Does Local Autonomy Breed Leviathans? An Empirical Examination of All Cities in the United States

• Author: Mikhail Ivonchyk
• DOI: http://doi.org/10.1177/10659129211049817
• Published date: 01 December 2022
• Date: 01 December 2022

Article

PoliticalResearchQuarterly

2022,Vol.75(4)1079–1095

©TheAuthor(s)2021

Articlereuseguidelines:

sagepub.com/journals-permissions

DOI:10.1177/10659129211049817

journals.sagepub.com/home/prq

DoesLocalAutonomyBreedLeviathans?

AnEmpiricalExaminationofAllCitiesin

theUnitedStates

MikhailIvonchyk

1



Abstract

USstatesgranttheirlocalunitsdifferentlevelsofautonomyinseveraldimensionsincludingfiscal,functional,structural,and

legaldiscretion.Thisstudyusesacomprehensive,multidimensionalmeasureofautonomytotestitsassociationwiththe

fiscalbehaviorofover19,000municipalitiesintheUnitedStates.Competingtheoreticalpredictionsrangefromsignificant

increasesingovernmentsize(Leviathanmodel),tonoeffect(median-votermodel),andevensmallergovernments

(institutionalcollectiveactionmodel).Quantileregressionanalysisisimplementedtotesttheassociationbetweenau-

tonomyandfiscalbehaviorfordifferentcitysizes.Theempiricalfindingsindicatethatcitieswithmoreautonomytendto

spendlessandhavelowertaxesanddebt.Thestrengthofthisrelationship,however,variesbycitysize.

Keywords

localautonomy,fiscalbehavior,Leviathanmodel,quantileregression,municipalbudget,municipaldebt,municipaltaxes

IntroductionandLiteratureReview

Thisstudyusesamultidimensionalconceptofautonomy

totestitsinfluenceonthefiscalbehaviorofallmunici-

palitiesintheUnitesStates.TheAmericanConstitution

doesnotmentionlocalgovernmentsdelegatingauthority-

givingpowerstostates(1,5).Forhistorical,political,and

culturalreasons,stateshavebeengranteddifferentde-

greesoflocalautonomy.

Theimplicationsofthisvariationforcitypolicymaking

havebeenstudiedsincethelate1800s,butmuchofthe

literaturehasbeen“descriptive,classificatoryandoften

historical”(Krane,Rigos,andHill2001,14)andhasrelied

oncrudemeasuresofautonomy,suchastheDillon’srule–

homeruledichotomy(Richardson,Gough,andPuentes

2003;Richardson2011).

TheconceptsofDillon’sruleandhomeruledonot

capturetheamountofdiscretionavailabletolocalunits

(Richardson2011),andgiventhecomplexitiesinvolvedin

thedistributionofpowerbetweenstateandlocalunits(Liner

1989),itisimpossibletouseseveralorderedcategoriesto

accuratelymeasurelocalautonomyinallstates(Krane,

Rigos,andHill2001,15–16).Instead,researchersargue

thatlocalautonomyisacontinuous,multidimensional

conceptwithoutbright-lineboundaries(Richardson2011).

Theydistinguishlegal,functional,personnel,structural,and

fiscaldimensions(Chapman2003;Krane,Rigos,andHill

2001;Wood2011;Zimmerman1981,1995)aswellasthe

importanceoflocalgovernmentsinthestateeconomyand

thecapacitytousetheirdiscretion(Wolman2008,Wolman

etal.2010).

Muchoftheliteraturehasbeenpreoccupiedwithonly

oneofthesedimensionsofautonomy,fiscaldiscretion,and

hasgeneratedmanyvaluableinsights(e.g.,Dove2016;

JoyceandMullins1991;Maher,Deller,andAmiel2011;

MullinsandCox1995;MullinsandJoyce1996;Mullins

2004;MullinsandWallin2004;MullinsandPagano2005;

Park,Maher,andEbdon2018;Park,Park,andMaher

2018;Rown2000;Shadbegian1998,1999;Stallmannetal.

2017).Themainpremiseofthisstudy,however,isthatall

dimensionsmustbetestedinconcerttofullyunderstandthe

roleofautonomyinlocalpolicymaking.Therehavebeen

severalattemptstosystematicallyoperationalizeand

measurelocalautonomyindifferentcontexts(e.g.,Ladner

etal.2019;LadnerandKeuffer2021;Vetter2007;Wolman

1

RockefellerCollegeofPublicAffairsandPolicy,UniversityatAlbany,

StateUniversityofNewYork,Albany,NY,USA

CorrespondingAuthor:

MikhailIvonchyk,RockefellerCollegeofPublicAffairsandPolicy,

UniversityatAlbany,StateUniversityofNewYork,Milne308,135

WesternAvenue,Albany,NY12203,USA.

Email:mivonchyk@albany.edu

etal.2010),butlittleprogresshasbeenmadeintesting

empiricallytheinfluenceofmultipledimensionsofau-

tonomysimultaneously.Wood’swork(2011)isoneofthe

exceptionshere,whichusedafour-dimensionalmeasureof

municipalautonomyintheUnitedStatesandfoundthat

municipalitieswithmoreautonomytendtoraiselesstaxes

andreceivelessstateaid,withthelevelofspendingand

debtnotaffected.However,thisstudyusedcitydataag-

gregatedatthestatelevelandomitsmunicipalitieswith

fewerthan25,000residents.

Inthemeantime,thereislittleempiricalevidenceto

facilitatethedebatesabouttheappropriateleveloflocal

autonomyanditsimplicationsforlocalpolicymaking.

Higherlevelsofgovernmentsinrecentdecadeshavebeen

imposingadditionalmandatesonmunicipalgovernments

expandingtheirresponsibilitieswithoutcorresponding

increasesintheircapacityandauthority(Krane,Ebdon,and

Bartle2004).Stateconstraintsonlocalauthoritymaybe

deemednecessarytoavoidlocalcorruption,patronage,and

oppressionofopposingviews(BrennanandBuchanan

1980;Martin1990),buttheopponentsnotethatlackof

autonomycanmakeitdifficultforlocalofficialstoreactto

political,social,environmental,andeconomicchanges

(Chapman2003),andthatmoreautonomyisnecessaryto

beabletofulfillthemandatesandadequatelymeetthe

desiresoflocalconstituents(Bunch2014).Intheabsence

ofclearempiricalevidencethatwouldhelpsortoutthese

opposingclaimsandelucidatetheimpactofmultipledi-

mensionsofautonomy,localpolicymakersarecaught

betweenevergrowingdemandsandinadequateauthorityto

meetthem.

Currentstudyaimstoprovideseveralcontributions.

First,itusesacompositeindexincorporatingsevendi-

mensionsofmunicipalautonomyinall50states.Second,it

usesanindividualcityastheunitofanalysisandincludes

over19,000citiesintheUnitedStates.Third,giventhe

diversityofcitiesinthecountryandthattheeffectofstate

limitationsmayvarybyjurisdictionsize(Mullins2004;

Rown2000),thisstudyimplementsaquantileregression

(QR)analysisallowingarichercharacterizationofthe

testablerelationshipatdifferentlevelsoftheoutcome

variables(CameronandTrivedi2009).

Relevanttheoreticalframeworksofferseveralalterna-

tivepredictionswithrespecttotheassociationbetween

municipalautonomyandfiscalbehavior.TheLeviathan

modelpositsthatlocalofficialsseektomaximizegov-

ernmentrevenuesandexpenditurestodistributerewardsto

theirpoliticalallies,andmoreautonomywillbeassociated

withinflatedspendingandhighertaxes.TheInstitutional

CollectiveAction(ICA)frameworksuggeststhatmore

independencefromstaterestraintsmayprovidetheflexi-

bilityneededtocreatemutuallybeneficialcollaborations

andgeneratecostsaving,whichmaybereflectedinlower

spendingandtaxation,holdingallelseequal.Accordingto

themedian-votermodel,however,democraticallyac-

countablelocalofficialsmakepolicychoicesconsistent

withvoterpreferencesandtheautonomyinfluenceonfiscal

behaviormaybeinsignificantafteraccountingforcitizens’

demands.

Thefindingsprovideempiricalsupporttothemedian-

voterandICAmodels,andoffernosupporttotheLeviathan

model.Morespecifically,theresultssuggestthatmunici-

palitieswithmoreautonomytendtohavelowerlevelsof

expenditures,taxation,anddebt.Theseresults,aspredicted

bytheliterature,varybythejurisdictionsize,butnoin-

dicationsofsignificantincreasessupportiveoftheLevia-

thanmodelwerefoundinanyoftheoutcomesatanycity

size.Consequently,thisstudysupportsthenotionthat,

assumingproperlyfunctioningdemocraticinstitutionsand

interjurisdictionalcompetition,localofficialsareunlikelyto

runamokandinflatecityspendingintheabsenceofstate-

imposedconstraints.

Thenextsectionlaysoutthetheoreticalpredictions

testedinthisstudy.Sectionthreedescribesthedata,variable

operationalization,andestimationmethodology.Empirical

findingsarediscussedinsectionfour.Thepaperconcludes

withadiscussionoftheresults,relevantpolicyimplications,

andopportunitiesforfuturestudies.

TheoryofLocalAutonomyandFiscalBehavior

Howcanautonomyinfluencecityfiscalchoices?Relevant

literaturedescribesseveralmodelsoflocalpolicymaking

providingalternativetheoreticalpredictionsaboutlocal

autonomyandfiscalbehavior(Craw2008;Krane,Ebdon,

andBartle2004;McCabeandFeiock2005).

TheLeviathanmodelreflectstheskepticismabout

voters’abilitytocontrolthegovernment(Brennanand

Buchanan1979,1980,1981,1984).Thismodelpredicts

thatintheabsenceofexternalconstraintslocalofficialswill

seektomaximizegovernmentrevenuesandusepublic

resourcestorewardtheirsupportersandensurereelection.

Theywilldistributefavorstotheirpoliticalalliesandsilence

theiropponents(Martin1990).Governmentinthismodelis

akintoanuncontrollableLeviathanwithexpansionary

tendenciesthatovertimeproducegrosslyinflatedbudgets.

State-imposedtaxandspendinglimitations(TELs)arethe

externalconstraintsnecessarytopreventtheLeviathanfrom

expandingbeyondwhattheconstituentsprefer(Brennan

andBuchanan1980).ThefindingthatTELsareeffectivein

reducinggovernmentsizeis“interpretedassupportiveof

thenotionthattheLeviathanmodelmaybeoperative”

(McCabeandFeiock2005;McGuire1999,129).Thus,if

theLeviathanmodelisrelevant,weshouldexpecthigher

taxesandexpenditureswithgreaterdegreesofcityau-

tonomy,holdingallelseequal.

TheICAframeworkpointstoanotherplausiblescenario

(AndrewandFeiock2010;Feiock2007,2013;Feiock,

1080PoliticalResearchQuarterly75(4)