Policy Research Working Paper 8497

The Economic Effects of Refugee Return and Policy Implications

Uri Dadush

Fragility, Conflict and Violence Global ThemeJune 2018

WPS8497P

ublic

Dis

clos

ure

Aut

horiz

edP

ublic

Dis

clos

ure

Aut

horiz

edP

ublic

Dis

clos

ure

Aut

horiz

edP

ublic

Dis

clos

ure

Aut

horiz

ed

Produced by the Research Support Team

Abstract

The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent.

Policy Research Working Paper 8497

This paper is a product of the Fragility, Conflict and Violence Global Theme, with the financial support of the German Federal Ministry for Economic Cooperation and Development and in coordination with the German Institute for International and Security Affairs (SWP). It is part of a larger effort by the World Bank to provide open access to its research and make a contribution to development policy discussions around the world. Policy Research Working Papers are also posted on the Web at http://www.worldbank.org/research. The author may be contacted at udadush@worldbank.org.

The recent surge in the number of forcibly displaced who cross international borders in search of protection has prompted interest in evaluating policies that achieve the possible “end points” of the phenomenon. As envisaged by United Nations High Commissioner for Refugees (UNHCR), these are the integration in the country of des-tination, relocation in a third country, and return to the country of origin. The focus of this paper is on the third aspect, namely the appropriate conduct of return policy viewed from the perspective of the host country. More spe-cifically, the main question is whether it is in the economic self-interest of host countries to return forcibly displaced persons. In addressing the question, four ancillary issues are to be addressed: (i) the macroeconomic impact of refugees

and of their return; (ii) the labor market impact of refugees and of their return, (iii) the fiscal impact of refugees and of their return; and (iv) how return policy should be for-mulated and executed. The available evidence and analyses allow this paper’s main conclusion, namely that the costs of hosting asylum seekers and refugees are front-loaded, while the benefits accruing from their integration into the labor market and the host economy typically take years to materialize. It follows that from the economic perspective their return after a short stay may represent a costlier option than continuing to invest in their successful integration. Countries with a flexible labor market, strong investment climate, and a welcoming attitude to immigrants tend to see the economic benefits of refugee inflows materialize faster.

TheEconomicEffectsofRefugeeReturnandPolicyImplicationsUriDadush1

SeniorFellow,OCPPolicyCenterandNon‐ResidentScholar,Bruegel

JELclassifications:C8,F22,O15,O19,R23,F51Keywords:forceddisplacement,refugees,asylum‐seekers,IDPs,returnees,affectedhostcommunities,UNHCR,IDMC,internationalmigration

1ThispaperwascommissionedbytheWorldBank’sFragility,ConflictandViolenceDepartmentDisplacementwiththefinancialsupportoftheGermanFederalMinistryforEconomicCooperationandDevelopmentandincoordination with the German Institute for International and Security Affairs (SWP). I am grateful forcomprehensive and very useful comments received from Xavier Devictor, Mona Niebuhr, and the peerreviewers,NailaAhmed,DavidKipp,andHansTimmer.CarolineBahnson,RebekahSmith,andPaoloVermealsoprovidedcommentsonpreviousdrafts.BingyingWuprovidedexcellentresearchassistance.

2

The recent surge in the number of forcibly displaced persons2 who cross internationalbordersinsearchofprotectionhaspromptedinterestinevaluatingpoliciesthatachievethepossible “end points” of the phenomenon. As envisaged by United Nations HighCommissioner forRefugees (UNHCR3), theseare the integrationof the forciblydisplacedpersonsinthecountryofdestination,relocationinathirdcountry,andreturntothecountryoforigin.Thefocusofthispaperisonthethirdaspect,namelytheappropriateconductofreturnpolicyviewedfromtheperspectiveofthehostcountry.The importance of the return issue is underscored by a recent report of the EuropeanCommission.AccordingtothelatestEUActionPlanonReturn(2017),during2014‐2015justover1millionirregularmigrantswereorderedtoleavetheEuropeanUnion.In2015/2016,theEUreceivedaround2.6millionasylumapplicationsforrefugeestatus4.ofwhichlessthan60%wereacceptedinthefirstinstance,implyingthatperhapsasmanyas1millionasylumseekers could be ordered to return. According to the latestWorld Bankmigration brief(2017) in the EU, the number of potential returnees – those denied asylum and otherdetectedundocumentedmigrants–rosefrom1.4millionin2011toover5millionin2016.IntheUnitedStates,thestockofpotentialreturneesrosefromaround1.5millionin2011to3millionin2016.Atthesametime,actualreturnratesremainwellbelowmandatedreturn.Forexample,lessthan40%offailedasylumseekersorderedtoreturnactuallyreturnedinrecentyears.Ofcourse,returnpolicyisnotlimitedtotheforcedreturnoffailedasylumseekers,asasylumseekersandrefugeessometimesreturnvoluntarily,andareroutinelyencouragedtodoso.Countriesalsosometimesaccordprovisionalasylumstatus,whichcanbewithdrawnwhenreturntothecountryoforiginisdeemedtobesafeagain.Thoughnotthemainobjectofthis

2“Non‐refoulement”,anormagreedinArticle33ofthe1951UNRefugeeConventionrequiresthat“nocontractingstateshallexpelorreturnarefugeeinanymannerwhatsoevertothefrontiersofterritorieswherehislifeorfreedomwouldbethreatened.”AsstatedintheWorldBank’s“ForciblyDisplaced”2017report,“the1951ConventionRelatingtotheStatusofRefugeesdefinesrefugeesintermsoftheirvulnerabilityduetothedenialorlossofstateprotection.‘Arefugeeisapersonwhoisoutside his or her country of nationality or habitual residence; has awell‐founded fear of beingpersecutedbecauseofhisorherrace,religion,nationality,membershipofaparticularsocialgroup,orpoliticalopinion;andisunableorunwillingtoavailhimorherselfoftheprotectionofthatcountry,ortoreturnthere,forfearofpersecution.Refugeestatusisnotaccordedtothosewhofleeintolerableeconomicconditions,poverty,famineorenvironmentalcatastrophes(unlesstheseareconsequencesoforrelatedtopersecution)”Refugeesincludeindividualsrecognizedunderthe1951ConventionrelatingtotheStatusofRefugees;its1967Protocol;the1969OAUConventionGoverningtheSpecificAspects of Refugee Problems inAfrica; those recognized in accordancewith theUNHCR Statute;individuals granted complementary formsofprotection; or thoseenjoying temporaryprotection.Since2007,therefugeepopulationalsoincludespeopleinarefugee‐likesituation.Asylum‐seekersareindividualswhohavesoughtinternationalprotectionandwhoseclaimsforrefugeestatushavenotyetbeendetermined, irrespectiveofwhen theymayhavebeen lodged.UNHCRalsorefers tointernallydisplacedpeople–i.e.peopleforcedtoleavetheirhomesbutremainintheircountryoforigin‐asforciblydisplaced:theyarenot,however,theobjectofthispaper.3SeeUNHCR2016foranoverviewofrecentdisplacementtrends4SeeAnnex1fordefinitions

3

paper,returnpolicyalsodealswithlargenumbersofimmigrantswhoarriveirregularlyforeconomicorfamilyreunificationreasonsanddonotapplyforasylum,andextendstothereturnofeconomicmigrantswhoarrivelegallybutwhosevisasexpireorwhoseservicesarenolongerneeded.Over and beyond the fraught politics, appropriate policies towards asylum seekers andrefugees–ofwhichreturnpolicyispart–entailthreetypesofconsideration:humanitarian,legal,andeconomic.Thispaperwill focusmainlyon theeconomic implicationsof returnwhilerecognizing that the legalandhumanitarianarean integralpartofdecisions.Morespecifically,themainquestionIaddressiswhetheritisintheeconomicself‐interestofhostcountriestoreturnforciblydisplacedpersons.The vast majority of refugees end up in developing countries such as Lebanon, Jordan,PakistanandTurkey,asubjecttreatedinrecentWorldBankreports(Devictor2016;Vermeetal2015).Thispaperinsteadaimstoaddressthereturnpolicyofadvancedcountriesthathaveseenasurgeinthearrivalofasylumseekers,suchasAustria,Germany,Greece,Italy,andSweden.WhiletheUnitedStatesisalsoahosttomanyforcedmigrants,themajorityareresettled, i.e. theirasylumapplicationsaredecideduponwhile theyare in thecountryoforiginorinathirdcountry,sotheissueofreturningfailedasylumseekersislesspressing.IntheUnitedStates,theissueofreturnisjustaspoliticallyprevalent,butitrelatesmainlytoirregulareconomicmigrantswhoinfiltratetheSouthernborderoroverstaytheirvisas5,andeventotheU.S.‐bornchildrenofthoseimmigrants,butnottoasylumseekersorrefugees.The weaknesses of data on the return of immigrants are well known: “If the data oninternationalmigrationaregenerallypoor,therecordingofreturnmigrationisfarworse”(Lucas 2005). The scarcity of analytical examinations of the forced displacementphenomenonisalsowellknown.Forexample,arecentIMFreporttitled“TheRefugeeSurgeinEurope:EconomicChallenges”containssome65bibliographicreferences,ofwhichnearlyalldealwithimmigrationgenerallyandonlyaboutfivedealmainlywithrefugees.(Aiyaretal.2016).Despitethese lacunas, Ibelievethat theavailableevidenceandanalysesassetoutbelowallow thispaper’smain conclusion, namely that the costs of hosting asylum seekers andrefugeesarefront‐loaded,whilethebenefitsaccruingfromtheirintegrationintothelabormarketandthehosteconomytypically takeyearstomaterialize. It followsthat fromthe

5Tight immigration restrictions exist both across the EU and in the United States, and they areextremeinthecaseofunskilledworkers.InboththeUSandEUtheresulthasbeenlargenumbersofundocumentedimmigrants.Forexample,theIOMestimatedthat3millionillegalimmigrantsresidedintheEUin2000,doublethenumberoftenyearsearlier,representingabout¼oftheofficialnon‐EUpopulation.TheINSestimatedthat7millionillegalimmigrantsresidedintheUSin2001.IntheUS,approximatelyasmanyillegalworkersenteredtheUSasdidunderofficialtemporaryworkvisas(Lucas, 2005). In the US, while efforts have been directed at attracting skilled immigrants anddenyingofficialentrytounskilledimmigrants, theeffecthasbeenthatunskilledimmigrantshavecontinuedtorepresentahighproportion,butillegally.

4

economicperspectivetheirreturnafterashortstaymayrepresentacostlieroptionthancontinuingtoinvestintheirsuccessfulintegration.Countrieswithaflexiblelabormarket,stronginvestmentclimate,andawelcomingattitudetoimmigrantstendtoseetheeconomicbenefitsofrefugeeinflowsmaterializefaster.Thepaperalsodevelopsthefollowingmessages:

Theeconomicanddemographicsituationacrossadvancedcountries6varies,asdothenumberandattributesof asylumseekerswho reach their shores.While somegeneralizationsarepossible,thereisno“onesizefitsall”returnpolicy,justasthereisnosingleoptimalimmigrationpolicy.However,allnationsaredefinedbywholiveswithin them, their lawsandcontrolofborders. So, there isa strong legal case forrejectingasylumapplicationsthatareclearlyspecious.Thisisnecessarynotonlyforretainingcontrolofimmigrationpolicybutalsotopreservetheintegrityandpoliticalviabilityoftheasylumsystem.

Inmanyadvancedcountriestodaythepurelyeconomiccaseforreturningmigrants,as distinct from the legal case – whether they are asylum seekers, failed asylumseekers,refugees,orirregulareconomicmigrants–isweak.Thisconclusionisdrawnfromacarefulexaminationofthemacroeconomic,fiscalandlabormarketimpactsontheadvancedhostcountryoftheinflowofimmigrants.Thebottomlineisthattheaggregate economic effects of migration, including those of refugees, in mostinstancesarebeneficial.

Morespecifically,theoverallfiscalimpactofrefugeeflowsislikelytobesmall.Itisnegativeatfirstduetothehighcostofrelocatingrefugeesandhostingthemuntiltheyfindjobsandlearnthelanguage.Sincemostrefugeesareyoung,studiessuggestthattheirfiscalimpactbecomespositiveonacash‐flowbasisafewyearsafterarrivalandpositiveonapresentvaluebasisafewyearsafterarrival.Becauseasylumseekersareoftenleftinlimboforprotractedperiods,andthepermissiontoworkisdelayedordenied, the fiscal costofhosting them ishigherand their contribution to thehosteconomyissmallerthanitcouldbe.

Mostrefugeesarelow‐skilled.Thebenefitofrefugeearrivalontheiremployers,onfirmsgenerally,onthepriceofmanynon‐tradableproductsandservices,andonthevastmajorityofnativeworkerswhodonotcompetedirectlywiththemandwhoarecomplementary to them in production, are likely to be felt at about the pace thatrefugeesareallowedtoandfindwork.Immigrationgenerally,includingofrefugees,tendstobesupportiveoflong‐termeconomicgrowth.Thedampeningeffectonthewages of low‐skilled nativeworkers of refugee arrival is likely to be small. Somecrowding out of public services may occur in refugees concentrated localities.However,thisislesslikelyincontextswherethepopulationisdecliningorhassloweddownmarkedly.

6Thesearethe“high‐incomecountries”asidentifiedbytheWorldBankusinganincomecriterion.

5

The economic effects of refugee return are only a part of the economic effects of

immigration, whether for work, family or protection reasons. While there areimportant considerations that distinguish forcibly displaced persons from others,policiesregardingthereturnofrefugeesmustbeframedwithbroaderpolicytowardsimmigrationinmind.

Inlightoftheobligationtohostrefugees,returnpoliciesmustalsorecognizethat,as

shownbelow,fewasylumseekersorrefugeesreturnvoluntarilyfromanadvancedcountrytoadevelopingcountry.Whenviewedfromthestandpointof the forciblydisplacedpersons,inmostinstancesthefinancialcaseforreturnisveryweakindeed,asthedisparityinexpectedearningbetweenstayingandreturningisbig,evenwhenthecountryoforiginissafe.Itfollowsthattrulyvoluntaryreturn(asdistinctfromwhatsomemightcall“softdeportation”)canonlybeexpectedinspecificinstancesorwherepersonalconsiderations,suchasfamilytiestothehomecountry,prevail.Theavailable literaturesuggests thatmigrants thatretainstrong familyandemotionalconnectionswiththeircountryoforiginandthathavebuiltupsavingsordevelopedneeded skills during their stay abroad are the most likely to return voluntarily.Financialconsiderationsalonewillrarelyplayadeterminantroleinspurringreturn.

Fewrefugeesreturnalsobecauseconflictsappeartohavebecomemoreprotracted

andmost, thoughnotall, countries thatare theoriginof largenumbersofasylumseekersarepatentlyunsafeandalsoincapableofaffordingthereturningmigrantalivelihood.Returnunderunfavorableconditionsmayonlyresultinrenewedattemptsatemigration.

Theexpectation thathumanitarianassistanceordevelopment tools (grants, loans,

policy advice) can play a significant role in promoting the return to countries inconflict is unrealistic. Countries of origin that are unsafe today, however, couldbecomesafeinthefuture,asconflictsareresolved.Inthatcase,thereturnflowofforciblydisplacedpersonsispossible,eventhough–inmostinstances–voluntaryreturnfromadvancedcountriestoapoorlocationisunlikelytobelarge.Forexample,SyrianrefugeeshousedprecariouslyinLebanonaremorelikelytoreturnthanthosebased in Sweden. Development policy can help returnees resettle successfully,acceleratereconstruction,andreducethelikelihoodofrecurrenceofhumanitariancrises.Returningrefugeescanplayarole–evenifitislikelytobeamodestone‐inthereconstructionanddevelopmenteffortintherightcircumstances,andalsobuildbridgesbetweenthecountryoforiginandthehost,creatingbusinessopportunitiesforboth.

Thepoliticaldebateoverwelcomingrefugeesisfraught,andactualpolicydecisionswill,inthefinalanalysis,havetonavigatethesedifficultwaters.However,ourpurposehereistoaddresstheissues,insofaraspossible,fromatechnicalstandpoint–i.e.tohelpinformthepoliticaldebatewithfactsandanalysisratherthanengaginginit.

6

Thereisaclearandcruciallegaldistinctionbetweenasylumseekersandrefugees.Onceanasylumseeker’scasehasbeenapprovedandhisstatusasarefugeehasbeenrecognized,thereisalegalobligationtohosthimorher.(seeFootnote2andAnnex1).Forthisreason,thispapergoestosomelengthtoclarifywhethertheunitofanalysisisanasylumseekerorarefugeewheneverthereisthepossibilityofconfusion.Inaddressingthequestionofwhetheritisintheeconomicself‐interestofthehostcountrytoencouragerefugeereturn,fourancillaryissuesaretobeaddressed:

‐ Themacroeconomicimpactofrefugeesandoftheirreturn‐ Thelabormarketimpactofrefugeesandoftheirreturn‐ Thefiscalimpactofrefugeesandoftheirreturn‐ Howreturnpolicyshouldbeformulatedandexecuted

Beforediscussingtheseissues,itisusefultoreviewtheevidenceontheprevalenceofreturn.PrevalenceofReturnReturnisawell‐establishedfeatureofmigration,andso isthefactthatreturnratesvarygreatly depending on country of origin anddestination.However, truly voluntary returnfromarichcountrytoapooroneisararephenomenon7.Evenwhenturneddownforasylumandofficiallyexpectedtoleave,mostasylumseekersdonotinfactdoso.(seebelow)Article3ofthe1951Conventionprohibitsexpulsion,return,orrefoulementofpersonstocountrieswherethereisasubstantialriskthattheywillfacetorture,inhumanordegradingtreatment.Thevoluntaryreturnofrefugeesisevenlesscommon.AccordingtoUNHCR,therewere19.5millionrefugees in2014andonly126,800returned to their countryoforigin.ThesearepredominantlyrefugeesthatfoundprecariousasyluminAfghanistan,SudanandIraqandreturnedtoneighboringcountriesoforigin.Thenumbersofreturneessawlittleincreasein2015 and approached 500,000 in 2016, mainly due to returnees to Afghanistan underpressuretoleaveneighboringPakistan8.

7Returnofmigrants fromamoredevelopedtoa lessdevelopedcountry isnotalways low.Manydevelopingcountrieshaveestablishedmodestprogramstoattractbackstudentswhogoabroadandskilledemigrants,buttheyhavehadlimitedsuccess.AninterestingexceptionisChina,asafeandlargemiddle‐incomeeconomy(percapitaincomehigherthan$13,000,PPPadjustedin2015)thathasexperiencedtheworld’shighestsustainedgrowthrateinrecentdecades.AccordingtothemostrecentreportoftheChineseMinistryofEducationabout80%ofChinesestudentsabroadreturnedtoChinain2016.8AccordingtoCrispandLong(2016)“thelast10yearsarebestdescribedas“adecadeofprotractedemergencies.”MillionsofnewrefugeeshavebeencreatedasaresultofintenseviolenceinBurundi,CentralAfricanRepublic,Iraq,Nigeria,SouthSudan,Syria,Ukraine,andYemen.Atthesametime,longstandingconflictsincountriessuchasAfghanistan,theDemocraticRepublicofCongo,Myanmar,Somalia,andSudanhavegoneunresolved.Asaresultofthesedevelopments,refugeenumbershavejumpedtoanall‐timehigh,whilerepatriationlevelshavedroppedtoanhistoriclow.”

7

ArecentWorldBankreport“SustainableRefugeeReturn”(2015)examined8casestudiesof large‐scale refugee return, namely return to Afghanistan, Angola, Bosnia‐Herzegovina,Burundi, Cambodia, Iraq, Liberia, and South Sudan. In all these cases, except Bosnia‐Herzegovina, refugees were hosted mainly in nearby developing countries – often in aprecariousconditionandunderpressuretoleave.Approximately350,000Bosnianrefugeesfound shelter in Germany, however, they were never granted permanent status.Furthermore,inthewakeoftheDaytonPeaceAccords,theywereunderstrongpressuretoleave, being denied of rights and support. Some 250,000 did so, returning to Bosnia orresettlinginthirdcountries9.Thesestatisticsonthereturnofrefugeescontraststronglywiththoserelatingtoeconomicmigrantsamongrelativelyadvancedcountries.Itisestimated,forexample,thatabout30%ofmigrantstotheUnitedStatesduringthegreatageofmigration1890‐1914,returned,andthat return rates varied from around 5% of those originating from Russia, Ireland, andScandinavia,toabout50%ofthoseoriginatingfromItaly.Post‐WorldWarII,guestworkerprogramshelpedtofacilitatelarge‐scalemigrationwithinEuropeandfromcountriessuchasMorocco,Turkey,andTunisia.Theseprogramscollapsedinthewakeofthefirstoilshockin 1973‐75, and return/circularmigration has been a prominent feature in Europe eversince.Forexample, in1975‐1985theoutwardflowof the foreignpopulation inGermanyexceededtheinflow.Insubsequentyears,netmigrationresumed,andcircularmigrationdidnotstopthemigrantstockfromincreasing,astheforeignpopulationinGermanyincreasedfromlessthan1millioninWestGermanyin1960to3millionin1970,4.5millionin1980,to7.5millioninunifiedGermanyin2000.UnliketheUnitedStates,Europeneverencouragedpermanentsettlement.Over1990‐2003,theforeignpopulationoutflowtoinflowratiowas73%inGermany.InBelgium,theratiowas49%,inDenmarkandSwedenitwas25%andintheNetherlands,itwas19%.Amongthemost accurate estimatesof return are from social registers available forFinlandandSweden,which enable a longitudinalmonitoring of individuals and a reliable sorting bycountryoforiginandbythenatureofmigration.Some30%ofthe1990immigrantcohortinFinlandhadleftwithin10years.While60%ofthosewhosecountryoforiginwastheOECDreturned,only20%returnedtoEstoniaandonly15%ofthoseoriginatingfromRussia.Only15%ofthosewhocameinasrefugees,typicallylow‐incomecountries,returned.AsimilarpictureisobtainedinSweden.Over1970‐1990,25%oftheimmigrantcohorthadleftwithin5years,withthereturnmorecommontoNordiccountriesorotherOECDcountriesandleastcommonamongrefugeesandimmigrantsfromlow‐incomecountries.The data shows that return is more likely the shorter the distance, the shorter thepermanenceandthewealthieristhecountryoforigin.Financialincentivesforreturnappeartohaveplayedonlyaminorrole(Dustmann1996).Basedonasurveyof273returneesto8origincountries,KuschminderandKoserconcluded:“returneeswhomigratedforeconomicreasonsweremore likely tobe reintegrated than thosewhomigrated forother reasons;returneeswho both had a sense of belonging to the community prior tomigration and9 http://www.dw.com/en/refugees‐reloaded‐lessons‐from‐germanys‐approach‐to‐bosnian‐war/a‐19021249

8

returned to the same community after migration were more likely to be reintegrated”(KuschminderandKoser,2015)ThefactthatveryfewasylumseekersandrefugeesreturnhometoadevelopingcountryisconfirmedbythemostrecentUNHCRandEurostatstatistics–imperfectandincompleteastheyare.(Table1and2) Table 1:Refugees,Asylum‐Seekers,And ReturnRefugees by Country/Territory ofAsylum,in2016(Adash ("‐") indicates that thevalue is zero,notavailableornotapplicable.Alldataareprovisionalandsubjecttochange.)Country/territory ofAsylum

Refugees Asylumseekers(pendingcases)

Returnedrefugees(tocountryoforigin,during2016)

Turkey 2,869,421 245,955 ‐Pakistan 1,352,560 4,856 7Lebanon 1,012,969 13,745 ‐Iran 979,435 91 11Uganda 940,835 41,880 1,192Ethiopia 791,631 1,964 1Jordan 685,197 35,615 ‐Germany 669,482 587,346 ‐Dem.Rep.oftheCongo 451,956 1,327 13,223Kenya 451,099 43,764 4Sudan 421,466 16,052 37,215Chad 391,251 1,909 28China 317,255 668 ‐France 304,546 62,771 ‐Tanzania 281,498 8,539 6UnitedStatesofAmerica 272,959 542,649 ‐Yemen 269,783 9,097 1SouthSudan 262,560 1,792 1Iraq 261,888 11,458 157RussianFederation 228,990 3,039 56Egypt 213,530 49,877 ‐India 197,851 9,219 ‐

9

Afghanistan 59,711 128 383,951Source:UNHCR2016DataonreturnfromEUcountriescompiledbyEurostatismorecomprehensivebutdoesnotdistinguishbetweenthereturnofeconomicmigrants–whichisplausiblythelargernumber‐andrefugees,andthereforerepresentatmostanupperboundonthereturnofrefugees.Even so, the number of returnees – whether forced or voluntary – is small except forGermanyandtheUnitedKingdomwhichhousealargepopulationofeconomicmigrants,and,inthecaseofGermany,oflargenumbersofasylumseekersfromcountriesintheBalkans.Table2:TypesofReturnfromEUMemberStatesin2015

Personsreturned

Voluntaryreturn

Enforcedreturn

Type ofreturnunknown

Germany 55,340 ‐ ‐ 55,340UK 50,590 ‐ ‐ 50,590France 18,245 5,920 12,325 0Greece 14,390 ‐ ‐ 14,390Spain 13,315 2,355 10,960 0Poland 12,930 12,080 850 0Sweden 9,830 7,285 2,545 0Netherlands 8,620 ‐ ‐ 8,620Hungary 5,975 210 5,765 0Belgium 5,835 3,310 2,525 0Austria 5,275 ‐ ‐ 5,275Italy 4,670 1,015 3,655 0Finland 3,365 ‐ ‐ 3,365Denmark 2,655 170 2,480 0Romania 1,995 1,810 180 0Croatia 1,940 1,250 690 0Cyprus 1,840 ‐ ‐ 1,840Lithuania 1,720 ‐ ‐ 1,720CzechRep. 1,715 ‐ ‐ 1,715Slovakia 1,230 670 560 0Latvia 1,030 695 340 0Slovenia 840 90 110 640Bulgaria 735 180 555 0Luxembourg 720 545 175 0Portugal 610 240 370 0Estonia 560 475 85 0Malta 465 285 180 0Ireland 365 115 250 0

10

TOTAL 226,800 38,700 44,600 143,495Source:EuroStat,citedbyESI2017

Dataontheactualreturnofthosewhoareorderedto leave is incomplete.Wheredata isavailable, the indication is thatwhilemany asylum applications are rejected, only smallnumbers of thosewith orders to leave actually leave the country. For instance, in 2011,Irelandissuedintotal940orderstoreturnduetofailedasylumapplications,butonly381returned,ofwhom152benefitedfromAssistedVoluntaryReturn(AVR)and52returnedvoluntarilywithoutAVR10.Startingin2011,Germanymadeincreasinguseofdeportation,fromlessthan8,000in2011toover20,000in2015,butthereasonfordeportationisnotspecified. In2015,35,514voluntary returnswereapproved for funding, butnearly¾ofthesewerenationalsoftheBalkans:Albania(about32%),Kosovo(about23%),andSerbia(about18%)11,andrelativelyfewwerefromdevelopingcountriesoutsideofEurope.

Asylum seekers resort to multiple strategies to avoid the return, including “procedureshopping”, defined as “the simultaneous, parallel or successive introduction of severalsimilar or different procedures (asylum, humanitarian regularization and medicalregularization)”.12AsreportedbasedonarecentEuropeanMigrationNetworkconference13,there are many challenges confronting the return of rejected asylum seekers: “publicresistance to return and political pressurenot to implement removals; strong individualresistancetoreturn;greaterdifficultiesinobtainingtraveldocuments,compoundedbythefact that asylum seekers are more frequently undocumented than other third‐countrynationals; and greater prevalence ofmedical cases among rejected asylum seekers thanamongotherreturnees.”Additionally,therearechallengesintermsofthesecuritysituationin countries of origin, the possibility of lodging late‐stage appeals and judicial reviews,combinedwithdifficultyinestablishingcontactwiththeauthoritiesofcountriesoforiginbefore theasylumprocedure isclosed.Theactual returnofasylum‐seekers iscontingentuponallthosefactorsmentionedabove.

10 Returning Rejected Asylum Seekers: Challenges and Good Practices in Ireland, July 2017https://www.esri.ie/pubs/RS65.pdf11 Migration, Integration, Asylum‐ Political Developments in Germany 2015https://ec.europa.eu/home‐affairs/sites/homeaffairs/files/what‐we‐do/networks/european_migration_network/reports/docs/annual‐policy/annual‐policy‐11a_germany_apr_part2_en.pdf12Anin‐housestudyoftheBelgianImmigrationOfficeinvolvedexaminationof86,000peoplewhoon average introduced an asylum application 1.55 times between 2010 to 2012. Taking allapplicationsofthesepersonsintoconsiderations,onaverageeachpersonapplied3.48times.45.5%ofpeopleapplyingforregularizationonhumanitarianormedicalgroundsin2012hadpreviouslylodgedanasylumapplication.1313 http://www.bamf.de/SharedDocs/Anlagen/EN/Publikationen/EMN/emn‐tagungsband‐rueckkehr‐2017.pdf;jsessionid=F32B96A388AD9CA9DEF211C9162499B8.2_cid294?__blob=publicationFile

11

Thecostofimplementingforcedreturnisoftenoverlooked.AnEUobserver14probeofsome100jointreturnflightscoordinatedbytheEU’sborderagencyFrontex,calculatedthatonaverage,itcosts€5,800todeportoneindividual.Thepricedependsonthedestinationoftheflight, its route, and the number of escort personnel needed. For instance, deporting toAlbaniafromGermanymaycostaround€1,000,whileforcedreturnstoNigeriamaycostupto €9,000. Unexpected circumstances may “lead to last‐minute changes like returneesabsconding,changedmedicalconditions,newasylumclaimsorotherlegalchallenges15.”Inextremecases,costscanspiraloutofcontrol,asinacasewhereSwissauthoritiesdecidedtorerouteaflightboardedbymigrantsfromTogo,spendingcloseto€90,000foreachpersonreturnedtoTogo.AreportbytheAmericanImmigrationCouncil(2013)estimatedthattheUnitedStatesspendsnearly$2billionayearonimmigrationdetention16andthatthedirectfiscalcostofremovingsome9millionillegalimmigrantscouldstretchintothehundredsofbilliondollars,notaccountingfortheadverseimpactonthebroadereconomy.

Contrarytocommonbelief,thegrantingofrefugeestatusisnotirreversible.AccordingtoArticle1Cofthe1951ConventionrelatingtotheStatusofRefugees,refugeestatusshallceasetoapplytoapersonundercertainconditions(“theCessationClause”),includingifheorshehasvoluntarilyre‐availedhimselfoftheprotectionofthecountryofhisnationalityorif“Hecannolonger,becausethecircumstancesinconnectionwithwhichhehasbeenrecognisedasarefugeehaveceasedtoexist,continuetorefusetoavailhimselfoftheprotectionofthecountry of his nationality…”. Thus, many developing country hosts have treated theacceptanceofrefugeesasatemporaryphenomenonpendingtheendofconflicts,andthisisalsoformallythepositionofcountriessuchasDenmark,GermanyandtheUnitedKingdomtoday.

Butisreturnbeneficialinthefirstplace–morespecifically,isitintheeconomicself‐interestofthehostcountry?Toaddressthisquestion,weneedtoconsiderthecounterfactual,whichistheeffectofintegratingrefugees.Theseissuesareconsideredinthenextthreesections.

TheEffectofRefugees(andofReturn)ontheMacro‐Economy17Itshouldbestatedattheoutsetthattheeconomicimplicationsofintegratingrefugeesarenot unconditional. They depend crucially on policies that the host countries pursue,especiallywhetherandhowsoonhost countriesallowrefugees toworkand thesupportgiven to them to enable them to find jobs. The arrival of large numbers of refugees canstimulatedemandintheshorttermandcontributetoanaccelerationoflong‐termeconomicgrowth in the right circumstances. Even when policies in the host country are strongly1415EUObserver,SkyrocketingcostsforreturningEUmigrants,https://euobserver.com/migration/137720https://euobserver.com/migration/13772016AmericanImmigrationCouncil,TheCostofImmigrationEnforcementandBorderSecurityhttps://www.americanimmigrationcouncil.org/research/the‐cost‐of‐immigration‐enforcement‐and‐border‐security17PartsofthissectiondrawonDadushandNiebuhr(2016)andDadush(2014)

12

supportiveoftheirrapid integrationin labormarkets,refugeestake longertomaketheircontributionfeltthaneconomicmigrants.Theircosttotheeconomy(transfersfromnatives)isfront‐loaded,whiletheircontributiontoGDP(valueadded)isback‐loaded.It is evident that the economic effect on thehost andon the origin countryof returningrefugeescannotbeseensimplyasthemirrorimageoftheirfirstdepartureandarrival.Notonlyarethecoststhatthehostcountryincursinabsorbingrefugeesheavilyfront‐loaded.Thetypicalestimateisthatitcostsabout$10,000tosupportarefugeeinthefirstyear.ThesecostsarehigherinEuropewherethevettingprocesstendstobedoneinlocothanintheUnitedStates,whererefugeesarrivehavingbeencleared inthirdcountriesandresettled(Legrain2016).Insubsequentyears,thenetcostofhostingrefugeesdependscruciallyontheirabilitytowork.Refugeesthemselvesareprofoundlytransformedbytheirexperience.Thecountrytheyreturntoisoftenverydifferentfromthecountrytheyleft.Nevertheless,ifone is toanalyze theeffectofreturn,onemustcompare it tothecounterfactual,which–exceptininstanceswheretherefugeeisresettledinathirdcountry–isthatheorshestaysandbecomesintegratedinthehosteconomy.Refugeesareconcentratedindevelopingcountries(seebelow),andrelativelyfewsettleinadvancedcountries.Lebanon,Jordan,andTurkeyhaveseencumulativenetrefugeeinflowsthatexceed16%,7%and3.6%oftheirpopulationrespectively.Althoughsomeadvancedcountrieshaveseencomparablylargeinflowsofrefugeesinthepast,andcopedsuccessfullywiththechallenge18,over2015‐2016mostOECDcountriesreceivedsmallannualinflowsofasylumseekersandrefugees,andthetotalstocktypicallyrepresentslessthan0.5%oftheirpopulation.Inmostinstances,inadvancedcountries,themacroeconomicandlabormarketeffectofrefugeesissmall,andthatofreturn,farsmallerstill.Table3:RefugeesHostedAsAShareOfNativePopulation(Tablecontinueonnextpage)Country Refugee

(2016)Percentageofpopulation

Australia 69,497 0.3%Austria 139,761 1.6%Canada 97,311 0.3%

18Germanyabsorbednearly12millionethnicGermanandotherrefugeesinthewakeofWorldWar2;WestGermanysawaverylargeinflowofrefugeesafterthefalloftheBerlinWallinadditiontotheinflowofaround2millionworkers fromEastGermany. Israel saw the inflowofnearly1millionSovietJewsinthewakeofthefalloftheWall,accountingfornearly4%ofthepopulationin1990andover3%in1991.ThecityofMiamisawaninflowofCubanrefugeesamountingtoabout[7%]ofitspopulationoverasix‐monthperiodin1980(seediscussionbelow).Immigrationrates(includingallimmigrants) have been much higher during some periods than they are today. For example,immigrationtoArgentinaadded29%tothepopulationin1901‐1910.Duringthatdecade,Canadareceived immigrants adding 17% to its population, and the United States 10% (Hatton andWilliamson1998).

13

France 304,507 0.5%Germany 669,408 0.8%Greece 46,381 0.4%Italy 147,302 0.2%Japan 2,512 0.0%Portugal 1,129 0.0%Spain 12,943 0.0%Sweden 230,103 2.3%UnitedKingdom 118,913 0.2%United States ofAmerica

272,898 0.1%

Jordan 685,178 7.2%Lebanon 1,012,954 16.9%Turkey 2,869,379 3.6%Source:Source:UNHCR,Refugees,includingrefugeelikesituationsTable4:AsylumSeekersHostedAsAShareOfNativePopulationCountry Asylum

Seekers(2016)

Percentageofpopulation

Australia 29,467 0.12%Austria 76,362 0.87%Canada 23,856 0.07%France 62,719 0.09%Germany 587,283 0.71%Greece 3,993 0.04%Italy 99,876 0.16%Japan 18,801 0.01%Portugal 805 0.01%Spain 20,360 0.04%Sweden 83,010 0.84%UnitedKingdom 43,436 0.07%United States ofAmerica

439,833 0.14%

Jordan 35,570 0.38%Lebanon 13,709 0.23%Turkey 245,896 0.31%Source:Source:UNHCR,Refugees,includingrefugeelikesituations

14

Thereare,nevertheless,goodreasonstoexploretheseeffects,overandbeyondthestandardargument that economic effects must be analyzed at the margin. First, a few advancedcountriesdoreceivelargenumbersofrefugeesinsomeyears.Second,eventhoseadvancedcountries that do not receive many refugees often attract large numbers of economicmigrantsand,overaspanofseveralyears,themacroeconomicandlabormarketeffectsofrefugeesbecomesimilartothoseofothermigrants,andalsoactincombinationwiththoseofothermigrantsincomplexways.Forexample,Chart1showsthat,inSweden,theshareofemployedamongmalerefugeesincreasessharplywiththetimeofstayandis15percentagepointslowerthanthatofworkmigrantsandofnativesevenafter12years.Bycontrast,Chart5showsthattheemploymentrateofrefugeesintheUnitedStatessurpassesthatofnativeswithinjust8years.Chart1:PercentageOfEmployed,ByDurationOfResidenceInSweden,Men,CohortArriving 1997‐99

Source:SwedishPublicEmploymentService(Arbetsformedlingen)Third,whileinflowsofrefugeesaretriggeredbyconflictandpersecutioninthecountryoforigin,andare tosomedegreeexogenousevents in thecountryofdestination19, there ismuchevidencethateconomicmigrationisendogenous,inthesensethateconomicmigrantsrespondtodemandfortheirservicesandtocompetitionfromotherworkers20.So,where

19Refugeesleavethecountryoforigintoescapedangerbutthensometimeshavetheoptiontodecidewheretheymoveto,sorefugeeinflowsarenotentirelyexogenous.20Theendogeneityofmigrationandinparticularitsresponsivenesstodemandinthehostcountrieshas been demonstrated in numerous studies (See Dadush 2016, for a summary) and is a well‐

15

immigration is believed to have positive macroeconomic and labor market effects,promotingthereturnofrefugeesatconsiderablefinancialandhumancostonlytohavethemreplacedbyeconomicmigrantsmaynotmakesense.Policiesrelatingtorefugeesneedtoaccountforthefactthat‐dependingonthehumancapitaltheybringandthattheybuildduringthetimetheyresideinthehostcountry–refugeescanbesubstitutesorcomplementstobothnativeworkersandothermigrants.Thus,whilerefugeepoliciesarehighlyspecific,(they must account for the special legal status of refugees, and for the humanitariandimensionofacceptingrefugees)theeconomicimplicationsofintegratingrefugeescannotbesimplyseparatedfromthoseofothermigrants.Themacroeconomic impact and labormarket impact of refugeeswill only be significantwhere the flowof refugee is large enough –which, in advanced countries, is the case ofSwedenandGermanyinrecentyearsorofIsraelintheearly1990s(arrivalofSovietJews).Inthefirstinstance,thespendingdedicatedtoabsorbingalargeflowofrefugeescanbeasourceofeconomicstimulustodemand.Forexample,recentOECDandIMFreportsestimatethatthedemandincreaserelatedtotherefugeeinflowinEuropewasabout0.1%ofGDPin2016(OECD2015;Aiyaretal.2016).Wheretheeconomyisalreadyatfullemployment,theeffectofspendingonrefugeescouldbeinflationaryatfirstbut–assumingrefugeesareallowedtowork–theyeventuallyexpandtheeconomy’sproductive capacity.While thedemand‐expandingeffectof refugee inflowmay be temporary and depends on its intensity over short periods and the stage of thebusinesscycle,thesupply‐expandingeffectareexpectedtobelong‐lasting,andtocumulateiftherefugeeinflowissustainedovermanyyears.Asrefugeesfindjobs,theyhaveeffectslikethoseofeconomicmigrants,whichistostimulateincreasedcapitalformation.Investmentisneededtoequipthenewworkerswithmachines,as well as to house them. Refugees are likely to induce expansion of sectors such asconstruction and utilities (electricity, water, etc.), which are among the most capitalintensive(See,forexample,UKOfficeforNationalStatistics,2013and2016).Theincreasedcapitalformationcanbefinancedfromincreaseddomesticsavingsorfromincreasedcapitalflowsfromabroad,asintheUnitedStatesinthesecondhalfofthe19thcentury(Hatton&Williamson1998)andIsraelinthe1990s(Cohen&Hsieh2001).Intoday’sGermanyandSweden,whicharenearfullemploymentandrunlargestructuralcurrent account surpluses, the increased domestic capital formation ismost likely to befinancedbydomesticsavings,implyingreducedcurrentaccountsurplusesor,equivalently,reducedcapitaloutflow.Togiveasenseofthemagnitudesinvolved,ifthenewrefugeeflowadds1%tothe labor forceand–assumingthat in the longrun, thecapital/laborratio is

recognized featureofmigration to theUnitedStates as far backas the19th century. (Hatton andWilliamson1998)

16

constant21and that investment reacts fairly quickly22– that prompts a proportional 1%additiontothecapitalstockover,say,5years.Bytheendofthe5years,thenationaloutputis1%higher,implyinganaccelerationoftheaverageannualgrowthrateof0.2%ayearuntiltheadjustmentiscompleted.Assuminganincrementalcapitaloutputratioof4–neartheaverage for advanced countries 23 – the arrival of refugees would imply increased netdomesticinvestmentequalto0.8%ayearfor5yearsuntiltheadjustmentiscompleted.Ofcourse,iftheflowofrefugeesissustainedat1%ayearofthelaborforceoveraperiodofyears, then further output growth is possible and newadditions to the capital stock arerequiredeachyear.Forexample,a1%ayearincreaseinthelaborforceeachyearover3years,requiresincreasednetinvestmentamountingto2.4%oftheinitiallevelbytheendofyear3andimpliesthat,bythen,thegrowthrateofoutputishigherby0.6%ayear,andsoon. The message is that the cumulative effect of refugee arrival on economic growth ispotentiallysignificant,especiallyinslow‐growingmatureeconomies.The arrival and on‐streaming of large numbers of inexpensive laborers is politicallycontroversial, but its potential positive effect on long‐term economic growth is wellestablishedindevelopmentthinking.Theinsightthattheinflowofabundantlaborcanraisetherateofreturn tocapitaland‐undercertainconditions ‐generateavirtuouscircleofgrowthismostcloselyassociatedwiththeNobelPrize‐winnerW.ArthurLewis.Althoughthe Lewis model is usually thought to apply to the movement of workers from thecountryside to the cities and factories in poor countries, its relevance is more general.AccordingtotheHarvardeconomichistorianCharlesKindleberger,theLewismodelhelpsexplainmuchofEurope’spost‐wareconomicmiracleinthe1950’sandearly1960’sand,also(ashe accuratelypredicted) the subsequent slowdown.Hewrote: “…themost importantfactorshapingtheremarkableeconomicgrowthsince1950hasbeentheavailabilityofalarge supply of labor. The labor has come from a high rate of natural increase (theNetherlands), fromtransfers fromagriculturetoservicesand industry(Germany,France,Italy), from the immigration of refugees (Germany), and from the immigration ofunemployed or underemployed workers from the Mediterranean countries (France,GermanyandSwitzerland)”.(Kindleberger1967)Kindlebergerunderlines the fact thatwhile the increased supplyof labor canhelp fostereconomicgrowth,itisnotpersesufficient.Inpost‐warEurope,theneedsofreconstructionand pent‐up consumer demand during thewar years provided the sufficient conditions.

21Theassumptionsthatthecapital‐outputratioisfairlyconstant,andthatthecapitalstockandthelabor‐force to grow tend to grow at similar rates over very long periods are long‐ establishedempiricalregularities(Harrod1939).

22 Recent research in advanced countries suggests that domestic investment is quite quicklystimulatedbyamigrationsurge,sothatwithinafewyearsthecapital‐laborratiotendstoreturntoitspriorlevelinthefaceofalabormarketshock.ThisconclusionissupportedbystudiessuchasOttavianoandPeri(2008)fortheUS,BrückerandJahn(2011)forGermany,CohenandHsieh(2000),andOrtegaandPeri(2009)inastudyof14OECDcountries.23UKOfficeofNationalStatistics2013

17

Today’spost‐financial crisisadvancedcountriesvarygreatly in termsof theirunderlyingdynamism and need/ability to absorb new workers. It is perhaps not surprising that acountrysuchasGermany,whichhaslowunemployment,isagingrapidly,iscompetitiveandgrowingsmartly,runsabalancedbudgetandalargecurrentaccountsurplushasbeenmoreopentotheprospectofreceivingrefugeesthanothers,whilestrugglingItalyhasnot,despiteexhibitingevenmoreunfavorabledemographictrendsthanGermany(seechart3).Itshouldbenotedthat,whiletheinflowofrefugeescanpromotefasterGDPgrowth,thatwillnot be necessarily reflected in the higher growth of GDP per capita,whichmanywouldconsidertobethemorerelevantmeasureofwelfare.However,theexpansionofthecapitalstockimpliedbythearrivalofnewworkersmaynotbetheonlywaythatimmigrationboostseconomic growth.Micro‐level studies suggest thatmigrantsmay also induce acceleratedproductivitygrowthbyprovidingadisproportionateshareofentrepreneursandinnovators,bytakingonjobsormovingtolocalitieswherenativeworkersarereluctanttogo,andbyprovidingasourceoflaborservicesthatrespondmorereadilytothebusinesscycle(Orhan&Senyücel2015).Noneofthesebenefits–exceptfortheinitialdemand‐expandingstimulustodemand–arelikelytoaccrueiftherefugeeisforcedorencouragedtoreturnearlyinthecycleofhisorhereconomicintegration.AshasbeenextensivelyarguedinOECD,WorldBankandIMFreports,policiesinadvancedcountriestowardsimmigration,includingtowardsrefugees,shouldalsotakeintoaccountthe degree to which immigration can help offset unfavorable demographic trends. ThisconsiderationisespeciallyimportantinEurope,wherethepopulationisagingrapidly.Theold‐agedependencyratio–theratioofpeopleover65andthoseofworkingage–isprojectedtoriserapidlyfrom30in2015to55in2050.Charts2and3(fromIMF2016)showprojectedpopulationandnetmigrationfortheEU28andforItaly,wherepopulationagingisadvancingmostrapidly,andwherethenaturalchangeinthepopulationturnednegativebackin2010.AccordingtoUnitedNationsPopulationDivisionprojections(mediumvariant),Italy’s20‐64cohort,thepopulationofworkingage,willdeclineby10%(!)over2015‐2030.Chart2EU28PopulationChanges(Millions)

18

Chart3ItalyPopulationChanges(Millions)

TheEffectonWagesofUnskilledWorkers,Employment,andUnemployment

The1951ConventionrelatingtotheStatusofRefugeesrequiresstatestoaccordrefugees

19

"themost favorable treatment"accordedtoanynon‐nationalsofa foreigncountry in thesamecircumstances,withregardtotherighttoengageinwage‐earningemployment,andrefugeesmustbeallowedtostartbusinessesandpracticeliberalprofessionsequallywithothernon‐nationals.However, a recent studyby theWorldBank’sKNOMAD initiativeofpractices in 20 countries (including Germany, Italy, the United Kingdom and the UnitedStates) hosting 70% of the world’s refugees found that “remarkable diversity in legalprovisionsandconstraintsonrefugees’righttowork.Arestrictiveapproachtotherighttowork prevails, and most states are reluctant to ease these restrictions. The majority ofrefugeesworkintheinformalsector,butundermuchlesssatisfactoryandmoreexploitativeconditions compared with nationals” (Zetter & Ruaudel 2016). Assessing the effect ofrefugeesonlabormarketsinthehostcountrymustrecognizethisreality.

Moreover, the vastmajority of studies on the impact ofmigration onhost country labormarkets relate to economic migrants, and refugees differ in a number of respects fromeconomicmigrants.Refugee flows tend tobemore concentrated than thoseof economicmigrantsintimeandspace,andarelesslinkedtolabormarketopportunities.Inadvancedcountries,economic immigrationmaybemorehigh‐skilledthanrefugees,and,asalreadyindicated,tendstobemorecircular.However,asrefugeesfindjobsandbecomeintegratedintothelabormarketofhostcountries,theireconomiceffectstendtobecomemoresimilartothoseofeconomicmigrants.

Asisgenerallythecaseoflow‐skilledeconomicmigrants,refugees,evenwhentheyarriveinlargenumbers,areunlikelytohaveasignificantimpactonthewagesofnativelow‐skilledworkersorontheirjobprospectsinanadvancedcountry24.Eventhoughittakeslongerforthemtofindjobs,likeeconomicmigrants,refugeescanstimulatethereturntocapitalandcontribute to higher real wages among natives, especially those of the vast majority ofnativeswithwhomtheydonotcompetedirectly.

Manyrefugeesareunskilled25andthefearthatlargeinflowsofunskilledmigrantsfromtheSouthwilltakejobsawayfromunskillednativesintheNorthiswidespread.Theliteratureonmigrationdealswiththisissueextensively,anditexaminesepisodesoflargeinflowsofeconomicmigrantsaswellasofrefugees.Considertheeffectoftheinflowoflargenumbersof predominantly unskilled migrants on the native skilled and unskilled workers. Asdiscussedintheprevioussection,theinflowofmigrantstendstoexpandfinaldemandandtostimulateinvestment,raisingthedemandforallworkersthroughthosechannels.Skillednatives,whoarecomplementarytotheunskilledimmigrants,willtendtoendupbetteroffin the new steady state because of increased investment and because they have more24ThisconclusiondoesnotnecessarilyapplyincountriessuchasJordanandLebanon,whichhavereceivedfarlargernumbersofrefugeesrelativetotheirpopulation,therefugeesspeakArabic,andtheycompetedirectlywithnativelow‐skilledlaborers.(Dadush,2015)25 Germany’s Federal Service for Migration and Refugees (BAMF)released a study “Asylumapplicants:socialstructure,qualificationsandemployability”in2015.Accordingtothestudy,18%asylumseekersinGermanyhasauniversitydegree,while20percenthaveattendedahighschool,approximately one‐third a secondary school and 22 percent a primary school. Seven percent ofmigrants have no formal education at all. https://www.rt.com/news/343753‐germany‐refugees‐males‐statistics/

20

unskilled workers to work with. However, the effect on unskilled natives is a prioriindeterminate,since,whilethey,too,benefitfromincreasedinvestment,theywillconfrontincreasedcompetitionfromtheunskilledmigrants.Whichof theseeffectsdominateswilldepend critically on the extent, towhich theunskilledmigrants are close substitutes fordomesticunskilledworkers,aquestionthathasbeenextensivelystudied.

Most–thoughnotall‐studiesoftheeffectofunskilledmigrationonthewagesofunskilledworkers find only small negative effects. The early literature on the subject typicallyconcluded that a 1 percent increase in the immigrant share in the population causes nodeclineinwagesoradeclineof0.1percent.These“areastudies”attemptedtoexploitthevariationinmigrationincidenceacrosscountries,ormoretypically,acrosslocalitiesintheUnitedStates26.

However,asBorjaspointedoutinalandmarkcontribution,theareastudiesoftenfailedtocontendwithaseriousendogeneityproblem,since the intensityofmigrants inaspecificlocality is itself influenced by wage and employment opportunities, tending to bias theestimatedeffectofmigrationonwagesseverelydownwards(2003).TheadditionalcriticismonstudiesthatcorrelatethevariationinimmigrationintensityacrosslocalitiesintheUnitedStateswithwagesisthatnatives(andothermigrants)compensateforincreasedmigrationbymovingelsewhere.

Borjas’critiquegaverisetomanystudiesthattriedtogetaroundtheendogeneityproblem.Forexample,OttavianoandPeri(2008)estimatedaproductionfunctionincludinggroupsofworkers classified by education, experience, gender andorigin to evaluate the degree towhich immigrants substitute forotherworkers.Theyreach theconclusion thatunskilledimmigrantsandnativesareimperfectsubstitutesintheUnitedStates,andsimilarstudiesreachedthesameconclusionintheUKandGermany27.Immigrantswhodonotspeakthelanguage, often cannot read and write in Latin script, have scant social networks, andrelatively low expectations, tend to get and do different jobs than unskilled natives, socompetewiththemonlyindirectly.Manyoftoday’srefugees,hailingfromSyria,Afghanistan,andpartsofSub‐SaharanAfricaconformtothisdescription.

Anothersolutiontotheendogeneityproblem,whichalsoturnsouttobehighlyrelevanttothe study of forced migration, was to examine instances of mass migration which arepromptedbypoliticalorotherexternaleventsandthereforearenotplausiblyinresponsetolabormarketconditionsinthereceivingcountry.

ThefirstofthesestudieswasDavidCard’sclassicanalysisoftheinfluxofCubanmigrantstoMiamiduringthe1980MarielBoatlift(namedaftertheportinCubawheretheboatlifttookplace)(1990).Cardfoundthat,althoughmanyoftheMarielimmigrantswereunskilledandtheyincreasedthelaborforceoftheMiamimetropolitanareaby7percentinlessthanoneyear,theyhadvirtuallynoeffectonthewageratesorunemploymentlevelsofless‐skilled26ApartiallistofthesestudiesincludesAltonjiandCard(1991),ButcherandCard(1991),LaLondeandTopel(1991),andSchoeni(1997)27Studieshavereachedsimilarconclusionsinothercountries:Manacordaetal.(2006)fortheUK,andforBrückerandJahn(2011),D’AmuriandPeri(2011)andFelbermayretal.(2010)forGermany

21

non‐Cubanworkers, includingblacks andotherHispanics. Effects onpreviously residentCubanworkerswerealsoverysmall.

Instead,Card’sstudyfindsthattheMiamieconomywasabletorapidlyabsorbthefloodofpredominantly low‐skilled, Cuban immigrants. He speculates that two factors may helpaccountfortherapidabsorptionoftheMarielimmigrants:presenceofindustriesinMiami,such as garments, able to expand and employ unskilled migrants, and a compensatoryreductionintherateofmigrationintoMiamifromtherestoftheUnitedStatesfollowingtheinfluxoftheMarielimmigrants.FollowingCard,manyotherinstancesofsuddenmigrationsurgeshavebeenanalyzedandyieldedcomparableresults(seeHunt1992;Carrington&DeLima1996)thoughsomefoundsomewhatgreatereffectsonnativewages,intherangeof1‐2%.ThedebatearoundtheMarielboatlifthasrecentlybeenreignitedbyanewattemptbyBorjastoestimateitseffects,whichconcludesthattheyweresizable,andwhoseresultswereagainconvincinglyrefutedbyClemensandothers28.

Current flows of asylum‐seekers in advanced countries are not large enough to havesignificant effects on the wages of the unskilled, with the possible exception of a fewcountries. TheUnited States, for example, has admitted a cumulative total of just over1millionrefugeessince2000,mostlythroughresettlementschemes,andmanyofwhomarenownaturalized,anannualaveragethatrepresentslessthan1/50of1%ofitspopulationin2017. Refugees admitted since 2000 also represent less than 3% of the foreign‐bornpopulation in theUnitedStates.Over2014‐2016,when thenumberofasylumseekers inadvancedcountries surged,onlyHungary,GermanyandSwitzerlandsawasylumseekersexceed1%ofthepopulation,andonlySwedensawthemexceed2%ofthepopulation.

Asismostclearlyseenindevelopingcountriesreceivinglargenumbersofrefugees,suchasJordan,LebanonandTurkey,theconcentrationofrefugeesinspecificlocalitiescan“crowdout”publicservicessuchasschoolsandhospitals,orcausethetransportinfrastructuretobecomecongested.Inadvancedcountries,wheretheinfrastructureiswelldevelopedandtherefugeeinflowismuchsmallerrelativetothenativepopulations,theseeffectscanbeobserved in specific neighborhoods, typically in low‐income areas. Mitigating theseconcentration effects while at the same time avoiding the temptation of encouragingrefugeestosettleindistantlocationswhereitisdifficulttofindworkandpeopledonotwanttostayposesagenuinechallengetopolicy‐makers.

At thesametime, it isclear thatunskilledmigrantscanreducethepriceofmanymarketservicesandalsoreducethecostofmanypublicservices(e.g.unskilledmigrantshelpcleanstreets).Animportantbenefitthatnatives‐whethertheyareskilledorunskilled‐derivefromunskilledmigrants,andonethatuntilrecentlywaslargelyneglected,isthattheyhelpreducethepricesofnon‐tradedgoodsandservicesthatnativesuseintensively,suchashomecare, foodpreparation,gardening,andconstruction.Cortes(2008)findsthatthesurge inimmigrationintheUnitedStatesduring1980‐2000mayhavereducedthepricesoftheseservicesbyabout10percent.Bycontrast,asarguedfurtherbelow,thearrivalofrefugeescanputupwardpressureonhousinginlocalitieswheretheyareconcentrated.Thiscanmake

28ClemensandHunt2017,Peri&Yasenov2017,Pritchett,2017andBorjas2017

22

low‐incomehousing lessaffordableevenas itrepresentsanetwealthgain forthenativepopulationthatownshousing.

Sincemany advanced countries,most notably in Europe, experience high and persistentunemploymentovermanyyears,evenwhengrowthisnearorabovelong‐termpotential,the worry that increased immigration will simply make the structural or cyclicalunemployment problem worse resonates widely. Brucker and Jahn (2011) consider aneconomywherecollectivebargainingpredominatesinsome“rigid”sectorsandwherethelabormarketisveryflexibleinothers.Theyfindthatmigrationcanincreaseunemploymentintherigidsector,whilehavingarelativelysmallnegativeeffectonwagesintherigidsector,andreducesthewageintheflexiblesectorwhereunemploymentremainslowbydefinition.

Inthelongrun,BruckerandJahnfind,theseeffectstendtodissipateasinvestmentresponds.Thus, themodel that assumesperfect labormarkets tends tooverestimate the impactofmigration on averagewages and to underestimate the effect on unemployment. Brücker(2011) conducted a meta‐analysis of studies examining the effect of immigration onunemploymentintheU.S.andEuropeandfoundthatingeneral,anincreaseinimmigrationby1percentofthepopulationleadstoanincreaseinunemploymentofnomorethan0.3percent.

Incontrast,studiesoftheU.S.labormarket,whichisamongthemostflexible,havefoundnosignificant effect of immigration on employment opportunities for native workers (Peri2012), including low‐skilled native workers, between 1990 and 2007. Peri and Sparber(2008)reportthatmigrantsinducetaskspecializationofnatives.Theyfoundthat,amongless‐educatedworkers,thosebornintheUnitedStatestendtohavejobsinmanufacturingormining,whilemigrantstendtohavejobsinpersonalservicesandagriculture,providinganexplanationforwhylow‐skilledmigrationhasalimitedimpactonemployment.

Infact,theshareofmigrantsamongtheless‐educatedisstronglycorrelatedwiththeextentof U.S.‐born worker specialization in communication tasks. In states with a heavyconcentration of less‐educated migrants, U.S.‐born workers have shifted toward morecommunication‐intensiveoccupations.Thosejobspayhigherwagesthanmanualjobs,andsosuchamechanismhasstimulatedtheproductivityofworkersbornintheUnitedStatesandgeneratednewemploymentopportunities.

Themainpointthatemergesfromthisbriefreviewisthatreturnpolicycannotabstractitselffromimmigrationpolicymoregenerallyorevenfromlabormarketpolicy.Countriesthathavemoreflexiblelabormarkets,forexample,arebetterabletoabsorbrefugeesandwillbelessinclinedtopressfortheirreturn.Wenowconsidertheeffectsofrefugeesonthepublicpurse,anissuethathasbeentheobjectofextensivepoliticaldebate.

TheFiscalImpactofRefugeeFlows

23

Theoverallfiscalimpactofrefugeeflowsislikelytobesmall.Itisnegativeatfirstastheyarecostlytohouseandsupportwhentheyfirstarriveandittakestimeforthemtolearnthelanguageand findwork,when theyare allowed towork. Sincemost refugeesareyoung,studiessuggestthattheirfiscalimpactbecomespositiveonacash‐flowbasisinlessthanadecade and positive on a present value basis a few years later. Like economicmigrants,refugees that aremore skilled tend to have amore favorable fiscal profile than the lessskilled29.However, at least one study reviewedbelow suggests that less‐skilled refugeesonce settled and started working, have a more favorable fiscal profile than low‐skillednativesbecausetheydrawlessonsocialbenefitprograms.Many studies of the fiscal impact of immigration exist,while very few analyze the fiscalimpactofrefugees.Statisticsrelatingspecificallytotheemploymentofrefugees,thetaxesthey pay and benefits they receive, and especially those that do so over the lifetime ofrefugees,arescarcelyavailable:“…whatrefugeeliteraturedoesexisttendstoconcernveryspecific populations, uses very small samples, relies on data from a small number ofcountrieswith high refugee totals, or focuses on very short‐term outcomes” (Evans andFitzgerald, 2017). Studies of the fiscal impact of general immigration, which includeimmigration forwork, to reunite families, aswell as refugees, help shed light on that ofrefugees–eventhoughtherearemajordifferencesbetweenthegroups.FiscalImpactofGeneralImmigrationandofRefugeesinParticularTheassessmentof the fiscal impactof immigrantsdependscruciallyon themethodologyusedandontheprecisequestionbeingasked.Thus,thefiscalimpactcanbemeasuredinagivenyearasismostcommonlydone,oroverthelifetimeoftheimmigrant,andaccountingornotforthedebtincurredbycurrentandfuturegenerationsofnatives,andaccountingornotforthegeneralequilibriumeffectsontheeconomy,includingtheimpactofimmigrationoneconomicgrowth, i.e.on investment, laborparticipationbynativesandontotal factorproductivity.Differentmethodscananddoyielddifferentanswers,notonlyquantitativelybutalsodirectionally.Still,theacademicconsensusonthefiscalimpactofimmigrationcanbepithilysummarizedas by Lee and Miller (AER, 2000) in a study of the United States: “it is clear that theconsequencesofimmigrationcanonlybeassessedoververylongtimehorizons.SomeU.S.‐bornchildrenofimmigrantsarrivingtodaywillstillbealive130yearsfromnow”and“theoverall fiscal consequencesof…immigration [are]quite small andshouldnotbeamajorconsiderationforpolicy…100,000moreimmigrantsperyearwouldinitiallyraisetaxesfornonimmigrants,andlaterreducethem,byamountslessthan1percentofcurrenttaxlevels”.Thatnumberof immigrants,by theway, representsabout0.33%of theUSpopulation in2000 – a proportion similar to the annual immigration flow inmost advanced countriestoday.An international survey of the annual fiscal impact of immigrants, including economicmigrants,familyreunificationandrefugees(OECD,2013)in2007‐2009concludedthatthere29UnitedStatesAcademyofSciencesReportonImmigration,2017

24

isconsiderablevariationacrosscountries,dependingonthecompositionoftheimmigrantpopulation,theirparticipationinthelaborforce,thestructureoftaxes,andthenatureandextent of social benefits provided. The net fiscal impact of immigrants across the OECDsamplein2007‐2009isonaverageslightlypositive(0.35%ofGDP),butrangesfrompositive2%inSwitzerland(whichhostsmanyskilledimmigrants)tonegative1%ofGDPinGermany(whichhostsmanyunskilledandolderimmigrants).ExcludingoutlierssuchasSwitzerland,there is no correlationbetween the shareof immigrants in thepopulation and the fiscalimpact,underscoringthefactthatotherfactorsthanthenumberofimmigrantsareatwork(IMF,2016),suchasagestructureoftheimmigrantstock,timesincearrival,etc.With this as background, we now turn to the fiscal impact of refugees. A rigorousexaminationof the fiscal impactof refugees requires the identificationof refugees. Sincehistoricalstatisticsonrefugeesarenotsufficientlyreliable,orarenotavailable,estimationsmustuseindirectmethodstoidentifyrefugeesthatcombinevarioussourcesandmakemanyassumptions.Typically,thisisdonebyincludingintherefugeecountallimmigrantsfromcountriesknowntohavebeentheoriginoflargenumbersofrefugeesduringyearsofconflictorpersecution.Below,weexamineexamplesofthreetypesofstudy,relatingtoagivenyear,relatingtothelong‐term,andrelatingtoaspecificlocality.Short‐termfiscalimpactofrefugeesTheSwedishcaseexaminedbyRuistisespeciallyinterestingbecauseSwedenacceptslargenumbers of refugees, and, since its welfare state is among the most developed, socialspendinginsupportofrefugeesisespeciallygenerous.TherefugeepopulationinSwedenisestimatedat5.1%ofthepopulation30:“Ifall15pre‐2004EUmemberswouldhavehadthesame rate of refugee immigration per capita as Sweden in 2005‐2014, total refugeeimmigrationinthesecountrieswouldhavebeen5.9millioninsteadoftheactual740,000.”(Ruist,2015)Hisassessmentof thenet transfer torefugees fromtherestof theSwedishpopulation in 2007 (a year chosen because it is the last yearwhere the needed data isavailable)amountedto1%ofSwedishGDP.Ofthistransfer,about20%reflectsthegreaterdraw of refugees on social support expenditures, including integration grants, languagetraining,diversetypesofsocialassistanceandindividualgrants,andexpenditureoncrimeand justice.Although refugeesdraw lessonpensions,disability, andeducationspending,theirdisproportionatecostontheformerservicesdominates.Thelion’sshareofthetransfertorefugees‐80%‐,however,reflectstheirlowercontributionthrough taxes, which is due in turn to the lower share of refugees that are employed,estimatedat56.5%ascomparedto78.5%amongnatives,andthelowerearningsofrefugeesthatareemployed,estimatedtobe40%lessthanthoseofnatives.However,policiestowardsrefugeesmust also recognize that time since arrivalmatters. As the chart below shows,

30 This number differs from UNHCR statistics‐ presumably because it is based on estimates ofrefugees arriving inpriordecades,manyofwhomhavebeen accorded some formof permanentresidencestatusorhavebeenaccordedSwedishcitizenship.

25

unlikenatives, refugees inSwedencannotquickly integrate into the labormarket inpartbecause they do not speak Swedish. Eventually, they do, even though their participationremainslessthannatives.ItisclearfromtheSwedishexampleandthatofothercountriesthatrefugeesorasylumseekersthatareleftinlegallimboandarenotallowedtoworkcosttheeconomyandthetaxpayerdearly31.AcrucialassumptionmadebyRuist,andwhichisalsothemostcontroversial,isthatrefugeesincur the full cost of pure public goods, including defense, infrastructure, and publicadministration,asnatives.Ifinstead,theassumptionisthatthecostofprovidingthesepublicgoodsdoesnotchangemateriallywiththearrivalofrefugees,thenettransferfromtherestoftheSwedishpopulationtorefugeesisreducedto0.2%ofGDP.Anothercaveat is that the fiscal impactofrefugees in2007maynotberepresentativeofother timeperiods. In2007,over2/3of theSwedishrefugeepopulationhadbeen in thecountrylessthan15years,andover1/3lessthan10years,anditwouldbereasonabletoassume that the fiscal impact of these relatively recent refugees would become morebeneficialovertime.Comparingtheeffectofrefugeesontheeconomywiththatofnativesoreconomicmigrantsatjustonepointintimeasisoftendone,canbemisleading.Ifsuchcomparisonsaredoneearlyonafterthearrivalofrefugees,asisthecaseinmanycountriestoday,theyareboundtoexaggeratethecostsandunderestimatethebenefitsofintegratingthem.Ifontheotherhand,comparisonsaredonemanyyearsaftertheyarrive,theyunderestimatethecostsandexaggeratethebenefitsofhostingrefugees.Instead,comparisonsmustbecarriedoutoverthelifetimeoftherefugeephenomenon,i.e.overoneortwogenerations,andconsideringthetimevalueofmoney.ThisisthemethodemployedbyarecentstudyofrefugeesintheUnitedStates,theobjectofthenextsection.Long‐TermFiscalImpactofRefugeesAsmentionedpreviously,thefiscalimpactofRefugeesdependscruciallyonthespeedwithwhichtheystartwork.Clearly,theinitialcostsofreceivingthemwillbehigheriftheyremaininlimboasasylumseekersforlongandarenotallowedtowork,whichisoftenthecaseinEuropean countries, for example. By contrast, in the United States, most refugees areresettledonce their applicationhasbeen clearedand they canworkalmost immediatelyuponarrival.EvansandFitzgerald(2017)examinetheexperienceof19,298refugeesagedbelow65thatenteredtheUnitedStatesbeforetheageof45.Thestudy,whichcoversa20‐year period to 2010‐2014, may not be representative of other countries, but is worthexaminingsinceitrepresentsperhapsthemostcomprehensiveattempttoexaminethefiscalimpactofrefugeesoveranextendedperiod.31OnereportestimatedthatCanadaspendsover$129millionannuallyongovernmenttransferstothosewhoare deemed as protected persons by Immigration and Refugee Board in Canada but still face numerousobstacles and delays before they could enjoy rights as permanent residents. (Coates & Hayward 2005) Ifautomaticlandingwereimplemented,thegovernmenttransferwoulddecreasewithanestimatedsavingsofover$101million.Thisresearchalsoshowsthatduetobarriersinthelabormarket,refugeesloseintotalover$344millioninforgoneincome.

26

TheyfindthatrefugeeswhoentertheU.S.beforetheageof14graduatefromhighschoolandcollegeatthesamerateastheU.S.born.Holdingeducationalattainmentconstant,thereisnodifferenceineconomicoutcomesbetweenrefugeeswhoarriveaschildrenandtheU.S.born.TheyalsofindthatevenrefugeeswhoenteredtheU.S.asteenstendtoachieveasimilareducational standard as the U.S. born, although it takes them longer to complete theirstudies.They suggest that, therefore, theeconomicdecision to invest in theeducationofrefugeechildrenisunlikelytodiffersignificantlyfromthedecisiontoinvestintheeducationofnatives.Refugeeswhoenteredasadults(18‐45)hadlowereducationalattainmentthanU.S.bornadultsatthetimeofthesurveyin2010‐2014:67%haveahighschooldegreeand23%haveacollegedegree,comparedwith91%and37%amongnatives;andtheirlaborearningsare$23,000comparedwith$35,000fornatives.RefugeeshavemuchlowerEnglishlanguageskills.Moreover,9.3%areonwelfareand39%areonSNAP(foodstamps)comparedto3%and4.8%ofnatives.However,68%of refugeesareemployed,aslightlyhigherrate thannatives, 67%. Controlling for age, gender and education, refugees attain higher laborparticipationandemploymentratesthantheU.S.bornafterabouttenyearsinthecountry.Chart5:OutcomesofRefugees thatEntered theU.S.atAges18‐45asaFunctionofYearsintheU.S.,ComparedtoU.S.BornAdults,18‐65

27

(Source:EvansandFitzgerald2017)

Evans and Fitzgerald estimate that the fiscal cost of resettling a refugee initially isapproximate$10,000.(2007)Tothiscostareaddedthoseofvarioussocialsafetynets,whichaverage about $6,000 a year over the first 8 years anddecline gradually thereafter. Therefugeetypicallyfindsworkafterabriefperiodandcontributesminimallytothetaxtake(sales,realestate,socialsecurity,Medicareandincometaxes)fromthefirstyear,withthetax take exceeding $6,000 each year, by year 8 after arrival. After year 8, the refugee isbudget‐positiveonacash‐flowbasisandhas‘paid‐back’byyear13orso.Usingadiscountrateof2%,theauthorsestimatethatthosewhoenterthecountrybetweenages18‐45payonaverage$21,000moreintaxesthantheyreceiveinbenefitsovera20‐yearperiod.

Becauseofdatalimitations,nocomparablecalculationiscarriedoutforimmigrantsenteringaftertheageof45.Giventhetimeittakesforrefugeestointegrateandtheirclosenesstoretirement(drawingonMedicare,socialsecurity,etc.),thenetfiscalimpactofolderrefugeesiscertaintobelessfavorable.However,thevastmajorityofrefugeesthatenteredtheUnitedStatesfrom1990to2014wereundertheageof45atthetimeofentry.TheIMFestimatesthat,oncurrenttrends,thearrivalofrefugeescouldleadtoadeclineof0.75%ofGDPofpension spending in2030,as refugeeswill contribute toGDPbutdraw

28

proportionallylittleonthepensionsystemincomingyears.HealthcarecostsasashareofGDPwouldalsobeexpectedtodecline.

TheAustraliangovernmentistheonlyoneknowntosystematicallymonitorandprojectthefiscalimpactofrefugeescomparedtoothervisacategoriesemployingamodelovera20‐year period. While immigrants across all visa categories other than “parent” and“humanitarian”makeapositivefiscalcontributionwithinayearortwoofarrival, thatofrefugeesisnegative,especiallyoverthefirst3years,andonlyturnspositiveafter20years.ItisnotablethattheemploymentrateofAustralianbornandofthoseborninmainlyEnglish‐speakingcountriesisessentiallythesame.ItisalsonotablethatmigrantsborninamainlyEnglish=speakingcountryaremore likely tobeemployed than thoseborn innotEnglishspeaking countries and that the difference is especially pronounced among the femalepopulation.Table 5 ‐ Employment status of civilian population aged 15 years or more, bybirthplace,2010–11Category Employment

rate(%)Unemploymentrate(%)

Participationrate(%)

Australianbornmales 71.1 5 74.8Australianbornfemales 59.2 5.1 62.4Australianborn 65.1 5 68.6MESC (Mainly English SpeakingCountries)bornmales

71.7 3.9 74.6

MESCbornfemales 58.5 5 61.6MESCborn 65.3 4.4 68.2NESC (Not English SpeakingCountries)bornmales

63.5 4.9 66.8

NESCbornfemales 47.6 6.6 51NESCborn 55.3 5.6 58.6Total 63.3 5.1 66.6Source:LaborForce,Australia(6291.0),AustralianBureauofStatistics,CitedbyTrendsinMigration:Australia2010–11FiscalImpactattheLocalLevelTheprevioussectionshavedealtwiththefiscalimpactofrefugeesatthenationallevel,butrefugeessettle inspecific localities.Whathappens to thebudgetwhenrefugeesarrive inlargenumbersatthelocallevel?OneexamplecomesfromHagstrom’sexaminationoftheMohawkValley (2000). This examplemay not be representative of situationswhere therefugees’arrivalisunplannedandrefugeesarestilltobevetted,itisneverthelessinstructive.Between1975and1999theMohawkValley(whichincludesOneidaCountyandwithinitthecityofUticainNewYorkState),resettled8,759refugees,mainlyfromBosnia,theformer

29

SovietUnion,andVietnam,accountingforabout3%ofthecounty’spopulation.Over1977to1999,thetotalpopulationofOneidaCountydeclinedby32,000,from327,000to295,000.Therefugeeinflowreplacedonlyapartofdepartures.ClosuresofmanufacturingbusinessesduetointernationalcompetitionandclosureofalargeAirForcebaseseverelyaffectedtheregion’s economy. Despite these setbacks, the region’s economy grewmodestly and theunemploymentratedeclinedtolowlevels.Hagstrom’sevaluationconsidersonlytheeffectofrefugeesonthecounty’sfiscalbalancesothatrefugeecostsbornebythefederalandstategovernmentsandtaxespaidbyrefugeestothe federal and state governments are excluded from the analysis. The cost of the localauthorityofsettlingarefugeehousehold(onaverageconsistingof3.22people)is$4,413,consisting mainly of education, nutrition support, and healthcare. The cost of nutritionsupportdropsquicklyas refugees found jobs, andof education, as children leave school,whilehealth‐carecosts(Medicaid)persist.Theeducationofrefugeechildrenisestimatedtocost less than average because of that, given the county’s declining population includingchildren,thereissparecapacityintheclassroom.Only3.7%ofrefugeesareover65onarrivalandmostrefugeesfindlow‐wagejobs($7.18anhour)thatrequirelittleskillandknowledgeofEnglish.Afterafewyears,asmallnumberofrefugeeshavefoundhigherpayingjobssuchasschoolteachersoraccountants.Thereislittleevidencethatrefugeeshavehurtthejobprospectsofnativeworkers.Despitethedeclineinpopulation,employmentinOneidaCountydidnotdecline,reflectingbothincreasedlaborforceparticipationandsharpreductionsinunemployment.The primary fiscal benefit accruing from refugees stems from their employment,consumption of local goods and services (sales taxes) and the significant boost to localproperty markets, including on housing prices (property taxes). Without the arrival ofrefugees,manyhousesmighthaveremainedunoccupied.Hagstrom’ssimulationssuggestthat,basedonadiscountrateof4%,thenetfiscalimpactofarefugeehouseholdbecomespositiveafter13yearsandispositivethereafter,assumingthatthehouseholdremainsintheregion.Underascenarioofaconstantstreamof750refugeesa year, net annual benefits for the county become positive after 15 years, while thecumulative benefits (considering the cost of settling newwaves of refugees every year)becomepositiveafter23years.ThoughHagstrom’sanalysisdoesnotaccount fornationwide fiscal impact, it isuseful inillustrating the importance of refugee inflows at the local level in a context of decliningpopulation.Theanalysisalsohighlightsthepotentialofregionalresettlementpolicies.Suchpoliciesmust,however,berealistic,i.e.encouragingrefugeestosettleinlocalitieswheretheycanfindjobsandarelikelytowanttosettleinthelongterm.Theprevioussectionshaveshownthat,inmostinstances,theeconomicandfiscalimpactofrefugees is typically small. It is negative at first and tends tobecomepositivewith time.Advanced countries need immigrants, including refugees, though some are better placed

30

todaytoabsorbthemthanothers,especiallyonaccountoftheirlabormarketpolicies.Thenext section considers what happens when refugees return voluntarily or are forced toreturn,andtriestodrawsomelessonsforreturnpolicy.ReturnPoliciesAs argued above, return policies should be setwith broader economic and immigrationobjectives inmind. Voluntary return of refugees from an advanced to a poor country isunlikelytobecomealarge‐scalephenomenon,notonlyonaccountofsecurityrisksbutalsoofthebigdifferencesinearningcapacitybetweenthecountriesoforiginanddestination.Inthissection,Iwillarguethatthedevelopmentdimensionofreturnpolicymattersandshouldbepursuedbutisveryunlikelytoyieldquickresults.Theoverallobjectiveofreturnpolicyshouldbetofosterinclusiveeconomicgrowthinthehostcountry,developmentofthecountryoforigin,andenhancethewelfareoftheforciblydisplaced In practice, the return policy is more often motivated by legal and politicalconsiderationsthanbybroadereconomicobjectives.Forexample,accordingtoEuropeanCommission’sreportonReturnofRejectedAsylumSeekers,“aneffectivereturnpolicyisofcrucial importance for themaintenance of trust in the EU’s asylum system as a systemprovidingprotectiontothosewhoneedit,whileensuringthereturnofthosewhodonot”(EC2016).Thissectionaddressesthreequestions:Whatshouldbetheinstrumentsofreturnpolicy?Isvoluntaryreturnpossible?Howcandevelopmentpolicyandreturnpolicyworkhandinhand?InstrumentsofReturnPolicyViewed through thenarrowestprism, a returnpolicyhas twodimensions – encouragingvoluntary return by providing financial incentives and other forms of resettlementassistance,andestablishingthe legalandenforcementmechanismsfordeportationwhennecessary.Toreturnirregularmigrants(includingfailedasylumseekers,aswellasthoseresiding in a country illegally), countries implement voluntary return programs. Theseprograms are recognized as potentially the most effective (in terms of financial cost,expenditureofpoliticalcapital,reducedfriction,thewelfareofthereturneeandtheimpacton the country of origin) as a means to return irregular migrants. Host countries haveAssistedVoluntaryReturnPrograms(AVRP)thathelpthirdcountrynationalsreturntotheircountryoforiginandresettle.Countriesmostlyencouragerejectedasylumseekerstoreturnvoluntarily, but coercive methods are used if the rejected asylum seekers refuse tocooperate. Measures of deportation range from detention to enforced removal throughsurpriseraids,etc.Inreality,afarbroaderandlessobvioussetofmeasurescanbeadoptedandareadoptedtoencouragereturn.Tostartwith,countriesofdestinationexerciseconsiderablediscretiononwhether they grant asylum, as canbe seen, for example, in the case ofAfghanor Syrianrefugees.InthecaseofasylumseekersfromSyria,positivedecisionstograntrefugeestatusrangefrom96%oraboveinmostEUmemberstatestolessthan60%inHungary,Italy,andRomania.(ESI2017)Officialpronouncementsstatingthenumberofrefugeescountriesare

31

willingtoadmitprovideanimportantsignaltoasylumseekersastothelikelihoodthattheirapplicationwillsucceed.Undertheirnon‐refoulementUNagreementobligations,countriesarecommittedtoshelteringallthosethatentertheirborderandwhichtheyrecognizearerefugees.Yetmosthostcountriessetspecifictargetsofhowmanytheywilladmit.Countries also decide on the duration and complexity of legal procedures governing thegrantingofasylumand, if inthefinal instance, it isnotgrantedandthemigrantdoesnotleaveoftheirownaccord,onthealacritywithwhichthedecisiontodeportisimplemented.Thoughacrossthehostcountriesmanydeportationdecisionsareimplementedwithindaysorweeks,thatisfarstillfromauniformpractice.Forexample,perarecentreportbytheSWPthinktankinBerlin,itisestimatedthatover215,000peopleofficiallyobligedtoleavewere living inGermanyattheendof July2016;ofthoseabout2/3wererejectedasylumapplicantsandanother1/4wereirregularmigrantswhohadoverstayedtheirvisas.About3/4 of those technically obliged to leave had obtained “toleration status”, so theirrepatriationhadbeensuspendedduetoobstaclesofsomekind32.Countries also decide on whether the recognition of refugee status is temporary andcontingent or whether it is permanent. For example, the United Kingdom’s granting ofrefugeestatusincludesa5‐yearreview,inaccordancewiththecessationclause,allowingforitswithdrawalifthesituationinthecountryoforiginimproves.Insomecountries,e.g.theUnited States, the granting of permanent residence (Green Card), paves the way tocitizenship. In many European nations granting refugee status and even permanentresidencestatus,itdoesnot.Amongthemostimportantmeasuresthatinfluencethedecisiontoreturnisthegrantingorwithholdingofaworkpermit.Asalreadymentioneddespitetherefugees’righttoworkasmandatedunderthe1951convention,whichstipulatesthattheyshouldhavethesamerighttoworkasothernon‐nationals,practicesvarywidely.Forexample,pertheOECD,inthefiveyearsfollowingtheoutbreakoftheSyriancrisis,only18,200workpermits,15,300studentresidencepermitsand72,000family‐reunificationpermitsweregrantedtoSyriansinthe35OECDmemberstates33.Inadditiontoworkpermits,theextentofassistanceinprovidingtherefugeesandasylumseekerswithfood,shelter,healthservicesandeducationforchildrenarealsopartofthepolicyinstrumentscaneitherto“push”themtoreturnorto“pull”themin,encouragingthemtostay.Whilemostcountriesrecognizethattherearelegalobligationsandmoralimperativetotreatasylumseekersandrefugeeshumanelyandtrytodosowithinlimitedmeans,thereareverylargedifferencesintheirtreatment.Someofthesedifferencesareobjective,inthesensethattheyreflectaffordabilityandlabormarketconditions,asismostobviouswhencomparingrelativelypoorcountryhostssuchasJordanandLebanon–whichreceivebyfarthelargestshareofrefugees–andadvancedcountryhosts.However,asdifferencesintreatmentacrossadvanced countries also show, they can reflect policydecisions (ornon‐decisions) aboutwhethertheywanttohelptherefugeeintegrateorencouragethemtoleave.32Angenendt,KippandMeier,201733OECDMigrationOutlook(2016)

32

This verywide and grey area of asylumpolicy is recognized in the academic and policyliterature, and experts have begun to distinguish between genuine voluntary return and“nominallyvoluntaryreturn”34or“soft‐deportation”35.Webber(2011)states:“Repatriationcannot be termed ‘voluntary’ where the alternative is utter destitution, with denial ofaccommodation, basic support and the opportunity towork, or the prospect of childrenbeingtakenintocareormonthsoryearsindetention.”An important aspect of return policy – and one that is sometimes overlooked – is therelationshipwiththecountryoforigin.Underinternationallaw,everycountryisobligedtoadmititsowncitizens.Inpractice,however,manyquestionsarise:whatprotocolshouldbefollowedwhenadeniedasylumseekerorrefugeeissupposedtoreturnunderthecessationclause?Howwilltherefugeebetreatedonreturnandhowcanheorshebehelpedtoresettle,andhowwill costsbe shared?Readmissionagreementsaredesigned to facilitate return.Theysetouttheconditionsthatcountriesareobligedtofollowtoreadmittheirowncitizensandsometimesthird‐countrycitizenswhohavepassedthroughtheirterritory.TheEUhasconcludedreadmissionagreementswith the followingcountries:HongKong,Macao,SriLanka,Albania,Russia,Bosnia,andHerzegovina,Macedonia,Georgia,Moldova,Montenegro,Pakistan,Serbia,Turkey,andUkraine.Negotiationsareon‐goingwithMorocco,whereasmandatesfortheEuropeanCommissiontonegotiatereadmissionagreementsexistalsoinrespectofChina,Algeria,andCapeVerde.Inaddition,EUmemberstateshavebilateralagreementsonreadmissionwithcountriesinandoutsideofEU.Forexample,asof2010,Germanyhas29agreementsand13ofthemarewithEUmemberstates.(CouncilofEurope,2010) Tripartite agreements, involving the country of origin, destination, and aninternationalagencysuchasUNHCRorIOMhavealsobeenreachedtofacilitatethereturnofirregularmigrantsandofrefugees.Though there is little doubt that these agreements play a useful role and increase thelikelihoodthatreturnwillbesustainable, thereareconcernsthattheycanalsobecomeavehicleforhumanrightsabuses.Whenafailedasylumseekerisforciblyrelocatedtoathirdcountry,theymaybedeprivedofthechancetosubmitanasylumapplicationandbecomesubject to “chain‐refoulement”. (Council of Europe 2010) For instance, under bilateralagreements,manynon‐LibyanirregularmigrantsarereturnedtoLibyaeachyear.Libyahasnotsignedthe1951GenevaConventionandlacksanasylumsystem.BeingreturnedtoLibyaasathird‐countrynationalmeansthatonemightnothaveachanceofsubmittinganasylumapplicationatall.Insummary,ifthedecisiontoadmitasylum‐seekersandtherateatwhichtheirclaimforrefugeestatusisrecognizedarepoliciesthatlieatthestartoftheprocess,programsthatincludenegotiationswithorigincountriesandfinancialincentivesforreturnlieattheotherend.Butalmostcertainlythemostimportantpoliciesthatinfluencethedecisionofaforcibly

34Gibbey(2008),citedinKuschminder35Leerkes,OsandBoersema(2017)

33

displacedperson to returnor to stay, andhis orher success in either endeavor, iswhathappensinbetween.Inthenextsection,weevaluatethelikelihoodofvoluntaryreturn.IsVoluntaryReturnPossible?

Thedecisiontoreturnvoluntarilyismadeaftercomparinginformationaboutconditionsandprospectsofcountriesoforiginandthehostcountries.(Koser1998,Faist1999,King2000)Policyinthehostcountrywithregardtoincentivesordisincentivestostayorreturnalsoplaysarole,(Bloch&Atfield2002)however,aspointedoutbythesurveyresearchdonebyBlack,KoserandMunkonVoluntaryReturnofrefugeesintheUK,thedecisiontoreturnismadebasedmoreonthesituationinthecountryoforiginthanpoliciesofhostcountries.(2004)

Three sets of factors determine the decision to return: structural, individual and policyinterventions.Structuralfactorsincludebothsituationsinthehostcountryandcountryoforigin. Individual factors include individual attributes like age and gender and socialrelationslikespouseandchildren.Policyincentivesalsoincludepolicyinterventionsbothfromthehostcountryandcountryoforigin.Basedontheabove‐mentionedsurvey,thetopthreefactorspotentialreturneesconsiderissecurity,employment,andhousing.Althoughthesampleissmall,itincludedmultiplegroupsofcountriesoforigin(Somalis,Tamils,Kurds,Afghans, Iranians Kosovans, and Chinese). Different nationalities emphasized differentconsiderations.Notsurprisingly,respondentsfromAfghanistanandSomaliareportedthatsecurityistheirgreatestconcern.

Ideally, the forciblydisplacedpersonwillhavesavedupsomecapitalandacquiredskillsabroadthatcanbeinvestedinthecountryoforigin.(Sorensenetal.2008)Asylumseekersorrefugeesthathaveresidedinthehostcountryforonlyabrieftimewillnothavebuiltupsavings,and–unlesstheyhaveaccesstohigherpayingjobs‐thehumancapitalandskillstheyacquirewhileabroadwillbelimitedormaynotfittheneedsoftheircountryoforigin.

Iftheyreturnvoluntarily,refugeestendtodosowithcautionandovertime.Afamilywillrelyonitsownnetwork,andoftensendascout, typicallyayoungman,toverifythatthegroundissecureandthattheirlandandhousecanberetrieved,andotherswillfollowbeforewomenandchildrenreturn.(Harildetal.2015)Ifthedecisionismadetoreturn,theaimofpolicymakersshouldbetoachieve“sustainablereturn”.Therearemanydefinitionsofsustainablereturn–hereisausefulonefromtheIOM“Sustainablereturnshouldbeunderstoodeitherasa)successfulreintegrationinthecountryoforigin,whichincludestheeconomic,socialandpsychologicalaspectsandthecapacityofthe individual tocopewithpush factors,botholdandnewonthesame levelas the localpopulation, or b) eventual legal remigrationmade possible by skills acquiredduring thereintegrationprocess”(2015)

34

Toachieve“sustainablereturn”expertstendtounderscorethe importanceofcontinuousandintensivecounseling,carefulmonitoringandresearch,aholisticviewofindividualandstructural influencing factors, including the situation in thecountriesoforigin, andclosecooperationbetweenallstakeholderswithinthehostcommunityaswellasthereturnees’countriesoforigin36.Evenafterthedecisiontoreturnhasbeenmade,continuingtoupdateinformationonsocio‐economicconditionswillhelpwithreintegration.Ideally,reintegrationassistanceshouldbetailoredtotheindividual’sprofileandthecountryoforigin’ssituation.Itcouldincludevocationaltraining,educationgrants,etc.Itispointedout“goodresultshavebeenachievedwhereassistanceisofferedtoboththeindividualandthecommunitytowhichheorsheisreturning”.(Harildetal.2015)Therearemanyreportssuggestingthatreturneesoftenfeeldiscriminatedagainstwhentheyreturntotheirhomecountries, reinforcing the importance of counseling. (Schuster&Majidi 2013) Return tounfavorableconditionsmayonlyresultinrenewedattemptstoemigrate.

a) Thesituationincountriesoforigin.

Thecountries thatarethe largestsourceofrefugees include:Afghanistan,Syria,Somalia,SouthSudan,Sudan,DemocraticRepublicofCongo,CentralAfricanRepublic,Iraq,Myanmar,and Eritrea (UNHCR, 2017). These countries account for the vast majority of refugeesworldwide.WewillalsoconsiderthreecountriesinEuropethatareanimportantsourceofasylum seekers in Europe, and whose recognition rate is low, and account for manyreturnees–theseareKosovo,SerbiaandAlbania.Isreturnpossibleandsafe,andifso,doestheeconomicsituationbodewelltoprovideessentials,namelyajob,housing,educationandreliablepublicservices?

Itishelpfultoidentifythreegroupsofcountries:onesthatarestillmiredinwarand/orhigh‐intensitycivilconflicts,countrieswheremedium‐intensityconflictpersistandarealreadyintheprocessofrebuilding,andcountrieswhicharenotinconflict.AccordingtotheArmedConflict Survey (IISS 2017), high‐intensity conflict is defined by frequent (daily) armedclashes between governments, government forces and insurgents, or among non‐statearmedgroupsthatcontrolterritory.Medium‐intensity,ontheotherhandrefersto“regulararmed clashes between governments, government forces and insurgents.” (IISS 2017)Withinthefirstcategory,accordingtotheArmedConflictSurveyin2016areAfghanistan,Syria,Somalia,SouthSudan,andIraq,whileMyanmar,CentralAfricanRepublic,DemocraticRepublicoftheCongo,andSudanareinmedium‐intensityconflict.(EritreaisnotincludedintheArmedConflictSurveyin2017,discussedinAnnex).MinoritiesinAlbania,KosovoandSerbiaareoftentheobjectsofdiscriminationbutcountriesarenotinconflict.

AcapsulesummaryofthesecurityandsocialsituationineachoftheothercountriesincludedinthissurveycanbefoundinAnnex2.Thesurveyillustrateshowimpracticalreturnistodayinthecountriesinconflict,suchasSyria,Afghanistan,andIraq,andvoluntaryreturninlarge

36 Best practice and new methods in return policy, July 2017,http://www.bamf.de/SharedDocs/Meldungen/EN/2017/EMN/20170504‐emnjahrestagung‐rueckkehr.html

35

number is difficult to conceive. Even here there is return but only when refugees findthemselvesinevengreaterperilabroad.Thiswasthecase,forexample,ofthe1.2millionIraqiswhofledtoSyria,manyofwhomreturnedwhenanevenmorethreateningcivilwarbrokeoutinSyria.Bycontrast,withtheimportantexceptionofpersecutedminorities,returntocountriesnotinconflictasintheWesternBalkansispossible.

SomekeyeconomicindicatorsrelatingtothethreegroupsofcountriescanalsobefoundinthetablesinAnnex2.Thetablesshowthat,whilesomecountries’economies–suchasSyriaand the Central AfricanRepublic are devastated by conflict – several of the countries inconflictshowsomeprogressonsocial indicators,andsustainedmoderategrowthand/orsporadicallyrapideconomicgrowth.Thissuggeststhateveninthosecountries,foraselectfew,orinsomelocations,voluntaryreturnisnotalwaysoutofthequestion.Butagain,theseareexceptions,andveryunlikelytobecomemassmovements.

b) FinancialIncentivesThistakesustothequestion‐canfinancialincentivesmakeadifferenceininducingfailedasylum seekers and refugees to return to these difficult environments? In a recent blog,Harvard'sLantPritchett37(aformerWorldBankresearcher)writes"thefactthattherearemassivewagegainstoworkersmovingfromdevelopingcountrytodevelopedcountrylabormarkets(ontheorderofPPPadjusted$15,000peryear,perlow‐skillworker)isnowoneofthemost robust facts in development economics. Thismagnitude ofwages gains,whichimplyworkerswouldincreasetheiraverageearningbyroughlyafactoroffour,isconsistentwith observational data, econometric comparisons of equivalent workers across labormarkets and adjusting for selection, occupational data, randomized experiments, andmacroeconomicdata.”Thisdifference inrealwagespersistseven though immigrantsarepaidlessandhaveloweremploymentratesthancomparablenatives.Howmuchwould it take to compensate theworkerwho returns home to a developingcountry?Forexample,PashaandAltaf(1987)estimatelifetimeearningsintheGulfintheearly1980’sandcomparedthemtohomeearningsinPakistan.Accountingforaverageageandotherobservables,theyconcludethatonlyifwagesof immigrantworkersintheGulfwerecutbyover50%,wouldtheyinducereturn.Applying Pritchett’s more recent estimate of the difference in earnings at home in adevelopingcountryandabroadinadevelopedcountry,iftheimmigrantisfullyemployedthroughouttheirstayandtheirworkingtimehorizonis25years,thepresentvalueoftheimmigrant’sforegoneearningsifheorshereturnsisapproximately$262,000atadiscountrateof3%.This sum is at least10 times larger than thehighestknowngrantoffered to

37 Lant Pritchett, Labor Mobility and Wages of the Rich Country Poor, Part One: Analysis andImplicationsoftheMarielBoatlift,CenterforGlobalDevelopment,June2017

36

encourage returnandperhaps100 times larger than the typical grant38as reported in arecentreportbyGerber(2017).Eventhisisanunderestimateoftheforegoneearningsonreturn since the calculation assumes that the immigrant will not receive a pension onretirementinthehostcountryandthatheorsheplacesnovalueonthefutureearningsoftheiroffspring,whicharehigheriftheystay.Thecaseofarefugeeisdifferent,sincerefugeestakelongertointegrateinlabormarketsthan typical migrants and are paid less on average. However, even applying a furtherdiscount to expected earnings of 30%39 and even if one assumes that the refugee willimmediatelyfindajobwhentheyreturn,theessentialmessageofthecalculationdoesnotchange,asthedifferenceinthepresentvalueofexpectedearningsremainsinthevicinityof$200,000. It is worth noting that, in instances where the refugee has access to welfarebenefitswhenheorsheisnotworking,aswellastohealthandeducationservices,thosebenefitsalonemaybeworthmorethanwhattheywouldearnathome.Thesecalculationsemploybroadaverages.Inpractice,thedifferencesinearningsstreamsvarygreatlydependingonthecountryoforiginanddestination.Comparablerealwagedatais not available, but can be approximated using data on per capita income, which arereportedinTable6.Thedifferenceinwagesbetweenthedevelopingandadvancedcountriesislikelytobeabout15%smallerthanissuggestedbythedataintheTable,reflectingthefact

38Gerver,M (2017)PayingRefuges toLeave.Political Studies. “In2007 theSwedishgovernmentprovided$7,150tofamiliesagreeingtorepatriatetoAfghanistan(UNHCR,2007).Ayearlater,theGhanaiangovernment,workingwiththeUN,gaverefugees$100toreturntoLiberia(Omata,2011).Soon after,Denmarkbeganoffering $18,700 to anyone returning to Iraq, Iran, and Somalia (TheTelegraph 2009). In 2010 the British National Party, in an election campaign, promised to give$78,000tomigrantsorrefugeeswhoagreedtoleavethecountry.TheBNPwasneverelected,butin2011theUKgovernmenthandedover$3,500incashto familiesagreeingtoreturntoZimbabwe(Weber, 2011). More recently, Australian Prime Minster Tony Abbott proposed paying asylumseekers $10,000 to go back to their countries of origin (Whyte, 2014) and in 2013 the Israeligovernmentbegangiving$3,500tothousandsofasylumseekerswhoagreedtorepatriate.Thosewho refused to repatriatewere given $3,500 to accept a one‐way ticket to Uganda, Rwanda, orEthiopia,wheretheywouldbeunabletoobtainanylegalstatus.Inallsixofthesecases,andmanymore(Riiskjaer,2008andBlacketal,2011)”

39UsingdatafromtheEuropeanSocialSurvey,andbasedonareviewoftherelevantliterature,anIMFreport(IMF,2016)concludesthat,afteraccountingforobservables(age,education, languageskills,hostcountry,time)refugeeshaveemploymentratesthatare22percentagepointslowerthanthoseofnativeworkersoverthe6yearsafterarrivaland2percentagepointslowerafter20years,and their earnings are, respectively, 26% and 2% lower. Combining the lowerwages and loweremploymentrates, refugeescanexpect their laborearnings tobeapproximately25% lower thancomparablenativesonaverageovertheirfirst20yearsinthecountry.However,sincetheexpectedearningsstreamisbackloaded,thepresentvaluedifferenceinearningsisgreaterstill.Applyingadiscountrateof3%,thepresentvalueofthedifferenceinearningsstreambetweenarefugeeandanativeiscloserto30%.

37

thatdevelopingcountriestendtohavesomewhatlowerlaborforceparticipationratesandalsoahigherlaborshareinnationalincomethanadvancedcountries40.Table6:GrossDomesticProductPerCapitaBasedOnPurchasing‐Power‐Parity(PPP),USDollars,2016Afghanistan 1,919Albania 11,840CentralAfricanRepublic 652DemocraticRepublicoftheCongo 773Eritrea 1,410France 42,314Germany 48,111Iraq 17,944Italy 36,833Kosovo

Myanmar 5,832Serbia 14,493SouthSudan 1,657Sudan 4,447Sweden 49,836Syria n/aUnitedStates 57,436Source:IMFWEO,WorldBankWorldDevelopmentIndicators

40Reflectinganaccountingidentity,averagewagesdifferfromaveragepercapitaincomesreflectinglabor force participation rates and the share of labor in national income: other things equal, thehigherthelaborforceparticipationrate,theloweriswagerelativetopercapitaincome,and,otherthingsequal, thehighertheshareof laborinnational income,thehigherthewagerelativetopercapitaincome.ThereferencedataforlaborforceparticipationistakenfromtheWorldBank(WDI,2016), which shows 59% from lower middle‐income countries (which in our view is the mostrepresentativeoftoday’ssituationincountriesthatarethelargestsourceofrefugees),and61%inhigh‐incomecountries. In a comprehensive surveybyGuerriero (2012) sheprovides apreferredestimateofthelaborsharewhichaccountsforself‐employmentincomeandwhichyields75%forlowermiddle‐incomecountriesand69%forhigh‐incomecountries.Theseestimates,takentogether,implythatwagesinlower‐middleincomecountries–includingself‐employmentincome‐are27%higherthanaveragepercapitaincomesonaccountofthefactthatthelaborshareinincomeishigherthanthelaborforceparticipationshare.Similarly,inhigh‐incomecountriestheaveragerealwageis13% higher than per capita income. Thus, the difference in realwages between developing andadvancedcountrieswouldbeexpectedtobesome15%smallerthanthatofpercapitaincomes.

38

Itisneverthelessclearthatamongcountriesthatarethesourceoflargenumbersofrefugees,except for Albania, Serbia, and Iraq, the difference in wages with respect to advancedcountries is far largerthanthatsuggestedbyPritchett.Forexample,averagepurchasing‐power‐adjustedwagesinFrancealmostcertainlyexceedthoseinAfricancountriesandinAfghanistanbyafactoroftenormore.IncomenumbersforSyriaarenotavailable.However,even before the war, Syria’s PPP‐adjusted per capita income was about 20% of that inSwedenorGermany.Bycontrast,thedifferenceinwagesbetweenFranceandIraqmaybeinthevicinityof2,suggestingthatavoluntaryreturntoapost‐conflictoil‐richIraqmaybearealisticpossibilityformanyIraqirefugees.Suchacourse,however,isunlikelyintheothercountriesincludedinthissurveyintheforeseeablefuture.Thisfinancialcalculationdoesnotexcludethepossibilityofindividualsreturningforpersonalreasons,orthepossibilitythatpost‐conflict rapid economic growth sustained overmany years couldmaterially changeprospectsinsomecountries.Thisleadsustoconsiderthedevelopmentpolicydimensionofreturn.ReturnandDevelopmentCandevelopmentaidcreatetheconditionsforrefugeestoreturnsuccessfullyandtowanttoreturn? The expectation that humanitarian assistance or development tools (capacitybuilding,grants,loans,andpolicyadvice)canplayasignificantroleinpromotingthereturntocountriesinconflictisunrealistic.Thepurposeofdevelopmentpolicyisnotandshouldnot be to encourage return. Indeed, it is quite possible that in poor countries advancingincomes will spur increased migration initially on account of the “migration hump 41 ”(Clemens&Hunt2017,Dadush,Demertzis&Wolff,2017).However,countriesoforiginthatareunsafe todaycouldbecomesafe in the future, as conflictsare resolved.Developmentpolicycanhelpthosethatdoreturnreintegratesuccessfully,acceleratereconstruction,andreducethelikelihoodofarecurrenceofhumanitariancrises.Returningrefugeescanplayarole–evenifitislikelytobeamodestone‐inthereconstructionanddevelopmenteffortintherightcircumstances.Refugeeflowsaretriggeredbyconflicts,especiallycivilconflicts.Theseconflictsthreatencivilians with bodily harm and arbitrary treatments as the rule of law breaks down.AccordingtoastudybyCollierandHoffer(1998),ofconflictsthatoccurredbetween1960and1992,conflictswereaccompaniedbyasharpdeclineinGDPpercapita,‐2.2%ayearversusacounterfactual.Theauthorsalsofindthattheincidenceofcivilconflictsisinverselyrelatedtopercapitaincomes,aftercontrollingforfactorslikelytoinduceconflict,suchasethnicfractionalization.Itwouldbeinterestingtoverifywhetherthecorrelationisasstrongasitusedtobe,sincethelargestsourcesofrefugeestodayarefromSyriaandIraq,whicharemiddle‐incomecountries.Still,asevident fromthedatapresentedabove, todaytherearealmostnoasylumseekersthathailfromhigh‐incomecountries,whiletherearemanythatoriginateinthelow‐incomecountries.

41Thereisconsiderableempiricalsupportforthenotionthatmigrationintensityislowatlowlevelsofincomepercapita,risesaspercapitaincome(PPPadjusted)risestowards$10000ayear,anddeclines thereafter.Mostcountries thatarea largesourceof refugeesexhibitpercapita incomesbelowthisthreshold.

39

Therearegroundsforhopingthatadevelopmentassistanceeffort–comprisedofaid,privateand public investment and the fostering of a business‐friendly economic environment –whichsucceedsinstimulatingsustainedeconomicgrowthcanreducethelikelihoodofcivilconflict. However, a policy still needs to contend with two issues. First, the scope ofdevelopmentagenciesamidconflictisevidentlylimited.Second,inthebestofcircumstances,the development thatmaterially changes theprobability of conflict is a long‐termmulti‐decadepropositionwhiletherefugeeproblemispressingtoday.Oneshouldnot,however,dismissthepossibilitythat–onceconflictabates(whetherbecauseofdiplomacyorexhaustionofoneorbothparties)–recoverycanbeveryrapid. In theirstudy,CollierandHofferfindthatpercapitaincometendstobouncebackafteraconflictandthe longer the conflict the sharper the recovery. Moreover, historical studies of largemigration flows suggest that emigration can create more employment opportunities forstayers42and/or increasewagesofstayers43. In theirstudyofmassmigration in the19thcentury,HattonandWilliamsonconcludethatemigrationfromIrelandandSwedenwasasignificantcontributortothegrowthofwagesinthosecountries,whichtendedtoconvergegradually towards thoseof theUnitedStates, andwereonlypartlymoderatedby capitaloutflowstowardstheNewWorld.(1998)Cantherecoverybesorapidastoattractlargenumbersofrefugeestoreturn?Thereisnosingleanswertothisquestion,asitdependsonthecountryoforigin’sendowment,itslevelof development, who left and why, who prevails in the war, the rate at which conflictdeescalates,etc.Onecanspeculate,forexample,thatahighermiddle‐incomecountrysuchasIraq–whichhasoneoftheworld’slargeoilendowments–andwhichhasseenmassivedisplacementof itspopulationanddestructionof its capital stock in recentyears,wouldpresent returnees, especially those who have skills and capital, with considerableopportunitiesinaperiodofreconstructionandrapidgrowth.Ontheotherhand,aresource‐poor low‐income country in Africa where the exodus was small as a proportion of itspopulationmayremainunattractivetoreturneesevenduringthereconstructionphase.It should be noted that the civilwar in Syria,which had a population of 22million, hasproducedanestimated4.8millionrefugeessince2011andresultedinmassive,impossibletomeasure, capital flight. If peace returned andwith‐it capital, the reconstruction effortwouldrequirelaborthatisinshortsupply.Still,evenin2010–beforethecivilwar–Syria’sPPP– adjustedper capita incomewasabout10% lower than Jordan’s.Assumingapost‐conflictSyriacaneventuallyreturntoalevelofincomepercapitaroughlyequaltothatofJordantoday,$9,000–thedifferencewithrespecttoJordanandLebanonwouldbesmall,thatwithTurkeyis2‐3to1andtoadvancedcountrieswouldbe4‐6to1.Thissuggeststhatvoluntarymass return toSyria fromneighboring countries suchas Jordan,Lebanon, andTurkeywhereSyrianrefugeesliveindireconditionsisquitepossible–dependingontheregimethatprevailsinthewar,andthatthiswillcreatetheneedforlarge‐scaledevelopment

42ExamplescomefromSriLankaandAlbania(Lucas2005)43ExamplescomefromPakistanandthePhilippines(Lucas2005)

40

assistanceforwhichagenciesshouldbeprepared.However, itappearsfar lesslikelythatSyrianswillreturnvoluntarilyfromthemoreadvancedEuropeancountries.Canreturnitselfcontributetodevelopment?Intheory,returncancontributetodevelopmentin three ways: innovation (skills, entrepreneurship, mores), capital, and networks thatstimulatetradeandinvestmentthroughdiaspora.Studiesshowthatmigrantsreturnforcomplexreasons,andthatincomeisnottypicallytheprimary reason; instead lifecycle considerations and opportunities for investment andentrepreneurshipseemmoreimportantinsomecontexts.(Gibson&McKenzie2011)Otherreasonsincludecontributingtothesocialandeconomicdevelopmentoftheirhomecountryinnumerousways.(Avle2014)Thereareseveralexamplesofreturneesbringingimportantskills to bear and using their capital and network in a relatively secure environment toconnectthehomecountrywiththeworldinimportantsectors,asinChina,India,andGhana.(Avle2014)Diasporanetworkscanalsobeanimportantsourceofbusinessopportunities,linkingthecountryoforiginandthehostcountryincreativeandlucrativeways.(Dadush2015)However,while inprinciplethemigrantbringsskills, this ismoreoftennotthecase,andwheretheydo,theskillsmaynotbetransferable.Manyreturneeshavestayedtoolittletimetoacquirenewskillsorreturnafteraverylongperiodandstruggletoadjusttotheworkculture of the origin country.Most Turkishworkers in Germany (90% according to onestudy,Paine1974)wereemployedinoccupationsthataddedlittletotheirhumancapital.Skillsacquiredmayalsonotbeapplicableorinsufficientdemand.Intheprefaceofaneditedvolumeonreturnanddevelopment,Kingwrites“Although it isrepeatedlyacknowledgedthatreturningmigrantscanbeaforceforeconomicinnovationandsocialchange,thegeneralconclusionisthattheextenttowhichthispotentialisrealizedisdeterminedverystronglybytheeconomicenvironmentoftheareainwhichtheyresettle.”(King1986)The available evidence on economicmigrants suggests that very few returnees want toreturn to work in agriculture, and few work in industry. Large numbers become self‐employed.Evidenceon return from theGulf and theexperienceof returnees inPakistansuggestthatnon‐employmentoverextendedperiodsiscommonandoftenvoluntaryastheyhadaccumulatedsufficientsavingsandcouldtaketimetofindtherightopportunitywhichoftenturnsouttobeself‐employment.(Lucas2005)Voluntaryreturneesthathavehadtimetobuildsavingsandskillsabroadandaremotivatedtoreintegrateintheircountryoforiginaremorelikelytocontributetodevelopment(Blacketal.2003).CasestudiesofeconomicmigrantsreturningtoItaly,Portugal,Spain,Yugoslavia,Pakistan,Morocco,Algeriaandothers, suggest that capitalbroughtback,which isoftendifficult todistinguish from remittances in statistics, is dedicated to support of family, housing,investmentinsmallbusinessessuchasashoporabar,generallyinservices–inthatorder.However,littlegoestoagricultureorindustry.So,theevidenceonthecontributiontothedevelopmentofreturningeconomicmigrantsismixedandmuchdependsonwhoreturnsandthecircumstancesinthecountryoforigin.

41

Thereislittlereasontobelievethatthecontributionofareturningrefugeeorasylumseeker–especiallyonewhohasnothadtimetobuildcapitalorskillandwhoreturnsininsecureconditions–willbeanybetter.ConclusionThispaperhasarguedthatthecostsofhostingrefugeesarefront‐loadedwhiletheeconomicandfiscalbenefitsthataccruefromtheireventualintegrationinthehosteconomyareback‐loaded.Theeconomicbenefitsassociatedwithrefugeeswillnotmaterializeunconditionally.The singlemost important step that countries can take tomaximize those benefits is toacceleratethevettingprocessandallowasylumseekersandrefugeestoworkasquicklyaspossible. In addition, modest investments in language training, counseling and revisingoverlyrestrictiveregulationsthat impedetherecognitionof therefugees’credentialscanpayhighdividends.There is a strong legal case tobemade for rejectingasylumapplications thatare clearlyspecious. However, in many advanced countries today, the purely economic case forreturningmigrants(asdistinctfromthelegalcase)isweak,oncost,macroeconomic,labormarket,andfiscalgrounds.Thisistrueformostcategoriesofmigrantsbutisespeciallytrueofasylumseekerswhohavebeenleftinlimboforaprotractedperiodandforrefugees(whocanbeencouragedtoleaveorforcedunderthecessationclause).Thereisno“onesizefitsall”returnpolicy, justasthere isnosingleoptimal immigrationpolicy. However, policies regarding the return of refugeesmust be framedwith broaderpolicytowardsimmigration.If,asisthecaseinmostadvancedcountriestoday,aginganddecliningnativepopulations imply that thedemand formigrants isbound to increase, itmakeslittlesensetoincurthehighfiscal,politicalandhumanitariancostsassociatedwithreturningrefugeesonlytofindthatmoreeconomicimmigrants,manyofwhomwillarriveillegally,willbeattracted.Returnthatistrulyvoluntary(asdistinctfrom“softdeportation”)canonlybeexpectedinspecificinstancesorwherepersonalconsiderationsprevail.Migrantsthathavebeengiventheopportunitytobuildskillsandsavingsandthatretainstrongfamilyand emotional connections with their country of origin are the most likely to returnvoluntarily.Most,thoughnotall,countriesthatarethesourceoflargenumbersofasylumseekersarepatentlyunsafeandincapableofaffordingthereturningmigrantalivelihood.Financial incentives are unlikely to play a significant role in spurring return, except inspecificinstanceswherepersonalconsiderationsprevail.Theexpectation thathumanitarianassistanceordevelopment tools (grants, loans,policyadvice) can play a significant role in promoting the return to countries in conflict isunrealistic.Thepurposeofdevelopmentpolicyisnotandshouldnotbetoencouragereturn.However,countriesoforiginthatareunsafetodaycouldbecomesafeinthefuture.Inthatcase,an increasedreturn flowofmigrants ispossible,even though– inmost instances–voluntary return from advanced countries to a poor location is unlikely to be large.Developmentpolicycanthenhelpthosethatdoreturnreintegratesuccessfully,acceleratereconstruction,andreducethelikelihoodofarecurrenceofhumanitariancrises.Returning

42

migrantscanplayarole–evenifitislikelytobeamodestone‐inthereconstructionanddevelopmenteffortintherightcircumstances.

43

Annex1:DefinitionsThereisalegalobligationunderthe1951UNConventiononRefugees(UNCOR),towhichnearly150UNmembersareparty,nottoreturntherefugeetohisorhercountryoforiginifthatplacestheminharm’sway(non‐refoulement).Atthesametime,thereisoftenconfusionin the broadpublic about the distinctionbetween an asylum seeker (who canbe legallyreturnedifhisorherasylumapplicationfails)andarefugee.Thereisalsoconfusionbetweeneconomicmigrants–whoareadmittedbasedonverydifferentcriteriathanarefugee,andwhomayarrivelegallyorillegally.Fordefiningpolicy,itisimportantattheoutsettoclarifythedistinctionbetweeneconomicandforcedmigrantsand,withinforcedmigrantsbetweenasylum seekers and refugees and, among returnees, between forced and voluntaryreturnees.AnnexTable1TermDefinitionTerm DefinitionEconomicmigration44 Economicmigrationisdefinedasachoiceto

move to improve thestandard of livingbygainingabetter‐paidjob.

Forced migration 45 and the forciblydisplaced

Forced Migration is “a general term thatrefers to the movements of refugees andinternallydisplacedpeople(thosedisplacedbyconflictswithintheircountryoforigin)aswell as people displaced by natural orenvironmentaldisasters,chemicalornucleardisasters,famine,ordevelopmentprojects.”.Peoplethatarepartofaforcedmigrationareforciblydisplaced.

RefugeeStatus Refugees include individuals recognizedunder the 1951 Convention relating to theStatus of Refugees, its 1967 Protocol, the1969 Organization of African Unity (OAU)ConventionGoverningtheSpecificAspectsofRefugeeProblemsinAfrica,thoserecognizedin accordance with the UNHCR Statute,individualsgrantedcomplementaryformsofprotection, and those enjoying temporaryprotection. The refugee population alsoincludespersonsinrefugee‐likesituations.(Arefugee,pertheConvention,isunableorunwillingtoreturntotheircountryoforiginowing to a well‐founded fear of beingpersecuted for reasons of race, religion,

44Source:BBC45Source:ColumbiaUniversity,ForcedMigrationLearningModule

44

nationality,membershipofasocialgroup,orpoliticalopinion)(UNHCR)

Asylum Asylum‐seekersare individuals who havesought international protection and whoseclaims for refugee statushavenotyetbeendetermined, irrespective ofwhen theymayhavebeenlodged.(UNHCR)

ForcedReturn Returnees who intendedpermanentmigrationbutwereforcedtoreturn. Theirpreference was to remain abroad butbecause of external factors they wererequiredtoreturn(Gmelch1980)

VoluntaryReturn Returneeswhoreturnvoluntarilyfulfilloneofthefollowingconditions:1)Heorshefacesaclearandopenchoiceeithertoreturntohisor her country of origin or to staypermanently and integrate into the hostsociety;or2)achoicebetweenreturningtothe country of origin now in a voluntaryfashion (perhaps with financial or otherincentives) or staying and risking forciblereturnlater46.

Resettlement Resettlementisthetransferofrefugeesfromanasylumcountrytoanotherstatethathasagreed to admit them and ultimately grantthemapermanentsettlement.(UNHCR)

Annex2:TheSituationinCountriesofOriginofAsylumSeekersandRefugeesAfghanistan46Morrison,2000

45

SecurityThe security situation in Afghanistan has seen some improvement in the past years;nevertheless,thecountryisstillmiredinconstantconflict.TheAfghangovernmentstrugglestoregaincontrolofalargepartofitsterritory.AccordingtoareporttotheUSCongressbytheUSSpecialInspector‐GeneralforAfghanistanReconstruction(SIGAR),atendofMarchin2017,just59.7%ofthenation’sdistrictswereinthecontroloftheAfghangovernment.Thegovernment’scontrolof407districtswasreducedfrom72%inNov.2015to57%inNov.2016.Thereportalsopaintedapictureofstruggle.Majorroadsthat linkthecountryareheavilycontested.NumerousstrikesinAfghanistan’smaincitiesandbombingsinthecapitalKabularesymptomsofthelackofcontrol47.UNreportsindicatethatin2015,atleast3,545civilianswerekilledand7,457wereinjured48.Thenumberwashigherin2016,settingtherecordat3,498deathsand7,920injured49.EconomicOpportunityAccordingtopredictionsbytheWorldBank,AfghanistanisexpectedtogrowitsrealGDPby2.2%in2016.Therateofeconomicexpansionwillaccelerateto2.6%and3.4%in2017and2018 respectively. However, the unstable political environment and the fragile securitysituation represent large risks. Hindering progress further is the extent of corruption50,whichaffectsdecisionsinallspheres:economicdevelopment,public investment,politicaldecisions, appointments, and recruitment. Recruitment for political positions has shownstrongethnicpreference51.SocialServicesTheAfghanChamberofCommerceandIndustries(ACCI)saidworkingconditionsin2016improvedcomparedto2015.Thefindingcomesfromasurveyconductedonemployment.Althoughthissurveyshowedsomepositiveimprovementinworkingconditions,however,security and infrastructure are stillworrying.According toACCI, private sector’s trust isreducingeveryday.ResidentsinAfghanistanraiseconcernsoveracuteforeigncurrencyshortage.UNHCRdoesnotpromotethereturntoAfghanistan.However,thosewhodecidetoreturnwillbeassisted.UNHCR’smonitoringsurveyshows that the typical cashgrantof$400perpersoncoversexpenditureforabouttwotothreemonths.Thevastmajorityofrefugees(75%)spentthecashgrantonmeetingimmediatehumanitarianneeds.15%wereabletomakeinvestments

47EIUCountryOutlook:Afghanistan,June20th201748AfghanistanHadRecordCivilianCasualtiesin2015,U.N.Says,Feb142016,DavidJolly,TheNewYork Times, https://www.nytimes.com/2016/02/15/world/asia/afghanistan‐record‐civilian‐casualties‐2015‐united‐nations.html49Afghanistan’sCivilianCasualtiesHitRecordHighin2016,PatriciaGossman,HumanRightsWatch,https://www.hrw.org/news/2017/02/06/afghanistans‐civilian‐casualties‐hit‐record‐high‐201650BBCMonitoringcitedbyEIU,Afghanistanpolitics:Dailyexaminesinsecurity,domesticchallenges.June15,201751BBCMonitoring,citedbyEIUreport:Afghanistanbusiness:Assessingoperationalchallengesforbusinesses,May4th2017

46

inbusinessorsustainablelivelihoods,11%saidtheycouldbuildahouse.“Thosewhoaresurveyedbytelephonethreemonthsafterthereturn,52%saidtheywereabletoreturntotheirprovinceoforigin.Theremaining48%saidtheychoseanotherprovinceduetolackofshelter,land,livelihoodopportunitiesorinsecurityintheirhomeareas.”UNHCRalsosharesconcernsoverthereintegrationneeds.It islargerthanthecapacityAfghanistanhasrightnow.Manyrefugeesarereturningtouncertainfuturesandhaveto“rebuildtheirlivesamidstincreasinglevelsofinternaldisplacement,insecurityandlevelsofviolencenotseensincethefalloftheTalibanin2002.”(2017)SyriaSecurityThewarinSyriahaskilledmorethan400,000peopleanduprooted12millionsinceitstartedin201152. “AsofMarch2017,more than5millionSyrianshave fled thecountryand6.3millionpeoplearedisplacedinternally.”Asofnow,Syriaisstillundergoingsevereattacksandairstrikes.AchievingceasefireisstillevadingSyria.Attheendof2016,aceasefirewasdeclaredbutnoneof theparties involvedhad respected it53.According toHumanRightsWatch’sreportonSyria,“unlawfulattacksonciviliansbyallpartiesoftheconflictpersistedthroughouttheyear.”“Governmentforcesusedatleast13typesofinternationallybannedclustermunitionsinover400attacksonopposition‐heldareasbetweenJuly2002toAugust2016, killing and injuring civilians, including children54.” BetweenMarch 2011 and June2016,at least12,679personsdiedfromtorture,abuse,starvation,beatinganddiseaseingovernmentdetention.EIU’seconomicoutlookfor2017‐2021predictedthattheintensityofconflictinSyriawouldabateismoreconfinedtoruralareas.EconomicOpportunitySyria’sfiscalsituationwillremaindire.Itsoilandtaxrevenuewillcontinuetobedepressed.Substantialneeds fromthemilitarywillpersist Infrastructure isdamagedbothby ISandcoalition airstrikes 55 . These factors hinder the efforts of reconstruction and economicdevelopment.Thewarbroughtenormousdamageto theeconomy,domesticdemandhasfallenduetoinparttorefugeeexodus,andtheoilsectorismostlyoffline.Thegovernmenthas financed its largedeficitbyprintingmoneyanddrawingdown foreign reserves.TheSyrianpoundhaslostfour‐fifthsofitsvalue,andreserveshavedroppedfrom$20billionto$1billionsince2010.TheIMFestimatesthatSyria’sGDPtodayislessthanhalfofwhatitwasbeforethewar56.InJuly2017,thewheatharvestwassettofallshortofgovernment

52CNNhttp://www.cnn.com/2013/08/27/world/meast/syria‐civil‐war‐fast‐facts/index.html53EIUSyriapolitics:QuickView‐Proposalforde‐escalationzonesdrawnupatAstanatalks54 Human Rights Watch (HRW) Syria, Events of 2016, https://www.hrw.org/world‐report/2017/country‐chapters/syria55EIUSyriaeconomy:QuickView‐Wheatharvestsettofallshortofgovernmentforecast,July201756Syria’snewwarmillionaires,Economist June2017https://www.economist.com/news/middle‐east‐and‐africa/21722833‐new‐elite‐peace‐would‐be‐bad‐business‐syrias‐new‐war‐millionaires

47

forecast, as military operations against IS had badly affected farmers in major grain‐producingareas,andISleviesheavytaxesonfarmersintheareasunderitscontrol57.Thelong‐termoutlookforSyria’sagriculturalsectorisslightlymorepositive,accordingtothenewestviewonSyriaeconomybyEIU.Therehasbeenanimprovementinconditionsinrebel‐heldareas,whichcouldpotentiallyleadtohigherproductionandincreasedeconomicactivity.SocialServicesThe infrastructure damage is enormous. In some instances, humanitarian aid sometimescannotreachtheaffectedpopulations.Aidagencieshavetogothroughextensivecontrolsbythe Syrian government to deliver life‐saving medicines. The UN reported that thegovernmentprevented80,000medicaltreatmentitems,“includingdiarrheakits,emergencyhealthkits,antibiotics,andothermedicinesfromgoingintobesiegedareas.”SomaliaSecurityCiviliansinSomaliaaresufferingfromabusesbyallwarringparties,andhavelimitedaccesstobasicservices.CitizensarelivinginthefearofmasssecuritysweepsbySomalia’snationalintelligenceagency,eventhoughithasnolegalmandatetoarrestordetainpeople.Childrenandwomenareinworstdanger.Securityforcesarbitrarilydetainandrecruitchildrenanduse them as informants to identify Al‐Shababmembers. Internally displacedwomen areextremelyvulnerabletorapebyarmedmen,“includinggovernmentsoldiersandmilitia.”Protection for thembasicallydoesnotexist58.According toUNHCR,Somaliahasover1.4million internallydisplacedpeople. In theyear2016, some24,500 refugees andasylum‐seekerswereregisteredinSomalia.RefugeereturnfromKenyabeganin2014.Voluntaryrepatriationnumberiscloseto40,000Somalinationalsfrom2014toDecember201659.EconomicOpportunityAccordingtoBerlin‐basedTransparencyInternational,Somaliaisoneoftheworld’smostcorruptcountries. ImprovedgovernancecouldenableSomalia’seconomy togrowon thebasis of its oil and gas reserves. Ongoing droughts continue to drive hungry and thirstyrefugees to surrounding countries, and large parts of the population are in need ofhumanitarian aid. The agriculture sector contributes to over two‐thirds of its GDPwhileindustryonlymakesupfor7%in201360.AccordingtotheIMF,Somaliahasaveryhighyouth

57EIUSyriaeconomy:QuickView‐Wheatharvestsettofallshortofgovernmentforecast,July201758 Human Rights Watch, Somalia Events of 2016, https://www.hrw.org/world‐report/2017/country‐chapters/somalia59UNHCR,Somalia,http://reporting.unhcr.org/node/2550?y=2016#year60 CIA The World FactBook, https://www.cia.gov/library/publications/the‐world‐factbook/geos/so.html

48

unemployment rate, contributing to irregular migration and participation in extremistactivities,includingAl‐Shabaab.Joiningmilitantjihadistgroupisviewedasanotherformofemployment61.SocialServicesThelackofinfrastructureandbasicservicehindersIDPsettlements.Ontopofthat“urbanareasarebeingoverwhelmedwithnewarrivals62”.Accesstobasicneeds,suchashealthandeducation,areunmet.SouthSudanSecuritySouthSudan’scivilwarbeganinDecember2013andcontinueswithseriousabusesagainstcivilians.ApeaceagreementwassignedinAugust2015buttheceasefirewasnotachieved63.OnMay 25th, 2017, South Sudan President declared a ceasefire. According to theWorldReport byHumanRightsWatch, South Sudanese “government soldiers killed, raped andtortured civilians as well as destroying and pillaging civilian property duringcounterinsurgencyoperationsinthesouthernandwesternpartsofthecountry,andbothsides committedabusesagainst civilians inandaround Jubaandotherareas.UNSpecialAdvisor on the Prevention of Genocide Adma Dieng said the ongoing violence hadtransformedintoan“ethnicwar”andwarnedofa“potentialforgenocide64”Ontopoftheprecarious living situation, security and logistical challenges posed constraints to thedelivery of much‐needed humanitarian assistance 65 . According to Council on ForeignRelations,theestimatednumberofpeoplekilledsinceDecember2013isover50,000,andover1.6millionpeopleareinternallydisplaced66.EconomicOpportunitySouth Sudan has abundant natural resources. Before its oil production fell sharply, thegovernmentreliesonoil for itsrevenue. Italsohasvery fertilesoilsandabundantwatersupplies.South Sudan has struggled with economic development since its independence and itseconomicconditionshavedeterioratedsinceJanuary2012whenthegovernmentdecidedto

61 IMF, Six Things to Know about Somalia’s Economy, April 11, 2017,http://www.imf.org/en/News/Articles/2017/04/11/NA041117‐Six‐Things‐to‐Know‐About‐Somalia‐Economy62UNHCR,Somalia,http://reporting.unhcr.org/node/2550?y=2016#year63 Human Rights Watch, World Report, https://www.hrw.org/world‐report/2017/country‐chapters/south‐sudan64 Human Rights Watch, World Report, https://www.hrw.org/world‐report/2017/country‐chapters/south‐sudan65UNHCR,SouthSudanhttp://reporting.unhcr.org/node/255366 Council on Foreign Relations (CFR) https://www.cfr.org/global/global‐conflict‐tracker/p32137#!/conflict/civil‐war‐in‐south‐sudan

49

shut down its oil production67 . UNHCR reported that in 2016, South Sudan’s economicsituationdeterioratedfurtherandthecostoflivingroseexponentially68.SocialServiceSouthSudanhasveryhighmortalitycausedbyAIDSandhighriskofinfectionofdiseases.Educationexpenditure is lowandthe literacyrate isalsovery low(27%in2009).SouthSudanhaslittleinfrastructure.AccordingtotheCIAFactbook,thereareapproximately200kilometersofpavedroads.Electricityisproducedmostlybycostlydieselgenerators.Goodsandservicesaremostlyimportedfromsurroundingcountries.69SudanSecuritySimilar to South Sudan, Sudan’s government forces have raped, killed civilians, anddestroyedhundredsofvillages.InSeptember2016UNfoundthatviolencehasdisplacedupto 190,000 people and many of them are not accessible to humanitarian agencies 70 .GovernmentforcesandarmedrebelsinSouthernKordofanandtheBlueNilecontinuetobeengagedinarmedconflictforthefifthyearin2016.Civiliansinpopulatedareasweresubjecttoindiscriminatebombing,especiallyduringMarchthroughJunein2016.Citizensalsofacearbitrarydetentions, ill‐treatment,andtorture.Sudan’sNationalIntelligenceandSecurityService is known for detaining activists, students, lawyers, doctors and those who areperceivedtobeneededinsomecapacitybythegovernment71.Manydetaineesarefacingill‐treatment.Females are subjected to sexualharassmentby securityofficers.According toUNHCR,thereare2.7millionpeopleofconcerninSudan(includesrefugees,asylum‐seekers,IDPs, returned refugees, returned IDPs, stateless persons, and other concern), a lowernumber than2015.About37,000refugees returned toSudan in2016.At the same time,therewereover2millionIDPs,over420,000refugeesandover16,000asylum‐seekersinothercountries.EconomicOpportunityOiloutputinSudanhasbeenlowduetocivilwar,poorinfrastructure,andlowproductivity.Compared to South Sudan, Sudan also has fewer oil resources; nevertheless, it still hasabundantresources72.Sudanranks186thof190statesintheWorldBank’sDoingBusiness

67 CIA Factbook, South Sudan, https://www.cia.gov/library/publications/the‐world‐factbook/geos/od.html68UNHCR,SouthSudanhttp://reporting.unhcr.org/node/255369 CIA Factbook, South Sudan, https://www.cia.gov/library/publications/the‐world‐factbook/geos/od.html70UNHCR,Sudan,http://reporting.unhcr.org/node/253571HumanRightsWatch,https://www.hrw.org/world‐report/2017/country‐chapters/sudan72EIUCountryOutlook,Sudan,June19th2017

50

Rankings, with particularly poor scores for starting a business, getting electricity andregisteringproperty.Agricultureslowedin2016andisexpectedtodobetterin2017.Lackofrainaffectstheoutcomeoftheagriculturalsector.CIA’sFactbook73reportsthat46.5%ofSudan’spopulationislivingbelowthepovertyline.Only35%ofitspopulationhasaccesstoelectricity.SocialServiceItisdifficultforthestaffofinternationalhumanitarianaidagenciestoobtaintravelpermitsinSudan,severelyhamperingthereliefeffort.TheDemocraticRepublicoftheCongo(DRC)SecurityThesecuritysituationintheDemocraticRepublicofCongohasbeenpoorsince2012.AnattempttointegrateaTutsirebelgroupintotheCongolesemilitaryfailedandpromptedthedefection and formation of the M23 armed group. The renewed conflict led to largepopulationdisplacementandhumanrightsabuses.Furthermore,thePresidentoftheDRC,JosephKabilaisbarredfromrunningforathirdterm,buttheDRCgovernmenthasdelayednationalelectionoriginally slated forNovember2016.The failure toholdelection fueledsporadicstreetprotestsbyKabila’sopponents74.AlthoughadealsignedbyrepresentativesofKabila’srulingpartysaidthepresidentialelectionwouldbeheldbeforetheendof2017,thePresidenthimselfhasnotendorsed thedeal.Governmentofficials repeatedlybannedoppositiondemonstrations,firedteargasandlivebulletsatpeacefulprotesters,shutmediaoutlets,andpreventoppositionleadersfrommovingfreely75.AccordingtoUNHCR,thereareover2millionIDPsintheDRC,andover450,000refugees.Inyear2016,therewereabout13,000returnedrefugeesand619,000returnedIDPs.“Dozensofarmedgroupsremainedactive ineasternCongo,manyof their commandershavebeen implicated inwar crimes,including ethnic massacres, killing of civilians, rape, forced recruitment of children andpillage.76”EconomicOpportunity

73 CIA the World Factbook, Sudan, https://www.cia.gov/library/publications/the‐world‐factbook/geos/su.html74 CIA, the World Factbook, https://www.cia.gov/library/publications/the‐world‐factbook/geos/cg.html75HumanRightsWatch, https://www.hrw.org/world‐report/2017/country‐chapters/democratic‐republic‐congo76 Human Rights Watch, DRC, https://www.hrw.org/world‐report/2017/country‐chapters/democratic‐republic‐congo

51

RegionalwarsintheDRCpreventeconomicrecovery.Still,EIU77predictsthatcopperexportandminingwillpickup.Lowlaborcost,highoregrades,andpotentialforsignificantfindswillcontinuetoattractinvestors.SocialservicesHealth,education,shelterandinfrastructure,andprotectionfromarmedconflictsarekeyobjectivesidentifiedbytheaidagenciesintheDRC78.TheCentralAfricanRepublicAccordingtoUNHCR,CARhasaboutover410,000IDPsin2016,over12,000refugeesandasylum seekers and over 34,000 returned refugees. The number of people of concern(458,607) has declined by half since 2014, but refugees and IDPs are still facing harshconditions.UNHCRsupportedthereturnof235,000IDPstotheirareasoforigin.IDPsinCARarefacingharshlivingconditions;only40%ofIDPhouseholdslivedinadequatedwellings.Lackofresourcesandaccesstoservicesledtolimitedreintegrationassistanceinareasofmajorreturn.Themostrecentceasefireagreementbetweenthegovernmentand13armedgroupsfailed2daysaftersigningtheagreement.AccordingtoEIUforecast,realGDPgrowthisexpectedtoreach4.6%in2017,fromanestimated4.3%in2016,as“thepresenceofanelectedgovernmenthelpstounlockdonorfunding,andsupportsamodestriseinpublicandprivateconsumption.”Growthisexpectedtopickupagainin2018,to5.1%,dueto“somediamond exports restart and foreign‐funded investment rises”. The speed of economicdevelopmentwillstillbehinderedbyinsecurityandthereturnofrefugees.IraqSecurityIraq’spoliticalandsectariandivisionsareprofoundasthecountryfacesenormouseconomicchallenges.According toUNHCR, thenumberofpeopleofconcern increased toaround5millionpeoplein2016.Over3.6millionareIDPsandover261,000arerefugees.WhilemanypredictthattheIraqiarmywillsooncompleteitsretakingofMosul(IS’smainstrongholdinIraq),thesecuritysituationinIraqstilllooksdire.TheInternationalFederationofJournalistsdeemsIraqasthedeadliestcountryintheworldforjournalists.AUNICEFreportdeemedIraqasoneofthedeadliestcountriesforchildren.ItscitizensarenotonlyfacingdangersbroughtbyISIS,butalsobythegovernment’scounterinsurgencymeasures.ISIScarriesoutnumerousattacks,bombings,andchemicalattacks;italsoimposesextremerestrictionsonIraqisundertheircontrol,especiallywomenandgirls.Theyarenotallowedtoleavetheirhomeswithoutamale’scompany;nordotheyhaveaccesstoadequatehealthcare.ISISalsobannedgirlsfromgoingtoschools.Yezidiwomenbecometargetsoftorture,rape,murderandsexualslavery.

77EIU,countryoutlook,DRC78UNHCR,DemocraticRepublicoftheCongo,http://reporting.unhcr.org/node/4874

52

TheIraqigovernmentfrequentlycarriesoutsummaryexecutions,enforceddisappearances,andmutilationofcorpses.Government‐affiliatedinstitutionsunlawfullydetaincitizensafterscreeninghundredsoffamilies.DeportationofresidentsanddestructionofArabhomesorevenvillagesoccur.Internallydisplacedpeoplearebeingrestrictedintheirmovement;theyarenot allowed to travel to a certain regionof Iraq and thedisputed territories andarerequiredtostayincamps79.Iraq’s largely state‐run economy is dominated by the oil sector. According to the CIA’sFactbook80,oilsectorprovides“morethan90%ofgovernmentrevenueand80%offoreignexchangeearnings.”EconomicOpportunityWiththeimprovementofsecurityandstabilityinIraq,itsGDPannualgrowthexceeded10%in 2016, which is the best performance in a decade. Iraq also was able to lower itsunemploymentratefromthehighest28%toabout15%inthepastfewyears81.Althoughgovernmentrevenuereliesheavilyontheoilsector,onlyasmallpercentageofIraqisareemployedinthesectorwhilethemajorityofIraqisstillreportlivinginpoverty82.Lackofessentialserviceslikeelectricity,cleanwater,sanitation,andhealthservicesisalsoamajorconcern for refugee’s return. Detailed statistics on Iraqi refugee returnees is not yetavailable,butmediareportthatsmallbutgrowinginfluxofreturneesisgoingbacktoIraqduetoasylumrejectionsinEurope.MyanmarAccordingtoUNHCR83,thetotalestimatedpopulationofconcerninMyanmarisabout1.3million, including 926,000 Stateless people and 375,000 IDPs. Myanmar’s nascentdemocracyanditsnewly‐electedgovernmentin2016,facesmajorpoliticalandeconomicchallenges. These include furthering the national reconciliation process, strengtheningrespectforhumanrights,protectingciviliansinarmedconflictandaddressingsomedeeplyrooteddivisionswithinthecountry84.RefugeesfromSouth‐EastMyanmarthatareresidinginThailandandtheregion’sIDPshavestillnotreturnedinnumber.TheNortheastregionofMyanmar is stillmired in fighting betweenmilitary and ethnic armed groups. People inRakhineStateinMyanmararestillindangerofarmedattacksandlockdowns.OperationsbyMyanmar security forces had led to 43,000 statelessMuslims to flee to Bangladesh anddisplacedestimated30,000withinMaungdawdistrict.EIU 85 predicts that Myanmar will achieve real GDP growth to average 7.3% a year,continuingtobeoneofthefaster‐growingeconomiesinAsia.

79HumanRightsWatch,Iraq,https://www.hrw.org/world‐report/2017/country‐chapters/iraq80CIAWorldFactbookhttps://www.cia.gov/library/publications/the‐world‐factbook/geos/iz.html81TradingEconomics,https://tradingeconomics.com/iraq/unemployment‐rate82UNDPhttp://www.iq.undp.org/content/iraq/en/home/countryinfo.html83UNHCR,Myanmar,http://reporting.unhcr.org/node/254184UNHCR,Iraq,http://reporting.unhcr.org/node/254785EIU,Myanmar,Keydevelopment,July6th2017

53

Eritrea“Sincegainingindependencein1993,EritreanshavecontinuedmigratingtoSudan,Ethiopia,Yemen, Egypt, or Israel because of a lack of basic human rights or political freedom,educationalandjobopportunities,ortoseekasylumbecauseofmilitarization86”UNHCRreported474,296Eritreanstoberefugeesandasylumseekersgloballyattheendof2015,about12percentofEritrea’sofficial3.6million‐population.ASwissimmigrationfact‐findingmissiontoEritrea’scapital,Asmara,concludedthat“proofof improved human rights conditions is still missing.” It also concluded, “involuntaryreturneescouldcountonimprisonmentandperhapstorture87.”Bylaw,Eritreanisrequiredto serve18months innational services starting at 18but in reality, the time couldbe adecade.Thefearofopen‐endedconscriptionisthemaindriverformigration.Conscriptsfaceharshtreatmentandattempttofleeissternlypunished.Theruleoflawisflimsyatbestandthe president is not constrained by institutional safeguards or processes. Free speech,expression, and association are heavily repressed. Detention of journalists and politicalopponents is not rare. Arbitrary imprisonment of less prominent citizens, as well aspersecutionofcitizenswhopracticereligionsthatEritreadiscouragesreturn.EIU88forecastsEritrea’srealGDPgrowthat4.9%in2017,buttheeconomyisperformingwell below its potential. Government mismanagement, lack of financial resources, anddroughtshindereconomicgrowthandhavecontributedtosoaringinflationoverthelasttwodecades.AccordingtoCIA’sWorldFactbook89,whileover80%ofEritreansareemployedinagriculture, agriculture only contributes to 12% of GDP in 2016. In 2004, 50% of thepopulationislivinginpoverty.In2013,only17%oftheruralareahasaccesstoelectricityand4millionpeoplearelivingwithoutelectricity.WesternBalkans:Albania,KosovoandSerbiaAlbania, Kosovo, and Serbia are part of theWestern Balkans, a group that also includesBosniaandHerzegovina,MacedoniaandMontenegro.Differentfromtheother10countriesmentionedaboveaslargesourcesofrefugees,thesethreecountriesarenotinconflict.Theyare also relatively small countries, with a combined population of around 12 million.Nevertheless, they make up for a non‐negligible part of the asylum seekers in Europe.PublicationbytheBalkanInvestigativeReportingNetwork(BIRN)‐BalkanInsightreported

86 CIA World Factbook, https://www.cia.gov/library/publications/resources/the‐world‐factbook/geos/er.html87 Human Rights Watch, Events of 2016, Eritrea, https://www.hrw.org/world‐report/2017/country‐chapters/eritrea88EIU,countryoutlook,May17th201789 CIA World Factbook, https://www.cia.gov/library/publications/resources/the‐world‐factbook/geos/er.html

54

thatAlbanianasylumseekersthinkthemigrantcrisisfromtheMiddleEastcouldalsofavortheirchancesofgettingasylumstatus90.Therealityisdifferent.Albania,Kosovo,andSerbia,areconsideredsafeandhaveverylowasylumapplicationapprovalrateinGermany(rangingfrom0.1%to0.3%inAugust201591)Incomparison,aroundthesametime,theapprovalrateforasylumseekersfromIraqandSyriawasaround89%.ThelowacceptancerateofasylumseekersinEuropehasdeterredapplicationsin2016.TheWesternBalkancountrieshaveundergoneamajoreconomictransformationoverthepast15years(Murgasovaelal2015)buttheirincomelevelsremainnear1/3or¼ofthewealthiestEuropeannations,andisnotfarfromthe$9000‐11000(PPPadjusted)whichisoftenassociatedwiththepeakmigrationtendency(Dadush,Demetris,Wolff,2017).Facingaweakexternalenvironmentandstructural frailties, in recentyears theWesternBalkancountries faced major challenges in boosting growth sustainably, reducing poverty, andredressing very large current account deficits, which are financed largely by workerremittances. The unemployment rate inAlbania and Serbia hovers around15%. Thoughsomecountrieshaveseenimprovementinthebusinessenvironmentandtakenmeasurestoimprovetheworkingofthelabormarket,inAlbaniaandSerbiapovertyhasincreasedsince2008. InKosovo,weakgovernanceseverelyhampersprivatesectordevelopmentandjobcreations.(Murgasovaelal2015).Over2017and2018,GDPGrowthintheWesternBalkansisexpectedtoacceleratewithrecoveryinEurope,toaround3.5%.Kosovowasinthecivilwarinthelate20thcentury,andthenumberofrefugeesandIDPsremainslarge.Theprotectionofhumanrightsandaccountabilityforwarcrimeshasmadeslowprogress.Roma,Ashkali,andBalkanEgyptianscontinuetofacedifficultiesacquiringpersonal documents, accessing health care, obtaining social assistance, and getting anadequate education. Between Jan. and Aug. 31st, 2016, the Kosovo Ministry of InternalAffairs registered4,534 forcedreturnees, among these forciblydeported,ofwhom1,474were children. Most minorities were deported from Germany and did not have muchassistanceafterreturningtoKosovo. HumanRightsWatchreportedthatminoritiesinSerbiafacedifficulttreatmentinallaspectsoflife.Discriminationtakesmanyforms,andsomeminoritiesfaceharassmentinhousing92.RomapeopleareoftenmentionedbyhumanrightsorganizationasthetargetsofpersecutionintheBalkanstatesbutarenotrecognizedasunderthreatbyauthoritiesincountrieslikeGermany.Policies tograntasylumstatushavebeen tightenedanddeportationandothermethodstopreventdeporteesfromre‐entryintoGermanyarebeingimplemented.Similarly, ethnic groups in Albania are discriminated against in many aspects of lives.MembersofRomani andBalkanEgyptian communitiesoften find themselves rejectedor

90 BalkanInsight, http://www.balkaninsight.com/en/article/large‐number‐of‐albanians‐still‐asking‐for‐asylum‐in‐germany‐10‐19‐201691 DW, 9‐25‐2015 Tougher laws against migrants from the Balkanshttp://www.dw.com/en/tougher‐laws‐against‐migrants‐from‐the‐balkans/a‐1874840692https://www.hrw.org/world‐report/2017/country‐chapters/serbia/kosovo#40e5f4

55

becomemarginalizedineducation,housing,employment,andhealthcare.Forinstance,someschoolsrefusetoacceptRomaniandBalkanEgyptianstudents,particularlyiftheyappearpoor.Whenadmitted,Romanistudentsareoftenseparatedphysicallyfromtherestoftheclass93.AnnexTable2‐CountrystatisticsHighIntensityConflictCountries(..indicatesdatanotavailable)Country‐HighIntensityConflict Afghanistan Syria Somalia South

SudanIraq

Population(million) 31.6 22.2 10.5 11.9 34.8GNI,Atlas$bn 21.4 .. .. 11.6 227.3GNIpercapita,PPP,$ 2,000 .. .. 1800 15100LifeExpectancyatbirthin2010 59 72 54 54 68LifeExpectancyatbirthin2014 60 70 55 56 69Labor force participation rate in2010(%ofpopulationages15+)

48 44 56 .. 42

Labor force participation rate in2014(%ofpopulationages15+)

48 44 56 .. 42

Grosscapitalformationin2000(%ofGDP)

12 17 .. .. 29

Grosscapitalformationin2010(%ofGDP)

18 28 .. 10 16

Grosscapitalformationin2014(%ofGDP)

18 .. .. 10 18

Individuals using the internet in2010(%ofpopulation)

4 20.7 1.3 7 2.5

Individuals using the internet in2014(%ofpopulation)

6.4 28.1 1.6 15.9 11.3

Merchandise trade in 2000 (% ofGDP)

72 44 .. .. ..

Merchandise trade in 2010 (% ofGDP)

35 65 .. .. 70

Merchandise trade in 2014 (% ofGDP)

41 .. .. .. 64

Electrification % of totalpopulationin2013,*in2012

43%* 96% 15% 1% 98%

RealGDPCAGR2001‐2010(yearsmayvary),%

*9% N/A N/A ***5.3% 3.1%

93http://www.refworld.org/topic,50ffbce528c,50ffbce52f8,58ec8a7e4,0,,ANNUALREPORT,ALB.html

56

RealGDPCAGR2011‐2016(yearsmayvary),%

**4.1% N/A N/A ****‐12.3%

7.3%

*2002‐2010, **2011‐2016

***2008‐2010,****2011‐2015

Source:TheWorldBankLittleDataBook2016,IMFWEO,CIAWorldFactbookAnnexTable3:Country StatisticsMedium IntensityConflictCountries (.. indicatesdatanotavailable)Country‐MediumIntensityConflict Sudan Central

AfricanRepublic

DemocraticRepublic oftheCongo

Myanmar

Population(million) 39.4 4.8 74.9 53.4GNI,Atlas$bn 67.3 1.6 28.7 68.1GNIpercapita,PPP,$ 3920 600 650 ..LifeExpectancyatbirthin2010 62 48 57 65LifeExpectancyatbirthin2014 63 51 59 66Laborforceparticipationratein2010(%ofpopulationages15+)

53 79 72 79

Laborforceparticipationratein2014(%ofpopulationages15+)

54 79 72 79

Gross capital formation in 2000 (% ofGDP)

25 11 14 12

Gross capital formation in 2010 (% ofGDP)

23 14 16 ..

Gross capital formation in 2014 (% ofGDP)

18 10 21 ..

Individualsusingtheinternetin2010(%ofpopulation)

16.7 2 0.7 0.3

Individualsusingtheinternetin2014(%ofpopulation)

24.6 4 3 2.1

Merchandisetradein2000(%ofGDP) 27 30 8 ..Merchandisetradein2010(%ofGDP) 33 22 48 ..Merchandisetradein2014(%ofGDP) 18 20 41 42Electrification % of total population in2013,*in2012

35% 3% 9% 52%

RealGDPCAGR2001‐2010,% 6.8% 2.5% 4.7% 12.0%RealGDPCAGR2011‐2016,% 3.4% ‐6.1% 6.8% 7.5%Source:TheWorldBankLittleDataBook2016,IMFWEO,CIAWorldFactbookAnnex Table 4: Country Statistics not In Conflict Countries (.. indicates data notavailable)

57

Country‐Notinconflict Eritrea Kosovo Serbia Albani

a

Population(million) 5.1 1.8 7.1 2.9GNI,Atlas$bn 2.3 7.3 41.5 12.9GNIpercapita,PPP,$ 1400 9300 13040 10980LifeExpectancyatbirthin2010 61 70 74 77LifeExpectancyatbirthin2014 64 71 76 78Laborforceparticipationratein2010(%ofpopulationages15+)

85 N/A 52 55

Laborforceparticipationratein2014(%ofpopulationages15+)

85 N/A 53 55

Gross capital formation in 2000 (% ofGDP)

22 N/A 11 32

Gross capital formation in 2010 (% ofGDP)

9 33 18 30

Gross capital formation in 2014 (% ofGDP)

.. 26 16 25

Individualsusingtheinternetin2010(%ofpopulation)

0.6 N/A 40.9 45

Individualsusingtheinternetin2014(%ofpopulation)

1 N/A 53.5 60.1

Merchandisetradein2000(%ofGDP) 72 N/A N/A 37Merchandisetradein2010(%ofGDP) 32 N/A 67 50Merchandisetradein2014(%ofGDP) .. N/A 81 58Electrification % of total population in2013,*in2012

32% 100 (in2016)

100(in2016) 100 (in2016)

Real GDP CAGR 2001‐2010 (years mayvary),%

0.9% 6.0% 4.4% 5.5%

Real GDP CAGR 2011‐2016 (years mayvary),%

*8.7% 3.0% 0.6% 2.1%

*2010‐2011

Source:TheWorldBankLittleDataBook2016,IMFWEO,CIAWorldFactbookReferences:Altonji,J.G.,&Card,D.(1991).Theeffectsofimmigrationonthelabormarketoutcomesofless‐skillednatives.InImmigration,trade,andthelabormarket(pp.201‐234).UniversityofChicagoPress.Aiyar, S., Barkbu,B., Batini,N., Berger,H.,Detragiache, E.,Dizioli, A., ...& Spilimbergo,A.(2016).TherefugeesurgeinEurope.EuropeIMFStaffDiscussionNote(SDN/16/02).

58

Angenendt,S.,Kipp,D.,MeierA.“MixedMigration”BetlesmannStiftung,2017Avle, S. (2014). Articulating and enacting development: Skilled returnees in Ghana’s ICTindustry.Infor‐mationTechnologies&InternationalDevelopment,10(4),1–13.Borjas,G.,&Bratsberg,B.(1996).WhoLeaves?TheOutmigrationoftheForeign‐Born.TheReviewofEconomicsandStatistics,78(1),165‐176.doi:10.2307/2109856Borjas,G.J.(2003).Thelabordemandcurveisdownwardsloping:Reexaminingtheimpactofimmigrationonthelabormarket.Thequarterlyjournalofeconomics,118(4),1335‐1374.Borjas,G.J.(2017).Thelaborsupplyofundocumentedimmigrants.LabourEconomics,46,1‐13.Black, Richard and Gent, Saskia (2006) Sustainable return in post‐conflict contexts.InternationalMigration,44(3).pp.15‐38.ISSN0020‐7985Black,R.,King,R.,&Litchaeld,J.,withAmmassari,S.,&Tiemoko,R.(2003).Transnationalmigration,returnanddevelopmentinWestAfrica(FinalResearchReport).Sussex,UK:SussexCenterforMigrationRe‐search,UniversityofSussex.Black,R.,Koser,K.,Munk,K.,Atfield,G.,D'Onofrio,L.,&Tiemoko,R.(2004).Understandingvoluntaryreturn.Black, R., & Gent, S. (2006). Sustainable Return in Post‐conflict Contexts.InternationalMigration,44(3),15‐38.Bloch,A.andAtfield,G.(2002)TheProfessionalCapacityofNationalsfromtheSomaliRegionsinBritain.ReportforRefugeeActionandIOM.Brücker,H.,&Jahn,E.J.(2011).MigrationandWage‐setting:ReassessingtheLaborMarketEffectsofMigration.TheScandinavianJournalofEconomics,113(2),286‐317.Butcher,K.F.,&Card,D. (1991). ImmigrationandWages:Evidence from the1980's.TheAmericanEconomicReview,81(2),292‐296.Card,D.(1990).TheImpactoftheMarielBoatliftontheMiamiLaborMarket.IndustrialandLaborRelationsReview,43(2),245‐257.doi:10.2307/2523702Carrington,W.J.anddeLima,R.1996.“TheImpactof1970sRepatriatesfromAfricaonthePortugueseLaborMarket.”IndustrialandLaborRelationsReview.Clemens,M.A.,&Hunt, J. (2017).TheLaborMarketEffectsofRefugeeWaves:ReconcilingConflictingResults(No.w23433).NationalBureauofEconomicResearch.

59

Council of Europe: Parliamentary Assembly,Readmission agreements: a mechanism forreturning irregular migrants, 16 March 2010,Doc. 12168,available at:http://www.refworld.org/docid/4bdadc1c3.html[accessed10August2017]Cohen, S., & Hsieh, C. T. (2001).Macroeconomic and labor market impact of RussianimmigrationinIsrael.Bar‐IlanUniv..Cortes, P. (2008). The effect of low‐skilled immigration onUS prices: evidence fromCPIdata.JournalofpoliticalEconomy,116(3),381‐422.Crisp, Jeff, and Katy Long. "Safe and Voluntary Refugee Repatriation: From Principle toPractice."J.onMigration&Hum.Sec.4(2016):141.Dadush,U. (2014). The effect of low‐skilled labormigration on the host economy.WorldBankKNOMADWorkingPaperSeries,(1).Dadush,U.(2015).Diaspora,DevelopmentandMorocco.PolicyPaperPP15/18,OCPPolicyCenter.Dadush,U.,&Niebuhr,M.(2016).Theeconomicimpactofforcedmigration(No.1605).OCPPolicyCenter.Dadush,U.,Demertzis,M.,&Wolff,G.B.(2017).Europe’sroleinNorthAfrica:development,investmentandmigration.BruegelPolicyContributionIssuen˚10|2017.Devictor, X. (2016).Forcibly Displaced: Toward a Development Approach SupportingRefugees,theInternallyDisplaced,andTheirHosts.WorldBank.Dustmann,C.(1996).Returnmigration:theEuropeanexperience.Economicpolicy,11(22),213‐250.EU, “On amore effective returnpolicy in the EuropeanUnion –A renewedAction Plan”March,2017EuropeanCouncil,ReturningRejectedAsylumSeekers:ChallengesAndGoodPractices InBelgium,Dec2016Evans,W.N.,&Fitzgerald,D.(2017).TheEconomicandSocialOutcomesofRefugeesintheUnitedStates:EvidencefromtheACS(No.w23498).NationalBureauofEconomicResearch.ESI"TheRefugeeCrisisThroughStatistics."EuropeanStabilityInitiative,30Jan.2017.Web.http://www.esiweb.org/pdf/ESI%20‐%20The%20refugee%20crisis%20through%20statistics%20‐%2030%20Jan%202017.pdf

60

"Eurostat."Database ‐ Eurostat. N.p., n.d. Web. 03 July 2017http://ec.europa.eu/eurostat/data/databaseFaist,T.(1999)TheCrucialMesolevel.InHammer,T.,Brochmann,G.,Tamas,K.andFaist,T.(eds.) InternationalMigration, ImmobilityandDevelopment,Berg,Oxford/NewYork,187‐217.Gerver,M.(2017).PayingRefugeestoLeave.PoliticalStudies,0032321716677607.Gibney, M. J. (2008). Asylum and the Expansion of Deportation in the UnitedKingdom.Governmentandopposition,43(2),146‐167.Gibson,J.,&McKenzie,D.(2011).Themicroeconomicdeterminantsofemigrationandreturnmigration of the best and brightest: Evidence from the Paciac. Journal of DevelopmentEconomics,95(1),18–29.doi:http://dx.doi.org/10.1016/j.jdeveco.2009.11.002Gmelch,G.(1980).Returnmigration.Annualreviewofanthropology,9(1),135‐159.Guerriero,M.(2012).Thelabourshareofincomearoundtheworld.Evidencefromapaneldataset.DevelopmentEconomicsandPublicPolicyWorkingPaper,32,2012.Harrod,R.F.(1939).Anessayindynamictheory.Theeconomicjournal,49(193),14‐33.Hagstrom,P.(2000).ThefiscalimpactofrefugeeresettlementintheMohawkValley.Clinton,NY:HamiltonCollege.RetrievedAugust,3,2006.Harild,N.,Christensen,A.,&Zetter,R.(2015).SustainableRefugeeReturn.Hatton, T. J., &Williamson, J. G. (1998).Theageofmassmigration:Causesand economicimpact.OxfordUniversityPressonDemand.Hunt, J. (1992). The impact of the 1962 repatriates from Algeria on the French labormarket.ILRReview,45(3),556‐572.Houte, M. V., & Davids, T. (2014). Moving back or moving forward? Return migration,developmentandpeace‐building.InternationalMonetaryFund“RefugeeSurgeinEurope:EconomicChallenges”,January2016InternationalOrganizationforMigration,Reintegration‐EffectiveApproaches(2015)InternationalInstituteforStrategicStudies(IISS)ArmedConflictSurvey2017Kindleberger, C. P. (1967).Europe's postwar growth: The role of labor supply. HarvardUniversityPress.

61

King,R.(1986).Returnmigrationandregionaleconomicdevelopment:anoverview.King, R. (2000) Generalizations from the History of ReturnMigration. In Ghosh, B. (ed.)ReturnMigration.JourneyorHopeorDespair?IOM/UNHCR,Switzerland,pp.1‐18.Koser,K.(1998)Informationandrepatriation,JournalofRefugeeStudies,10(1):1‐19.Koch,A.(2013).Assistedvoluntaryreturnschemes.ForcedMigrationReview,44,65.Kuschminder,K(2017)TakingStockofAssistedVoluntaryReturnfromEurope:DecisionMaking,ReintegrationandSustainableReturn–TimeforaparadigmshiftLaLonde,R.J.,&Topel,R.H.(1991).Labormarketadjustmentstoincreasedimmigration.InImmigration,trade,andthelabormarket(pp.167‐199).UniversityofChicagoPress.Lee, R., &Miller, T. (2000). Immigration, social security, and broader fiscal impacts.TheAmericanEconomicReview,90(2),350‐354.Leerkes,A.,Os,R.,&Boersema,E.(2017).Whatdrives‘softdeportation’?Understandingtherise in Assisted Voluntary Return among rejected asylum seekers in theNetherlands.Population,SpaceandPlace.Legrain, P. (2016). Refugees Work: A humanitarian investment that yields economicdividends.TentFoundationandOpenNetwork.Liebig,T.,&Mo,J.(2013).ThefiscalimpactofimmigrationinOECDcountries.InternationalMigrationOutlook2013,125‐189.Lucas,R.E. (2005).Internationalmigrationandeconomicdevelopment:Lessons from low‐incomecountries.EdwardElgarPublishing.Morrison,J.(2000).ExternalevaluationofthevoluntaryreturnprojectforrefugeesintheUnitedKingdom.London:RefugeeAction.Murgasova,Z.,Ilahi,N.,Miniane,J.,Scott,A.,&Hollar,I.V.(2015).TheWesternBalkans:15yearsofeconomictransition.InternationalMonetaryFund.Nadler,Kovacsetal"ReturnMigrationandRegionalDevelopmentinEurope”2016OECD (2013) International Migration Outlook 2003, Chapter 3, The Fiscal Impact ofImmigration in OECD Countries,http://www.globalmigrationgroup.org/system/files/Liebig_and_Mo_2013.pdfOECD,HowwilltherefugeesurgeaffecttheEuropeaneconomy?(2015).RetrievedJune28,2017, from https://www.oecd.org/migration/How‐will‐the‐refugee‐surge‐affect‐the‐European‐economy.pdf

62

Orhan,O.,&SenyücelGündoğar,S.(2015).EffectsoftheSyrianrefugeesonTurkey.ORSAMReport,195.Ortega,F.,&Peri,G.(2009).Thecausesandeffectsofinternationalmigrations:EvidencefromOECDcountries1980‐2005(No.w14833).NationalBureauofEconomicResearch.Ottaviano,G.I.,&Peri,G.(2008).Immigrationandnationalwages:Clarifyingthetheoryandtheempirics(No.w14188).NationalBureauofEconomicResearch.Paine,S.(1974).Exportingworkers:theTurkishcase.Pasha,H.A.,&Altaf,M.A.(1987).Returnmigrationina life‐cyclesetting:AnexploratorystudyofPakistanimigrantsinSaudiArabia.PakistanJournalofAppliedEconomics,6(1),1‐23.Peri,G.(2012).Theeffectofimmigrationonproductivity:EvidencefromUSstates.ReviewofEconomicsandStatistics,94(1),348‐358.Peri, G., & Sparber, C. (2009). Task specialization, immigration, and wages.AmericanEconomicJournal:AppliedEconomics,1(3),135‐169.Peri,G.,&Yasenov,V.(2017).TheLaborMarketEffectsofaRefugeeWave:SyntheticControlMethodMeetstheMarielBoatlift.Plaza,S.,&Ratha,D.(2011).HarnessingdiasporaresourcesforAfrica.DIASPORAfor,1.Ruben, R., Van Houte, M., & Davids, T. (2009). What Determines the Embeddedness ofForced‐Return Migrants? Rethinking the Role of Pre‐and Post‐ReturnAssistance.InternationalMigrationReview,43(4),908‐937.Ruist,J.(2015).Thefiscalcostofrefugeeimmigration:theexampleofSweden.PopulationandDevelopmentReview,41(4),567‐581.Schuster,L.&Majidi,N.(2013).Whathappenspost‐deportation?TheexperienceofdeportedAfghans.Migrationstudies,1(2),pp.221‐240.doi:10.1093/migration/mns011Schoeni,R. F. (1997).TheEffectsof Immigrationon theEmploymentandWagesofNativeWorkers:Evidencefromthe1970sand1980s.RAND.Sorensen,N.N.,VanHear,N.,&Engberg‐Pedersen,P.(2003).Themigration‐developmentnexus:evidenceandpolicyoptions.TheMigration‐DevelopmentNexus.Strand,A.,Akbari,A.,&Chaudhary,T.W.(2008).Returnindignity,returntowhat?ReviewofthevoluntaryreturnprogrammetoAfghanistan.CMI.

63

UNHighCommissionerforRefugees(UNHCR),GlobalTrends:ForcedDisplacementin2015,20 June 2016,available at: http://www.refworld.org/docid/57678f3d4.html[accessed 6July2017]UNHCR,TheRefugeeConvention,1951,availableat:http://www.unhcr.org/4ca34be29.pdfUNHCRPopulationStatistics‐Data‐TimeSeries.N.p.,n.d.Web.03July2017.UNHCR,Protection,R.(2007).A10‐PointPlanofaction.The10‐PointPlaninaction,15.UNHCRToughchoices forAfghanrefugeesreturninghomeafteryears inexile.RetrievedAugust 29, 2017, from http://www.unhcr.org/en‐us/news/briefing/2017/2/589453557/tough‐choices‐afghan‐refugees‐returning‐home‐years‐exile.htmlUN‐ Summary Overview Document. (2016, September 20). Retrieved July 6, 2017, fromhttps://refugeesmigrants.un.org/sites/default/files/public_summary_document_refugee_summit_final_11‐11‐2016.pdfVan Houte, M. And M. De Koning. 2008. Towards a Better Embeddedness? MonitoringAssistance to Involuntary Returning Migrants from Western countries. Nijmegen /Amsterdam:CIDIN/AMIDSt.Verme,P.,Gigliarano,C.,Wieser,C.,Hedlund,K.,Petzoldt,M.,&Santacroce,M.(2015).ThewelfareofSyrianrefugees:evidencefromJordanandLebanon.WorldBankPublications.Webber,F.(2011).Howvoluntaryarevoluntaryreturns?Race&Class,52(4),98‐107.AiyarAiyarZetter,R.,&Ruaudel,H. (2016).Refugees’ right toworkandaccess to labormarkets–Anassessment.