Sunteți pe pagina 1din 2

Definition

Econometricsistheapplicationofstatisticalmethodstoeconomicdataandisdescribedasthe
branchofeconomicsthataimstogiveempiricalcontenttoeconomicrelations.Econometricsisan
amalgamofeconomictheory,mathematicaleconomics,economicstatistics,andmathematical
statistics.Economictheorymakesstatementsorhypothesesthataremostlyqualitativeinnature;
while,econometricsgivesempiricalcontenttomosteconomictheory.Forexample,microeconomic
theorystatesthat,otherthingsremainingthesame,areductioninthepriceofacommodityis
expectedtoincreasethequantitydemandedofthatcommodity.Thus,economictheorypostulatesa
negativeorinverserelationshipbetweenthepriceandquantitydemandedofacommodity.Butthe
theoryitselfdoesnotprovideanynumericalmeasureoftherelationshipbetweenthetwo;thatis,it
doesnottellbyhowmuchthequantitywillgoupordownasaresultofacertainchangeintheprice
ofthecommodity.Itisthejoboftheeconometriciantoprovidesuchnumericalestimates.
Econometricsallowseconomiststosiftthroughmountainsofdatatoextractsimplerelationships.
Econometriciansformulateastatisticalmodel,usuallybasedoneconomictheory,confrontitwith
thedata,andtrytocomeupwithaspecificationthatmeetstherequiredgoals.Precisely,
Econometricsisthequantitativeanalysisofactualeconomicphenomenabasedontheconcurrent
developmentoftheoryandobservation,relatedbyappropriatemethodsofinference.hefirst
knownuseoftheterm"econometrics"(incognateform)wasbyPolisheconomistPaweCiompain
1910.RagnarFrischiscreditedwithcoiningtheterminthesenseinwhichitisusedtoday.

Methodology

Econometricsusesafairlystraightforwardapproachtoeconomicanalysis.Althoughthereare
severalschoolsofthoughtoneconometricmethodology,thetraditionalorclassicalmethodology,
whichstilldominatesempiricalresearchineconomicsandothersocialandbehaviouralsciences
proceedsas:

1) Statementoftheoryorhypothesis:Thefirststeptoeconometricmethodologyistolookat
asetofdataanddefineaspecifichypothesisthatexplainsthenatureandshapeoftheset.
Forexample,Keynespostulatedthatthemarginalpropensitytoconsume(MPC),therateof
changeofconsumptionforaunit(say,adollar)changeinincome,isgreaterthanzerobut
lessthan1.
2) Specificationofmathematicalmodelofthetheory:Aneffectivemodeloutlinesaspecific
mathematicalrelationshipbetweentheexplanatoryvariableandthedependentvariable
beingtested.Amodelissimplyasetofmathematicalequations.Amathematicaleconomist
mightsuggestapreciseformofthefunctionalrelationshipregardingthephenomenon
subjectingthehypothesis.
Forexample,thefunctionalformoftheKeynesianconsumptionfunctionbeas:
Y=1+2X 0<2<1
whereY=consumptionexpenditure
X=income,and
1and2knownastheparametersofthemodel,theinterceptandslope
coefficients.Theslopecoefficient2measurestheMPC.
3) Specificationofthestatistical(oreconometric)model:Thepurelymathematicalisof
limitedinteresttotheeconometrician,foritassumesthatthereisanexactordeterministic
relationshipbetweentheexplanatoryvariableandthedependentvariable.Butrelationships
betweeneconomicvariablesaregenerallyinexact.Toallowfortheinexactrelationships
betweeneconomicvariables,theeconometricianwouldmodifythedeterministic
mathematicalmodelbyincludingdisturbanceterm(errorterm),whichisarandom
(stochastic)variablehavingwelldefinedprobabilisticproperties.Thedisturbancetermmay
wellrepresentalltheeffectivefactorsthatarenottakenintoaccountexplicitly.
Forexample,theeconometricmodelofKeynesianpostulatebeas:
Y=1+2X+u
whereuistheerror(disturbance)term
4) Obtainingthedata:Toestimatetheeconometricmodel,thatis,toobtainthenumerical
valuesoftheparametersofthemodel,weneeddata.
Forexample,wemayuseU.S.economicdatafortheperiod19811996,fromEconomic
ReportofthePresident(1998),toestimatenumericvaluesof1and2.
5) Estimationoftheparametersoftheeconometricmodel:Havingthedata,thenexttaskisto
estimatetheparametersofthemodel.Thenumericalestimatesoftheparametersgive
empiricalcontenttothemodel.
Forexample,usingregressionanalysistechniqueandthedata,weobtaintheestimatesof
1and2as:184.08and0.7064.
Thus,theestimatedconsumptionfunctionis:
Y=184.08+0.7064Xi
ThehatontheYindicatesthatitisanestimate
6) Hypothesistesting:AccordingtopositiveeconomistslikeMiltonFriedman,atheoryor
hypothesisthatisnotverifiablebyappealtoempiricalevidencemaynotbeadmissibleasa
partofscientificenquiry.So,assumingthatthefittedmodelisareasonablygood
approximationofreality,wehavetodevelopsuitablecriteriatofindoutwhetherthe
obtainedestimatesareinaccordwiththeexpectationsofthetheorythatisbeingtested.
Suchconfirmationorrefutationofeconomictheoriesonthebasisofsampleevidenceis
knownashypothesistesting.
Forexample,asnotedearlier,KeynesexpectedtheMPCtobepositivebutlessthan1.Inour
example,wefoundtheMPCtobeabout0.70.Butbeforeweacceptthisfindingas
confirmationofKeynesianconsumptiontheory,wemustenquirewhetherthisestimateis
sufficientlybelow1toconvinceusthatthisisnotachanceoccurrenceorpeculiarityofthe
particulardatawehaveused.Inotherwords,is0.70statisticallylessthan1?Ifitis,itmay
supportKeynestheory.
7) Forecastingorprediction:Ifthechosenmodeldoesnotrefutethehypothesisortheory
underconsideration,wemayuseittopredictthefuturevalue(s)ofthedependentvariable
(or,forecastvariable)Yonthebasisofknownorexpectedfuturevalue(s)oftheexplanatory
(orpredictor)variableX.
Forexample,puttingtheGDPvaluefor1997(7269.8billiondollars)ontheobtained
Keynesianmodel,weobtain:
Y1997=184.0779+0.7064(7269.8)
=4951.3167
Thus,giventhevalueoftheGDP,theaverageforecastconsumptionexpenditureisabout
4951billiondollars.
8) Usingthemodelforcontrolorpolicypurposes:Itispossible,andofagreatinterest,touse
anestimatedmodelforcontrol,orpolicy,purposes.Byappropriatefiscalandmonetary
policymix,thegovernmentcanmanipulatethecontrolvariable(s)toproducethedesired
levelofthetargetvariable.

S-ar putea să vă placă și