COGITO: Migration, Wages and (un)Employment

Labour economics classically predicts that wages are determined by supply and demand. Theoretically, if demand remains constant, emigration should increase wages in countries of origin (decreasing labour supply) and decrease wages in receiving countries (increasing labour supply). A malthusian vision of the economy also assumes that the arrival of new-comers in a market of limited jobs will leave some workers without employment or drive previously employed ones out of their jobs. The view that immigration decreases natives’ wages and generates unemployment are widely held. An alternative theory suggests that immigrants also consume in the destination country, hence increasing local demand, and/or that local production expands, now that additional labor is available. We need to try and empirically distinguish the competing theories from each other. We also need to examine if this is what is happening on empirical grounds at local levels?

Causality is Difficult to Establish

Some studies show that the wage curve corresponds to what is implied by the first type of theory: immigration will decrease wages by 3 to 4 % for a 10% increase in workers in a given skill-experience group.

But economists find it difficult to identify the impact of migration on wages or unemployment, both in countries of origins and in host countries. Why? Causal relations are hard to establish because migrants often move to receiving countries when labour market conditions are good and improving there. Thus, observing that wages do not decrease when migrants arrive may both mean that they are not impacted by migration, or that wage increase due to better labour market conditions compensate the impact. The same applies to unemployment: context matters and market dynamics strongly determine the impact of immigration or emigration.

Apart from problems to establish a causal relationship, migration effects are also specific to time and place, to economic conditions and dynamics in host and receiving countries. Thus results will vary according to the type of data (e.g. micro vs macro) used, the scale (city, national, inter-national), the theoretical model assumed, and the timeframe (which years, and short- vs long run) considered.

Empirical evidence in OECD countries has recently shown that the overall effect of immigration is either neutral or slightly positive on average natives’ wages and has no overall impact on natives’ employment. But these results need refining: when looking at differential impact along the remuneration distribution, negative effects for low skilled workers and positive effects for highly skilled workers are observed, and any declines in the wages and employment in the short run is compensated by rising wages and employment in the long run.

The Impact Differs According to Employment Levels

The UK example (1983-2000) shows that immigration had no statistically significant effect on the overall employment outcomes of UK-born workers.

It also shows that wages of poorer native workers decrease (the 20th percentile of the distribution) whereas wages of the richer native workers grow (above the 40th percentile) when immigrants arrive. This echoes an important dimension of immigration: the fact that migrants downgrade considerably upon arrival (both in wages and level of employment) and that their location in the wage distribution (crowding the bottom) is generally below their actual skill level. Empirical evidence shows that this initial mismatch explains the wage impact. Similar studies carried out in the UK after the 2008 crisis confirm the lack of any impact of migration on natives’ unemployment in aggregate during the 2008 economic crisis.

Further empirical case studies support the idea that fully opening the border to immigration from neighbouring countries increased immigrants to Switzerland only by 4% of the labour force over 8 years. Such an increased inflow did not have significant aggregate effects but generated differentiated effects across the wage ladder: highly educated workers benefited through higher wages, while middle-educated ones experienced employment losses.

A US based study showed that in the period from 1990 to 2006 immigration had a small effect on the wages of native workers with no high school degree (between 0.6% and +1.7%). It also had a small positive effect on average native wages (+0.6%) and a substantial negative effect (−6.7%) on wages of previous immigrants in the long run.

Studies Have Contradictory Results

Recent controversies have questioned the existing consensus that the impact of immigration on average native-born workers is small. Classical historical studies using city-based data had used mass sudden inflows of refugees to establish that immigration does not decrease wages (Card, 1990); but may slightly increase unemployment. A study confirmed the distributional effect of refuge inflows on wages (negative on low wages and positive on high wages).

But other influential contributions have found a significant negative effect of immigration on the wages of low skilled natives in the U.S. Another study across OECD countries suggested the contrary: no impact of immigration on the wages of low skilled workers since the 1990s but negative impact of emigration from OECD countries. Significances of wage drops or increases were contested in both cases.

What Needs to be Considered

Understanding the mechanisms that are beneficial for labour markets, for natives and migrants, understanding how much and more crucially for how long labour markets are affected in the host countries, requires more than the simpler question “does immigration decrease natives’ wages”. It takes further insight into labour market mechanisms such as:

• Complementarity between natives vs immigrants. As demonstrated by empirical evidence in California from the 1960s onwards, because immigrants are imperfect substitutes for natives with similar education and age, they tend to stimulate, rather than harm, the demand and wages of most US native workers. 
• Internal mobility (geographical and professional) of workers both migrants and natives adapting to labour demands (within countries and at a regional scale for the Schengen space). Findings in the United States which could be transferred to the EU Schengen zone or ECOWAS in West Africa suggests that except for a reduced number of native workers, most wage effects resulting from more migration are probably dissipated quite fast. 

To sum it up, the evidence from refugee waves shows negative short-term effects on some native workers’ wages and employment in some times and places but no effect in others, and does not support claims of large overall negative effect even on low skilled workers.

A fully referenced article was first published in Sciences Po's Cogito Research Magazine, Migration, Diversity, and Mobility dossier.

Florian OSWALD is an Assistant Professor of Economics at the Department since 2015.

His research lies primarily in the fields of Housing, Urban, Macro and Labour Economics. He is also interested in the computational techniques required for structural estimation of microeconometric models.

More about Florian OSWALD and his research

Hélène THIOLLET is a CNRS researcher at the Center for Internanational Studies (CERI).

She examines migration policies in developing countries. She is particularly interested in the Middle East and sub-Saharan Africa.

More about Hélène THIOLLET and her research