Are migrants paid more or less than natives for doing the same or similar work and, if they are paid differently, can we be sure that it is due to their migrant status rather than to other differences between migrants and natives, such as their productivity levels? This is a long-standing question in labour economics, and one which remains largely unresolved due to the inadequacies of existing data. The standard or traditional story is that migrant workers are paid less than their native counterparts, but that the gap closes over time due to assimilation into the host country. This assimilation often entails learning skills that are particular to the host country, such as language skills, or overcoming initial barriers to practicing the occupation for which they were trained in their home country (Dustmann and Fabri 2003, Friedberg 2000). Nevertheless, a wage penalty usually persists and this is often attributed to labour market discrimination due to employer preferences.
But economic theory does not necessarily predict that all migrants will receive lower pay than natives. In fact, in a global market for highly talented labour, employers will seek to recruit the top talent, whether it is native to the country or not. One may expect highly talented workers to move only in circumstances where they receive very good offers which make it worthwhile leaving their home country. Thus, it seems likely that migrants will be drawn from the top of the ability spectrum, commanding a premium over natives.
Why is this a difficult question to answer?
There are two very big drawbacks to most of the existing studies that make it very hard to know whether migrants really experience a wage penalty or a wage premium.
- The first is that they compare workers across occupations, making it difficult to establish whether any wage differences reflect migrant status or are, in fact, due to occupational or skill differences.
- The second major drawback is that they lack individual-level labour productivity, so they are unable to quantify the contribution to the wage gap of labour productivity differentials between natives and migrants.
In a new study, we overcome problems with earlier studies by analysing wage differentials between professional football players in Italy (see Bryson et al 2012). Because migrant and native footballers are clearly substitutes for one another, and because we have individual-level information on what players do in each game, we are able to isolate the effects of migrant status on wages.
A wage premium for migrants
Using data tracking players in Italy's Series A and B for the period 2000-8, we find a substantial wage premium for migrant players relative to their native (Italian) counterparts. The raw gap – without controlling for differences between players – is a huge 62%. But this falls dramatically when accounting for differences in demographic characteristics, position on the field and, crucially, on-field labour productivity (using 20 measures, the richest set of measures used in studies of footballer performance). A migrant wage premium of between a quarter and a third is still apparent even among players in the same team and is only partially accounted for by individual labour productivity.
Two obvious questions arise.
- First, why do migrants receive a premium?
- Second, do Italian teams benefit from paying these higher wages to foreign players?
We explore two potential reasons for the migrant wage premium. The first is that those who migrate to Italy are superstar football players. ‘Superstars’ has two specific meanings in economics discussed in two seminal papers (Rosen 1981; Adler 1985). Sherwin Rosen thinks of superstars as workers who are more productive than their counterparts, while Michael Adler thinks of them as popular among consumers, such that they attract bigger audiences. In both cases, they tend to exist in settings where an individual's productivity or popularity results in very big revenue gains for the employer, as occurs in televised sports or music concerts where the individual's productivity is ‘scaled up’ by the technological setting in which the worker performs. The second potential reason for a migrant wage premium is the possibility that native workers are taking a wage hit for remaining in their home country.
We show that the differential partly reflects the superstar status of migrant workers, but it also reflects domestic workers' preferences for working in their home region. Among the Italians, those playing close to their place of birth suffer a particularly large wage penalty. One can think of being ‘local’ as an amenity for which they are prepared to take a compensating wage differential. Alternatively, one might say these players are forced to accept this wage in the face of employer monopsony power, something that does not affect the migrant workers.
Do football clubs benefit from this migrant talent? The answer is – yes. We estimate players' contributions to team wins and fan attendance at games which, in our data, are the key determinants of team revenues. We do so by looking at changes in points won and crowd attendance within each team season-on-season and comparing it to changes in the share of migrants in the team. We find that a 1% increase in the standard deviation of migrant share increases crowd attendance at games by just under 1%. The effect on the number of points a team earns is of a similar magnitude. These increases in team points and crowd attendance with the rise in the percentage of migrants in the team are strong indicators of migrants' superstar status.
Adler, M (1985), “Stardom and talent”, American Economic Review 75, 208-212
Bryson, A, R Simmons, and G Rossi (2012), Why Are Migrants Paid More?, NIESR Discussion Paper No. 388 and CEP Discussion Paper No. 1134
Dustmann, C and F Fabbri (2003), “Language proficiency and the labour market performance of immigrants in the UK”, Economic Journal 113, 324-F341.
Friedberg, R (2000), “You can’t take it with you: Immigrant assimilation and the portability of human capital”, Journal of Labor Economics 18, 221-225.
Rosen, S (1981), "The economics of superstars", American Economic Review 71, 845-858.