The use of natural experiments in migration research

Data on rapid, unexpected refugee flows can credibly identify the impact of migration on native workers’ labor market outcomes

Central Bank of the Republic of Turkey, and IZA, Germany

one-pager full article

Elevator pitch

Estimating the causal effect of immigration on the labor market outcomes of native workers has been a major concern in the literature. Because immigrants decide whether and where to migrate, immigrant populations generally consist of individuals with characteristics that differ from those of a randomly selected sample. One solution is to focus on events such as civil wars and natural catastrophes that generate rapid and unexpected flows of refugees into a country unrelated to their personal characteristics, location, and employment preferences. These “natural experiments” yield estimates that find small negative effects on native workers’ employment but not on wages.

Natural experiments in immigration: instances of
            sudden, rapid refugee inflows

Key findings


Refugee flows into a country are generally due to reasons unrelated to the immigrants’ location and employment preferences.

Refugee flows bring in a massive number of immigrants within a short period.

The location of refugees within the host country is generally determined by the host government based on security, logistic, and social concerns.

From the host country perspective, refugee flows are mostly unexpected events that can be considered immigration shocks.


Refugee inflows to a region can trigger an outflow of native workers from the refugee-receiving regions to other regions in the host country, creating new selectivity problems.

The skill composition of refugees may affect their impact on native workers’ labor market outcomes.

The impact of refugees may also depend on the existing stock of immigrants, which affects the absorption capacity of local labor markets.

In the longer term, the occupational distribution of the refugee population may be influenced by the relative returns across occupations in the host labor market.

Author's main message

Countries are concerned that immigration may cause the employment and wages of native workers to fall. To estimate such causal effects, the decision on whether and where to migrate has to be randomly assigned. A close substitute is to exploit natural experiments, such as sudden and rapid refugee flows. Estimates based on refugee flows find larger short-term impacts than long-term impacts, since local labor markets tend to adjust in the long term. Thus, policymakers should focus on targeted labor market policies and social programs that jointly facilitate the integration of refugee workers into local labor markets in the short term.

Full citation

Full citation

Data source(s)

Data type(s)