Forschungsinstitut zur Zukunft der Arbeit Institute for the Study of Labor DISCUSSION PAPER SERIES Sexual Violence, Title IX and Women’s College Enrollment IZA DP No. 10345 November 2016 Dave E. Marcotte Jane E. Palmer
Forschungsinstitut zur Zukunft der ArbeitInstitute for the Study of Labor
DI
SC
US
SI
ON
P
AP
ER
S
ER
IE
S
Sexual Violence, Title IX and Women’sCollege Enrollment
IZA DP No. 10345
November 2016
Dave E. MarcotteJane E. Palmer
Sexual Violence, Title IX and Women’s College Enrollment
Dave E. Marcotte American University
and IZA
Jane E. Palmer
American University
Discussion Paper No. 10345 November 2016
IZA
P.O. Box 7240 53072 Bonn
Germany
Phone: +49-228-3894-0 Fax: +49-228-3894-180
E-mail: [email protected]
Any opinions expressed here are those of the author(s) and not those of IZA. Research published in this series may include views on policy, but the institute itself takes no institutional policy positions. The IZA research network is committed to the IZA Guiding Principles of Research Integrity. The Institute for the Study of Labor (IZA) in Bonn is a local and virtual international research center and a place of communication between science, politics and business. IZA is an independent nonprofit organization supported by Deutsche Post Foundation. The center is associated with the University of Bonn and offers a stimulating research environment through its international network, workshops and conferences, data service, project support, research visits and doctoral program. IZA engages in (i) original and internationally competitive research in all fields of labor economics, (ii) development of policy concepts, and (iii) dissemination of research results and concepts to the interested public. IZA Discussion Papers often represent preliminary work and are circulated to encourage discussion. Citation of such a paper should account for its provisional character. A revised version may be available directly from the author.
IZA Discussion Paper No. 10345 November 2016
ABSTRACT
Sexual Violence, Title IX and Women’s College Enrollment Sexual violence has long been a problem on college campuses, yet federal policies to protect students have largely been ineffectual. Spurred by student grievances, the U.S. Department of Education’s Office of Civil Rights recently began investigating how sexual assault cases were handled at a number of institutions under the Title IX provisions of the Education Amendments of 1972. These investigations focus attention on specific colleges’ responses to cases of sexual violence and raise the specter that these institutions may fail to properly investigate allegations or punish perpetrators. In this paper, we examine the implications of these investigations on college enrollment, particularly for women. We combine institution-level panel data on enrollment by age and gender, with information on Title IX investigations to study changes in women’s college enrollment. We estimate that enrollment of women at colleges under Title IX investigation declined by 16 to 22 percent. The declines are consistent with both declining matriculation and retention of female students. JEL Classification: I24, J16 Keywords: college, educational attainment, sexual assault Corresponding author: Dave E. Marcotte School of Public Affairs American University 4400 Massachusetts Avenue NW Washington, DC 20016-8070 USA E-mail: [email protected]
IZA Discussion Paper No. 10345 November 2016
NON-TECHNICAL SUMMARY
We assess the impact of high profile federal investigations of the handling of sexual assault claims at colleges and universities in the United States. Notably, we examine the impact of these investigations on student enrollment, rather than on outcomes one might expect would be most directly affected, such as rates of sexual assaults, supports for victims, and punishments for perpetrators. Sexual victimization and how a university handles those claims affect students beyond those directly involved. Our focus is broad because the climate and perceptions of a campus can affect the student body generally, and especially female students. We find that enrollment of women of all ages drops following a Title IX investigation, with no similar changes for men. We estimate that the college enrollment decisions of more than 30,000 women changed as a result of a Title IX investigation, and more than two-thirds of this is due to declining retention at colleges under investigation. Our paper illustrates that attention to the appearance of insufficient protections and processes at specific colleges had clear effects on the enrollment of women. That many women “voted with their feet” by leaving or avoiding institutions under investigation makes clear the stakes are high for colleges and universities. Since female students make up a sizeable majority of undergraduates, assuring that their safety and protection are paramount concerns will be essential for any college or university that hopes to thrive.
1
Collegeanduniversitycampuseshavelongbeenplaceswheresexual
violenceagainststudentsisendemic.Thoughnotoriouslydifficulttomeasure,1
studiesconsistentlyestimatethatapproximately18to20percentofundergraduate
womenarevictimsofsexualassaultduringcollege(NationalInstituteofJustice,
2008).Whilehighprofilecaseshaverecentlyfocusedattentionontheissue,2sexual
violenceoncollegecampusesisnotnew.Indeed,themostimportantfederal
legislationaimedatprotectingcollegestudents,theCleryAct,waspassedmorethan
25yearsago.Unfortunately,thereislittleevidencethatfederalpoliciesintendedto
reducetheincidenceofsexualassaultorprotectvictimshavebeeneffective.
Thepersistenceoftheproblemalongwithperceptionsofinstitutionalfailure
ledstudentactivistsinthe2000stopushtheU.S.DepartmentofEducationto
investigateclaimsthatthefailuretoprotectvictimsofsexualassaultatsome
collegesresultedinunequalaccesstoeducationalopportunityforwomen.The
adventofthisuseofequalaccessrequirementsoffederallawasameanstobring
attentionandremediationhasresultedinasubstantialshiftinfederalpolicyand
publicattentiontotheissue.Inthispaperwedescribethegrowthintheuseof
investigationsofindividualcollegesanduniversitiesbytheU.S.Departmentof
Education’sOfficeofCivilRights(OCR)inresponsetocomplaintsfiledunderTitleIX
oftheEducationAmendmentsof1972--establishedtoprohibitdiscriminationon
1SeePalmerandPerrotti(2016)andFedina,Holmes&Backes,2016fordiscussionsofmeasurementissues.2ExamplesincludeaColumbiaUniversitystudent’syears-longprotestbycarryinghermattressasasymbolofhervictimizationandtheuniversity’sresponse(e.g.NYTimes,Sept.21,2014)andtheturmoilattheUniversityofVirginiafollowingalater-retractedarticleinRollingStone.
2
thebasisofsexinallactivitiesatcollegesanduniversitiesbenefitingfromfederal
funding.3
Becausetheseinvestigationsarebasedoncomplaintsthatinstitutionsdid
notproperlyinvestigateoradjudicateinstancesofsexualviolence,theymaysignal
tocurrentandprospectivestudentsthatacollegeoruniversityisnotsafeordoes
nottakesexualassaultseriously.Inthispaper,weassessthispossibilityby
examiningwhetherenrollmentatcollegesanduniversitiesisaffectedbythe
initiationofaTitleIXinvestigationforfailuretohandleacredibleallegationof
sexualviolence.Sincewomenareoverwhelminglythevictimsofsexualviolence,we
examinewhetherenrollmentofwomendecreasesataninstitutioncomparedto
menoverthesameperiod.
Todothis,wecompileddataonenrollmentoffull-timeundergraduatesby
genderandageatallbachelordegreegrantingpublicandnot-for-profitcollegesand
universitiesfrom2005through2014.WemergeddataonwhereandwhenOCR
openedTitleIXinvestigations,obtainedfromtheChronicleofHigherEducation’s
TitleIXTrackerdatabase.4Usingthesecombineddata,weestimateevent-study
modelsofenrollmentchangeswithinauniversitybygenderandagegroups
followingtheopeningofaninvestigation.Wefindthatenrollmentofwomenofall
agesdropsfollowingaTitleIXinvestigation,withnosimilarchangesformen.The
enrollmentdeclinesforyoung(17to19yearsold)undergraduatewomenmaybe
3Thiseffectivelymeansallcollegesanduniversities,becauseoftheimportanceoffederalstudentfinancialaid,includingPellGrantsandsubsidizedloans.4http://projects.chronicle.com/titleix/
3
duetodecliningapplicationandmatriculationrates.Buttheslightlyhigherdeclines
amongwomenintheirearly20sislikelyduetodropoutandtransfer.
Webeginwithanintroductiontotheissue,includingdescribingthescope
anddimensionsofsexualviolenceoncampus.Wethenbrieflydescribethe
institutionalandjudicialprocessesthroughwhichvictimscanseekredress,along
withthedevelopmentofTitleIXinvestigationsofstudentallegationsofundue
treatment.
Background
Campussexualassaulthaslongbeenthesubjectofstudybyresearchers
fromthefieldsofcriminology,psychology,socialworkandrelatedfields.Estimates
ofsexualvictimizationcansometimesbedifficulttocompare,andcancause
confusionamongnon-researchers.Thedifficultiesarisebecauseresearchersuse
differentreportingperiods,havedifferentratesofresponseinsurveys(where
subjectsareaskednecessarilysensitiveandpotentiallytraumaticquestions),and
researchersmeasuresexualvictimizationdifferently(Fedina,Holmes&Backes,
2016;Palmer&Perrotti,2016).
Forexample,thefirstmajorstudyonsexualassaultvictimizationamong
collegewomenwasKoss,GidyczandWisniewski(1987),inwhichtheauthors
estimatedthatmorethanoneinfourcollegewomenhadexperiencedattemptedor
completedrape.WhileKossetal.(1987)measuredvictimizationsincetheageof14,
notjustsincedateofcollegematriculation,theirestimateswerequitesimilarto
laterestimatesfocusingonlyonvictimizationduringcollege.Forexample,the
4
NationalCollegeWomenSexualVictimizationstudyestimatedthat20–25%of
womenwouldexperienceattemptedorcompletedrapewhileattendingcollege
(Fisher,Cullen&Turner,2000).Morerecently,Krebs,etal.(2007)foundthat,
withinrandomsamplesattwolargepublicuniversities,approximately1in4
womenexperiencedattemptedorcompletedsexualassaultusingforceor
incapacitationsinceenteringcollege.5Inthesameyear,Kilpatricketal.(2007)
reportedthat5.2%ofanationalsampleof2,000collegewomenexperienced
completedrapeusingforceorincapacitationinthepastacademicyear,whichover
fouryearswouldbecomparabletoearlierestimates.
Notethatthereferenceperiod(since14,sinceenteringcollege,past
academicyear)andhowvictimizationwasmeasured(sexualassault,attempted
rape,completedrape)variesforeachstudy.Inaddition,underreportingmaybea
largerissueforsurveysthataskaboutrape,ratherthanabroadermeasurelike
unwantedsexualactivityorsexualassault.Still,nomatterthemeasure,the
estimatesaredisturbinglyhigh.
Althoughimportantissuesinconsistentlymeasuringsexualviolenceacross
studiesremain(Fedina,etal.,2016;Kruttschnitt,Kalsbeek&House,2014;Palmer&
Perrotti,2016),thereissolidevidenceontheeffectsofsexualviolenceonvictims.
Traumaassociatedwithexperiencingsexualassaultcanresultindepression,
anxiety,post-traumaticstressdisorderandarangeofphysicalhealthissuesthat
5TheKossetal.andFisheretal.studiesfocusedonmeasuringincidentsthatwouldlegallybedefinedasrape.TheKrebsetal.studymeasuredincidencesof“sexualassault”,whichisbroaderthan“rape,”butstillincludesillegalbehaviorsrelatedtounwantedsexualactivity(includingrapeandattemptedrape)withoutconsent.
5
mayimpactastudent’sabilitytocompleteadegree(Campbell,Dworkin,&Cabral,
2009;Kilpatrick,etal.,2007).
PolicyContext
Thehistoryoffederallegislationtoprotectcollegestudentsfromcriminal
victimizationbeginswiththeCleryActof1990.ThemainprovisionofClery
requirescollegesanduniversitiestomakecrimestatisticspubliclyavailableand
issuetimelywarningsofanyongoingthreatstothecampuspopulation.Institutions
arealsorequiredtoprovidevictimswithsupportintheaftermathofthecrimeand
implementpoliciesandprocedurestoensurethatvictimsaretreatedfairlyduring
crimereportingandadjudicationprocesses(USDepartmentofEducation,2016).
Yet,institutionshaveadisincentivetoencouragestudentstoreport,since
crimestatisticsarepubliclyavailableandmayaffectpublicperceptionsofthe
institutionandfutureenrollment.Forotherreasons,too,criticshavedescribedthe
CleryActassymbolicratherthansubstantial(Fisher,Hartman,Cullen&Turner,
2002).Thestatisticsthatcollegesanduniversitiesreportarenotinclusiveofall
crimes,becausetheyareonlyrequiredtoreportcrimesthatoccurredonornear
campusorcampusproperty.Inaddition,thefailuretodistinguishbetweencrimes
inwhichstudents(oremployees)arevictimsorperpetratorshasmeantthat
reporteddatadoesnotdistinguishbetweendangerouscampusesanddangerous
citiesinwhichcollegesaresituated.Afurthercomplicationarisesbecauseofthe
generalreluctanceofstudentstoreportsexualassaulttoauthorities(Fisheretal.,
2000).Asaresult,highersexualassaultstatisticsmaybeindicatorsofacampus
6
climateconducivetoincreasingstudentwillingnesstoreportassaults,ratherthan
underlyingvictimizationrates(Cantalupo,2011;Palmer&Alda,2016).
TheViolenceAgainstWomenAct(VAWA),firstauthorizedin1994aspartof
theViolentCrimeControlandLawEnforcementAct(H.R.3355),supplemented
federallawtoprotectstudents(andothers)oncollegeanduniversitycampuses
(Dunn,2013;Schroeder,2013).In1999,thefirstCampusGrants,usingVAWA
funds,wereawardedtosupportinitiativestopreventandrespondtoviolentcrimes
againstwomenoncollegecampuses.Inadditiontograntingfundssuchprograms,
thelawalsorequirescollegesanduniversitiestooffersexualassaultprevention
programsforallincomingstudentsandnewemployees.Inaddition,VAWAoutlines
severalproceduralrequirementsrelatedtoadjudicationprocessincasesofsexual
violencetoassuretransparency,andalsoconfidentialityforthevictim.
TheUseofTitleIXtoCombatSexualViolence
In2011,theU.S.DepartmentofEducation’sOfficeofCivilRightsreleaseda
“DearColleagueLetter”(DCL)6thatclarifiedtherequirementsofTitleIXinthe
contextofsexualviolenceatallpublicandprivateschools,collegesanduniversities
thatreceivefederalfinancialassistance.TitleIXisacivilrightslawpassedaspartof
theHigherEducationAmendmentsof1972toassureequalaccesstoeducation.The
initialimplementationofthislawfocusedonfemalestudents’accesstosportsand
equitableathleticfacilities.Overtime,thelawwasinterpretedtoincludeother
formsofsex-specificdiscriminationthataffectfemalestudents’equalaccessto
6Seehttp://www2.ed.gov/about/offices/list/ocr/letters/colleague-201104.html
7
education.TheOCRisresponsibleforinvestigatinganycomplaintthatan
educationalinstitutionviolatedTitleIX.
TheDCLclarifiedthatTitleIXrequiresapromptandequitableinvestigation
iftheschoolisawareofanallegationofsexualharassmentorsexualviolence.Ifa
school“knowsorreasonablyshouldknow”ofaninstanceofsexualharassmentor
sexualviolence,itisrequiredtoimmediatelycompletea“prompt,thoroughand
impartial”investigation.Anyadjudicationprocessmustuseapreponderanceof
evidencestandard,whichisalowerburdenofproofthanthebeyondareasonable
doubtstandardusedforsexualassaultcasesinthecriminaljusticesystem.Ifthe
adjudicationprocessresultsinafindingthatsexualviolenceoccurred,the
institutionmust“takeimmediateactiontoeliminatethehostileenvironment,
preventitsreoccurrence,andaddressitseffects”(DCL,2011,p.15).Ifavictim
believesanyofthesestepswerenotfollowedadequately,sheorhecanelecttofilea
complaintwithOCRwithin180days.IfOCRdeterminesthecollegeoruniversity
doesnotrespondinpromptlyandequitably,itmayindicatetheinstitutionviolated
TitleIX.Schoolsthatarefoundtobenon-compliantcanfacefines,losefederal
fundingandberequiredtotakestepstoremedyorcorrecttheirresponsetosexual
violence.
TitleIXandPublicAwareness
TheOCR’sactionwaslargelyinresponsetogrowingpublicattentiontothe
issueofcampussexualassaultandhowitwasbeinghandled.In2010,National
PublicRadioandtheCenterforPublicIntegrityreleasedaseriesonfailureby
collegestoprotectwomenfromcampusrape.Theirinvestigationsfoundthat
8
collegesrarelyexpelsexualassaultperpetratorsandpreventionprogramsputthe
responsibilityonwomentopreventrape.Theypresentedtheirfindingsdirectlyto
theAssistantSecretaryforOCR,whocommittedtomoreaggressivelyaddress
sexualviolenceoncollegecampuses(NPR,2010).Morerecentexamplesofpublic
attentionincludethe2015documentaryfilmTheHuntingGround,whichoffered
severalsurvivors’perspectivesofhowtheiruniversitiesdidnotproperlyrespondto
theirallegationsofsexualassault,andfollowstwoactivistswhobegantoteachtheir
peershowtofileacomplaintwiththeOCR.Thesameyearawidelyreadbookonthe
handlingofsexualviolencecasesbytheUniversityofMontanaandthelocalpolice
helpedtofurtherraisethepublic’sawarenessoftheinstitutionalprotectionismand
inadequateresponsethatmanystudentsfacewhentheyreportsexualviolenceto
universityadministrators(Krakauer,2015).
Formany,therealizationthatsexualviolencewasendemiconcollege
campusesandthatinstitutionswerenotproperlyrespondingtothesecrimes,was
somewhatofashocktothegeneralsenseofrealadvancesforwomeninhigher
education.Theshareofwomenenrollinginandgraduatingfromcollegehaslong
beengrowing.Women’sundergraduateenrollmentsurpassedmen’senrollmentin
thelate1970sandby2009,therewereapproximately132womenenrolledin
Americancollegesanduniversitiesforeach100men(Audetal.2011).However,the
campussettingisauniqueplaceforbothraisingtheriskofsexualassaultandfor
cloudingtheprocessofprotectingvictimsandpunishingcriminals.
TheOCR’sincreaseinTitleIXinvestigationshavefocusedsubstantial
attentiononissuesofclimateandsafetyforwomen.Eachnewinvestigationis
9
widelycoveredinnationalnews,andtheChronicleofHigherEducationhasdevoted
specialattentiontoeachcase.7InMayof2014,forthefirsttimeever,OCRmade
publicthelistof55highereducationalinstitutionswithopenTitleIXinvestigations
relatedtosexualviolence.AsofSeptember2016,thereare277opensexualviolence
investigationsat214collegesanduniversities.
UnlikeClery,whichprovidesgeneralinformationaboutcrimestatistics,Title
IXinvestigationshavefocusedattentiononsexualassaultatspecificinstitutions,
andhowtheseschoolshandlevictims’complaints,andhowtheytreatoffenders.
Consequently,TitleIXinvestigationsmaycauserealconcernamongprospectiveand
currentstudentsaboutschoolclimate.Thesecasesmaybeseenaswarningsigns
abouttheextentofsexualvictimizationonagivencampusandthedegreetowhich
students’complaintsaretakenseriouslyandjudiciouslyprocessed.Below,we
describeourmethodsforexaminingwhethertheopeningofaTitleIXinvestigation
affectsenrollmentatacollegeoruniversity.
DataandModel:
TostudytheimpactofaTitleIXinvestigationonstudentenrollment,weuse
datafromtheIntegratedPostsecondaryEducationDataSystem(IPEDS)for
academicyears2006-07through2013-14.Werestrictouranalysestopublicand
non-profitinstitutionsthatgrantbachelor’sdegrees.Therefore,weomitallfor-
profitandtwo-yearcolleges.TheIPEDSinstitutionaldataincludesinformationon
thecostofattendanceineachyearincludingtuitionandfeesandfinancialaid,along
withotherfactorsthatwouldaffectenrollment.Theseincludeselectivityas7http://projects.chronicle.com/titleix/
10
measuredbyadmissionrateandgraduationrates,measuredastheproportionof
matriculatingfreshmenreceivingabachelor’sdegreewithinsixyears.Wealso
includemeasuresofuniversitytypes:distinguishingbetweenresearch-intensive
(Ph.D.granting)universities,andthoseofferingmainlymaster’sdegreesastheir
highestdegrees,separatefromtraditionalbachelor’sgrantingliberalartscolleges.
Onallofthesedimensionswedistinguishbetweennon-profitandpublic
universities.
TheIPEDSalsoprovidesdetailedenrollmentdatabyage,gender,andlevelof
study(undergraduatevs.graduate)foreachinstitution.Thisenrollment
(headcount)dataisrecordedintheFallofanacademicyearandisusuallyreported
annually,howeveritisonlymandatorybiennially.So,forsomesmallercollegeswe
observegroup-specificenrollmentonlyeveryotheryear.8Wecombinethe
institutionalandenrollmentdatatoconstructadatasetofyear-specificenrollment
inallpublicandnon-profitcollegesthatgrantbachelorsdegrees,byageandgender
between2006and2014.
Tothisdataset,weaddinformationonTitleIXinvestigationsfromthe
ChronicleofHigherEducation’sTitleIXTrackerdatabase,whichcompiles
informationfromFreedomofInformationActrequestsoftheU.S.Departmentof
Education.9TheinformationoncasesintheTitleIXTrackervariesbyinstitution.
However,allcasefilesincludeadateonwhichtheOCRdeterminedthatacomplaint
hadmeritandopenedaninvestigationintothehandlingofacomplaintofsexual
violencebyaninstitution.8Wediscusstheimplicationsofthisforourempiricalmodelsbelow.9http://projects.chronicle.com/titleix/
11
TomeasureimpactsofTitleIX,wegenerateindicatorsofwhetheracollege
oruniversityhadaninvestigationopeninanacademicyear.Weassignan
investigationtoanacademicyearifOCRnotifiedtheinstitutionbyMayofthatyear.
So,wewouldclassifyanoticeissuedinJune2011asoccurringinthe2011-12
academicyear.Werecognizethatanindicatorvariableofacaseopenedinan
academicyearisacrudemeasureofenrollmenteffects:Itmighttaketimefor
matriculatingorprospectivestudentstoreactorcurrentstudentstotransferifa
TitleIXinvestigationwasasourceofconcern.Or,itcouldbethatcasesofsexual
violencethatarebroughttotheU.S.DepartmentofEducationaresourcesofupsetat
auniversitybeforeanyTitleIXinvestigationislaunched.Tomeasurethese
possibilities,weestimatemodelswithbothleadingandlaggingindicatorsofTitleIX
investigations.
WeusethesedatatoestimatetheimpactofTitleIXinvestigationson
enrollmentusingmodelsofthefollowingtype:
!"#"$%!!&'( = * + ,-.&( + /'012!34.&('
+ 5' + 6( + !& + 7&'( !
wherethedependentvariableisthelogofenrollmentoffull-timestudentsatschool
s,inyeartforeachoffourgroups(g).Thegroupsare1)Womenfrom17-19years
old,2)Menfrom17-19yearsold,3)Women20to24yearsold,and4)Menfrom20
to24yearsold.Enrollmentisloggedbecauseitishighlypositivelyskewed.We
includeaseriesofcontrolvariables(Xst)likelytoaffectenrollment,suchastuition
andfeesandthepercentofstudentsreceivingaid;theadmissionand(6-year)
graduationratesandtheshareofallstudentsinundergraduateversusgraduate
12
programs.TitleIXstisanindicatorofwhetherinstitutionswasunderinvestigationin
academicyeart.Thecoefficientsofinterestarethed’swhichmeasureenrollment
changesforeachgroupcoincidentwithaTitleIXinvestigation.Wecontrolforboth
yearandinstitutionfixedeffects,sothed’smeasurechangesineachgroup’s
enrollmentwithinaninstitutionattheoutsetofaninvestigation,netofyearand
groupspecificenrollmenteffectsmoregenerally.Allstandarderrorsareclustered
atthecollege/universitylevel,andwerestrictouranalysestocollegeswithatotal
enrollmentofmorethan100students.
Ourmodelsaresimilartothestandarddifference-in-differences
parameterizationofamodelwhereinonecomparesgroupsaffected/unaffectedby
anevent/policy.Sincewomenareoverwhelminglythevictimsofsexualviolence,
andthevastmajorityofTitleIXinvestigationsareinresponsetocasesbroughtby
women,onemightexpectdifferentresponsesfromfemaleandmalestudents.
However,theenrollmentofmalestudentsmayalsobeaffectedbythenegative
attentionbroughtontoacollegebyaninvestigation.So,weestimateenrollment
responsesforallgroups,andtestforbetween-groupdifferences.
Twoillustrativecasesofhowfemaleandmalestudentsmightrespondtoa
TitleIXinvestigationareMichiganStateandtheUniversityofMontana.Figure1
showsenrollmentofyoungwomenandmenatMichiganState.BeforetheOCR
openedaTitleIXinvestigationforMichiganState’shandlingofasexualassaultofa
student(representedbytheverticalline),theundergraduatestudentbodywas
majorityfemale.Leadinguptotheinvestigation,thenumberoffemalestudentsfell
(byabout800,or10percent),whilethenumberofmalestudentsgrew.Thisisin
13
contrasttopatternsattheUniversityofMontana(Figure2),whereenrollmentfell
forbothmenandwomen.Montanawasthesubjectofintensemediascrutiny
followingtwoaccusationsofrapeagainstfootballplayers(seeKrakauer,2015).
Figures1and2illustrateanotherimportantissue.BecauseTitleIX
investigationsareinitiatedafteranunsatisfactorycampusadjudicationprocess,
theylagassaultsandthepublicityandproteststhatcanembroilacollegeintheir
wakeandduringtheprocess.Asaresult,wealsoestimatemodelsthatinclude
indicatorsfortheyearsbeforeaTitleIXinvestigation.Wealsoincludeanindicator
oftheyearafter,toassesswhetherenrollmenteffectsareshort-lived.
Figure1
6000
6500
7000
7500
Men
7600
7800
8000
8200
8400
Wom
en
2006 2008 2010 2012 2014Academic Year
Women MenVertical line indicates date of Title IX case opening
Michigan State Univ.FT Undergrads < 20 yrs. Old - by Gender
14
Figure2
1100
1200
1300
1400
Men
1500
1600
1700
1800
1900
Wom
en
2006 2008 2010 2012 2014Academic Year
Women MenVertical line indicates date of Title IX case opening
Univ. of MontanaFT Undergraduates < 20 yrs. Old - by Gender
Results
Figure3providesacomparisonofthe56collegesanduniversitiessubjectto
TitleIXinvestigationsbylatespringof2014.Collegessubjecttoinvestigationwere
morelikelytobelarge,research-intensive(R1)universities,eitherpublicorprivate.
Approximatelythree-quartersoftheinstitutionsunderTitleIXinvestigationwerein
thiscategory.Lower-tier,lessselectiveschools(R2)werelessrepresentedamong
schoolsunderinvestigation.
15
Figure3
Publ. R1
Publ. R2
Priv. R1
Priv. R2
Priv. BA
Publ. BA
Publ. R1
Publ. R2
Priv. R1
Priv. R2
Priv. BA
Publ. BA
No Investigation Title IX Investigation
Authors calcutations from IPEDS
By Title IX StatusType and Control of College & University
InTable1weprovideadditionalcomparativeevidenceofschoolsunder
investigationandthosethatwerenot.SchoolsunderTitleIXinvestigationare,on
average,muchlarger,theyaremoreselectiveandhavehighergraduationrates.
Schoolsunderinvestigationadmitabouthalfofapplicants,comparedto63percent
forotherschools,andSATscoresatthe75thpercentileareafullstandarddeviation
higher.Remarkably,1in4schoolsunderTitleIXinvestigationarestateflagship
universities.
InTable2wepresenttheresultsofourmainempiricalmodel.Thefirst
columnpresentstheresultsofourmodelspecifiedabove.Thenexttwocolumns
differonlyinhowwemodelgeneralenrollmentchangesovertime.Model2includes
alineartimetrend,ratherthancontrollingforyearfixed-effects.Model3allows
16
eachage/gendergrouptohaveitsowntrend.Model1islessrestrictivethan2,
becausegrowthinenrollmentisnotparametric.Model3ispreferredifonebelieves
thatenrollmentgrowthwouldbeexpectedtodifferbetweenage/gendergroups.
Sincecollegefixedeffectsareincludedinallmodels,theseestimatesare
robusttochangesinunmeasuredinstitutionalattributesthatmightaffect
enrollment.Theidentifyingassumptionisthattherearenounmeasuredchangesat
auniversitythatperfectlycoincideswiththeopeningofaTitleIXinvestigation.Of
course,itispossiblethataninvestigationspursresponsefromauniversity’s
administration.Thisisnotaconfounder,however.Rather,itwouldbepartand
parceloftheinvestigation.
Incolumn1,weestimatethatenrollmentforwomenunder20yearsofage
declinesapproximately16percentwhenauniversityisunderTitleIXinvestigation,
withevenlargerenrollmentdeclinesamongundergraduates20andolder(both
estimatesaresignificantatconventionallevels).Wefindnosimilardeclinesin
enrollmentformen.Indeed,weestimatethatenrollmentamongmenincreases,
thoughnotforoldermen.Oneexplanationforthispatternisthatenrollment
changesfor17to19year-oldcollegestudentsisdrivenmainlybychangesin
matriculationrates.Recallthatschoolsunderinvestigationarequiteselective.So,a
declineintherateatwhichwomenapply,acceptoffersandenrollmightactually
benefityoungmenasspotsattheseinstitutionsopenup.Forstudentsbeyondthe
Freshmenyear,changesinenrollmentarelikelytobemainlydrivenbydropout.So,
therelativelylargeenrollmentdeclineforwomenisbadforthewomeninvolved,
17
andbadfortheinstitution,sincethesestudentscannotbesoreadilyreplaced.The
resultsfromModels2and3aresubstantiallyandstatisticallysimilar.
InTable3,weexaminewhetherenrollmenteffectsofTitleIXinvestigations
differbycollegeselectivity.Todoso,wegroupinstitutionsintothreecategories,
basedontheiradmissionsrates.Wedefineschoolsas“highlyselective”iftheyadmit
lessthanhalfofapplicants;“Selective”schoolsadmit50to75percentofapplicants,
and;“Lessselective”schoolsadmitmorethan75percent.Inallcases,theopeningof
aTitleIXinvestigationdecreasesenrollmentforwomennotintheirteens–though
theeffectisincreaseswithselectivity.Onlyforthehighlyselectiveschoolsisthere
alsoaneffectonenrollmentofwomenintheirteens.And,onlyattheseschoolsis
thereanincreaseinenrollmentformen.Thispatternisconsistentwiththe
possibilitydescribedabove,thatTitleIXinvestigationsandtheatmospherethey
signaldissuadeyoungwomenfrommatriculating,buttheirplacesaretakenatthese
highlyselectiveschoolsbyyoungmen.
Next,weexaminewhetherenrollmentataninstitutionchangesevenbeforea
TitleIXinvestigation,andwhetheranyenrollmentlossesarerecoupedinthe
subsequentyear.Todothis,were-estimateourmainmodel,butincludelead/lag
indicatorsforeachstudentgroup.Tosummarizethe12coefficientsofinterest,we
plottheresultsinFigure4.Thefigureillustratesthecoefficientsand95%
confidenceintervalsfortheindicatorvariablesforeachgrouponeyearbefore,the
yearof,andoneyearafteraTitleIXinvestigationwasopened.Recallthatthese
modelscontrolforinstitution,academicyear,andstudentgroupfixedeffects.Sothe
coefficientsmeasurechangesinenrollmentforagroupoverandabovechangesfor
18
thatgroupinotheryearsandchangesforsamegroupinotheruniversitiesduring
thesameyear.Weincludeareferencelineatzero–theexpectedvalueiftherewas
nounusualchangeinenrollmentforagroup/institution/year.
Figure4
-.4-.2
0.2
.4C
hang
e in
ln(E
nrol
lmen
t)
Yr. Before IX IX Year Yr. After IX
Women < 20 Men < 20Women 20+ Men 20+
Change net of university, year and age/gender fixed effects
Estimates & 95% Confidence IntervalsEnrollment Effects Before/During/After Title IX: by Gender/Age Group
Wefindnoevidencethatenrollmentsdeclineintheyearleadingupto,orthe
yearafteraTitleIXinvestigationislaunched.Oneinterpretationofthefactthat
thereisnodropinenrollmentintheleadingacademicyearisthatitprovidesatype
offalsificationcheck.OurmainestimatesoffromTable2arenotdrivenbydeclines
inenrollmentamongwomenthatsimplyhappentobecoincidentwithaTitleIX
investigation.However,onemighthaveexpectedaleadingeffectifthepublicityofa
highprofilecaseprecededtheOCR’saction.However,ourmaincoefficientmaybe
19
pickingupsomeofthissincestudentenrollmentlikelychangesslowly(asaccepted
studentsdeclineandcurrentstudentstransferorfailtoreturn).
DiscussionandImplications
Inthispaperweassesstheimpactofhighprofilefederalinvestigationsofthe
handlingofsexualassaultclaimsatcollegesanduniversitiesacrossthecountry.
Notably,weexaminetheimpactoftheseinvestigationsonstudentenrollment,
ratherthanonoutcomesonemightexpectwouldbemostdirectlyaffected,suchas
ratesofsexualassaults,supportsforvictims,andpunishmentsforperpetrators.One
reasonthatwefocusonthisbroaderoutcomeisduetothedifficultiesinherentin
measuringvictimization(giventheextentofunderreportingbyvictims).Butour
focusonenrollmentisalsopurposeful.Sexualvictimizationandhowauniversity
handlesthoseclaimsaffectstudentsbeyondthosedirectlyinvolved.Ourfocusis
broadbecausetheclimateandperceptionsofacampuscanaffectthestudentbody
generally,andespeciallyfemalestudents.
Usingaseriesofdifference-in-differencesestimateswefindthatenrollment
ofwomendropssharplyinresponsetotheinitiationofaTitleIXinquiry.Further,
usingatriple-differencestylecomparisonofwomenincontrasttomenat
institutionswith/withoutTitleIXinvestigations,wecanruleoutthepossibilitythat
subsequentenrollmentdeclinesweredrivenbyotherfactorsthatmadethese
institutionslessattractiveoptionsforstudents.Theenrollmentpatternswefind
followingtheinitiationofTitleIXinvestigationsareconsistentwithbothdeclinesin
therateofmatriculationandretentionoffemalestudents.
20
WeestimatethatenrollmentofteenagewomenatcollegesunderTitleIX
investigationdeclinedby16percent,whiletheenrollmentofundergraduatewomen
from20to22yearsolddeclinedmorethan20percent.Ifweuseenrollmentmeans
fortheseschools,thisimpliesadeclineof330matriculation-agedwomen,and730
retention-agedwomenattheaverageschoolunderTitleIXinvestigation.These
numbersaredriveninbythefactthatTitleIXschoolsareoftenstateflagship
universities,withlargeenrollments.By2014,ourestimatesimplythecollege
enrollmentdecisionsofmorethan30,000womenchangedasaresultofaTitleIX
investigation.Tobesure,thisisasmallnumberrelativetothemorethan6.25
millionwomenenrolledasfull-timeundergraduatesatthetime.Nonetheless,the
enrollmentchangesinducedbyTitleIXcouldcertainlynegativelyaffectthe
prospectsofthoseinvolved,sincemorethantwo-thirdsoftheenrollmenteffectis
duetodecliningretention.
OurresultsalsosuggestthattheenrollmentshiftsduetoTitleIX
investigationsareshort-lived,andessentiallydisappearafterayear.Thiscouldbe
becausenews-cyclesareoftenshort.Or,itcouldbethatthestudentsmostupsetby
aninvestigationleaveimmediately,andthoseremainingarelesscompelledtodoso.
Amoreoptimisticinterpretationtotheshort-termenrollmenteffectsitthat
studentsatacampusunderinvestigationperceiverealchangesbeingmadein
response.
Itisimportanttorecognizethattheenrollmenteffectsidentifiedherepertain
onlytotheearly2010s.InthisperiodoftheuseofTitleIXtoinvestigatesexual
assaultwasnew,andthecollegesanduniversitiesunderinvestigationwere
21
relativelyfew.Intheshorttimesincetheperiodstudiedhere,thenumberofTitleIX
investigationshasincreasedten-fold.Withsomanycasesopen,andopening,college
applicantsandstudentstodayhavefeweroptionstoavoidacampusthathasnot
beenunderinvestigation.
Itremainstobeseeniftherecentattentiontotheissueaffectsratesofsexual
violenceoncollegecampusesintheaggregate.Ourpaperillustratesthatattention
totheappearanceofinsufficientprotectionsandprocessesatspecificcollegeshad
cleareffectsontheenrollmentofwomen.Thatmanywomen“votedwiththeirfeet”
byleavingoravoidinginstitutionsunderinvestigationmakesclearthestakesare
highforcollegesanduniversities.Sincefemalestudentsmakeupasizeable
majorityofundergraduates,assuringthattheirsafetyandprotectionareparamount
concernswillbeessentialforanycollegeoruniversitythathopestothrive.The
federalrequirementsofCleryandthe2013recentreauthorizationofVAWAmean
thatcollegesanduniversitiesnationwideareincreasingtheamount,extentand
qualityofsexualassaultpreventionprogrammingtheyofferoncampus.Our
findingsindicatethattheoutcomesoftheseeffortsareconsequential.
22
References
Aud,S.,Hussar,W.,Kena,G.,Bianco,K.,Frohlich,L.,Kemp,J.,Tahan,K.(2011).eConditionofEducation2011(NCES2011-033).U.S.DepartmentofEducation,NationalCenterforEducationStatistics.Washington,DC:U.S.GovernmentPrintingOffice.
Campbell,R.,Dworkin,E.,&Cabral,G.(2009).Anecologicalmodeloftheimpactofsexualassaultonwomen'smentalhealth.Trauma,Violence,&Abuse.Cantalupo,N.C.(2011).Buryingourheadsinthesand:lackofknowledge,Knowledgeavoidanceandthepersistentproblemofcampuspeersexualviolence.LoyolaUniversityChicagoLawJournal,43,205.CenterforPublicIntegrity.2010.https://www.publicintegrity.org/2010/02/24/4360/lack-consequences-sexual-assaultFedina,L.,Holmes,J.L.,&Backes,B.L.(2016).CampusSexualAssaultASystematicReviewofPrevalenceResearchFrom2000to2015.Trauma,Violence,&Abuse,1524838016631129.Dunn,L.L.(2013).AddressingSexualViolenceinHigherEducation:EnsuringCompliancewiththeCleryAct,TitleIXandVAWA.Geo.J.Gender&L.,15,563.Fisher,B.S.,Cullen,F.T.,&Turner,M.G.(2000).TheSexualVictimizationofCollegeWomen.ResearchReport.Koss,M.P.,Gidycz,C.A.,&Wisniewski,N.(1987).Thescopeofrape:incidenceandprevalenceofsexualaggressionandvictimizationinanationalsampleofhighereducationstudents.Journalofconsultingandclinicalpsychology,55(2),162.Krakauer,J.(2015).Missoula:Rapeandthejusticesysteminacollegetown.NewYork,NY:Doubleday.Krebs,C.P.,Lindquist,C.H.,Warner,T.D.,Fisher,B.S.,&Martin,S.L.(2007).Thecampussexualassault(CSA)study:Finalreport.Washington,DC:NationalInstituteofJustice,USDepartmentofJustice.Kruttschnitt,C.,Kalsbeek,W.D.,&House,C.C.(2014).Estimatingtheincidenceofrapeandsexualassault.Washington,DC:NationalResearchCouncil.Kilpatrick,D.G.,Resnick,H.S.,Ruggiero,K.J.,Conoscenti,L.M.,&McCauley,J.(2007).Drug-facilitated,incapacitated,andforciblerape:Anationalstudy.NationalCriminalJusticeReferenceService.
23
NationalPublicRadio.(2010).SeekingJusticeforCampusRapes.WashingtonDC:NationalPublicRadiowiththeCenterforPublicIntegrity.Retrievedfromhttp://www.npr.org/series/124073905/seeking-justicefor-campus-rapes.Richards,T.&Kafonek,K.(2016).Reviewingstatelegislativeagendasregardingsexualassaultinhighereducation:Proliferationofbestpracticesandpointsofcaution.FeministCriminology,11(1),91-129.Schroeder,L.P.(2013).Cracksintheivorytower:HowtheCampusSexualViolenceEliminationActcanprotectstudentsfromsexualassault.Loy.U.Chi.LJ,45,1195.U.S.DepartmentofEducation.(2016).TheHandbookforCampusSafetyandSecurityReporting,2016Edition.Washington,D.C.:OfficeofPostsecondaryEducation.Retrievedfromhttp://www.ed.gov/admins/lead/safety/campus.html.USDepartmentofJustice,NationalInstituteofJustice.(2008).MeasuringFrequency.Retrievedfromhttp://www.nij.gov/topics/crime/rape-sexual-violence/campus/pages/measuring.aspx
24
!
Table!1!!!!Descriptive!Statistics!of!Colleges/Universities:!!By!Title!IX!Status
Variable No!Title!IX!Case Title!IX!Case
Enrollment!(Headcount) 8022.8 17321.59351.2 15449
Admission!Rate 0.632 0.5310.182 0.242
Graduation!Rate! 0.533 0.7270.186 0.17
State!Flagship?!(0/1) 0.032 0.250.175 0.437
Tuition/Fees $17,711 $21,48611920 14859
Math!SAT!(75th!ptile) 589.5 664.967.9 82.5
Reading!SAT!(75th!ptile) 581.9 653.167.1 78
n=888 n=56
25
Table&2&&&&&&Enrollment&Effects&of&Title&IX&Investigations
Variable Model&1 Model&2 Model&3Title&IX&x&Women&<&20 ?0.159** ?0.166** ?0.124
0.069 0.068 0.068Title&IX&x&Men&<&20 0.189* 0.181** 0.144
0.082 0.082 0.082Title&IX&x&Women&>=&20 ?0.242*** ?0.250*** ?0.221***
0.075 0.074 0.074Title&IX&x&Men&>=&20 0.114 0.106 0.081
0.062 0.062 0.063Admission&Rate 0.045 0.038 0.033
0.027 0.027 0.027Pct.&Receiving&Aid 0 0 0.001
?0.001 ?0.001 ?0.001Graduation&Rate&(6?year 0.127 0.129 0.149**
0.077 0.076 0.076Bach&Deg.&As&%&of&All&Degrees 0.175*** 0.144** 0.146**
0.066 ?0.066 ?0.073Tuition&&&Fees&($1,000s) ?0.002 ?0.003 ?0.002
0.002 0.002 0.001Time&Trend 0.017*** 0.012***
0.002 0.002Time&Trend&x&Men&<&20 0
0.002Time&Trend&x&Women&>=&20 0.005**
0.002Time&Trend&x&Men&>=&20 0.008***
0.003
Institution&FE? Yes Yes YesStudent&Group&(Age/Gender)&FE? Yes Yes YesYear&FE? Yes No NoTime&Trend? No Yes YesStudent&Group&Time&Trend? No No Yes
H o :&Women&<&20&=&Men&<&20&(p?value) <0.01 <0.01 0.012
H o :&Women&20+&=&Men&20+&(p?value) <0.01 <0.01 <0.01
R?squared 0.817 0.817 0.898N 30171 30171 29771
**&p<0.05,&***&p<0.01
All&models&control&for&institution&and&student&group&(age/gender)&fixed&effects
26
Table&3&&&&&&Title&IX&Enrollment&Effects:&By&College&Selectivity
Variable Highly&Selective Selective Least&Selective
Title&IX&x&Women&<&20 D0.207* D0.1 D0.193
0.099 0.093 0.134
Title&IX&x&Men&<&20 0.325*** 0.188 0.001
0.094 0.128 0.108
Title&IX&x&Women&>=&20 D0.314** D0.255* D0.103*
0.097 0.126 0.052
Title&IX&x&Men&>=&20 0.158 0.084 0.149*
0.082 0.093 0.065
Admission&Rate 0.089 0.066 D0.076
0.133 0.059 0.09
Pct.&Receiving&Aid 0.001 0 0
0.001 0.001 0.001
Graduation&Rate&(6Dyear) 0.213 0.056 0.086
0.207 0.07 0.107
Bach&Deg.&As&%&of&All&Degrees D0.117 0.165* 0.107
0.177 0.071 0.09
Tuition&&&Fees&($1,000s) 0.003 D0.004 D0.001
0.003 0.003 0.003
Institution&FE? Yes Yes Yes
Student&Group&(Age/Gender)&FE? Yes Yes Yes
Year&FE? Yes Yes Yes
H o :&Women&<&20&=&Men&<&20&(pDvalue) <0.01 0.1978 0.2594
H o :&Women&20+&=&Men&20+&(pDvalue) <0.01 0.4289 0.0584
RDsquared 0.79 0.83 0.802
N 6169 14658 9344
*&p<0.05,&**&p<0.01,&***&p<0.001