Spain: From massive immigration to vast emigration?

Izquie do, Ma io; Jimeno, Juan F.; Lacues a, Ai o A icle Spain: F om massi e immig a ion o as emig a ion? IZA Jou nal o Mig a ion P o ided in Coope a ion wi h: IZA – Ins i u e o Labo Economics Sugges ed Ci a ion: Izquie do, Ma io; Jimeno, Juan F.; Lacues a, Ai o (2016) : Spain: F om massi e immig a ion o as emig a ion?, IZA Jou nal o Mig a ion, ISSN 2193-9039, Sp inge , Heidelbe g, Vol. 5, Iss. 10, pp. 1-20, h ps://doi.o g/10.1186/s40176-016-0058-y This Ve sion is a ailable a : h ps://hdl.handle.ne /10419/149457 S anda d-Nu zungsbedingungen: Die Dokumen e au EconS o dü en zu eigenen wissenscha lichen Zwecken und zum P i a geb auch gespeiche und kopie we den. Sie dü en die Dokumen e nich ü ö en liche ode komme zielle Zwecke e iel äl igen, ö en lich auss ellen, ö en lich zugänglich machen, e eiben ode ande wei ig nu zen. So e n die Ve asse die Dokumen e un e Open-Con en -Lizenzen (insbesonde e CC-Lizenzen) zu Ve ügung ges ell haben soll en, gel en abweichend on diesen Nu zungsbedingungen die in de do genann en Lizenz gewäh en Nu zungs ech e. Te ms o use: Documen s in EconS o may be sa ed and copied o you pe sonal and schola ly pu poses. You a e no o copy documen s o public o comme cial pu poses, o exhibi he documen s publicly, o make hem publicly a ailable on he in e ne , o o dis ibu e o o he wise use he documen s in public. I he documen s ha e been made a ailable unde an Open Con en Licence (especially C ea i e Commons Licences), you may exe cise u he usage igh s as speci ied in he indica ed licence. h p://c ea i ecommons.o g/licenses/by/4.0/ RESEARCH Open Access Spain: om massi e immig a ion o as emig a ion? Ma io Izquie do, Juan F. Jimeno and Ai o Lacues a * * Co espondence: ai o .lacues a@ bde.es Banco de España, Mad id, Spain Abs ac La ge immig a ion lows du ing he 1995–2007 pe iod inc eased he weigh o o eigne s li ing in Spain o 12 % o he o al popula ion. The apid inc ease in unemploymen associa ed wi h he G ea Recession and he subsequen Eu opean deb c isis, subs an ially changed mig a ion lows, so ha , om he beginning o he 2010s, Spain expe ienced posi i e ne ou lows. In his pape , we ake on h ee asks. Fi s , we show ha sensi i i y o mig a ion lows o unemploymen is simila be ween Spania ds bo n in Spain and o eigne s ha en e ed in Spain du ing he las 20 yea s. Second, we es ima e he impo ance o pas ne wo k e ec s o o eigne s and he ecen ne wo k e ec s o Spania ds in shaping hese mig a ion lows. Finally, we will use hese es ima ions o conjec u e ha he apid c ea ion o ne wo ks o Spania ds ab oad migh keep hei emig a ion a es g owing, hus coun e balancing he e ec s o ou lows due o he ecen all in he unemploymen a e. JEL classi ica ion: F22, J64, J61 Keywo ds: Mig a ion in lows and ou lows, Unemploymen , Ne wo ks 1 In oduc ion Jus be o e he G ea Recession, Spain ecei ed massi e mig a ion in lows ha con ib- u ed o an a e age annual popula ion g ow h o 1.4 % be ween 2000 and 2007 and in- c eased he weigh o he o eign popula ion om 2 o 12 %. Howe e , he e ec s o he G ea Recession on he Spanish labo ma ke we e ema kable: F om peak ( hi d qua e o 2007) o ough ( i s qua e o 2014), employmen ell by 18.3 %, while he unemploymen a e eached a maximum o 26.9 % ( i s qua e o 2013). No su p is- ingly, mig a ion in lows and ou lows changed signi ican ly since 2010. The ques ions ha we add ess in his pape a e wo old. Fi s , we es ima e o wha ex en changes in mig a ion lows o o eigne s and na ionals esponded di e en ly o he wo sening in he labo ma ke si ua ion. Second, we discuss whe he he changes in mig a ion lows should be expec ed o pe sis and, he e o e, whe he Spain could be in ansi ion om massi e immig a ion o o eigne s o as emig a ion o Spania ds. To answe hese ques ions, we ca e ully look a he beha io o mig a ion in lows and ou lows o di e en popula ion g oups. Indeed, i is in e es ing o see ha du ing he c isis, while o eigne s expe ienced emig a ion a es abo e 6 %, only 0.1 % o Spania ds bo n in Spain mo ed ab oad. Ei he mig a ion lows o Spania ds we e less sensi i e o unemploymen han hose o o eigne s o he o me aced la ge © 2016 Izquie do e al. Open Access This a icle is dis ibu ed unde he e ms o he C ea i e Commons A ibu ion 4.0 In e na ional License (h p://c ea i ecommons.o g/licenses/by/4.0/), which pe mi s un es ic ed use, dis ibu ion, and ep oduc ion in any medium, p o ided you gi e app op ia e c edi o he o iginal au ho (s) and he sou ce, p o ide a link o he C ea i e Commons license, and indica e i changes we e made. Izquie do e al. IZA Jou nal o Mig a ion (2016) 5:10 DOI 10.1186/s40176-016-0058-y emig a ion cos s due o ins ance o a lack o ne wo ks ab oad. The unde lying mo i - a ion is impo an since each explana ion has di e en long- e m consequences on mi- g a ion lows. I Spania ds mig a ion lows a e no e y sensi i e o changes in unemploymen de elopmen s, some hing ha was con i med in he case o in e nal mig a ion in Spain (An olín and Bo e 1997; Ben olila and Dolado 1991), mig a ion ou lows ab oad migh no acqui e he su icien en i y o gene a e any ele an ne - wo k e ec s ha could lowe mig a ion cos s. Hence, unde his scena io, la ge mig a- ion ou lows would be unlikely. On he con a y, i mig a ion lows ab oad we e sensi i e o changes in unemploymen and ne wo ks e ec s s a ed o play some ole, hese ou lows would con inue g owing, e en i some imp o emen in he labo ma ke si ua ion in Spain ook place. The e a e wo easons why we do no neglec his possi- bili y up on . One is he apid c ea ion o ne wo ks obse ed in Spain du ing he a - i al o la ge in lows in he expansiona y pe iod jus be o e he c isis. Ano he is he Spanish expe ience o la ge mig a ion ou lows ab oad du ing he pe iod 1950–1970, which eached a maximum o 0.7 % o popula ion (Fig. 1) e en a a ime when GDP pe capi a was g owing abo e 5 % pe yea (P ados de la Escosu a 1993). In sho , in his pape , we ake on h ee asks. Fi s , we y o disen angle whe he he sensi i i y o mig a ion lows wi h espec o unemploymen is di e en be ween Spania ds and o eigne s esiding in Spain. Second, we es ima e he impo ance o pas ne wo k e ec s o o eigne s and he ecen ne wo k e ec s o Spania ds in shaping hese mig a ion lows. Finally, we will use hese es ima ions o hypo hesize abou u u e mig a ion ou low scena ios. The s uc u e o he pape closely esembles hese h ee goals. Fi s , we desc ibe he da a sou ces used o measu e mig a ion in lows and ou lows. We hen ocus on he mig a ion in lows o o eign na ionals, and on he ou - lows o Spania ds bo n in Spain, o es ima e hei esponses o economic condi ions, conside ing also hei sociodemog aphic composi ion. Finally, we es ima e he emig a- ion cos s and conclude wi h some commen s on he implica ions o hese mig a ion lows o u u e po en ial g ow h. 2 Da a 2.1 Sou ces Da a on g oss mig a ion lows in Spain a e qui e limi ed in e ms o bo h de ails and ime span. Un il ecen ly, da a on mig a ion ou lows came exclusi ely om eco ds o passenge s lea ing he coun y by sea o ai and om in o ma ion on o icial bila e al p og ams o o ganized mig a ion o Eu ope. This is why mos o he s udies on Spanish emig a ion had o ely on des ina ion coun y in o ma ion. 1 I was no un il 1998 ha an o ganized a emp was made o keep eco ds o mig a ion in lows and ou lows based on municipal egis e s, which yielded he mig a ion module o he Es adís ica de Va iaciones Residenciales, a ailable since 2002. Howe e , his s a is ical sou ce has some d awbacks, because al hough o eign na ionals a i ing in Spain ha e an incen i e o en oll on he municipal egis e , hey do no p o ide in o ma ion needed o upda e i when hey lea e he coun y. To co ec his, since 2006, o eign na ionals ( om non-EU coun ies and hose who do no ha e a pe manen esidence pe mi ) a e e- qui ed o enew hei egis a ion e e y 2 yea s; hose who ail o do so a e conside ed o ha e le he coun y and a e d opped om he egis e . Since 2009, he Na ional Izquie do e al. IZA Jou nal o Mig a ion (2016) 5:10 Page 2 o 20 S a is ics O ice (INE) also uses al e na i e su eys o ob ain in o ma ion on exi s o EU ci izens and mig an s wi h pe manen esidence pe mi s in o de o assign an es ima ed depa u e da e o hose exi s. This in o ma ion is he basis o he da a on ou lows a he Es adís ica de Mig aciones p o ided by he INE since 2008. As o mig a ion in lows and ou lows o Spanish na ionals, hey a e be e cap u ed by he municipal egis e s. Howe e , he in o ma ion on ou lows is no ee o p ob- lems, since i elies on en olmen s on Spanish consula es and embassies ab oad. In his case, some delay is e y likely be ween he da e o mig a ion and he egis a ion, and i is also likely ha only pe manen mo es ab oad a e egis e ed. 2 To o e come some o he d awbacks o o icial s a is ics, we use a combina ion o s a is ical sou ces. Fo en ies in o Spain, we ely on he Es adís ica de Va iaciones Residenciales, which p o ide longe se ies. Fo ou lows, we use as he Es adís ica de Mig aciones, which p o ides be e in o ma ion du ing he a ailable sample pe iod. Bo h da ase s p o ide in o ma ion on gende , age, na ionali y, coun y o bi h, Fig. 1 Mig a ion lows o and om Spain as a unc ion o unemploymen a e Izquie do e al. IZA Jou nal o Mig a ion (2016) 5:10 Page 3 o 20 p o ince o o igin (des ina ion), and coun y o des ina ion (o igin) o mig an s. 3 In he case o o eign na ionals lea ing Spain, we assume ha he coun y o des ina ion coincides wi h he coun y o bi h, since a ailable e idence sugges s ha his is a good app oxima ion o eali y. 4 Mo eo e , by es ic ing one des ina ion o each place o bi h, he empi ical s a egy is simple . Da a on he popula ion o o eign na ionals in he o igin coun y is ob ained om he Wo ld Economic Ou look da abase. Da a on he popula ion o Spanish and o eign na ionals esiding in Spain is ob ained om municipal egis e s (Pad ón de Habi an es). 5 In o ma ion on he s ock o na ionals esiding ab oad is aken om he egis e s o Spanish consula es and embassies ga he ed by he INE since 2009 (Pad ón de Españoles Residen es en el Ex anje o), which p o ide in o - ma ion on he coun y o bi h, p o ince o las esidence, p o ince o bi h and coun y o des ina ion. 2.2 Main his o ical and s ylized ac s Spain had ne mig a ion ou lows h oughou mos o he 20 h cen u y (Table 1). A he beginning o he 20 h Cen u y, he e we e huge ou lows o Sou h Ame ica, mos ly o A gen ina, while mig a ion in lows we e negligible. Those mo es we e igge ed by se e al ac o s: (i) he ee mo emen laws enac ed in Spain and in des ina ion coun- ies in he second hal o he 19 h cen u y; (ii) a g owing popula ion, due o he de- c ease in mo ali y a he end o he 19 h cen u y ha was ollowed by a subsequen dec ease in e ili y (Spain’s popula ion almos doubled be ween 1857 and 1950, om 15 million o 27 million); (iii) economic s agna ion, wi h he a e age annual g ow h a e o GDP pe capi a in eal e ms es ima ed o ha e been sligh ly below 1 % be ween 1850 and 1950; and (i ) a u bulen poli ical si ua ion in he pe iod 1898–1939. How- e e , mig a ion ou lows came o a s ands ill in he wake o he Fi s Wo ld Wa and he global c isis o he 1930s and a e he Spanish ci il wa (1936–1939) mig a ion om Spain was banned. When mig a ion ab oad was eely allowed again in 1946, ini- ially, almos 100 % o he o al ou lows wen o Sou h Ame ica, al hough hey we e smalle han hose egis e ed a he beginning o he XX h Cen u y. Since 1950 and wi h ega d o mig a ion lows, he e a e h ee di e en pe iods. 2.2.1 Ea ly emig a ion (1950–1989) Ea ly in his pe iod, he main des ina ion was Sou h Ame ica, which a ac ed some 50 housand mig an s pe yea , ollowed in he 1960s by Eu ope (mos ly F ance, Ge many, and Swi ze land), wi h a e age annual lows o a ound 170 hou- sand mig an s (see Table 1). In he 1960s, some 80 % o o al emig an s wen o Eu ope, eaching a peak o 7 pe housand o he o al popula ion mid-decade. Emig a ion o Eu ope was mos ly d i en by he sho age o unskilled wo ke s o ill jobs in ag icul u e o manu ac u ing in he des ina ion coun ies. F om he mid- 1960s, hese emig a ion lows declined, due i s o economic g ow h in Spain and subsequen ly o he highe ba ie s o immig a ion e ec ed in he des ina ion coun- ies ollowing he oil c isis o he ea ly 1970s. Howe e , ne posi i e ou lows con- inued, al hough a a slowe pace, du ing he i s hal o he 1980s when Spain was s ill su e ing signi ican employmen losses. Izquie do e al. IZA Jou nal o Mig a ion (2016) 5:10 Page 4 o 20 Table 1 Immig a ion and emig a ion by na ionali y and coun y o bi h Flow (pe sons) Ra io espec o popula ion (‘000) To al Spanish na ionali y Spanish na ionali y Fo eign na ionali y To al Spanish na ionali y Spanish na ionali y Fo eign na ionali y Bo n in Spain Bo n ab oad To al To al Bo n in Spain Bo n ab oad To al To al Immig a ion 1996 29.895 9.359 3.850 13.209 16.686 0.8 0.2 6.8 0.3 28.6 1997 57.877 15.401 6.860 22.261 35.616 1.4 0.3 10.3 0.5 59.2 1998 81.227 15.881 8.151 24.032 57.195 2.0 0.4 13.7 0.6 89.8 1999 127.365 17.488 10.755 28.243 99.122 3.2 0.5 18.1 0.7 132.3 2000 362.468 17.608 13.979 31.587 330.881 8.9 0.5 21.8 0.8 358.1 2001 414.772 9.522 11.202 20.724 394.048 10.1 0.2 16.7 0.5 287.5 2002 483.260 17.836 22.339 40.175 443.085 11.6 0.5 31.7 1.0 224.0 2003 470.010 19.209 21.277 40.486 429.524 11.0 0.5 28.4 1.0 161.2 2004 684.561 19.941 18.776 38.717 645.844 15.8 0.5 23.9 1.0 212.8 2005 719.284 18.476 18.097 36.573 682.711 16.3 0.5 21.9 0.9 183.0 2006 840.844 18.944 18.929 37.873 802.971 18.8 0.5 21.5 0.9 193.8 2007 958.266 19.010 18.722 37.732 920.534 21.2 0.5 19.9 0.9 203.7 2008 726.009 17.068 16.713 33.781 692.228 15.7 0.4 16.1 0.8 131.4 2009 498.977 15.887 13.748 29.635 469.342 10.7 0.4 12.1 0.7 83.1 2010 464.443 15.658 17.451 33.109 431.334 9.9 0.4 14.3 0.8 75.0 2011 454.686 18.678 19.726 38.404 416.282 9.6 0.5 14.8 0.9 72.4 2012 370.515 17.767 16.638 34.405 336.110 7.8 0.4 11.4 0.8 58.6 2013 342.390 18.197 17.157 35.354 307.036 7.3 0.5 11.0 0.9 55.4 2014 380.659 20.205 18.950 39.155 341.504 8.1 0.5 11.0 0.9 68.0 Emig a ion Pe iod o ea ly emig a ion 1960s 1960 105.420 3.5 1961 176.821 5.7 1962 216.381 6.9 1963 200.539 6.3 1964 230.124 7.2 1965 203.609 6.3 1966 155.093 4.8 1967 89.484 2.7 Pe iod o ecen emig a ion 2002 26.102 3.572 29.674 0.67 5.07 0.74 2003 13.870 2.120 15.990 0.35 2.83 0.40 2004 10.985 2.171 13.156 0.28 2.76 0.33 2005 15.914 3.376 19.290 0.40 4.08 0.48 2006 17.900 4.142 22.042 0.45 4.70 0.54 2007 22.527 5.564 28.091 0.57 5.91 0.69 2008 288.432 25.461 8.044 33.505 254.927 6.2 0.64 7.75 0.82 48.38 2009 380.118 26.334 9.656 35.990 344.128 8.1 0.66 8.53 0.88 60.92 2010 403.379 29.204 10.953 40.157 363.221 8.6 0.73 8.97 0.97 63.19 2011 409.034 40.150 15.321 55.472 353.562 8.7 1.00 11.49 1.34 61.47 Izquie do e al. IZA Jou nal o Mig a ion (2016) 5:10 Page 5 o 20 2.2.2 Immig a ion: he boom (1990–2007) and ecen ends (2008–2014) F om he ea ly 1990s, and mos no iceably a e 1997, Spain became a des ina ion coun y o immig an s. In lows inc eased s eadily, om unde 30 housand pe yea in 1996 o 958 housand in 2007, when o eigne s amoun ed o mo e han 12 % o he o al popula ion (see Table 1). Du ing he expansion, o eign immi- g an s we e mos ly Eu opeans, closely ollowed by La in Ame icans (mos ly Pe u- ians and Boli ians) and A icans (mos ly Mo occans). The onse o he c isis b ough abou a sudden shi in his end, and in 2008 and 2009, Eu opean en ies came o a sudden s ands ill, al hough since 2010, hei sha e has e u ned o p e- c isis le els. Du ing hose yea s, he numbe o immig an s om he Ame icas con inued o decline, while he numbe o immig an s om A ica and Asia ose somewha as a sha e o he o al. As o in lows o Spanish na ionals, du ing he 1980s and he ea ly 1990s, mos o hem we e Spania ds bo n in Spain, a clea sign o e u n mig a ion. Howe e , la e in he 1990s, he composi ion changed due o a big inc ease in he numbe o Spania ds bo n ab oad coming o Spain. Since Spanish na ionali y is acqui ed h ough pa en al na ionali y, ega dless o he coun y o bi h, i is likely ha many o eigne s (in he sense o people who had ne e li ed in Spain be o e) we e immig a ing unde Spanish na ionali y, so ha hese in lows should no be conside ed as he consequence o e u n mig a ion. Mo eo e , in lows o Spanish na ionals bo n in Spain we e ela i ely inelas ic o economic condi ions, in con as wi h he in lows o o eigne s. As o he impac o he c isis, he e was a ise in he numbe o Spanish na ionals e u ning om Asia and A ica, while in he case o Spania ds bo n ab oad, he mos no iceable de elopmen is a u he inc ease in he sha e o immig an s om he Ame icas. 2.2.3 Emig a ion: he c isis (2008–2014) Mig a ion ou lows s a ed o inc ease in 2007, and since 2010, ou lows ha e amoun ed o mo e han 400 housand pe yea (sligh ly below 1 % o he o al domes ic popula ion), which is, bo h in absolu e and ela i e e ms, he highes le el o emig a- ion in Spanish his o y (Table 1). This is mos ly due o he high mobili y o o eign na ionals. Indeed, hey make up he g ea majo i y o mig a ion ou lows: in 2012, app oxima ely 5 % o o eigne s esiding in Spain le he coun y, while less han 1 pe housand o Spania ds bo n in Spain emig a ed. Fo he la e , mig a ion ou lows a e s ill smalle han hose egis e ed in he 1960s, 6 al hough be ween 2011 and 2014, ou - lows ha e no been o se by in lows, so Spain is now eco ding ne mig a ion ou lows o na ionals o he i s ime since he 1970s (ne ou lows in 2012 a e es ima ed a a ound 70 housand). Table 1 Immig a ion and emig a ion by na ionali y and coun y o bi h (Con inued) 2012 446.606 38.749 18.518 57.267 389.339 9.4 0.97 12.64 1.38 67.87 2013 532.303 48.136 25.193 73.329 458.974 11.3 1.20 16.11 1.76 82.75 2014 409.344 50.249 28.536 78.785 330.559 8.8 1.26 16.49 1.89 65.80 Sou ces: Emig a ion om Spain du ing he pe iod o ea ly mig a ion is om Nadal (1984) and Ga cia Fe nandez (1965). I is compu ed as he sum o emig a ion o Ame ica based on passenge s on ships and planes and emig a ion o Eu ope based on des ina ion coun ies’s a is ics popula ion in ha pe iod is om census da a 1960 and 1970. The yea s in be ween a e he esul o a geome ic in e pola ion immig a ion om municipal egis e s (Es adís ica de Va iaciones Residenciales). Emig a ion om municipal egis e s (Es adís ica de Va iaciones Residenciales) un il 2007 and Es adís ica de mig aciones since 2008 Popula ion since 1996 comes om municipal egis e s (Pad on Con inuo) Izquie do e al. IZA Jou nal o Mig a ion (2016) 5:10 Page 6 o 20 Mo eo e , since 2007, he e a e also ne mig a ion ou lows o Spanish-bo n Spania ds; he numbe s a e low, bu mig a ion ou lows o his popula ion g oup a e accele a ing. P elimina y da a a ailable o he i s hal o 2015 poin o a u he inc ease in exi s, up o 32,980, while 51,267 was he numbe o he comple e yea 2014. Mos o eigne s wen o Eu ope and Sou h Ame ica, while Spania ds bo n in Spain o e whelmingly decide o mig a e o Eu ope. Al hough he c isis has had li le impac on he choice o des ina ion coun ies o o eigne s, in he case o Spania ds bo n in Spain, i has inc eased he sha e o ou lows o Eu ope (mos ly Ge many and he UK) and o he USA. 3 Es ima ing he impac o unemploymen on mig a ion in lows and ou lows Apa om many geopoli ical ac o s (wa s, na u al ca as ophes, sea ch o poli ical asylum), economic ac o s play a e y impo an ole a de e mining in e na ional mig a ion in lows and ou lows. A e y wide li e a u e has highligh ed many o hem, such as wage di e en ials, employmen condi ions, mig a ion cos s, bo h pecunia y and cul u al, he di e si ica ion o isk wi hin amily membe s, di e ences in ela i e p ices be ween hos and home coun y, he accumula ion o human capi al, he imp o emen o he heal h s a us, o he willingness o each a sa ings a ge o o e come capi al cons ain s in he home coun y (Massey e al. 1993; Dus mann and Weiss 2007). Mo e ecen ly, global alue chains ha e also gene a ed a new so o empo a y mo emen s. Finally, he consolida ion o in e na ional mig a ion is usually ela ed o ne wo k e ec s ha suppo ansna ional mo emen s gene a ed by pa icula ins i u ions in he hos and he ecei ing coun y. In o de o analyze he Spanish case, we ocus on he di e en ial o employmen con- di ions be ween o igin and des ina ion coun ies. Ou s a egy is simila o G ogge and Hanson’s (2011), who ela e he log odds o esiding in coun y h o a pe son om coun y sis hough o be de e mined by absolu e di e ences in ea nings be ween he wo coun ies and by he cos o mig a ing ha is idiosync a ic o ha pa icula coun- y o o igin and des ina ion pai . Howe e , when conside ing mig a ion o o om Spain, ea nings do no appea o be a good p oxy o economic oppo uni ies, since high unemploymen a es ha e been p e alen , anging be ween 8 and 26 % in he pe iod 2000–2015. Indeed, changes in unemploymen appea o be mo e app op ia e han changes in wages o measu e how economic oppo uni ies e ol e o e ime, pa - icula ly when, as has happened du ing he cu en ecession, wages ha e eac ed slowly o he wo sening o he economic si ua ion due o signi ican eal and nominal igidi ies. 7 Since Ha is and Toda o (1970), many o he pape s in he li e a u e ha e in- cluded unemploymen a e di e en ials as ei he push o pull ac o s o explain mig a- ion lows (see, o ins ance Za odny 1997; Pede sen e al. 2008; Beine e al. 2013; Be oli e al. 2013a). In he spi i o G ogge and Hanson’s (2011) app oach, mig a ion lows a e assumed o espond symme ically o changes in ela i e economic oppo uni ies, so ha he e - ec s on mig a ion lows o a change in ela i e economic condi ions in one pa icula coun y should disappea comple ely when he ini ial economic condi ions a e es o ed. In his ega d, he ecen expe ience in Spain shows ha his migh no be he case. The sha p d op in unemploymen ( om a ound 23 o 11 %) be ween 1995 and 2001 d o e up he sha e o o eigne s in he Spanish economy. Subsequen ly, unemploymen Izquie do e al. IZA Jou nal o Mig a ion (2016) 5:10 Page 7 o 20 emained oughly cons an , bu he o eign popula ion con inued o g ow. In 2007, he Spanish unemploymen a e headed up again, bu he o eign popula ion con inued o inc ease, be o e declining sligh ly in 2012. I seems, he e o e, ha analyzing changes in he s ock o o eign na ionals in Spain in he cu en si ua ion needs a mo e lexible speci ica ion. Fo ha eason, we es ima e he e ec s o economic condi ions on bo h en ies o o eigne s and exi s o Spania ds sepa a ely (see Fig. 1). The sha p decline in unemploymen be ween 1995 and 2001 p omp ed an inc ease in he numbe o in lows o o eigne s in Spain. These in lows con inued o ise un il 2007, e en hough he unemploymen a e was qui e s eady, due o dec easing mig a ion cos s and as esul o he inc ease in he s ock o mig an s in Spain ha p ecluded a gene alized p e e ence o Spain a he han o he al e na i e des ina ions (Be oli e al. 2013b). As om 2007, in lows dec eased as unemploymen ose. I is no ewo hy ha he impac o changes in unemploymen on mig a ion in- lows is simila o ha obse ed in he 1990s, bu a a highe le el o unemploymen . In u n, in hese las 4 yea s, ou lows o bo h o eign and na ionals ha e also inc eased as unemploymen has isen a a simila pace. One would be emp ed o ask whe he he cu en le el o ou lows o Spania ds has been enough o dec ease he cos s o emig a ion o his pa icula socioeconomic g oup as i happened o o eigne s be ween 2001 and 2007. In pa icula , Fig. 2 illus a es ha a e a small decline in unemploymen a es in 2014, emig a ions o Spania ds bo n in Spain kep g owing whe eas hose o o eigne s al eady declined. In pa icula , he emig a ion a e o Spania ds bo n in Spain (pe housands) inc eased in 2014 om 1.2 o 1.35 whe eas ha o o eigne s dec eased om 83 o 65 ( o u he de ails, see Izquie do e al. 2014). This desc ip i e e idence sugges s ha e en hough bila e al in lows and ou lows migh espond o economic condi ions as heo y p edic s, changes in cos s o mig a ion may blu he con empo aneous esponses o he s ock o mig an s o economic condi- ions somewha . To es quan i a i ely he impo ance o hese mechanisms, we ollow Be oli e al (2013a) and ela e he log odds o immig a ing (I) ( ac ion o en ies om one coun y o bi h o one egion di ided by he co esponding popula ion esiding in ha coun y o o igin). 8 We also do he same o emig a ing (E) ( ac ion o exi s om one coun y o bi h o one coun y o des ina ion di ided by he co esponding popu- la ion esiding in ha egion) o unemploymen di e en ials (U) and o he cos s o im- mig a ion/emig a ion be ween o igin (h) and des ina ion (s). 9 Using he supe sc ip o deno e o eigne s and e o deno e Spania ds, ou eg ession speci ica ions a e: lnI hs ¼α0þα1Uh −Us ðÞþλ c þλhchþλscsð1Þ lnEe hs ¼γ0þγ1Uh −Us ðÞþκ c þκhchþκscsð2Þ In subsequen speci ica ions, we will allow di e en elas ici ies o he unemploymen a e a o igin and des ina ion. We p oxy he cos o emig a ion (c) in di e en ways (sepa a e dummies o each o igin coun y and o each des ina ion egion o dummies o ime dummies coupled wi h a dyad coun y/ egion). We also include as co a ia es he ime dummies (c ). The li e a u e has iden i ied wo p oblems wi h speci ica ions such as (1)–(2): i s , he po en ial bias in oduced by he exis ence o ze os in emig a ion/immig a ion a es Izquie do e al. IZA Jou nal o Mig a ion (2016) 5:10 Page 8 o 20 4.2 Mig a ion cos s and ne wo k e ec s Columns 1–3 o Table 7 show ha no ma e he way we speci y he incidence o ela- i e unemploymen in o igin and des ina ion coun y, mig a ion a es inc ease when- e e he s ock o mig an s is highe a des ina ion. In pa icula , con olling o unemploymen a es, pai s o coun ies/ egions ha ha e 1 pp o a highe s ock o mi- g an s gene a e a highe mig a ion in low a e o 0.8 pp. This is a well-es ablished ac ; simila esul s a e in Ha on and Williamson (2002) and Pede sen e al. (2008) al hough hei co esponding es ima es (0.43 and 0.36, espec i ely) a e sligh ly lowe han ou s. This could be he case because he ecen immig a ion his o y in Spain s a ed in 1995 wi h almos no immig an se led in he coun y, some hing ha may ha e gene a ed a highe po en ial o endogenous g ow h since conges ion e ec s a e nil. This ne wo k e ec is e en mo e p onounced o Spania ds bo n in Spain (see col- umns 4–6). In pa icula , con olling o unemploymen , pai s o egions o o igin and coun y o des ina ion ha ha e 1 pp o a highe s ock o Spania ds gene a e a highe immig a ion a e o 1.5 pp. No ice ha esul s migh be a ec ed by he ac ha he in- o ma ion egis e ed in he Spanish consula es and embassies could be a e y poo p oxy o se lemen s ab oad, since incoming mig an s do no egis e immedia ely, some hing ha does no happened wi h o eigne s coming o Spain since hey ha e many incen i es o egis e in he municipal egis e s as soon as possible. 16 As i has been said, emig a ion by Spania ds bo n in Spain is a new phenomenon and ne wo k e ec s could be s a ing o ope a e a ull po en ial (embassy egis e s eco d jus unde 750,000 Spania ds bo n in Spain and li ing ab oad in Janua y 2015, which is less han 2 % o he popula ion). As a consequence, his mechanism gene a es a coun- e balance o he ecen ela i e imp o emen o he Spanish unemploymen a e Table 6 Log odds o emig a ing om Spain by Spania ds bo n in Spain. CCE Exi o Spania ds bo n in Spain CCE Va iables (1) (2) (4) (5) U(coun y) −0.07086 −0.06150 −0.07091 −0.06206 (0.00658)** (0.00735)** (0.00656)** (0.00730)** U(CCAA) 0.02133 0.04422 (0.00410)** (0.00330)** Time dummies Yes Yes Yes No Coun y dummies Yes No Yes No Region dummies Yes No Yes No Pai coun y/ egion dummies No Yes Yes Yes Auxilia y eg esso s No No Yes Yes Cons an −7.258 −1.298 −7.758 −1.870 (0.57809)** (0.17405)** (0.58854)** (0.17853)** R-squa ed 0.79 0.32 0.79 0.33 Obse a ions 2267 2267 2267 2267 Sou ce: The dependen a iable is he loga i hm o exi s om Es adís ica de Va iaciones Residenciales by egion o esidence and coun y o des ina ion (2002–2012) o e he popula ion in he egion o esidence (pad on) in housands. Unemploymen a es o egions in Spain a e om he labou o ce su ey and unemploymen in o igin coun ies a he WEO. The auxilia y eg esso s a e he c oss coun y a e age o all dependen and independen a iables. Obse a ions a e weigh ed by he popula ion o he egion o o igin in Spain. S anda d e o s in pa en heses *p< 0.05; **p< 0.01 Izquie do e al. IZA Jou nal o Mig a ion (2016) 5:10 Page 15 o 20 leading o s ill pe sis en ly high emig a ion a es o Spania ds. To analyze he quan i a- i e impo ance o his ac o , we could ake he p e ious esul s (Table 8) as g an ed and y o es ima e he educ ion in he unemploymen a e ha would s abilize emi- g a ion a es. This is s aigh o wa d; using Eq. (4) yields: ΔUh −Us ðÞ¼ γ2 γ1 ΔShs −1ð6Þ Du ing he las yea s, he s ock o Spania ds li ing ab oad has been inc easing s ead- ily, and in he las 2 yea s, he inc ease o he a io has been a ound 0.7 (see Fig. 3). Thus, once we apply he abo emen ioned 0.7 o Eq. (6), one can unde s and why emi- g a ion a es con inued inc easing o Spania ds bo n in Spain a he same ime ha unemploymen in Spain dec eased (Fig. 2). In pa icula , in o de o s op mig a ion ou - lows, he unemploymen a e should ha e dec eased 4 pp in 2014, ins ead o he ac ual d op o 1.75 pp. Mo eo e , i is expec ed ha in 2015, emig a ion a es will con inue inc easing since unemploymen a e acco ding o o icial igu es only dec eased 2.3 pp, some hing insu icien o compensa e o he size o his es ima ed ne wo k e ec . Table 7 Log odds o immig a ing o Spain by o eigne s o emig a ing om Spain by Spania ds bo n in Spain as a unc ion o s ock o people om he same na ionali y in des ina ion coun y En y o o eigne s Exi o Spania ds bo n in Spain Fixed e ec s Fixed e ec s Va iables (1) (2) (3) (4) (5) (6) U(CCAA)-U(coun y) −0.02813 0.04627 (0.00001)** (0.00574)** U(coun y) 0.04739 0.04741 −0.07865 −0.07894 (0.00002)** (0.00002)** (0.00927)** (0.00926)** U(CCAA) −0.00946 0.02630 (0.00002)** (0.00726)** S ock o immig an s pai coun y/ egion by popula ion in o igin (in housands) 0.08518 0.08869 0.08976 0.14919 0.14924 0.14922 (0.00011)** (0.00011)** (0.00011)** (0.00479)** (0.00479)** (0.00479)** Time dummies Yes Yes Yes Yes Yes Yes Coun y dummies Yes Yes Yes Yes Yes Yes Region dummies Yes Yes Yes Yes Yes Yes Pai coun y/ egion dummies No No No No No No Auxilia y eg esso s No No No No No No Cons an −5.282 −8.393 −5.991 −6.012 −4.424 −5.178 (0.00035)** (0.00021)** (0.00058)** (0.12207)** (0.07732)** (0.22205)** R-squa ed 0.88 0.88 0.88 0.80 0.80 0.80 Obse a ions 6542 6542 6542 1080 1080 1080 Sou ce: In columns 1 o 3, he dependen a iable is he loga i hm o en ies om Es adís ica de Va iaciones Residenciales by coun y o bi h and egion o des ina ion in a pa icula yea (1998–2012) o e he popula ion in he coun y o bi h in he co esponding yea (WEO) in housands. In columns 4 o 6, he dependen a iable is he loga i hm o exi s om Es adís ica de Va iaciones Residenciales by egion o esidence and coun y o des ina ion (2002–2012) o e he popula ion in he egion o esidence (pad on) in housands. Unemploymen a es o egions in Spain a e om he labou o ce su ey and unemploymen in o igin coun ies a he WEO. The auxilia y eg esso s a e he c oss coun y a e age o all dependen and independen a iables. Obse a ions a e weigh ed by he popula ion o he o igin coun y in (1)-(3) o he egion o o igin in Spain in (4)-(6). S anda d e o s in pa en heses *p< 0.05; **p< 0.01 Izquie do e al. IZA Jou nal o Mig a ion (2016) 5:10 Page 16 o 20 To analyze he consis ency o he p e ious esul s o o he speci ica ions, we es ima e he endogenous componen o mig a ion ollowed by o eigne s as in Eq. (6). This new app oach is in o ma i e since, as men ioned p e iously, immig a ion lows o o eigne s s a ed in a momen when he e we e no o eigne s in Spain, some hing simila o wha is happening nowadays wi h Spania ds going ab oad. Es ima es show he pe sis ence o immig a ion lows in Spain in he pe iod o he c ea ion o he o eigne ne wo ks wi h an au o eg essi e pa ame e o 0.6. Applying his pa ame e o he cu en emig a ion a es o Spania ds bo n in Spain leads o simila conclusions o he ones p esen ed be- o e. In pa icula , he unemploymen a e in Spain should ha e dec eased 5 pp in 2014 Table 8 Log odds o immig a ing o Spain by o eigne s En y o o eigne s A ellano Bond Va iables (1) (2) (3) U(CCAA)–U(coun y) −0.02620 (0.00109)** U(coun y) 0.02344 −0.00210 (0.00415)** (0.00423) U(CCAA) −0.03037 (0.00123)** Lag. Log odds o immig a ing o Spain by o eigne s 0.61734 0.59858 0.59844 (0.00598)** (0.00663)** (0.00654)** Cons an −1.876 −2.279 −1.661 (0.03436)** (0.03930)** (0.04614)** R-squa ed 0.88 Obse a ions 5988 5988 5988 Sou ce: The dependen a iable is he loga i hm o en ies om Es adís ica de Va iaciones Residenciales by coun y o bi h and egion o des ina ion in a pa icula yea (1998–2012) o e he popula ion in he coun y o bi h in he co esponding yea (WEO) in housands. Unemploymen a es o egions in Spain a e om he labou o ce su ey and unemploymen in o igin coun ies a he WEO. The auxilia y eg esso s a e he c oss coun y a e age o all dependen and independen a iables. Obse a ions a e weigh ed by he popula ion o he o igin coun y. S anda d e o s in pa en heses *p< 0.05; **p< 0.01 Fig. 3 E olu ion o Spania ds bo n in Spain esiding ab oad Izquie do e al. IZA Jou nal o Mig a ion (2016) 5:10 Page 17 o 20 in o de o s op he g ow h in emig a ion a es in ha pa icula yea . Mo eo e , in 2015, emig a ion a es will con inue inc easing unless he unemploymen a e in Spain would dec ease by a ound 3 pp. 5 Conclusions This pape p o ides a i s look a he da a on mig a ion in lows and ou lows in Spain du ing he G ea Recession. Ou esul s hin a a signi ican change in he scale and composi ion o mig a ion in lows and ou lows. In e ms o scale, Spanish and o eign na ionals show qui e a simila esponse o unemploymen de elopmen s, and he con- inuing low exi a e o Spania ds bo n in Spain can only be a ibu ed o small ne wo k e ec s o Spanish emig an s. Howe e , as we lea ned om he apid c ea ion o ne - wo ks o o eign na ionals in Spain ha a ac ed many immig an s o Spain du ing he housing boom, his is a phenomenon ha could de elop qui e apidly, ega dless o u u e labo ma ke de elopmen s. In pa icula , he apid c ea ion o ne wo ks migh coun e balance he e ec o a all in unemploymen a es o a ound 3–4 pp o he ol- lowing yea . The possibili y o ne wo k e ec s s a ing o come in o play o Spanish emig an s, so ha many ou lows would become pe manen , is a po en ial h ea o Spanish po en ial g ow h. This is especially ue gi en ha ecen o eign mig an s seem o be posi i ely selec ed on educa ion (see Izquie do e al. 2014). 17 This inding hin s a he possibili y o he s a o a signi ican b ain d ain, which could exace ba e he e ec s o he c isis on po en ial ou pu i i we e o las oo long. Endno es 1 See, o ins ance, Ga cia Fe nandez (1965), Nadal (1984), and Ins i u o Español de Inmig ación (1973). 2 Since 2008, he INE has used al e na i e s a is ical echniques o imp o e he in o - ma ion held by municipal egis e s and has s a ed o compile s a is ics on mig a ions whose mic oda a a e no a ailable o us. 3 We ha e epea ed all ou compu a ions using he Es adís ica de Va iaciones Resi- denciales, ob aining quali a i ely simila esul s. Those calcula ions a e a ailable upon eques . 4 Acco ding o he New Immig a ion Su ey conduc ed in Spain in 2007, mo e han 85 % o immig an s planning o lea e in he nex 5 yea s epo ed ha hey in ended o e u n o hei bi h coun y. Mo eo e , acco ding o he Es adís ica de Mig aciones, he g ea majo i y o mig an s o di e en o igins (mo e han 70 %) epo ha hey e- u n o hei bi h coun y. See he companion pape o mo e in o ma ion on his issue. 5 To ob ain socioeconomic in o ma ion o he s ock, we use Labou Fo ce Su ey da a. 6 When e u n mig a ion by o eign na ionals was nonexis en . 7 On he adjus men o unemploymen and wages ac oss Spanish egions in p e ious ecessions, see Ben olila and Jimeno (1998). Ha ing said ha , unemploymen igu es migh be a wo se p oxy o economic oppo uni ies in o he coun ies. Ne e heless, we p e e ed o use he same a iable in bo h Spain and o he po en ial o igins/ des ina ions. Izquie do e al. IZA Jou nal o Mig a ion (2016) 5:10 Page 18 o 20 8 The co esponding popula ion o o eigne s is he coun y o o igin popula ion; o Spania ds o one pa icula egion in one pa icula coun y i is he popula ion esiding in ha coun y and who a e om ha pa icula egion. The a es will be always exp essed in housands o people. 9 Gi en he la ge dispa i ies in unemploymen a es ac oss Spanish egions (see Cha 1), we conside he log odds o esiding in a Spanish egion (we ha e access o da a on 17 egions) o one pe son o a pa icula coun y (we ha e access o da a on some 80 coun ies). 10 I he e is co ela ion be ween unemploymen a es o al e na i e des ina ion coun- ies, he es ima ed coe icien o he impac o unemploymen a e in he o igin coun- y on mig a ion lows is biased upwa ds. 11 In he case o o eign na ionals, o en ies, we use a sample o 31 coun ies o bi h ou o 80 po en ial coun ies, which esul s in less han 2 % o ze os in egion/coun y o o igin/yea be ween 1998 and 2012, and o depa u es, we keep 37 coun ies o bi h o he pe iod 2008–2012 ou o 99 po en ial coun ies, which esul s in less han 5 % o ze os. Fo ou lows o Spanish na ionals, we keep 30 ou o 97 po en ial des ina- ions o he pe iod 2008–2012, which esul s in 5 % o ze os. 12 Resul s a e a ailable upon eques . 13 A e age co ela ion o e ime o ac oss egion unemploymen is 86 % be ween wo egions, wi h he excep ion o he Basque Coun y which has an a e age co ela ion wi h he es o 68 %. 14 The s ock is compu ed as he a io be ween he s ock o people esiding ab oad and he s ock o people who did no mig a e (in housands). 15 Be oli e al. (2013b) analyzed a case s udy o Ecuado ian na ionals mo ing o Spain, showing he impo ance o changes in u u e expec a ions o inc ease he will- ingness o mo e o a coun y. 16 In pa icula , one migh p e e no o lag he s ock o Spania ds bo n in Spain esiding ab oad, since i is al eady lagged due o delays on he momen o egis a ion. Howe e , o he sake o compa abili y wi h p e ious esul s, we ha e p e e ed o main ain he same speci ica ion o bo h o eigne s and Spania ds as in Eq. (4). 17 Due o he lack o p ope da a, he pape canno edo he exe cise assuming di e en cos s o mig a ion by educa ional a ainmen . Howe e , using al e na i e in o - ma ion om he Labou Fo ce Su ey, Izquie do e al (2014) conclude ha he e is posi i e selec ion in hose households esiding in Spain ha send a membe o he household ab oad. Compe ing in e es s The IZA Jou nal o Mig a ion is commi ed o he IZA Guiding P inciples o Resea ch In eg i y. The au ho s decla e ha hey ha e obse ed hese p inciples. Acknowledgemen s We would like o hank he anonymous e e ee and he edi o o he use ul ema ks. Responsible edi o : Denis Fougè e Recei ed: 11 No embe 2015 Accep ed: 28 Ma ch 2016 Re e ences An olín P, Bo e O. Regional mig a ion in Spain: he e ec o pe sonal cha ac e is ics and o unemploymen , wage and house p ice di e en ials using pooled c oss-sec ions. Ox o d Bull Econ S a . 1997;59(2):215–35. A ellano M, Bond S. Some es s o speci ica ion o panel da a: Mon e Ca lo e idence and an applica ion o employmen equa ions. Re Econ S ud. 1991;58(2):277. Izquie do e al. IZA Jou nal o Mig a ion (2016) 5:10 Page 19 o 20 Beine M, Bou geon P, B icongne JC. Agg ega e luc ua ions and in e na ional mig a ion, CESi o Wo king Pape Se ies 4379, CESi o G oup Munich; 2013 Ben olila S, Dolado JJ. Misma ch and in e nal mig a ion in Spain, 1962–1986. In: Padoa-Schioppa F, edi o . Misma ch and Labou Mobili y. Camb idge: Camb idge Uni e si y P ess; 1991. Ben olila S, Jimeno JF. Regional unemploymen pe sis ence (Spain, 1976–1994). Labou Econ. 1998;5:25–51. Be oli S, B ücke H, Fe nandez-Hue as J. The Eu opean c isis and mig a ion o Ge many: expec a ions and he di e sion o mig a ion lows, IZA discussion pape s 7170, Ins i u e o he S udy o Labo (IZA). 2013a. Be oli S, Fe nandez-Hue as J, O ega F. C ossing he bo de : sel -selec ion, ea nings, and indi idual mig a ion decisions. J De Econ. 2013b;101:75–91. Dus mann C, Weiss Y. Re u n mig a ion: heo y and empi ical e idence om he US. B J Indus Rela . 2007;45(2):236–56. Ga cia Fe nandez J. La Emig ación Ex e io de España. Mad id: Edi o ial A iel; 1965. G ogge J, Hanson GH. Income maximiza ion and he selec ion and so ing o in e na ional mig an s. J De Econ. 2011;95:42–57. Ha is JR, Toda o MP. Mig a ion, unemploymen and de elopmen : a wo-sec o analysis. Am Econ Re . 1970;60(5):126–42. Ha on T, Williamson JG. Wha undamen als d i e wo ld mig a ion? NBER Wo king Pape No. 9159; 2002 Ins i u o Español de Emig ación. Pano ámica de la Emig ación. Edi o ial Minis e io de T abajo; 1973 Izquie do M, Jimeno JF, Lacues a A. Los Flujos Mig a o ios de España Du an e la C isis Economic Bulle in Banco de España; 2014 Massey D, A ango J, Hugo G, Kouaouci A, Pelleg ino A, Taylo JE. Theo ies o in e na ional mig a ion: a e iew and app aisal. Popul De Re . 1993;19(3):431–66. Nadal J. La Población Española (Siglos XVI a XX). Edi o ial A iel; 1984 O ega F, Pe i G. The e ec o income and immig a ion policies on in e na ional mig a ion, Mig a ion S udies, 1, 1–28; Ox o d Uni e si y P ess; 2013 Pede sen PJ, Py liko a M, Smi h N. Selec ion and ne wo k e ec s? Mig a ion lows in o OECD coun ies 1990-2000. Eu Econ Re . 2008;52(7):1160–86. P ados de la Escosu a L. Long- un economic g ow h in Spain since 1800: an in e na ional pe spec i e. In: Szi mai E, Van A k B, Pila D, edi o s. Explaining Economic G ow h. Ams e dam: No h Holland; 1993. p. 285–99. Za odny M. Wel a e and he loca ional choices o new immig an s Economic Re iew –Fede al Rese e Bank o Dallas; Second Qua e 1997. 1997. Submi you manusc ip o a jou nal and benefi om: 7 Con enien online submission 7 Rigo ous pee e iew 7 Immedia e publica ion on accep ance 7 Open access: a icles eely a ailable online 7 High isibili y wi hin he fi eld 7 Re aining he copy igh o you a icle Submi you nex manusc ip a 7 sp inge open.com Izquie do e al. IZA Jou nal o Mig a ion (2016) 5:10 Page 20 o 20