) design complemented by event study specifications and Propensity Score Matching (PSM) proposed for methodological rigor. Platform mechanisms (ranking algorithms, reputation systems, information asymmetry) are conceptually incorporated to illustrate how they may moderate GenAI’s distributional impacts. This study provides a replicable GenAI task typology, advances theoretical understanding of digital labor market inequality, and outlines implications for platforms, gig workers, and policymakers.

Comments

02-FutureofWork

Share

COinS
 
Jul 5th, 12:00 AM

Task Heterogeneity and the Income Effects of Generative AI in Online Freelancing Markets

Generative AI (GenAI) is reshaping platform-mediated online freelancing markets, yet its heterogeneous income distribution effects across task types and the moderating role of platform mechanisms remain underexplored. This study develops a task-heterogeneity framework to examine GenAI’s potential impacts on within-category income dispersion across AI-substitutable, AI-complementary, and AI-native freelance tasks, as well as the income quantile heterogeneity of these effects. A 23-month panel dataset from Upwork is planned for future empirical implementation, with a Difference-in-Differences (DiD) design complemented by event study specifications and Propensity Score Matching (PSM) proposed for methodological rigor. Platform mechanisms (ranking algorithms, reputation systems, information asymmetry) are conceptually incorporated to illustrate how they may moderate GenAI’s distributional impacts. This study provides a replicable GenAI task typology, advances theoretical understanding of digital labor market inequality, and outlines implications for platforms, gig workers, and policymakers.