Can regional policies shape migration flows?

Our empirical analysis focuses on the effect of regional policies migration attraction factors in Europe. We employ a regression discontinuity design to assess causal relationship between reception large amounts public funds and flows EU-15 regions. In highly-subsidised regions, we find increase share foreign citizens from less-developed countries when compared low-subsidised regions with similar pre-treatment characteristics. The shows that such an is due positive impact European policy job market opportunities as well improvement goods supply. El análisis empírico de este estudio se centra en el efecto las políticas regionales sobre los factores atracción la migración Europa. Se empleó un diseño regresión discontinua para evaluar relación entre recepción grandes cantidades fondos públicos y flujos migratorios regiones UE-15. En altamente subvencionadas, encontró gran aumento proporción ciudadanos extranjeros procedentes países menos desarrollados, comparación con poco subvencionadas características similares antes del tratamiento. muestra que dicho debe al impacto positivo política europea oportunidades mercado laboral, así como mejora oferta bienes públicos. ?????????????????????????????????????????????????????????EU15?????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????? From economic point view, migrations are expression factor mobility across international borders and, therefore, globally efficiency production processes. However, social perspective, they complex events, often costly wasteful, which require resources flexibility both those emigrating who welcome immigrants. Faced dramatic migratory high-income (International Organization for Migration, 2019), strong heterogeneity literature has investigated what mechanisms guide choice migrate final destination, help developed govern this phenomenon. highlights driven mainly by opportunities. 11 seminal contribution Todaro (1969), people compare conditions their native country possibly faced host country. This comparison leads income-maximizing behavior so single agents, migrant, take into account differentials wages employment countries. Whereas embracing macroeconomic perspective emphasises key features labour different geographical areas, corresponding microeconomic framework (Todaro, 1976), represents investment human capital. Thus, location cost–benefit considering returns moving, net cost traveling, assimilating new culture unknown language, among other things. welfare system (Borjas, 1999, see also Conte & Mazza, 2019 survey), healthcare services (Borjas Hilton, 1996; Preston, 2014) provide migrants attain better living standards. Further dimensions represented natural amenities (Glaeser, Kolko, Saiz, 2001; Graves, proximity lakes green landscape, religious ethnic differences. hypothesis some these might be modified availability use Structural Cohesion Funds (EUF). EUF foster structural homogeneity context infrastructure, education, markets. theoretical entry based Hatton Williamson (2005), where individual more likely migrate, higher compensating differential is, namely, wage destination minus home country, lower fixed is. An important addition model development so-called magnet hypothesis, according select receiving depending generous benefits can get. Borjas (1999) programmes offered many US states have become Concerning countries, represent focus our analysis, De Giorgi Pellizzari (2009) estimate extent generosity affects decisions migrants. conclusions magnets but relatively weak, role conditions, unemployment rate level wages. paper suggests critical issue variation institutions EU will generate changes spatial distribution flows. recent report World Economic Forum (Schwab, 2017), inadequate or limited urban infrastructure indicated causes (push factors), whereas affordable accessible (including health care, utilities, transport) pull factors. context, there clear government intervention (see, example, Cebula, 2002). Policies stimulate growth, effects market, affecting attract migrants, however constrained limits placed spending also, state funding. relax constraints thus partially change direction space. There two main channels through influence flows, one direct indirect. channel concerns financing forms welfare, health, poverty immigration. programming period 2000–2006, dedicated activities was almost quarter. indirect regards growth therefore well-being region, attracting Several papers show Policy had performance convergence income wealth others, Basile, Castellani, Zanfei, 2008; Becker, Egger, Von Ehrlich, 2010; Pellegrini, Terribile, Tarola, Muccigrosso, Busillo, 2013). For Pellegrini et al. (2013) it estimated average 0.6% 0.9% per year financed Therefore, although certainly not taken deciding region do possible underlying choice. Estimating crucial importance because expands range information available policy-makers overall EUF. identification question. very task, first need identification, separates multiple migrations, second sought could weak statistically irrelevant. deals evaluation significance, counterfactual framework. looking margin (or discontinuity): given ex ante economically presence absence substantial funding affect differences degree attractiveness present allocation gives us opportunity (RDD), quasi-experimental method high internal validity (Lee Lemieux, 2010). Comparing scenario arising under situation (what would happened if been implemented), isolates confounding induced (see Mohl Hagen, 2010), then empirically evaluates existence significance robust way. To end, exploit rule transfers: less capita GDP purchasing power standards (PPS) below 75% average, qualify Objective 1 (Ob.1) status, receive most (called “treated” regions). assume non-eligible just above threshold “non-treated” regions), valid (Cerqua 2018). analyse foreigners without citizenship split estimation continent origin. results significant Ob.1 mostly number non-European immigrants coming specifications RDD hold even access sea bordering belonging vulnerable arrival legal illegal makes twofold contribution. First, vast meta-analysis Dall'Erba Fang, link arguably under-researched topic, data scarce. relevant exceptions Kessler, Hansen, Lessmann (2011), adopt residential political choice, Eggert, Larch (2014), whose geography model. analyses rely country-level data, — nature inequalities within lead explanatory models used migrations. created dataset containing at NUTS 2 years 2001 2011, intra- extra-European best knowledge, time set up. Second, already investigate Becker al., 2013), far aware approach never adopted Given increasing budget devoted since mid-1970s Europe, policy's shaping structured follows. its potential presented Section 2. methods discussed 3. 4, while 5 assesses robustness estimates. section concludes paper. invested locally 1975 improve competitiveness slow-growth correct unbalance. relevance widely increased 1988 landmark reform, followed accession Greece, Spain, Portugal widening disparities over previous 15 years. Since then, ceased increase, reduction goal Policy. Indeed, primarily assigned compensate them preconditions growth—infrastructure, accessibility, care (Camagni Capello, 2015). Nonetheless, targets changed time. initial priority unemployment, industrial reconversion, modernisation agriculture. priorities extended include terms innovation, environmental quality, poverty, shown composition types expenditure. 1989–1993 2000–2006 period, allocated supporting capital (e.g., training, inclusion) passed 20.6% 24.5%. infrastructures (from 1.6% 14%), directed non-environmental decreased 36.3% 30.9%) (European Commission, 2014). Interestingly, contributed programmes, which, turn, raised capacity least (named 2007–2013 period), defined level, PPS average. eligibility computed before beginning actual interventions. Commission basis values observed 1994–1996. fact forcing variable measured interventions guarantees treatment assignment mimics randomisation process around threshold. consisted 70% total allocations amounting €136 billion. 22 remaining part 2000-2006 targeted did comply “75 percent rule”, were suffering problems backwardness decline difficulties, long-term rural underdevelopment, low population density). received much aid intensity head respect (Pellegrini subsequent periods registered small rules, means still intact 2020; Ferrara, McCann, Stelder, 2017). It worth noting rule” allows procedure located cut-off suitable work same characteristics treatment. These following sections. territorial interest 2010 nomenclature migration, census provided Eurostat 2011 33 consider (EU-12) 2004 2004. peculiarities strategy lack EU-12 2001. glean ascribed, part, local EUF, facilities supply highly-financed accurate reporting EU-countries, (those belong EU), Using escape regarding fragmented definition constitutes limiting kind Westoff Frejka, 2007). 44 study, proxy migration. Although proxies foreign-born data), usually employed, available, immigration considered (Coleman, 2003). While concerning fully Belgium, Germany, French overseas-departments available: missing retrieved using figures Hellenic Statistical Authority (https://panorama.statistics.gr/en/) Statistics Belgium (http://www.statbel.fgov.be), respectively. Regarding collection implemented, German Federal Office collected micro-census covering 800,000 persons, sample size 1% 55 Data trace out interval censuses. Due reunification Germany occurred 1989, 1987. Schwarz, 2001). enables retrieve nationality fill data. Lastly, four overseas territories (Guadeloupe, Martinique, Reunion, Guyane) gathered Atlas National des Populations Immigrées each published Institut Statistique Etudes Economiques. relative censuses calculate percentage difference absolute groups. Such variables outcome variables. Unfortunately, directly refer EU-15, EU-27. EU-27 additional origin splitting Africa, Asia, Central South America, North Oceania. several demographic covariates databases Cambridge Econometrics Eurostat, linked decisions. are: density, 65, divided active population, 66 individuals working (in employment) residing region. productivity (gross value added (GVA) hour worked), hours worked employee, primary sector, secondary sector. As “untreated” status period. 77 starts 2001, i.e., after concern bulk spent including up end non-Ob.1 expenditures than two-thirds budget. Furthermore, payments Roemisch, 2016). continuous measure test sensitivity binary gain insights Before carrying map plot start investigating patterns. Figure illustrates geographic foreigner deciles (Panel A) B). instance, decile consists shares 1.3%. light blue. tenth 9.2% 26.6%. They dark navy. intermediate shades highest localised core EU, UK, Ireland. Ten later, observe somewhat pattern steep Italian, Portuguese, Spanish Panel C depicts allowing identify trends A striking feature figure top (the ones marked orange) concentrated periphery EU-15. reports amount during 2001–2010 Both panels 1, suggesting foreigners. peripheral tend consideration. evident 3, kernel-weighted polynomial smoothing regression. line points foreigners, flattens only right-hand side fund distributions. Nevertheless, rigorous econometric determine whether spurious correlation. With intent, rest paper, larger exogenous rule. perform exploits (calculated three period) qualified funds, considerable (which funds) comparisons funds). economies levels attributes, science technology endowments, institutional quality (Iammarino, Rodríguez-Pose, Storper, 2019). sharp exploited via inter alia, Hahn, Todd, van der Klaauw, Lee compares lying closely either delivers (LATE), threshold, (i.e., good randomised. words, equivalent random point. overcomes inverse causality problem, implicitly analysis. begins graphical evidence. simple way evaluate dependent (pre-treatment capita) 4 plots against graph, sharply treated untreated Each superimposes fit quadratic (estimated separately point). clearly that, seems citizens. evidence showing formal zero. obtained specification detailed subsection 3.2 Table 1. Columns (1) (2) citizenship, X. negative 5% specifications. ?2.23 corresponds 4.84% 6.45% 2011. columns, (columns (3) (4)) extra-Europeans (5) (6)). all coefficients positive, non-Europeans zero analysed brought about wide expansion non-Europeans. referring EU-27, columns (7) (8), check no confirming come So far, looked groups citizenships, non-European. informative conceal derive rather addition, natives reduce underestimating reasons, complement group estimates reported findings arrivals migrants—especially countries—rather citizenship. 35,000 sizable, especially averag