Labor Market Conditions and Geographic Mobility in the Eurozone.
The Eurozone has survived an economic crisis that began a decade ago. The global financial crisis in 2007-2008 was followed by a deep recession in 2008-2009, a European debt crisis in 2010, and a second recession in the Eurozone in 2012-2013. To avoid a collapse of the European monetary union (EMU), the Eurozone countries, that were severely hit by the crisis (Greece, Ireland, Portugal, Spain, and Cyprus), received financial assistance. Yet, recovery has been uneven. Some countries are still struggling to cope with the consequences of massive job losses. The Eurozone unemployment rate was 9.1% in 2017 (from 7.5% in 2007 to 12% in 2013), but the highest unemployment rates were in Greece at 21.5%, Spain at 17.2%, Italy and Cyprus at 11.1%, whereas the lowest unemployment rates were in Germany at 3.8%, Malta at 4.0%, and the Netherlands at 4.9%. (1)
The severity of the Eurozone crisis shows that mechanisms of adjustment to shocks within EMU are lacking or don't work properly. In this respect, the literature about optimum currency areas (OCA) is useful for drawing some lessons (Krugman 2012; Eichengreen 2014; Gibson et al. 2014; Aizenman 2018). According to this literature, the costs of joining a monetary union and giving up the exchange rate as an economic instrument could be minimized if some other mechanisms of adjustment to shocks were available. Geographic labor mobility is among such mechanisms, and its extent is considered as a criterion for optimality (Mundell 1961; Kenen 1969). (2)
In this paper, we analyze geographic mobility in the Eurozone. We provide a measurement of mobility in the Eurozone and an assessment of its responsiveness to national labor market conditions. Thus, we address the following questions: What is the extent of mobility in the Eurozone? Do people move from one country to another in response to an increase in unemployment? Does mobility across the Eurozone respond to national wage developments as well? How did the crisis affect mobility?
In theory, labor mobility can help prevent the persistence of unemployment in a country hit by an adverse shock as long as unemployed people move to other countries. Besides, within a monetary union, a common currency ought to facilitate labor mobility between member countries thanks to price transparency and reduced transaction costs (elimination of conversion fees). (3)
In reality, however, it is common knowledge that labor mobility is low in Europe in comparison with the experience of the USA, Canada or Australia (Heinz and Ward-Warmedinger 2006; European Commission 2015). There is also evidence that the role of labor mobility in labor market adjustments is limited in Europe, but it has increased over time (Dao et al. 2014; Jauer et al. 2014; Beyer and Smets 2015; Arpaia et al. 2016). It is worth noting that empirical studies relate to mobility flows in the European Union (EU), especially between the latter and the rest of the world, but not in the Eurozone.
Yet, from the standpoint of the analysis of optimum currency areas, it is interesting to know whether citizens in the Eurozone migrated to another country in the Eurozone or elsewhere during the crisis. For example, the entry of Spanish citizens in Germany was about 34,300 in 2014 against an average of inflows of about 8100 per year before the crisis (1999-2007). Likewise, inflows of Greek citizens in Germany peaked at 32,600 in 2012 against 12,600 on average before the crisis. In contrast, Irish and Portuguese people moved to non-EU countries: for the most part, Irish citizens headed toward Australia (inflows of about 6200 in 2014 against an average of 1300 before 2008), while Portuguese citizens went to Switzerland (a peak of about 19,900 in 2013 against 10,000 on average before 2008). (4)
The original feature of our work lies precisely in the fact that we analyze mobility in the Eurozone. It is a threefold contribution.
First, our approach is based on a direct measure of migration: we use data on stocks of foreign population by nationality to compute net inflows of migrants between Eurozone countries. (5) This measure provides valuable information about mobility flows in the Eurozone. In contrast, in most studies (e.g., Eichengreen 1993; European Commission 2011; Jauer et al. 2014), mobility is measured indirectly by migration-induced population growth (crude rate of migration). (6) With such a measure, the country of origin of migrants is not known. There are also studies that use direct measures of mobility based on bilateral flows. However, panel data include non-European countries (European Commission 2015; Beine et al. 2017) or Germany only as the destination country (Bertoli et al. 2013).
Second, we calculate three net migration rates for Eurozone countries. (7) These rates differ depending on the group of countries of origin of migrants: Eurozone countries, EU countries that are not members of the Eurozone (8), and non-EU countries. They are useful to compare mobility flows in the Eurozone with those between Eurozone countries and other countries from the EU or the rest of the world.
Last, we apply the model proposed by Pissarides and McMaster (1990) to estimate the response of net migration rates to national labor market conditions. Eichengreen (1993) and the European Commission (2011) used this model, but they measured mobility with crude rates of migration. Our regression tests are an attempt to determine whether mobility flows in the Eurozone changed during the crisis and reacted to national differences in unemployment or wage developments. (9)
Our approach certainly has some limitations. We study mobility between countries and not between regions. We do not use regional data because there is no direct measure of mobility flows at the regional level in the EU. Besides, due to data availability issues, our dataset includes 14 out of 19 Eurozone countries and covers the 1999-2015 period. The fourteen countries are the first eleven countries that adopted the euro in 1999 (Austria, Belgium, Finland, France, Germany, Ireland, Italy, Luxembourg, Netherlands, Portugal and Spain), plus Greece (joining the Eurozone in 2001), Slovenia (2007), and Slovakia (2009). The five countries that are not included in the dataset are Cyprus and Malta (the two islands have been Eurozone members since 2008), and the three Baltic countries (Estonia adopted the euro in 2011, Latvia in 2014, and Lithuania in 2015). The problem of missing data is particularly acute for these countries. (10)
It is important to note that, in our empirical analysis, the 14 Eurozone countries are destination countries. The countries of origin of migrants are the 19 Eurozone countries, the EU countries (except Eurozone countries) and non-EU countries (the rest of the world as a whole). Our three measures of net migration rates are ultimately based on an aggregate of bilateral net inflows between the 14 Eurozone countries and each group of countries of origin (Eurozone, EU countries outside the Eurozone, non-EU countries).
Unfortunately, it is not possible to study the impact of the creation of the euro on labor mobility because the sample period is too short. We look at the impact of the crisis instead. Furthermore, we do not examine the mobility of workers but of people. Our empirical approach is indeed based on data of foreign population of all ages, because there is a lack of data on migrants of working age from Eurozone countries. (11)
The paper is organized as follows. In "Related Literature" section, we review the related literature. In "Methodology" section, we explain the methodology, and in "Data" section, we present the data. We comment our estimation results in "Estimation Results" section and conclude in "Conclusion" section.
Our purpose is not to study the determinants of migration. It is instead to determine whether mobility of people responds to national labor conditions in the Eurozone. (12)
The theoretical background rests on the literature about optimum currency areas (OCA). Mundell (1961) advocated the role of factor mobility in a currency union as an alternative instrument to a flexible exchange rate. Kenen (1969) explained how labor mobility between regions could alleviate regional unemployment problems by decreasing labor supply in the region of origin which is in decline and increasing labor supply in the destination region which is in expansion. This mechanism of adjustment works well if skills of migrants prove to be suitable to those that are required in the destination region. In addition, as long as mobility increases the capital-labor ratio in the region of origin, the required fall in real wages for workers who stay is reduced (Eichengreen 1990). Employment conditions could even improve for people who stay if migrants still consume home products once they are settled in the host country (Farhi and Werning 2014). Indeed, labor demand could increase thanks to higher exports.
Yet, labor mobility may not be helpful if there are labor market imperfections, which prevent wages from adjusting. It can even aggravate the problems of regions of origin if the latter suffer from brain drain (Eichengreen 2014). Skill-selective migration (i.e., the higher propensity to migrate for highly-skilled workers) increases disparities in regional income per capita (Fratesi and Riggi 2007). The overall impact on regions of origin is ambiguous because there is a U-shaped migration pattern with respect to skill levels due to asymmetric information (Akkoyunlu and Vickerman 2001), and return migration needs to be accounted for (Holland and Paluchowski 2013). (13)
From an empirical standpoint, studies, that were carried out before the launch of EMU, concluded that labor mobility across European countries or regions was not sufficiently high to play a significant role as a mechanism of adjustment to asymmetric shocks. In this respect, Europe was far from being an optimum currency area in comparison with the USA (De Grauwe and Vanhaverbeke 1993; Eichengreen 1993; Decressin and Fatas 1995). More recent studies have relied on the methodology of Blanchard and Katz (1992), as Decressin and Fatas (1995) did earlier: a vector autoregressive model is estimated in order to capture the role of labor mobility in absorbing the effects of a shock on regional (national) employment. In this kind of approach, mobility is taken as a residual. The main conclusion is that the role of mobility is still limited in Europe, but it has somewhat increased over time (Dao et al. 2014; Beyer and Smets 2015; European Commission 2015; Arpaia et al. 2016). In contrast, it has decreased much in the USA. (14)
Our work belongs to another strand of the literature, which seeks to determine to what extent labor mobility between countries (regions) reacts to national (regional) disparities in labor market conditions. The latter are often described by relative unemployment and wage rates. The relative unemployment rate is defined as the difference between the unemployment rate in the destination country and the average unemployment rate of a reference group (the sample of countries). The relative wage rate is computed similarly.
Some works are based on the crude rate of migration as a measure of mobility. At a country level, for a panel of 22 EU countries over the 1991-2009 period, the European Commission (2011) finds that an increase of one percentage point in relative unemployment leads to a decrease of 0.25 percentage point in net inflows, whereas an increase of one percentage point in relative real wage leads to an increase of 0.07 percentage point in net inflows. These results apply to a sample that includes noneuro countries and years before the creation of the euro. At a regional level, for a group of 17 Eurozone countries and the 2006-2011 period, Jauer et al. (2014) find that the influence of relative unemployment is very small (-0.017) and that of relative GDP per capita (0.005) is not statistically significant.
A few studies are based on a direct measure of bilateral flows of migrants. Beine et al. (2017) estimate equations of gross inflows between 30 developed countries over the 1980-2010 period. They use a dummy variable for EMU and find that it increases mobility significantly. The European Commission (2015) considers flows between 38 destination countries and 163 origin countries over the 1992-2011 period. The effect of EMU as a dummy variable is not robust to the inclusion of country fixed effects (nor is it for relative GDP per capita). A variable of interaction between EMU and relative unemployment has a high statistical significance (the estimated coefficient is - 0.18).
Eichengreen (1993) and implicitly the European Commission (2011) used the model of Pissarides and McMaster (1990) to analyze the issue of optimum currency areas in Europe from the point of view of labor mobility. We apply this model to the Eurozone. The intuition goes as follows. In terms of incentives, individuals, who contemplate moving to other countries, compare the gains and costs of geographic mobility.
Costs depend on individual characteristics such as gender, age, education, skills, household composition, living accommodation. They are also influenced by institutional impediments that take the form of linguistic and cultural barriers, administrative difficulties in transferring social security rights from one country to another, inefficiencies in the housing markets. In addition, some policies may diminish incentives (increase the opportunity cost) to move out, such as generous unemployment benefits (Hassler et al. 2005). (15) Also, costs of geographic mobility can be reduced within a currency union as long as countries share a single currency which lowers transaction costs.
For given mobility costs, the gains from moving are increasing with job opportunities, in particular with the probability of being hired and earning a higher income in the destination country. A country with relatively lower unemployment rate or higher average wage is then expected to experience higher net inflows of migrants (inflows minus outflows). The empirical method thus consists of regressing net inflows of migrants in a destination country on relative unemployment rate and relative average wage between the destination country and other countries:
[m.sub.i,t] = [[beta].sub.1] [m.sub.i,t-1] + [[beta].sub.2] ([[mu].sub.i] + [[beta].sub.3] [([DELTA][w.sub.i] - [DELTA][bar.w]).sub.t-1] + [[gamma].sub.i] + [[gamma].sub.i] (1)
where [m.sub.i,t] denotes the net migration rate in destination country i at time t. It is computed by dividing net inflows by the average population of the destination country during the year. The resulting net migration rate is expressed per 1000 people, A positive (negative) rate means that inflows are above (below) outflows.
The presence of lagged net inflows [m.sub.i,t-1] can be justified by the persistence of patterns of bilateral migration. (16) The variable [([u.sub.i] - [bar.u]).sub.t-1] is the lagged relative unemployment rate which is defined as the difference between the unemployment rate in country i and the average unemployment rate of the group of countries considered (the Eurozone in our case). Similarly, the variable [([DELTA][w.sub.i] - [delta][bar.w]).sub.t-1] is the lagged relative average real wage growth which is defined as the difference between country i's average real wage growth and the average real wage growth of the reference group (Eurozone). (17)
We expect that the sign of the coefficient [[beta].sub.2] is negative and that of [[beta].sub.3] is positive. There is one lag in the explanatory variables to account for delays in the response of potential migrants to changing conditions (they base their decision to move on information about economic conditions which is available with a lag). (18) The problem of reversed causality is attenuated by the fact that at each period the extent of bilateral migration flows is small compared to the size of national labor markets (Beine et al. 2017).
Country fixed effects ([[gamma].sub.i]) are included to account for the impact of time-invariant factors that are not included in the model but influence the decision to move (e.g., distance or common language). Time fixed effects ([[gamma].sub.t]) can be added to capture determinants of mobility that are time-specific.
In robustness checks of estimation results of Eq. (1), we used a specification in which wage growth is replaced by GDP per capita growth. Indeed, in the literature, some studies use GDP per capita as a proxy for wages because of a lack of data (Jauer et al. 2014; European Commission 2015). We also constructed dummy variables to test the influence of some European institutional features about the free movements of people, such as the Schengen area and some transitional arrangements in the context of the latest EU enlargements (see next section).
With regard to estimation methodology, we could use the Least Square Dummy Variable (LSDV) procedure. However, due to the limited number of observations, especially the time dimension (T) of sub-periods before and after the crisis, the estimated coefficients may be biased. Indeed, in dynamic panel data models, the LSDV estimation procedures are asymptotically valid only when the number of observations in the time dimension is large. Nevertheless, it has been shown that the LSDV estimator, although inconsistent, has a relatively small dispersion in comparison with various consistent estimators such as the Generalized Method of Moments (GMM) and the Instrumental Variables (IV) (Kiviet 1995; Judson and Owen 1999).
Here, GMM and IV estimators are not well suited for our empirical strategy because the number of cross-sectional units (N) is small. Bruno (2005) found that in the case of small N, the bias-corrected LSDV (LSDVC) emerged as the preferred estimator in dynamic panel data fixed effects models. And in the case of small T, Kiviet (1995) found in a Monte Carlo study that LSDVC estimators were not outperformed by any other estimator whatever the size of the sample. In the same way, Judson and Owen (1999) found that the LSDVC estimator was overall the best choice for such models. Therefore, we use the LSDVC estimator.
Our dataset comprises 14 Eurozone countries. The sample period is 1999-2015. We use data on total population and stocks of foreign population by nationality from Eurostat database. We supplement existing data with data collected from various sources: OECD International Migration Database; United Nations Global Migration Database; Directorate-General Statistics and Economic Information of Belgium; Holland et al. (2011), Adsera and Pytlikova (2015). For each country, we compute net bilateral inflows of migrants by country of origin by taking the difference of stocks between two consecutive years. The use of stocks of foreign population implies that net inflows correspond to mobility of foreigners but not mobility of natives. Therefore, we compute net inflows of natives by subtracting net inflows of foreigners from total migration (i.e., net migration plus statistical adjustment in Eurostat). We then add net inflows of natives to net inflows of foreigners. As a result, our measure of net inflows for each Eurozone country of our sample includes not only mobility of foreigners but also mobility of citizens of the country considered.
Next, for each of these 14 countries, we compute aggregates that represent net inflows of migrants by group of countries of origin. We define three main groups: the world, the EU, and the Eurozone. (19) Finally, we make the sum of aggregates of all 14 Eurozone countries to work with a panel dataset. The three broad measures of net inflows are the following:
* "Mobility with the world" stands for net inflows of migrants in the 14 Eurozone countries from non-EU countries.
* "Mobility with the EU" represents net inflows of migrants in the 14 Eurozone countries from EU countries that are not members of the Eurozone.
* "Mobility in the Eurozone" corresponds to net inflows of migrants in the 14 Eurozone countries from the 19 Eurozone countries (including the fourteen). This is the novelty of our approach.
Furthermore, we take into account some European institutional features which are related to the free movements of people and workers in the EU. (20) In the Schengen area, border controls are abolished. All fourteen Eurozone countries, except Ireland, are members of the Schengen area (Greece since 2000, Finland since 2001, Slovenia and Slovakia since 2008). It is also worth knowing that EU member states were allowed to apply restrictions on the free movement of workers from eight member states ("EU8") that joined the EU in 2004, because they feared that there would be massive inflows of workers and that this would disrupt their national labor markets. (21) Such restrictions could be maintained during a 7-year transitional period, and could also be applied to two new member states ("EU2") that joined the EU in 2007 (Bulgaria and Romania), and to Croatia that joined the EU in 2013. Table 8 in Appendix shows the various dates at which such restrictions were lifted. In our sample of 14 Eurozone countries, two countries are new EU member states (Slovenia and Slovakia). In some old EU member states (Austria, Belgium and Germany), the removal of restrictions on workers from EU8 took place at the end of the 7-year transitional period (in 2011), which happened to be right in the middle of the Eurozone crisis.
This information is used to build three dummy variables for each Eurozone country i:
* "Schengen" takes the value of 1 if the Schengen agreement is into force in country i, 0 otherwise.
* "FreeEU8" takes the value of 1 if free access to labor market of country i is granted to citizens of 8 new member states (from the EU enlargement in 2004), 0 if restrictions still apply.
* "FreeEU2" takes the value of 1 if free access to labor market of country i is granted to citizens of 2 new member states (from the EU enlargement in 2007), 0 if restrictions still apply.
With regard to macroeconomic variables, the source is the AMECO database of the European Commission. We use the unemployment rate (Eurostat definition). The average real wage variable is computed by dividing nominal compensation per employee (1000 EUR) by the price deflator of private final consumption expenditure. Real GDP per capita is gross domestic product at 2010 reference levels per head of population (1000 EUR). In robustness checks, relative GDP per capita growth is the difference between GDP per capita growth in destination country i and the average GDP per capita growth in the Eurozone. Summary statistics are shown in Table 9 in Appendix.
Figure 1 shows that the dispersion of unemployment rates across Eurozone decreased before the crisis and increased during the crisis. It started to fall in 2013. In contrast, the dispersion of average wage and that of GDP per capita were relatively stable between 1999 and 2015.
Figure 2 indicates that mobility of people in the Eurozone is slightly lower than mobility with the world (non-EU countries) but higher than mobility with the rest of the EU. On average, it amounts to 1.6 [per thousand] of population over the sample period (1999-2015). It doesn't change much over time, and there is no perceptible decreasing trend during the crisis period. Mobility with other EU countries (1.3 [per thousand]) is lower than mobility with non-EU countries (1.7 [per thousand]), despite EU enlargements (as pointed out by the ECB, 2012).
At country level (Fig. 3), net migration rates increased in the countries that were least affected by the crisis (Austria, Germany, and Luxembourg). Not surprisingly, they declined in the countries that were severely hit by the crisis: in particular, in Greece (net outflows), Portugal (to a lesser extent), Ireland (but net migration rates turned positive again with economic recovery in 2015), and Spain (net outflows concerned mobility with the world only). In this respect, Fig. 5 in Appendix shows that during the crisis period, Spain did not face net outflows of Spanish citizens but foreigners. This may mitigate concerns about brain drain (see "Related Literature" section). In Ireland, net outflows mostly involved Irish people, but there was a reversal in 2015. On the contrary, concerns apply in Greece and Portugal where net outflows of natives were still not reversed in 2015.
Figure 4 displays mobility in the Eurozone and relative unemployment. Net migration rates tend to be positive in countries where the unemployment rate is below the average unemployment rate in the Eurozone, and negative in countries where the unemployment rate is above the average.
We report estimation results of Eq. (1) for each of the three measures of mobility: with the world (Table 1), with the EU (Table 2), and in the Eurozone (Table 3). The first three columns of each table display results with country fixed effects and the last three columns show results with both country and time fixed effects. (22)
Results can be summarized as follows:
* Considering both country and time fixed effects over the whole sample period (1999-2015), the response of mobility flows to relative unemployment is stronger for mobility with non-EU countries (-0.50) than with EU countries (-0.35) and in the Eurozone (-0.24).
* During the crisis period (2008-2015), the response of mobility to relative unemployment is significantly stronger than during the whole period with regard to mobility with EU countries (Table 2) and mobility in the Eurozone (Table 3). It is still lower regarding mobility in the Eurozone (-0.41) than mobility with EU countries (-0.48) and the world (-0.43).
* The estimated coefficient of the relative wage growth variable is not statistically significant (and sometimes with a sign that is different from the expected one), except for mobility with the world before the crisis (0.44). This result is not surprising because, as we said before, the dispersion of average wage across the Eurozone is relatively stable over the sample period (see Fig. 1 in the previous section).
Results from tests with country fixed effects only and results from tests with both country and time fixed effects are, for the most part, consistent with each other. We therefore display results from robustness checks with both country and time fixed effects.
In Table 4, the relative wage growth variable is replaced by the relative GDP per capita growth variable. The influence of this variable on mobility is not statistically significant for mobility with the EU and in the Eurozone. In contrast, the estimated coefficient of this variable is statistically significant in the case of mobility flows with the world (0.30), especially during the period before the crisis (0.63).
The effects of dummy variables about EU agreements (Schengen, FreeEU8, FreeEU2) are considered in Tables 5, 6, and 7. In the case of mobility with the world (Table 5), free access of workers from EU2 countries plays a strong and significant role in net inflows in Eurozone countries during the 2008-2015 period (the estimated coefficient is 2.47). The latter result could reflect the fact that some citizens from non-EU countries found it easier to move to some Eurozone countries after the removal of restrictions to free movement of workers from Bulgaria and Romania. As for mobility with the EU (Table 6), the free movement of workers from EU8 countries plays a strong and positive role in net inflows (the estimated coefficient is statistically significant over the whole sample period only). However, the dummy variables do not play any role with regard to mobility in the Eurozone (Table 7). In particular, it is not surprising that the removal of restrictions on the access of workers from EU8 countries to national labor markets has no significant effect on mobility in the Eurozone during the sample period, because Central and Eastern European countries that are members of the Eurozone are small countries, and the removal did not take place at the same time in all countries of the Eurozone.
Finally, our results are not strictly comparable with those of the European Commission (2011), European Commission (2015), because we use a direct measure of mobility in the Eurozone. We find a stronger influence of relative unemployment on mobility, and this influence is stronger and significant during the crisis period (2008-2015) than during the whole sample period (1999-2015).
We have studied the influence of national labor market conditions on mobility flows of people in the Eurozone during the crisis. The originality of our approach is to use a direct measure of mobility that is based on stocks of foreign population by nationality. We could thus compute net inflows of people between Eurozone countries. We found that mobility in the Eurozone was relatively low during the 1999-2015 period. It was lower than mobility between Eurozone countries and non-EU countries, but relatively higher than mobility with other EU countries that are not members of the Eurozone. It did not decline during the crisis. It responded to national differences in unemployment, especially during the crisis period, but not to national differences in wage developments.
It remains that labor mobility is unlikely to be the best option to cope with a sharp drop in employment when all countries are faced with a severe shock at the same time (even if consequences turn to be country-specific). National differences in unemployment can persist over time if industrial structures and job requirements differ greatly from one country to another. Labor mobility may not be a proper mechanism of adjustment to shocks if skill-biased migration flows widen regional income disparities. Furthermore, costs of mobility for unemployed people should not be overlooked.
In terms of policy implications, the EU could take new measures to help unemployed persons move to places with better job opportunities. The European Social Fund and the European Globalization Adjustment Fund are instruments of EU labor market policies that provide support to people who lose their jobs and help them upgrade their skills and find a job. Some allowances of these EU programs could be earmarked for facilitating labor mobility across the EU. For example, they could serve as a financial assistance for expenses that are caused by a lack of automatic recognition of diplomas and qualifications (language proficiency tests, translation and authentication of documents, credential evaluations). Furthermore, the portability of social benefits throughout the EU needs to be reinforced. Indeed, in practice, the transfer of acquired rights to another country may prove to be difficult.
Amid other policy options that would help people (and economies) adjust to adverse shocks, there are measures that improve the functioning of labor markets (skills mismatch, labor taxation, regulations of working time, wage setting), service industry (regulation of entry, opening hours, price controls), housing markets (construction permits, housing shortage, access to affordable housing), and financial sector (access to banking services).
In future research, our dataset could be used in a gravity model to study whether national differences in language and social protection systems help or impede geographic mobility of people in the Eurozone and in the EU.
Acknowledgements We are grateful to Alberto Bagnai, Paul Wachtel, the journal's co-editor, three anonymous referees, Andreea Stoian, Cecile Couharde, Etienne Farvaque, participants at the INFER Conference (Bordeaux, 2017), and participants at the GdRE Conference (Paris, 2017), for useful suggestions. This work is part of a research contract (PACaPe) headed by Xavier Chojnicki (LEM-CNRS, University of Lille), and as such, it has benefited from the financial support of the Directorate General for Foreigners in France (DGEF) of the Ministry of the Interior.
See Fig. 5, Tables 8 and 9.
Adsera, A., and M. Pytlikova. 2015. The role of language in shaping international migration. The Economic Journal 125: F49-F81.
Aizenman, J. 2018. Optimal currency area: A twentieth Century Idea for the twenty-first century? Open Economies Review 29 (2): 373-382.
Akkoyunlu, S., and R. Vickerman, 2001. Migration and the efficiency of european labour markets. In Spatial change and interregional flows in the integrating Europe. Contributions to economics, ed. J. Brocker, and H. Herrmann. Physica-Verlag HD.
Alesina, A., R. Barro, and S. Tenreyro, 2002. Optimal currency areas. NBER Macroeconomic Annual 17: 301-55.
Arpaia, A., A. Kiss, B. Palvolgyi, and A. Turrini. 2016. Labour mobility and labour market adjustment in the EU. IZA Journal of Migration 5 (21): 1-21.
Beine, M., F. Docquier, and C. Ozden. 2011. Diasporas. Journal of Development Economics 95: 30-41.
Beine, M., P. Bourgeon, and J.C. Bricongne. 2017. Aggregate fluctuations and international migration. The Scandinavian Journal of Economics.
Bertoli, S., H. Briicker, and J. Fernandez-Huertas Moraga. 2013. The European crisis and migration to Germany: Expectations and the diversion of migration flows. IZA discussion paper no. 7170.
Beyer, R., and F. Smets. 2015. Labour market adjustments in Europe and the US: how different? Economic Policy 30 (84): 643-682.
Blanchard, O., and L. Katz. 1992. Regional evolutions. Brookings Papers on Economic Activity 1: 1-75.
Boltho, A., and W. Carlin. 2013. EMU's problems: Asymmetric shocks or asymmetric behavior? Comparative Economic Studies 55: 387-403.
Bruno, G.S.F. 2005. Estimation and inference in dynamic unbalanced panel-data models with a small number of individuals. The Stata Journal 87: 361-366.
Chiswick, B., and P. Miller (eds.). 2015. Handbook of the economics of international migration. Amsterdam: Elsevier B.V.
Decressin, J., and A. Fatas. 1995. Regional labor market dynamics in Europe. European Economic Review 39: 1627-55.
De Grauwe, P., and W. Vanhaverbeke. 1993. Is Europe an optimum currency area? Evidence from regional data. In Policy Issues in the Operation of Currency Unions, ed. P. Masson, and M. Taylor, 111-129. Cambridge: Cambridge University Press.
Dellas, H., and G. Tavlas. 2009. An optimum-currency-area odyssey. Journal of International Money and Finance 28 (7): 1117-37.
Dao, M., D. Furceri, and P. Loungani. 2014. Regional labor market adjustments in the United States and Europe. IMF working paper no 14/26.
European Central Bank (ECB). 2012. Euro area labour markets and the crisis. Structural issues report, October.
Eichengreen, B. 1990. One money for Europe? Lessons from the US currency union. Economic Policy 10: 117-87.
Eichengreen, B. 1993. Labor markets and European monetary unification. In Policy issues in the operation of currency unions, ed. P. Masson, and M. Taylor, 130-62. Cambridge: Cambridge University Press.
Eichengreen, B. 2014. The Eurozone crisis: the theory of optimum currency areas bites back. Notenstein Academy white paper series.
European Commission. 2011. Labour market developments in Europe 2011. European Economy 2/2011.
European Commission. 2015. Labour market and wage developments in Europe 2015. Report coordinated by Strauss R Arpaia A and Duiella M.
Farhi, E., and I. Werning. 2014. Labor mobility within currency unions. NBER working paper no. 20105.
Frankel, J., and A. Rose. 1998. The endogeneity of the optimum currency area criteria. The Economic Journal 108: 1009-25.
Fratesi, U., and M.R. Riggi. 2007. Does migration reduce regional disparities? The role of skill-selective flows. Review of Urban and Regional Development Studies 19(1): 78-102.
Gallin, J. 2004. Net migration and state labor market dynamics. Journal of Labor Economics 22 (1): 1-21.
Gibson, H.D., T. Palivos, and G.S. Tavlas. 2014. The crisis in the euro area: An analytic overview. Journal of Macroeconomics 39: 233-39.
Hassler, J., J.V. Rodriguez Mora, K. Storesletten, and F. Zilibotti. 2005. A positive theory of geographic mobility and social insurance. International Economic Review 46 (1): 263-303.
Heinz, F.F., and M. Ward-Warmedinger. 2006. Cross-border labour mobility within an enlarged EU. ECB occasional paper no 52.
Holland, D., T. Fic, A. Rincon-Aznar, L. Stokes, and P. Paluchowski. 2011. Labour mobility within the EU--The impact of enlargement and the functioning of the transitional arrangements. NIESR report.
Holland, D., and P. Paluchowski. 2013. Geographical labour mobility in the context of the crisis. European Employment Observatory Ad-hoc Request.
Jauer, J., T. Liebig, J. Martin, and P. Puhani. 2014. Migration as an adjustment mechanism in the crisis? A comparison of Europe and the United States. OECD social, employment and migration working papers No. 155.
Judson, R.A., and A.L. Owen. 1999. Estimating dynamic panel data models: A guide for macroeconomists. Economics Letters 65: 9-15.
Kalemli-Ozcan, S., B. Sorensen, and O. Yosha. 2001. Economic integration, industrial specialization and the asymmetry of macroeconomic fluctuations. Journal of International Economics 55: 107-37.
Kenen, P. 1969. The theory of optimum currency areas. In Monetary problems of the international economy: An eclectic view, ed. R.A. Mundell, and A.K. Swoboda, 41-60. Chicago: Chicago University Press.
Kiviet, J.F. 1995. On bias, inconsistency, and efficiency of various estimators in dynamic panel data models. Journal of Econometrics 68: 53-78.
Krugman, P. 1993. Lesson of Massachusetts for EMU. In Adjustment and growth in the European Monetary Union, ed. F. Torres, and F. Giavazzi, 241-61. Cambridge: Cambridge University Press.
Krugman, P. 2012. Revenge of the optimum currency area. NBER Macroeconomics Annual 27: 439-448.
Mundell, R.A. 1961. A theory of optimum currency areas. American Economic Review 51 (4): 657-65.
Pedersen, P., M. Pytlikova, and N. Smith. 2008. Selection and network effects--Migration flows into OECD countries 1990-2000. European Economic Review 52: 1160-86.
Pissarides, C., and I. McMaster. 1990. Regional migration, wage and unemployment: empirical evidence and implications for policy. Oxford Economic Papers 42 (4): 812-31.
Saks, R., and A. Wozniak. 2011. Labor reallocation over the business cycle: New evidence from internal migration. Journal of Labor Economics 29 (4): 697-739.
Tatsiramos, K. 2009. Geographic labour mobility and unemployment insurance in Europe. Journal of Population Economics 22: 267-83.
Wagner, H. 2014. Can we expect convergence through monetary integration? (New) OCA theory versus empirical evidence from European integration. Comparative Economic Studies 56: 176-199.
Zaiceva, A., and K.F. Zimmermann. 2008. Scale, diversity, and determinants of labour migration in Europe. Oxford Review of Economic Policy 24 (3): 267-83.
Florence Huart  * Medede Tchakpalla 
[mail] Florence Huart
 LEM-CNRS, Universite de Lille, Cite scientifique (SH2), 59655 Villeneuve d'Ascq Cedex, France
(1) Source: AMECO database, European Commission.
(2) The OCA criteria include economic features of countries that lower the probability of asymmetric shocks (diversification of production, trade interdependence, and business cycle synchronization), and mechanisms of adjustment to shocks (price and wage flexibility, factor mobility, and fiscal integration). For a review of the literature, see Delias and Tavlas (2009).
(3) This effect can be linked to the idea that the fullfilment of OCA criteria is endogenous to monetary integration (Frankel and Rose 1998; Alesina et al. 2002). The idea itself is nonetheless controversial (Krugman 1993; Kalemli-Ozcan et al. 2001). For a critical appraisal of the literature in the light of EMU, see Boltho and Carlin (2013), Wagner (2014), and Aizenman (2018).
(4) Source: OECD International Migration Database.
(5) In this paper, the terms "mobility" and "migration" are interchangeable. Net inflows of migrants are the difference between inflows of people (the number of immigrants) and outflows of people (the number of emigrants). If net inflows are negative, it means that outflows exceed inflows.
(6) The crude rate of migration is the difference between the total change and the natural change (births minus deaths) of the population. This difference is divided by the average population during the year and expressed per 1000 people.
(7) The net migration rate is the ratio of net inflows of migrants to the average population of the destination country over a given year. It is expressed per 1000 people.
(8) There are 9 out of 28 EU countries that are not members of the Eurozone: Bulgaria, Croatia, Czech Republic, Denmark, Hungary, Poland, Romania, Sweden, and United Kingdom.
(9) The analysis of how mobility in turn influences labor market adjustments is beyond the scope of this paper (a short review of this subject is made in "Methodology" section).
(10) No data on stocks of foreign population by nationality is available in the case of Cyprus and Malta. It is, therefore, not possible to compute any flows for any year. As for the other countries, the lack of data is such that we could not compute flows before 2011 for Estonia, between 2007 and 2011 for Latvia, and before 2014 for Lithuania. Given the late entry to the Eurozone, the analysis would not have produced any meaningful results if the dataset had contained these countries. In other respects, despite missing data in the case of France, we still could compute net inflows of migrants from other Eurozone countries for at least 7 years (of which 5 years are part of the crisis period).
(11) The Eurostat database includes data from the European Union Labor Force Survey on foreign population by citizenship, age, and labor status, but the definition of country of citizenship is very large (EU-28, EU-15, Total).
(12) A good introduction to the literature about international migration is the handbook edited by Chiswick and Miller (2015). Among studies about labor mobility in Europe, Zaiceva and Zimmermann (2008) explore the historical background and the enlargement of the EU to Central and Eastern European countries. Holland and Paluchowski (2013) examine the context of the recent crisis.
(13) This is currently an unsettled issue because data on skill levels of migrants in the EU is patchy (missing data for some countries) and not available for a period of time sufficiently long to draw some reliable conclusions.
(14) In this matter, Saks and Wozniak (2011) argue that interstate migration in the USA responds to common factors to all locations (national business cycle) rather than local economic conditions.
(15) At the same time, unemployment benefits lower search costs to find a job and may thus not be associated with lower mobility (Tatsiramos 2009).
(16) For example, the latter can be explained by network effects (Pedersen et al. 2008; Beine et al. 2011).
(17) Pissarides and McMaster (1990) preferred the use of wage growth to wage levels, because they obtained better results from the regression analysis. Eichengreen (1993) found that, for Italy, the wage variable should be in levels and not in first difference. We also did the tests with wage levels and found that conclusions are not affected by the definition of this variable (results are available upon request).
(18) In contrast, Gallin (2004) and Bertoli et al. (2013) add some variables to the equation to take into account the role of expectations of future economic conditions. Results are yet inconclusive in the European case.
(19) We take into account the changing composition of the Eurozone (from 11 countries in 1999 to 19 countries in 2015).
(20) We use online information which is provided by various Directorates-General in the European Commission.
(21) The eight new member states are Central and Eastern European Countries: Czech Republic, Estonia, Hungary, Lithuania, Latvia, Poland, Slovenia, and Slovakia. There were no restrictions for Cyprus and Malta, given the tiny size of their population.
(22) The software Stata does not supply the R-squared value when one uses the XTLSDVC command (LSDVC estimator). For the reader, we provide it by calculating the correlation between the observed response and the predicted response and then squaring it.
Caption: Fig. 1 Dispersion of unemployment rate, average wage and GDP per capita in the Eurozone. Panel data: 14 Eurozone countries. Note: the relative standard deviation is the ratio of the standard deviation to the mean. It is used here because the variables are expressed in different units. Source: own calculations
Caption: Fig. 2 Net migration rates in 14 Eurozone countries (average, per 1000 of people). Panel data: 14 Eurozone countries. Source: own calculations
Caption: Fig. 3 Net migration rates in individual countries (per 1000 of people). Note: the vertical line stands for 2008. Source: own calculations
Caption: Fig. 4 Mobility in the Eurozone countries and relative unemployment. Panel data: 14 Eurozone countries. Note: the vertical line stands for 2008, the left axis for net migration, and the right axis for relative unemployment. Source: own calculations
Caption: Fig. 5 Net migration rates of foreign nationals and citizens of the reporting country (per 1000 people). Source: own calculations
Table 1 Mobility with the world (dependent variable: net migration rate) (1) (2) (3) 1999-2015 1999-2007 2008-2015 Net inflows (t - 1) 0.34 *** 0.32 ** 0.17 (0.08) (0.13) (0.13) Relative unemployment -0.49 *** -0.45 * -0.34 (t - 1) (0.10) (0.23) (0.23) Relative wage growth 0.01 0.31 ** 0.02 (t - 1) (0.08) (0.13) (0.10) Observations 190 84 93 Country fixed effects Yes Yes Yes Time fixed effects No No No [R.sup.2] 0.23 0.83 0.30 (4) (5) (6) 1999-2015 1999-2007 2008-2015 Net inflows (t - 1) 0.35 *** 0.29 ** 0.16 (0.08) (0.14) (0.13) Relative unemployment -0.50 *** -0.41 * -0.43 * (t - 1) (0.10) (0.25) (0.23) Relative wage growth -0.05 0.44 ** -0.30 (t - 1) (0.13) (0.18) (0.24) Observations 190 84 93 Country fixed effects Yes Yes Yes Time fixed effects Yes Yes Yes [R.sup.2] 0.27 0.09 0.35 Panel data (14 Eurozone countries), LSDVC estimator with country and time fixed effects Standard errors in parentheses. * p < 0.1; ** p < 0.05; *** p < 0.01 Table 2 Mobility with the EU (dependent variable: net migration rate) (1) (2) (3) 1999-2015 1999-2007 2008-2015 Net inflows (t - 1) 0.30 *** 0.20 0.01 (0.06) (0.12) (0.12) Relative unemployment -0.35 *** -0.79 -0.44 ** (t - 1) (0.11) (0.52) (0.21) Relative wage growth -0.01 0.09 -0.10 (t - 1) (0.09) (0.26) (0.11) Observations 183 82 89 Country fixed effects Yes Yes Yes Time fixed effects No No No [R.sup.2] 0.17 0.35 0.12 (4) (5) (6) 1999-2015 1999-2007 2008-2015 Net inflows (t - 1) 0.31 *** 0.27 ** -0.02 (0.06) (0.13) (0.12) Relative unemployment -0.35 *** -0.73 -0.48 ** (t - 1) (0.11) (0.54) (0.22) Relative wage growth -0.09 0.10 -0.37 * (t - 1) (0.14) (0.38) (0.20) Observations 183 82 89 Country fixed effects Yes Yes Yes Time fixed effects Yes Yes Yes [R.sup.2] 0.24 0.11 0.13 Panel data (14 Eurozone countries), LSDVC estimator with country and time fixed effects Standard errors in parentheses. * p < 0.1; ** p < 0.05; *** p < 0.01 Tabic 3 Mobility in the Eurozone (dependent variable: net migration rate) (1) (2) (3) 1999-2015 1999-2007 2008-2015 Net inflows (t - 1) 0.26 *** 0.21 0.16 (0.07) (0.15) (0.12) Relative unemployment -0.23 *** -0.41 -0.40 *** (t - 1) (0.08) (0.47) (0.15) Relative wage growth O.OO 0.15 -0.18 (t - 1) (0.09) (0.21) (0.12) Observations 197 90 94 Country fixed effects Yes Yes Yes Time fixed effects No No No [R.sup.2] 0.30 0.15 0.27 (4) (5) (6) 1999-2015 1999-2007 2008-2015 Net inflows (t - 1) 0 29 *** 0.20 0.16 (0.07) (0.14) (0.12) Relative unemployment -0.24 *** -0.39 -0.41 ** (t - 1) (0.09) (0.48) (0.16) Relative wage growth -0.05 0.28 -0.43 (t - 1) (0.15) (0.27) (0.27) Observations 197 90 94 Country fixed effects Yes Yes Yes Time fixed effects Yes Yes Yes [R.sup.2] 0.33 0.19 0.28 Panel data (14 Eurozone countries), LSDVC estimator with country and time fixed effects Standard errors in parentheses. * p < 0.1; ** p < 0.05; *** p < 0.01 Table 4 Mobility and GDP per capita growth (dependent variable: net migration rate) Group of countries of origin World 1999-2015 1999-2007 2008-2015 Net inflows (t - 1) 0 0.35 ** 0.18 (0.08) (0.15) (0.13) Relative -0.46 *** -0.60 ** -0.33 unemployment (t - 1) (0.10) (0.25) (0.23) Relative GDP per 0.30** 0.63** 0.26 capita growth (t - 1) (0.13) (0.32) (0.18) Observations 190 84 93 Country fixed effects Yes Yes Yes Time fixed effects Yes Yes Yes [R.sup.2] 0.29 0.09 0.36 EU 1999-2015 1999-2007 2008-2015 Net inflows (t - 1) 0.31 *** 0.23 * -0.01 (0.06) (0.13) (0.12) Relative -0.33 *** -0.84 * -0.40 * unemployment (t - 1) (0.11) (0.49) (0.22) Relative GDP per 0.05 -0.63 0.00 capita growth (t - 1) (0.18) (0.62) (0.18) Observations 183 82 89 Country fixed effects Yes Yes Yes Time fixed effects Yes Yes Yes [R.sup.2] 0.24 0.08 0.13 Eurozone 1999-2015 1999-2007 2008-2015 Net inflows (t - 1) 0.29 *** 0.17 0.21 * (0.07) (0.14) (0.13) Relative -0.22 ** -0.65 -0.31 ** unemployment (t - 1) (0.09) (0.47) (0.15) Relative GDP per 0.07 -0.52 0.19 capita growth (t - 1) (0.17) (0.43) (0.18) Observations 197 90 94 Country fixed effects Yes Yes Yes Time fixed effects Yes Yes Yes [R.sup.2] 0.34 0.16 0.35 Panel data (14 Eurozone countries), LSDV estimator with country and time fixed effects Standard errors in parentheses. * p < 0.1, ** p < 0.05, *** p < 0.01 Table 5 Mobility with the world and EU agreements (dependent variable: net migration rate) (1) (2) (3) 1999-2015 1999-2007 2008-2015 Net inflows (t - 1) 0.35 *** 0.29 ** 0.16 (0.08) (0.14) (0.13) Relative unemployment -0.51 *** -0.41 -0.43 * (t - 1) (0.10) (0.25) (0.23) Relative wage growth -0.05 0.44 ** -0.30 (t - 1) (0.13) (0.18) (0.24) Schengen -0.99 -0.35 (1.13) (2.27) FreeEU8 FreeEU2 Observations 190 84 93 Country fixed effects Yes Yes Yes Time fixed effects Yes Yes Yes [R.sup.2] 0.26 0.09 0.35 (4) (5) (6) 1999-2015 1999-2007 2008-2015 Net inflows (t - 1) 0.35 *** 0.29 ** 0.17 (0.08) (0.14) (0.13) Relative unemployment -0.50 *** -0.41 -0.34 (t - 1) (0.10) (0.26) (0.24) Relative wage growth -0.09 0.45 ** -0.29 (t - 1) (0.13) (0.18) (0.23) Schengen FreeEU8 0.78 -0.35 -0.22 (0.75) (1.05) (1.53) FreeEU2 0.67 0.47 2.47 ** (0.65) (1.51) (1.04) Observations 190 84 93 Country fixed effects Yes Yes Yes Time fixed effects Yes Yes Yes [R.sup.2] 0.28 0.08 0.32 Panel data (14 Eurozone countries), LSDVC estimator with country and time fixed effects Standard errors in parentheses. * p < 0.1; ** p < 0.05; *** p < 0.01. The variable "Schengen" is time-invariant over 2008-2015 and therefore discarded Table 6 Mobility with the EU and EU agreements (dependent variable: net migration rate) (1) (2) (3) 1999-2015 1999-2007 2008-2015 Net inflows (t - 1) 0.31 *** 0.27 ** -0.02 (0.06) (0.13) (0.12) Relative unemployment -0.36 *** -0.71 -0.48 ** (t - 1) (0.11) (0.54) (0.22) Relative wage growth -0.09 0.10 -0.37 * (t - 1) (0.14) (0.39) (0.20) Schengen - 1.30 0.60 (1.26) (3.85) FreeEU8 FreeEU2 Observations 183 82 89 Country fixed effects Yes Yes Yes Time fixed effects Yes Yes Yes [R.sup.2] 0.20 0.12 0.13 (4) (5) (6) 1999-2015 1999-2007 2008-2015 Net inflows (t - 1) 0.29 *** 0.26 * -0.04 (0.06) (0.13) (0.12) Relative unemployment -0.33 *** -0.76 -0.34 (t - 1) (0.11) (0.56) (0.24) Relative wage growth -0.18 0.09 -0.44 ** (t - 1) (0.15) (0.39) (0.20) Schengen FreeEU8 2.19 ** 1.34 2.10 (1.00) (1.92) (1.81) FreeEU2 0.42 -2.67 1.49 (1.19) (3.00) (1.00) Observations 183 82 89 Country fixed effects Yes Yes Yes Time fixed effects Yes Yes Yes [R.sup.2] 0.24 0.13 0.07 Panel data (14 Eurozone countries), LSDV estimator with country and time fixed effects Standard errors in parentheses. * p < 0.1; ** p < 0.05; *** p < 0.01. The variable "Schengen" is time-invariant over 2008-2015 and therefore discarded Table 7 Mobility in the Eurozone and EU agreements (dependent variable: net migration rate) (1) (2) (3) 1999-2015 1999-2007 2008-2015 Net inflows (t - 1) 0.29 *** 0.20 0.16 (0.07) (0.14) (0.12) Relative unemployment -0.25 *** -0.39 -0.41 ** (t - 1) (0.09) (0.49) (0.16) Relative wage growth -0.05 0.28 -0.43 (t - 1) (0.15) (0.27) (0.27) Schengen -1.18 0.89 (1.61) (3.97) FreeEU8 FreeEU2 Observations 197 90 94 Country fixed effects Yes Yes Yes Time fixed effects Yes Yes Yes [R.sup.2] 0.31 0.19 0.28 (4) (5) (6) 1999-2015 1999-2007 2008-2015 Net inflows (t - 1) 0.30 *** 0.21 0.20 (0.07) (0.15) (0.12) Relative unemployment -0.24 *** -0.46 -0.46 ** (t - 1) (0.09) (0.50) (0.18) Relative wage growth -0.05 0.31 -0.30 (t - 1) (0.15) (0.27) (0.26) Schengen FreeEU8 0.15 -0.91 -3.34 * (1.17) (1-84) (1.94) FreeEU2 0.79 -0.61 1.11 (1.06) (2.83) (1.24) Observations 197 90 94 Country fixed effects Yes Yes Yes Time fixed effects Yes Yes Yes [R.sup.2] 0.33 0.20 0.28 Panel data (14 Eurozone countries), LSDV estimator with country and time fixed effects Standard errors in parentheses. * p < 0.1; ** p < 0.05; *** p < 0.01. The variable "Schengen" is time-invariant over 2008-2015 and therefore discarded Table 8 Dates at which restrictions on the access to national labor markets were lifted EU8 EU2 Croatia Austria May 2011 January 2014 July 2018 Belgium May 2011 January 2014 July 2018 Germany May 2011 January 2014 July 2018 Greece May 2006 January 2009 July 2015 Spain May 2006 January 2009 * July 2015 Finland May 2006 No restrictions No restrictions France July 2008 January 2014 July 2015 Ireland No restrictions January 2012 No restrictions Italy August 2006 January 2012 July 2015 Luxembourg November 2007 January 2014 July 2015 Netherlands May 2007 January 2014 July 2018 Portugal May 2006 January 2009 No restrictions Slovenia No restrictions No restrictions July 2018 Slovakia No restrictions No restrictions No restrictions EU8: Czech Republic, Estonia, Hungary, Lithuania, Latvia, Poland, Slovenia, Slovakia EU2: Bulgaria and Romania * Restrictions were reinstated in Spain for Romanian citizens from 22.07.2011 until 31.12.2013 Table 9 Summary statistics Variable Mean SD Min Max N Net migration rate with 1.73 3.45 -10.1 16.6 207 the world (per 1000 people) Net migration rate with 1.30 4.69 -14.8 38.2 200 the EU (per 1000 people) Net migration rate in the 1.59 4.41 -16.1 20.1 211 Eurozone (per 1000 people) Unemployment rate (percent) 9.01 4.68 1.90 27.5 238 Real average wage (1000 EUR) 35.8 12.4 6.6 61.0 238 Real GDP per capita (1000 EUR) 30.5 15.4 7.7 84.4 238 Panel data: 14 Eurozone countries, 1999-2015
|Printer friendly Cite/link Email Feedback|
|Author:||Huart, Florence; Tchakpalla, Medede|
|Publication:||Comparative Economic Studies|
|Date:||Jun 1, 2019|
|Previous Article:||Introduction: 2017 INFER 19th Annual Conference, Bordeaux.|
|Next Article:||Monetary Union, Competitiveness and Raw Commodity Dependence: Insights from Africa.|