WorkingPaperNo.1023ClimateChangeandFiscalMarksmanship:EvidenceFromanEmergingCountry,IndiabyLekhaChakrabortyLevyEconomicsandInternationalInstituteofPublicFinance,MunichAjayNarayanJha(Former)GovernmentofIndiaandFinanceCommissionJiteshYadavandBalamuralyB(Former)NationalInstituteofPublicFinanceandPolicy(NIPFP)andAmandeepKaurNationalInstituteofPublicFinanceandPolicy(NIPFP)July2023ThispaperisanabridgedversionofthepaperinvitedforInternationalInstituteofPublicFinance(IIPF)MeetingsinUtahStateUniversity,August14–16,2023.ThanksareduetoProfessorPinakiChakraborty(AsianDevelopmentBank)forhisvaluablecomments.TheLevyEconomicsInstituteWorkingPaperCollectionpresentsresearchinprogressbyLevyInstitutescholarsandconferenceparticipants.Thepurposeoftheseriesistodisseminateideastoandelicitcommentsfromacademicsandprofessionals.LevyEconomicsInstituteofBardCollege,foundedin1986,isanonprofit,nonpartisan,independentlyfundedresearchorganizationdevotedtopublicservice.Throughscholarshipandeconomicresearch,itgeneratesviable,effectivepublicpolicyresponsestoimportanteconomicproblemsthatprofoundlyaffectthequalityoflifeintheUnitedStatesandabroad.LevyEconomicsInstituteP.O.Box5000Annandale-on-Hudson,NY12504-5000http://www.levyinstitute.orgCopyright©LevyEconomicsInstitute2023AllrightsreservedISSN1547-366X1ABSTRACTAccordingtothetheoryofefficientmarkets,economicagentsuseallavailableinformationtoformrationalexpectations.Therationalexpectationshypothesisassertsthatinformationisscarce,theeconomicsystemgenerallydoesnotwasteinformation,andthatexpectationsdependspecificallyonthestructureoftheentiresystem.Fiscalmarksmanship—theaccuracyofbudgetaryforecasting—canbeoneimportantpieceofsuchinformationthatrationalagentsmustconsiderinformingexpectations.Againstthebackdropoffiscalrules,ourpaperexploresthebudgetaryforecasterrorsofclimatechange–relatedpublicspendinginIndia.Thefiscalrulesstipulatethatfiscaldeficit–to–GDPratioshouldbemaintainedat3percent.However,inthepost-COVIDfiscalstrategy,amedium-termfiscalconsolidationpathof4.5percentfiscaldeficit–to–GDPisenvisionedby2025–26.Withinthisfiscalconsolidationframework,weanalyzedthebudgetcredibilityoffiscalcommitmentsforclimatechangeinIndia.Weanalyzedthefiscalbehavioralvariablesintermsofbias,variation,andrandomness,andcapturedthesystemicvariationsinbudgetaryforecastrelatedtoclimatechangeforaperiod2017–18to2020–21acrosssectors.Weidentifiedthesectorswheresystematiccomponentsofforecastingerrorsarerelativelyhigherthanrandomcomponents,whereminimizingerrorsthroughalteringthefiscalbehavioralmodelsisdonebyrevisingtheassumptionsandbyapplyingbetterforecastingmethods.Astate-leveldecompositionofthepublicspendingrevealedthatdisaggregatedfiscalspaceavailablefordevelopmentalspendingconstitutesaround60percentofthetotal.However,identifyingthespecificallytargetedpublicspendingrelatedtoclimatechangeacrossallstatesandanalyzingitsfiscalmarkmanshipcanfurtherthesubnationalinferences.KEYWORDS:FiscalMarksmanship,BudgetForecastErrors,ClimateChange,StateFinancesJELCODES:H30,H50,H70,Q5821.INTRODUCTIONAccordingtothetheoryofefficientmarkets,economicagentsuseallavailableinformationtoformrationalexpectations.Therationalexpectationshypothesisassertsthatinformationisscarce,theeconomicsystemgenerallydoesnotwasteinformation,andthatexpectationsdependspecificallyonthestructureoftheentiresystem.Fiscalmarksmanship—theaccuracyofbudgetaryforecasting—canbeoneimportantpiecethatrationalagentsmustconsiderinformingexpectations.Thesignificantvariationsbetweenactualrevenueandexpenditurefromtheforecastedbudgetarymagnitudescouldbeanindicationoffiscalpolicyobjectivesnotbeingoptimizedorattained.Inthiscontext,theroleofbudgetestimatesneedstobeemphasizedasafiscalsignal.Thispointhasgainedmuchmomentumespeciallywhenexpectationsarebased,notonwhathashappenedinthepast,butonthedatarelatingtothefuture.Thatis,ifexpectationsarerationalratherthanadaptive,itistheestimateoftaxesandspendinginanygivenbudget—theex-antedata,nottheobserveddata,availableonlywithalag—thatwillbeusedbyforward-lookingprivateagentswhobasetheirdecisions,inwholeorinpart,onfiscalvariables.Againstthebackdropoffiscalrules—legallymandatedundertheFiscalResponsibilityandBudgetManagementAct(FRBMA)—inIndia,ourpaperexploresthebudgetforecasterrorsofclimatechange–relatedpublicspendinginIndia.TheFRBMAstipulatesthatthefiscaldeficittoGDPratioshouldbemaintainedat3percent.However,inthepost-covidfiscalstrategy,amedium-termfiscalconsolidationpathof4.5percentfiscaldeficit-GDPisenvisionedfor2025–26.Withinthisfiscalconsolidationframework,weanalyzethebudgetcredibilityoffiscalcommitmentsforclimatechangeinIndia.ThisisparticularlyimportantagainstthebackdropofCOP27recentlyheldinEgypt.Thepaperanalyzesthefiscalbehavioralvariablesintermsofbias,randomness,andsystematicvariationsinbudgetaryforecast(forecasterrors)relatedtoclimatechange–relatedspendingforaperiod2017–18to2020–21.3InIndia,fiscalarithmetichasthreestages:theannouncementofBudgetEstimate(BE);thegovernment’sannouncementoftheRevisedEstimate(RE)thenextyear,afterreviewandrevision;andfinallythepublicationoftheActuals(actualspending)withalagofoneyearortwo.WeanalyzewhetherthereisacorrelationbetweenBEandRE,andbetweenBEandActualsorasignificantdeviationbetweenthesethree,relativetoclimatechange–relatedspending.2.REVIEWOFTHELITERATUREThepoliticaleconomyofbudgetdeficitsandothermacro-fiscalvariablesstartedgainingattentioninthe1990s(AlesinaandPerotti1995;Blanchard1990).However,oneoftheearlierdiscussionsoffiscalforecasterrorswasmadebyAllan(1965)inthecaseofBritain.AccordingtoAllan,fiscalmarksmanshipwasimportantduringthattimebecausethemarginforerrorwaslimitedgiventhetrade-offbetweeninflationandfullemployment.Insuchascenario,accuratepredictionsofbudgetaryestimateswereimportantformeetingfiscalpolicytargetsoffullemploymentwithoutundesirablyhighinflation.Davis(1980),followinguponAllan’sstudy,usedalongertimeseries(1951–78).Auld(1970)hasdoneafiscalmarksmanshipexerciseforCanadaforthepost-warperiod(through1968).Auldsaysthatifthegovernmentistofinanceitslong-rangeprograms,accuratepredictionsareimportant.Morrison(1986)hasdoneafiscalmarksmanshipexerciseintheUnitedStatesfortheyears1950–83.Cassidy,Kamlet,andNagin(1989)analyzedtherevenueforecastbiasesinthecontextofEurope.Goodfiscalmarksmanshipcanbeoneimportantpieceofavailableinformationrationalagentsmustconsiderinformingexpectations.Inthiscontext,theroleofbudgetestimatesneedstobeemphasizedaswhatDavis(1980)referstoasafiscalsignal,notingthatbudgetestimateshaveanimportant“signaleffect”foroutsideforecastersandanalysts,withparticularattention,inrecentyears,giventotheestimatedborrowingrequirement.Ifexpectationsarerationalratherthanadaptive,itistheestimateoftaxesandpublicexpenditureinanygivenbudget—theex-antedata,nottheobserveddata—thatwillbeusedbyforward-lookingprivateagentswhobasetheirdecisions,inwholeorinpart,onfiscalvariables(Morrison1986).4Inthecontextoftheeurozone,BrückandStephan(2005)haveestimatedthepoliticaleconomydeterminantsofbudgetdeficitforecasterrors.Theirfindingsshowthatpolitics,electoralcycles,andtheinstitutionaldesignsofgovernmentsaffectthequalityoffiscalforecasts.TheirfindingsagainstthebackdropoftheStabilityandGrowthPact(SGP)suggestmalignedincentivesfor“unobservablefiscaleffort”(BeetsmaandJensen2004)byeurozonegovernments(comparedtootherOECDgovernments)inreportingtheirbudgetdeficitspriortoelections.Theyexplainedthefiscalbehaviorunderthreecycles(i.e.,anelectoralforecastcycle,apartisanforecastcycle,andaninstitutionalcycle),applyingpaneleconometrictechniquestotheanalysisoftheforecasterrorsofbotheurozoneandnon-eurozoneOECDeconomies.Theirfindingssuggestthattheforecasterrorsalignwithelectioncyclesineurozonecountries.RullánandVillalonga(2018),inthecontextoftheSGP,haveexaminedtherelationshipbetweenfiscalrulesandbudgetaryforecastsbyanalyzingthesignificanceofpoliticalandinstitutionalvariablesintheeurozone.Theirfindingsshowthatthelevelofpublicsectordebtiscrucialinexplainingbudgetaryforecasterrors.Theelectoralmandate,politicalorientationofrulingparties,taxautonomy,andpercapitarevenuearetheothersignificantdeterminantsofforecasterrors.Thisstudytooktheliteratureforwardtosubnationaltiersofgovernmentin15Europeancountries,unliketheearlierstudiesinthecontextoftheeurozonethatconfinedtheiranalysistoamacroeconomicperspectiveatthenationalgovernmentlevel.TheSGPthereforecreatesincentivesforcreativebudgetarydeficitforecastspriortoelectioncycles(Strauch,Hallerberg,andHagen2004).Giuriato,Cepparulo,andBarberi(2016)analyzedthequalityofthefiscalforecastsof13eurozonecountriesbyusingannualforecastsfortheperiod1999–2013againstthebackdropofstabilityandconvergenceprograms.Theyfoundthatiffiscalrulescountertheexecutive’smonopolyonfiscalforecasting,strengtheningthelegislature’sformalpowersnegativelyinfluencesthefiscalforecastaccuracy.PinaandVenes(2011)analyzedthebudgetbalanceforecastspreparedby15EuropeancountriesintheirExcessiveDeficitProcedure(EDP)reporting.Theyfoundthatgrowthsurprises,fiscalinstitutions,electionscycles,formsoffiscal5governance,andnumericalexpenditurerules(unlikedeficitanddebtrules)affecttheforecasterrors.TherehavebeenanumberoffiscalmarksmanshipexercisesinthecaseofIndia(KumariandBhattacharya1988).InoneoftheearlierattemptsatanalyzingbudgetaryestimatesinIndia(fortheperiod1956–64),SamuelandRangarajan(1974)undertookananalysisoftwocomponentsofthestateandunionbudgets’capitalexpenditureonconstructionandindustrialdevelopment(theanalysiswaslimitedtothesetwobecauseofthescopeofthesubjectmattertheyweredealingwith).Inthisstudy,theanalysisofforecastingerrorswasbasedlargelyongraphsplottingtheactualexpenditureandthebudgetestimates.Intheiranalysis,itisstatedthat,whileinbothcomponentsthecentralgovernment’sbudgetestimatewasmoreaccuratecomparedtothestates,thisdifferencewasattributedtothedifferenceinbudgetaryprocess’sefficiency.Asher(1978)performedamorecomprehensivefiscalmarksmanshipexerciseforIndiafortheperiod1967–76,forboththerevisedandbudgetestimates.Thestudyshowedthat,duringthatperiod,boththerevenuesandexpenditureswereconsistentlyunderestimated.However,itwasobservedthattheextentoftheerrorontheexpendituresidewaslarger.ChakrabartyandVarghese(1982)haveuseddatafrom1970–80.Oneofthemajorfindingsofthatstudywasthatbothrevenuesandexpenditureareunderestimated.Pattnaik(1990)hasdoneafiscalmarksmanshipexerciseusingTheil’sindexfortheperiod1951–89.Thestudyobservesthattheerrorsintherevisedestimatesarelowerthantheerrorsinthebudgetestimates(althoughtherearelargeerrorsinboth).Itstatedthattheerrorsintheestimatesarelargelysystematicinnatureforboththeentiretimeperiodaswellasforsmallertimeperiodswithinthewhole(thesystematicerrorsweregreatestfortheperiod1981–89).MorerecentstudiesonfiscalmarksmanshipinIndiahaveadifferentconclusion.AstudydonebyNitinandRoy(2015)usingdatafrom1990–2012observesthatthesourceoferrorincomponentssuchastaxrevenue,nontaxrevenue,interestpayments,defenserevenueexpenditure,andfiscaldeficitwereprimarilyduetorandomerror(definedintheirpaperasbeingwhentheproportionoftherandomerrorisgreaterthanthebiascomponentsortheerrorin6variance).Therestofthecomponents—suchassubsidyexpenditure,capitalexpenditure,andnon-debtcapitalreceipts—hadahighersystematicerror(meanerrorandslopeerror).Averyinterestingpointmadeinthepaperisthat,whilethereisanattempttohavefiscalconsolidationbycontrollingexpenditure,thepredictabilityofexpenditureisquitelowcomparedtorevenue.Inasimilarstudy,ChakrabortyandSinha(2018)undertookafiscalmarksmanshipexercisefortheperiod1990–2017anddrewasimilarconclusion.Atrendthatcanbeobservedbasedontheempiricalliteraturefrom1951to1990isthatthesystematiccomponentoftheerrorwashigher,while,from1990to2017,therandomcomponentishigher.Itisworthnotingthattheabovestudiesarebasedonthefederalgovernment’sdata.ShresthaandChakraborty(2019)haveexaminedthefiscalmarksmanshipinthecontextofIndia’sstates.TheirstudyfocusedonKeralaandidentifiedforecasterrorswithrespecttotaxrevenueprojections.Intherecentempiricalliterature,thefiscalforecasterrorsareanalyzedagainstthebackdropoffiscalrules.Thepoliticaleconomyoffiscalforecastsatthesubnationalleveldependsonthetaxautonomyandthenatureoftheintergovernmentalfiscaltransfermechanism.Thetaxautonomyisheterogeneousacrossstates.Theintergovernmentalfiscaltransfersmaybeprogressiveifthetransferisdesignedtooffsettheinterstatefiscaldisabilities.InIndia,theFinanceBill2018hasincorporatedafewclauses(clauses207–10)toamendtheFRBMActof2003,withspecialemphasisontheeliminationofreferencesto“revenuebalance”andusingfiscaldeficitasanoperationalparameter(ChakrabortyandChakraborty2018).Againstthesepolicychanges,itispertinenttoanalyzetheimpactoffiscalrulesonfiscalmarksmanshipofmacro-fiscalvariablesinIndia.BuiterandPatel(2010)haveanalyzedfiscalrulesinIndia,howevertheeffectoffiscalrulesonfiscalmarksmanshipinthecontextofIndiahasnotbeenanalyzed.Asmentionedabove,NitinandRoy(2015)haveanalyzedthenormativefiscalassessmentsofIndia’sFinanceCommission,andtherealizationoffiscalpolicywithregardtothecentralgovernment’sfinancesovertheperiod1990–2012.7TherecentempiricalliteratureonfiscalmarksmanshipishighlyconfinedtotheIndiannationalgovernment’sforecasterrors(ChakrabortyandSinha2018;NitinandRoy2015).Therehasbeenvirtuallynoefforttoundertakeafiscalmarksmanshipexerciseatthestatelevel,exceptbyChakraborty,Chakraborty,andShrestha(2020).Inthispaper,weattempttodoafiscalmarksmanshipexerciseforclimatechange–relatedspending,analyzingthemagnitudeoftheerrorsandsubsequentlyexaminingthenatureoftheerrors.Thisisdoneintwoways:firstwecheckwhethertheerrorsareoverestimatesorunderestimates,andthenwechecktheextentofsystematicandrandomcomponentsinthesefiscalforecasterrors.3.METHODOLOGYOFFISCALMARKSMANSHIPThedataisorganizedfromthefinanceaccountsofvariousstatesandtheCentralStatisticsOffice(CSO).Theforecasterrorisdefinedasthedeviationbetweenthepredictedbudgetestimates(BE)orrevisedestimates(RE)andtheactual.TheMeanErrorThemeanerror(ME)referstotheaveragedifferencebetweentheforecastandtheactual.TheMEhasbeencalculatedbytakingtheaverageofthedifferencebetweenthepredictedvalues(ofbothBEandRE)andtheactuals.WehavedividedtheMEbythesumoftheactualsforthereferenceperiod.TheMEisacrudemeasureoftheforecast’squality,aspositiveandnegativeerrorscanoffseteachother,therebynotgivingustheexactmagnitudeoferror.However,theMEisanindicatorofpossiblebiasintheforecast.TheRootMeanSquareErrorTherootmeansquarederror(RMSE)isameasureoftherelativesizeoftheforecasterror.Inthispaper,tocalculatetheRMSE,themeansquarederror(MSE)istakenoverthereferenceperiodafterwhichthesquarerootoftheMSEiscalculated.Whilethiswillgiveusthemagnitudeoferror,itwillnotgiveanyinformationonthedirectionoftheerror,i.e.,whethertheerrorispositiveornegative.WehavetakentheRMSEasaproportionofthesumofactualsof8thereferenceperiod.Itreflectsthefactthatlargeforecasterrorsaremoresignificantthansmalldifferences.Theil’sInequalityCoefficients(U)Theil’sinequalitycoefficient(U)isusedtoanalyzethemeasureofaccuracyofthebudgetforecasts.Theil’sinequalitycoefficientisbasedontheMSE(U1).TheforecasterrorofTheil(1958)isdefinedas:U1=(1)WhereU1istheinequalitycoefficient,Ptisthepredictedvalue,Atistheactualvalue,andnisthenumberofyears.Thisinequalitycoefficientrangesfromzerotoone.WhenPtisequaltoAtforallobservations(aperfectforecast),U1equalszero.1U1hasbeendecomposedinordertoindicatesystematicandrandomsourcesoferror.Thesystematiccomponentisfurtherdividedintotheproportionofthetotalforecasterrorduetobias1Theil’ssecondequationfortheinequalitycoefficientusesarevisedmeasureofforecasterror.Theil’s(1966,1971)revisedmeasureofinequalityisasfollows:U2=ThismeasurehastheadvantagethatthedenominatordoesnotcontainPandtheinequalitycoefficientdoesnotdependontheforecast.Inaperfectforecast,U2equalstozero.U2doesnothaveanupperbound.AmorerigorousmeasureofTheil’sinequalitystatisticsisalsousedbyincorporatingthelagsintheactualsandthedifferenceofthepredictedvaluefromthelagoftheactualstocapturethemagnitudeoferror:U3=wherea=At-At-1,Pt=Pt-At-1,andn=numberofyearsååå+-222/1/1)(/1ttttAnPnAPnåå-22/1)(/1tttAnAPnååå+-222][/1][/1][/1atnPtnatPtn9andtheproportionoftotalforecasterrorattributabletounequalvariation.Thederivationofequation(2)isgivenindetailinDavis(1980).1=(2)Inequation(2),PandAaremean-predictedandmean-actualchanges,respectively;SpandSaarethestandarddeviationsofpredictedandactualvalues,respectively;andristhecoefficientofcorrelationbetweenpredictedandactualvalues.Thefirstexpressionontherighthandside(RHS)inequation(2)istheproportionofthetotalforecasterrorduetobias.Itrepresentsameasureoftheproportionoferrorduetooverpredictionorunderpredictionoftheaveragevalue.ThesecondexpressionoftheRHSinequation(2)istheproportionoftotalforecasterrorattributabletounequalvariation.Inotherwords,itmeasurestheproportionoferrorduetooverpredictionorunderpredictionofthevarianceofthevalues.ThethirdexpressionoftheRHSinequation(2)measurestheproportionofforecastingerrorduetorandomvariation.Thefirsttwosourcesoferroraresystematic;presumablytheycanbereducedbyimprovedforecastingtechniques,whiletherandomcomponentisbeyondthecontroloftheforecaster.4.MAGNITUDEOFFORECASTERRORSACROSSIDENTIFIEDSECTORSThedataonbudgetestimates,revisedestimates,andactuals,startingfrom2017–18until2020–21,forallclimateadaptation–relatedschemes,isextractedfromtheDetailDemandforGrants,UnionBudgetsdocumentsoftheGovernmentofIndia.Thereexistsahugevariationintheexpenditureincurredbyvariousministriestowardsadaptation-relatedprograms(Chakrabortyetal.,forthcoming).ååå--+--+--22222)(/1.)1(2)(/1)()(/1)(ttttttAPnSaSprAPnSaSpAPnAP10ThevaluesofU1,U2,andU3forvariousministriesareprovidedinTable1.U1takesonavaluebetween0and1.Therefore,itcanbedeterminedfromTable1thatthemagnitudeoferrors—inministriessuchastheMinistryofConsumerAffairs,FoodandPublicDistribution,andtheMinistryofScienceandTechnology—isquitesignificantataround0.5.However,theMinistryofScienceandTechnologydevotesarelativelyscantbudgetonadaptation-relatedprograms.Incontrast,theMinistryofConsumerAffairs,Food,andPublicDistributionspendssignificantlyonadaptation-relatedprograms.U1forBEwasreportedtobelowestfortheMinistryofRoadTransportandHighways(0.04)asshowninTable1.ThevalueofU1forREwashighestfortheMinistryofScienceandTechnology(0.66)andnegligiblefortheMinistryofLawandJustice,MinistryofHeavyIndustries,andMinistryofSteel.Table1:FiscalMarksmanship:Theils’InequalityStatistic(U)NameofMinistry/DepartmentTheils'U(BE,Actual)Theils'U(RE,Actual)U1U2U3U1U2U3MinistryofAgricultureandFarmersWelfare0.1150.2590.3930.0760.1610.301DepartmentofAtomicEnergy0.0980.1970.7320.0220.0440.123MinistryofAyush0.1680.3841.1320.0260.0530.312MinistryofChemicalsAndFertilizers0.1740.3151.1290.0170.0350.085MinistryofCoal0.1550.3540.8750.0510.1070.334MinistryofCommerceandindustry0.0800.1540.8250.0420.0850.441MinistryofConsumerAffairs,FoodandPublicDistribution0.4960.7670.9770.1280.2390.240MinistryofDevelopmentofNorthEasternRegion0.1870.4410.7050.0680.1410.227MinistryofEarthSciences0.1310.2901.2450.0190.0390.149MinistryofEducation0.0640.1360.8410.0280.0580.509MinistryofEnvironment,ForestsandClimateChange0.1200.2621.3460.0130.0260.130MinistryofExternalAffairs0.0950.2020.5340.0130.0250.110MinistryofFinance0.1800.3880.9780.0360.0740.159MinistryofFisheries,AnimalHusbandryandDairying0.0610.1260.3700.0070.0130.025MinistryofFoodProcessingIndustries0.2220.5290.4890.0950.2050.25311MinistryofHealthandFamilyWelfare0.0960.1781.1020.0130.0260.110MinistryofHeavyIndustries0.1970.4470.6770.0000.0000.000MinistryofHomeAffairs0.0760.1520.6130.0190.0370.112MinistryofHousingandUrbanAffairs0.1100.2120.5950.0380.0760.047MinistryofInformationandBroadcasting0.3431.0420.5590.0040.0090.007MinistryofJalShakti0.0870.1821.2910.0290.0580.428MinistryofLawandJustice0.1110.2150.5590.0000.0010.002MinistryofMicro,SmallandMediumEnterprises0.0780.1590.5780.0050.0100.023MinistryofMines0.2220.4710.5330.2140.4430.507MinistryofMinorityAffairs0.0830.1780.7960.0660.1370.606MinistryofNewandRenewableEnergy0.2230.5440.9480.0540.1140.287MinistryofPanchayatiRaj0.1740.4110.9750.0130.0260.088MinistryofPetroleumandNaturalGas0.1270.2750.5710.0210.0430.090MinistryofPower0.1230.2561.0220.0810.1580.493MinistryofRailways0.4500.6480.9660.2360.6090.422MinistryofRoadTransportandHighways0.0380.0740.3520.0160.0330.131MinistryofRuralDevelopment0.1480.2701.0100.0020.0040.011MinistryofScienceandTechnology0.4961.6940.8710.6633.2350.871MinistryofSkillDevelopmentandEntrepreneurship0.1200.2700.6730.0320.0660.255MinistryofSocialJusticeandEmpowerment0.0650.1280.5270.0080.0160.051DepartmentofSpace0.1000.2070.8840.0040.0080.026MinistryofSteel0.3030.6230.5750.0000.0000.000MinistryofTextiles0.0750.1520.6720.0250.0500.096MinistryofTourism0.1960.4080.4470.0000.0000.000MinistryofTribalAffairs0.0730.1480.9420.0040.0080.009MinistryofWomenandChildDevelopment0.1310.2931.1110.0920.0920.519Source:(Basicdata),FinanceAccounts(variousyears),GovernmentofIndia5.PARTITIONINGTHEBUDGETFORECASTERRORSTable2givestheresultsobtainedafterpartitioningtheforecasterrorsinbudgetestimatesintosystematicandrandomcomponents.Systematicerrorcanbeimprovedupon,butrandomisbeyondtheforecaster'scontrol.Inthecaseofbudgetestimates,theMinistryofScienceand12TechnologyandtheMinistryofInformationandBroadcastingreportedthehighestsystematicerrors,whereastheMinistryofSocialJusticeandEmpowermentreportedthelowestsystematicerrorataround0.07(Table2).Table2:PartitioningtheSourcesofForecastErrors:BiasandRandomComponentsNameofMinistryBIASUnequalRandomMinistryofAgricultureandFarmersWelfare0.630.330.04DepartmentofAtomicEnergy0.010.400.59MinistryofAyush0.460.160.38MinistryofChemicalsandFertilizers0.230.510.25MinistryofCoal0.610.090.30MinistryofCommerceandindustry0.300.390.31MinistryofConsumerAffairs,FoodandPublicDistribution0.080.570.36MinistryofDevelopmentofNorthEasternRegion0.730.030.23MinistryofEarthSciences0.540.030.43MinistryofEducation0.700.250.05MinistryofEnvironment,ForestsandClimateChange0.540.020.45MinistryofExternalAffairs0.500.010.49MinistryofFinance0.050.300.65MinistryofFisheries,AnimalHusbandryandDairying0.360.010.63MinistryofFoodProcessingIndustries0.440.090.47MinistryofHealthandFamilyWelfare0.600.190.22MinistryofHeavyIndustries0.440.020.54MinistryofHomeAffairs0.000.220.78MinistryofHousingAndUrbanAffairs0.070.270.66MinistryofInformationandBroadcasting0.490.500.00MinistryofJalShakti0.350.010.64MinistryofLawandJustice0.040.400.55MinistryofMicro,SmallandMediumEnterprises0.070.050.88MinistryofMines0.230.570.20MinistryofMinorityAffairs0.610.010.38MinistryofNewandRenewableEnergy0.650.000.35MinistryofPanchayatiRaj0.790.030.18MinistryofPetroleumandNaturalGas0.240.220.5313MinistryofPower0.160.190.65MinistryofRailways0.290.680.03MinistryofRoadTransportandHighways0.070.300.63MinistryofRuralDevelopment0.290.640.07MinistryofScienceandTechnology0.890.110.00MinistryofSkillDevelopmentandEntrepreneurship0.930.010.06MinistryofSocialJusticeAndEmpowerment0.030.040.93DepartmentofSpace0.090.020.90MinistryofSteel0.070.150.78MinistryofTextiles0.020.120.86MinistryofTourism0.250.300.45MinistryofTribalAffairs0.070.050.88MinistryofWomenAndChildDevelopment0.640.060.30Source:(Basicdata),FinanceAccounts(variousyears),GovernmentofIndiaTable2givestheresultsobtainedafterbifurcatingtheerrorsinrevisedestimatesintosystematicandrandomcomponents.Inthecaseofrevisedestimates,theMinistryofRuralDevelopmentandtheMinistryofRailwaysreportedthehighestsystematicerrorswhereastheMinistryofTourismandtheMinistryofSteelreportedthelowestsystematicerrorataround0.07(Table2).Forbothbudgetandrevisedestimates,thereisspaceforimprovementinforecastingerrorsincethesystematiccomponentisgreaterthantherandomcomponentinamajorityofministries.Thisimpliesthatthefiscalmarksmanshipmaybeenhancedbyusingmoreeffectivepolicyinnovationstomanagethetightfiscalspacewithinthefiscalregulations.6.SUBNATIONALFISCALSPACEFORCLIMATECHANGECOMMITMENTSTheStateActionPlans(SAPs)forclimatechangecommitmentsarenothomogeneous,andeachstateinIndiahaspreparedtheSAPasperthespecificitiesofclimatechange–relatedrisksanduncertainities.Theestimatesintheprevioussection,however,areconfinedtotheDemandforGrantsanalysisofnationalbudgets.Giventheprincipleofsubsidiarity,thedecisionsrelatedtoclimatechangeconsiderations—especiallyadaptation—needtobetakenatthelevelclosesttothepeople.Themeticulousanalysisofallthedetaileddemandsforgrantsacrossallstatesof14Indiascannedfortheintensityofidentifiedcomponentsofadaptationisataskbeyondthescopeofthepresentpaper.However,attheaggregatelevel,weidentifiedtheplausiblediscretionaryfiscalspaceavailableattheaggregateleveltothestategovernmentstoundertaketheclimatechangecommitments.Theclassificationofbudgetarytransactionsintodevelopmental(economicservicesandsocialservices)andnon-developmental(generalservicesincludinginterestpayments,salaries,andpensions)forthepurposeofidentifyingtheplausiblefiscalspaceforclimatechangecommitments.Attheaggregatestatelevel,theReserveBankofIndiaanalysisofpublicexpenditureacrossthestatesofIndiarevealedthatdevelopmentalspendingisaround60percentoftotalpublicexpenditure(Table3).Table3:StateLevelDevelopmentandNon-DevelopmentExpenditureas%ofAggregatePublicExpenditureYearDevelopmentNon-DevelopmentOthersTotal(incrores)123452004–052,86,473.01,85,152.081,803.05,53,428.0(51.8)(33.5)(14.8)(100.0)2005–063,30,044.11,90,020.641,616.85,61,681.6(58.8)(33.8)(7.4)(100.0)2006–073,92,165.02,11,872.453,242.96,57,280.3(59.7)(32.2)(8.1)(100.0)2007–084,64,462.02,33,232.854,629.67,52,324.4(61.7)(31.0)(7.3)(100.0)2008–095,67,086.22,54,981.460,265.28,82,332.8(64.3)(28.9)(6.8)(100.0)2009–106,37,731.13,07,547.070,051.710,15,329.8(62.8)(30.3)(6.9)(100.0)2010–117,20,354.73,57,287.481,087.611,58,729.7(62.2)(30.8)(7.0)(100.0)2011–128,52,405.64,01,059.498,147.313,51,612.3(63.1)(29.7)(7.3)(100.0)2012–139,72,256.54,46,878.91,15,119.415,34,254.8(63.4)(29.1)(7.5)(100.0)2013–1410,76,452.25,04,548.41,25,144.017,06,144.515(63.1)(29.6)(7.3)(100.0)2014–1513,25,989.25,66,467.41,33,326.020,25,782.5(65.5)(28.0)(6.6)(100.0)2015–1615,84,006.26,29,349.31,46,873.223,60,228.7(67.1)(26.7)(6.2)(100.0)2016–1718,31,163.87,10,365.11,66,686.427,08,215.3(67.6)(26.2)(6.2)(100.0)2017–1818,77,392.38,25,774.02,21,432.929,24,599.2(64.2)(28.2)(7.6)(100.0)2018–1921,00,801.69,44,483.72,92,428.133,37,713.3(62.9)(28.3)(8.8)(100.0)2019–2021,63,340.610,05,162.73,26,499.334,95,002.6(61.9)(28.8)(9.3)(100.0)2020–2122,64,470.710,63,162.23,69,859.436,97,492.3(61.2)(28.8)(10.0)(100.0)2021–22(BE)29,11,369.412,87,938.24,23,804.246,23,111.7(63.0)(27.9)(9.2)(100.0)2021–22(RE)29,22,422.812,40,854.34,33,289.045,96,566.0(63.6)(27.0)(9.4)(100.0)2022–23(BE)32,34,504.414,18,957.34,79,783.651,33,245.2(63.0)(27.6)(9.3)(100.0)Note:RE:RevisedEstimates.BE:BudgetEstimates.:IncludesexpenditureonrevenueandcapitalaccountandloansandadvancesextendedbystategovernmentandUTs.:IncludesGrants-in-AidandContributions(CompensationandAssignmentstoLocalBodies),DischargeofInternalDebtandRepaymentofLoanstotheCentre.Figuresinparenthesesarepercentagetototal.Datafrom2017–18onwardsincludeDelhiandPuducherryalso.Source:RBIandBudgetdocumentsoftheStategovernments(variousyears)Followinganopen-endedapproach,thepublicexpenditureforclimatechange,specificallyadaptation-relatedspendingcanbeidentifiedbasedon(i)cropimprovementandresearch;(ii)droughtproofingandfloodcontrol;(iii)forestconservation;(iv)povertyalleviationandlivelihoodpreservation;(v)ruraleducationandinfrastructure;(vi)health;(vii)riskfinancing;and(viii)disastermanagementacrossallstatesofIndia.Thisidentificationisthecrucialpreludetoasubnational,fiscalmarksmanshipanalysisofclimatecommitmentsbyallthestategovernments.State-specificmappingofadaptation-relatedfiscalspaceanditsmarksmanshipisanareaoffutureresearch.167.CONCLUSIONThepaperconductstheministry-wisefiscalmarksmanshipanalysisforclimatechangespending.Thesourcesoferrors,disaggregatedintobiasedness,unequalvariation,andrandomcomponentsareanalysedacrosssectors.Inthesectorswherethesystematiccomponentsofforecastingerrorsarerelativelyhigher,theerrorscanbereducedbyusingbetterforecastingmethods.Astate-leveldecompositionofthefiscalmarksmanshipestimatestounderstandthesourcesoferrors—systemicorrandombias—isanareaoffutureresearch,whichwouldbeconductedonlyaftersortingouttheintertemporalcomparabilityissuesinthedetaileddemandforgrantsacrosssectors.Inthispaper,theanalysisisconfinedtoidentifyingonlythefiscalspacefordevelopmentalspendinganditisrevealedthatthediscretionaryfiscalspaceavailableforplausibleclimatechangecommitmentsinthedevelopmentalspendingcategoryconstitutesaround60percentoftotalspending.Identifyingthespecificallytargetedpublicspendingrelatedtoclimatechangeacrossallstatesandanalyzingitsfiscalmarksmanshipcanfurtherthesubnationalinferences,whichisanareaoffutureresearch.17REFERENCESAlesina,A.,andR.Perotti.1995.“ThePoliticalEconomyofBudgetDeficits.”IMFStaffPaper42(1):1–31.Allan,C.M.1965.“FiscalMarksmanship,1951–63.”OxfordEconomicPapers(NewSeries)17(2):317–27.Auld,D.A.L.1970.“FiscalMarksmanshipinCanada.”TheCanadianJournalofEconomics3(3):507–11.Asher,M.G.1978.“AccuracyofBudgetaryForecastsofCentralGovernment,1967–68to1975–76.”EconomicandPoliticalWeekly13(8).Blanchard,O.1990.“SuggestionsforaNewSetofFiscalIndicators.”OECDEconomicsDepartmentWorkingPapersNo.39.Beetsma,R.M.W.J.,andH.Jensen.2004.“Mark-UpFluctuationsandFiscalPolicyStabilizationinaMonetaryUnion.”JournalofMacroeconomics26:357–76.Brück,T.,andA.Stephan,2005.“DoEurozoneCountriesCheatwiththeirBudgetDeficitForecasts?”Europa-UniversitätViadrinaFrankfurt(Oder)PostgraduateResearchProgrammeWorkingPaperNo.2005.5.Frankfurt:EuropeanUniversityViadrinaFrankfurt(Oder).Cassidy,G.,M.S.Kamlet,andD.S.Nagin.1989.“AnEmpiricalExaminationofBiasinRevenueForecastsbyStateGovernments.”InternationalJournalofForecasting5(3):321–31.Chakraborty,L.,A.N.Jha,A.Kaur,J.Yadav,andBalamuralyB.,(forthcoming).“COP27andPublicExpenditureforIndia’sFirstNationalAdaptationCommunication.”NIPFPWorkingPaper.NewDelhi:NationalInstituteofPublicFinanceandPolicy(NIPFP).Chakraborty,L.,andP.Chakraborty.2018."Federalism,FiscalAsymmetriesandEconomicConvergence:EvidencefromIndianStates."NIPFPNo.232.NewDelhi:NationalInstituteofPublicFinanceandPolicy(NIPFP).Chakraborty,L.,andD.Sinha.2018.“HasFiscalRuleChangedtheFiscalMarksmanshipofUnionGovernment?”NIPFPWorkingPaperNo.234.NewDelhi:NationalInstituteofPublicFinanceandPolicy(NIPFP).Chakraborty,L.,P.Chakraborty,andR.Shrestha.2020.“BudgetCredibilityofSubnationalGovernments:AnalyzingtheFiscalForecastingErrorsof28StatesinIndia.”LevyEconomicsInstituteWorkingPaperNo.964.NewYork.18Chakrabarty,T.K.,andW.Varghese.1982.“TheGovernmentofIndia'sBudgetEstimation:AnAnalysisoftheErrorComponents.”ReserveBankofIndia,OccasionalPapers3(2):175–90.Davis,J.M.1980.“FiscalMarksmanshipintheUnitedKingdom,1951–1978.”TheManchesterSchool48(2):111–28.Kumari,A.,andB.B.Bhattacharya.1988.“BudgetForecastsofCentralGovernmentRevenueandExpenditure:ATestofRationalExpectation.”EconomicandPoliticalWeekly23(26).Giuriato,L.,A.Cepparulo,andM.Barberi.2016."FiscalForecastsandPoliticalSystems:ALegislativeBudgetingPerspective.”PublicChoice,Springer.168(1):1–22(July).Morrison,R.J.1986.“FiscalMarksmanshipintheUnitedStates:1950–1983.”TheManchesterSchool54(3):322–33.Nitin,K.,andR.Roy.2015.”FinanceCommissionofIndia’sAssessment:APoliticalEconomyContentionbetweenExpectationsandOutcomes.”AppliedEconomics48(2):73–88.Pattnaik,R.K.1990.“FiscalMarksmanshipinIndia.”ReserveBankofIndiaOccasionalPapers11(3).Pina,A.,andN.Venes.2011.“ThePoliticalEconomyofEDPFiscalForecast:AnEmpiricalAssessment.”EuropeanJournalofPoliticalEconomy27(3):534–46.Rullán,X.O.,andJ.R.Villalonga.2018.“DeterminantsofRegionalBudgetForecastErrorsinFederalEconomies:Spain1995–2013.”ReviewofPublicEconomics226(4):85–121.Samuel,P.,andC.Rangarajan.1974.Short-TermInvestmentForecasting.Noida:MacmillanIndia.Shrestha,R.,andL.Chakraborty.2019.“PractisingSubnationalPublicFinanceinanEmergingEconomy:FiscalMarksmanshipinKerala.”NIPFPWorkingPaperNo.261.NewDelhi:NationalInstituteofPublicFinanceandPolicy.Strauch,R.,M.Hallerberg,andJ.V.Hagen.2004.“BudgetaryForecastsinEurope–TheTrackRecordofStabilityandConvergenceProgrammes”ECBWorkingPaperNo.307.Germany:EuropeanCentralBank(ECB)Theil,H.1958.EconomicForecastsandPolicy.Amsterdam:NorthHolland.———.1966.AppliedEconomicForecasting.Amsterdam:NorthHolland.Willem,H.Buiter,andUrjit,R.Patel.2010."FiscalRulesinIndia:AreTheyEffective?."NBERWorkingPapers15934.NationalBureauofEconomicResearch,Inc.
VIP
