Contracts between Government and State Enterprises

ShirleyandXu1JournalofLaw,Economics,andOrganization,October1998.Information,IncentivesandCommitment:AnEmpiricalAnalysisofContractsbetweenGovernmentandStateEnterprises*byMaryM.ShirleyandLixinColinXuTheWorldBankAbstract:Thisarticleanalyzestheexperiencewithperformancecontractsbetweendevelopingcountrygovernmentsandthemanagersoftheirstateownedenterprises.Itidentifieshowproblemsofinformationasymmetry,incentives,andcommitmentcanleadtoshirking.Itappliesthisconceptualframeworktoasampleof12contractswithmonopolystateenterprisesin6developingcountriesandfindsthatallsufferfromseriouscontractingproblemsandthereisnopatternofimprovedperformancethatcanbeattributedtothecontracts.ShirleyandXu2I.IntroductionAcriticalchallengefacingdevelopingcountriesistoimprovetheperformanceofgovernment,includingstate-ownedenterprises(SOEs).SOEsareresponsibleforanaverageof10percentofGDPindevelopingmarketcountries,andmuchmoreinformersocialisteconomiesandsuchdevelopingcountriesasAlgeria(65percent),Egypt(34percent),Guyana(41percent)orZambia(31percent).SOEdeficitshavebeenanimportantsourceoffiscalproblemsandinefficientstateenterprisesareadragongrowth(WorldBank,1995).Incentivecontracts,similartothecontractsbetweenmanagementandownersthatarewidelyusedinprivatecompanies,havebeensuggestedasatooltoimprovetheperformanceofSOEs(Jones,1991;Trevedi,1990)andcentralgovernmentbureaus(Mookherjee,1997).SubstantialresourceshavebeensunkintothedesignandenforcementofsuchcontractsforSOEs,yetthefewassessmentstodateshowmixedresults(see,forexample,Shirley,1989,1991;Nellis,1989;Trivedi,1990).Thisarticleanalyzestheexperiencewithincentivecontractsthroughcasestudiesof12SOEsinsixcountriesandfindsthattheyfailedtoimproveperformance.Itshedslightonsomereasonswhycontractsseemtoworklesswellingovernmentsettingsthanintheprivatesector.Writtencontractsbetweengovernmentsandstate-ownedenterprises(SOE)haveenjoyedavoguesincethemid-1980s,thankspartlytotheirembracebytheWorldBankandbilateraldonors.Oursurveyofdevelopingcountriesfound565suchcontractsforsomeofthelargestSOEsin32countriesplusanother103,000inChinaasofJune,1994ShirleyandXu3(table1).Wedefinedthesecontracts,whichwetermedperformancecontracts,asnegotiated,writtenagreementsbetweengovernmentsandthemanagersoftheirSOEswhichspecifyexplicittargetsthatSOEmanagementpledgestoachieveinagiventimeframe,andwhereperformanceismeasuredattheendofaspecifiedperiod.Thenextsectiondevelopsaconceptualframeworktoanalyzehowthreecontractingproblems--information,incentivesandcommitment--mightaffecttheimpactofincentivecontractsonSOEperformance.SectionIIIanalyzeshowasampleofcontractsdealtwitheachofthecontractingproblemsandSectionIVmeasuresthesuccessorfailureofthecontracts.SectionVconcludeswithanassessmentoftheimplicationsforincentivecontractsinpublicsettings.II.ConceptualFrameworkDrawingontheworkofSappington(1991),LafontandTirole(1986,1993)andWilliamson(1976and1985),1wehypothesizedthathowwellacontractdoesinimprovingSOEeconomicperformancedependsonhowwellitaddressesthreeinterrelatedproblems:informationasymmetry,incentives,andcommitment.Agencyproblemsarisebecausetheprincipalhasimperfectinformation;theagent’sactionscannotbeobservedbutcanonlybeimperfectlyinferredonthebasisofwhatcanbeobserved.Inthecaseofincentivecontractstheprincipalcannotmeasureaccuratelytheeffortexpendedbytheagentordistinguishtheeffectsofeffortfromotherfactorsaffectingperformance.Theagentfacesadisutilityofeffortandcanusetheinformationsheholdsexclusivelytoshirk.Toinducetheagenttocomplywiththeprincipal’sgoals,contractsarewrittenwhichincludepromisesofincentivesbasedonachievement.ShirleyandXu4Sinceitiscostlyfortheprincipaltopaytheincentive,shewilltrytomaximizeherwelfarewhileminimizingtheamounttransferredtothemanageroftheenterprise.InthecaseofSOEs,thecostsofincentivesincludeforegonerevenues(whichabureaucratmighthaveusedtoexpandherpowerorapoliticiantorewardhersupporters),thecostsofmonitoringandevaluatingtheSOE’sperformance,andanydistortionsintroducedbyassociatedtaxincreases.Themoreimperfecttheprincipal’sinformation,thelowertheprincipalwillsettherewardinordertoreducethecostassociatedwithpayinganincentivewithoutanygainsinperformance.Asaresulttherewardmaybesettoolowtomotivatemanagerstoimproveperformance.Incentivecontractswithgovernmentactorsconfrontadditionalproblemsbeyondtheusual.SOEshavenoclearresidualclaimants;insteadtheyaresubjecttomanyprincipalswhoimposemultiple,oftenconflictingobjectivesandconstraintsontheenterprise(Jones,1990).SomeofthesemultipleprincipalsmayderivebenefitsfromobjectiveswhichruncontrarytoimprovingSOEefficiency.Whileinsomecasesthesecontrarygoalsmaymaximizesocialwelfare,weassumewithShapiroandWillig(1990)thatgovernmentisnotalwaysbenevolent.Forexample,politiciansmaybenefitpoliticallyfrominstructingSOEstomaximizeemployment(ShleiferandVishny,1994).BureaucratsmightbenefitfrominstructingSOEstoengageinactivitiesthatincreasetheirpower,prestigeorperks(Niskanin,1991).Manyoftheseobjectivesarelikelytobemuchhardertocontractuponthanprofitmaximization(LaffontandTirole,1993).Moreover,tosafeguardtheirrents,th