In most countries, international migration is the most debated facet of human mobility. The number of asylum applications lodged in 2015 in EU Member States exceeded 1.3 million. This placed migration policy in the forefront of the global policy debate, as is often the case after each immigration peak. However, the concern over migration policy can be seen as the by-product of a long-lasting process of increasing immigration that has been occurring for the last 50 years or so. In the US and in the EU15, the stock of legal international immigrants has been multiplied by 4 since 1960. In the EU15, the stock of immigrants originating from developing countries has been multiplied by 8.7 since 1960, and by 2.3 since 1990 (Ozden et al. 2011).
Qualitatively, the underlying global root drivers of these trends are known – demographic imbalances, economic inequality, increased globalization, political instability, climate change, etc. – and the role of migrant networks in generating dynamic multiplier effects is well understood (Beine et al. 2011). Digital transformation in developed economies has also increased the demand for highly skilled workers. In some areas, such as Luxembourg, the local supply of such workers is insufficient. Only a few studies have empirically investigated the long-run causes of migration in a global setting (see Hatton and Williamson 1994, 2003, 2005; Dao et al. 2017). Hence, little is known about the relative contribution of each root cause of migration in explaining the long-run trends in the size and structure of migration. In addition, International migration laws/policies have been rarely accounted for in empirical studies, due to the difficulty to classify and code them.
We plan to develop new approaches to explain long-term trends in international migration and anticipate future pressures. We will extend existing quantitative methodologies (e.g., Lutz et al. 2014) to dissect the inter-relationships between migration, population growth, education decisions and economic development. To shed light on the effectiveness, legal consistency and coherence of policy reforms, ACROSS could make use of data on immigration policy in the context of the IMPALA project conducted a few years ago in Luxembourg (Beine and Souy 2016). We propose to carry out empirical analysis of the long-term determinants of migration. For example, concerning highly skilled workers, the role of push and pull factors can be addressed by using original data on job vacancies in the EU on the one hand, and microdata on EU Blue Card candidates (i.e., potential migrants) on the other hand. Given the complex interdependencies between migration, population and income, we will also develop multi-country, micro-founded general equilibrium models with overlapping generations of individuals to understand the forces at work (see Delogu et al. 2018; Docquier and Machado 2018; Burzynski et al. 2019a,b).
Secondly, existing statistics emphasize the nonpermanent nature of many migration decisions. In a recent report, the OECD (2008) estimates that 20 to 50% of immigrants leave the host country within the first five years after arrival, depending on the countries and time periods considered. Dustmann and Görlach (2016) estimate that immigrant out-migration rates are substantial and larger from European destination countries than from the US, Canada or Australia. This highlights the rising temporary nature of European human mobility, which raises new challenges in terms of integration policies, social cohesion, taxation, housing, etc. We plan to shed light on the determinants of migration temporariness and its economic and social implications (e.g., Delpierre and Verheyden 2014a). This requires conducting new empirical studies focusing on the determinants of temporariness, and developing new models that jointly endogenize migrants’ duration of stay and the consequences for the host countries.
In many countries, internal/interregional migration is an even more sizeable phenomenon. Bell et al. (2015) find that internal migration intensity is 60 times more important than international migration worldwide. Official cross-country databases on infra-national mobility are imprecise, and are difficult to reconcile with coarse international flow data. With a few exceptions (Beine and Coulombe 2018, Burzynski et al. 2019a, 2019b), little is known about the interplay between internal and international migrations. We plan to go beyond the state of the art in combining traditional and new sources of data. For instance, researchers have relied on Big Data using cell phone data – see for instance Beine et al. (2019) in the case of refugees in Turkey. Other approaches rely on worldwide opinion surveys on migration intentions and geo-referenced data on population changes. Gridded data sources are available and allow researchers to better connect spatial shocks and mobility decisions (Rango and Laczko, 2017; Blumenstock et al. 2015). These new data sources can potentially revolutionize the understanding of the drivers of migration flows. They also raise important challenges in terms of storage, analysis, estimation, visualization, and information privacy; this requires developing new methodologies for processing and analyzing them (e.g., data mining, machine learning algorithms). We ambition that the exploratory analyses of Big Data conducted in this DTU will serve as a foundation for a larger data infrastructure project focusing on the measurement of human mobility in virtually continuous time and space.
Linked to internal movements, daily cross-border commuter flows have a strong incidence on the economy of receiving and sending countries and regions. In general, immigration and cross-border commuting increase the process of urbanization in the receiving country (Bertinelli and Black, 2004). This is particularly the case in Luxembourg or in the Alpine region (e.g., Geneva, Basel, Salzburg, Monaco, Trieste, Milan). In 2018, 191,998 of Luxembourg employees were commuting from their country of residence, which ranks Luxembourg as the third country in Europe hiring most cross-border workers living in a different country in absolute numbers, behind Germany and Switzerland (EC, 2019)! The importance of these flows raises the question of their long-run sustainability (Picard and Worrall 2018). Again, little is known about the interplay between commuter flows and residential mobility. We propose to develop new location choice models that jointly endogenize decisions about place of work and place of residence, especially in case where there is important wage and job-opportunity differentials on both sides of the border. Analytical and simulation models can help understanding the effect of non-integrated land use or transport policies on the spatial segmentation of land/housing markets, over/sub densification at borders and “elastic migrants” behavior (moving across borders because of housing markets, but keeping all daily activities on the original side). On the applied side, we plan to explore the possibility to construct two-dimensional nested logit models for work and residence, and estimate (or calibrate) them using data on commuters and migrants.