View in EDS HTML Full Text PDF Full Text

How Does Having a Football Program Affect Enrollment at Small Private Colleges?

Saved in:

Bibliographic Details
Title:	How Does Having a Football Program Affect Enrollment at Small Private Colleges?
Language:	English
Authors:	Clay Collins, E. Frank Stephenson
Source:	Education Economics. 2026 34(2):311-320.
Availability:	Routledge. Available from: Taylor & Francis, Ltd. 530 Walnut Street Suite 850, Philadelphia, PA 19106. Tel: 800-354-1420; Tel: 215-625-8900; Fax: 215-207-0050; Web site: http://www.tandf.co.uk/journals
Peer Reviewed:	Y
Page Count:	10
Publication Date:	2026
Document Type:	Journal Articles Reports - Evaluative
Education Level:	Higher Education Postsecondary Education
Descriptors:	Small Colleges, Private Colleges, Enrollment, Team Sports, College Athletics, Gender Differences
DOI:	10.1080/09645292.2025.2501068
ISSN:	0964-5292 1469-5782
Abstract:	This paper examines the effects of fielding college football on enrollment and other aspects of NCAA Division III (DIII) institutions. Unlike larger Division I universities, DIII schools, which are primary private colleges, do not award athletic scholarships. Using two-way fixed effect regressions and coarsened exact matching approaches, we examine the presence of a football program on colleges' enrollment, admissions standards, and finances. We find that having a football team is associated with significant increases in total enrollment as well as increases in both male and female enrollment. We find no evidence that football negatively affects admissions rate or college revenues.
Abstractor:	As Provided
Entry Date:	2026
Accession Number:	EJ1500944
Database:	ERIC
Full text is not displayed to guests. Login for full access.

FullText	Links: – Type: pdflink Url: https://content.ebscohost.com/cds/retrieve?content=AQICAHj0k_4E0hTGH8RJwT4gCJyBsGNe_WN95AvKlDbXJGqwxwGSUbXZOWafbCZimeYP3TZGAAAA4zCB4AYJKoZIhvcNAQcGoIHSMIHPAgEAMIHJBgkqhkiG9w0BBwEwHgYJYIZIAWUDBAEuMBEEDGMTR_qMwfMMVSAGWAIBEICBmxwqHFi5ieQq6mXGXGEqb0OXjLIuw9W2E1zKgXNKpUyQD3Vt7uU6X-yeZuWyujfRA_bZ9O5m7XiD98y5rDmLka99gXgJJ3aFA7tczSikFdRH4DiAJNovalSrVeXhEOg44ehhljonQ4TXBP5VCRAfQOXOddEh99nL7S6SLzERa5Cbi1T6pTAG92rCaSrvlWEgnVOtninY8YKmUA3d Text: Availability: 1 Value: <anid>AN0192371729;ede01apr.26;2026Mar23.03:20;v2.2.500</anid> <title id="AN0192371729-1">How does having a football program affect enrollment at small private colleges? </title> <p>This paper examines the effects of fielding college football on enrollment and other aspects of NCAA Division III (DIII) institutions. Unlike larger Division I universities, DIII schools, which are primary private colleges, do not award athletic scholarships. Using two-way fixed effect regressions and coarsened exact matching approaches, we examine the presence of a football program on colleges' enrollment, admissions standards, and finances. We find that having a football team is associated with significant increases in total enrollment as well as increases in both male and female enrollment. We find no evidence that football negatively affects admissions rate or college revenues.</p> <p>Keywords: College enrollment; university sports; college football</p> <hd id="AN0192371729-2">Introduction</hd> <p>While football programs sponsored by large universities such as Alabama, Clemson, and Michigan draw the largest crowds, have lucrative broadcast packages, and receive the most public attention, many smaller colleges also have football programs. The large university programs compete in the National Collegiate Athletic Association's (NCAA) Division I (DI), but the NCAA also has Division II (DII) and Division III (DIII).[<reflink idref="bib1" id="ref1">1</reflink>] DII is comprised primarily of regional state universities, which, like DI institutions offer athletic scholarships to student-athletes. By contrast, DIII members are mostly smaller private colleges, and institutions competing in this division are not allowed to offer athletic scholarships. For athletes, DIII is a 'pay to play' model, but DIII student-athletes are eligible for any need-based or merit scholarships offered by their institutions so many receive substantial financial aid.</p> <p>Although DIII football is much lower profile than programs affiliated with large universities, the past decade has seen several small colleges add football programs. Examples include Berry College in Georgia, Hendrix College in Arkansas, Keystone College in Pennsylvania, and Stevenson University in Maryland. In many cases, adding enrollment is specifically cited as the primary justification for adding football.[<reflink idref="bib2" id="ref2">2</reflink>] Conversely, institutions such as Occidental College in California and Principia College in Illinois have discontinued football programs (see Table 1 for a list of private institutions adding or dropping DIII football within the time period analyzed in this paper).</p> <p>Table 1. DIII schools that added/dropped football.</p> <p> <ephtml> &lt;table&gt;&lt;thead valign="bottom"&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Added football&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Dropped football&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;School&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Year added&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;School&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Year dropped&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;/thead&gt;&lt;tbody&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Alvernia University (PA)&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;2018&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Blackburn College (IL)&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;2009&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Anna Maria College (MA)&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;2009&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Concordia College at Moorhead (MN)&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;2014&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Berry College (GA)&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;2013&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Earlham College (IN)&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;2019&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Birmingham Southern College (AL)&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;2007&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Iowa Wesleyan College (IA)&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;2017&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Crown College (MN)&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;2008&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Maranatha Baptist University (WI)&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;2017&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Finlandia University (MI)&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;2015&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Occidental College (CA)&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;2020&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Gallaudet University (DC)&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;2007&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Principia College (IL)&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;2009&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;George Fox University (OR)&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;2014&lt;/p&gt;&lt;/td&gt;&lt;td /&gt;&lt;td /&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Hendrix College (AR)&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;2013&lt;/p&gt;&lt;/td&gt;&lt;td /&gt;&lt;td /&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Hilbert College (NY)&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;2022&lt;/p&gt;&lt;/td&gt;&lt;td /&gt;&lt;td /&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Illinois College (IL)&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;2014&lt;/p&gt;&lt;/td&gt;&lt;td /&gt;&lt;td /&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Keystone College (PA)&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;2020&lt;/p&gt;&lt;/td&gt;&lt;td /&gt;&lt;td /&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;LaGrange College (GA)&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;2006&lt;/p&gt;&lt;/td&gt;&lt;td /&gt;&lt;td /&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Misericordia University (PA)&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;2012&lt;/p&gt;&lt;/td&gt;&lt;td /&gt;&lt;td /&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Pacific University (OR)&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;2010&lt;/p&gt;&lt;/td&gt;&lt;td /&gt;&lt;td /&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Saint Vincent College (PA)&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;2007&lt;/p&gt;&lt;/td&gt;&lt;td /&gt;&lt;td /&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Southwestern University (TX)&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;2013&lt;/p&gt;&lt;/td&gt;&lt;td /&gt;&lt;td /&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Stevenson University (MD)&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;2011&lt;/p&gt;&lt;/td&gt;&lt;td /&gt;&lt;td /&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;The College of Saint Scholastica (MN)&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;2008&lt;/p&gt;&lt;/td&gt;&lt;td /&gt;&lt;td /&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;University of New England (ME)&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;2018&lt;/p&gt;&lt;/td&gt;&lt;td /&gt;&lt;td /&gt;&lt;/tr&gt;&lt;/tbody&gt;&lt;/table&gt; </ephtml> </p> <p>This paper's purpose is to provide a systematic analysis of the effects football'spresence on enrollment at private DIII colleges. In addition to examining overall enrollment, the paper also examines enrollment effects by sex. Belkin ([<reflink idref="bib2" id="ref3">2</reflink>]) reports that one factor motivating schools to add football is boosting enrollment of male students, as many small private colleges have female-to-male ratios well over 1:1. Moreover, this paper estimates the effects of sponsoring football on the percentage of students who are black since many colleges are concerned about diversifying their student bodies. In addition, the paper considers football programs' effects on the percentage of applicants accepted and on colleges' retention rates to provide insight into football's effects on academic quality. Finally, the paper analyzes football's effects on overall revenue and expenditures and on students awarded financial aid since a common criticism is that football's facilities, coaching staff, and travel might make it a money-losing endeavor for colleges.</p> <p>This is one of the first studies examining the effect of Division III (non-scholarship) football on college attendance. Anecdotally (Belkin [<reflink idref="bib2" id="ref4">2</reflink>]; Demirel [<reflink idref="bib6" id="ref5">6</reflink>]), smaller, private colleges and universities have begun college football programs in recent years in bids to increase enrollment. We find that, in general, these strategies can be successful in increasing student enrollment, with both male and female student enrollment increasing by a factor larger than the average football team.</p> <p>Before turning to the paper's analysis, the next section summarizes existing research on the effects of colleges fielding football programs.</p> <hd id="AN0192371729-3">Literature review</hd> <p>The existing research on intercollegiate athletics' effects on colleges and universities falls into three categories. First, papers such as Humphreys ([<reflink idref="bib11" id="ref6">11</reflink>]) and Jones ([<reflink idref="bib16" id="ref7">16</reflink>]) focus on the effect that high-level football programs have on state appropriations and find that football is associated with increased appropriations. Second, papers such as Turner, Meserve, and Bowen ([<reflink idref="bib23" id="ref8">23</reflink>]), Tucker ([<reflink idref="bib22" id="ref9">22</reflink>]), Holmes, Meditz, and Sommers ([<reflink idref="bib10" id="ref10">10</reflink>]), Humphreys and Mondello ([<reflink idref="bib12" id="ref11">12</reflink>]), and Anderson ([<reflink idref="bib1" id="ref12">1</reflink>]) examine the relationship between college sports and alumni giving. While these papers vary in focus (e.g. giving associated with athletic success or giving by former student-athletes), all find positive relationships between athletics and alumni giving.</p> <p>Third, and most relevant for this paper, are studies examining the so-called 'Flutie effect,' a colloquial term for the notion that athletic success increases student interest in institutions. This area of research is extensive – examples include Pope and Pope ([<reflink idref="bib18" id="ref13">18</reflink>], [<reflink idref="bib19" id="ref14">19</reflink>]), Perez ([<reflink idref="bib17" id="ref15">17</reflink>]), Collier et al. ([<reflink idref="bib5" id="ref16">5</reflink>]), and Eggers, Groothuis, and Redding ([<reflink idref="bib7" id="ref17">7</reflink>]) – and tends to find a positive relationship between athletic success and student interest and/or quality. Reinforcing the notion that athletic success or prestige can increase student interest in an institution are recent papers such as Eggers et al. ([<reflink idref="bib8" id="ref18">8</reflink>], [<reflink idref="bib9" id="ref19">9</reflink>]) and Johnson and McCannon ([<reflink idref="bib15" id="ref20">15</reflink>]) which find that bad publicity associated with athletics negatively affects student interest and/or quality. Much of this stream of research is only tangentially related to our paper because it focuses on indicators of athletic success such as going to a bowl game rather than the decision of whether or not to offer a football program. Most of this literature also focuses on large universities rather than the small private institutions found in DIII.</p> <p>However, there are three papers somewhat closely related to ours. First, Caudill, Hourican, and Mixon ([<reflink idref="bib4" id="ref21">4</reflink>]) examine ten schools that either added or eliminated a football program between 1997 and 2015. They find that the size of applicant pools and the ACT scores of incoming students decrease following the discontinuation of football; however, their results show no statistically significant effects associated with adding a football program. Although Caudill, Hourican, and Mixon ([<reflink idref="bib4" id="ref22">4</reflink>]) analyze the effects of adding or dropping a football program, their study differs from ours by not looking directly at enrollment and by focusing on a small number of schools that are generally regional public universities (e.g. University of South Florida, Old Dominion University, and UNC-Charlotte) rather than small private institutions.</p> <p>Unlike Caudill, Hourican, and Mixon ([<reflink idref="bib4" id="ref23">4</reflink>]), Segura and Willner ([<reflink idref="bib20" id="ref24">20</reflink>]) focus exclusively on Division III institutions, but like Caudill, Hourican, and Mixon ([<reflink idref="bib4" id="ref25">4</reflink>]) they examine indicators other than student enrollment. Instead, they focus on the relationship between fielding a football team and a school's median SAT score and its graduation rate. Their results indicate that having a football program is positively related to both median SAT scores and graduation rates. Third, Willner ([<reflink idref="bib25" id="ref26">25</reflink>]) looks at the relationship between football and applications for admission to private Division III institutions over the period 2003–2014. He finds different results for male and female applicants, with a football program increasing the number of male applicants but decreasing the number of female applicants. Although the overall effect on applications is negative, a football program could be a useful tool for increasing applications from male students. However, as with Caudill, Hourican, and Mixon ([<reflink idref="bib4" id="ref27">4</reflink>]) and Segura and Willner ([<reflink idref="bib20" id="ref28">20</reflink>]), Willner ([<reflink idref="bib25" id="ref29">25</reflink>]) does not examine the effect on actual enrollment.[<reflink idref="bib3" id="ref30">3</reflink>] Likewise, none of the three papers somewhat related to ours considers the effects of football on institutions' revenues and expenses.</p> <p>Our paper focuses on small, private colleges forfourreasons. First, as summarized earlier, this is a segment of college athletics that has received relatively scant attention yet has experienced much churn in schools adding or discontinuing football programs. Second, state universities receive additional, often substantial, financial support from their states and therefore have a fundamentally different business model. For them, the decision to add or discontinue football or make other athletic changes may be caused by increasing or declining taxpayer support. Third, public and private universities use separate accounting standards, which would make financial information between the two schools difficult to compare.[<reflink idref="bib4" id="ref31">4</reflink>]</p> <p>Finally, and perhaps most importantly, while declining enrollment is a concern for all institutions, the small colleges which are the majority of DIII are facing some of the biggest challenges. Over a dozen DIII colleges have closed since 2020 (Weaver [<reflink idref="bib24" id="ref32">24</reflink>]). Jacob, McCall, and Stange ([<reflink idref="bib14" id="ref33">14</reflink>]) found that students place a high value on consumption amenities when selecting colleges. Sports programs, including football, are one of those amenities (Eggers et al. [<reflink idref="bib8" id="ref34">8</reflink>]). While some students view football participation as an amenity, others enjoy watching football or engaging in social activities held in conjunction with football games. If college football's campus presence provides valuable amenity for many students, it could be a way for schools to boost enrollment.</p> <hd id="AN0192371729-4">Estimation</hd> <p>This paper uses two estimation strategies to examine football's effect on enrollment. We begin with a two-way fixed effects (TWFE) regression specification to examine the naïve effects of football on enrollment. We refer to this approach as naïve because it ignores the endogeneity of a school's choice to field or not fielda football program; we take up the issue of endogeneity with a second estimation approach discussed below. The TWFE regression is as follows:</p> <p>Graph</p> <p> <ephtml> &lt;math xmlns="http://www.w3.org/1998/Math/MathML"&gt;&lt;mtable columnalign="right left" rowspacing=".5em" columnspacing="thickmathspace" displaystyle="true"&gt;&lt;mtr&gt;&lt;mtd /&gt;&lt;mtd&gt;&lt;mi&gt;Enrollmen&lt;/mi&gt;&lt;msub&gt;&lt;mi&gt;t&lt;/mi&gt;&lt;mrow&gt;&lt;mi&gt;it&lt;/mi&gt;&lt;/mrow&gt;&lt;/msub&gt;&lt;mo&gt;=&lt;/mo&gt;&lt;mi&gt;&amp;#945;&lt;/mi&gt;&lt;mo&gt;+&lt;/mo&gt;&lt;msub&gt;&lt;mtext fontfamily="times"&gt;&amp;#946;&lt;/mtext&gt;&lt;mn&gt;1&lt;/mn&gt;&lt;/msub&gt;&lt;mi&gt;Footbal&lt;/mi&gt;&lt;msub&gt;&lt;mi&gt;l&lt;/mi&gt;&lt;mrow&gt;&lt;mi&gt;it&lt;/mi&gt;&lt;/mrow&gt;&lt;/msub&gt;&lt;mo&gt;+&lt;/mo&gt;&lt;msub&gt;&lt;mi&gt;&amp;#947;&lt;/mi&gt;&lt;mn&gt;1&lt;/mn&gt;&lt;/msub&gt;&lt;mi&gt;PercentAdmitte&lt;/mi&gt;&lt;msub&gt;&lt;mi&gt;d&lt;/mi&gt;&lt;mrow&gt;&lt;mi&gt;i&lt;/mi&gt;&lt;mo&gt;(&lt;/mo&gt;&lt;mrow&gt;&lt;mi&gt;t&lt;/mi&gt;&lt;mo&gt;&amp;#8722;&lt;/mo&gt;&lt;mn&gt;1&lt;/mn&gt;&lt;/mrow&gt;&lt;mo&gt;)&lt;/mo&gt;&lt;/mrow&gt;&lt;/msub&gt;&lt;mo&gt;+&lt;/mo&gt;&lt;msub&gt;&lt;mi&gt;&amp;#947;&lt;/mi&gt;&lt;mn&gt;2&lt;/mn&gt;&lt;/msub&gt;&lt;mi&gt;Tuitio&lt;/mi&gt;&lt;msub&gt;&lt;mi&gt;n&lt;/mi&gt;&lt;mrow&gt;&lt;mi&gt;i&lt;/mi&gt;&lt;mo&gt;(&lt;/mo&gt;&lt;mrow&gt;&lt;mi&gt;t&lt;/mi&gt;&lt;mo&gt;&amp;#8722;&lt;/mo&gt;&lt;mn&gt;1&lt;/mn&gt;&lt;/mrow&gt;&lt;mo&gt;)&lt;/mo&gt;&lt;/mrow&gt;&lt;/msub&gt;&lt;mo&gt;+&lt;/mo&gt;&lt;/mtd&gt;&lt;/mtr&gt;&lt;mtr&gt;&lt;mtd /&gt;&lt;mtd&gt;&lt;msub&gt;&lt;mi&gt;&amp;#947;&lt;/mi&gt;&lt;mn&gt;3&lt;/mn&gt;&lt;/msub&gt;&lt;mspace width="thickmathspace" /&gt;&lt;mi&gt;PercentAwardedFinancialAi&lt;/mi&gt;&lt;msub&gt;&lt;mi&gt;d&lt;/mi&gt;&lt;mrow&gt;&lt;mi&gt;i&lt;/mi&gt;&lt;mo&gt;(&lt;/mo&gt;&lt;mrow&gt;&lt;mi&gt;t&lt;/mi&gt;&lt;mo&gt;&amp;#8722;&lt;/mo&gt;&lt;mn&gt;1&lt;/mn&gt;&lt;/mrow&gt;&lt;mo&gt;)&lt;/mo&gt;&lt;/mrow&gt;&lt;/msub&gt;&lt;mo&gt;+&lt;/mo&gt;&lt;msub&gt;&lt;mi&gt;&amp;#947;&lt;/mi&gt;&lt;mn&gt;4&lt;/mn&gt;&lt;/msub&gt;&lt;mi&gt;PercentBlac&lt;/mi&gt;&lt;msub&gt;&lt;mi&gt;k&lt;/mi&gt;&lt;mrow&gt;&lt;mi&gt;i&lt;/mi&gt;&lt;mo&gt;(&lt;/mo&gt;&lt;mrow&gt;&lt;mi&gt;t&lt;/mi&gt;&lt;mo&gt;&amp;#8722;&lt;/mo&gt;&lt;mn&gt;1&lt;/mn&gt;&lt;/mrow&gt;&lt;mo&gt;)&lt;/mo&gt;&lt;/mrow&gt;&lt;/msub&gt;&lt;mo&gt;+&lt;/mo&gt;&lt;/mtd&gt;&lt;/mtr&gt;&lt;mtr&gt;&lt;mtd /&gt;&lt;mtd&gt;&lt;msub&gt;&lt;mi&gt;&amp;#947;&lt;/mi&gt;&lt;mn&gt;5&lt;/mn&gt;&lt;/msub&gt;&lt;mi&gt;PercentHispani&lt;/mi&gt;&lt;msub&gt;&lt;mi&gt;c&lt;/mi&gt;&lt;mrow&gt;&lt;mi&gt;i&lt;/mi&gt;&lt;mo&gt;(&lt;/mo&gt;&lt;mrow&gt;&lt;mi&gt;t&lt;/mi&gt;&lt;mo&gt;&amp;#8722;&lt;/mo&gt;&lt;mn&gt;1&lt;/mn&gt;&lt;/mrow&gt;&lt;mo&gt;)&lt;/mo&gt;&lt;/mrow&gt;&lt;/msub&gt;&lt;mo&gt;+&lt;/mo&gt;&lt;msub&gt;&lt;mi&gt;&amp;#947;&lt;/mi&gt;&lt;mn&gt;6&lt;/mn&gt;&lt;/msub&gt;&lt;mi&gt;PercentWome&lt;/mi&gt;&lt;msub&gt;&lt;mi&gt;n&lt;/mi&gt;&lt;mrow&gt;&lt;mi&gt;i&lt;/mi&gt;&lt;mo&gt;(&lt;/mo&gt;&lt;mrow&gt;&lt;mi&gt;t&lt;/mi&gt;&lt;mo&gt;&amp;#8722;&lt;/mo&gt;&lt;mn&gt;1&lt;/mn&gt;&lt;/mrow&gt;&lt;mo&gt;)&lt;/mo&gt;&lt;/mrow&gt;&lt;/msub&gt;&lt;mo&gt;+&lt;/mo&gt;&lt;msub&gt;&lt;mi&gt;&amp;#947;&lt;/mi&gt;&lt;mn&gt;7&lt;/mn&gt;&lt;/msub&gt;&lt;mi&gt;CoreRevenue&lt;/mi&gt;&lt;msub&gt;&lt;mi&gt;s&lt;/mi&gt;&lt;mrow&gt;&lt;mi&gt;i&lt;/mi&gt;&lt;mo&gt;(&lt;/mo&gt;&lt;mrow&gt;&lt;mi&gt;t&lt;/mi&gt;&lt;mo&gt;&amp;#8722;&lt;/mo&gt;&lt;mn&gt;1&lt;/mn&gt;&lt;/mrow&gt;&lt;mo&gt;)&lt;/mo&gt;&lt;/mrow&gt;&lt;/msub&gt;&lt;mo&gt;+&lt;/mo&gt;&lt;/mtd&gt;&lt;/mtr&gt;&lt;mtr&gt;&lt;mtd /&gt;&lt;mtd&gt;&lt;msub&gt;&lt;mi&gt;&amp;#947;&lt;/mi&gt;&lt;mn&gt;8&lt;/mn&gt;&lt;/msub&gt;&lt;mi&gt;CoreExpense&lt;/mi&gt;&lt;msub&gt;&lt;mi&gt;s&lt;/mi&gt;&lt;mrow&gt;&lt;mi&gt;i&lt;/mi&gt;&lt;mo&gt;(&lt;/mo&gt;&lt;mrow&gt;&lt;mi&gt;t&lt;/mi&gt;&lt;mo&gt;&amp;#8722;&lt;/mo&gt;&lt;mn&gt;1&lt;/mn&gt;&lt;/mrow&gt;&lt;mo&gt;)&lt;/mo&gt;&lt;/mrow&gt;&lt;/msub&gt;&lt;mo&gt;+&lt;/mo&gt;&lt;msub&gt;&lt;mi&gt;&amp;#948;&lt;/mi&gt;&lt;mn&gt;1&lt;/mn&gt;&lt;/msub&gt;&lt;mi&gt;Religiou&lt;/mi&gt;&lt;msub&gt;&lt;mi&gt;s&lt;/mi&gt;&lt;mrow&gt;&lt;mi&gt;it&lt;/mi&gt;&lt;/mrow&gt;&lt;/msub&gt;&lt;mo&gt;+&lt;/mo&gt;&lt;msub&gt;&lt;mi&gt;&amp;#950;&lt;/mi&gt;&lt;mi&gt;i&lt;/mi&gt;&lt;/msub&gt;&lt;mo&gt;+&lt;/mo&gt;&lt;msub&gt;&lt;mi&gt;&amp;#951;&lt;/mi&gt;&lt;mi&gt;t&lt;/mi&gt;&lt;/msub&gt;&lt;mo&gt;+&lt;/mo&gt;&lt;msub&gt;&lt;mi&gt;&amp;#1013;&lt;/mi&gt;&lt;mrow&gt;&lt;mi&gt;it&lt;/mi&gt;&lt;/mrow&gt;&lt;/msub&gt;&lt;/mtd&gt;&lt;/mtr&gt;&lt;/mtable&gt;&lt;/math&gt; </ephtml> (<reflink idref="bib1" id="ref35">1</reflink>)</p> <p> <emph>Enrollment</emph> is the total number of enrolled students at school i for academic year t. Descriptive statistics are reported in Table 2. The variable of interest is a dummy for whether the school fielded a football team. The football dummy variable is coded as 1 if the college fielded a football team that competed at the DIII level in the fall of that academic year.[<reflink idref="bib5" id="ref36">5</reflink>] As noted earlier, schools that add or drop football during the sample period are listed in Table 1.</p> <p>Table 2. Summary statistics.</p> <p> <ephtml> &lt;table&gt;&lt;thead valign="bottom"&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Variable&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Obs.&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Mean&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Std. dev.&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Min&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Max&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;/thead&gt;&lt;tbody&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Total enrollment&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;6,024&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;2,800&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;3,394&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;256&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;59,144&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Pct. admitted&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;5,970&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;65.31&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;19,323&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;3&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;100&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Football&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;6,248&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;0.573&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;0.495&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;0&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;1&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Religiously affiliated&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;6,248&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;0.607&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;0.488&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;0&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;1&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Tuition&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;5,021&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;34,619&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;10,323&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;7,000&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;66,490&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Pct. awarded financial aid&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;5,996&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;92.32&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;14.00&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;18&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;100&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Retention rate (full-time students)&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;6,024&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;78.77&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;10.36&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;28&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;100&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Pct. black&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;6,024&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;8.533&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;9.173&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;0&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;99&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Pct. Hispanic&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;6,024&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;6.817&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;6.329&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;0&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;59&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Pct. women&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;6,024&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;59.65&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;14.74&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;0&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;100&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Core revenues (millions)&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;6,019&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;128.0&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;474.6&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;0.016&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;13,897&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Core expenses (millions)&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;6,019&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;107.5&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;364.0&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;6.163&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;7,870&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;/tbody&gt;&lt;/table&gt; </ephtml> </p> <p>We employ several control variables, but because enrollment decisions (by both students and schools) are often made using past information, we use a one-year lag for each variable. We also include the percentage of students admitted. A possible argument against having a college football program is that it would lead to a reduction in academic standards and the college's selectivity if the school caters to lower-(academic)-quality athletes at the expense of higher-quality non-athletes. However, as shown in other works, a football program may allow a school to increase its selectivity as the number of applicants increases. Unfortunately, some colleges stopped requiring standardized tests for admissions decisions, so a measure of test scores is infeasible as an indicator of academic quality.[<reflink idref="bib6" id="ref37">6</reflink>] Admissions rates could reflect many factors, but under some circumstances may reflect a college's selectivity and academic quality. Thus, the admissions rate is included as a covariate, but we are well aware of its limitation as a clean indicator of academic quality.</p> <p> <emph>Tuition,</emph> the total tuition and fees for the academic year, and the percentage of students who receive financial aid are included to control for each institution's cost of attendance. Also included are variables for the percentage of total enrolled students who are black, Hispanic, and women which might affect some students' enrollment decisions. We also incorporate a school's reported core revenues and expenses since a school's financial stability and resources may influence enrollment.[<reflink idref="bib7" id="ref38">7</reflink>] We also include a dummy variable if a school's website indicates it is affiliated with any religious denomination.[<reflink idref="bib8" id="ref39">8</reflink>] Lastly, school (</p> <p>Graph</p> <p> <ephtml> &lt;math xmlns="http://www.w3.org/1998/Math/MathML"&gt;&lt;msub&gt;&lt;mi&gt;&amp;#950;&lt;/mi&gt;&lt;mi&gt;i&lt;/mi&gt;&lt;/msub&gt;&lt;mo&gt;)&lt;/mo&gt;&lt;/math&gt; </ephtml> and year (</p> <p>Graph</p> <p> <ephtml> &lt;math xmlns="http://www.w3.org/1998/Math/MathML"&gt;&lt;msub&gt;&lt;mi&gt;&amp;#951;&lt;/mi&gt;&lt;mi&gt;t&lt;/mi&gt;&lt;/msub&gt;&lt;/math&gt; </ephtml> ) fixed effects are included to control for any school-specific effects or year-specific effects (e.g. macroeconomic conditions).</p> <p>Our sample consists of 372 private DIII colleges and universities from 2005/06 to 2022/23 academic years.[<reflink idref="bib9" id="ref40">9</reflink>] Institution data comes from the U.S. Department of Education's Integrated Postsecondary Education Data Systems (IPEDS). This provides information on school size, tuition, revenues and expenses (FASB), and demographic information on the student body. DIII institutions were identified using the NCAA's website. As noted earlier, summary statistics are reported in Table 2.[<reflink idref="bib10" id="ref41">10</reflink>] Many institutions were not in DIII for the entirety of the sample, with some moving to or from DI or DII and others entering or leaving the NCAA. In these cases, the institution only appears in the sample during the years it is a DIII member.[<reflink idref="bib11" id="ref42">11</reflink>]</p> <p>As noted earlier, an important issue with the TWFE approach is the endogeneity of treatment. The decision for a school to field football is not random and is often done in response to (actual or prospective) changes (or lack thereof) in the control variables. Thus, there is a possibility that estimated parameters from the TWFE model will be biased by endogeneity. To address the potential endogeneity of colleges' choosing whether to have a football team, we follow Blackwell et al. ([<reflink idref="bib3" id="ref43">3</reflink>]) with a Coarsened Exact Matching (CEM) design.</p> <p>Unlike other methods to balance covariates to ensure equal sampling, such as propensity score matching, CEM can assume that balancing properties hold throughout the entirety of the sample, as opposed to simply on average. CEM can strictly dominate other equal percent bias-reducing matching methods in reducing imbalance, estimation error, and bias (Iacus, King, and Porro [<reflink idref="bib13" id="ref44">13</reflink>]). Therefore, CEM is more likely to produce a more accurate average treatment on treated (ATT). However, CEM requires the removal of all observations that are not within a coarsened stratum for which there exists a treated and control unit, leading to a significant reduction in observations. In our context, the relatively large number of existing observations, combined with the relatively low variance of the sample prevents a catastrophic decline in observations. Even the most restrictive sample employed contains over 1,200 observations. Matching is based on the lagged control variables listed in the equation above, as well as the religion dummy, with the addition of school and year fixed effects. As a robustness check, we also repeat the CEM model confined to data before 2020 to eliminate any effects associated with the Covid-19 pandemic.</p> <p>Next, we turn to several extensions examining the effect of football on the enrollment of various subsets of the student population. As noted earlier, a stated motivation for many schools to sponsor football is to obtain a more balanced female-male ratio. Hence, we repeat the CEM with the dependent variable confined to either male enrollment or female enrollment. Likewise, fielding a football program may be part of some institutions' actions to have more racially diverse student bodies, so we repeat the CEM model with the dependent variable defined as the black student percentage.</p> <p>Since there are also concerns about football's effects on schools' student bodies and finances, our last set of models repeats the CEM estimation with the acceptance rate, the retention rate, the percentage of students receiving financial aid, core revenues, core expenses, tuition, and total cost to students (measured by the difference between tuition price and average grant aid awarded) as dependent variables.[<reflink idref="bib12" id="ref45">12</reflink>]</p> <p>For all alternative dependent variables, the same matching variables are used in each specification (the religion dummy and lagged values for percent admitted, total enrollment, tuition, percent awarded financial aid, percent black, percent Hispanic, percent women, core revenues, and core expenses). However, we exclude the lagged dependent variable from the matching covariates in each specification. For example, when matching based on the percentage of students who receive financial aid, the lagged percentage of students awarded financial aid is excluded. Similarly, when matching based on core revenues the lagged value of core revenues is excluded.</p> <hd id="AN0192371729-5">Results</hd> <p>Table 3 reports results for total enrollment. In all columns (and all subsequent tables of estimation results), standard errors clustered at the school level are in parentheses. Likewise, school and year fixed effects are included in all columns (and all subsequent tables of estimation results) but, for brevity, are not reported. The <emph>R</emph><sups>2</sups> for all models is very high because of the inclusion of school fixed effects. Since many small private colleges emphasize establishing a residential learning community in which most students live on campus, few small private schools have large swings in enrollment because of their quasi-fixed residential capacity. Columns 1 and 2 are naïve estimates using TWFE; columns 3 and 4 report results from CEM estimation.</p> <p>Table 3. Two-way fixed effects and CEM estimation results.</p> <p> <ephtml> &lt;table&gt;&lt;thead valign="bottom"&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Variable&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;TWFE&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;CEM&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Total enrollment&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Total enrollment&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Total enrollment&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Total enrollment &amp;#8211; pre-2020&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;(1)&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;(2)&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;(3)&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;(4)&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;/thead&gt;&lt;tbody&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Football&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;214.5&amp;#42;&amp;#42;&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;171.3&amp;#42;&amp;#42;&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;321.4&amp;#42;&amp;#42;&amp;#42;&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;380.4&amp;#42;&amp;#42;&amp;#42;&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;(90.98)&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;(79.27)&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;(105.3)&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;(109.7)&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Religiously affiliated&lt;/p&gt;&lt;/td&gt;&lt;td /&gt;&lt;td char="."&gt;&lt;p&gt;95.34&lt;/p&gt;&lt;/td&gt;&lt;td /&gt;&lt;td /&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td /&gt;&lt;td&gt;&lt;p&gt;(158.8)&lt;/p&gt;&lt;/td&gt;&lt;td /&gt;&lt;td /&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Lagged controls&lt;/p&gt;&lt;/td&gt;&lt;td /&gt;&lt;td /&gt;&lt;td /&gt;&lt;td /&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Pct. admitted&lt;/p&gt;&lt;/td&gt;&lt;td /&gt;&lt;td char="."&gt;&lt;p&gt;1.827&lt;/p&gt;&lt;/td&gt;&lt;td /&gt;&lt;td /&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td /&gt;&lt;td&gt;&lt;p&gt;(1.890)&lt;/p&gt;&lt;/td&gt;&lt;td /&gt;&lt;td /&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Tuition&lt;/p&gt;&lt;/td&gt;&lt;td /&gt;&lt;td char="."&gt;&lt;p&gt;0.016&amp;#42;&amp;#42;&lt;/p&gt;&lt;/td&gt;&lt;td /&gt;&lt;td /&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td /&gt;&lt;td&gt;&lt;p&gt;(0.008)&lt;/p&gt;&lt;/td&gt;&lt;td /&gt;&lt;td /&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Pct. awarded financial aid&lt;/p&gt;&lt;/td&gt;&lt;td /&gt;&lt;td&gt;&lt;p&gt;&amp;#8722;0.542&lt;/p&gt;&lt;/td&gt;&lt;td /&gt;&lt;td /&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td /&gt;&lt;td&gt;&lt;p&gt;(1.236)&lt;/p&gt;&lt;/td&gt;&lt;td /&gt;&lt;td /&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Pct. black&lt;/p&gt;&lt;/td&gt;&lt;td /&gt;&lt;td char="."&gt;&lt;p&gt;4.295&lt;/p&gt;&lt;/td&gt;&lt;td /&gt;&lt;td /&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td /&gt;&lt;td&gt;&lt;p&gt;(11.129)&lt;/p&gt;&lt;/td&gt;&lt;td /&gt;&lt;td /&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Pct. Hispanic&lt;/p&gt;&lt;/td&gt;&lt;td /&gt;&lt;td&gt;&lt;p&gt;&amp;#8722;14.36&amp;#42;&lt;/p&gt;&lt;/td&gt;&lt;td /&gt;&lt;td /&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td /&gt;&lt;td&gt;&lt;p&gt;(8.034)&lt;/p&gt;&lt;/td&gt;&lt;td /&gt;&lt;td /&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Pct. women&lt;/p&gt;&lt;/td&gt;&lt;td /&gt;&lt;td char="."&gt;&lt;p&gt;0.693&lt;/p&gt;&lt;/td&gt;&lt;td /&gt;&lt;td /&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td /&gt;&lt;td&gt;&lt;p&gt;(1.624)&lt;/p&gt;&lt;/td&gt;&lt;td /&gt;&lt;td /&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Core revenues (millions)&lt;/p&gt;&lt;/td&gt;&lt;td /&gt;&lt;td&gt;&lt;p&gt;&amp;#8722;0.004&lt;/p&gt;&lt;/td&gt;&lt;td /&gt;&lt;td /&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td /&gt;&lt;td&gt;&lt;p&gt;(0.089)&lt;/p&gt;&lt;/td&gt;&lt;td /&gt;&lt;td /&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Core expenses (millions)&lt;/p&gt;&lt;/td&gt;&lt;td /&gt;&lt;td char="."&gt;&lt;p&gt;2.702&amp;#42;&amp;#42;&amp;#42;&lt;/p&gt;&lt;/td&gt;&lt;td /&gt;&lt;td /&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td /&gt;&lt;td&gt;&lt;p&gt;(0.072)&lt;/p&gt;&lt;/td&gt;&lt;td /&gt;&lt;td /&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Constant&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;957.7&amp;#42;&amp;#42;&amp;#42;&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;508.9&lt;/p&gt;&lt;/td&gt;&lt;td /&gt;&lt;td /&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td /&gt;&lt;td&gt;&lt;p&gt;(84.49)&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;(426.6)&lt;/p&gt;&lt;/td&gt;&lt;td /&gt;&lt;td /&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;School FEs&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Yes&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Yes&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Yes&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Yes&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Year FEs&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Yes&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Yes&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Yes&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Yes&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Obs.&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;6,024&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;4,632&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;1,580&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;1,236&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Adjusted &lt;italic&gt;R&lt;/italic&gt;&lt;sup&gt;2&lt;/sup&gt;&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;0.972&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;0.985&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;0.977&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;0.984&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;/tbody&gt;&lt;/table&gt; </ephtml> </p> <p>1 Note: standard errors clustered by school in parentheses; , , and ** denote 0.1, 0.05, and 0.01 significance levels, respectively. For CEM models, matching based on religious affiliation and lagged values of tuition, pct. awarded financial aid, pct. black, pct. Hispanic, pct. women, pct. admitted, core revenues, and core expenses.</p> <p>As a baseline, column 1 only uses school and year fixed effects, without including any of the covariates such as lagged demographic variables or being religiously affiliated. Having a football team is associated with an additional 215 students, a bit more than twice the size of a typical DIII football roster of about 90–100 players. Column 2 includes the religion dummy and lagged controls for between-school and between-year variations in enrollment, tuition, and demographic variables. With a few exceptions, most notably a strong positive relationship between lagged core expenditures (which include student aid) and enrollment, the estimated coefficients on the covariates are not statistically different from zero.[<reflink idref="bib13" id="ref46">13</reflink>],[<reflink idref="bib14" id="ref47">14</reflink>] The estimated effect of football remains positive and statistically significant, though the magnitude falls by about one-fifth to 171 additional students. Modifying the model to omit various controls (such as the religion dummy, revenues, expenses, or demographic variables) does not change results.</p> <p>As discussed earlier, the results in columns 1 and 2 might be contaminated by endogeneity so we now turn to the CEM results reported in column 3.The CEM estimation usesexactly the same control variables from column 2 for the matching procedure. Examining column 3 indicates that the estimated effect of football remains positive and statistically significant using CEM instead of TWE. Indeed, the estimated coefficient on football, 321 additional students, is nearly twice as large as the TWFE results contained in column 2.</p> <p>Since our sample period includes the Covid-19 pandemic, column 4 reports results from a robustness check in which we repeat the CEM estimation with 2020, 2021, and 2022 excluded. The estimated coefficient on football remains statistically significant and increases to 380 additional students. Hence, our results in columns 1–3 are not driven by the inclusion of the pandemic in our sample; if anything, estimating the model over a time-period with the pandemic included reduces the estimated relationship between football and enrollment.</p> <p>To assess the validity of the CEM approach, Table 4 shows the balance table for the CEM approach used for the regression in column 3 of Table 3. Following Blackwell et al. ([<reflink idref="bib3" id="ref48">3</reflink>]), we find the L1 measure of univariate imbalance for each variable as described by Iacus, King, and Porro ([<reflink idref="bib13" id="ref49">13</reflink>]) for each variable used. The goal of matching is to reduce the univariate distance from the unmatched to the match sample, and the CEM regression used generally performs well. We also show that the global multivariate L1 distance reduces from the unmatched to the matched samples. In all models shown, the multivariate L1 distance between unmatched and matched samples reduces by a minimum of 12%.</p> <p>Table 4. Balance Table for CEM Results (Table 3, Column 3).</p> <p> <ephtml> &lt;table&gt;&lt;thead valign="bottom"&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Variable&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;L1 Univariate Imbalance (Unmatched)&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;L1 Univariate Imbalance (Matched)&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;/thead&gt;&lt;tbody&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Religiously affiliated&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;0.026&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;0.000&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Lagged tuition&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;0.140&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;0.043&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Lagged pct. awarded financial aid&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;0.048&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;0.061&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Lagged pct. black&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;0.172&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;0.160&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Lagged pct. Hispanic&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;0.093&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;0.064&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Lagged pct. women&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;0.385&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;0.151&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Lagged core revenues&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;0.031&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;0.163&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Lagged core expenses&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;0.056&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;0.221&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Total multivariate L1 imbalance&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;0.997&lt;/p&gt;&lt;/td&gt;&lt;td char="."&gt;&lt;p&gt;0.848&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;/tbody&gt;&lt;/table&gt; </ephtml> </p> <p>To test how the presence of football affects student demographics, Table 5 shows CEM results using a series of different demographic variables as the dependent variable. Column 1 uses the total enrollment of male students as the dependent variable. Column 2 uses the total enrollment of female students. Column 3 uses the percentage of students who are black. Matching is based on the same controls as in Table 3, with the relevant lagged dependent variable from the matching criteria removed in each specification.</p> <p>Table 5. Sub-population changes.</p> <p> <ephtml> &lt;table&gt;&lt;thead valign="bottom"&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Variable&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Total enrollment &amp;#8211; men&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Total enrollment &amp;#8211; women&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Pct. black&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;(1)&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;(2)&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;(3)&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;/thead&gt;&lt;tbody&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Football&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;146.1&amp;#42;&amp;#42;&amp;#42;&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;178.87&amp;#42;&amp;#42;&amp;#42;&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;0.479&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;(34.10)&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;(65.29)&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;(0.393)&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;School FEs&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Yes&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Yes&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Yes&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Year FEs&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Yes&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Yes&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Yes&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Obs.&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;3,321&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;3,321&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;2,202&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;&lt;italic&gt;R&lt;/italic&gt;&lt;sup&gt;2&lt;/sup&gt;&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;0.962&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;0.948&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;0.927&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;/tbody&gt;&lt;/table&gt; </ephtml> </p> <p>2 Note: clustered (school) standard errors in parenthesis. , , and ** correspond to 0.1, 0.05, and 0.01 significance levels, respectively. Matching based on religious affiliation and lagged values of tuition, pct. awarded financial aid, pct. Women (except in column 2), pct. black (except in column 3), pct. Hispanic, core revenues, and core expenses.</p> <p>The results show an increase in both male and female student enrollment. The male effect of 146 students is roughly 1.5 times the size of a typical DIII football team. The female effect of 179 is larger than the male effect, supporting the anecdotes in Belkin ([<reflink idref="bib2" id="ref50">2</reflink>]) that some colleges adding football were doing so to continue attracting female students. Recall that Willner ([<reflink idref="bib25" id="ref51">25</reflink>]) found a positive relationship between DIII football and applications from men but a negative relationship between DIII football and applications from women. Our results showing an increase in both male and female enrollment suggest that analyzing applications to determine the effects of DIII football is not a good approach for determining actual enrollment effects. By offering football, the school generates a greater consumption amenity, which attracts both female students and male students (not all of whom play football).</p> <p>As for the effect of football on the black percentage of the student body, the results in column 3 show a positive coefficient but it is not statistically different from zero. This suggests that having a football team has little influence on the student body's racial composition and that adding football will have limited effects on diversity.</p> <p>Table 6 contains results from models focused on indicators of academic quality or financial effects. Column 1 reports results on the percentage of applicants admitted and column 2 uses the retention rate as the dependent variable. In both columns, the estimated coefficients are not statistically significant. Thus, there is no evidence that fielding football causes colleges to increase admission rates or causes retention rates to fall. Similarly, columns 3–7 indicate having a football team has no statistically significant effect on the percentage of students awarded financial aid, core revenues, core expenses, tuition, or cost to students. We find no evidence that tuition increases for schools having football teams. This might be surprising, as football may increase the amenity value of attending the school, which could be reflected in a rise in tuition. When looking at the average cost to students, we find little evidence that the cost of attending schools that have football teamsincreases. This may be a helpful complement to tuition as many smaller schools may have a high tuition price but award comparably large amounts of financial aid to students. We find no evidence that the total financial burden on students changes, which also may suggest that the school is not partaking in other measures alongside sponsoring football to boost enrollment, such as offering larger financial aid packages.</p> <p>Table 6. Admission, retention, and financial effects.</p> <p> <ephtml> &lt;table&gt;&lt;thead valign="bottom"&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Variable&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Pct. admitted&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Retention rate&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Pct. awarded financial aid&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Core revenues (millions)&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Core expenses (millions)&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Tuition&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Cost to students&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;(1)&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;(2)&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;(3)&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;(4)&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;(5)&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;(6)&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;(7)&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;/thead&gt;&lt;tbody&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Football&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;&amp;#8722;2.288&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;0.099&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;&amp;#8722;0.576&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;6.073&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;2.343&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;5.510&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;&amp;#8722;440.1&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;(3.038)&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;(1.109)&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;(1.501)&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;(3.707)&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;(2.206)&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;(535.1)&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;(846.5)&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;School FEs&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Yes&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Yes&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Yes&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Yes&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Yes&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Yes&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Yes&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Year FEs&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Yes&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Yes&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Yes&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Yes&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Yes&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Yes&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;Yes&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;Obs.&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;2,849&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;1,464&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;2,009&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;2,609&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;2,609&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;2,204&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;2,164&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td&gt;&lt;p&gt;&lt;italic&gt;R&lt;/italic&gt;&lt;sup&gt;2&lt;/sup&gt;&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;0.706&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;0.862&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;0.747&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;0.941&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;0.991&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;0.959&lt;/p&gt;&lt;/td&gt;&lt;td&gt;&lt;p&gt;0.799&lt;/p&gt;&lt;/td&gt;&lt;/tr&gt;&lt;/tbody&gt;&lt;/table&gt; </ephtml> </p> <p>3 Note: clustered (school) standard errors in parenthesis. , , and ** correspond to 0.1, 0.05, and 0.01 significance levels, respectively. Matching based on religious affiliation and lagged values of pct. admitted (except in column 1), pct. awarded financial aid (except in column 3), core revenues (except in column 4), core expenses (except in column 5), and tuition (except in columns 6 and 7).</p> <p>Although the estimated coefficients in the core revenue and core expense models should be treated with caution since they are not statistically significant, it is worth noting that the estimated coefficient on core revenue is more than twice as large as the effect on core expenses, thereby suggesting that having a football program might improve a college's financial position. Dividing the estimate coefficients on football in the core revenue and expense models by the estimated increase in enrollment reported in column 3 of Table 3 suggests that each additional student attracted by football increases core revenues by about $19,000 while increasing core expenses by only about $7,300.</p> <hd id="AN0192371729-6">Conclusion</hd> <p>Over the past 15 years, several small private colleges have added football programs, and a few have discontinued football programs. An important consideration is how football affects an institution's overall enrollment. If fielding a team of, say, 90 players displaces 90 other students who would have attended an institution, then doing so adds significant costs for coaches, equipment, and travel while adding little or no revenue. Conversely, if fielding a team of 90 players results in a substantial increase in enrollment, then the program likely covers its expenses and may help the college's overall financial position. Our results indicate that having a football team increases enrollment by more than 300 students thereby supporting the notion that football can be beneficial for college enrollments and finances. Moreover, our results indicate that fielding a football program is not inimical to academics, at least as measured by acceptance rates and retention rates. (We again acknowledge that acceptance and retention rates are affected by factors other than academic quality, so these results should be treated with caution.) In sum, having a football program appears to be a successful strategy used by some small private colleges to compete in the crowded and highly competitive higher education marketplace.</p> <hd id="AN0192371729-7">Acknowledgements</hd> <p>We would like the thank participants of the 2021 Southern Economics Association annual meeting for their helpful comments and the Berry College Richards Undergraduate Research Scholarship for their support. We are grateful to the editors and two anonymous referees for their insightful comments.</p> <hd id="AN0192371729-8">Disclosure statement</hd> <p>No potential conflict of interest was reported by the authors.</p> <ref id="AN0192371729-9"> <title> Notes </title> <blist> <bibl id="bib1" idref="ref1" type="bt">1</bibl> <bibtext> Another governing body, the National Association of Intercollegiate Athletics (NAIA), also sponsors intercollegiate sports competitions including football. The NAIA currently has approximately 240 member institutions, most of which are small, private schools with religious-oriented missions and/or denominational affiliations. NAIA members are allowed to give athletic scholarships.</bibtext> </blist> <blist> <bibl id="bib2" idref="ref2" type="bt">2</bibl> <bibtext> See, for example, Demirel ([6]).</bibtext> </blist> <blist> <bibl id="bib3" idref="ref30" type="bt">3</bibl> <bibtext> One reason that focusing on actual enrollment is more important than applications for admissions is that high school students' application behavior may change over time. For example, Soodik ([21]) reports that 36 percent of students seeking admission for fall 2015 applied to seven or more colleges whereas only 17 percent seeking admission for fall 2005 did so. Colleges may also influence the number of applications received in various ways such as by increasing or decreasing application fees, choosing to accept or not accept the Common Application, or choosing whether or not to require lengthy essays or interviews as part of the admissions process.</bibtext> </blist> <blist> <bibl id="bib4" idref="ref21" type="bt">4</bibl> <bibtext> Public universities use the Government Accounting Standards Board (GASB), while private universities use the Financial Accounting Standards Board (FASB).</bibtext> </blist> <blist> <bibl id="bib5" idref="ref16" type="bt">5</bibl> <bibtext> One important note involving the COVID-19 pandemic. Many colleges and athletic conferences cancelled athletics during the 2020 COVID-19 pandemic. In the cases where a school's football team plays in the 2019 and 2021 seasons, but not the 2020, we code 'football' as 1 for 2020, with the assumption that a team did <emph>exist</emph> that year. In cases where a team is not fielded in 2020 and 2021, we record both 2020 and 2021 as zero.</bibtext> </blist> <blist> <bibl id="bib6" idref="ref5" type="bt">6</bibl> <bibtext> For robustness, we ran some (unreported) regressions examining the total number of applications and average ACT scores of admitted students, with the caveat that these variables see a lower reporting rate in IPEDS compared to others in our sample. Though we see some limited evidence that male applicants increase in the presence of football, our results are generally insignificant.</bibtext> </blist> <blist> <bibl id="bib7" idref="ref17" type="bt">7</bibl> <bibtext> In IPEDS, core revenue includes tuition and fees, government grants, private grants, and net investment income, and core expenses include instructional expenses, research, student support, and net grant aid to students.</bibtext> </blist> <blist> <bibl id="bib8" idref="ref18" type="bt">8</bibl> <bibtext> A possible drawback of the religion dummy is that religious affiliation can function as a spectrum, with various levels of intensity. There is no way toparameterize how religious a school is. However, separating religious affiliation by faith (Catholic, Methodist, Baptist, Presbyterian, Reformed Church) does not alter the results reported below. Removing religious affiliation entirely also does not change results.</bibtext> </blist> <blist> <bibl id="bib9" idref="ref19" type="bt">9</bibl> <bibtext> 2022/23 is the most recent available. 2006/07 is the first year we could include because IPEDS reformed many of its variables in 2005 and our control variables are lagged one year.</bibtext> </blist> <blist> <bibtext> A few summary statistics warrant comment. The largest enrollment, revenue, and expenses belong to New York University, a large research university that competes at the NCAA DIII level. There are also several all-female and two all-male (Hampden-Sydney College and Wabash College) institutions, thus the percentage of students who are women ranges from 0 to 100.</bibtext> </blist> <blist> <bibtext> An 'NCAA DIII member' is defined as fielding a sport as a member of an NCAA Division III conference or as a provisional NCAA DIII member. There are also a few schools that compete in DIIIfor some sports but DI in others. For example, Johns Hopkins University is D1 for lacrosse while being DIII for football, and Colorado College is D1 for ice hockey and women's soccer, but DIII for other sports (it does not sponsor a football team). These schools are dropped from the sample.</bibtext> </blist> <blist> <bibtext> Because acceptance rates and retention rates can be affected by factors in addition to student quality, the results for these models should be interpreted cautiously and be considered only as possible indicators of academic quality if no other relevant factors have changed.</bibtext> </blist> <blist> <bibtext> Unsurprisingly, core revenues and core expenses are highly correlated (r = 0.91). However, only including one variable (revenues or expenses) does not alter the effect of the football coefficient.</bibtext> </blist> <blist> <bibtext> Additionally, the significance of the coefficient related to tuition is likely caused by a small number of research universities in the sample, which have both high tuition prices and high enrollment.</bibtext> </blist> </ref> <ref id="AN0192371729-10"> <title> References </title> <blist> <bibtext> Anderson, Michael L. 2017. " The Benefits of College Athletic Success: An Application of the Propensity Score Design." Review of Economics and Statistics 99 (1): 119 – 134. https://doi.org/10.1162/REST_a_00589.</bibtext> </blist> <blist> <bibtext> Belkin, Douglas. 2016. "To Boost Male Enrollment, Some Schools Turn to Football." The Wall Street Journal. November 14, 2016. https://<ulink href="http://www.wsj.com/articles/gender-gap-try-football-1479087946">www.wsj.com/articles/gender-gap-try-football-1479087946</ulink>.</bibtext> </blist> <blist> <bibtext> Blackwell, Matthew, Stefano Iacus, Gary King, and Giuseppe Porro. 2009. " cem: Coarsened Exact Matching in Stata." The Stata Journal 9 (4): 524 – 546. https://doi.org/10.1177/1536867X0900900402.</bibtext> </blist> <blist> <bibtext> Caudill, Steven B., Shannon Hourican, and Franklin G. Mixon. 2018. " Does College Football Impact the Size of University Applicant Pools and the Quality of Entering Students? " Applied Economics 50 (17): 1885 – 1890. https://doi.org/10.1080/00036846.2017.1380289.</bibtext> </blist> <blist> <bibtext> Collier, Trevor, Nancy Haskell, Kurt W. Rotthoff, and Alaina Baker. 2020. " The "Cinderella Effect": The Value of Unexpected March Madness Runs as Advertising for the Schools." Journal of Sports Economics 21 (8): 783 – 807. https://doi.org/10.1177/1527002520944437.</bibtext> </blist> <blist> <bibtext> Demirel, Evin. 2013. "The D-III revolution: How America's most violent game may be saving liberal arts colleges." SB Nation. October 1, 2013. https://<ulink href="http://www.sbnation.com/longform/2013/10/1/4786810/diii-football-revolution">www.sbnation.com/longform/2013/10/1/4786810/diii-football-revolution</ulink>.</bibtext> </blist> <blist> <bibtext> Eggers, Austin F., Peter A. Groothuis, and Parker T. Redding. 2021. " The Flutie Effect: The Influence of College Football Upsets and National Championships on the Quantity and Quality of Students at a University." International Journal of Sport Finance 16 (2): 59 – 68. https://doi.org/10.32731/jsf.2021.a927049.</bibtext> </blist> <blist> <bibtext> Eggers, Austin F., Peter A. Groothuis, Parker Redding, Kurt W. Rotthoff, and Michael Solimini. 2019. " The Negative Effect of NCAA Football Bowl Bans on University Enrolment and Applications." Applied Economics 51 (54): 5870 – 5877. https://doi.org/10.1080/00036846.2019.1630708.</bibtext> </blist> <blist> <bibtext> Eggers, Austin F., Peter A. Groothuis, Parker Redding, Kurt W. Rotthoff, and Michael Solimini. 2020. " Universities Behaving Badly: The Impact of Athletic Malfeasance on Student Quality and Enrollment." Journal of Sports Economics 21 (1): 87 – 100. https://doi.org/10.1177/1527002519859416.</bibtext> </blist> <blist> <bibtext> Holmes, Jessica A., James A. Meditz, and Paul M. Sommers. 2008. " Athletics and Alumni Giving: Evidence from a Highly Selective Liberal Arts College." Journal of Sports Economics 9 (5): 538 – 552. https://doi.org/10.1177/1527002507313896.</bibtext> </blist> <blist> <bibtext> Humphreys, Brad R. 2006. " The Relationship Between big-Time College Football and State Appropriations for Higher Education." International Journal of Sport Finance 1 (2): 119 – 128.</bibtext> </blist> <blist> <bibtext> Humphreys, Brad R., and Michael Mondello. 2007. " Intercollegiate Athletic Success and Donations at NCAA Division I Institutions." Journal of Sport Management 21 (2): 265 – 280. https://doi.org/10.1123/jsm.21.2.265.</bibtext> </blist> <blist> <bibtext> Iacus, Stefano M., Gary King, and Giuseppe Porro. 2011. " Multivariate Matching Methods That are Monotonic Imbalance Bounding." Journal of the American Statistical Association 106 (493): 345 – 361. https://doi.org/10.1198/jasa.2011.tm09599.</bibtext> </blist> <blist> <bibtext> Jacob, Brian, Brian McCall, and Kevin Stange. 2018. " College as Country Club: Do Colleges Cater to Students' Preferences for Consumption? " Journal of Labor Economics 36 (2): 309 – 348. https://doi.org/10.1086/694654.</bibtext> </blist> <blist> <bibtext> Johnson, Candon, and Bryan C. McCannon. 2022. " Athletics and Admissions: The Impact of the Penn State Football Scandal on Student Quality." Journal of Sports Economics 23 (2): 200 – 221. https://doi.org/10.1177/15270025211039444.</bibtext> </blist> <blist> <bibtext> Jones, Willis A. 2015. " High-level Football and Appropriations to Universities: Are Sports-Crazed Representatives Responsive to NCAA Divisional Affiliation? " Journal of Education Finance 40 (4): 438 – 455. https://doi.org/10.1353/jef.2015.a596198</bibtext> </blist> <blist> <bibtext> Perez, Stephen J. 2012. " Does Intercollegiate Athletics Draw Local Students to a University? " Journal of Sports Economics 13 (2): 198 – 206. https://doi.org/10.1177/1527002511404509.</bibtext> </blist> <blist> <bibtext> Pope, Devin G., and Jaren C. Pope. 2009. " The Impact of College Sports Success on the Quantity and Quality of Student Applications." Southern Economic Journal 75 (3): 750 – 780. https://doi.org/10.1002/j.2325-8012.2009.tb00930.x.</bibtext> </blist> <blist> <bibtext> Pope, Devin G., and Jaren C. Pope. 2014. " Understanding College Application Decisions: Why College Sports Success Matters." Journal of Sports Economics 15 (2): 107 – 131. https://doi.org/10.1177/1527002512445569.</bibtext> </blist> <blist> <bibtext> Segura, Jerome, III, and Jonathan Willner. 2019. " Athleticism in NCAA D-III: It Ain't Only Football That Matters." Journal of Sports Economics 20 (7): 929 – 958. https://doi.org/10.1177/1527002519832070.</bibtext> </blist> <blist> <bibtext> Soodik, Nicholas. 2017. "College Choice Overload." Insidehighered. com. December 3, 2017.. https://<ulink href="http://www.insidehighered.com/admissions/views/2017/12/04/high-school-students-are-applying-too-many-colleges-essay">www.insidehighered.com/admissions/views/2017/12/04/high-school-students-are-applying-too-many-colleges-essay</ulink>.</bibtext> </blist> <blist> <bibtext> Tucker, Irvin B. 2004. " A Reexamination of the Effect of big-Time Football and Basketball Success on Graduation Rates and Alumni Giving Rates." Economics of Education Review 23 (6): 655 – 661. https://doi.org/10.1016/j.econedurev.2004.03.001.</bibtext> </blist> <blist> <bibtext> Turner, Sarah E., Lauren A. Meserve, and William G. Bowen. 2001. " Winning and Giving: Football Results and Alumni Giving at Selective Private Colleges and Universities." Social Science Quarterly 82 (4): 812 – 826. https://doi.org/10.1111/0038-4941.00061.</bibtext> </blist> <blist> <bibtext> Weaver, Karen. 2024. "In A Rising Tide of College Closures, Impact On Division III Athletics Becomes Clear." Forbes. April 30, 2024. https://<ulink href="http://www.forbes.com/sites/karenweaver/2024/04/30/in-a-rising-tide-of-college-closures-impact-on-division-iii-athletics-becomes-clear/">www.forbes.com/sites/karenweaver/2024/04/30/in-a-rising-tide-of-college-closures-impact-on-division-iii-athletics-becomes-clear/</ulink>.</bibtext> </blist> <blist> <bibtext> Willner, Jonathan. 2019. " Private Universities and NCAA D-III Athletics as a General Recruiting Tool." International Advances in Economic Research 25 (3): 293 – 307. https://doi.org/10.1007/s11294-019-09745-4.</bibtext> </blist> </ref> <aug> <p>By Clay Collins and E. Frank Stephenson</p> <p>Reported by Author; Author</p> </aug> <nolink nlid="nl1" bibid="bib11" firstref="ref6"></nolink> <nolink nlid="nl2" bibid="bib16" firstref="ref7"></nolink> <nolink nlid="nl3" bibid="bib23" firstref="ref8"></nolink> <nolink nlid="nl4" bibid="bib22" firstref="ref9"></nolink> <nolink nlid="nl5" bibid="bib10" firstref="ref10"></nolink> <nolink nlid="nl6" bibid="bib12" firstref="ref11"></nolink> <nolink nlid="nl7" bibid="bib18" firstref="ref13"></nolink> <nolink nlid="nl8" bibid="bib19" firstref="ref14"></nolink> <nolink nlid="nl9" bibid="bib17" firstref="ref15"></nolink> <nolink nlid="nl10" bibid="bib15" firstref="ref20"></nolink> <nolink nlid="nl11" bibid="bib20" firstref="ref24"></nolink> <nolink nlid="nl12" bibid="bib25" firstref="ref26"></nolink> <nolink nlid="nl13" bibid="bib24" firstref="ref32"></nolink> <nolink nlid="nl14" bibid="bib14" firstref="ref33"></nolink> <nolink nlid="nl15" bibid="bib13" firstref="ref44"></nolink>
Header	DbId: eric DbLabel: ERIC An: EJ1500944 AccessLevel: 3 PubType: Academic Journal PubTypeId: academicJournal PreciseRelevancyScore: 0
IllustrationInfo
Items	– Name: Title Label: Title Group: Ti Data: How Does Having a Football Program Affect Enrollment at Small Private Colleges? – Name: Language Label: Language Group: Lang Data: English – Name: Author Label: Authors Group: Au Data: <searchLink fieldCode="AR" term="%22Clay+Collins%22">Clay Collins</searchLink><br /><searchLink fieldCode="AR" term="%22E%2E+Frank+Stephenson%22">E. Frank Stephenson</searchLink> – Name: TitleSource Label: Source Group: Src Data: <searchLink fieldCode="SO" term="%22Education+Economics%22"><i>Education Economics</i></searchLink>. 2026 34(2):311-320. – Name: Avail Label: Availability Group: Avail Data: Routledge. Available from: Taylor & Francis, Ltd. 530 Walnut Street Suite 850, Philadelphia, PA 19106. Tel: 800-354-1420; Tel: 215-625-8900; Fax: 215-207-0050; Web site: http://www.tandf.co.uk/journals – Name: PeerReviewed Label: Peer Reviewed Group: SrcInfo Data: Y – Name: Pages Label: Page Count Group: Src Data: 10 – Name: DatePubCY Label: Publication Date Group: Date Data: 2026 – Name: TypeDocument Label: Document Type Group: TypDoc Data: Journal Articles<br />Reports - Evaluative – Name: Audience Label: Education Level Group: Audnce Data: <searchLink fieldCode="EL" term="%22Higher+Education%22">Higher Education</searchLink><br /><searchLink fieldCode="EL" term="%22Postsecondary+Education%22">Postsecondary Education</searchLink> – Name: Subject Label: Descriptors Group: Su Data: <searchLink fieldCode="DE" term="%22Small+Colleges%22">Small Colleges</searchLink><br /><searchLink fieldCode="DE" term="%22Private+Colleges%22">Private Colleges</searchLink><br /><searchLink fieldCode="DE" term="%22Enrollment%22">Enrollment</searchLink><br /><searchLink fieldCode="DE" term="%22Team+Sports%22">Team Sports</searchLink><br /><searchLink fieldCode="DE" term="%22College+Athletics%22">College Athletics</searchLink><br /><searchLink fieldCode="DE" term="%22Gender+Differences%22">Gender Differences</searchLink> – Name: DOI Label: DOI Group: ID Data: 10.1080/09645292.2025.2501068 – Name: ISSN Label: ISSN Group: ISSN Data: 0964-5292<br />1469-5782 – Name: Abstract Label: Abstract Group: Ab Data: This paper examines the effects of fielding college football on enrollment and other aspects of NCAA Division III (DIII) institutions. Unlike larger Division I universities, DIII schools, which are primary private colleges, do not award athletic scholarships. Using two-way fixed effect regressions and coarsened exact matching approaches, we examine the presence of a football program on colleges' enrollment, admissions standards, and finances. We find that having a football team is associated with significant increases in total enrollment as well as increases in both male and female enrollment. We find no evidence that football negatively affects admissions rate or college revenues. – Name: AbstractInfo Label: Abstractor Group: Ab Data: As Provided – Name: DateEntry Label: Entry Date Group: Date Data: 2026 – Name: AN Label: Accession Number Group: ID Data: EJ1500944
PLink	https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=eric&AN=EJ1500944
RecordInfo	BibRecord: BibEntity: Identifiers: – Type: doi Value: 10.1080/09645292.2025.2501068 Languages: – Text: English PhysicalDescription: Pagination: PageCount: 10 StartPage: 311 Subjects: – SubjectFull: Small Colleges Type: general – SubjectFull: Private Colleges Type: general – SubjectFull: Enrollment Type: general – SubjectFull: Team Sports Type: general – SubjectFull: College Athletics Type: general – SubjectFull: Gender Differences Type: general Titles: – TitleFull: How Does Having a Football Program Affect Enrollment at Small Private Colleges? Type: main BibRelationships: HasContributorRelationships: – PersonEntity: Name: NameFull: Clay Collins – PersonEntity: Name: NameFull: E. Frank Stephenson IsPartOfRelationships: – BibEntity: Dates: – D: 01 M: 01 Type: published Y: 2026 Identifiers: – Type: issn-print Value: 0964-5292 – Type: issn-electronic Value: 1469-5782 Numbering: – Type: volume Value: 34 – Type: issue Value: 2 Titles: – TitleFull: Education Economics Type: main
ResultId	1