The bulk of European Union (EU) population growth now comes from immigration – and in the
face of increasing competition from immigrant workers and other globalising forces, antiforeign
sentiment and the demand for protection has grown. But new research by Professors
Joshua Angrist and Adriana Kugler, published in the June 2003 Economic Journal, suggests
that attempts to protect native jobs with tough labour market restrictions and limits on new
businesses are likely to fail. Indeed, when it comes to employment, these institutional rigidities
may actually be counter-productive.
The researchers note that two of the most controversial economic policy questions in Europe
are whether immigrant workers take jobs from European natives and the employment
consequences of stringent government regulation of labour and product markets. Examples of
such regulations include high minimum wages and payroll taxes, high firing costs and costly
bureaucratic delays for entrepreneurs. The rationale for many of these regulations is to
protect European workers from an American-style free-for-all (or so it is said), in which the
lowest paid workers win.
The rapid increase in immigration to EU Europe in the 1990s provides a unique opportunity to
test the thesis that inflexible markets protect native workers from immigrant competition.
Previous attempts to estimate the effects of immigration on native employment have often
been confounded by the fact that immigrants tend to go where the jobs are. This may mask
any job losses due to immigration. For example, Angrist and Kugler find that migration by EU
natives from one EU country to another is positively correlated with employment levels in the
host country.
Their study overcomes this difficulty by using the Balkan wars in the early and late 1990s as a
natural experiment that sharply shifted non-EU inflows to EU Europe. During the Bosnia and
Kosovo wars, EU countries relatively close to the former Yugoslavia were flooded with
immigrant workers, while countries farther away were mostly unaffected. Since immigrants
migrated mostly for political reasons and people naturally tend to migrate to closer places, this
helps to control for the fact that immigrants can usually be found in high-growth regions.
Over the period 1995-9, roughly one third of male immigrants aged 20-59 from non-EU
countries were from the former Yugoslavia. Yugoslavs accounted for about 30% of asylum
seekers in EU countries during the peak war years (1992 and 1998-9), and the estimates
suggest that the wartime upturn in immigration was 30-40% larger for every 500 miles closer
a country was to the conflict. Compare, for example, Graz in Austria with Liege in Belgium.
As a check that the differential impact of the war years by distance was indeed due to the
wars, the authors show that there is no corresponding pattern in within-EU migration. While
outsiders often see EU countries as less than immigrant-friendly, Austria absorbed 100,000
Bosnian refugees in 1992-5, most of whom were granted work permits. Similarly, Italy
experienced a tripling of foreign employment in 1990-7.
Results that correlate immigration rates with native employment show mostly modest negative
effects that are often insignificant. Estimates that exploit the Balkan wars as a natural
experiment to estimate immigrant effects are larger (more negative), though again not always
significant. A more intriguing picture emerges when the impact of immigrants on native
employment is allowed to vary according to the restrictiveness of local labour and product
market institutions.
These estimates suggest that as restrictions on labour and product market flexibility increase,
immigrants are more likely to take native jobs and not less. For example, an increase in the
severity of labour standards – a measure that captures minimum wages and employment
protection – by the difference in standards between Denmark and Belgium is predicted to
increase immigration-induced job losses by about 50%.
So why do protective institutions apparently fail to protect native workers? The authors
suggest a number of possible explanations:
First, inflexibility may decrease the natural level of employment – and lower employment
levels make the impact of a given number of immigrants more substantial.
Second, minimum wages or inflexible union wage scales make immigrants an especially
attractive alternative to natives when the opportunity to employ immigrants arises.
Finally, on the product market side, the ability of entrepreneurs to open new businesses that
exploit cheap immigrant labour would ordinarily act to undo any negative effects of
immigration on employment. Ultimately, immigration might even stimulate employment. But if,
as in France, business entry is a long and costly process, this mechanism is disabled.
ENDS
Notes for Editors: 'Protective or Counter-Productive? Labour Market Institutions and the Effect
of Immigration on EU Natives’ by Joshua Angrist and Adriana Kugler is published in the June
2003 issue of the Economic Journal.
Angrist is in the Department of Economics at the Massachusetts Institute of Technology
(MIT); Kugler is at the Universitat Pompeu Fabra in Barcelona.
For Further Information: contact Josh Angrist on +1-617-253-8909 (fax +617-253-1330; email:
angrist@mit.edu); or RES Media Consultant Romesh Vaitilingam on 0117-983-9770 or
07768-661095 (email: romesh@compuserve.com).