Romer Rosenthal (1978) - Political Resource Allocation, Controlled Agendas and the Status Quo

Political Resource Allocation, Controlled Agendas, and the Status Quo Author(s): Thomas Romer and Howard Rosenthal Source: Public Choice, Vol 33, No (1978), pp 27-43 Published by: Springer Stable URL: http://www.jstor.org/stable/30023066 Accessed: 22/08/2009 13:33 Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use Please contact the publisher regarding any further use of this work Publisher contact information may be obtained at http://www.jstor.org/action/showPublisher?publisherCode=springer Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission JSTOR is a not-for-profit organization founded in 1995 to build trusted digital archives for scholarship We work with the scholarly community to preserve their work and the materials they rely upon, and to build a common research platform that promotes the discovery and use of these resources For more information about JSTOR, please contact support@jstor.org Springer is collaborating with JSTOR to digitize, preserve and extend access to Public Choice http://www.jstor.org Articles Political controlled resource agendas, allocation, and the status quo ThomasRomerandHowardRosenthal* Introduction Economicanalysisrequiresmodellingpoliticalas well as marketresource allocation.Voting institutions,in particulartwo-candidatemajorityrule elections and voting on motions,have been a primaryfocus of recent In the case of a singlegoodto be allocatedpoliticanalyticaldevelopments lead to 'single-peakedness' of voterpreferences ally, standardassumptions over the set of alternatives When,in choosingbetweena pairof available alternatives,every voter votes for his preferredalternative,the allocative is the 'Condorcetpoint'or politicalallocationmost desiredby equilibrium the medianvoter(Bowen,1943;Black,1958;RikerandOrdeshook,1973) Thisresultconcerningthe dominanceof the medianvoter'sidealallocation dependsimportantlyon the natureof competitionin the allocation process.In the context of the politicalallocationof economicgoods,the 'medianvoter'outcomeis typicallyjustifiedon the basisof anunderlying but usuallyunmodeled- processof politicalcompetitionbetweentwo candidatesfor elective office, whereinthe dominantstrategyfor each candidateis to offerto providethe levelof publicspendingthatcorresponds to the medianvoter'sidealexpenditure Sucha viewof equilibrium undermajorityrule(whenequilibrium exists) of politicalprocesses.Manysuchprocesses, may be very unrepresentative those relatedto collectiveexpendituredetermination, particularly may be more appropriately characterized as ones in which some grouphas the powerto make a proposalto the voters,andtherebyset the agenda.This group,whichwe call the agendasetter,by havingmonopolypoweroverthe proposalplacedbeforethe electorate,canconfrontthe voterswitha 'takeit or leaveit' choice.Because'competitive' substitutesto the setter'sproposal are not offered, the medianvoter cannot simply 'hold out' until the Condorcetpointis proposed Whenthe setterhasmonopolypower,votersareforcedto choosebetween * GraduateSchool of IndustrialAdministration,Carnegie-Mellon University,Pittsburgh, Pa Financialsupport was provided,in part, by The SpencerFoundation.We thank M Harris,J Lave, S Salop, and participantsin workshopsat the Board of Governorsof the FederalReserveandat Queen'sUniversityfor helpful comments Public Choice33 (1978) 27-43 All rightsreserved Copyright@ 1978 MartinusNijhoffPublishersbv, TheHague/Boston/London Publicchoice 28 the setter'sproposalandthe statusquo or fall-backposition.Thestatusquo is the situationthatprevailsif votersrejectthe setter'sproposedalternative The ruledetermining the statusquo or fall-backpositionis generallyspecified by law, andis not subjectto the setter'scontrol.Examplesof fall-back andthe previousyear'sexpenditure positionsarezeroexpenditure To providesome structurefor this 'monopoly'type of process,we analyze a simple model of collective expendituredeterminationwith behavior.Thismodelcorresponds agenda-setting quitecloselyto the situation in many local jurisdictionswhere some collectiveexpendituresare determinedthroughthe interactionof citizen-voters anda committeeor a bureauchargedwith the provisionof publicservices.Typically,the bureau/ committee formulatesa proposalfor the comingperiod'sexpenditures Thisbudgetproposalis then subjectto approvalor defeatin a referendum of thejurisdiction's residents.1 In this paper,we areconcernedwith a singlevote on a tax-expenditure decision.We thus not considerdynamic,sequentialaspectsof the politicalprocess,nor we investigatelogrollingandcoalition-formation Voters are characterizedas behavingindividualistically, makingvoting decisionsindependentlyof others.Decisionsare madein a world of certainty, so that we not deal with questionsarisingfrom incomplete turnoutand lack of informationaboutvoterpreferences Neglectingthese aspectsof the problemdoesnot meanwe feel they areunimportant Rather, our intentionis to explorethe implications of agendacontrolperse, leaving to futureworkthe furtherelaboration of the basicstructure.2 To focus as sharplyas possibleon the role of the setter,we characterize the setter as havinga preferencefor the largestfeasibleexpenditure.The motivationis exploredin detailby justificationfor thisviewof bureaucratic Niskanen(1971, chapter4; 1975) Althoughtherehas been somecriticism of the budget-maximizing assumptionwhen appliedto bureaucratsin general(see, for example,BretonandWintrobe,1975;andMargolis,1975), we find this characterization of setterbehaviorparticularly for appropriate the situationsthat concern us here The setter in public expenditure referendais typicallyan interestedprofessional,such as a school superintendent,who may quite sincerelybelievethat provisionof the service suppliedby his agencyis importantfor the community'swelfare.Conseunitsof supplyquitehighly.A setter, quently,he wouldvalueincremental of course,may also find directpersonalsatisfaction(pecuniaryand nonpecuniaryin source)frombeingin chargeof an agencywitha largebudget It may be arguedthat neitheraltruismnorprivateincentivesneedmakethe setterliterallya budget-maximizer Whileacknowledging and agreeingwith this contention,we not believethatintroduction of complicating factors - structuring the problemin termsof maximizing the 'discretionary budget' or maximizinga setter'sutility function,someof whosearguments depend positivelyon budgetsize - wouldprovidesignificantlygreateranalytical Volume33 Issue4 29 in exploringthe implicainsightin this context.Weareinterestedprimarily tions of the expenditure-maximizing assumption,since it highlightsquite the of the setter.3,4 sharply importance setterobviouslyprefersa levelof Althoughan expenditure-maximizing the median that desired than voter,to passthe proposed by supplyhigher the approvalof at least the setter needs under voting, majority expenditure interestedin the way that the level of half the voters.We are particularly supply that the voters approvedependson the statusquo position.We show that the status quo point stronglyaffectsthe allocation.In other - thedeterminants orhistorical structure background words,the institutional are or fall-back of the statusquo very importantto the outposition the seeminglyperverseand We obtain election the of come expenditure the voters of a that result maybe betteroff and largemajority paradoxical allocativeefficiencymay be more nearly achievedwhen the setter can imposesometax withoutvoterapprovalthanwhenthe setterrequiresvoter approvalfor all taxes.The reasonfor this is that, in effect, zero taxesalso meanszeroexpenditure.The settercan then use his monopolypowerover the agendato threatenthe voters with facingthe consequencesof zero Wealsoshow expenditureif they failto approvea highlevelof expenditure thatthe setterhasa morecomplexproblemthansimplychoosinga proposal that inducesa 'Yea'vote by a uniquelydefined'median'voter,evenin the context of a singleelectionwith full turnoutandperfectinformationfor the setter in section 2, the behaviorof votersand We begin by characterizing, statics of changesin the statusquo are analyzedin setter Comparative relatedto educational section3 andillustratedin section4 withdiscussions In a morespeculativevein,we also financeandpublicworksexpenditures of usingzero-basebudgetingin an commentbrieflyon possibleimplications setters-areactive.The paper environmentwhere expenditure-maximizing concludeswithindicationsforthe directionof futureinvestigation Analyticalframework voter5 2.1 Theindividual Wedealwitha setN = {1, 2, , n} of voters.Eachvoteri EN hasa strictly quasi-concavepreferencefunction Ui(C', G'), defined over all pairsof i's consumptionof in (C, G) C' represents goods(C, G) andnondecreasing a bundleof privategoodsandG' his consumptionof a collectivelyprovided good Thisgoodmay be a purepublicgood,a privategood,or somemixed is that it is financedcollectivelyand allogood Its essentialcharacteristic betweencollectiveexpenditures catedpolitically.The relationship (in units G' is givenby: of privateconsumption good)E andi's consumption G' = fi(E) Publicchoice S30 Wetakefi(E) to be increasing andweaklyconcave.It followsthat: u' (C',E) U' [C',f' (E)] is strictlyquasi-concave andnon-decreasing in (C',E).6 Supplyof the collectivegood is financedfrom (e.g., local) taxes and, possibly,from other (e.g., state or federal)revenuesources.We definea tax structureas a rule that determines,giventhe level of expenditureE, the tax paymentsof each votersi For a giventax structure,each collective expenditurelevel is associatedwith some maximumfeasibleprivate for eachvoter consumption The statusquo for voter i is represented by a (Ci, E) pairthat we call q' This is a point associatedwith someconstitutionally prescribed'reversion rule' that specifiesthe level of expenditure(and the reversiontax is voteddown structure)thatoccursif a proposedalternative For a giventax structure,the alternativesto the statusquo imposea constraint: (1) Ci < T'(E) on voteri The functionTi is determined by the waytaxesareapportioned amongthe votersand the availabilityof outsiderevenue.Weassumethat T' is non-increasing and weaklyconcavefor all voters.Individualutility maximization impliesthat, for givenE, (1) will hold withequality.Letting A' denotetheset of alternatives to thestatusquo facingvoteri, we have: A= (C',E) : C' = T4(E) (2) Theutilityforvoteri of alternative collectiveexpenditure is givenby: V'(E) u' [ Y'(E), E ] (3) A straightforward aboutthe utility funcconsequenceof our assumptions tionsandalternative setsis: Lemma1 Vi(E) is single-peaked in E Specifically,Vi(E) is strictlyincreasfor E H and for E > , for alliE N, where strictly decreasing < inp EV E is the voter's'most-preferred' oridealexpenditure givenTi Any proposalby the setter(otherthanthe statusquo) impliesa point in A' for eachvoter.If the financingof the statusquo involvesthe sametax structureas the financingof alternatives to the statusquo, then qi E A' for all i If, on the otherhand,in proposinga changein total expenditure from the statusquo, the proposingbody or setteris constrained to using Volume33 Issue4 31 a differentfinancingstructure,thenqi maynot be in the set of alternatives.7 Individualsare askedto vote 'Yea'or 'Nay'on a proposedexpenditure E If the statusquo for voteri is q!, then he votes 'Yea'on the proposed if: alternative >ut(qi) Vi(E) - andunimportantly - assumea Otherwise,he votes 'Nay'.(Wearbitrarily 'Yea'vote in the caseof indifferencebetweenthe proposalandthe status quo.) Lemma2 Let B'(q'o)C A' designatethe set of proposalsvoteri approves A' designatethe set of proposals when status quo is qi0 and B'(q) the when status some is '1) quo approved C_ differentpoint qi If u'(qi then B'(q') B(qI) u'(qi), C_ Thisfollowsdirectlyfromour assumptions aboutui(-) andA' Thelemma q o Increasing U c/ /3 E 0" I S I I 1.2 Io I E E2 Ei E3 E4 Expenditure on Political Good for individualvoter Figure1 Decision-making Public choice 32 is illustratedin Figure1 Withpointqgoon contourIOthe voterwill support any expenditurelevel betweenEl and E4 For q, on contourIi, only expendituresbetweenE2zandE3 wouldbe supported.Note alsothat for statusquossuchthatu'(q')> V'(E'),B'(q') is empty.(Thisis the casewith q+ in Figure1.) 2.2 Thesetter'sbehavior The setter or groupmakingexpenditureproposalsis assumedto know voters'preferences For a givenexpenditureproposalE, definebi(E)= if voteri votes'Yea'andbi(E) = if 'Nay'.Thenthe setter'sobjectiveis to: maximize E subjectto n bi(E)> 0.5n (4) i= The rest of the paperconsidershow changesin the statusquo affectthe solutionof thismaximization we areconcernedwith problem.In particular, how the solutionchangeswith increasesin the statusquo levelof expenditure somecomparative Changesin the statusquoexpenditure: statics Let Q = {q', q2, , qn} be the set of statusquopointsundera givenfallwhenthe statusquo backrule,andlet E(Q) denotecollectiveexpenditures is Q Definition:An alternativeexpenditureE' is viableagainsta statusquo Q if: (a) the numberof voters for whom V'(E') > u'(q'), i EN, is greaterthan0.5n; and (b) E' > E(Q);i.e the setterprefersE' overE(Q) The setter'sproblemis trivialfor statusquos withoutviablealternatives: stick with the statusquo The more interestingcases involvestatusquos Ourdiscussionthereforefocuseson these whichdo haveviablealternatives cases 3.1 Dominantstatusquopoints Definition.A statusquo Qo= {qO, , qn} dominatesanotherstatusquo u Q, : q, , iffu (qi) >.u (q' )for alli EN, anduq o)>u'(q) for somei E N Proposition1 Considertwo statusquosQo andQ,, eachwith at leastone viablealternative.Let the solutionto the setter'sproblem(4) by E*(Qo) when the statusquo is Qo and E*(Q1) whenthe statusquo is Ql If Qo dominatesQI, thenE*(Qo) E*(QI) solves(4) for E(Qo) > E(Q1) But then,fromLemma3, E*(Qo) would at get least as manyvotes againstE(QI) as againstE(Qo),implyingthat E*(QI) is not a solutionto (4) forE(Qi) Thecontradiction impliesthat > E*(QI) E*(Qo) 3.3 Otherstatusquopoints A more explicit considerationincludingprivateconsumptionbundlesas well as expenditurelevels allowsus to developa somewhatmoregeneral resultwhichincludesProposition2 as a specialcase.Weconsiderchangesin statusquo expenditures suchthatstatusquopointslie on or abovethe Ci = Ti(E)locusforeachvoter.WegeneralizeLemma3 to: Lemma3' Let Qo and Q, be two distinctstatusquos, with collective expendituresE(Qo) and E(Q1), respectively,and E(Qo) > E(Qi) Let C9 and C' be the privateconsumptionof voteri understatusquosq' and q , respectively Supposethat: >c C1 - c T'[E(Q1)] for all iEN < T'[E(Q1) - T[E(Qo)] for alli EN (5) (6) Volume33 Issue4 35 Considera proposedexpenditureE' (which allows voter i to obtain (T'(E'), E') E A') such that E' > E(Qo) > E(Q,) Thenthe proposalE' cannotreceivemore'Yea'voteswhenE(Qo) is the statusquoexpenditure thanwhenE(Q1)is the statusquo expenditure Remark.Condition(5) requiresthat the statusquopointq1 lie on or above the alternative locus Condition(6) statesthat the tax cost to voteri of the increasein expenditurefrom E(Qi) to E(Qo) underthe status quo (i.e., C - C') be no greaterthanwouldbe the tax cost of sucha change underthe alternativetax structure[i.e., T'(E(Qi))- Ti(E(Qo))] Note that conditions (5) and (6) togetherimply that Co >I T'[E(Qo)] for alli E N Proofof Lemma3' Considerthreeexhaustivecases: If ui(q > u(qo), then V'(E') < u'(qi) andvoteri willvoteagainstE' underboth statusquos.Thisfollowsfromthe conditionsof the lemma, the concavityof T'(E), andthe strictquasi-concavity of preferences (for details,seeAppendix) If > V'(E'), thenvoteri willvoteagainstE' underboth > u'(qo) u'(ql ) status quos If u'(ql) > u'(q ) and u'(q' ) < Vi(E'), then voter i will votefor E' when the statusquo is q' , and vote eitherfor or againstE' whenthe statusquois qo E' cannotget more 'Yea'votes againstE(Qo) thanagainst Consequently, the lemma E(Q1).Thisestablishes With Lemma31 and argumentssimilarto those used to demonstrate Proposition2 we canprovethe following: Proposition2': Considertwo distinctstatusquosQo andQ1,eachwithat least one viablealternative,andsatisfyingconditions(5) and(6) Suppose that proposalE*(Qo) solves(4) whenthe statusquo is Qo withstatusquo expenditureE(Qo), andproposalE*(Q1) solves(4) whenthe statusquois Q with statusquo expenditureE(Qj) < E(Qo) ThenE*(Q1) > E*(Qo) Discussionandexamples The crucialfeatureof the allocationprocesswe areexaminingis the existTheabilityto to entry'in the formulationof alternatives ence of 'barriers controlthe agendagivesthe settera monopolypowerwhichhe canexploit to an extent that dependson the statusquo By facingthe voterswith a 'take-it-or-leave-it' choice, the setterexercisesa threatover the voters The worse the statusquo, the greaterthis threatand, consequently,the Publicchoice 36 greaterthe gain to the setterfrom being able to proposethe alternative of The remainderof the paperis devotedto elucidationand amplification ourresultsasthey applyto a varietyof situations 4.1 A potentialtestof themodel A numberof statesuse referendato determinepublicschoolbudgets.In the state of Oregon,for example,local school districtbudgetsand their implicittax ratesmust be approvedby the votersannuallyif the board's proposalexceedsa statutoryamountthat can be leviedwithoutgoingto the voters Simplifyingsomewhat,this statutorymaximumexpenditure is givenby: E(Q) = (BASE)(1.06)t-1916 (sic) E(Q) is, in effect, the statusquo expenditurefor the local schooldistrict, and status quo points lie in the set of alternatives.BASEis a number directlyrelatedto the district'sexpenditurein the year 1916, and t is the currentcalendaryear Since BASEvarieswidely acrossdistricts,so does and is outE(Q) Moreover,E(Q) is independentof currentexpenditures, is satisfied- whichis side the setter'scontrol.Providedsingle-peakedness to publiceducalikelywheretherearefew privateor parochialalternatives tion - our resultswouldpredictthat,ceterisparibus,currentexpenditures shouldbe less in districtsthat had substantialtax basein 1916 thanin the many districtsthat havelow BASEandhencemustface the voterswith a low E(Q) - in many casesE(Q) = - each year Underthe 'competitive'assumption,variationsin the statusquo should not affectthe outcomeof a referendum An econometricmodelthattakes andsocio-economic into accountdemographic factorsandexplicitlyincorporatesthe variationsin the statusquo wouldprovidean empiricaltest of ourhypotheses.Wearecurrentlyengagedin developingsucha test.9 4.2 Example:publicworksexpenditures As an illustrationof some of the propertiesof the controlled-agenda process, considera highwaydepartmentthat proposesto replacean existing bridgewith a new structure.Any constructionof a new bridgeinvolves destructionof the old facility.The new bridgewouldbe paidfor out of a specialassessment.The existingbridgeyieldsutilityu'(q') to voteri Supbureauwere to submitan expenditure pose an expenditure-maximizing to the voters Our between proposal examplefocuseson the relationship the qualityof the old bridge(as measuredby the statusquo utilities)and the approvedexpenditureon the newstructure Weconsidera 'community'with threevoters,withpreferences overnew and expendituresgivenby V1(E), V2(E), V3(E),respectively(see Figure assumea utility scalesuchthatu' (q') = 2) For simplicity,we arbitrarily Volume33 Issue4 37 u2(q2) = u3(q3), for all the statusquos underconsideration.Increasing u(q) then impliesmovingto a statusquothatdominatesthe previousstatus quo.10 U4 I -4- " U, - o I - - :I) V,(E) o0 Vs(E) E2 E Expenditureson NewStructure,E Figure Approvedexpenditureand status quo utility (heavy line shows approved expenditure) In Figure2, we measureutility on the verticalaxis and expenditure on the horizontalaxis.Withstatusquo utilityuo, for example,the largest expenditurethat voter would approveis E4 Any higherexpenditure would leavehim worse off thanhe is with the statusquo.E4 wouldnot, however,be approvedby a majority,sincewith statusquo utilityuo, both voters2 and3 wouldvote againstE4 In fact, the highestexpenditurethat to voter2 willpassagainstuo is E', the largestexpenditureacceptable The heavyblackline in the figureis the graphof the solutionto the setter'sproblem(4) as a functionof the statusquo In accordwith the results of Section 3, the approvedexpendituredecreasesas status quo qualityincreases.For statusquo valuesaboveu4, no positiveamountof new expenditurewould be approved.(The discontinuitiesthat occurat status quo valuesus and u4 are a consequenceof the smallnumberof willusuallybe voters.Withmanyvoters,the plot of approvedexpenditures 'close'to continuous.) In contrastto the outcomeof this controlled-agenda process,standard Publicchoice 38 cost-benefitanalysissuggeststhat,providedthatthe newbridgehaspositive net benefits(treatingas a cost the loss of the utilityprovidedby the old bridge),the optimalscaleof the new bridgeshouldbe independentof the valueof the old bridge.Note alsothat, for statusquo valuesbelowuI, the approvedexpenditureis such that the voterswouldunanimously prefera reductionin expenditure,under the tax/financingarrangements for the new project.[For statusquosbelowu1, expenditure E3 (whichhappensto be voter3's idealexpenditure is preferredunanimously to anyE > ] This examplealso points out that the voter whose preferencesare decisivefor the setterdependson the statusquo.As the heavysolutionline indicates,at low levels of the statusquo voter2 is decisive.As the status quo improves,first voter 1, then voter 3, then voter again,and finally voter againbecome decisive.Consequently,the criticalrole Niskanen (1971, chapters13-14) and othersascribeto the 'middledemandgroup' in thiscontext, (or, in general,voterwith medianidealpoint)is misleading and would ariseonly as a resultof specialformsascribedto the V'(E) functions In the usualmajorityrule situation,the medianvoterdoes, of course, havea centralrole In ourexample,if thereis no statusquoalternative, and the 'competitive' is operating,majority processoutlinedin the Introduction ruleleadsto an expenditureof E2, thatmostdesiredby voter2 Evenwith a statusquo, this level,if proposed,wouldbe selectedas longasthe status quo for each voteris not better than u2 At statusquosjust aboveuz, voters and both preferthe statusquo to E2 and would combineto defeatF2 In fact,cyclicalmajoritiesariseanda Condorcetequilibrium fails to exist until the statusquo valuepassesu4 At thispoint,the statusquois a Condorcetwinner If decisivenessis a valuedattributeof choicemechanisms, the example illustratesa general advantageof controlledagendasover competitive in expenditure, majorityrule Even thoughpreferencesare single-peaked the presenceof a statusquopointwill frequentlysufficeto robcompetitive majorityrule of an equilibrium,whereasthe controlledagendaalways createsa decision 4.3 Agendascontrolled by high demandgroups The agendamay not actuallybe controlledby an expenditure-maximizer but by the groupwith the highestidealpoint."1If this groupis the setter andcan passa budgetbeyondits idealpoint,givensingle-peakedness, it can alsopassits idealpoint.Consequently, we mayreformulate (4) as: maxE subjectto E >Eh n and Z bi(E) > 0.5n i= (4') Volume33 Issue4 39 whereEh is the most preferredlevel of expenditurefor the voterwith the highestidealexpenditure This modificationto the objectivefunctiononly triviallymodifiesthe solution.The solutionto the publicworksexamplein Figure2 is voter3's ideal point, E3, for statusquo utilityless thanul For greaterstatusquo utility,the previoussolutionapplies 4.4 Implications for 'zero-base budgeting' Goingbeyondour formalstructure,we note that althoughpublicprojects are rarelybuilt by voters,they are frequentlyundertakenby legislatures reviewprocess.If proposalsfor an agency'sbudget usinga Niskanen-type arecontrolledby a high-demand group,our resultssuggestthat the agency may well prefera reversionrulewhichmakesthe statusquo fairlylow In the absenceof a carefulreviewmechanismin whichlow- and moderatedemand groups have the opportunityto enter the policy-formulation process,lowerstatusquoswill tendto leadto higherapprovedexpenditures concernoverthe level of Respondingto what appearsto be widespread publicspendingby the U.S government, pendingCongressional legislation Thisprocess aimsto institutea newreviewprocessfor federalexpenditures of new budget authorityfor Government would require'authorizations and at least activities everyfive years [andwould] establisha programs zero-base review and evaluationof Government and for procedure programs activitiesevery five years.'12The majorpurposeof such legislationreferredto, with varyingdegreesof precisionand vividness,as 'zero-base review','zero-basebudgeting',or 'sunsetlegislation'- is, of course,to andoutdatedprograms Mostcriticsof the legislation, eliminateunnecessary while agreeingwith its goals,havefocusedon the enormityof the reviewing task andon the undesirability of havingprogramssuchas nationaldefense or the federaljudiciarysystemgo to a zerobasependingreview Drawingon our results,we suggest,in addition,thatformanyprograms whosecontinuationis deemeddesirable,zero-basereviewand- evenmore so - zero-basebudgetingmay tend to increasethe monopolypowerof the likelyif, swamped agencies.Thisseeminglyperverseoutcomeis particularly reviewprocess(or by crisessuchas Vietnamor with the newly-instituted committeesbecomelax with respectto appropriaCongressional Watergate), tions for activitieswhosebudgetsa majoritywouldbe unwillingto havego to zero.The threatof a zerobudgetactuallyoccurringfor caseswhenthis event wouldbe a very bad outcome(a low-rankedstatusquo) may then bureau(whosemotto workto the advantageof an expenditure-maximizing in this case may well be 'reculerpour mieux sauter').As long as agendasetting power remainswith the agencies,for importantareasof public spendinga non-zeroreversionrulewouldthereforeproveto be a moreeffectivewayto controlthe size of the budget Publicchoice 40 Concluding remarks The need for furtherelaborationof the basicmodelof this paperis indicated by the importanceof referendaand directvotingfromBerkeleyto Berne and by the likelihoodthat thereis more than a grainof truthin Niskanen'sidea that high interestgroupscontrolagendasin legislatures While the efficiency propertiesof controlledagendasmay indeed be troubling,it is not clear that other democraticprocedures,such as conventionalmajorityrule, are alwayspreferable.Moreover,controlled agendasappear(like dictators,beneficientand otherwise)to minimize decisioncosts Our model of politicalresourceallocationfocuseson the effects of monopolypowerin the formof controlledagendas.As in othermonopoly situations,it would certainlybe in the interestof somegroupsto devote resourcesto attemptto reducethe extent of monopolypower.In particular, low-demandgroups would make significantgains if they could attenuatethe powerof a high-demand setter.The incentiveto formlowandmoderate-demand coalitionsis certainlypresent,thoughsuchcoalitionformationmay be verycostly - especiallyif the barriersto entryarehigh A most interestingextensionof our analysiswouldbe a kindof 'imperfect competition'model of the political resourceallocationprocess which recognizesthe role of agendasettersandallowsfor somecompetitionfor that position.Wouldsuch a processapproach,as its dynamicequilibrium, the 'medianvoter' allocation?Or would some of the setter'smonopoly powerbe preserved? To answerthese questionsrequiresfurtherexplicit modelingof the political allocation process Incorporatingimperfect informationand uncertainty,as well as sequentialelements,arenecessarystepstowardthe of equilibrium characterization Appendix Elaborationof Step of Proof of Lemma3' E' - E(Qo) Let XE , so that XE(Q1) + (1 - X)E' =E(Qo) E - E(Q ) Fromconcavityof Ti: Ti[E(Qo)] > X Ti[E(Ql)j +(1 - X) T'(E') , so that Ti[E(Qo)] - Ti[E(QZ1)J> (1 - X)T'(E') -(1 - X)Ti[E(Q1)](A1) Using(6): Co - XC' - {T'[E(Qo)I - T'[E(Q,)]} > (1 - X) Cf Volume 33 Issue 41 and(Al): - )Cf C - xCf - {(1 - X)Ti(E') - (1 2>(1 A)Ti[E(Q1)]} = > (1 - X) {Cf Ti[E(Ql)] Co XC' -(1 X)Ti(E') > (A2) i.e.: C > XC + (1 - X)T'(E'), and therefore: u'(qi) = u'[C, E(Qo)] > u'i[XC + (1 - X)Ti(E'), E(Qo)](A3) From strict quasiconcavityof u': ui[xC + (1 - X)Ti(E'),E(Qo)] > {ui[Ci, E(Qi ), u[ Ti(E'), E']} i.e.: ui[XCI +(1 - X)T'(E'), E(Qo)] > [ui(q'), Vi(E')] (A4) By assumption: u(q') u'(q') This, togetherwith (A3) and (A4), implies: u (qo) > Vi(E') Notes In the areaof publiceducationalone (the most importantitem of collectivelyprovided goods at the local level), Holcombe(1975) lists twenty-two states in the U.S.A that requirea referendumprocessto determinelocal school taxes and budgets In additionto public expenditureprocessescharacterizedby directreferenda,the controlled agendaframeworkis an appealingmodel for votingon California-stylepopular initiatives, where there are significantcosts to generatingnew alternatives.We also suggest,followingNiskanen(1971), that settersmay play an importantrole in legislative voting, although we recognizethat legislativeprocessesmay differ significantly from referendumvoting in the extent of opportunitiesfor logrolling,coalition-formation, andgeneratingnew alternatives An importantquestion ariseswhen the proposingpartyis an elected boardratherthan a civil servicebureau.One might contend that the desirefor reelectionmightnot lead the board to make expenditure-maximizing proposals.Yet it might just as well be arguedthat the personalinvestmentof time and money neededto obtain reelectionto the board resultsin the board beingdrawnonly from 'highinterest'individuals.If this Publicchoice 42 is the case, an entry barriermay allow board membersto disregardthe threat of challengefrom 'low interest'individuals.Anotherpossibilityis that the boardmembers maximizethe budget subjectto an upper bound, either as a hedge on reelectionor as an expressionof personalpreference.As we indicatein section 4, the analysisis not materiallyaffected by takingthe setter as the group or person with the highest ideal expenditure,ratherthanactuallyexpenditure-maximizing Oblique support for viewing the setter as a budget-maximizingentrepreneurcomes from evidence,presentedby Edwards(1977), that managersof privatefirmsoperating in a regulatedindustrybehaveas expense-maximizers Edwardsalso finds that increasbehavior ing monopolypowerintensifiesexpense-preference The assumptionsthat lead to Lemma are by now standardin the literature.We develop them mainly as a convenient way of indicating the notation that will be followed in the paper If G is a pure public good, Gi = E For publiclyprovidedprivategoods, such as socialized health care, a simple formulationwould be G' = aiE with < < 1, whereai representsindividuali's perceptionof his 'benefitshare' This might be the case if, for example,the statusquo involvesdifferentamountof outside financingthan wouldbe availableunderthe proposedalternative The suggestionsof an anonymousrefereehelpedabbreviatethe proof of Lemma3 For an attemptto test the 'competitive'model, see Inman(1978) 10 The assumptionof equal utilities for a given status quo enablesus to carryon the discussion using only one diagram.It is by no means necessaryfor the results we are illustrating.In the currentcontext, what mattersis that votersagreeon the rankingof statusquos 11 This assumptionis used in Niskanen'smodel (1971, chapter14) 12 The quote is from the title of a prominentrecent legislativedocument of this type, S.2925, the 'GovernmentEconomyand SpendingReformAct of 1976' References Black, D The Theoryof CommitteesandElections.Cambridge,CambridgeUniversity Press,1958 Bowen, H.R 'The Interpretationof Voting in the Allocationof EconomicResources.' QuarterlyJournalof Economics(November),1943 Breton, A., and R Wintrobe.'The EquilibriumSize of a Budget-Maximizing Bureau: A Note on Niskanen'sTheory of Bureaucracy.'Journal of Political Economy (February),1975 Edwards, F.R 'ManagerialObjectives in Regulated Industries:Expense-Preference Behaviorin Banking.'Journalof PoliticalEconomy (February),1977 Holcombe, R.G A Public Choice Analysis of MillageIssue Elections for Financing Public Schools UnpublishedPh.D thesis,VirginiaPolytechnicInstituteand State University(1975) Volume33 Issue4 43 Inman, R.P Testing Political Economy's "As If" Proposition:Is the MedianVoter ReallyDecisive?'PublicChoice(Summer),1978 Margolis,J 'Comment [on Niskanen].' Journal of Law and Economics (December), 1975 Niskanen,W Bureaucracyand RepresentativeGovernment.Chicago,Aldine-Atherton, 1971 and Politicians.'Journalof LawandEconomics(December), Niskanen,W 'Bureaucrats 1975 Riker,W., and P Ordeshook.An Introductionto PositivePolitical Theory.Englewood Cliffs,N.J., Prentice-Hall,1973 THE QUARTERLY REVIEW OF ECONOMICS AND BUSINESS SPECIAL ISSUE ON INFLATION AND CURRENT VALUE ACCOUNTING Vol 19 Spring 1979 No ARTICLES Studies in Inflation Accounting: A Taxonomization Miklos A Vasarhelyiand Edward F Pearson Approach Powers On the Comparabilityof Alternative Valuation Bases .Marian the Effects of in Accounts Inflation Reporting Company G Piper in the United Kingdom .A Inflation Accounting Development of Theory and Hanns-Martin W Schoenfeld Practice in Continental Europe Evaluationof Alternative Methods of Income Measurementvia a Criterion of ManagerialAbility Ronald D Picur and James C McKeown Depreciation, Inflation, and Capital Formation in Public Utilities: One PossibleApproach Toward a Solution Milford S Tysseland and Devinder K Gamdhi Price-ChangeAccounting Models and DisaggregatedMonetary Gains and Losses Leroy D Brooks and Dale Buckmaster Limitationsof Replacement Cost Harold Bierman,Jr and Ronald E Dukes The Impact of SEC Replacement Cost Reporting Requirements: An Analysis James H Scheiner and Wayne J Morse on Income, Dividends, Their Measures: Cost Impact Replacement and Investment Return Linda H Kistler and ClairmontP Carter ReplacementCost Accounting: Application to the PharmaceuticalIndustry Frank i Fabozzi and Lawrence M Shifirin THE QUARTERLY REVIEW OF ECONOMICS AND BUSINESS is published by the Bureau of Economic and Business Research, College of Commerce and Business Administration, University of Illinois Subscription rates are $9.00 a year for individuals and $14.00 a year for organizations and associations The single-copy price is $4.00 Manuscripts and communications for the editors and business correspondence should be addressed to the QUARTERLY REVIEW OF ECONOMICS AND BUSINESS, 428 Commerce West, University of Illinois, Urbana, Illinois 61801  ... dependson the statusquo By facingthe voterswith a 'take-it-or-leave-it' choice, the setterexercisesa threatover the voters The worse the statusquo, the greaterthis threatand, consequently ,the Publicchoice... setter'sproposalandthe statusquo or fall-backposition.Thestatusquo is the situationthatprevailsif votersrejectthe setter'sproposedalternative The ruledetermining the statusquo or fall-backpositionis... greaterthan the statusquo wouldbe preferred to the statusquoby a simplemajorityof viableagainstthe statusquo voters,andhencethereis no expenditure lessthanmedianidealexpenditure B Statusquoexpenditure The