Economics of Migration Jan Fidrmuc Brunel University
Introduction n Important, and very contentious, aspect of economic integration and globalization q q 175 mn (2. 9%) int’l migrants in 2000, 190 mn today (WB Migration Database, 2007) Most (37%) from LDC to DC; 24% LDC, 16% DC DC Immigrants: 8 -12% of population in US, Germany, France, UK; 18 -21% Canada & Australia; 38% HK Emigrants: 5 -10% of Mexico, Afghanistan, Morocco, UK, Algeria, Italy, Germany, Bangladesh, Turkey; 0. 5 -0. 9% China, US, India
Introduction n n EU Single Market: free movement of labor Migrants (foreign born): 11. 7% of EU 15 population in 2005 (OECD) q n Approximately 1/3 EU foreigners EU enlargements in 2004 and 2007 q q q Forecasted East-West migration 3 -4 mn Temporary restrictions imposed by most EU 15 countries for up to 7 years Large E-W influx to Austria, Germany, Spain, UK and Ireland
Outline 1. 2. 3. 4. 5. 6. 7. 8. Introduction Models of migration Economic Impact of migration: Theory and evidence Brain Drain; Remittances Labor-market Performance of Migrants Politics of Immigration EU Enlargement and East-West Migration Conclusions
Economics of Migration n n Most common type of migration: LDC to DC Revealed preference: migrants move iff they expect to be better off at destination Yet, migrants often suffer occupational downgrading, end up in poorly-paid informal jobs or remain unemployed Ex-ante vs ex-post: Harris-Todaro Model q Michael P. Todaro, AER 1969; and John R. Harris and Michael P. Todaro, AER 1970.
Harris-Todaro Model n n Original focus: rural-urban migration in LDC Rural residents move to urban regions despite already high unemployment there Puzzle: migration continues although it makes (some) rural migrants worse off HT model: migrants motivated by expected returns q Expected returns may be different from actually realized returns
Harris-Todaro Model n Two regions: urban and rural q q n n Full employment in rural region Involuntary urban unemployment q q n Rural wage: w. R (farming) Urban wage: w. U>w. R Fraction q of urban workers hold jobs 1 -q are unemployed and have zero earnings Urban wages downward rigid q Minimum-wage rules, unionization, or b/c workers must acquire residence/work permits
Harris-Todaro Model n n Workers are risk neutral Migration continues as long as: w. U*q>w. R Migration is optimal despite unemployment Migration from LDC to DC similar case
Harris-Todaro Model: Implications Urban job creation (government spending) raises q migration more attractive Improving education in rural areas may increase migration if educated rural worker face higher q Rising rural wages reduce incentive to migrate to urban region 1. 2. 3. n However, if migration costly, rising rural incomes may relieve liquidity constraints on migration
Other Models of Migration: Beyond Income Differentials n n n Stark (The Migration of Labor, 1991) Households vulnerable to idiosyncratic shocks that are region or sector-specific Migration household members exposed to different regional shocks Risk diversification through pooling of household members’ income remittances Migration optimal even without income differentials if individuals risk averse consumption smoothing through risk sharing
Other Models of Migration: Roy-Borjas Model n n Roy (OEP 1951); Borjas (AER 1987) Consider two countries, A and B q q n n Identical mean earnings Different income distributions: returns to human capital higher in A Individual returns to migration depend on one’s skills Skilled workers fare better in A
Other Models of Migration: Roy-Borjas Model n Migration patterns: q q n Skilled migration to A Unskilled migration to B Returns to human capital important also when mean earnings not identical q DC – LDC migrants often highly skilled professionals and managers
Economics of Migration: Impact n n Trade theory: free trade, free capital mobility and free labor mobility should have similar effects on the economy Yet, migration more controversial than either free trade or capital mobility Popular view: immigrants displace native workers and/or drive down wages Is this consistent with theory and evidence?
Economic Impact of Migration: Theory n n Dustmann et al. , EJ 2005; Dustmann et al. , Ox. Rev. Ec. Pol 2008) Two countries: Home and Foreign One output good, price set at world market Two types of labor: skilled and unskilled q n Labor supplied inelastically Supply of capital perfectly elastic q Interest rate set at world markets
Economic Impact of Migration: Theory n If skill composition of immigration the same as that of natives q q q n No labor-market effect of migration This is because capital supply is elastic Economy adjusts to immigration by importing capital no change in relative endowments Migrants in DC predominantly unskilled compete with native unskilled workers only q Consider case with only unskilled immigrants
Economic Impact of Migration: Theory n n n L and L* unskilled workers in Home and Foreign Demand for labor given by MPL initial wage w 0 and w*0; w 0>w*0 Migration equalizes wages: w 1=w*1 Unskilled workers in Home worse off Migrants and unskilled workers in Foreign better off
Economic Impact of Migration: Theory MPL* w 0 w 1 w*0 o L L* M L+M o*
Economic Impact of Migration: Theory n n n What about overall impact on Home? Capital supplied elastically no impact for owners of capital Skilled labor in Home becomes scarcer relative to unskilled labor Skilled-wage premium goes up Overall effect: average earnings go up q q Immigration surplus: unskilled workers paid less than their marginal product Net gain accrues to skilled workers
Economic Impact of Migration: Theory MPL Immigration surplus MPL* w 0 w 1 w*0 o L L* M L+M o*
Economic Impact of Migration: Theory n Immigration has important distributional implications q q n This can have important political implications Note: if labor supply flexible, migration leads to unemployment in addition to (or instead of) lower unskilled wages Note: reverse holds for Foreign: skilled labor becomes less abundant and skilled workers lose out
Economic Impact of Migration: Heckscher-Ohlin Model n n n Multiple heterogenous output goods Free and competitive trade goods prices set at world markets All countries have access to the same technology Skilled and unskilled labor, supplied inelasticly Capital supply elastic All immigrants unskilled
Economic Impact of Migration: Heckscher-Ohlin Model n n Output mix determined by relative factor endowments Immigration pressure on unskilled wages to fall output of goods produced by unskilled labor goes up Wages of skilled and unskilled labor unchanged as long as goods prices constant Immigration absorbed through changes in output mix (Rybczinski Theorem) q No labor-market impact at all
Economic Impact of Migration: Evidence n Most studies: no or mildly negative impact of migration on natives’ wages or employment q q q Card (EJ 2005) US data, Dustmann et al. (EJ 2005; Ox. Rev. EP 2008) UK data Borjas (QJE 2003, NBER WP 2005): labor-market impact of migration is mitigated by out-migration of natives (US data) Card (EJ 2005): little evidence of natives’ migration being driven by immigration
Economic Impact of Migration: Evidence n n n Dustmann et al. (Ox. Rev. EP 2008) UK data Immigrants predominantly low skilled impact on wages different alongside natives’ wage distribution Elasticity of natives’ wages with respect to immigration rate q q q Low-wage earners: -0. 5 at 10 th percentile Positive for most: 0. 6 at median, 0. 35 on average Insignificant for high-wage earners (from 95 th percentile)
Economic Impact of Migration: Evidence (Dustmann et al. , 2008)
Economic Impact of Migration: Evidence n n Impact on employment (Dustmann et al. , EJ 2005) UK data: zero effect overall Positive effect for high-educated natives (high -school diploma and higher) Negative effect for intermediate-educated, negative but insignificant for unqualified Effects for high and intermediate educated approximately cancel each other in aggregate
Economic Impact of Migration: Evidence from Natural Experiments n Friedberg and Hunt (JEP 1995): Large-scale immigration episodes little long-term impact on labor markets q q q French and Portuguese decolonization Cuban immigration to the US during the Mariel boatlift Russian-Jewish immigration to Israel in 1990 s
Economic Impact of Migration: Israel n n 1990 s: more than 1 million ethnic Jews immigrated to Israel from the FSU Israeli population in 1989: 4. 6 million Migration driven by economic hardship and political unrest Approximately two-thirds of these immigrants highly skilled
Economic Impact of Migration: Israel n FSU immigration no long-term effect on wages or employment of natives q n Friedberg (2001 QJE), Gandal, Hanson and Slaughter (2004 EER), and Cohen and Paserman (2004 CEPR DP 4640): Cohen and Paserman (2004): negative shortterm effect (elasticity -0. 1 to -0. 3) on wages (but not on employment) q Effect disappears in 4 -7 years
Economic Impact of Migration: Israel n n n Gandal et al. : global technology changes increased demand for skilled labor This helped Israeli economy absorb immigrant influx Cohen and Hsieh (2000 mimeo): immigration followed by large influx of capital (borrowing) q Consistent with standard neoclassical growth model
Economic Impact of Migration: Israel n n Eckstein and Weiss (2003 IZA DP 710): substantial initial occupational downgrading of FSU immigrants but wages increase rapidly No return on their imported skills in short run Lower return to education, same return to experience and higher return to unobserved skills, than native Israelis, in long run Because of lower return to imported skills, immigrants’ wages never catch up with natives’ wages
Economic Impact of Migration: Germany n n n Re-unification of Germany large migration flow from East to West Frank (2007 mimeo): no overall effect on wages or unemployment in West Germany But: important distributional effects Employment of less educated workers, bluecollar workers and foreign nationals declined Wages of workers in non-traded-goods and service sectors increased
Immigration and Crime n n Immigrants often associated with high crime rates Theory: ambiguous relationship q q q Immigrants fare poorly in labor market, but: face higher detection probability (prejudice) and stricter punishment (sentence & deportation)
Immigration and Crime n Bianchi, Buonanno & Pinotti (2009 BI wp) q n OLS: elasticity of crime to immigration: 0. 1 q n Especially for property crime (theft and robbery) Result may be driven by endogeneity q n Immigration & crime in IT provinces, 1990 -2003 Eg immigrants more to high-crime areas because of low cost of housing IV: no significant effect on total crime or property crime, significant effect on robberies q Robberies 1. 5% of total crime only
Brain Drain n n Migration of skilled workers from LDC to DC Docquier et al. (IZA DP 2005): brain drain estimates, 1990 -2000 q q q n World weighted-average skilled migration rate 5. 3% vs unskilled rate 1. 1% LDC: 7% vs 0. 3% Latin America 11%, Africa 10. 4%, Asia 5. 5% Traditional view: brain drain reduces stock of human capital lower potential for growth
Brain Drain n Skilled immigrants often subject to occupational downgrading But face better employment prospects than unskilled immigrants Liquidity constraints q q Migration is costly Skilled migrants better able to afford the cost
Brain Drain: Reassessment n n n Mountford (1997 JDE), Fan and Stark Individuals under-invest in education because they ignore social returns Education raises probability of emigration q q n n Higher expected private return to education Greater incentive to invest in education Emigration uncertain some skilled workers remain in LDC Brain drain may raise LDC stock of human capital better prospects for growth
Remittances n Large inflows, esp. for developing countries: q n n Top recipients in 2008: Mexico, China and India: $25 -27 bn Mexico: remittances 1/3 of formal wage income in 2006 (Vargas-Silva, RDE 2009) q n Remittances one third of exports and greater than FDI (Barajas et al. , 2009, IMF WP 09/153) US: 18 mn people of Mexican origin Poland: 2 mn Poles abroad $6 bn in 2007 (NBP report)
Remittances n China: remittances of rural migrants large q n 15% of agricultural income of selected provinces in 1992 (Wu & Zhou 2005) 1995 survey in Jinan and Shandong (Liu & Reilley, Apllied Economics 2004) q q Rural migrants remit RMB 2110 p. a. 36% of earnings 85% of rural migrants transfer remittances
Remittances: Impact n n n Remittances increase household disposable income Rozelle, Taylor and de. Brauw (AER P&P 1999): remittances increase agricultural productivity in rural China (Hebei and Liaoning) Adams and Page (WB WPS 3179): remittances reduce poverty in LDC q Elasticity: -0. 19 with respect to emigration rate and -0. 16 with respect to remittances-to-GDP ratio
Remittances: Impact n Barajas et al. (2009, IMF WP 09/153): q q q Remittances to 84 countries over 1970 -2004 Effect on growth is insignificant or even negative Interpretation: remittances alleviate poverty and increase consumption but not investment
Remittances and the Dutch Disease n n n DD typically associated with revenue from export of natural resources (origin: North-Sea natural gas exports in the Netherlands) Large receipts of foreign currency XR appreciates loss of competitiveness manufacturing exports fall while imports rise Vargas-Silva (RDE 2009): remittances cause appreciation of real exchange rate in Mexico
Politics of Immigration n n Migration (and trade) winners and losers among natives If immigrants predominantly unskilled, unskilled natives lose out and skilled workers gain Losers may need to be compensated (redistribution of gains from winners) Otherwise, economic integration may not be politically feasible
Politics of Immigration n Wages reflect relative abundance of each factor of production Consider again skilled vs unskilled labor DC: skilled labor relatively abundant q q n Immigrants predominantly unskilled Skilled workers likely to emigrate LDC: unskilled labor relatively abundant q q Immigrants predominantly skilled Unskilled labor likely to emigrate
Politics of Immigration n n Attitudes depend on relative factor endowments and redistributional impact of immigration DC: immigrants predominantly unskilled q q n n Skilled wage goes up Unskilled wage falls Skilled workers should favor immigration Unskilled workers should oppose immigration
Politics of Immigration n n LDC: immigrants predominantly skilled Skilled workers should oppose immigration Unskilled workers should be in favor Attitudes on free trade determined similarly
Politics of Immigration n n O’Rourke and Sinnott (2005 EJPE), Mayda and Rodrik (2005 EER) and Mayda (2005) use large multi-country individual-level survey dataset to investigate individual attitudes on trade and migration Skilled individuals more in favor of immigration (free trade) q More so in rich countries
Politics of Emigration: Home Country n n Emigration experience may affect one’s political opinion and attitudes Spilimbergo (CEPR DP 5934): q n n UNESCO database on international student flows: 1950 -2003 Share of students studying abroad increases democracy in home country But only if students study in democratic countries
Politics of Emigration: Home Country n n Fidrmuc and Doyle (CEPR DP 4619): Voting behavior of Czech and Polish emigrants in home-country elections Emigrant votes differ from home country votes and also across host countries Votes for pro-reform and left-wing parties depend on host-country characteristics Migrants adapt to institutional environment q q Level and tradition of democracy Extent of economic freedom
Political Impact of Immigration n Living in economically liberal and democratic countries should have a favourable impact on migrants from less developed countries q n Autocratic regimes often restrict their citizens’ freedom to travel q n Migrants espouse liberal attitudes while living abroad North Korea, Turkmenistan, Zimbabwe Autocracies that tolerate free travel often more liberal former Yugoslavia
Migrant’s Labor-market Performance n n Migrants typically subject to substantial occupational downgrading Human capital poorly portable q q n Eastern European Jews in Israel: low return on imported education and experience (Friedberg JLE 2000; Eckstein and Weiss, 2003 IZA DP 710) Destination-country education raises return also to home-country education (Friedberg 2000) Immigrants catch up slowly and imperfectly
Occupational Downgrading: UK (Dustmann et al. , 2008)
Migrant’s Labor-market Performance n US/UK: immigrants from Latin American and Eastern Europe suffer more downgrading than immigrants from industrialized countries q q n UK: Drinkwater et al. (2006 IZA DP 2410) US: Mattoo et al. (JDE 2008) Language skills important q Immigrants who speak destination-country language earn up to 20% more (Chiswick and Miller, 2002 JPop. E; 2007 IZA DP 2664)
EU Enlargement and East-West Migration n EU enlargements in May 2004 and Jan 2007 Pre-enlargement debate in academia, policymaking and popular press: fear of mass migration, welfare shopping and displacement effects in labor market Result: transitional restrictions on free movement of workers (2+3+2 formula) q q Exceptions in 2004: UK, Ireland Sweden More countries removed restrictions later
EU Enlargement and East-West Migration (Zaiceva & Zimmerman, Ox. Rev. EP 2008) n Fear of mass migration: high unemployment:
EU Enlargement and East-West Migration (Zaiceva & Zimmerman, Ox. Rev. EP 2008) n Fear of mass migration: low wages:
East-West Migration: Predictions (Zaiceva & Zimmerman, Ox. Rev. EP 2008) n 1. 2. Since early 1990 s – more than 30 studies forecasting East-West migration Predictions based on intentions to move to the West using surveys Econometric models using historical data for countries other than CEECs (“double outof-sample extrapolations”) q Migration experience after the Southern EU enlargement or other countries (e. g. immigration to Germany from a broad sample of countries)
East-West Migration: Predictions n Boeri and Bruecker (2000) q q n n n Estimate model of immigration to Germany over 1968 -98 Use it to predict East-West migration to the EU Total net immigration 335 ths to the EU 15 Most predicted to go to Germany and Austria: 218 ths and 40. 5 ths, respectively; UK to receive some 15 ths.
EU Enlargement and East-West Migration (Zaiceva & Zimmerman, Ox. Rev. EP 2008)
East-West Migration: Predictions n n n Surveys of willingness to migrate (WTM) Drinkwater (2003): WTM to the EU highest among the young and those with high skills and foreign-language skills Expected impact: q q mainly positive impact for destination countries potentially adverse impact on source countries (brain drain)
EU Enlargement and East-West Migration (Zaiceva & Zimmerman, Ox. Rev. EP 2008) n Actual migration flows difficult to estimate q q q n UK and Ireland: no restrictions, large influx (Gilpin et al. , 2006, Blanchflower et al. , 2007, UK Home Office Accession Monitoring Reports; Doyle et al. , 2006) Sweden: no restrictions, little immigration Austria, Germany: restrictions, large influx Why? Push and pull factors: Economic factors, geographic proximity, language, networks, demand-driven (IE), re-directed from other countries…
EU Enlargement and East-West Migration (Zaiceva & Zimmerman, Ox. Rev. EP 2008)
EU Enlargement and East-West Migration (Zaiceva & Zimmerman, Ox. Rev. EP 2008) n Main sending countries: q q q n UK: Poland, Slovakia, Lithuania, Ireland: Poland, Lithuania, Latvia Sweden: Poland, Lithuania, Estonia Sectoral distribution of immigrants: q q q UK: hotels/catering, manufacturing, agriculture/construction Ireland: construction, manufacturing, hotels/catering Sweden: health care, trade, manufacturing
EU Enlargement and East-West Migration (Zaiceva & Zimmerman, Ox. Rev. EP 2008)
EU Enlargement and East-West Migration (Zaiceva & Zimmerman, Ox. Rev. EP 2008)
EU Enlargement and East-West Migration (Zaiceva & Zimmerman, Ox. Rev. EP 2008)
East-West Migration: Impact (Zaiceva & Zimmerman, Ox. Rev. EP 2008) n n n No negative impact on receiving countries’ economies Occupational downgrading common No evidence of an impact on unemployment If displacement of natives (Ireland) no rise in aggregate unemployment but “upgrade” jobs for nationals No wage pressure, not even in manufacturing sector highest share of new immigrants
East-West Migration: Impact (Zaiceva & Zimmerman, Ox. Rev. EP 2008) n n Reduced or falling wage growth rates mainly follow pre-enlargement trends UK: immigration has increased supply by more than it has increased demand q Lower inflationary pressures and lower natural rate of unemployment.
East-West Migration: Public Perceptions (Zaiceva & Zimmerman, Ox. Rev. EP 2008)
East-West Migration: Public Perceptions (Blanchflower, EJ 2009) Balance 60 Per cent 12 Unemployment expectations over the next 12 months: EU 15 (3 month average - advanced 12 months - LHS) 50 11 40 10 30 9 20 8 10 7 0 Unemployment rate (RHS) --- 1985 -2007 average -10 1985 6 5 1988 1991 1994 1997 2000 2003 2006
East-West Migration: Public Perceptions Balance 60 Unemployment expectations over the next 12 months: Germany (3 month average - advanced 12 months - LHS) Per cent 14 50 12 40 10 30 8 20 6 10 4 0 -10 Unemployment rate (RHS) --- 1985 -2007 average 0 -20 1985 1988 1991 2 1994 1997 2000 2003 2006
East-West Migration: Public Perceptions Balance 60 Unemployment expectations over the next 12 months: UK (3 month average - advanced 12 months - LHS) 50 Per cent 12 11 10 40 9 30 8 20 7 10 6 5 0 -10 Unemployment rate (RHS) --- 1985 -2007 average -20 1985 4 3 2 1988 1991 1994 1997 2000 2003 2006
East-West Migration: Public Perceptions Balance 60 Unemployment expectations over the next 12 months: Ireland Per cent 20 (3 month average - advanced 12 months - LHS) 50 40 15 30 20 10 10 0 -10 5 -20 -30 Unemployment --- 1985 -2007 average rate (RHS) -40 1985 1988 1991 1994 1997 2000 2003 2006 0
East-West Migration: UK Experience n n n Annual gross inflow of A 8 nationals over 200 ths The stock of A 8 migrants estimated to be around 500 ths by the end of 2006 (Blanchflower, Saleheen and Shadforth, 2007) 65 -70% of A 8 immigrants are Polish Most work in low-skilled occupations and earn low wages 80% are below 35; 60% are males
East-West Migration: UK Experience n Unlike migrants from other countries, relatively low share come to London (around 10%) Main destinations: Anglia, Midlands, London Anecdotal evidence: some immigrants highly skilled No welfare tourists: by 2007, less than 1, 000 A 8 nationals per quarter approved to receive income support and job-seekers allowance q n n
East-West Migration: UK Experience WRS registrations as WRS a % of home Registrations population (000 s) U Rate (2004) Emp Rate (2004) GDP per head (2005)* Czech Rep. 0. 28 28. 9 8. 3 64. 2 5, 200 € Estonia 0. 47 6. 2 9. 7 63. 0 4, 000 € Hungary 0. 19 18. 9 6. 1 56. 8 5, 000 € Latvia 1. 43 32. 8 10. 4 62. 3 3, 100 € Lithuania 1. 85 62. 8 11. 4 61. 2 2, 500 € Poland 1. 02 394. 2 19. 0 51. 7 4, 200 € Slovakia 1. 13 61. 2 18. 2 57. 0 4, 200 € Slovenia 0. 03 0. 6 6. 3 65. 3 11, 400 € 0. 560 -0. 257 -0. 711 Correlation
East-West Migration: UK Experience Life Satisfaction Bulgaria Czech Rep Estonia Hungary Latvia Lithuania Poland Romania Slovakia Slovenia UK 2004 2. 06 2. 82 2. 74 2. 44 2. 52 2. 55 2. 81 2. 32 2. 59 3. 17 3. 22 2006 1. 99 2. 92 2. 74 2. 50 2. 62 2. 80 2. 33 2. 70 3. 09 3. 18
East-West Migration: UK Experience WRS Applications Intended length of stay Less than 3 months 12 months ending March 2007 Per cent 126, 100 55% 3 to 5 months 3, 840 2% 6 to 11 months 7, 605 3% 1 to 2 years 10, 520 5% More than 2 years 21, 225 9% Do not know 58, 480 26% Total 227, 770 100%
East-West Migration: UK Experience n n 1 -2% (at least) of Polish, Slovak, Latvian and Lithuanian populations lived and worked in the UK at some time between 2004 and 2007 (WRS statistics only) East-West migration helped lower unemployment in A 8 countries q n Some caused labor shortages, especially in agriculture and construction Gilpin et al. (2006, DWP WP 29): relationship between change in share of A 8 migrants and change in regional unemployment
East-West Migration: UK Experience n n n Drinkwater, Eade and Garapich (2006 IZA DP 2410): A 8 migrants highly skilled but have low returns to skills and experience Wages increase with years since migration A 8 migrants earn 30% less than migrants from other European countries (those from English-speaking countries earn 30% more) when not controlling for occupation No evidence as to whether the extent of occupational downgrading is falling over time
East-West Migration: UK Experience n If occupational downgrading persists: q q n n Destination countries do not realize the migrants’ full contribution to their GDPs Migrants fail to receive wages corresponding to their human capital Occupational downgrading may become permanent and persist even upon return to the home country (scarring) Source countries suffer brain drain Overall, everyone loses Time will show if this will be the case.
Summary n Factors underlying patterns of migration flows: q q q n Differentials in (expected) earnings Returns to human capital Risk sharing at household level Economic impact of migration q q Host country: little or none aggregate labor-market impact but important distributional implications Home country: may be harmful because of brain drain and/or Dutch disease
Summary n Political impact q q n Host country: distributional implications winners and losers political backlash against labor mobility and integration Home country: spread of liberal norms and values EU enlargement q q Large and unprecedented migration flows within Europe Little or no negative labor-market impact on host countries (so far)
Скачать презентацию Economics of Migration Jan Fidrmuc Brunel University
560a1e92901b80a5939e90d75c00fca5.ppt