SERIES

IZA DP No. 5710

PAPER

A Flexicurity Labour Market in the Great Recession: The Case of Denmark

Torben M. Andersen

DISCUSSION

May 2011

Forschungsinstitut zur Zukunft der Arbeit Institute for the Study of Labor

A Flexicurity Labour Market in the Great Recession: The Case of Denmark

Torben M. Andersen
Aarhus University CEPR, CESifo and IZA

Discussion Paper No. 5710 May 2011

IZA P.O. Box 7240 53072 Bonn Germany Phone: +49-228-3894-0 Fax: +49-228-3894-180 E-mail: iza@iza.org

Any opinions expressed here are those of the author(s) and not those of IZA. Research published in this series may include views on policy, but the institute itself takes no institutional policy positions. The Institute for the Study of Labor (IZA) in Bonn is a local and virtual international research center and a place of communication between science, politics and business. IZA is an independent nonprofit organization supported by Deutsche Post Foundation. The center is associated with the University of Bonn and offers a stimulating research environment through its international network, workshops and conferences, data service, project support, research visits and doctoral program. IZA engages in (i) original and internationally competitive research in all fields of labor economics, (ii) development of policy concepts, and (iii) dissemination of research results and concepts to the interested public. IZA Discussion Papers often represent preliminary work and are circulated to encourage discussion. Citation of such a paper should account for its provisional character. A revised version may be available directly from the author.

IZA Discussion Paper No. 5710 May 2011

ABSTRACT A Flexicurity Labour Market in the Great Recession: The Case of Denmark*
Flexicurity labour markets are characterised by flexible hiring/firing rules, generous social safety net, and active labour market policies. How can such labour markets cope with the consequences of the Great Recession? Larger labour shedding is to be expected and this strains the social safety net and increases the demands on active labour market policies. This paper takes a closer look at the labour market consequences of the crisis for Denmark. It is found that employment adjustment is not particularly large in international comparison, although it has more weight on the extensive (number of employees) than the intensive (hours) margin. The level of job creation remains high, although job creation is pro-cyclical and job-separation counter-cyclical. As a consequence most unemployment spells remain short. This is critical since a persistent increase in unemployment will affect the financial balance of the model severely. Comparative evidence does not, however, indicate that flexicurity markets are more prone to persistence. Crucial for this is the design of the social safety net and in particular the active labour market policy. However, the larger inflow into activation raises questions concerning the possibility of maintaining the efficiency of the system.

JEL Classification: Keywords:

J01

flexicurity, employment protection, unemployment insurance, active labour market policy

Corresponding author: Torben M. Andersen School of Economics and Management Department of Economics University of Aarhus DK - 8000 Aarhus C Denmark E-mail: tandersen@econ.au.dk

Comments and suggestions from Michael Svarer, Coen Teulings and participants at the CPB-ROA conference on flexibility of the labour market, Den Haag, January 2011 are gratefully acknowledged.

1.

Introduction

Priortothefinancialcrisistherewasmuchfocusontheconceptofflexicurity.Theideathatitispossibleto ensureflexibilityforemployersandsecurityforemployers,withoutimpairinglabourmarketflexibilityand socialbalances,receivedsubstantialattention.TheEUcommissionevenproposedthatmembercountries shouldfollowtheflexicurityapproach,althoughthedefinitionoftheconceptwassobroadastomakethe precisemeaningunclear. Denmarkhasoftenbeenreferredtoasaflexicuritycaseduetothecombinationofflexiblehiringandfiring rulesaswellasageneroussocialsafetynet.Anequallyimportantelementofthemodelisasetofactive labourmarketpoliciesattachingconditionstotheclaimingofunemploymentandsocialbenefits,andalso includingprogrammestoenhancequalificationsandthusjobfindingchancesforjobless(seee.g.Andersen andSvarer(2007)). Intherunuptothefinancialcrisisunemploymentwaslow,andthemodelwasthustakentobewell functioning.ThecrisishasinducedaGreatRecessionwhichhasalsoaffectedDenmarkseverely.Outputhas beenfallingbyalmost6%fromtoptothrough(2008.3to2009.4)andtheunemploymentratehas increasedbysome34percentagepoints.Howcantheflexicuritymodelcopewithsuchlargeshocks?Since thecrisishasastrongglobalelementandaffectedmanycountriessimilarlyatroughlythesametime,these developmentsprovideanopportunitytodrawsometentativelessonsontheimportanceoflabourmarket institutionsandpolicies.Despiteitsdireconsequencethefinancialcrisisproduceasemicontrolled experimentontheimplicationsofpoliciesandinstitutionsfortheresponseoflabourmarketstoadeep crisis. Inaflexicuritylabourmarketwithflexiblefiringrulesitistobeexpectedthattheimmediateeffecton unemploymentislarge.Sinceitiseasytoshedlabouritisanimmediateimplicationthattheemployment tooutputresponseshouldbelarger,andinthiswaythemodelisvulnerabletoaggregateoutputshocks. Figure1:Unemployment,DenmarkandOECD,1970.12010.3

Datasource:www.sourceoecd.org

TheunemploymentratesfortheOECDareaandDenmarkaregiveninfigure1usingOECDdefinitions.Itis seenthatunemploymentinDenmarkpriortothecrisiswasmuchlowerthantheOECDaverage,butalso thatitincreasedmoresteeplywiththeonsetofthecrisis.AlthoughunemploymentinDenmarkhas increasedmorethaninmanyothercountries,itisstilllowininternationalcomparison.Though,thereal questioninaflexicuritylabourmarketisnotthesensitivityofunemploymenttotheaggregatestanceofthe economy,butwhetheritisassociatedwithlargesocialcosts,andwhetherittendstopersist. Thekeyquestionisthuswhetherthesocialsafetynetiscapableofabsorbingthisincreaseinjoblessness andalsowhetherthehighlevelofjobcreationandturnovercharacterizingthelabourmarketcanbe maintained.Thelatterisimportanttopreventtheincreaseinunemploymentfromproducingamore persistentincreaseinunemployment(longtermunemployment)orreductioninlabourforceparticipation. OnecanputitinthewaythattheGreatRecessionisanultimatetestoftheflexicuritymodel.Canitcope withthisandinduceareasonablequickreturntolowerunemploymentorwillitproduceapersistent increaseinjoblessnesswhichinturnviatheimplicationsforpublicfinanceswillmakethesituation untenable? Thereisnocoherenttheoreticalmodellingofallaspectsoftheflexicuritymodel,butitbringstogethera hugeliteratureontheroleofemploymentprotectionlegislation(EPL),unemploymentinsuranceschemes (UIB)andactivelabourmarketpolicies(ALMP)1.Itisbeyondthescopeofthispapertoprovideasurveyand referencestothisveryvastliterature,andthereforeonlyafewselectivepointsandreferencesaremade.A keytradeoffisbetweenemploymentprotectionlegislationprotectingjobs2andunemploymentbenefits providinginsurancetothoseloosingjobs.Thetwocanthusbeseenasalternativewaysofprotection workers(BlanchardandTirole(2008)andBoeri,CondeRuiz,andGalasso(2006)),althoughwithpotentially largeimplicationsforlabourmarketperformance3.Whileemploymentprotectionlegislationtendsto createamoresharpdistinctionbetweenprotectedandnonprotectedworkers,generousunemployment insuranceraisesanissueofmaintainingjobsearchincentives.However,activelabourmarketpoliciescan beusedtocounteracttheseeffectsbyassociatingrequirementstoreceivingunemploymentbenefits,see AndersenandSvarer(2010).Theflexicuritylabouristhusseenasacasewithlargeflexibilityforfirmsin adjustingtheirlabourforce,relativegenerousunemploymentbenefits,andactivelabourmarketpoliciesto maintaintheincentivestructureinthelabourmarketandtoovercomeobstaclesforemployment.Ifit worksithassomeattractiveproperties,butitmaybeparticularlyvulnerabletosignificantdropsin unemployment.Thiswillproduceasteepincreaseinexpendituresonunemploymentbenefitsandactive labourmarketpoliciesandthusbringthefinancialviabilityoftheschemeatstake.Thekeyquestionis whethertheflexicurityschemeisparticularlyvulnerabletolargenegativeshocks,andinherentlytendsto producepersistenceinadjustment. 1 DavoineandKeuchnigg(2010)isaninterestingcombinationofallthreeelementsinasettingwithsafeandrisky
jobs,andunemploymentconcentratedinhighproductiveandvolatilesectors.SeealsoBrown,MerklandSnower (2009) 2 TheliteratureonEPLsuggeststhatemploymentfluctuationsaredamped(seee.g.Nickell(1986))whiletheeffecton longrunemploymentismoreuncertain(BentolilaandBertola(1990)). 3 Thereisalsoaliteraturepointingtotheendogenousadoptionofindustrialstructuretolabourmarketinstitutions andpoliciesseeCuntandMelitz(2007).SeealsoLommerudandStraume(2010)onhowlabourmarketinstitutions affecttechnologyadoption.

Thispapertakesafirstlookattheseissuesintheperspectiveoflabouradjustmentsinresponsetothe greatrecession.Itisobviouslytooearlytomakeafinalcallonthesequestions,butitispossibletomakea midtermevaluationofhowtheDanishflexicuritymodelhascopedwiththecrisisinordertojudge whetheritisonastableorunstabletrack.Thispaperattemptstodothis,andstartsoutinSection2witha briefoutlineofsomekeycharacteristicsoftheDanishflexicuritymodel.Thelabourmarketconsequences ofthecrisisarelaidoutinSection3,andSection4turnstoamoredetailedanalysisoftheimplicationsand adjustmentprocessesreleasedfromacomparativeperspective.Section5offerssomeconcludingremarks.

2.TheDanishflexicuritymodel
ThemainingredientsoftheDanishflexicuritymodelarewellknown,andthefollowinggivesabrief overviewrelevantforthelatterdiscussion.Moredetailedaccountscanbefoundine.g.Andersenand Svarer(2007),JrgensenandKongshjMadsen(2007),andStrbyJensen(2008).Table1summarises countryindicatorsonemploymentprotection(EPL),unemploymentbenefitgenerosity(UIB)andactive labourmarketpolicies(ALMP). Table1:Employmentprotection(EPL),unemploymentbenefitgenerosity(UIB)andactivelabourmarket policies(ALMP) EPL UIB ALMP USA 0.2 LUX 87 DEN 4.8 UK 0.7 NZL 72 FIN 3.6 CAN 0.8 NET 71 BEL 3.6 IRE 1.0 SPA 69 NET 3.6 AUS NOR POL JPN DEN CH FIN AUT NET SWE BEL GER NZL ITA FRA SPA LUX GRE 1.1 1.1 1.5 1.6 1.8 1.9 2.1 2.1 2.4 DEN FRA SWE BEL GER AUT FIN CAN IRE 68 67 66 65 64 61 60 52 50 45 42 42 38 37 33 33 28 28 SWE GER SPA IRE FRA AUT NOR NZL CAN POL AUS ITA UK LUX JPN GRE USA CH 3.5 3.1 2.7 2.6 2.6 1.8 1.8 1.7 1.6 1.4 1.4 1.3 0.9 0.9 0.7 0.7 0.6 0.4

2.6 JPN 2.6 2.6 2.7 2.8 3.0 3.2 POL AUS NOR ITA GRE CH

3.3 UK 3.3 USA

Source:EPLmeasuredbyOECDepl1indexmeasuredonscalebetweenzeroandthree,UIBisthefirstyearreplacementrate,and ALMPasexpendituresonactivelabourmarketpoliciesasashareofGDP. Source:www.sourceoecd.organdOECD(2009).

2.1.Hiringandfiringrules Forhiringandfiringrulesthekeydistinctionisbetweenbluecollar(hourlypaid)orwhitecollar (funktionrer)workers. Forbluecollarworkersdismissalrulesaresettledincollectiveagreements4.Forworkerswithshorttenure thereisnodismissalperiod,whiletheremaybesomeforworkerswithlongertenure.Asanexample workersinthebuildingandconstructionsectorhaveaperiodofnoticeof3weeksafter1years employment,5weeksafter3yearsemploymentand7weeksafter5yearsofemployment5.Forworkers thenoticeis1weekafteroneyearemployment,and2weeksafter5yearsemployment. Thebasicruleforwhitecollarworkers6isthatthereisonemonthsnoticefordismissalswithinthefirst6 monthsofwork,andthisperiodisextendedby1monthper3yearsofemploymentuptoamaximumof6 monthsnotice(afteremploymentin8yearsand7months).Itispossibletobehiredaswhitecollarworker onprobationforaperiodupto3monthsinwhichcasethedismissalnoticeis14days.(Funktionrlovens paragraf2).Theemployeecanquitthejobwithonemonthsnotice. Incaseofdismissalofawhitecollarworkerwhohasbeenemployedinthecompanyinthelast12,15or18 years,theemployershouldcompensatetheworkerby1,2or3monthssalary,respectively.Itispossibleto negotiateseverancepaymentsaspartofthecontract. Foralllayoffstheemployercoversunemploymentbenefitsforthefirstthreedaysforallworkerswho withinthelastfourweekshavebeenemployedforatleasttwoweeks(Godtgrelsesdagsbetalinger).These aretheonlyformalseverancepayments.Notethatthishasbeenintroducedtoreducethenumberofshort termunemploymentspells. 2.2.Unemploymentinsuranceandthesocialsafetynet TheunemploymentinsurancesysteminDenmarkisavariantoftheGentmodelwithUIfundsand voluntaryindividualmembership.Contributionratesaredeterminedpoliticallyandarethesameacrossall UIfunds.TheUIfundsaresubsidizedbythepublicsector(inacountercyclicalway).Accordingly,the marginalcostsofincreasingunemploymentarefullytaxfinanced.Membershipisopentopersonswith relevantqualificationsforthespecificUIfund,orviaregularworkwithinitsarea.Theunemployment benefitcannotexceed90%ofthepreviouswage(calculatedoverthelastyear)oragivencap(currently about22.300eurosannually,taxableincome),anditisindexedtogeneralwagedevelopments7.The maximumdurationofbenefitsis4years(from2011:2years),andtheentitlementtobenefitscanbe regainedbyregularworkforatleast6months(2011:12months)withinthelast36months.Moreover, certainactivationrequirementsareassociatedwithclaimingofbenefits(seebelow).WhenUIbenefits
InDenmarkthereisatraditionofhavingmostlabourmarketissuessettledbythepartiesinthelabourmarketrather thanbylegislation. 5 http://www.workinfo.dk/daDK/EMNER/ID117/ID127/ID133.aspx 6 Inthecaseofsicknessformorethan120dayswithinthelastyeartherearespecialrulesfordismissalswithone monthsnotice 7 Thecurrentindexingformulaisfromalawenactedin1990(revised2003)accordingtowhichalltransfersare indexedonthebasisoftheannualwageincreasestwoyearsearlier.Iftheincreaseisabove2%,apartof0,3%is transferredtoafund(satsreguleringspuljen)whichisspentoninitiativesaimingatimprovingtheconditionsfor peopleontransfers.
4

expire,theindividualwouldnormallybeeligibleforsocialassistancewhichisalsothecasefor unemployedwhoarenotmembersofanUIfund.Thesocialassistanceschemeisrathercomplicatedsince thebenefitlevelamongotherthingsdependsonageandmaritalstatus,andinaddition,therearevarious meanstestedsupplements.Itisthereforedifficulttogeneralizeonthefallintransferincomeupon transitionfromUIbenefitstosocialassistance.Apersonreceivingthemaximumunemploymentbenefit wouldexperienceanincomereductionof2040%. Itisanimplicationofthisschemethatthereplacementrateisstronglydependentonpreviousincome, sincethecapimpliesthatthe90%compensationonlyappliesforlowincomegroups.InFigure2,the replacementrateisdepictedasafunctionofincomeforDenmarkandotherselectedcountries.Itisseen thatDenmarkstandsoutbyhavingahighreplacementrateforlowincomegroups,butnotforhigher income.Hence,whentheDanishUIsystemischaracterizedasverygenerous,ithastobemadewitha proviso,andtheflexicuritycharacterizationofDenmarkismostfittingforlowincomegroups. Figure2:Incomedependentreplacementrates,Selectedcountries
1

0,9
Danish APW

0,8

0,7 rate of compensation

0,6

0,5

0,4

Denmark Belgium Sweden Netherlands Finland

0,3

0,2

0,1

0 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 210 220 230 Daily wage in Euros

Note:Replacementrateinunemploymentinsuranceschemescalculatedonthebasisofdailyincome,andforsinglepersonsin countrieswherebenefitsaredependentonsocioeconomicvariables.Dailywagesaretruncatedatrelevantminimumwagelevels (forcountrieswithnolegalminimumwagethenegotiatedminimumwagesareused). Source:AndersenandSvarer(2007)

2.3.Activelabourmarketpolicies Forpersonseligibleforunemploymentbenefitsorreceivingsocialassistanceorstartaid8thebasicrules areasfollows9: Unemployedbelowtheageof30havetherightanddutytoanactivationofferafternolaterthan13weeks ofunemployment.Theactivityshouldhaveanuninterrupteddurationof6months. Forpersonsbelow25yearswithoutalabourmarketrelevanteducation(andwithoutchildren)the activationrequirementisaneducationundernormalconditions(studygrants).Inaddition,thelevelof socialassistanceisreducedafter6monthsunemploymenttothesamelevelasthestudygrant. Forpersonsabove30therightanddutytoanactivationofferisnolaterthanafter9monthsof unemployment.Though,theperiodisonly6monthsforpersonswithunemploymentinsuranceandabove 60years. Afterthefirstofferthereisarightanddutytoanewofferforeachconsecutiveperiodof6monthson publicsupport,andthedurationoftheoffershouldbeatleast4weeks. Unemployedeligibleforunemploymentinsurancewithanunemploymentperiodof30monthshavearight anddutytoanactivationofferfortheremainingpartofthe4yearunemploymentbenefitperiod(fulltime activation). Theactivationoffercanbeinoneofthreeformswhichallhavespecificrulesonduration:I)counsellingand requalification.Thisincludesshortcounsellingandassessmentprogrammesaswellasspecialprojectsand educationintheordinaryeducationalsystem,II)jobtraining(virksomhedspraktik),thismaybeinthe publicsectororaprivatecompany,andisusedforpersonswherethereisaneedtoclarifythepossiblejob prospects,orforpersonswhoduetolackofqualificationsmayhavedifficultiesinfindingajobonnormal conditions,III)employmentwithwagesubsidy,thismaybeatapublicorprivateemployerandisusedfor retrainingtoupgradethequalifications(specificskills,languageorsocialskils)ofunemployed.About2/3of allactivationsoffersareoftypeI,andtheothertwotypesconstituteabout1/6each. Thebasicingredientsoftheinteractionbetweenthesocialsafetynetandactivelabourmarketpoliciesin Denmarkareillustratedinfigure3showingboththesocialsafetynet(theupperpart)andtheactivation policies(thelowerpart)dependingonthedurationoftheunemploymentspell.Thusthereisduration dependenceinboththebenefitlevelandtheactivationrequirements.Whilethedurationdependencein benefitsisweak(longdurationofunemploymentbenefits),theeffectiveprofiledisplaysmoretime dependenceduetoasteppingupoftheactivationrequirementwithdurationofanunemploymentspell10.

Forthelattertwogroupsthisappliesiftheirproblemislackofajob. http://www.ams.dk/Reformerogindsatser/Indsatser/Kontaktogaktivering/Beskaeftigelsesrettetaktivering.aspx. 10 Itshouldbenotedthatadditionalemploymentconditionalitieshavebeenbuildintothescheme.Eligibilityforthe fullsocialassistancethushasconditionsonpastemployment,ashaschildsupport.

9

 Figure3:Danishsocialsafetynetincaseofunemployment Replacementrate%
Unemploymentbenefits RegainingeligibilityforUIB:6 monthsregularworkwithinthe last36months Lowincomegroups Socialassistance

90

60

APW

61218243036424854606672

o n

Active job

Activationoffer: rightandduty

Fulltime activation

AcActive Active aastive job job h

Activationoffer: rightandduty

t h s

Activationrequirement
Note:Diagramappliestoanindividualfulfillingtherequirementsforunemploymentbenefitsatthemomentofbecoming unemployed.Socialassistanceiscalculatedonthebasisofthebenefitsforasingleperson.Theactivationrequirementsarethe requirementsforpersonsabovetheageof30.APWreferstothereplacementrateforanaverageproductionworker.

3.Labourmarketconsequencesofthegreatrecession
Denmarkwasalikemostcountriesseverelyaffectedbythefinancialcrisis.Priortothecrisistheeconomy wasbooming,andtheunemploymentratewasrecordlow,cf.figure1.Therewereclearsignsof overheating,buttherewaspoliticaldelayinacceptingtheneedtoundertakecontractionarypolicies.The crisisimpliedadrasticandswiftchangeinthesituation,andhencetheDanishdevelopmentinrecentyears hasdisplayedaboombustpattern. Theincreaseinemploymentbeforethefinancialcrisiswassignificantlymorethanimpliedbythehistorical relationbetweenoutputandemployment,cf.figure4.Thisclearlyreflectedasituationwithshortageof labourandfirmsfacingrecruitmentproblems.Althoughtherewasanincreasinginflowofmigrantworkers, therewereclearsignsofexcessdemand.Thismayinparthavebeenstrengthenedbyfirmshoardinglabour inlightofthedifficultiesinrecruitingnewworkers.Thissituationrapidlychangedwiththeonsetofthe

crisis.Duetothehoardingoflabourtheinitialemploymentresponsewassharp,andlargerthanwhat wouldimmediatelyfollowfromtheoutputdecline,cf.figure4.Therehasthusbeensomeexcessvolatility Figure4:Actualandpredictedemployment,1991.42010.3

Note:PredictedemploymentfollowsfromanestimatedOkunsrelationforthesampleperiod1991.4to2007.4,i.e.thesample perioddoesnotincludethecrisis.Estimatedequation:ln(et)=a+bln(yt)+cln(yt1)+dln(yt2)+eln(yt3). Source:Ownestimationbasedondatafromwww.statistikbanken.dk

ofemploymentrelativetotheoutputpath,butthelevelofemploymentisnowalignedtooutputaccording tothehistoricalrelationbetweenthetwo,cf.figure4. Figure5:Output,totalhoursworkedandemploymentinthegreatrecession:2008.32010.3,index 2008.3=1

Note:Employmentandtotalhoursworkedfromnationalaccounts Source:www.statistikbanken.dk

 Itisinterestingtonotethattheadjustmentoflabourinputhasmainlybeenintheextensivedimension, thatis,therehasbeenasharpdeclineinemploymentbutonlyamodestchangeinworkinghours,cf.figure 5.Theoutputdropisabout6%andemploymentdecreasedalmostinthesameorderofmagnitudewitha lagof12quarters.Thedeclineintotalhoursworkedisonlymarginallylargerthanthatofemployment showingthattheadjustmentofworkinghourshasnotplayedamajorroleintheDanishcase. Figure6:Increaseingrossunemploymentandfallinemployment:2008.4

2010.3
Source:www.statistikbanken.dk

Thelabourmarketproblemcreatedbythecrisiscanbeassessedeitherfromthefallinemploymentand hencetheneedforjobcreationtoreturntopreviousemploymentlevels,oritcanbeassessedintermsof theincreaseinunemployment,i.e.thenumberwhoareexplicitlylookingforajob.Thesedifferentangles leavequitedifferentmessages.Theincreaseinunemploymenthasbeenmuchsmallerthanthedeclinein employment,cf.figure6.Grossunemploymentincludesunemployedandparticipantsinactivation receivingunemploymentbenefitsorsocialassistance.Roughlythedecreaseinemploymentisdoublethe increaseingrossunemployment.Thehugedifferencemaybeattributedtothreefactors.First, unemploymentstatisticsmayunderestimatethetruerateofjoblessnesssincenotallunemployedare eligibleforunemploymentbenefitsorsocialassistance.Priortothecrisistherewasadeclinein membershipinunemploymentinsurancefunds,andthusanincreaseinthenumberofnoninsured. Second,thestrainedlabourmarketsituationhasinducedincreasedenrolmentinvariousformsof education.Thepreviouspatternorfadamongyouthofworkingforsomeperiodasunskilledbefore commencingeducationhasbeenmoredifficultduetothestrainedlabourmarketsituation.Changesin enrolmentcanaccountforabout1/3ofthedifferencebetweenthefallinemploymentandtheincreasein grossunemployment.Finally,theremayhavebeenanoutflowofmigrantworkers.Therisein unemploymenthasaffectedvariousgroupsquitedifferently.Asformostcountriesthesharpestincreasein unemploymentisseenformales,youth,andunskilled.

10

Figure7:Registered,grossandAKUunemployment

Source:www.statistikbanken.dk

Thedifficultiesinassessingthestateofthelabourmarketisalsoreflectedinfigure7showingthree differentconceptsofunemployment.Thefirstisthenumberofregisteredunemployedi.e.unemployed receivingunemploymentbenefitsorsocialassistance.Thegrossunemploymentaddsthenumberofpeople inactivationprogrammes,andasexpectedthedifferencebetweengrossandregisteredunemployment hasincreasedinresponsetothecrisis.Finally,theAKUunemploymentisasurveybasedmetricof unemployment.Whilethelatterisassociatedwithvariousmeasurementproblemsthewideningdifference betweentheAKUandgrossunemploymentmayreflectthatnotalljoblessareeligibleforunemployment benefitsorsocialassistance. Thelowunemploymentratepriortothecrisisledtosomewagepressure.Howevertheconsumerreal wageincreasedmorethantheproducerrealwagereflectingtermsoftradegains(figure8a).Atfirst,this reversedatendencyforthewagesharetodecline(figure8b),andthenimpliedanincreaseintheshare whichpeakedwhentheunemploymentratewasatthebottom.Throughthe2000stherehasthusbeena clearcyclicalpatterninthewageshare,reflectingmainlysomevariationsinlabourproductivity.

11

Figure8:Realwagesandwageshare,2000.12010.3(2000.1=1)

Note: (a)The wage is wage index for private sector, the consumer real wage is found by deflating by the implicit consumption deflatorandtheproducerrealwagebytheimplicitdeflatorforgrossvalueadded.(b)Thewageshareisfortheentireeconomy. (c)hourlylabourproductivityfortheprivatesector(privatebyerhverv),and(d)wagecompetitivenessmeasuredbyrelativeunit labourcosts. Source:www.statistikbanken.dkandwww.sourceoecd.org

Intheperioduptotheonsetofthecrisis,labourproductivitywasfallingreflectingthehoardingoflabour (figure3).Theadjustmentofemploymenthasimpliedareadjustmentoflabourproductivitywhichisnow almostbacktotrend(figure8c).Whilewageincreaseshavebeenrelativemoderatedespitethelowlevelof unemployment,itremainsafactthatwagecompetitivenesshasdeteriorated(figure8c).Thisistosome extenttobeexpectedduetotheextraordinarylowlevelofunemployment(figure1)butalsoreflectsthat trendproductivityhasbeengrowingmoreslowlyinDenmarkthanforitscompetitors.

12

4.Adjustmentandburdensharing
TheadjustmentprocessintheDanishlabourmarketisofinterestinitsownright,butalsoinabroader perspectiveofthelessonstobedrawnontheroleofflexicuritylabourmarketinstitutions.Thefollowing considertheseissuesbyaddressingsomekeypointswheretheDanishdevelopmentsareseenin comparativeperspectiveandrelatedtolabourmarketinstitutionsandpolicies. Figure9:Adjustmentoftotallabourinputrelativetooutputchangedecomposedinhoursand employmentadjustment.

Note:Changemeasuredfrompeaktotroughduringthe200809recession(quarterlydata).Totalchangeinlabourinputis decomposedintochangesinhoursperworkerandnumberofemployees. Datasource:OECD(2010).

4.1.Adjustinglabourinput Itistobeexpectedthatemploymentismoresensitivetooutputinacountrywithmorelaxemployment protection.Asalreadynotedabove,theadjustmentoflabourinputinDenmarkhasmainlybeenviathe numberofemployees,andworkinghourshasonlybeenmoderatelychanged.Withlaxemployment protectionlegislationandageneroussocialsafetynet,thecostsofadjustinglabouralongtheextensive marginaresmaller,whilestrictemploymentprotectionmaymakehoursadjustmentacheapermodeof adjustment. Theadjustmentoflabourinputrelativetotheoutputchangeisinfigure9decomposedinthechangein hoursandemploymentforOECDcountries.Thereisquitesubstantialvariationintheadjustmentoflabour inputrelativetotheoutputchange,butalsosubstantialdifferencesintheroleplayedbyhoursrelativeto employment.Denmarkdoesnotstandoutamongthecountrieswiththelargestchangesinlabourinput relativetooutput,andmoreovertherelativelargeburdentakenbyemploymentisseenagainhere. Asnoted,therelativeburdenoflabourinputadjustmentbetweenhoursandemployeesdependsonthe relativecostsofchangingalongtheintensiveandextensivemargin.Thestrictnessofemployment protectionlegislationisimportantforthesecosts,andhoursadjustmentsarelikelytoberelativemore importantthemorestricttheEPLregulationis.However,inthecaseofsufficientlylargechangesinoutput

13

andthusneededlabourinputitmaybeinevitabletoadjustthelabourforce,evenifitiscostly.Itisseen fromfigure10thatthereisatendencythatcountrieswithmorestrictEPL,alargeradjustmentburdenfalls onhoursthanonemployees.TheexceptionsaretheoutliersSpainandPortugalhavinglargechangesinthe numberofemployeesrelativetotheoutputchangedespiterelativestrictEPLregulation. Figure10:ShareoflabourinputadjustmentviahoursandEPL

Note:Seenotetofigure9. Datasource:OECD(2009)

Asimpleregression,cf.table2,showsasexpectedthatthehighertheEPLindexis,themorelabouris adjustedviachangesinworkinghoursratherthaninthenumberofemployees.Unemploymentbenefit generositytendstolowertheroleofworkinghoursasanadjustmentmechanism.Thelargertheoutput change,thelessthehoursshare,reflectingthattheneededadjustmentscannotbeachievedwithoutalso changingthenumberofemployees.However,theexplanatorypowerisweak,andalargeunexplained variationintherelativeroleoftheintensiveandextensivemarginoflabourinputremains. Table2:Shareofhoursinlabouradjustment EPL UIB ALMP 0.20 0.02 0.11 (0.27) (0.92) (0.37) 0.24 0.31 0.07 (0.26) (0.91) (0.39) R2 0.05 0.18

Outputchange NI 0.15 (0.11)

Constant 4.07 (3.36) 5.55 (3.41)

Note:allvariablesinlogsNI=notincluded.Shareofhoursinlabouradjustmentasinfigure9,EPLasintable1,andoutputchange isthenumericalvalueofthechangeinoutput.Estimationfor17OECDcountries.

Thesmallerthechangeinlabourinputrelativetotheoutputdecline,themorelabourproductivityfalls. Thishasimplicationsfortheadjustmentovertime.Ifoutputrecoversthereareidleresourcesmakingthis

14

possibleuptosomelevelwithoutincreasinglabourinput.Ifoutputdoesnotrecover,adelayedlabour inputadjustmentistobeexpected. Figure11:Jobcreationanddestruction,OECDCountries,200005

Note:Percentageratesadjustedbyindustrycomposition Datasource:OECD(2009)

4.2.Inflowandoutflowfromunemployment Inalabourmarketinwhichitiseasytodismisspeopleitistobeexpectedtofindalargeinflowinto unemployment.Atthesametimeflexibilityisassociatedwithahighlevelofjobturnoverandthusexit fromthepoolofunemployed.ThisisanimportantcharacteristicoftheDanishlabourmarket,andfigure11 showsthatthegrosslevelofbothjobcreationanddestructionishighincomparativeperspective. Grossflowsarethuslargeinthelabourmarketimplyingthatmanyareaffectedbyunemploymentbutin mostcasesonlyforashortperiod.Figure12showsthesurvivalrateforthegroupofunemployedattwo differentdates,atthepeakpriortothecrisis(2008.2),andinthemidstoftherecession(2009.3).As

15

Figure12:Survivalfunctionunemploymentincludingactivation.2008.2and2007.3

Note:Survivalrateappliestothegroupofunemployedandinactivation Source:www.jobindsats.dk

expectedtheexitfromunemploymentisquickerintheboomthanintherecessionperiod,reflectingthat thejobfindingrateisprocyclical.However,eveninthemidstofthefinancialcrisistheexitratefrom unemploymentisstrikinglyhigh.After13weeksabout60%haveleftthegroupofunemployed,andafter 26weeks80%.Hence,evenduringtherecessiontherearelargegrossflowsinandoutofjobsandmany areaffectedbyunemploymentonlyforashortperiod.Thelargeincidenceofshorttermunemployment remainsdespitetheincreaseinunemployment. Figure13:Unemploymentintensitynumberaffectedbyunemploymentdependingondurationand fulltimeequivalents(2009q3)

Source:www.statistikbanken.dk

16

Theanatomyofunemploymentrevealedbyconsideringthenumberofpersonsaffectedandfulltime equivalentsisshowninfigure13basedondatafrom2009.3.Intermsofthenumberofpeopleaffectedby unemploymentwithintheyearabout1/3areunemployedforlessthan30%oftheyear,andintermsof fulltimeequivalentsthiscontributesonlyabout10%tototalunemployment.About22%ofthoseaffected byunemploymentwithintheyearareunemployedforatleast80%oftheyear,andintermsoffulltime equivalentstheyaccountforoftotalunemployment.Inshort,thenumberofpeopleaffectedby unemploymentwithintheyearismorethantwicetheunemploymentinfulltimeequivalents,duetothe highincidenceofshorttermunemployment.Manyareaffectedbyshortspellsofunemployment,butthe burdenofunemploymentrestsonasmallgroupoflongtermunemployed.

Thehighgrossflowsinthelabourmarketarealsoreflectedintheperceptionpeoplehaveconcerningthe possibilitiesoffindingajobifbeinglaidoff.InEurobarometersurveyscloseto70%answerthattheyare fairlyorveryconfidentthattheycanfindajobiftheyarelaidoff.ThisisthehighestrateforanyEUcountry (Eurobarometer(2010)).Thisshowsthattheriskassociatedtoaparticularjobisnotassociatedwitha generalperceivedriskconcerningthepossibilityoffindingajoborbeinginemployment.Thisisconsistent withthelargegrossflowsandthehighjobfindingrate,cf.above. Figure14:EPLandshortandlongtermunemployment

Note:Shorttermunemploymentisdurationlessthan3months,longtermunemploymentisdurationlargerthan6months. Datasource:www.sourceoecd.org

Thehighincidenceofshorttermunemploymentandrelativelyfewlongtermunemployedisrelatedto labourmarketinstitutions.Figure15belowshowstheshareofshorttermandlongtermunemployment andtheEPLindex.ItisseenthatalabourmarketwithlowEPLtendstohaveahighincidenceofshortterm unemployment,butalowleveloflongtermunemployment.Thisreflectsthelargergrossflow,andthus,a smallerconcentrationofunemploymentonamarginalizedgroupoflongtermunemployed.Thisis

17

consistentwiththeevidencereportedinfigure13thatshorttermunemploymentisadominanttypeof unemploymentinDenmark. Table3:Shortandlongtermunemploymentandlabourmarketcharacteristics EPL UIB ALMP Constant Shortterm 0.31 0.28 0.07 2.64 unemployment (0.13) (0.47) (0.19) (1.72) Longterm 0.36 0.57 0.02 5.70 unemployment (0.14) (0.49) (0.20) (1.80) R2 0.27 0.29

Noteestimationsfor21OECDcountries.Shareofshortandlongtermunemploymentmeasuredasinfigure14,andEPL,UIBand ALMPasinTable1.Allvariablesareinlogs.Standarddeviationsinparenthesis.

Table3confirmsthathighEPLtendstoimplythatshorttermunemploymentconstitutealessershareof totalunemployment,whilelongtermunemploymentconstitutesalargershare.Unemploymentbenefit generositytendstohavetheoppositeeffect(thoughnotsignificant).Thiscanbeinterpretedasshowing thattemporarylayoffsaremosteasywithbenefitgenerosity,andthatthisdominatespossiblelongterm disincentiveeffects(seebelow).Activelabourmarketpoliciesdonotinfluencethesplitofunemployment betweenshortandlongtermunemployment. Figure15:Burdenofunemploymentfallingonyouth,2009

Note:Thefiguresshowtheshareoftheincreaseinunemploymentbetween2008and2009whichisduetoanincreaseinyouth unemployment. Datasource:www.sourceoecd.org

Itisastylizedfactthatanincreaseinunemploymentisassociatedwithadisproportionatelargeincreasein youthunemployment.Thisreflectsprocyclicalhiringsrates.Inlabourmarketswithlargergrossflowsitis thustobeexpectedthatyouthunemploymentislessstronglyaffected.Figure15showstheshareofthe unemploymentincreaseduetothecrisiswhichisfallingonyouth(belowage25),anditisseenthat Denmarkisamongthecountrieswitharelativelowshareoftheunemploymentincreasecarriedbyyouth.

18

Thismaybeattributedbothtothelargergrossflowsinthelabourmarketandthushigherjobfindingrates andtothestrongrequirementsonunemployedyouth,cfsection2. 4.3.Persistenceandthesocialsafetynet11 Aconcerninthepresentsituationisthattheincreaseinunemploymentbecomespersistentaswasthe caseinmanycountriesduringthe1970sand1990s.Possiblecausesofpersistenceinthelabourmarket includedepreciationofhumancapitalincreasingwiththelengthofunemploymentspells,changesinthe wagesettingmechanismifitisprimarilyaffectedbyinsiders(theemployed)withlittleweightgiventothe outsiders(unemployed),orareductionofproductioncapacityasaresponsetothecrisis.Thekeyquestion hereiswhetherthesesourcesofpersistencearestrengthenedbythesocialsafetynet.VanderNoord et.al.(2006)findaweakpositiverelationbetweenpersistencemeasuredbythehalflivesofoutputgaps andsocialexpendituresasashareofGDP. LjungqvistandSargent(1998)describeagenerouswelfarestateasatimebombinthesensethatitmay operateefficientlyintranquiltimesbutbevulnerabletoturbulencewhicheasilytranslatesintopersistent unemployment12.Thelatteriscausedbyweakenedjobsearchactivitiesandhigherreservationwagesdue toageneroussocialsafetynet.Inparticular,shockstendtodepreciateskillsandthusrequireworkersto acceptawagecuttofindanewjob,butunemploymentbenefitsdependingonpastwagestendtocreate inertiaintheadjustmentofreservationwages.Asaconsequencethesafetynethinderstheprocessof restructuringtheeconomy.Itisalsoanimplicationofageneroustaxfinancedsocialsafetynetthatthe levelofmobilityacrossjobsislowered(LjungqvistandSargent(1995)).Thismaycontributetoreduce frictionalunemployment,butinducehigherstructuralunemploymentinasituationwithturbulence. Adifferentexplanationofpersistencehasbeenadvancedbypointingtotheroleofnormsincounteracting theincentiveeffectsofageneroussocialsafetynet(seee.g.Lindbeck(1995)andLindbecket.al(2003)).A strongnormtobeselfsupportingcounterstheeconomicincentivescreatedbyagenerousscheme. Allowingforthenormtobeendogenousanddepending(possiblywithalag)onthenumberofindividuals beingselfsupportingimpliesthatageneroussocialsafetynetcanbemaintainediftheemploymentrateis high.However,ifemploymentfallsduetoe.g.aseverebusinesscycledownturn,normsmaybeeroded, andthewelfarestateiscaughtinasituationwithpersistentnonemploymentandfiscalproblems. Whilethereisavoluminousempiricalliteratureaddressingtheroleofvariousinstitutionalfactorsincluding thesocialsafetynetforlabourmarketperformance(seee.g.Blanchard(2006)forasurveyanddiscussion), thereareveryfewstudieswhoexplicitlyaddressesthepersistenceissue.Twodifferentconceptualissues areatstakenamelyontheonehandthestructuralunemploymentrate,andontheotherhandthe responsivenessofthelabourmarketstoshocks.Thelatterinvolvesboththeimpacteffect(volatility)and theadjustmentprocess(persistence).Theseissuesareclearlyhighlyrelevantinthecurrentsituationwhere therehavebeenlargedecreasesinemployment.Thesechangesareirreversible,butitiscrucialtominimize theextenttowhichthistranslatesintopersistentreductionsinemployment.Accordinglythefollowing
ThissectionbuildsonAndersen(2011) Apossibilityofmultipleequilibriaalsoariseswhentakingintoaccountthefinancingofthesafetynet.Similarlyif incentiveproblemsarecounteredbycostlymonitoring,theeffectivenessofsuchmonitoringislargeatlowlevelsof unemploymentreinforcingthissituation,andoppositelyinasituationwithhighunemployment(Ljungqvistand Sargent(1995)).
12 11

19

takesacloserlookattheempiricalsupportforthesocialinsurancesclerosishypothesis,whichisvery importantforaflexicuritylabourmarket. Itisnotstraightforwardhowtomeasurepersistenceintheadjustmentprocess.Ideallyonewouldwantto separateexogenouspersistence(drivenbypersistenceinshocks)fromendogenouspersistence(drivenby adjustmentmechanismsinthesystem).Obviouslythisisverydifficultandwillinvariablyrelyonidentifying assumptionswhichmaybeopenfordebate.Itisbeyondthescopeofthispapertogointodetailswiththis, andratheramoresimplisticapproachispursuedbypresentingvariousmetricsofpersistence. Fortheperiod19702007figure9givesboththeautocorrelationandasignmetricofpersistencein unemploymentfor24OECDcountriesfortheemploymentrate(agegroup1664).Variousothermeasures ofpersistence,alsoappliedtounemploymentandemploymentratesarereportedinAndersen(2011). Whiletherearesomevariationsacrossthedifferentmeasures,theyarestronglycorrelated.Thesedifferent measuresleavethesamerelativerankingofthecountriesandhencetheparticularchoiceofmetricisof lessimportance.Itisanimplicationthatthereisstrongpersistenceforanumberofcountries,thatis,the adjustmentprocessislonglasting,cf.figure16. Figure16:UnemploymentpersistenceOECDcountries

Note:Dataisfor23OECDcountries1970to2007.ThesignmetricofpersistenceiscalculatedbasedonformulainDiasand Marques(2005).Bothmeasuresarecomputedfortheunemploymentrate.Thecorrelationbetweenthetwomeasuresis:0.61 Data:www.sourceoecd.org

Itisacrucialquestionwhetherthereisanyrelationbetweenthevolatilityandpersistenceinthelabour market.Isitthecasethatlabourmarketsexposedtovolatileshocksalsodisplaymorepersistence,orthe reverse?Figure17presentstwomeasuresofvolatilityinthelabourmarket,namely,thestandarddeviation ofunemploymentandthegrossjobflows(sumofjobseparationsandcreations).Forbothmeasuresthere isaweakpositivecorrelationwiththepersistencemeasure.Hence,itdoesnotseemthatcountrieswith lowvolatilityaremoreexposedtopersistenceintheadjustmentprocess,neitherisitclearthatmore volatilityisassociatedwithmorepersistence.

20

Figure17:Labourmarketvolatilityandpersistence

Note:Persistenceistheautocorrelationforunemploymentovertheperiod19702007.Volatilityisthestandarddeviationofthe unemploymentrateovertheperiod19702007.Grossjobflowsisthesumofhiringandseparationratesovertheperiod200007. Data:www.sourceoecd.org

Thepersistencemeasuresreportedaboveareopentovariousinterpretationproblemsandanalternative waytoapproachthesameproblemwouldbetolookatcasesoflargeunemploymentcrises.InAndersen (2011)alargeemploymentcrisesisdefinedasafallof3percentagepointsormoreintheemploymentrate withinathreeyearperiod,andthisleaves18sucheventsamongOECDcountriesovertheperiod1970 2007.Allthesecasesdisplayverystrongpersistenceinthesensethattherearenocaseswhere employmenthasrecoveredwithinfiveyearsandonlyfewinstanceswhereitisthecasewithintenyears aftertheonsetofthecrisis.Thereisthusevidencethatdeepemploymentcrisestendtobehighly persistent. Thenextstepistoconsiderwhetherthereareanyempiricalregularitieslinkingpersistencetopolicydesign andinstitutionsinthelabourmarket.Thisisadifficultendeavoursincethemetricsofpersistenceare impreciseandsinceitisdifficulttocharacteriseandsummarizepoliciesandinstitutionsinafewsimple measures.Thefollowingtakestwodifferentapproachestothisissuebothconsideringautomaticstabilizers asasummarymetricoftheextentofthesocialsafetynet,andbyconsideringvariousindicatorsrelatedto policiesandinstitutions. Automaticstabilizersplayanimportantroleinthemacroliteraturesincetheymeasuretheextenttowhich incomevariationsareabsorbedbythepublicbudgetandinthiswaycushionsincometoshocks.The quantitativesizeoftheautomaticstabilizersreflectsinasummarywaytheconsequencesofhowthesocial safetynetanditsfinancingisarranged.Hence,itcanbetakenasasimplesummarymeasureoftheextent ofthesocialsafetynetinthewidemeaningofincludingthestatedependenciesoftransfer,taxesand expenditures.Figure17givesacrossplotofautomaticstabilizersandunemploymentpersistence.Thereis noclearrelationbetweenthetwo.Interestinglythegroupofcountrieswiththestrongestautomatic

21

stabilizers(DEN,SWE,NET)hasalowerlevelofpersistencethanalargegroupofcountrieswithmiddle sizedautomaticstabilizerssuggestinganonlinearrelationship13. Figure18:Automaticstabilizersandunemploymentpersistence

Note:AutomaticstabilizersmetricsarefromvanderNoord(2000),Unemploymentpersistencemeasuredbytheautocorrelation 2 overtheperiod19702007,cf.figure9.Regressionline:y=0.25+0.33x,R =0.26. Source:www.sourceoecd.organdvanderNoord(2000).

Table4:Simplecorrelationmeasuresofpersistenceandpolicy/institutionalindicators Measureof EPL persistence Autocorrelation 0.10 (0.02) HP 0.14 autocorrelation (0.05) Signmetic 0.07 (0.02) UIB 0.06 (0.05) 0.13 (0.14) 0.03 (0.06) ALMP 0.01 (0.02) 0.02 (0.06) 0.04 (0.02) Constant 0.08 (0.18) 0.03 (0.51) 0.02 (0.21) R2 0.66 0.31 0.47

Note:Firstorderautocorrelationcoefficientforunemploymentrate,HPfilteredunemploymentrate,andthesignmetricfor persistence.Datafor21OECDcountriesEPL,UIBandALMPmeasuredasinTable1.Standarddeviationsonparanthesis. Source:www.sourceoecd.org

Findingthatpersistenceisonlyweaklyrelatedtothesizeofautomaticstabilizersmayreflectthatthelatter isatoosummarymeasure,andthataclearerpictureemergesbyconsideringindicatorsoflabourmarket institutionsandpolicies.Inthefollowingtheroleofemploymentprotectionlegislation,generosityofthe unemploymentinsurancescheme,andactivelabourmarketpoliciesareconsidered.Table4showsthe

Therelationbetweenpersistenceandtheautomaticstabilizersmaybenonlinear.Estimatingunemployment persistence(y)onautomaticstabilizers(x)inlevelsandsquaredyields(tvaluesinparanthesis)yields: y = 0.67 +0.92x 0.77x2 ;R2=0.26 (0.11) (0.45) (0.49)
13

22

resultsofasimpleregressionusingdifferentmetricsofpersistenceinunemployment.Itisseenthatmore strictemploymentprotectionlegislationtendstoproducemorepersistence,consistentwiththefinding thattheEPLtendstoincreasetheshareoflongtermunemploymentintotalunemployment,cf.table3. TheeffectofUIBissomewhatsurprisinglytoreducepersistence(thoughnotsignificant),andactivelabour marketpolicieshaveanuncleareffect. Importantforthepresentdiscussionthereisnoevidencesupportingthatcountrieswithamoreextended welfarestatesufferfrommorepersistence.Nonethelessasnotedabovethisisnotimplyingthat persistenceisunimportantforcountrieswithextendedwelfarestate.Quitethecontrarythestrong automaticbudgetresponsesimplythatpersistentdeclinesinemploymentrateswillhavedramatic consequencesforpublicfinances.Itisinterestingtonotethatthecountrieswithextendedwelfarestate areamongthecountrieswhichpriortothefinancialcrisisdidmosttoconsolidatepublicfinancesand undertakereformstoaddressfiscalsustainabilityproblemsarisingfromageing14. 4.4.MinimizingpersistenceachallengeforALMP ActivelabourmarketpoliciesareextensivelyusedandareanintegralpartoftheDanishflexicuritymodel. Theactivationrequirementsservetomaintainfocusonjobsearch,strengthenjobsearchincentives,and contributetoovercomequalificationbarriersforemployment.TheALMPisthusimportantinensuringthat asteepincreaseinunemploymentdoesnottranslateintopersistentunemployment.However,ALMPis alsostrainedwhenunemploymentincreases,thereisalargeinflowintoprogrammes,andtheymaybeless effectiveinarecession. Itisdifficultempiricallytodiscernthepreciseeffectofactivationpolicies.Moststudiestendtofocusonthe directeffectintermsoflockingineffectandpostprogrammeeffectsintermsofemployment.Whilethese areimportantelements,activationhasalsocrucialindirect(generalequilibrium)effects,seeAndersenand Svarer(2010)).Activationispartoftheeligibilityconditionsforsocialtransfers(unemploymentbenefits andsocialassistance)andthusservestoattachstrongerrequirementstoreceivingbenefits,cf.figure3. Thismayinturnbothhaveanincentiveeffectloweringthereservationrequirementstojobs(typeofjob, location,employer,wageetc).Thisissometimestermedathreatormotivationeffect.Thereisarelated screeningeffectsinceitismoredifficulttoclaimbenefitswhilenotbeinginterestedinworkorworkingin theblacksector.Relatedtotheabovetheremayalsobeawageeffect,sinceactivationnotonlyhasa directeffectonparticipantsbutalsoaffectsthefallbackpositionofemployediftheybecomeunemployed. Understandardassumptionsthisleadstowagemoderation.Hence,aproperassessmentoftheeffectsof activationisverydifficultandcareshouldbetakeninmakingconclusionsfrompartialanalysis. VariousassessmentshavebeenmadeofALMPinDenmarkandthegeneralfindingsarematchingfindings fromstudiesforothercountries,seee.g.Kluve(2006)andCardet.al.(2010).InRosholmandSvarer(2008) itisshownthatingeneralactivationhasalockingineffect.However,privatejobtrainingandeducation improvesqualifications,whilethereisnostatisticalsignificanteffectfrompublicjobtrainingandother formsofactivation.Thisstudyalsofindsathreat/motivationeffectfromallformsofactivationwhichtends toshortenthedurationofunemploymentspells.Overallitisassessedthatactivationwhentakinginto

InEuropeanCommission(2009)onfiscalsustainabilitytheScandinaviancountriesarefoundtobeamongthe countrieswiththesmallestsustainabilityproblems.
14

23

accountthelockingin,thequalificationandthemotivationeffectcontributestohigherjobfindingrates andthuslowerdurationofaverageunemploymentspells. PositivelongrunemploymenteffectsfromprivatejobtrainingarealsofoundinJespersenet.al.(2008), whilethereisnosucheffectfrompublicjobtrainingoreducationalactivationprogrammes. DanishEconomicCouncil(2007)considersthemotivation,lockinginandqualificationeffectofallformsof activation.Ingeneralactivationhasalockingineffect.Thereisapositiveemploymenteffectdueto improvedqualificationsfromprivatejobtrainingbutanegativeeffectfromotherformsofactivation.In additionapositivemotivationeffectisfound.Thestudyisnotabletofindanypositiveeffectsof educationalactivation,notevenafter56yearsafterparticipation. EducationalactivationprogrammesareanalysedinChristensenandJacobsen(2009)withparticularfocus onthelockinginandmotivationeffect.Thereisnoshortorlongrun(5years)positiveemploymenteffect fromordinaryeducationforpersonsbecomingunemployedin2002.However,positiveeffectsarefound forpersonswhobecameunemployedin1995.Forbothgroupstherearepositiveeffectsofprivatejob training.Itisalsofoundthattheeffectsofallformsofactivationarebetterwhenunemploymentislow (lesslockingin),butthedirecteffectoftheprogrammeisnotdifferentacrossbusinesscyclesituations. TheprogramEarlystart(Hurtigtigang)hasmadeitpossibletomakeanexperimentaldesignofthe effectsofactivelabourmarketpolicy.Inthisprogrammehalftheparticipantsreceivedtheactivationoffers asprescribedbyexistingruleswhiletheotherreceivedanearlyandmoreintensiveintervention.The programwasfirstlaunchedin20056andlaterrepeated200809toallowforamorespecificevaluationof theappliedactivities.Thegeneralfindingisthattheearlyinterventionhascontributedtoenhance employmentprospects,correspondingtoashorteningofunemploymentdurationspellsby3weeks(see Graversenet.al.(2007),Rosholm(2008),andRosholmandSvarer(2009)).Costbenefitanalysesshowthat theincreasedemploymentcancoverthecostsofrunningtheprogramme. Itisaparticularchallengetomaintaineffectivenessofactivationpoliciesinaperiodwithincreasing unemployment.First,theeffectsofactivationpoliciesmaybebusinesscycledependent,andlesseffective inadownturn.Second,thereisavolumeincreaseinparticipantswhichmayleadtodifficultiesinboth meetingthetimingrequirements(seeSection2),andalsoatwisttowardslesseffectiveprogramme activitiesduetoshortageofrelevantopenings.Finally,thecompositionofthegroupofunemployed changesandtargetingbecomesmoredifficult.Amongnewlyunemployedthereisalargegroupwhichis likelytofindanewjobquickly,andthereareotherswhichareatriskoflongtermunemployment.Ahigh levelofactivationmaythuscarryahighdeadweightcost. TheseproblemsarepresentinDenmark.Therehavebeenproblemsinmeetingtherequirements stipulatedintheruleswithregardtothedeadlinesforofferingofvariousinitiatives.Moreover,thevolume increasehasledtoincidenceswherethecontentand/orthequalityoftheactivityhavebeencalledinto question.Figure19showsthatthenumberofparticipantsinactivationhastripledsince2008,andthere hasbeensomedeclineintheaveragedurationofprogrammeactivities.

24

Figure19:Activationdegreeandaverageduration

Source:www.jobindsats.dk

Therearealsosignsthattheeffectivenessofactivationislowerduetothepresentsituationinthelabour market.Figure20displaystheshareinemploymentaftervarioushorizonsforparticipantsinactivation programmes.Itisseenthatthereisacleardeclineinrecentquarters.Thisisthedirecteffectoflowerjob findingrates,anditisaconcernsinceitincreasestheriskthatalargergroupbecomeslongterm unemployed. Figure20:Shareinemploymentafteractivation

Datasource:www.jobsindsats.dk

25

Note,thatthechallengeofmaintainingextensiveactivationpoliciesduringarecessionhastwosides.One istheeconomicarisingfromfirstacostincreaseandeventuallyfromtheproblemsarisingiflongterm unemploymentincreases.Theotheristhatthepoliticalsupportforactivelabourmarketpoliciesmay weakeninaperiodwithlowlabourdemandandexamplesofnotwellfunctioningactivationprogrammes.

5.Concludingremarks
TheDanishlabourmarkethasbeenseverelyaffectedbytheGreatRecession.However,incomparative perspectivetheperformanceisstillcharacterizedbybelowaverageunemployment,largegrossflows,and nostrongertendenciesforunemploymenttobecomepersistent.However,openissuesremains,in particularconcerningtheeffectivenessoftheactivelabourmarketpoliciesinadeeprecession.Balancing thecostpressurewiththevolumeincreaseandmaintainingefficiencyofprogrammeactivitiesisnotan easytaskinaperiodwithalowerjobfindingrate.Howeverthisiscrucialinpreventinganincreaseinlong termunemploymentwhichbothhassocialcostsbutalsowillaffectpublicfinancesseverely.Itistooearly tomakeacallonthesequestions. IncomparativeperspectivetheDanishcasefitsintoapicturewherecountrieswithlaxEPLtendstocarrya largeburdenofadjustmentoflabourinputviathenumberofemployeesratherthanworkinghours. Unemploymentbenefitgenerositytendstogointhesamedirection.LessstrictEPLregulationalsotendsto implythatshorttermunemploymentconstitutesarelativelylargershareofoverallunemploymentthan longtermunemployment.Youthunemploymentalsotendstobelessaffectedbyrecessionsincountries withlaxEPL,andchangesinunemploymentingeneraltendtobelesspersistent.Theevidenceonwhat characterizesthesuccesscasesislessconclusive.Whilethereissomeindicationthatgenerous unemploymentinsuranceincombinationwithactivelabourmarketpoliciesmakesitpossibletoachieve betterrisksharingwithoutimpairinglabourmarketflexibility,theevidenceismuchtooinconclusiveto makestrongstatementsonthis.Tosettletheseissuemoreresearchisneeded.

References
Andersen,T.M.andM.Svarer,2007,FlexicurityLabourmarketperformanceinDenmark,CESifo EconomicStudies. Andersen,T.M.andM.Svarer,2010,Theroleofworkfareinstrikingabalancebetweenincentivesand insuranceinthelabourmarket,WorkingPaper. Andersen,T.M.,2010,Collectiverisksharing:Thesocialsafetynetandemployment,WorkingPaper Andersen,T.M.,2011,Employmentcrises,WorkingPaperinpreparation Bentolila,S.,andBertola,G.,1990,FiringCotsandLabourdemand:HowbadisEurosclerosis?Reviewof EconomicsStudies57,381402. Blanchard,O.,2006,EuropeanUnemployment:Theevolutionoffactsandideas,EconomicPolicy21,559. Blanchard,O.,andJ.Tirole,2008,Thejointdesignofunemploymentinsuranceandemployment protection:Afirstpass,JournalofEuropeanEconomicAssociation,6,4577. 26

Boeri,T.,CondeRuiz,J.I.,andV.Gallaso,2006,ThePoliticalEconomyofFlexicurity,WorkingsPaper2006 15FEDEA. Brown,A.,andC.Merkl,andD.Snower,2009,IncentivesandComplementaritiesinFlexicurity,WP1526 KielInstitutefortheWorldEconomy. Card,D.,J.KluveandA.Weber,2010,ActiveLabourMarketPolicies:AMetaAnalysis,EconomicJournal 120:F452F477. Christensen,R.N.,Jacobsen,R.H,(2009).Analyserafeffektenafaktiveringogvoksenogefteruddannelse forforsikredeledige.CEBR. Cunt,A.,andM.Melitz,2007,Volatility,labourmarketflexibilityandthepatternofcomparative advantage,NBERWorkingPaper13062. Davoine,T.,andC.Keuschinigg,2010,FlexicurityandJobReallocation,WorkingPaper,UniversityofSt. Gallen. DanishEconomicCouncil(2007).Danskkonomi,forr2007. Duval,R.,J.Elmeskov,andL.Vogel,2006,StructuralPoliciesandEconomicResiliencetoShocks,OECD WorkingPaper. Eurobarometer,2010,March. EuropeanCommission,2009,Sustainabilityreport2009,EuropeanEconomy9,Bruxelles. GraversenB.K.,Damgaard,B.ogRosendahl,A.(2007).Hurtigtigangevalueringafetforsgmed entidligogintensivbeskftigelsesindsatsforforsikredeledige.SFIrapport07:10. IMF,2010,WorldEconomicOutlook,Washington Jespersen,S.,JacobR.MunchogSkipper,L.(2008).CostsandBenefitsofDanishActiveLabourMarket Programmes,LabourEconomics15,pp.859884. Jespersen,S.,Jacobsen,R.H.andBge,M.(2006).Evalueringafaktiveringsindsatsenforforsikredeledige medsrligfokuspuddannelsesaktivering.CEBR. Jrgensen,H.,andP.KongshjMadsen,2007,FlexicurityandBeyondFindinganewagendaforthe EuropeanSocialModel,DJFPublishingCopenhagen. Kluve,J.,2006TheEffectivenessofEuropeanActiveLaborMarketPolicy,IZADiscussionPaper,2018. Lindbeck,A.,1995,HazardousWelfareStateDynamics,AmericanEconomicReview,Papersand Proceedings,May,1995,915. Lindbeck,A.,andS.NybergandJ.W.Weibull,2003,SocialNormsWelfareStateDynamics,Journalofthe EuropeanEconomicAssociation,1,533542.

27

Ljungqvist,L.,andT.J.Sargent,1995,TheSwedishunemploymentexperience,EuropeanEconomicReview, 39,104370. Ljungqvist,L.,andT.J.Sargent,1998,TheEuropeanUnemploymentDilemma,JournalofPoliticalEconomy, 106(3),514550. Lommerud,K.E.,andO.R.Straume,2010,Employmentprotectionvs.flexicurity:Ontechnologyadoptionin unionizedfirms,CEPRworkingpaper7919. Nickell,S.J.,1986,Dynamicmodelsoflabourdemand,HandbookofLabourEconomics,Vol2,O. AshenfelterandD.Layard,NorthHolland. OECD,2009,EmploymentOutlook,Paris. OECD,2010,EmploymentOutlook,Paris. OECD,2010b,Labourmarketsandthecrisis,OECDEconomicsDepartmentWorkingPapers,756,Paris RosholmM.(2008).ExperimentalEvidenceontheNatureoftheDanishEmploymentMiracle. Rosholm,M.andSvarer,M.(2008).EstimatingtheThreatEffectofActiveLabourMarketProgrammes, ScandinavianJournalofEconomics,110(2),385401. Rosholm,M.andM.Svarer,M.(2009),Kvantitativanalyseafhurtigigang2,Arbejdsmarkedsstyrelsen. StrbyJensenC.,2008,EmploymentrelationsinDenmarkExplainingflexicurity,Workingpaper, DepartmentofSociology,UniversityofCopenhagen. VanderNoord,P.,2000,Thesizeandroleofautomaticstabilizersinthe1990sandbeyond,OECDworking paper230. VanderNoord,P.,N.Girouard,andC.Andr,2006,SocialSafetyNetsandStructuralAdjustment,OECD WorkingPaper517.

28