43
Guide to scoring evidence using the Maryland Scientific Methods Scale Updated June 2016

Guide to scoring evidencemethod, scoring them on the quality of their implementation. This scoring guide plays several important roles. Firstly, consistent with our commitment to openness,

  • Upload
    others

  • View
    44

  • Download
    0

Embed Size (px)

Citation preview

Page 1: Guide to scoring evidencemethod, scoring them on the quality of their implementation. This scoring guide plays several important roles. Firstly, consistent with our commitment to openness,

Guide to

scoring evidenceusing the Maryland

Scientific Methods ScaleUpdated June 2016

Page 2: Guide to scoring evidencemethod, scoring them on the quality of their implementation. This scoring guide plays several important roles. Firstly, consistent with our commitment to openness,

Contents

1. Introduction 1

2. Impact evaluation 3

Using Counterfactuals 3

3. SMS 5 methods 6

Randomised Control Trial (RCT) 6

4. SMS 4 methods 10

Instrumental variables (IV) 10

Regression Discontinuity Design (RDD) 13

5. SMS 3/4 methods’ 17

6. SMS 3 methods 18

Difference-in-differences (DiD) 19

Panel data methods 20

Propensity Score Matching (PSM) 23

7. SMS 2 methods and below 26

Cross-sectional regression 26

Before-and-after 27

Additionality (not SMS scoreable) 28

Impact modelling (not SMS scoreable) 29

Appendix 1: Quick-scoring scoring guide for the Maryland Scientific Method Scale 30

Appendix 2: Mixed methods scoring 4 or 3. 33

Hazard Regressions 33

Heckman two-stage correction (H2S)/ Control function (CF) 36

Appendix 3: Arellano-Bond method 39

00

Page 3: Guide to scoring evidencemethod, scoring them on the quality of their implementation. This scoring guide plays several important roles. Firstly, consistent with our commitment to openness,

Guide to scoring evidence using the Maryland Scientific Method Scale - Updated June 2016 1

01

Introduction

TheWhatWorksCentreforLocalEconomicGrowth(WWG)producessystematicreviewsoftheevidencebaseonabroadrangeofpoliciesintheareaoflocaleconomicgrowth.Animportantstepinthereviewprocessistheassessmentofwhetheranevaluationprovidesconvincingevidenceonlikelypolicyimpacts.OurassessmentisbasedonthescoringofpapersontheMarylandScientificMethodsScale(SMS),whichrankspolicyevaluationsfrom1(leastrobust)to5(mostrobust)accordingtotherobustnessofthemethodusedandthequalityofitsimplementation.Robustness,asjudgedbytheMarylandSMS,istheextenttowhichthemethoddealswiththeselectionbiasesinherenttopolicyevaluations.

ThisdocumentexaminesawidevarietyofcommonlyemployedmethodsandexplainshowweplacethemontheMarylandSMS.Itthenlooksatanumberofexamplesofpolicyevaluationsforeachmethod,scoringthemonthequalityoftheirimplementation.

Thisscoringguideplaysseveralimportantroles.Firstly,consistentwithourcommitmenttoopenness,itservesasdocumentationastohowwerankstudiesbyrobustnessforourseriesofsystematicreviewsandenablesustobeconfidentthatweareachievingalevelofconsistencyinourrankingacrosstheteam.Secondly,itactsasascoringhandbookforanyonewantingtoassesstherobustnessofaparticularpolicyevaluation.Thisprovidesusefulguidanceonhowmuchweighttoputonaparticularpieceofevidence.Finallyitcanhelporganisationsundertakingevaluationseitherinassessingthemaftercompletionorinhelpingchoosebetweenmethodologiesbeforehand.

Thisdocumentisnosubstituteforbettertechnicaltrainingandexpertadvice.Butitshouldhelpthosewithsomeknowledgeofevaluationtechniquestobetterunderstandrecentadvancesandthewaythatwetreattheseinoursystematicreviews.It’salsoimportanttonotethattherankingofindividualstudiesisnotanexactscienceandofteninvolvesadegreeofjudgement.Statisticaltestingcanonlytakeussofarinassessingthesuitabilityofagivenmethodandthequalityofitsapplicationinpractice.Whenassessinganindividualstudy(includingthespecificexamplesthatwediscusshere)thereisalwaysscopeforsomedisagreementontheexactranking.Indeed,anyonewhohasattendedanacademicseminarwillknowtheextenttowhichsuchissuescanbehotlydisputed.Thatsaid,onaverageourscoringwilltendtoproducerankingsonwhichmanyevaluationexpertswouldbroadly

Page 4: Guide to scoring evidencemethod, scoring them on the quality of their implementation. This scoring guide plays several important roles. Firstly, consistent with our commitment to openness,

Guide to scoring evidence using the Maryland Scientific Method Scale - Updated June 2016 2

agree.Ofcourse,whenpullingtogetherourevidencereviewswelookattheoverallbalanceoftheevidenceandit’sthereforehighlyunlikelythataspecificrankingonanygivenstudywillinfluencetheconclusionsthatwereachonpolicyeffectiveness.

Theexamplesthatweusearedrawnfromawiderangeofstudies.Notallofthemarespecificallyfocussedonlocaleconomicgrowthbutinallcasestheydemonstrateapproachesthatcouldfeasiblybetakeninfutureevaluations.

Page 5: Guide to scoring evidencemethod, scoring them on the quality of their implementation. This scoring guide plays several important roles. Firstly, consistent with our commitment to openness,

Guide to scoring evidence using the Maryland Scientific Method Scale - Updated June 2016 3

Impact evaluation

Governmentsaroundtheworldincreasinglyhavestrongsystemstomonitorpolicyinputs(suchastheamountofloansguaranteedtoSMEs,whichpolicymakershopewillcreatenewjobs)andoutputs(suchasthenumberoffirmsthathavereceivedloanguarantees).However,theyarelessgoodatidentifyingpolicyoutcomes(suchastheeffectofprovidingaloanguaranteeonfirmemployment).Inparticular,manygovernment-sponsoredevaluationsthatlookatoutcomesdonotusecrediblestrategiestoassessthecausalimpactofpolicyinterventions.

Bycausalimpact,theevaluationliteraturemeansanestimateofthedifferencethatcanbeexpectedbetweentheoutcomefor(inthisexample)firms‘treated’inaprogramme,andtheaverageoutcometheywouldhaveexperiencedwithoutit.Pinningdowncausalityisacruciallyimportantpartofimpactevaluation.Estimatesofthebenefitsofaprojectareoflimitedusetopolicymakersunlessthosebenefitscanbeattributed,withareasonabledegreeofcertainty,tothatproject.

Thecredibilitywithwhichevaluationsestablishcausalityisthecriteriononwhichthisreviewassessestheliterature.

Using CounterfactualsEstablishing causality requires the construction of a valid counterfactual–i.e.whatwouldhavehappenedtoprogrammeparticipantshadtheynotbeentreatedundertheprogramme.Thatoutcomeisfundamentallyunobservable,soresearchersspendagreatdealoftimetryingtorebuildit.Thewayinwhichthiscounterfactualis(re)constructedisthekeyelementofimpactevaluationdesign.

A standard approach is to create a counterfactual group of similar places, people or businesses not undertaking the kind of project being evaluated. Changesinoutcomescanthenbecomparedbetweenthe‘treatmentgroup’(thoseaffectedbythepolicy)andthe‘controlgroup’(similarplaces,peopleorbusinessesnotexposedtothepolicy).

A key issue in creating the counterfactual group is dealing with the ‘selection into treatment’ problem.Selectionintotreatmentoccurswhenparticipantsintheprogrammedifferfromthosewhodonotparticipateintheprogramme.

02

Page 6: Guide to scoring evidencemethod, scoring them on the quality of their implementation. This scoring guide plays several important roles. Firstly, consistent with our commitment to openness,

Guide to scoring evidence using the Maryland Scientific Method Scale - Updated June 2016 4

Anexampleofthisproblemforaccesstofinanceprogrammes,asinourexampleabove,wouldbewhenmoreambitiousfirmsapplyforsupport.Ifthishappens,estimatesofpolicyimpactmaybebiasedupwardsbecauseweincorrectlyattributebetterfirmoutcomestothepolicy,ratherthantothefactthatthemoreambitiousparticipantswouldhavedonebetterevenwithouttheprogramme.

Selectionproblemsmayalsoleadtodownwardbias.Forexample,firmsthatapplyforsupportmightbeexperiencingproblemsandsuchfirmsmaybelesslikelytogroworsucceedindependentofanyadvicetheyreceive.Thesefactorsareoftenunobservabletoresearchers.

So the challenge for good programme evaluation is to deal with these issues, and to demonstrate that the control group is plausible.Iftheconstructionofplausiblecounterfactualsiscentraltogoodpolicyevaluation,thenthecrucialquestionbecomes:how do we design counterfactuals? Thisscoringguideexplainsthewaysinwhichweassesshowwellresearchershavedoneinansweringthisquestion.InparticularitexplainshowwerankimpactevaluationsbasedonanadjustedversionoftheMarylandScientificMethodsScale(SMS)todothis.1TheSMSisafive-pointscalerangingfrom1,forevaluationsbasedonsimplecrosssectionalcorrelations,to5forrandomisedcontroltrials.Box1providesanoverviewofthescaleandhighlightssomeoftheapproachesthatwediscussinmoredetailintheremainderofthisguide.

1 Sherman,Gottfredson,MacKenzie,Eck,Reuter,andBushway(1998).

Page 7: Guide to scoring evidencemethod, scoring them on the quality of their implementation. This scoring guide plays several important roles. Firstly, consistent with our commitment to openness,

Guide to scoring evidence using the Maryland Scientific Method Scale - Updated June 2016 5

Box 1: Our robustness scores (based on an adjusted Maryland Scientific Methods Scale)

Level 1:Either (a) across-sectional comparison of treated groups with untreated groups, or (b) a before-and-after comparison of treated group, without an untreated comparison group.Nouseofcontrolvariablesinstatisticalanalysistoadjustfordifferencesbetweentreatedanduntreatedgroupsorperiods.

Level 2:Use of adequate control variables and either (a) across-sectional comparison of treated groups with untreated groups, or (b) a before-and-after comparison of treated group, without an untreated comparison group. In(a), controlvariablesormatchingtechniquesusedtoaccountforcross-sectionaldifferencesbetweentreatedandcontrolsgroups.In(b),controlvariablesareusedtoaccountforbefore-and-afterchangesinmacrolevelfactors.

Level 3:Comparison of outcomes in treated group after an intervention, with outcomes in the treated group before the intervention, and a comparison group used to provide a counterfactual (e.g. difference in difference). Justificationgiventochoiceofcomparatorgroupthatisarguedtobesimilartothetreatmentgroup.Evidencepresentedoncomparabilityoftreatmentandcontrolgroups.Techniquessuchasregressionand(propensityscore)matchingmaybeusedtoadjustfordifferencebetweentreatedanduntreatedgroups,buttherearelikelytobeimportantunobserveddifferencesremaining.

Level 4:Quasi-randomness in treatment is exploited, so that it can be credibly held that treatment and control groups differ only in their exposure to the random allocation of treatment.Thisoftenentailstheuseofaninstrumentordiscontinuityintreatment,thesuitabilityofwhichshouldbeadequatelydemonstratedanddefended.

Level 5: Reserved for research designs that involve explicit randomisation into treatment and control groups, with Randomised Control Trials (RCTs) providing the definitive example. Extensiveevidenceprovidedoncomparabilityoftreatmentandcontrolgroups,showingnosignificantdifferencesintermsoflevelsortrends.Controlvariablesmaybeusedtoadjustfortreatmentandcontrolgroupdifferences,butthisadjustmentshouldnothavealargeimpactonthemainresults.Attentionpaidtoproblemsofselectiveattritionfromrandomlyassignedgroups,whichisshowntobeofnegligibleimportance.Thereshouldbelimitedor,ideally,nooccurrenceof‘contamination’ofthecontrolgroupwiththetreatment.

Note:TheselevelsarebasedonbutnotidenticaltotheoriginalMarylandSMS.Thelevelsherearegenerallyalittlestricterthantheoriginalscaletohelptoclearlyseparatelevels3,4and5whichformthebasisforourevidencereviews.

Page 8: Guide to scoring evidencemethod, scoring them on the quality of their implementation. This scoring guide plays several important roles. Firstly, consistent with our commitment to openness,

Guide to scoring evidence using the Maryland Scientific Method Scale - Updated June 2016 6

SMS 5 methods

SMS5methodsrequirefullrandomisationofprogrammeparticipation.Thiscanonlybedonebyarandomisedcontroltrial.AssuchthereisonlyonemethodinthiscategoryontheSMS.Randomisation,properlyapplied,meansthereisnoselectionintothetreatment.Thisensuresthattherearenodifferencesbetweenthetreatmentgroupandcontrolgroupeitheronobservable(e.g.age)orunobservable(e.g.ability)characteristics.Anydifferencepost-treatmentmustthereforebeaneffectofthetreatment.Evidenceofthistypeisthehighestqualityofevidenceandisconsideredthe‘goldstandard’forpolicyevaluations.

Randomised Control Trial (RCT)AnRCTisdefinedbyrandomassignmenttoeitherthetreatmentorcontrolgroup.AtypicalRCTwillinvolvethefollowingsteps.First,theoriginalprogrammeapplicantsmaybepre-screenedoneligibilityrequirements.Second,alottery(computerrandomisation)assignsapercentageoftheeligibleapplicants(usually50%)tothecontrolgroupandtheremaindertothetreatmentgroup.Third,baselinedataiscollected(eitherfromanexistingdatasourceorfromabespoke‘baseline’survey).Fourth,thetreatmentisapplied.Fifthandfinally,dataiscollectedsometimeafterthetreatment(again,eitherfromanexistingdatasourceorabespoke‘follow–up’survey).Becauseindividualsarerandomlyassignedtothetreatmentandcontrolgroups,thereisnoreasonthattheyshoulddifferoneitherobservableorunobservablecharacteristicsinthebaselinedata.Thismeansthatdifferencesintheoutcomevariablesinthepost-treatmentdatawillbeentirelyattributabletothetreatmenti.e.arefreefromselectionbias.ForthisreasonanRCTscoresthemaximumscoreof5ontheMarylandscale.

InorderforanRCTtoachievethemaximumSMS5scoreonimplementation,threecriteriamustbemet.Firstly,therandomisationmustbesuccessful.Whetherornotthisisthecase,isoftentestedusing“balancingtests”thatcomparetreatedandcontrolindividualsinthebaselinedataonarangeofcharacteristics.Ifrandomisationhasbeensuccessfulthenthebalancingtestsshouldshownosignificantdifferencesbetweenthetwogroups.Secondly,attritionmustbecarefullyaddressed.Attritionhappenswheneverindividualsdropoutfromthestudye.g.donotcompletetreatmentordonotprovidefollow-updata(whenthisiscollectedusingafollow-upsurvey).Theavailabilityoffollow-updataisoftenaparticularproblemforindividualsinthecontrolgroupsincetheyhavelessincentive

03

Page 9: Guide to scoring evidencemethod, scoring them on the quality of their implementation. This scoring guide plays several important roles. Firstly, consistent with our commitment to openness,

Guide to scoring evidence using the Maryland Scientific Method Scale - Updated June 2016 7

tostayinthestudy.Ifdropoutisonarandombasis,thenattritionisnotanissue.However,theremaybeelementsofselectionbiasfordroppingout.Forinstance,itmaybethecasethelessskilledindividualsinacontrolgroupchoosetoleaveanunemploymenttrainingstudy.Thiswouldmaketheremainingcontrolgroupmoreskilledonaverageincreasingthelikelihoodthattheyfindemployment.Inturn,thiswouldleadtoadownwardlybiasedtreatmenteffect.Forthisreason,policyevaluatorsshouldpaycarefulattentiontotheissueofattrition.Ifattritionisliabletoselectionbias,thenevaluatorsshouldsatisfactorilyaddresstheissue.

Forinstance,inastudythatlooksattheimpactofneighbourhoodcrimeonthepropensityofrefugeestocommitcrimethemselves,ifcertainindividualsdropoutofthestudybecausetheymoveabroad,thenthereisaproblemofpotentiallybiasedattrition.Toovercomethis,theauthorscouldlookintowhatfactorsleadrefugeestogoabroad,andincludethesecontrolsintheirstudy(wediscussanexampleofthisbelow).Thirdly,theexperimentshouldbedesignedsuchthatthetreatmentdoesnotspillovertothecontrolgroupi.e.contaminationmustnotbeissue.Inorderforthecontrolgrouptobeasuitablecomparator,itmustnotinanywaybeexposedtothetreatment.Ifthisconditionisnotrespected,thenthetreatmenteffectmaybedownwardlybiasedbecausethecontrolgrouppartlybenefitsfromthetreatment.Contaminationmayhappenif,forexample,individualsinagivenneighbourhoodaregivenfinancialmanagementadvice,andtheysharetheirknowledgewithpeopleinotherneighbourhoodswhoaresupposedtogountreated.

MethodMaximum SMS score (method, implementation)

Adjusted SMS score (method, implementation)

Randomised Control Trial (RCT)

a.k.a.

FieldExperiment

5,5if

• Randomisationissuccessful

• Attritioncarefullyaddressedornotanissue

• Contaminationnotanissue

5,4if

• Oneofthecriteriaisseverelyviolated

• 5,3if

• Twoormoreofthecriteriaareseverelyviolated

RCT Example 1 (SMS 5, 5)

A2013paperbyJensetal.publishedintheAmericanEconomicReviewevaluatestheimpactofneighbourhoodqualityonlifeoutcomes.2Inthiscase,theevaluationisdifficultbecauseitispossiblethattheresidentsofdisadvantagedneighbourhoodshaveworseoutcomesbecauseofindividualorfamilycharacteristics.Toanextent,certaintypesoffamilies“self-select”intobadneighbourhoods,meaningthatthereisaproblemofselectionbias.Inthisinstance,theauthorssurpassthisissuebyevaluatingtheimpactofMovingtoOpportunity(MTO),aprogrammethatgavefamilieswithindisadvantagedneighbourhoodstheopportunitytomovetomoreaffluentareasonarandombasis.InorderforthestudytoachievethemaximumSMSscoreof5,thethreecriteriamustbemet.

1. Randomisation is successful

Pass

TheauthorsmakeaconvincingargumentinfavourofthetrulyrandomnatureoftheMTOprogramme.Inapreliminarystage,interestedresidentsofpoorneighbourhoods(4,604low-incomepublichousingfamilies)enrolledinMTO.Thesefamilieswererandomlyassignedtooneofthreegroups:theExperimentalgroup(whichreceivedhousingvouchersthatsubsidisedprivate-marketrentsinlow-

2 Jens,L.,Duncan,G.,Gennetian,L.,Katz,L.,Kessler,R.,Kling,J.,Sanbonmatsu,L.(2013).Long-TermNeighborhoodEffectsonLow-IncomeFamilies:EvidencefromMovingtoOpportunity.AmericanEconomicReview,226-231.

Page 10: Guide to scoring evidencemethod, scoring them on the quality of their implementation. This scoring guide plays several important roles. Firstly, consistent with our commitment to openness,

Guide to scoring evidence using the Maryland Scientific Method Scale - Updated June 2016 8

povertycommunities),theSection8group(whichreceivedunconstrainedhousingvouchers),andthecontrolgroup.Balancingtestsshowthattreatmentandcontrolgroupsaresimilaraccordingtoarangeofobservablecharacteristics,meaningthattherandomisationwassuccessful.

2. Attrition carefully addressed

Pass

Around90percentoftheindividualsthatwereoriginallyenrolledintheprogramwereinterviewed10to15yearsafterthebaselineyear(from1994to1998).Thisisafairlyhighrate,sothatsampleattritionisnotaprobleminthisinstance.

3. Contamination not an issue

Pass

Thevoucherswereinnowaytransferable,meaningthatcontaminationisnotanissueinthiscase.

Given that all three criteria are satisfied, this study achieves SMS 5 for implementation.

RCT Example 2 (SMS 5, 5)

A2014paperbyDammandDustmannpublishedintheAmericanEconomicReviewevaluatestheimpactofearlyexposuretoneighbourhoodcrimeonsubsequentcriminalbehaviour.3Isolatinghowneighbourhoodsaffectyoungpeople’spropensitytocommitcrimesisdifficult,asitispossiblethatchildrengrowingupincrime-riddencommunitiescomefromfamilyenvironmentsthatarethemselvesconducivetocriminalbehaviour.Toovercomethisproblem,theauthorsexploitaneventinwhichasylum-seekingfamilieswererandomlyassignedtocommunitieswithdifferinglevelsofcrimebytheDanishgovernment.

1. Randomisation is successful

Pass

Uponbeingapprovedforrefugeestatus,asylum-seekerswereassignedtemporaryhousinginoneofDenmark’s15counties.Withinthecounty,thecouncil’slocalofficeassignedeachfamilytoamunicipality.Thisassignmentwasrandom,althoughcouncilswereawareofbirthdates,maritalstatuses,numberofchildren,andnationalities.Theauthorshighlightthefactthatthecouncilinnowaybaseditsdecisiononthefamily’seducationalattainment,criminalrecord,orfamilyincome,simplybecauseitwasnotprivytothisinformation.Italsodidnotknowanythingaboutthefamilybesideswhatwaswritteninthequestionnaire.Furthermore,thefamily’spersonalpreferenceswerenottakenintoaccount.Thebaselinebalancingtestconfirmsthatthetreatmentandcontrolgroupswereindeedobservablysimilarbeforethetreatment.

2. Attrition carefully addressed

Pass

Ofthesampleof5,615refugeechildren,975leftDenmarkbeforetheageof21.Furthermore,anadditional215childrenwerenotobservedineveryyearbetweenarrivalandage21.Nonetheless,theauthorsfindnosignificantrelationshipbetweenleavingDenmarkornotbeingobservedallyearsandtheoutcomevariables.Thisindicatesthatthereisnosignificantselectionbiasinherenttodroppingoutofthestudy.

3 Damm,A.&Dustmann,C.(2014).DoesGrowingUpinaHighCrimeNeighborhoodAffectYouthCriminalBehavior?.AmericanEconomicReview,1806-1832.

Page 11: Guide to scoring evidencemethod, scoring them on the quality of their implementation. This scoring guide plays several important roles. Firstly, consistent with our commitment to openness,

Guide to scoring evidence using the Maryland Scientific Method Scale - Updated June 2016 9

3. Contamination not an issue

Pass

Althoughfamilieswererandomlyassignedtoamunicipalityandwereencouragedtostaythereforatleastthedurationofthe18-monthintroductoryperiod,theywerenostrictrelocationrestrictions.Thismeansthatitispossiblethatfamiliesthatreceivedthetreatment(beinginaneighbourhoodwithmorecrime)didnotnecessarilyadheretothetreatment,andfamiliesthatwerecontrols(livedinsaferneighbourhoods)mayhavebeenexposedtothetreatment.Theauthorsfindthataround80percentofthefamiliesstayedintheirassignedareaafterthefirstyear,andhalfofallfamiliesstayedintheassignedareaaftereightyears.

Given that all three criteria are satisfied, this study achieves SMS 5 for implementation.

RCT Example 3 (SMS 5, 3)

A2012reportbyDoyleevaluatestheimpactofPreparingforLife(PFL),aprogrammethatprovidessupporttofamiliesfrompregnancyuntilthechildisoldenoughtostartschool.4Itislikelythatthereisasignificantselectionintotreatment–moreconcernedparentsmaychoosetojointheprogramme.Therefore,inordertoaccuratelyunderstandtheeffectofthetreatment,theauthoremploysarandomisedcontroltrial.

1. Randomisation is successful

Fail

PFLisspecifictocertaindeprivedcatchmentareasinIreland.Withinthesecatchmentareas,52percentofallpregnantwomenparticipatedintheprogramme,withtheremaining26percentrejectingtheoffer,andanother22percentnothavingbeenidentified.Inasubsequentstage,PFLrecipientswererandomlyassignedtooneoftwogroups:hightreatmentandlowtreatment.Anobservationallysimilarcommunitywaslaterusedasacontrol,whichimpliesthattreatmentwasnot,infact,randomlyassigned.

2. Attrition carefully addressed

Fail

Theauthorsdonotdiscusstheextenttowhichtreatedwomendroppedoutoftheprogramme.

3. Contamination not an issue

Fail

GiventhatalargepartofPFLentailstheprovisionofinformation(i.e.mentoringanddevelopmentinformationpacks)withinacommunity,itispossiblethatneighbourssharethecontentsoftheirtreatment.Thisisparticularlytrueforthehighandlowtreatmentgroups,astheyliveinthesamecommunity.

Given that all three criteria are violated, this study achieves SMS 3 for implementation.

4 Doyle,O.,&UCDGearyInstitute PFL EvaluationTeam.(2012). PreparingforLifeEarlyChildhoodInterventionAssessingtheEarlyImpactofPreparingforLifeatSixMonths.

Page 12: Guide to scoring evidencemethod, scoring them on the quality of their implementation. This scoring guide plays several important roles. Firstly, consistent with our commitment to openness,

Guide to scoring evidence using the Maryland Scientific Method Scale - Updated June 2016 10

SMS 4 methods

SMS4methodsarecharacterisedbytheexploitationofsomesourceof‘quasi-randomness’.Thatis,randomnessthathasnotbeendeliberatelyimposedbutarisesbecauseofsomeotherreason.Usually,thismeansidentifyinghistorical,socialornaturalfactorsthatresultinpolicybeingimplementedinawaythatistosomeextentrandom.Byidentifyingcaseswhereapolicywasimplementedtosomeextentrandomly,theSMS4methodstrytoensurethatthetreatmentandcontrolgroupsaresimilaronobservableandunobservablecharacteristics.However,unlikecontrolledrandomisation,theyneedtomakethecasethattheresultingvariationintreatmentistrulyrandom.Ifthisargumentfailsinpracticethentreatmentandcontrolgroupsdifferandthiscanleadustoincorrectconclusionsabouttreatmenteffects.Itisforthisreasonthatthesemethodsscore4,ratherthanamaximum5,ontheSMSscale.

Instrumental variables (IV)Tosolvetheproblemofselectionbias,apolicyevaluationcanusetheinstrumentalvariables(IV)method.Thisapproachentailsfindingsomethingthatexplainstreatmentbuthasnodirecteffectontheoutcomeofinterest(andisnotrelatedtoanyotherfactorsthatmightdeterminethatoutcome).Thisfactor,the‘instrument’,substitutesforthetreatmentvariablethatmayitselfbecorrelatedwithothercharacteristicsthataffectoutcomes(andthuscauseselectionbias).Forinstance,whenlookingattheimpactofhighwaysonpopulationdensitywithincities,itispossiblethatnotonlymighthighwaysimpactpopulationdensity,butthattheconstructionofhighwaysmightrespondtochangesinpopulationdensity.Agoodwaytocircumventthisproblem,then,istoemployanIV.Forexample(discussedfurtherbelow)Baum-Snow(2007)usesaplannedUShighwaygridthatwasplannedin1947(butnotnecessarilybuilt)asaninstrumentforactualUShighways.Thelogicbehindthisinstrumentisthatitisessentiallyrandominitsassignment(asfaraspopulationdensitygoes),becausethe1947gridwasplannedformilitaryandtradepurposes.ThiselementofrandomnessisessentialforasuccessfulIV.Inthissense,itisawayofsimulatingtherandomassignmenttotreatmentthatisdonewithrandomisedcontroltrials.Asuccessfulinstrumenthasthepotentialtoeliminatedifferencesbetweentreatmentandcontrolgrouponobservableorunobservablecharacteristicsandthusscoreamaximumof4ontheSMSscale.Inordertoachievethisscore,the

04

Page 13: Guide to scoring evidencemethod, scoring them on the quality of their implementation. This scoring guide plays several important roles. Firstly, consistent with our commitment to openness,

Guide to scoring evidence using the Maryland Scientific Method Scale - Updated June 2016 11

instrumentmustsatisfythreemaincriteria.Firstly,theinstrumentmustbeexogenous,orrandominassignment.Secondly,itmustberelevant,orcrediblyrelatedtothevariablethatitreplaces.Thirdly,itmustbeexcludable,meaningthatitdoesnotdirectlyimpacttheoutcome.

MethodMaximum SMS score (method, implementation)

Adjusted SMS score (method, implementation)

Instrumental Variable (IV)

a.k.a.

Two-StageLeastSquares(2SLS)

4,4ifinstrumentis:

• Relevant(explainstreatment)

• Exogenous(notexplainedbyoutcome)

• Excludable(doesnotdirectlyaffectoutcome)

Scoredasperunderlyingmethodif:

• Instrumentisinvalid

• e.g.crosssectionwithinvalidIVscores2;difference-in-differencewithinvalidIVscores3(seebelow)

IV Example 1 (SMS 4, 4)

A2007paperbyNathanielBaum-SnowpublishedintheQuarterlyJournalofEconomicsevaluatestheimpactofhighwaysonsuburbanisation.5Thisisadifficultquestiontoevaluatebecauseofthepotentialpresenceofselectionbias:highwaysthatwerebuiltmayhavebeenconstructedtoaccommodatesuburbanisation.Inordertoovercomethisproblem,theauthorusestheIVmethod,withhighwaysthatwereoriginallyplanned(butnotbuilt)in1947asaninstrumentforactualhighways.InordertoachievethemaximumSMSscoreof4,theinstrumentmustsatisfythethreecriteria.

1. Instrument is relevant

Pass

Thereisnodoubtthattheoriginal1947planisarelevantinstrumentbecausethecurrenthighwaysystemispartlybasedonit.

2. Instrument is exogenous

Pass

The1947planisanexogenousinstrumentbecauseitwasnotplannedwithsuburbanisationinmindbuttoconnectcitiesfortradepurposes.Onethreattoexogeneityisthatitmayhavebeenthecasethatlargercitiesweregivendenserhighwaysystemsinthe1947plan.Thereforetheauthorcontrolsforthe1947populationofeachcity.Conditionalonthisimportantcontrol,theinstrumentisconvincinglyexogenous.

3. Instrument is excludable

Pass

Byvirtueofthe1947gridbeingsimplyaplan,thereisnowayitcouldhavedirectlyaffectedsuburbanization,hencetheexclusionrestrictionholds.

Given that all three criteria are satisfied, this study achieves SMS 4 for implementation.

5 Baum-Snow,N.(2007).DidHighwaysCauseSuburbanization?.QuarterlyJournalofEconomics,775-805.

Page 14: Guide to scoring evidencemethod, scoring them on the quality of their implementation. This scoring guide plays several important roles. Firstly, consistent with our commitment to openness,

Guide to scoring evidence using the Maryland Scientific Method Scale - Updated June 2016 12

IV Example 2 (SMS 4, 4)

A2013paperbyCollinsandShesterevaluatestheimpactofurbanrenewalprogrammesoneconomicoutcomesinvariousU.S.cities.6Theseprogrammesareparticularlyproblematicintermsofevaluationbecauseitmaybethatpoorercitiesself-selectintotreatmentbecausetheyaremoreinneedofurbanrenewalschemes,orconversely,thatrichercitiesself-selectintotreatmentbecausetheycanaffordtoundertakemoreurbanrenewal.Toovercometheseissues,theauthorsusevariationinthetimingofwhenstatesapprovedtheurbanrenewallawsasaninstrument.

1. Instrument is relevant

Pass

Duringtheroll-outofthepolicy,theinstrumentofstate-leveldelaysstronglypredictedwhetherornoturbanrenewalprogrammeswereinplace.

2. Instrument is exogenous

Pass

Theauthorsarguethatthestates’decisiontoallowurbanrenewallawsissomewhatrandominitsprecisetiming.Theyconcede,however,thatstates’receptivenesstofederalinterventionmaybeanindicatorofliberalness,whichmayinturnbepositivelyrelatedtoeconomicprogress.Thus,controlsforstateconservatismareincluded.Conditionalonthiscontrol,theargumentappearssound.

3. Instrument is excludable

Pass

Theyarguethatiftimingofstateapprovalofurbanrenewallawsdirectlyaffectedurbaneconomicoutcomes,thenitwouldalsoaffecttheeconomyoftherestofthestate.Findingthattheruralareasofstatesthatpassedthelawslaterwerenoeconomicallydifferentfromtheruralareasofstatesthatpassedthelawsearlier,theauthorsconvincinglydemonstratethattheexclusionrestrictionholds.

Given that all three criteria are satisfied, this study achieves SMS 4 for implementation.

IV Example 3 (SMS 4, 2)

A2011paperbyNishimuraandOkamuroevaluatestheimpactofindustrialclustersonthenumberofpatentsafirmappliesfor.Thisevaluationisproblematicbecausefirmsthatchoosetolocatewithinanindustrialclusterareprobablyinherentlydifferentfromthosethatdonot.Toovercomethis,theauthorsusefirmageasaninstrumentforclusterparticipation.Thelogicbehindthisinstrumentisthatsmallerfirmsareattractedtoindustrialclusters.Firmsizeis,inturn,relatedtofirmage.Furthermore,theauthorsarguethattheageofafirmissomewhatrandomgiventhatthenumberofpatentsthatafirmappliesfordoesnotaffectitsage.

1. Instrument is relevant

Pass

Theauthorsarguethatonlysmallandmediumfirmsgravitatetowardsclustersbecauselarge,internationalfirms,arelessinterestedinregionalcollaborations.Theydefendthatthesizeofafirmis,inturn,significantlyrelatedtoitsage.Thisappearsareasonableargument.

6 Collins,W.&Shester,K.(2013).SlumClearanceandUrbanRenewalintheUnitedStates.AmericanEconomicJournal:AppliedEconomics,239-273.

Page 15: Guide to scoring evidencemethod, scoring them on the quality of their implementation. This scoring guide plays several important roles. Firstly, consistent with our commitment to openness,

Guide to scoring evidence using the Maryland Scientific Method Scale - Updated June 2016 13

2. Instrument is exogenous

Fail

Thisinstrumentisnotexogenousbecausefirmageisnotrandomlyassigned;onlyrelativelysuccessfulfirmssurvivebeyondacertainage.

3. Instrument is excludable

Fail

Theauthorsclaimthatfirmagedoesnotdirectlyinfluencethenumberofpatentsafirmappliesfor.However,youngfirms(e.g.start-ups)maybemoreinnovativeandapplyformorepatents.Conversely,oldfirmsaretypicallylargerandmayhaveagreaterbudgetforR&Dthatcouldleadtopatents.Ifeitheristruethentheexclusionrestrictionisviolated.Theauthorsdonotaddresstheseconcerns.

Given that only one of the three criteria is satisfied, this study is scored according to its underlying method (cross sectional with controls) and achieves SMS 2 for implementation.

Regression Discontinuity Design (RDD)Thismethodcanbeappliedincaseswheretherearecut-offsfortreatmenteligibility.Thesediscontinuitiesintreatment(wherebyunitsaretreatedaslongastheyareabove/belowacertainobservablethreshold)canbeexploitedtogeneratequasi-randomassignmentintotreatment.Essentially,thisinvolvescomparingunitswhoarejustabovethethreshold(andhencetreated)tothosethatarejustbelowthethreshold(andhenceuntreated).Whilstingeneraltheunitsthataretreatedaredifferenttothosethatarenot,theunitsthatarejusteithersideofthecut-offarelikelytobesimilar(bothintermsofobservableandunobservablecharacteristics).Thismakestreatmentaroundthecut-offalmostrandom.Giventhatthistypeofstudyexploitsacertainrandomnessintreatment(aswiththeIVmethod),itensuresthatthetreatmentgroupandcontrolgrouparesimilaronobservableandunobservablecharacteristics,thereforewarrantingamaximumofSMS4.

MethodMaximum SMS score (method, implementation)

Adjusted SMS score (method, implementation)

Regression Discontinuity Design (RDD)

4,4if:

• Discontinuityintreatmentissharp(e.g.stricteligibilityrequirement)orfuzzydiscontinuitymethodused

• Onlytreatmentchangesatboundary

• Behaviourisnotmanipulatedtomakethecut-off

Scoredasperunderlyingmethodif

• Discontinuityconditionsseverelyviolated

• e.gcrosssectionwithinvalidIVscores2;difference-in-differencewithinvalidIVscores3(seebelow)

However,inorderforthestudytoachievethismaximumscore,threeassumptionsmustbesatisfied.Firstly,thediscontinuityintreatmentmustbesufficientlysharp.Thismeansthatthethresholdfortreatmentmustbeclearandenforcedinpractice,sothatwhenindividualsaboveandbelowthethresholdarecomparedtheresearcherisinfactcomparingtreatedtountreatedindividuals.Secondly,onlythetreatmentshouldchangeattheboundary.Thisisimportanttoensurecomparabilitybetweentreatmentandcontrolgroups.Thisassumptionwouldbeviolatedif,forexample,firmsthathadlessthantenemployeeshadtopaylowerpayrolltaxesinadditiontobeingeligibleforthegrant.Finally,behaviourmustn’tbemanipulatedaroundthecut-off,soastoguaranteethatindividualsaround

Page 16: Guide to scoring evidencemethod, scoring them on the quality of their implementation. This scoring guide plays several important roles. Firstly, consistent with our commitment to openness,

Guide to scoring evidence using the Maryland Scientific Method Scale - Updated June 2016 14

thethresholdareinfactcomparableineverywayexceptfortheirexposuretotreatment.Thisisparticularlyrelevantifthethresholditselfinducesunitstochangetheirbehaviour–forexample,firmscouldartificiallyensurethattheyhavelessthantenemployeeseventhoughtheywouldotherwisehiremore,justbecausetheywanttobeeligibleforthegrant.Ifthishappens,thenitcannotbecrediblyheldthattreatmentisrandomaroundtheboundary.

RDD Example 1 (SMS 4, 4)

A2013paperbyPop-ElechesandUrquiolapublishedintheAmericanEconomicReviewevaluatestheimpactofschoolqualityoneconomicoutcomes.7Thisisaparticularlydifficultquestiontoaddress,aspupilsofbetterschoolsareacademicallymorecapableinthefirstplace,andthereforemorelikelytogetbetterresultsregardlessofschoolquality.Toovercomethisselectionbias,theauthorsemploytheregressiondiscontinuitymethod.Theyexploitadiscontinuityintreatment,wherebystudentsthatachieveaboveacertaingradepointaverageareadmittedintoahigherqualityhighschool,whilethosethatachievejustbelowtherequiredgradeaverageattendalowerqualityhighschool.InorderforthestudytoachievethemaximumSMSscoreof4,thethreecriteriamustbemet.

1. Discontinuity in treatment is sufficiently sharp

Pass

InRomania,highschooladmissionsaredeterminedsolelybyschoolaveragesandexaminationresults.Thestandardsapplytoallstudentsinthesamemanner.Therefore,itcanbearguedwithcertaintythatthediscontinuityissharp,inthatthosethatscorebelowthethresholddefinitelydonotgetaplaceinthebestschool,andthosethatscoreabovedefinitelydo.

2. Only treatment changes at the boundary

Pass

Becausethisstudylooksatschoolquality,itcanbeplausiblyheldthattheonlythingthatchangesattheboundaryisthequalityofschool,sincestudentsthatsurpassthecut-offdonotgetfurtherbenefits(suchasmeritawardsforexample).

3. Behaviour is not manipulated to make the cut-off

Pass

Inorderforthisconditiontohold,itcannotbe,forinstance,thatstudentsjustabovethethresholdimploredtheirteachersforslightlybettergrades,asthatwouldmaketheminherentlydifferenttothosethatdidnot.Forthiscase,itisunlikelythatthishappenedgiventhatthethresholdisonlyknownonceexamresultsareout.

Given that all three criteria are satisfied, this study achieve SMS 4 for implementation.

RDD Example 2 (SMS 4, 4)

A2014studybyAndersonpublishedintheAmericanEconomicReviewevaluatestheimpactofpublictransporttransitonhighwaycongestion.8Inthiscase,itisdifficulttoisolatethedirectionofcausality,asthenumberofpeoplewhousepublictransportcanaffecthighwaycongestion,butat

7 Pop-Eleches,C.&Urquiola,M.(2013).GoingtoaBetterSchool:EffectsandBehavioralResponses.AmericanEconomicReview,1289-1384.

8 Anderson,M.(2014).Subways,Strikes,andSlowdowns:TheImpactsofPublicTransitonTrafficCongestion.AmericanEconomicReview,2763-2796.

Page 17: Guide to scoring evidencemethod, scoring them on the quality of their implementation. This scoring guide plays several important roles. Firstly, consistent with our commitment to openness,

Guide to scoring evidence using the Maryland Scientific Method Scale - Updated June 2016 15

thesametime,highwaycongestioncanalsoaffectthenumberofpeoplethatusepublictransport.Toovercomethisissue,theauthorsexploitanaturalexperimentwherebyLosAngelespublictransportworkerssuddenlywentonstrikefor35days.Morespecifically,theyusetheRDDmethodtocomparehighwaycongestionjustbeforethestrikestartedandhighwaycongestionjustafterthestrikestarted.

1. Discontinuity in treatment is sufficiently sharp

Pass

Here,thediscontinuityisverysharp.Fromonedaytoanother,transitworkerswentonstrike,meaningthattheyworkednormallyoneday,anddidnotworkatallthenextday,andthedaysthereafter.Thisledtothealmostcompleteshutdownofpublictransportservices,withminorexceptionsofsomecontract-operatedbusservices.Theauthorsarguethattheseemergencyserviceswereonlyusefulforaverysmallfractionofpublictransportusers.

2. Only treatment changes at boundary

Pass

Thestrikewasspurredbyadisagreementbetweenthetransportmechanicsandtheemployersovercontributionstoahealthcarefund.Theauthorsarguethatthetimingofthestrikeisexogenousbecauseitstartedjustaftertheexpirationofa60-daycourt-orderedinjunctionthatprohibitedstriking.

3. Behaviour is not manipulated to make the cut-off

Pass

Itispossiblethatifpeoplewereexpectingthestrike,theychangedtheirworkarrangementstoadjusttotheinterruptions.Forinstance,itcouldbethatpeoplestartedworkingfromhomewhentheyotherwisewouldn’thave.Ifthiswerethecase,theeffectsofpublictransportonhighwaycongestionwouldbedownwardlybiased.However,theauthorsarguethatthestrikeledtoanabruptandunexpectedhaltinservice,meaningthatitisunlikelythatpeopleanticipatedit.

Giventhatallthreecriteriaaresatisfied,thisstudyachievesSMS4forimplementation.

RDD Example 3 (SMS 4, 2)

A2007reportbyHaeglandandMoenevaluatestheimpactofanR&DtaxcreditonactualfirmR&Dinvestment.9Inprinciple,thisisatoughpolicytoevaluateasfirmsoptintoreceivingthetaxcredit.Toovercomethisproblem,theauthorsusetheRDDmethodandexploitthefactthatfirmsthatinvestedupto4millionNOK(about£342,000)receivedan18percenttaxcreditoneveryNOKspent,whilefirmsthatinvestedabovethisamountreceivedthesamecredituptothe4millionthreshold,butnocreditthereafter.

1. Discontinuity in treatment is sufficiently sharp

Fail

Thediscontinuityseemsextremelyfuzzy,asfirmsabovethethresholdstillgetthetreatment,buttoalesserextent.Thisisbecausefirmsthatspendmorethanthethresholdstillgetataxcreditforthefirst4millionNOKspent,butdonotreceiveacreditforanyamountinvestedbeyondthat.Forexample,afirmthatinvests4million(belowthreshold)getsan18percenttaxcredit(i.e.720,000,about

9 Haegland,T.,&Moen,J.(2007).InputAdditionalityintheNorwegianR&DTaxCreditScheme.DiscussionPaper,StatisticsNorway.

Page 18: Guide to scoring evidencemethod, scoring them on the quality of their implementation. This scoring guide plays several important roles. Firstly, consistent with our commitment to openness,

Guide to scoring evidence using the Maryland Scientific Method Scale - Updated June 2016 16

£61,500)butafirmthatinvests4.1million(abovethreshold)alsogets720,000whichis17.5percent.Thusthereisvirtuallynodifferenceintreatmentoneithersideoftheboundary.

2. Only treatment changes at boundary

Pass

Theredoesnotseemtoanysortofotherelementofvariationatthe4millionNOKcut-off.

3. Behaviour is not manipulated to make the cut-off

Fail

Inthiscase,firmshaveaclearincentivetospendlessthan4millionNOKinR&D,aseveryNOKspentbeyondthatthresholdisnotsubsidised.Therefore,itispossiblethatonlymuchlargerandmoresuccessfulfirmsspendbeyondthethreshold,whilesmaller,lesssuccessfulfirmsdeliberatelyspendlessthantheyotherwisewouldhaveintheabsenceofthe4millioncut-off.

Given that only one of the criteria is satisfied, this study achieves SMS 2 for implementation.

Page 19: Guide to scoring evidencemethod, scoring them on the quality of their implementation. This scoring guide plays several important roles. Firstly, consistent with our commitment to openness,

Guide to scoring evidence using the Maryland Scientific Method Scale - Updated June 2016 17

SMS 3/4 methods

Thereareanumberofmethodswhichmayscorea4ora3dependingontheparticularwayinwhichtheyarespecified.ThisisaparticularissueforhazardregressionsandHeckmanselection(andother‘controlfunction’-type).Becausetheseapproachesarenotfrequentlyencounteredinexistingstudiesoflocaleconomicgrowth,andbecauseaspectsoftheseapproachesareparticularlytechnical,we’verelegatedconsiderationofthemtoAppendix2.Theseapproachesmayseeincreasedapplicationinfuture,butinthemaintextwefocusonthemorefrequentlyencounteredapproaches.

05

Page 20: Guide to scoring evidencemethod, scoring them on the quality of their implementation. This scoring guide plays several important roles. Firstly, consistent with our commitment to openness,

Guide to scoring evidence using the Maryland Scientific Method Scale - Updated June 2016 18

SMS 3 methods

WedefineSMS3methodstobetheminimumstandardforourreviews.Thesemethodsmustinvolveacomparisonoftreatedunitsagainstacontrolgroupbefore-and-afterthetreatmentdate.Thesemethodscontrolforselectiononobservablecharacteristics,andthroughabeforeandaftercomparison,eliminateanyfixedunobservabledifferencebetweentreatmentandcontrolgroup.However,theydonotinherentlycontrolforunobservabledifferencesthatdovarywithtime.Forthisreason,theyscoreamaximumof3ontheSMS.Inordertoachievesuchascore,however,theymustattempttosomehowcontrolfortime-varyingunobservableselectionbias.Thisisoftendonethroughtheconstructionofavalidcontrolgroup.Furthermore,theyshouldcontrolforobservabledifferencesbetweenthetreatmentandcontrolgroupsusingsomeformofregression,matching,controlforselection,orcontrolvariables.

Difference-in-differences (DiD)TheDIDmethodentailscomparingatreatmentandcontrolgroupbeforeandaftertreatment.Morespecifically,thetreatmenteffectiscalculatedbyfirstevaluatingthechangeintheoutcomevariableforthetreatedgroup,andthensubtractingthechangeininthecontrolgroupoverthesameperiod.Here,thecontrolgroupprovidesthecounterfactualgrowthpathi.e.whatwouldhavehappenedinthetreatmentgrouphaditnotbeentreated.Thisismuchbetterthanasimplebeforeandaftertreatmentcomparison,becauseitaccountsforthefactthatchangesinoutcomecanbeduetomanydifferentfactorsandnotjustthetreatmenteffect.Moreover,becauseDIDsubtractsthedifferencesbetweentreatmentandcontrolbothbeforeandafterthetreatment,iteffectivelycontrolsforanyunobserved,time-invariantdifferencesbetweenthetwogroups.However,itdoesnotaccountforunobservabledifferencesbetweenthetwothatvarywithtime.Forthisreason,itachievesamaximumofSMS3.Inordertoachievethisscore,twomaincriteriamustbesatisfied.Firstly,itmustbecrediblyarguedthatthetreatmentgroupwouldhavefollowedthesametrendasthecontrolgroup.Secondly,theremustalsobeaknowntimeperiodfortreatmentsothatthegroupscanbecomparedbeforeandafterthetreatment.

06

Page 21: Guide to scoring evidencemethod, scoring them on the quality of their implementation. This scoring guide plays several important roles. Firstly, consistent with our commitment to openness,

Guide to scoring evidence using the Maryland Scientific Method Scale - Updated June 2016 19

MethodMaximum SMS score (method, implementation)

Adjusted SMS score (method, implementation)

Difference in differences (DID)

a.k.a.

Diffindiff

3,3if

• Controlgroupwouldhavefollowedsametrendandtreatmentgroup

• Knowntimeperiodfortreatment

3,2if

• Eitherofthecriteriaisnotsatisfied

DiD Example 1 (SMS 3, 3)

A2011studybyCurrie,Greenstone,andMorettievaluatestheimpactofhazardouswastesiteclean-upsonthebirthoutcomesofmotherslivingwithin2,000metresofthesite.10Theyusebirthstomotherswholivebetween2,000and5,000metresawayfromthesiteasacontrolgroup.InorderforthestudytoachievethemaximumSMSscoreof3,itmustsatisfythetwocriteria.

1. Treatment group would have followed same trend as control group

Pass

Firstly,thecontrolgroupischosenbecauseitissufficientlyfarawaysothatitisunaffectedbytreatment,becausehazardouswastesitesimpactpeoplethroughdirectcontact,fumes,andwatersupply,andonlythoselivingcloseenoughtoit(i.e.within2,000metres)areaffected.Secondly,thecontrolgroupwaschosenbecauseitissimilartothetreatmentgroup,asthepeoplelivingwithin2,000metresofawastesitearecomparabletothoselivingslightlyfartheraway.Evenso,theauthorsrecognisethatmotherslivingextremelyclosetothesitemaybeslightlydifferenttomotherslivingabitfartheraway.Forthisreason,theycontrolformaternaleducation,race,ethnicity,birthorder,andsmoking,aswellasotherneighbourhoodcharacteristics.

2. Known time period for treatment

Pass

Becausetheauthorsusedatafrom154sitesthatwerecleanedupbetween1989and2003,itishardtopinpointprecisedatesforeachclean-up.However,theauthorsusedatafromtheEnvironmentalProtectionAgency,whichincludesthedatewhenthesitewasaddedtotheNationalPriorityList,whentheclean-upwasinitiated,andwhenitwascompleted.

Given that the two criteria are satisfied, this study achieves SMS 3 for implementation.

DiD Example 2 (SMS 3, 3)

A1993studybyCardandKruegerevaluatestheimpactofaNewJerseyminimumwageincreaseontheemploymentlevelsoffastfoodrestaurants.11Itishardtoaccuratelygaugetheeffectsofsuchanincrease,asitislikelythatchangesinemploymentarecausedbyotherfactors.Toovercomethisproblem,theauthorsemployadifference-in-differencesapproachwherebytheycomparetheemploymentinNewJerseywiththeemploymentinPennsylvania,aneighbouringstatethatdidnotincreaseitsminimumwage.

1. Treatment group would have followed same trend as control group

10 Currie,J.,Greenstone,M.,Moretti,E.(2011).SuperfundCleanupsandInfantHealth.MITCEEPRWorkingPapers.11 Card,D.,Krueger,A.(1993).MinimumWagesandEmployment:ACaseStudyoftheFast-FoodIndustryinNewJersey

andPennsylvania.AmericanEconomicReview,Vol.84,No.4.,pp.772-793

Page 22: Guide to scoring evidencemethod, scoring them on the quality of their implementation. This scoring guide plays several important roles. Firstly, consistent with our commitment to openness,

Guide to scoring evidence using the Maryland Scientific Method Scale - Updated June 2016 20

Pass

TheauthorsmakeaconvincingargumentthatNewJerseyandPennsylvaniaare,infact,comparablestates.Theyanalysetherestaurantcompositionofeachandfindthattherearenosignificantdifferencesbetweenthetwo.Furthermore,becausetheyareneighbouringstates,itcanbecrediblyheldthattheyareexposedtothesamemacroeconomicshocks.

2. Known time period for treatment

Pass

Theminimumwageincreasedfrom$4.25to$5.05perhouronthe1stofApril1992.Theauthorsshowthatthenewminimumwageslawwas,infact,bindinginNewJersey,whichmeansthatwecanreasonablyassertthatthetreatmentwasalsosignificant.

Given that the two criteria are satisfied, this study achieves SMS 3 for implementation.

DiD Example 3 (SMS 3, 2)

A2005studybyValentinandLundJensenevaluatestheimpactofalawthattransferredpatentrightsfromindividualresearcherstouniversitiesoncollaborativedrugdiscoveryresearchinDenmark.12Becausethechangeinresearchfollowingthelawcouldbeattributedtomanydifferentcauses,theauthorsemployaDiDapproachusingSwedenasthecontrolgroup.

1. Treatment group would have followed same trend as control group

Fail

TheauthorsarguethattheDanishandSwedishbiotechindustriesaresimilarinsize,history,responsetothe2001high-techbubble,numberofinventions,andamountofinventorsmobilisedtobringinventionsabout.TheyarguethattheonlydifferencebetweenthemisthatDanishscientiststendtoworkinfirms,whileSwedishscientiststendtobeacademics.Tocontrolforthis,theyemployavariablefortheratioofbiopharmaceuticaltosmallmoleculepatents(asbiopharmaceuticalresearchtendstobedoneinuniversities,whilesmallmoleculeresearchtendstobedoneinfirms).However,theauthorsdonotadequatelyaddressthefactthatSwedenandDenmarkaresubjectedtodifferentmacroeconomicclimates,andhavedifferentsetsoffederallaws.Giventhattheydidnotuseanycontrolstotackletheseissues,itcannotbecrediblyheldthattheirtreatmentandcontrolgroupschangedinsimilarways,meaningthattheyarenotdirectlycomparable.

2. Known time period for treatment

Pass

ThelawthattransferredtheownershipofpatentsfromindividualresearcherstouniversitieswasenactedinDenmarkin2000.

Given that only one of the criteria is satisfied, this study achieves SMS 2 for implementation.

Panel data methods Thesemethodsusedatathatfollowsthesameindividualsovertimeallowingtheresearchertocontrolforthingsthatremainconstantforeachindividualacrosstime.Thiscanbedoneinavarietyofways.Themostcommonisthefixedeffects(FE)approachwhichentailsincludingdummyvariables(the

12 Valentin,F.&Jensen,R.(2005).EffectsonAcademia-IndustryCollaborationofExtendingUniversityPropertyRights.WorkingPaper.

Page 23: Guide to scoring evidencemethod, scoring them on the quality of their implementation. This scoring guide plays several important roles. Firstly, consistent with our commitment to openness,

Guide to scoring evidence using the Maryland Scientific Method Scale - Updated June 2016 21

“fixedeffects”)forbothindividualcharacteristicsandanythingthathappenedinagiventimeperiod(timedummies).Alternatively,thefirstdifferences(FD)approachcanbeused,wherebytheoutcomeofapreviousperiodissubtractedinthefinalregression.Thisessentiallymeansthateverythingthathappenedinthepreviousperiodisremovedfromthefinalregression,thereforecancellingoutanythingthatremainsconstantacrosstime.13Inthiscase,becauseeverythingthatremainsconstantacrosstimeisdifferencedout,itisunnecessarytoincludefixedeffects.

Paneldatamethodssuccessfullycontrolforobservableandunobservablecharacteristicsthatremainconstantthroughouttime.However,theydonotaccountforunobservablecharacteristicsthatchangewithtime.Forthisreason,theycanachieveamaximumSMS3.Inordertoachievethemaximumscore,twomaincriteriamustbemet.Firstly,yearlyfixedeffectsmustbeaccountedfor,sothatthetreatmentisnottaintedbythecontextofagiventimeperiod.Secondly,whenusingthefixedeffectsmethod,thefixedeffectsmustbeattheunitofanalysis.Thismeansthatifthesampleisofpeople,thenthefixedeffectsmustrefertothem.Lastly,althoughpaneldatamethodscontrolforcharacteristicsthatdonotchangewithtime,theydonotcontrolforcharacteristicsthatdonotchangewithtime.Forthisreason,itisimportanttoincludeothertime-varyingcontrolsthatmayalsoaffecttheoutcome.

MethodMaximum SMS score (method, implementation)

Adjusted SMS score (method, implementation)

Panel methods

Panel Fixed Effects (FE)

3,3if

• Fixedeffectisattheunitofobservation

• Yeareffectsareincluded

• Appropriatetime-varyingcontrolsareused

3,2if

• Oneormoreofthethreecriteriaisnotsatisfied

First Differences (FD)

3,3if

• Yeareffectsareincluded

• Appropriatetime-varyingcontrolsareused

3,2if

• Eitherofthetwocriteriaisnotsatisfied

FE Example 1 (SMS 3, 3)

A2011studybyDinkelmanpublishedintheAmericanEconomicReviewevaluatestheimpactofelectricityroll-outontheemploymentoutcomesofruralcommunities.14Inthiscase,theevaluationmaybetaintedbyselectionbias,asitispossiblethatpoliticallyimportantorgrowingareasarefavouredoverothers(thereversecouldalsobetrue).Toovercomethisissue,theauthoremploysafixedeffectsstrategy.InorderforthisstudytoachievethemaximumSMSscoreof3,itmustsatisfythethreecriteria.

1. Fixed effect is at the unit of analysis

Pass

Theauthorincludesavariableforcommunityfixedeffects.Giventhattheunitofobservationisthecommunity,thiscriterionismet.

13 SometimesresearchersthinkthatoutcomesinpreviousperiodsmaydirectlyaffectoutcomestodaywhichcausecomplicationsinpaneldatasettingsthataresometimesaddressedusingtheArellano-Bondmethod(seeAppendix3).

14 Dinkelman,T.(2011).TheEffectsofRuralElectrificationonEmployment:NewEvidencefromSouthAfrica.AmericanEconomicReview,3078-3108.

Page 24: Guide to scoring evidencemethod, scoring them on the quality of their implementation. This scoring guide plays several important roles. Firstly, consistent with our commitment to openness,

Guide to scoring evidence using the Maryland Scientific Method Scale - Updated June 2016 22

2. Year effects are included

Pass

Controlsfortimetrendswithinthedistrictareincluded.

3. Appropriate time-varying controls used

Pass

Inthisparticularexample,itislikelythattherearemanykeyvariablesthatvaryovertime,andarethereforenotcapturedbythefixedeffects.Forthisreason,theauthorincludesavectorofcommunitycovariateswhichincludehouseholddensity,fractionofhouseholdslivingbelowthepovertyline,fractionofwhiteadults,educationalattainmentofadults,shareoffemale-headedhouseholds,andfemaletomaleratio.

Given that all three criteria are satisfied, this study achieves SMS 3 for implementation.

FE Example 2 (SMS 3, 2)

A2013studybyAtasoyanalysestheimpactofbroadbandinternetexpansiononemployment.15Thisevaluationposesthesameissuesofselectivityastheelectricityroll-outexample.Toovercomethis,theauthoremploysthefixedeffectsstrategy.

1. Fixed effect is at the unit of analysis

Pass

Theauthorderivescounty-levelbroadbandadoptionratesusingzip-codedata.Hesubsequentlyusescountyfixedeffects,whicharetherelevantunitofobservation.

2. Year effects are included

Pass

Yeardummyvariablesareincluded.

3. Appropriate time-varying controls used

Fail

Theauthorincludescontrolsfortime-variantpopulationcharacteristicssuchasincome,age,gender,race,anddensity.However,heneglectscertaincounty-levelcharacteristicsthataffectemployment,andarenotcapturedbythecountyfixedeffectsbecausetheychangewithtime.Thesecontrolscanincludetransportationinfrastructure,investmentineducationandjobtraining,andleveloffundinggrantedbythestate.

Given that only two of the criteria are satisfied, this study achieves SMS 2 for implementation.

15 Atasoy,H.(2013).TheEffectsofBroadbandInternetExpansiononLabourMarketOutcomes.ILRReview.

Page 25: Guide to scoring evidencemethod, scoring them on the quality of their implementation. This scoring guide plays several important roles. Firstly, consistent with our commitment to openness,

Guide to scoring evidence using the Maryland Scientific Method Scale - Updated June 2016 23

FE Example 3 (SMS 3, 2)

A1997studybyGuellecanddelaPotterieevaluatestheimpactofgovernmentsubsidiesonR&Dexpenditure.16Inthiscasethereisaclearpotentialproblemofselectionbias,asfirmsthatchoosetousethesubsidiesaredifferentfromfirmsthatdonot.Toovercomethis,theyusethefirstdifferencesmethod.InorderforthisstudytoachievethemaximumSMSscoreof3,thetwocriteriamustbesatisfied.

1. Year effects are included

Pass

Theauthorsincludeayeardummyvariableinthefinalregression.

2. Appropriate time-varying controls used

Fail

Theauthorsdonotincludeanytime-varyingcontrols.Thismeansthatwhatisthoughttobetheimpactofthegovernmentsubsidiesmayinfactbeduetothingslikegrowthinthenumberoffirmemployees.

Given that only one of the criteria is satisfied, this study achieves SMS 2 for implementation.

Box 2: Fixed effects in the cross-sectional data context

Sometimes,cross-sectionalstudieswillusetheterm“fixedeffects”.However,itisimportanttodistinguishfixedeffectsinthiscontext,andfixedeffectsinthepaneldatacontext.Whilepaneldataconsidersanumberofindividualsacrosstime,cross-sectionaldataonlyconsidersindividualsinacertaintimeperiod.Forthisreason,panelfixedeffectsrefertowhatremainsconstantacrosstimeforaparticularindividual,andcross-sectionalfixedeffectssimplyrefertocontrols.Forinstance,ina2014studybyGibbonsetal,theauthorslookattheimpactofnaturalamenitiesonpropertyprices.Althoughtheirdataiscross-sectional,theyinclude“TravelToWorkArea(TTWA)levelfixedeffects”,whichinthiscasemeanscontrolsfortheTTWAthepropertyislocatedin.

Itisworthnotingthatevencross-sectionalstudiesthatincluderelevantcontrolscanonlygetamaximumSMSof2.

Propensity Score Matching (PSM)Thismethodentailsmatchingeverytreatedunitwithanobservationallysimilaruntreatedunit,andthenlookingatthedifferencesbetweenthemtogaugethetreatmenteffect.Thetreatedanduntreatedunitsarenotmatchedaccordingtowhethertheyhavesimilarcharacteristicsperse,butaccordingtoapropensityscore.Thisscoreisessentiallytheprobabilitythataunitistreated,accordingtoasetofcharacteristics.Thepropensityscoreistypicallycalculatedusingaregressionthatdealswitha0to1probabilityastheoutcomevariable(e.g.alogisticorprobitmodel).However,evenso,thismethoddoesnotentirelycorrectforselectionbiasbecauseitonlyaccountsforobservablecharacteristics.Forthisreason,ifusedinapurelycross-sectionalcontext,itcanachieveamaximumSMS2.Inordertoachievethisscore,itmustbecrediblyheldthatthematchingcriteriaarerelevanttoselectionintotreatment.Furthermore,akeyissuewithPSMisthatunitsthatareactuallytreatedaremorelikelytohavehigherpropensityscores,whileunitsthatareactuallyuntreatedaremorelikelytohavelower

16 Guellec,D.&delaPotterie,B.(1997)DoesgovernmentsupportstimulateprivateR&D?OECDEconomicStudies

Page 26: Guide to scoring evidencemethod, scoring them on the quality of their implementation. This scoring guide plays several important roles. Firstly, consistent with our commitment to openness,

Guide to scoring evidence using the Maryland Scientific Method Scale - Updated June 2016 24

propensityscores.Therefore,havingasufficientlylargeareaofoverlap(wherebytreatedunitsarematchedwithsimilaruntreatedunits)isessentialforthemethodtowork.InorderforthemethodtobeofSMS3quality,itmustbeusedwithothermethodsthatattempttocontrolforunobservablecharacteristics.Forinstance,itcanbecombinedwithmethodsthatexploittimevariationintreatment(e.g.DID).Inthiscase,inorderforthemethodtoachieveSMS3,thecriteriaforDIDmustberespected.

MethodMaximum SMS score (method, implementation)

Adjusted SMS score (method, implementation)

Propensity Score Matching (PSM)

a.k.a.

Matching

With DID or panel method

3,3if

• Matchingcriteriasatisfied(seebelow)

• DIDorpanelcriteriasatisfied

3,2if

• Oneofthecriteriaisviolated

Cross-sectional

2,2if

• Goodmatchingvariables(i.e.relevanttoselection)

• Significantcommonsupport

2,1if

• Oneofthecriteriaisviolated

PSM Example 1 (SMS 3, 3)

A2012studybyBuscha,Maurel,Page,andSpeckesserevaluatestheimpactofhighschoolemploymentoneducationaloutcomes.17Thisquestionisdifficulttoevaluatebecausehighschoolstudentsthathaveapart-timejobmaynotbedirectlycomparabletotheirnon-workingpeers;itcouldbethattheyaremorehard-workinganddrivenandthereforehavebettergradesanyway,oritcouldbethattheycomefromdisadvantagedbackgroundsandthereforehaveworsegrades.Toovercomethefactthattreatedanduntreatedhighschoolstudentsmaybeintrinsicallydifferent,theauthorsemploythePropensityScoreMatchingapproachincombinationwiththeDIDmethod.InorderforthisstudytoachievethemaximumSMSscoreof3,thethreecriteriamustbesatisfied.

1. Good matching variables

Pass

Studentsarematchedaccordingtosocio-economicbackground,familybackground,schoollevelandregionalvariables,parentalexpenditure,schoolabsenteeismandconflict,alcoholanddrugissues,andeducationalaspirationsatgrade8.

2. Significant common support

Pass

Thereisaverylargeareaofcommonsupport,meaningthatfewobservationsweredroppedduetolackofsuitablematch.

3. Treatment group would have followed same trend as control group

Pass

17 Buscha,F.,Maurel,A.,Page,L.,Speckesser,S.(2012).TheEffectofEmploymentwhileinHighSchoolonEducationalAttainment:AConditionalDifference-in-DifferencesApproach.OxfordBulletinofEconomicsandStatistics,380-396.

Page 27: Guide to scoring evidencemethod, scoring them on the quality of their implementation. This scoring guide plays several important roles. Firstly, consistent with our commitment to openness,

Guide to scoring evidence using the Maryland Scientific Method Scale - Updated June 2016 25

Usingthematchingmethod,theauthorsuseobservationallysimilarstudentsascontrolsforworkingstudents.Althoughunobserveddifferencesmayremain,itcanbereasonablyarguedthatthetwogroupswilllikelyfollowthesametrend.

4. Treatment date known and singular

Pass

Theauthorsuseadatasetthattracksstudentsfromwhentheywere8thgradein1988,until1992whentheygraduatedhighschool.Thedatasetincludesinformationforthreewaves:1988,1990,and1992.Becausestudentsarenotlegallyallowedtoworkbeforetheageof14,thedatasethaspreandposttreatmentinformationforthosethatchoosetoworkduringtheirhighschoolyears.Therefore,becausestudentsareonlylegallyallowedtoworkfrom1990onwards,thetreatmentdateisknownandsingular.

Given that the four criteria are satisfied, this study achieves SMS 3 for implementation.

PSM Example 2 (SMS 2, 2)

A2011studybyGhalib,Malki,andImaievaluatestheimpactofmicrofinanceonruralhouseholdpoverty.18Inthiscasethereisaclearproblemofselectionbiasbecausehouseholdsthatapplyformicrofinancemaybeintrinsicallydifferentfromthosethatdonot.Forinstance,theymayhavemoreentrepreneurialdriveorsimplybemorefinanciallyliterate.Toovercomethisproblem,theauthorsemployPSM.

1. Good matching variables

Pass

Treatmentandcontrolgroupsarematchedcharacteristicsthatincludeageofadults,childdependencyratio,accesstoelectricity,homeownershipstatus,consumptionofluxuryfood,percentageofliterateadults,andavailabilityoftoilets.Thesevariablesareimportantinthesensethattheycaptureafamily’slevelofwealthandeducation,anditcanbereasonablyheldthatpoorerhouseholdsselectintotreatment(orapplyformicrocreditinthiscase).

2. Significant common support

Pass

Whenanalysingtheoverlapbetweentreatmentandcontrol,theauthorsfindthatonly11observations(ofthe1,132)aredroppedduetoalackofsuitablematch.

Given that the two criteria are satisfied, this study achieves SMS 2.

18 Ghalib,A.,Malki,I.,Imai,K.(2011).TheEffectofEmploymentwhileinHighSchoolonEducationalAttainment:AConditionalDifference-in-DifferencesApproach.RIEBDiscussionPaperSeries.

Page 28: Guide to scoring evidencemethod, scoring them on the quality of their implementation. This scoring guide plays several important roles. Firstly, consistent with our commitment to openness,

Guide to scoring evidence using the Maryland Scientific Method Scale - Updated June 2016 26

SMS 2 methods and below

Cross-sectional regressionCross-sectionalregressionsentailusingadatasetthatfeaturesmanydifferentindividualsatonepointintime.Thismethodsimplycomparestreatedindividualswithuntreatedindividuals,withouttakingintoaccountthedifferentcharacteristicsofthetwogroupsthatmayinfluencetheoutcome.Anattempttoaccountfordifferencesbetweentreatmentandcontrolgroupscanentailtheinclusionofcontrolvariables.However,evenifsomeobservablecharacteristicsarecontrolledfor,unobservabledifferencesbetweenthegroupsremain.Forthisreason,thismethodcanonlyachieveamaximumSMS2.Inordertoachievethisscore,however,adequatecontrolvariablesmustbeused.

MethodMaximum SMS score (method, implementation)

Adjusted SMS score (method, implementation)

Cross-section regression 2,2if

• Adequatecontrolvariablesareused

2,1if

• Inadequatecontrolvariables

Cross-Sectional Regression Example 1 (SMS 2, 2)

A2012studybySarzynskievaluatestheimpactofpopulationsizeonthelevelofacity’spollution.19Theauthorusesadatasetthatfeaturesasampleof8,038citiesworldwidefortheyear2005.InorderforthestudytoachievethemaximumscoreofSMS2,thecriterionmustbesatisfied.

1. Adequate control variables are used

Pass

Besidespopulation,othercityvariablessuchasGDPpercapita,populationdensity,growthrate,climate,anddevelopmentstatusareincludedintheregression.Here,itisimportanttothinkabout

19 Sarzynski,A.(2012).BiggerIsNotAlwaysBetter:AComparativeAnalysisofCitiesandtheirAirPollutionImpact.UrbanStudies,3121,3138.

07

Page 29: Guide to scoring evidencemethod, scoring them on the quality of their implementation. This scoring guide plays several important roles. Firstly, consistent with our commitment to openness,

Guide to scoring evidence using the Maryland Scientific Method Scale - Updated June 2016 27

whetherthereareanyimportantvariablesthatareomittedandmaythereforebiastheresults.Inthisparticularcase,theauthorseemstohaveincludedanexhaustivelistoffactorsthatmayinfluencepollutionlevels.

Given that the criterion is satisfied, this study achieves SMS 2 for implementation.

Cross-Sectional Regression Example 2 (SMS 2, 1)

A2007studybyPateletal.evaluatestheimpactoflivinginanurbanversusruralsettingonmentalhealth.20Theyuseacross-sectionaldatasetofhouseholdsinMaputoandCuambainMozambique.

1. Adequate control variables are used

Fail

Toestimatetheimpactoflivinginaruralsettingonmentalhealth,theauthorsperformasimplebivariateregression.

Giventhatthecriterionisnotrespected,thisstudyachievesSMS1forimplementation.

Before-and-AfterThismethodoftenentailsusingatimeseriesdatasetforwhichoneindividualistrackedacrosstime.Anexampleofthisisthevariationofaspecificcountry’sGDPoveranumberofdifferentyears.Inthiscase,itispossiblethattheindividualisobservedbeforethetreatmentandafterthetreatment.However,the“after”doesnotnecessarilycapturethepuretreatmenteffect-itispossiblethatothercontextualfactorsaffectedtheoutcome,aswellastheindividual’scharacteristics.Forthisreason,thismethodcanachieveamaximumSMSscoreof2.Inordertoachievethisscore,however,relevantcontrolvariablesmustbeused.

MethodMaximum SMS score (method, implementation)

Adjusted SMS score (method, implementation)

Before-and-after 2,2if

• Adequatecontrolvariablesareused

2,2if

• Inadequatecontrolvariables

Before-and-After Example 1 (SMS 2, 2)

A2014studybyBlankandEgginkevaluatestheimpactofdifferentgovernmentalhealthpoliciesonhospitalproductivity.Theyexploitatimeseriesdatasetthatfeaturestheoverallproductivityofdutchhospitalsfrom1972to2010.Itisworthnotingthatalthoughthismightseemlikeapaneldataset(becausethereareobviouslymanydifferenthospitals),itisessentiallyatimeseriesdatasetsimplybecausethereisnovariationintreatment-i.e.allhospitalsaresubjectedtothesamecentralgovernmentpolicies.Theonlyvariationintreatmentinthiscaseistemporal,asdifferentpolicieswereadoptedindifferenttimeperiods.InorderforthisstudytoachievethemaximumSMSof2,thecriterionmustberespected.

20 Patel,V.,Simbine,A.,Soares,I.Weiss,H.,Wheeler,E.(2007).PrevalenceofSevereMentalandNeurologicalDisordersinMozambique:aPopulation-BasedSurvey.Lancet,1055-1060.

Page 30: Guide to scoring evidencemethod, scoring them on the quality of their implementation. This scoring guide plays several important roles. Firstly, consistent with our commitment to openness,

Guide to scoring evidence using the Maryland Scientific Method Scale - Updated June 2016 28

1. Adequate control variables are used

Pass

Inthisparticularcase,theauthorsdocontrolforthingslikethepriceofinputs,wages,andnumberofadmissions.

Given that the criterion is satisfied, the study achieves SMS 2 for implementation.

Before and After Example 2 (SMS 2, 1)

A2000studybyCormanandMocanevaluatestheimpactofdruguseandlevelsoflawenforcementoncrime.21Morespecifically,theylookathowthenumberofarrestsandpoliceofficers,aswellasdruguse,impactthenumberofmurders,assaults,robberies,andburglaries.TheyexploitadatasetthatfeaturesmonthlycrimeratesforNewYorkCity.Here,the“treatment”isthevariationinarrests,numberofpoliceofficers,andhospitalisationsrelatedtodruguse.

1. Adequate control variables are used

Fail

Theauthorsdonotincludecontrolsforotherfactorsthatmayinfluencecrimerates.Forinstance,itispossibletheunemploymentlevelsorinflationmayimpactcrime.Becausethesefactorsarenotaccountedfor,itmaybethecasethatthetreatmenteffectsarenotentirelyattributabletonumberofarrests,policeofficers,anddruguse.

Given that the criterion is not satisfied, the study achieves SMS 1 for implementation.

Additionality (not SMS scoreable) Thismethodentailsaskingparticipantsinaprogrammeaboutthechangesinoutcomethattheyattributetotheprogrammeeffects.AnexamplewouldbetoaskafirminreceiptofanR&Dgrant‘doyouthinkyoudomoreR&Dasaresultofthegrant?’.Ifsay75%offirmsanswer‘yes’thentheremaining25%isconsidered‘deadweight’i.e.R&Dthatwouldhavehappenedanyway.Whilstthismethoddoesacknowledgedeadweight,thecounterfactualisbasedpurelyonopinion:whatthefirmthinkswouldhavehappened.Giventhelackofobservablecounterfactual,thismethodisnotSMSscoreable.

Additionality Example

A2013reportcommissionedbytheDepartmentforBusinessInnovationandSkillsevaluatestheimpactoftheEnterpriseFinanceGuaranteescheme(EFG),aprogrammethatprovidedSMEswithanadditionalsourceoffunding.22Themethodologyofthisevaluationrelieschieflyonself-reportedsurveyinformation.Morespecifically,EFGfirmswereaskedtorespondtoquestionnairesthataskedabouthowtheirbusinessesbenefitedfromtheprogramme.Self-reporteddataareoftentheobjectofbias,andforthisreason,thisstudyisnotSMSscoreable.

21 Corman,H.,&Mocan,H.(2000).ATime-SeriesAnalysisofCrime,Deterrence,andDrugAbuseinNewYorkCity.AmericanEconomicReview,584-604.

22 Allinson,G.,Robson,P.,Stone,I.(2013).EconomicEvaluationoftheEnterpriseFinanceGuarantee(EFG)Scheme.DepartmentforBusinessInnovation&Skills.

Page 31: Guide to scoring evidencemethod, scoring them on the quality of their implementation. This scoring guide plays several important roles. Firstly, consistent with our commitment to openness,

Guide to scoring evidence using the Maryland Scientific Method Scale - Updated June 2016 29

Impact modelling (not SMS scoreable)Impactmodellingisanapproachtoproduceexanteorexpostestimatesoftheeconomicimpactofapolicyorevent.Estimatesaremadeusing‘modellingassumptions’aboutthewayinwhicheventsorpolicesimpactonlocalareas.Forexample,impactmodelsadjustestimatesusingscalefactorstoaccountforissuessuchasdisplacementeffects–wherespendingassociatedwithalocaleventisoffsetfromspendingelsewhereorinanothertimeperiod.Sinceestimatesarebased(partlyorentirely)onassumptions,ratherthanobservedoutcomes,theapproachisparticularlyamenabletoexantestudiesofimpactwhereoutcomesarenotyetobservableinthedata(e.g.modellingthepotentialeconomicimpactsofafestivalorotherbigevent).

However,theresultsfromsuchmodellingprocessesareverysensitivetotheassumptionsmade,implyingthatinaccurateassumptionscanleadtoveryunreliableestimates.Inpractice,theassumptionsusedtendnotbebasedonstrongevidenceandareoftennotrelevanttothespecificcaseathand.Forexample,thereisnoreasontoexpectlargelocalmultipliereffectsinalocaleconomythatisalsopartofthelargernationaleconomy,andnostraightforwardwaytodividetheimpactsacrossspace.Similarly,itisneverreallypossibletoknowaccuratelyhowmuchdisplacementoccurs(e.g.inthefestivalexample,howmanyindividualswhowouldhavevisitedduringthefestival,avoidthefestivalandeithervisitanothertimeornotatall).Sincethismethoddoesnotprovideareasonablecounterfactual,itisnotSMSscoreable.

Impact modelling Example 1

AstudycommissionedbyChelmsfordCityCouncilevaluatestheimpactoftheVFestival(amusicfestival)onthelocaleconomy.Theystartbycalculatingthefestival’stotaldirectexpenditureanditsincidenceonthecity,county,andregion.Todothis,theauthorsusesurveydataontheexpenditurebyMetropolisMusic(theorganisersoftheevent),festivalcontractors,andvisitors.Inordertoestimatetheimpactofthisexpenditureonthelocalarea,theymakevariousassumptionsaboutthesizeofthelocalmultipliereffect,leakageeffectsnddisplacementeffects.Overallthedirectexpenditure(netimpacts)areestimatedat£7.4m(£6.6m)fortheEastofEnglandregion,£7.2m(£7.4m)forEssexcountyand£6.6m(£8.2m)forthecityofChelmsford.Notably,thenetimpactishigherthanexpenditureinChelmsfordbutlowerthanexpenditureintheregion.ThisreflectsthefactsomeexpenditureinChelmsfordisprobablypartlydisplacedfromthegreaterregion(50%ofvisitorstothefestivalcamefromtheEastofEnglandregion)whereasthelocalmultiplierisassumedtobethesame(about1.45)forthecity,countyandregion.

Inpracticethough,itisverydifficulttoensuretheseassumptionsarecorrect.Forexample,onewaytheseassumptionscouldfalldownisthatitislikelythatthelocalmultiplierissmallerinasinglecitythaninanentireregion.Alternatively,leakageanddisplacementeffectsmaybemuchlargerthanthought.ForChelmsford,itisassumedthatonlyaround12.8%ofexpenditurewouldhavehappenedanywayorleavesthecity.Giventhatthegreatmajorityofexpenditureisonfoodonthefestivalsite,thisfactorcouldbeanunderestimateif,forexample,alotofthecateringfirmscomefromoutsideofChelmsford.

Insummary,makingbroadassumptionsabouteffectsisoftenastraightforwardwaytoestimateimpactbut,duetothemanysourcesforerror,itisnosubstituteforobservingactualchangeinoutcomes(e.g.localfirmGVA)comparedwithacontrolgroup.SincethismethoddoesnotprovideacounterfactualitisnotSMSscoreable.

Page 32: Guide to scoring evidencemethod, scoring them on the quality of their implementation. This scoring guide plays several important roles. Firstly, consistent with our commitment to openness,

Guide to scoring evidence using the Maryland Scientific Method Scale - Updated June 2016 30

Appendix 1: Quick-scoring guide for the Maryland Scientific Method Scale

MethodMaximum SMS score (method, implementation)

Adjusted SMS score (method, implementation)

Randomised Control Trial (RCT)

a.k.a.

FieldExperiment

5,5if

• Randomisationissuccessful

• Attritioncarefullyaddressedornotanissue

• Contaminationnotanissue

5,4if

• Oneofthecriteriaisseverelyviolated

5,3if

• Twoormoreofthecriteriaareseverelyviolated

Instrumental Variable (IV)

a.k.a.

Two-StageLeastSquares(2SLS)

4,4ifinstrument

• Relevant(explainstreatment)

• Exogenous(notexplainedbyoutcome)

• Excludable(doesnotdirectlyaffectoutcome)

Scoredasperunderlyingmethodif

• Instrumentinvalid

• e.gcrosssectionwithinvalidIVscores2;difference-in-differencewithinvalidIVscores3(seebelow)

Regression Discontinuity Design (RDD)

4,4if

• Discontinuityintreatmentissharp(e.g.stricteligibilityrequirement)orfuzzydiscontinuitymethodused

• Onlytreatmentchangesatboundary

• Behaviourisnotmanipulatedtomakethecut-off

Scoredasperunderlyingmethodif

• Discontinuityconditionsseverelyviolated

• e.gcrosssectionwithinvalidIVscores2;difference-in-differencewithinvalidIVscores3(seebelow)

Difference in differences (DID)

a.k.a.

Diffindiff

3,3if

• Controlgroupwouldhavefollowedsametrendandtreatmentgroup

• Knowntimeperiodfortreatment

3,2if

• Eitherofthecriteriaisnotsatisfied

Panel methods

Panel Fixed Effects (FE)

3,3if

• Fixedeffectisattheunitofobservation

• Yeareffectsareincluded

• Appropriatetime-varyingcontrolsareused

3,2if

• Oneormoreofthethreecriteriaisnotsatisfied

First Differences (FD)

3,3if

• Yeareffectsareincluded

• Appropriatetime-varyingcontrolsareused

3,2if

• Eitherofthetwocriteriaisnotsatisfied

Arellano-Bond (AB)

3,3if

Yeareffectsareincluded

• Appropriatetime-varyingcontrolsareused

3,2if

• Oneofthetwocriteriaisnotsatisfied

Page 33: Guide to scoring evidencemethod, scoring them on the quality of their implementation. This scoring guide plays several important roles. Firstly, consistent with our commitment to openness,

Guide to scoring evidence using the Maryland Scientific Method Scale - Updated June 2016 31

MethodMaximum SMS score (method, implementation)

Adjusted SMS score (method, implementation)

Hazard Regressions

Mixed Proportional Hazards (MPH)

4,4if

• Keyassumptionof‘noanticipation’holds

• Variationintiming(e.g.peoplestarttrainingatdifferenttimesrelativetobecomingunemployed)

4,3if

• Anticipationoftreatmentlikely

• Littlevariationintiming

4,2if

• Neitherofthecriteriaissatisfied

Proportional Hazards (PH)

3,3if

• Adequatecontrolgroupisestablished

• Treatmentdateisknownandsingular

3,2if

• Oneofthetwocriteriaisnotsatisfied

Heckman Two-Stage Approach (H2S)

Or

Control Function (CF)

With IV 4,4if

• SelectionequationincludesanIVthatsatisfiesthethreecriteria(seeIV)

Scoredasperunderlyingmethodif

• Instrumentinvalid

• e.gcrosssectionwithinvalidIVscores2;difference-in-differencewithinvalidIVscores3(seebelow)

With DID or panel method

3,3if

• Selectionequationincludesrelevantobservablevariables

• DIDorpanelcriteriamet

3,2if

• Oneofthecriteriaisnotsatisfied

Cross-sectional

2,2if

• Selectionequationincludesrelevantobservablevariables

2,1if

• Selectionequationdoesnotincluderelevantobservablevariables

Propensity Score Matching (PSM)

a.k.a.

Matching

With DID or panel method

3,3if

• Matchingcriteriasatisfied(seebelow)

• DIDorpanelcriteriasatisfied

3,2if

• Oneofthecriteriaisviolated

Cross-sectional

2,2if

• Goodmatchingvariables(i.e.relevanttoselection)

• Significantcommonsupport

2,1if

• Oneofthecriteriaisviolated

Cross-sectional regression 2,2if

• Adequatecontrolvariablesareused

2,1if

• Inadequatecontrolvariables

Before-and-after 2,2if

• Adequatecontrolvariablesareused

2,1if

• Inadequatecontrolvariables

Page 34: Guide to scoring evidencemethod, scoring them on the quality of their implementation. This scoring guide plays several important roles. Firstly, consistent with our commitment to openness,

Guide to scoring evidence using the Maryland Scientific Method Scale - Updated June 2016 32

MethodMaximum SMS score (method, implementation)

Adjusted SMS score (method, implementation)

Stated effects/impact

a.k.a.

Additionality

Statedexperiences

NotSMSscoreable

Impact modelling NotSMSscoreable

Page 35: Guide to scoring evidencemethod, scoring them on the quality of their implementation. This scoring guide plays several important roles. Firstly, consistent with our commitment to openness,

Guide to scoring evidence using the Maryland Scientific Method Scale - Updated June 2016 33

Appendix 2: Mixed methods scoring 4 or 3.Asdiscussedinthemaintext,thereareanumberofmethodsthatarelessfrequentlyencounteredinlocaleconomicgrowthevaluationswhichmayscorea4ora3dependingontheparticularwayinwhichtheyarespecified.Thissectionexaminesthesemethodsandhowwescorethem.

Hazard Regressions Hazardfunctionsareusedtofindouttheimpactofapolicyonanoutcomethatrepresentsaduration.Theyarecommoninlaboureconomics,wheretheyareoftenusedtoanalysethevariablesthatimpactstrikeorunemploymentduration.Thehazardfunctioncalculatestheprobabilitythatanindividualwillleaveagivenstate(forinstance,unemployment)atagivenmomentintime.Proportionalhazardmodelsacknowledgethatthedependentvariableisafunctionofthetreatment,someobservedexplanatoryvariables(suchasageoreducation),arandomvariablethataccountsforindividualheterogeneity,andabase-linehazard.Thisbase-linehazardisessentiallythe“average”hazardfunction(i.e.propensitytoexitunemployment).Thismethoddealswithselectionintheprogrammeonobservablecharacteristicsbycontrollingfortheeffectofthesefactorsonthehazardrate.However,sinceselectionintothetreatmentgroupmayoccuronunobservablevariables,abiasmaypersist.BecausethismethodcontrolswellforobservablesbutisnotabletodealwithunobservablesitscoresamaximumofSMS3.However,inorderforthemethodtoachieveSMS3,twocriteriamustbesatisfied.Firstly,acontrolgroupmustbeused,anditmustbecrediblyarguedthattreatmentgroupwouldhavefollowedthesametrendasthiscontrolgroup,haditnotbeenexposedtotreatment.Secondly,theremustalsobeaknownandsingulartreatmentdatesothatthegroupscanbecomparedbeforeandafterthetreatment.

AnextensionofProportionalHazardmodelsistheMixedProportionalHazard(MPH)model.TheMPHexploitstimevariationintreatmenttoconstructacontrolgroup.Morespecifically,ithingesonthefactthatindividualsareexposedtotreatmentatdifferenttimes.Therefore,individualsthatreceivedthetreatmentatthebeginningoftheirunemploymentspellarecomparedtoindividualsthatreceivedthetreatmentafewmonthsaftertheyenteredunemployment.Here,theperiodwheretheindividualdidnothavetreatmentisusedasacontrolgroup.Thebasicideaisthattheindividualthatgottreatmentfromtheoutsetandtheindividualthatgottreatmentafewmonthslaterareverysimilarbecausetheybothself-selectedintotreatment,theonlydifferencebeingthepointatwhichtheyactuallygotit.BecausetheMPHfeaturesacontrolgroupthatisplausiblyunobservablyandobservablysimilartothetreatmentgroup,itcanachieveamaximumofSMS4.However,inorderforittoachievethismaximum,itmustmeettwokeycriteria.Firstly,individualscannotanticipatetreatment.Forinstance,ifanindividualknowstheyaregoingtoreceivetraininginamonth’stime,theymaynotsearchforjobsasintenselyastheydidbeforeknowing.Thismeansthatthe“control”groupiseffectivelytaintedandtheimpactofthepolicywillthereforelikelybeoverstated.Secondly,theremustbevariationintimingofindividuals’treatment.TheMPHmodelessentiallycomparesindividualsthatgotthetreatmentstraightawaywiththosethatgotthetreatmentabitlater.Forthisreason,itisextremelyimportantthatthisvariationexists.

Page 36: Guide to scoring evidencemethod, scoring them on the quality of their implementation. This scoring guide plays several important roles. Firstly, consistent with our commitment to openness,

Guide to scoring evidence using the Maryland Scientific Method Scale - Updated June 2016 34

MethodMaximum SMS score (method, implementation)

Adjusted SMS score (method, implementation)

Hazard Regressions

Mixed Proportional Hazards (MPH)

4,4if

• Keyassumptionof‘noanticipation’holds

• Variationintiming(e.g.peoplestarttrainingatdifferenttimesrelativetobecomingunemployed)

4,3if

• Anticipationoftreatmentlikely

• Littlevariationintiming

4,2if

• Neitherofthecriteriaissatisfied

Proportional Hazards (PH)

3,3if

• Adequatecontrolgroupisestablished

• Treatmentdateisknownandsingular

3,2if

• Oneofthetwocriteriaisnotsatisfied

MPH Example 1 (SMS 4, 4)

A2008paperbyLalive,vanOurs,andZweimullerevaluatestheimpactofactivelabourmarketprogrammesinSwitzerland.23Theseprogrammesaimtohelptheunemployedbyprovidingassistanceinsearchingforjobs.Aswithmostunemploymentprogrammes,thisisparticularlyhardtoevaluateastheunemployedarelikelytobedifferentfromtheemployed.Toovercomethis,theauthorsemployaMixedProportionalHazard(MPH)approach.InorderforthestudytoachievethemaximumSMSscoreof4,thetwocriteriamustbesatisfied.

1. No anticipation of treatment

Pass

Inthisparticularcase,itisarguedthatanticipationdoesnotplayasignificantrolebecauseSwissparticipantswereonlynotifiedaboutactualparticipationonetotwoweeksinadvance.Furthermore,theauthorshighlightthatinSwitzerlandtherearepenaltiesforjobseekersthatreducetheirsearcheffortsinanticipationofprogrammeparticipation.Lastly,theypointtothefactthattheprogrammewasoversubscribed,sojobseekerscouldnotaccuratelyknowwhenandiftheywouldbeparticipating.

2. Variation in timing

Pass

Theprogrammedoesnotspecifythatindividualsshouldbeunemployedforacertainamountoftimeinordertobeeligible.Thismeansthattheauthorscanstudytheeffectsoftheprogrammeforindividualsthathavebeenunemployedfordifferenttimeperiods.

Given that the two criteria are satisfied, this study achieves SMS 4 for implementation.

23 Lalive,R.,VanOurs,J.,Zweimuller,J.(2008).TheImpactofActiveLabourMarketProgrammesonTheDurationofUnemploymentinSwitzerland.EconomicJournal,235-257.

Page 37: Guide to scoring evidencemethod, scoring them on the quality of their implementation. This scoring guide plays several important roles. Firstly, consistent with our commitment to openness,

Guide to scoring evidence using the Maryland Scientific Method Scale - Updated June 2016 35

MPH Example 2 (SMS 4, 3)

A2006studybyHujerandZeissevaluatestheimpactofajobcreationschemeonemployment.24Policiesthatseektoimproveindividuals’chancesoffindingajobareparticularlyhardtoevaluatebecauseunemploymentisoftendeterminedbypeople’sobservedandunobservedcharacteristics.Accordingly,theauthorsusetheMPHmodeltoovercomethisissue.

1. No anticipation of treatment

Fail

Itisextremelyunlikelythatindividualsdidnotknowbeforehandthattheywouldbeparticipatinginthejobschemeinthefuture,meaningthatthisconditiondoesnothold.Itisthereforelikelythatparticipantsstoppedsearchingforjobsasintenselywhentheyfoundoutthattheywouldbeparticipatinginthejobcreationscheme.

2. Variation in timing

Pass

Peoplethatparticipateintheprogrammedosowithvaryingamountsoftimeinunemployment.Inpractice,thismeansthattherearen’tanyeligibilityrequirementsintermsofunemploymenttime,thereforeimplyingthattheeffectoftheprogrammeonpeoplewithdifferingtimesofunemploymentcanbestudied.

Given that only one of the criteria is satisfied, this study achieves SMS 3 for implementation.

PH Example 3 (SMS 3, 3)

A2013studybyPalaliandvanOursevaluatestheimpactoflivingclosetoacoffee-shoponcannabisuse.25Theauthorsuseahazardfunctiontolookathowlongpeople“survive”withoutusingcannabis.Furthermore,theyexploitthefactthatcoffee-shopsbecamewidespreadduringthe1980s/90s,andthatthereweren’tanybeforethen.Theythereforeusecohortsthatwereinprimedrug-usingageaftertheriseofcoffee-shopsasthetreatmentgroup,andcohortsthataretoooldtobeaffectedcoffee-shopsasacontrol.

1. Adequate control group is established

Pass

Inthiscase,thetreatmentgroupisthecohortbornbetween1974and1992,whilethecontrolgroupisthecohortbornbetween1955and1973.Itcouldbeargued,however,thatthedifferentcohortsareexposedtodifferentculturaltrends,andarethereforeinherentlydifferenttoeachother,particularlywithrespecttocannabisuse.Tocounterthis,theauthorscontrolforthingslikereligiousaffiliation,urbanversusruralresidence,andmigrantstatus.

2. Treatment date is known and singular

Pass

In1980,theDutchgovernmentpubliclyannounceditspolicyoftoleranceofcoffee-shops.Thisledtoarapidincreaseinthenumberofestablishments,sothatbythemid-1990stherewerearound

24 Hujer,R.,Zeiss,C.(2006).TheEffectsofJobCreationSchemesontheUnemploymentDurationinEastGermany.IABDiscussionPapers.

25 Palali,A.,&VanOurs,J.(2013).DistancetoCannabis-ShopsandAgeofOnsetofCannabisUse.CentERDiscussionPaper,2013-2048.

Page 38: Guide to scoring evidencemethod, scoring them on the quality of their implementation. This scoring guide plays several important roles. Firstly, consistent with our commitment to openness,

Guide to scoring evidence using the Maryland Scientific Method Scale - Updated June 2016 36

1500coffee-shops.Itcanbereasonablybelievedthatthisannouncementisasingularandtraceabletreatment.

Given that the two criteria are satisfied, this study achieves SMS 3 for implementation.

PH Example 2 (SMS 3, 2)

A1990studybyGundersonandMelinoevaluatestheimpactofpublicpolicyonstrikeduration.26Becausetheoutcomevariablehastodowithduration,theauthorsuseahazardregression.InordertoachievethemaximumSMSscoreof3,thetwocriteriamustbesatisfied.

1. Adequate control group is established

Fail

Inthiscase,theauthorsdonotattempttofindacontrolgroup.Instead,theysimplyrunaregressionwherebythestrikedurationsofdifferentunitswithdifferentpublicpoliciesarecompared.Therefore,itcanbeheldthattheauthorsdonotcomparetreatedunitswithuntreatedunits,butcompareunitswithdifferentintensitiesoftreatment.

2. Treatment date is known and singular

Fail

Thereisnoattempttodoacomparisonbeforeandafterthetreatment.Instead,placeswithinitialandfixeddifferinglevelsoftreatmentarecompared.

Given that none of the criteria are satisfied, this study achieves SMS 2 for implementation.

Heckman two-stage correction (H2S)/ Control function (CF)Thesemethodsuseatwo-stageapproachtoovercomeselectionbias.Thefirststageentailscalculatinga“selectionequation”thatincludesthecharacteristicsthatmayleadsomeonetoselectintotreatment.Inthesecondstage,elementsofthisselectionequationareincorporatedintothefinalregression(theonethatincludesthetreatmenteffectandoutcomevariableofinterest).Inthisway,theinclusionoftheselectionequationeffectively“absorbs”anyofthepre-existingselectionbias.However,althoughthesemethodsprovideapotentialsolutionforselectionbias,theydonotnecessarilyachievethisaim.Thisisbecausethequalityofthesemethodsislargelydependentonthetypesofvariablesthatareincludedintheselectionequation.Iftheselectionequationincludesahigh-qualityinstrument,thenitcanbeheldthattheselectionequationsuccessfullyaccountsforanyunobservableselectionbias,meaningthatthesemethodscanachieveamaximumofSMS4.Thisscorecanbeachievediftheinstrumentintheselectionequationsatisfiesthethreecriteriaforvalidinstruments(exogenous,relevant,andexclusionary).However,iftheselectionequationonlyincludesobservablecharacteristicsandnoinstrument,thentheunobservableelementofselectionbiasremainsanissue.Inthiscase,thesemethodscanachieveamaximumSMS2.Thismaximumscorecanbeachievedifthevariablesintheselectionequationadequatelyexplainselectionintotreatmentthatisbasedonobservablecharacteristics.However,evenifonlyobservablecharacteristicsareincludedintheselectionequation,ifthesemethodsarecombinedwithothermethodsthatattempttocontrolforunobservablecharacteristics(forinstance,DIDorpaneldatamethods),thentheycanachieveamaximumSMSscoreof3.Inordertoachievethisscore,theymustsatisfythecriteriaforthemethodthattheyareusedincombinationwith.

26 Morley,G.&Melino,A.(1990).TheEffectsofPublicPolicyonStrikeDuration.JournalofLaborEconomics

Page 39: Guide to scoring evidencemethod, scoring them on the quality of their implementation. This scoring guide plays several important roles. Firstly, consistent with our commitment to openness,

Guide to scoring evidence using the Maryland Scientific Method Scale - Updated June 2016 37

MethodMaximum SMS score (method, implementation)

Adjusted SMS score (method, implementation)

Heckman Two-Stage Approach (H2S)

Or

Control Function (CF)

With IV 4,4if

• SelectionequationincludesanIVthatsatisfiesthethreecriteria(seeIV)

Scoredasperunderlyingmethodif

• Instrumentinvalid

• e.gcrosssectionwithinvalidIVscores2;difference-in-differencewithinvalidIVscores3(seebelow)

With DID or panel method

3,3if

• Selectionequationincludesrelevantobservablevariables

• DIDorpanelcriteriamet

3,2if

• Oneofthecriteriaisnotsatisfied

Cross-sectional

2,2if

• Selectionequationincludesrelevantobservablevariables

2,1if

• Selectionequationdoesnotincluderelevantobservablevariables

H2S with IV Example (SMS 4, 3)

A2013studybyChen,Sheng,andFindlayevaluatestheimpactofindustry-levelforeigndirectinvestment(FDI)ontheexportvalueofdomesticfirmsinChina.27Thereisaclearproblemofselectionbias,asindustriesthatreceiveFDIarelikelydifferenttothosethatdonot.Toovercomethis,theauthorsemploytheHeckmantwo-stageapproach.Furthermore,theyuseFDIinflowstoASEANcountries(perindustry)asaninstrumentforFDIinflows,andincludethisvariableintheselectionequation.

Instrument is:

1. Instrument is relevant

Pass

ItcanbeplausiblyheldthatFDIinChinaiscloselyrelatedtotheFDIinflowsofASEANcountriesasawhole.

2. Instrument is exogenous

Fail

WhetherornotaparticularindustryattractsFDIisnotrandom.Forthisreason,itcannotbereasonablyarguedthattheinstrumentisexogenous.

3. Instrument is exclusionary

Pass

FDIinASEANcountriesdoesnotdirectlyimpacttheexportvalueofChinesefirms.

Given that only two of the criteria are satisfied, this study achieves SMS 3 for implementation.

27 Chen,C.,Sheng,Y.,Findlay,C.(2013).ExportSpilloversofFDIonChina’sDomesticFirms.ReviewofInternationalEconomics,841-856.

Page 40: Guide to scoring evidencemethod, scoring them on the quality of their implementation. This scoring guide plays several important roles. Firstly, consistent with our commitment to openness,

Guide to scoring evidence using the Maryland Scientific Method Scale - Updated June 2016 38

CF with DID Example (SMS 3, 3)

A2006studybyAndrenandAndrenevaluatestheimpactofvocationaltrainingonemployment.28Toovercomeissuesofselectionbias,theyemploytheC.F.method.Furthermore,itcontrolforunobservableselectionintotreatment,theyconstructacontrolgroupofindividualsthatdidnotreceivethetraining.Theythencomparetreatedandcontrolindividuals’propensitytoleaveunemploymentbeforeandafterthetraining.

1. Selection equation includes relevant observable variables

Pass

Theauthorsincludevariablessuchasage,sex,education,whethertheindividualhaschildrenornot,andregionofresidence.Withrespecttoobservables,thisseemstobeanadequatelist.

2. Control group would have followed same trend as treatment group

Pass

TheauthorshadaccesstoarichdatabaseofallunemployedindividualsinSweden.Theythereforechoseindividualsthatwereobservationallysimilartotreatedindividualstoformtheircontrolgroup.Accordingly,itcanbecrediblyarguedthattreatmentandcontrolgroupsareatleastobservationallysimilar,andthereforewilllikelyfollowthesametrends.

3. Treatment date is known and singular

Pass

Althoughdifferentindividualsstartedandendedthevocationaltrainingprogrammeatdifferenttimes,itisreasonabletopresumethatthereisaclearandsingulardateoftreatmentforeachperson.Thisisbecausevocationaltrainingprogrammeshaveaclearstartdate,sothatthereareno“partial”treatmenteffects.

Giventhatthethreecriteriaaresatisfied,thestudyachievesSMS3forimplementation.

H2S Cross-sectional Example (SMS 2, 2)

A2009studybyHammarstedtevaluatestheimpactoffutureearningsonthedecisiontoseekself-employmentinsteadofwage-employment.29Thedecisiontobecomeanentrepreneurislikelytaintedbyproblemsofself-selection–thosethatchoosetoforgowageemploymentareprobablyinherentlydifferenttothosethatdonot.Forthisreason,theauthorusestheHeckmantwostageapproachtocorrectforselectionbias.However,becausethestudyonfeaturesobservationsfortheyear2003,itispurelycross-sectional.

1. Selection equation includes relevant observable variables

Pass

Theauthorincludesanextensivelistofcontrolvariablesthataccountfordifferencesineducationandlocation.

Giventhatthecriterionissatisfied,thisstudyachievesSMS2forimplementation.

28 Andren,T.&Andren,D.(2006).AssessingtheEmploymentEffectsofVocationalTrainingUsingaOne-FactorModel.AppliedEconomics,2469-2486.

29 Hammarstedt,M.(2009).PredictedEarningsandthePropensityforSelf-Employment.InternationalJournalofManpower,349-359.

Page 41: Guide to scoring evidencemethod, scoring them on the quality of their implementation. This scoring guide plays several important roles. Firstly, consistent with our commitment to openness,

Guide to scoring evidence using the Maryland Scientific Method Scale - Updated June 2016 39

Appendix 3: Arellano-Bond methodTheArellano-Bond(AB)methodtakesasimilarapproachtoeitherthefixedeffectsorfirstdifferencesmethodbutincludesalagoftheoutcomevariableintheregression..However,ABrecognisesthatthelagmaybecorrelatedwiththeerrorterm,meaningthatthecoefficientsmaybebiased.Forthisreason,afurtherlagisusedasaninstrumentfortheoriginallag.

MethodMaximum SMS score (method, implementation)

Adjusted SMS score (method, implementation)

Panel Methods

Arellano-Bond (AB)

3,3if

• Yeareffectsareincluded

• Appropriatetime-varyingcontrolsareused

3,2if

• Oneofthetwocriteriaisnotsatisfied

AB Example 1 (SMS 3, 3)

A2010studybyBayona-SaezandGarcia-MarcoevaluatestheimpactoftheEurekaprogramme(aEuropeaninitiativetosupportR&D)onfirmperformance.30AswiththeevaluationofotherR&Dsubsidyprogrammes,thereisprobablyaproblemofselectionbiasasfirmsthatusethesubsidiesarelikelytobedifferentfromthosethatdonot.Toovercomethis,theauthorsusetheArellano-Bondmethod.

1. Year effects are included

Pass

Theauthorsincludeayearvariableinthefinalregression.

2. Appropriate time-varying controls used

Pass

Inthisparticularcase,itislikelythatthereareimportantvariablesthataffectfirmprofitabilityandvarywithtime.Accordingly,theauthorsincludeatime-varying“firmsize”variable.

Given that the two criteria are satisfied, this study achieves SMS 3 for implementation

30 Bayona-Sáez,C.&García-Marco,T.(2010).AssessingtheEffectivenessoftheEurekaProgram.ResearchPolicy,1375-1386.

Page 42: Guide to scoring evidencemethod, scoring them on the quality of their implementation. This scoring guide plays several important roles. Firstly, consistent with our commitment to openness,

The What Works Centre for Local Economic Growth is a collaboration between the London School of Economics and Political Science (LSE), Centre for Cities and Arup.

www.whatworksgrowth.org

Page 43: Guide to scoring evidencemethod, scoring them on the quality of their implementation. This scoring guide plays several important roles. Firstly, consistent with our commitment to openness,

ThisworkispublishedbytheWhatWorksCentreforLocalEconomicGrowth,whichisfundedbyagrantfromtheEconomicandSocialResearchCouncil,theDepartmentforBusiness,InnovationandSkillsandtheDepartmentofCommunitiesandLocalGovernment.ThesupportoftheFundersisacknowledged.TheviewsexpressedarethoseoftheCentreanddonotrepresenttheviewsoftheFunders.

Everyefforthasbeenmadetoensuretheaccuracyofthereport,butnolegalresponsibilityisacceptedforanyerrorsomissionsormisleadingstatements.

Thereportincludesreferencetoresearchandpublicationsofthirdparties;thewhatworkscentreisnotresponsiblefor,andcannotguaranteetheaccuracyof,thosethirdpartymaterialsoranyrelatedmaterial.

June2016

WhatWorksCentreforLocalEconomicGrowth

[email protected]@whatworksgrowth

www.whatworksgrowth.org

©WhatWorksCentreforLocalEconomicGrowth2016