Foreign workers are a defining component of the Luxembourg economy and labour market. By 2006, seven in ten private sector workers in Luxembourg were foreign — of which three immigrant residents and four cross-border workers. Luxembourg is the EU country relying most on foreign labour to fuel its domestic economic activity.
With the support of the Fonds National de la Recherche in its CORE programme, this project set out to examine differences in pay between Luxembourg nationals and foreign workers and analyzed the implications thereof for earnings inequality in the country. InWIn — a project which started in 2011 — has now delivered its results.
State-of-the art methodologies and a reference point on inequality and wage distribution differences between native and foreign workers in Luxembourg
This project’s ambition was to exploit large-scale data sources on wage and earnings to undertake a thorough examination of earnings inequality and wage differences between Luxembourg nationals, immigrant residents, and cross-border workers in Luxembourg. Analyses were meant to pay particular attention to detailed wage structure differences (beyond standard examination of differences ‘at the mean’) and to overall wage inequality impacts. Researchers involved aimed to apply and develop state-of-the-art methodologies, thereby offering a new empirical reference point for these issues.
Foreign workers’ wages are lower than natives’… in the middle of the wage distribution
Results confirm earlier evidence that foreign workers are paid less than native-born employees. The native worker advantage (defined as the extra wage obtained by natives compared to foreign workers with similar characteristics and/or jobs) is low or inexistent among for the lowest paid workers (near the minimum wage) and among workers in the top of the wage structure, but it is positive and large for the middle half of native workers. The pattern holds against both immigrants and cross-border workers, although the latter do not catch up so much to highest wages.
Foreign workers do not inflate wage inequality, quite the contrary (at least in the short run)!
Foreign workers are sometimes perceived as contributing to wage inequality. InWIn results contradict this view and find no indication that immigrants inflate indicators of inequality (at least in the short term). The only significant exception is for non-EU immigrants — not more than 10 percent of immigrants — that appear to drive wage dispersion upwards by their concentration at the top of the wage distribution. But cross-border workers and all other immigrant groups tend to affect inequality downwards, if at all.
Earnings inequality increased surprizingly little since 1988
The project also exploited a large-scale administrative dataset to document trends in earnings inequality between 1988 and 2009, twenty years of rapid economic growth. Perhaps surprisingly given the major evolution of the economy and employment composition in this period, only a modest increase in overall earnings inequality between 1988 and 2009 was observed. This apparent stability is however the net result of somewhat more complex underlying changes, with marked increases in ‘persistent’ inequality among cross-border and immigrant workers and between foreign and native workers, but a decrease in ‘transitory’ inequality among native workers. Such results possibly hint at the role of strict labour market regulations and collective bargaining institutions in holding back earnings inequality, at least in a period of fast economic growth and soaring demand for labour.
Exploiting large-scale data on resident and non-resident workers
The project exploited two rich data sources which, unlike many surveys conducted in Luxembourg, provide information about both residents and non-resident workers: the 2002 and 2006 Luxembourg Structure of Earnings Surveys collected by STATEC and a large-scale extraction of anonymized administrative records from the social security administration covering 1998 to 2009. Analyses undertaken in the project are prime illustrations of the benefits of access to both large-scale administrative registers and detailed surveys for research on wages and inequality.
The development of statistical methods as research objective
On a more technical note, the project also had as central objective the development and refinement of a number of statistical, quantitative models for analysis of wage differentials, earnings dynamics and inequality on the basis of large-scale micro-data.
Fertile ground for future research
We trust the project lived up to its ambitions to provide an empirical reference point and methodological innovations. The subject matter remains fertile ground for research however! First of all, the availability of new, more up-to-date data should allow updating empirical evidence to more recent years, a period of slower growth and higher unemployment (the project’s research covers data up to 2009). Also many of the factors determining the differences in pay documented in the project remain unknown. The wage advantage in favour of native workers need not reflect mere ‘discrimination’. It is possible to ascribe the different sorting of foreign workers across different occupations and the unexplained difference in wage levels to other factors such as unobserved productive characteristics (e.g., multiple language proficiency), workers’ preferences, asymmetric wage bargaining power — especially at entry in Luxembourg. The extent to which these factors account for the wage gap remains however largely unknown.
Finally, the project has not touched upon the long-term impacts of the presence of foreign workers on employment and earnings of natives — a topic that remains largely contended among experts — and which requires predicting credibly what would have happened had there been no foreign workers in Luxembourg, an ambitious exercise which was beyond the scope of the present project but a challenge that ought to be taken up in the future!