Who Beneﬁts From Multiple Choice(s)? - The Equilibrium Impacts of Test-Optional College Admissions - Jack Kelly - Senior Essay

Sep 14, 2026

Uploaded by Lute

A rapidly increasing number of American colleges and universities have implemented test-optional admissions policies, allowing applicants to withhold SAT/ACT scores and sparking extensive public debate with limited formal study. This paper estimates equilibrium effects by comparing test-optionality to a full-disclosure counterfactual and builds a framework to simulate policy switches using test-optional status quo data. Using proprietary applicant data from 2017–2019, results indicate minimal changes in underrepresented minority admissions shares and non-submitters, driven by limited outcome movement and demographically similar beneficiaries and harmed students.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages50

Reading Time2 h 6 min

AccessFree to download

Contents

1 Introduction

2 Related Literature

3 Institutional Details, Data Description, and Summary Statistics

4 Simulating Admissions Under Full Disclosure

5 Testing Models of Test-Optionality

6 Conclusion

Who Beneﬁts From Multiple Choice(s)?: The Equilibrium Impacts of Test-Optional College Admissions Jack Kelly􀀃 Senior Essay, Yale University Dept. of Economics, April 2022 Abstract A rapidly increasing number of American colleges and universities have implemented test-optional admissions policies that allow applicants to withhold their SAT/ ACT scores, occasioning a lively public dialogue but little formal scholarship. This paper estimates the equilibrium effects of such policy by asking which students are more likely to get in under test-optionality than when required to submit scores (what I term “full disclosure” admissions) . I develop a general framework that can be used to simulate the effect of a switch to full disclosure given data from a test-optional status quo, and apply this to a proprietary data set of all applicants from a test-optional college between 2017 and 2019 . My simulation results suggest that a switch to fulldisclosure would have little impact on the share of admitted students who are underrepresented minorities and the share who are non-submitters. My ﬁndings indicate that full disclosure has minimal effects on these quantities both because (i) few students see their admissions outcomes change and (ii) among those whose outcomes do change, those students who are beneﬁtted and those who are harmed by the switch to full disclosure are demographically similar. I further analyze different models of behavior under test-optionality to examine how universities may behave in the absence of scores. 􀀃 B.A. Candidate, Economics & Mathematics, e-mail: [jack.kelly@yale.edu](jack.kelly@yale.edu). I would like to thank my advisor, Professor Jason Abaluck, for constant support, insight, and patience throughout the process of writing this senior essay. I never could have navigated the curcuitous nature of either the research process or the SOM stairwells without his guidance. I also beneﬁtted greatly from the assistance of the senior essay teaching fellow Natalia Drozdoff, and from conversations regarding this paper with Professors Mira Frick and Eduardo Dvila. Professor Dvila also was kind enough to employ me as a a research assistant of his for over a year a half, a period during which he taught me an immense amount, ranging from the beautiful subtleties of normative economics to how to pronounce“dplyr.” I thank the Economics Department's Tobin and Scarf programs for making such research opportunities possible. I also wish to thank the Mellon Senior Research Grant for ﬁnancially supporting this project. This paper uses proprietary data from a test-optional college; I would like to extend my deep gratitude to the admissions department at this college for agreeing to share their anonymized data with me and for replying to every question I emailed them with inspiring speed and thoroughness. My friends and family have provided vital emotional support throughout the essay-writing process. I particularly wish to thank my mother, Madelyn Kelly, who, along with constant encouragement, has texted me every single article about test-optional admissions that has appeared in a major news outlet since I told her about the nature of my project. Had I chosen any other topic, I know she would have done the same. Any errors are my own. Contents 1 Introduction 3 2 Related Literature 6 3 Institutional Details, Data Description, and Summary Statistics 8 3.1 Institutional Details ........................................... 8 3.2 Preliminary Data Cleaning and Baseline Sample ............................ 9 3.3 Available Variables ........................................... 9 3.4 Classifying Test-Optional Applicants .................................. 9 3.5 Test-Score Availability Among Test-Optional Applicants ....................... 10 3.6 Descriptive Statistics .......................................... 10 3.6.1 Frequency of Non-Submission ................................. 10 3.6.2 Application Outcomes an

FAQs

What question does this paper address about test-optional admissions?

It estimates which students are more likely to be admitted under test-optionality compared with a full-disclosure setting where SAT/ACT scores are required.

How is the full-disclosure counterfactual simulated?

The paper develops a general framework that uses data from a test-optional status quo to simulate what would happen if the policy switched to full disclosure.

What are the main findings regarding underrepresented minorities and non-submitters?

A switch to full disclosure would have little impact on the admitted shares of underrepresented minorities and on the share of non-submitters, with minimal effects for these quantities.

Sep 14, 2026

Uploaded by Lute