Resource Automation + AI Mitigating Harm + Bias

Popular Support for Balancing Equity and Efficiency in Resource Allocation

This study examines public attitudes toward balancing equity and efficiency in algorithmic resource allocation, using online advertising for SNAP enrollment as a case study.

Organizations: Cornell University, Code for America, Stanford University, Harvard Kennedy School

Authors: Allison Koenecke, Eric Giannella, Robb Willer, Sharad Goe

Published Year: 2023

View Source

Download Resource

The researchers analyze the trade-off between efficiency (maximizing total enrollments) and equity (ensuring proportional representation of Spanish-speaking applicants) in GetCalFresh’s digital outreach.

Through experiments and a national survey, they find broad support for prioritizing equity, even at the cost of lower overall enrollment. Their findings challenge the prevailing efficiency-first paradigm in algorithmic decision-making, suggesting that public preferences favor inclusive resource distribution