Prestige in Numbers: How Test Scores and Choices Reveal School Rankings

TL;DR

This paper proposes a new, transparent ranking method for schools based on applicants' choices and test scores, reflecting collective student preferences and resisting manipulation, demonstrated with MBA application data.

Contribution

It introduces a revealed-preference approach to ranking schools using application choices and test scores, offering a manipulation-resistant and customizable alternative to traditional rankings.

Findings

Rankings strongly correlate with established MBA rankings (ρ=0.72).

Method reflects collective applicant judgments rather than expert opinions.

Approach is resistant to institutional manipulation.

Abstract

This paper introduces a novel revealed-preference approach to ranking colleges and professional schools based on applicants' choices and standardized test scores. Unlike traditional rankings that rely on data supplied by institutions or expert opinions, our methodology leverages the decentralized beliefs of potential students, as revealed through their application decisions. We develop a theoretical model where students with higher test scores apply to more selective institutions, allowing us to establish a clear relationship between test score distributions and school prestige. Using comprehensive data from over 490,000 GMAT test-takers applying to U.S. full-time MBA programs, we implement two ranking methods: one based on monotone functions of test scores across schools, and another using score-adjusted tournaments between school pairs. Our approach has distinct advantages over traditional rankings: it reflects the collective judgment of the entire applicant pool rather than a small group of experts, and it utilizes data from an independent testing organization, making it resistant to manipulation by institutions. The resulting rankings correlate strongly with leading published MBA rankings ($\rho = 0.72$) while offering the additional benefit of being customizable for different student subgroups. This method provides a transparent alternative to existing ranking systems that have been subject to well-documented manipulation.