What does economic research suggest for immigration policy?

In recent years, immigration has become one of the most prominent political speech topics in developed nations. While the immigration debate is wider than its economics implications (or other objective facts), it is important that we turn to data to clarify some of the misperceptions about immigration in order to better understand political stances and policy debates.

The international immigrants’ stock has grown since at least 1995, increasing from 174 million in 1995 to 272 million in 2019, which represents around 3.5% of the total world population, according to the International Organization for Migration.

While this group represents a small part of the world’s population, they can make up for a significant part of high-income countries’ population. For instance, in Switzerland, foreign-born represent 29.7%, while in the United States (U.S.) and Germany, the top 2 destination countries, they represent 13.6% and 16.1%, respectively.

It is in this very objective data that we can find the first misperception about immigration, according to Nobel laureates Abhijit Banerjee and Esther Duflo, surveys of natives revealed that respondents overestimate not only the share of immigrants in their population (estimated 26% in Italy while the actual number was closer to 10%), but also have a wrong perception of their origin and skills. It is not difficult to consider that just these errors can significantly shape where people stand on either side of the debate.

Another misunderstanding can be found in one of the main arguments related to labor market. Among immigration opponents it is believed that immigration has detrimental effects for the natives in the labor market; namely, it generates unemployment and downward pressure for wages due to the increased competition from the new workers.

The former reasoning suggests an unequivocal conclusion regarding immigration’s effects; economic research suggests a much more nuanced one, after considering, among others, the differences in workers’ skills and the labor market rules that are set in place.

Within a neoclassical framework, immigration’s most immediate effect is an increase in the labor supply, and such a change in supply would decrease the economy’s average wage, ceteris paribus. Nevertheless, this is a short-term result, and as labor demand adapts, theory predicts that the negative impact of the average wage is mitigated in the long run. Of course, such results depend, among others, on whether wages are flexible or not, and how fast firms can change their production or output mix.

While theory predicts that there will be no harm on average, this does not preclude the possibility of certain groups being negatively affected by the immigration flows. This arises from the fact that workers differ in specific characteristics, such as skill level, and consequently belong to different labor markets within an economy. Thus, some natives might face more competition than others, depending on the incoming workers’ profile.

Theory predicts that natives with similar skills to those of the immigrants will face a reduction in their wages due to the increased competition. In contrast, natives with complementary skills will raise, as their productivity increases from collaborating with the new workers (NAS,2017).

It is important to highlight that immigrants might not be perfect substitutes for natives. Even with similar education levels, immigrants and natives could differ in their language and communication skills (Edo,2019). This has two important implications: (1) immigrants are closer substitutes to previous immigrants than natives, thus any detrimental effects should be smallest for the latter (Edo, 2019), and (2) natives who compete with immigrants should specialize in tasks where they use their competitive advantage (NAS, 2017).

But what does data tell us about immigration?

The empirical studies of immigration rely on counterfactuals, that is what would have happened in the absence of immigration. This is done through comparing the outcomes of areas (or groups of workers) with similar characteristics, except for the fact their exposure to immigration.

One of the most influential studies is the one by Card (1990) about the exodus of Mariel, an unexpected immigration of 125’000 Cubans to Miami during 1980 that increased the city’s labor force by 7% (Banerjee & Duflo, 2019).

Card compared Miami residents’ salaries before and after the exodus with the outcomes observed for other similar cities and estimated the effect of this influx. His results showed no significant differences between the cities, meaning that the wages and employment level of native workers were not affected by the new residents (Edo, 2019) either immediately or a few years after their entry. Another important finding was that it highlighted that prior Cuban immigrants were more likely to be negatively affected due to their similarity with the recent entrants (Banerjee & Duflo, 2019).

Besides from the Mariel’s exodus, other empirical studies of immigration that rely on different types of methodologies and counterfactuals, such as comparably skilled groups or other unexpected influxes, have shaped the consensus that on average there is little effect of native wages and employment, but there can be winners and losers among natives (Blau & Hunt, 2019; Edu, 2019).

The effects on subgroups (and even on average) vary for each region due, among others, to labor market rules and differences in skill composition of immigrants and native population. For example, in the U.S., the expert consensus is that the most likely to be negatively affected are prior migrants, as frequently they are the closest substitutes for the new migrants, followed by high-school dropouts, as they have similar qualifications to the low-skilled workers, a group that represents a large share of the U.S. immigrants (Blau & Hunt, 2019).

Nevertheless, in other countries, such as Switzerland, no negative effect has been found for this group. Favre (2011) suggests this could be due to market rigidities, such as the minimum wages included in the collective labor agreements adopted to prevent wage decreases in this group after the agreement of free movement of persons was enacted.

When there is complementarity between native’s and immigrant’s skills, positive effects can be found. For instance, Mitaritonna et al. (2017) found a positive effect on the average wage of natives in a French department. The two channels that potentially explained the results were: (1) that immigrants were more skilled than the native population, so they could contribute to innovation and (2) the specialization in different and complementary tasks could have improved efficiency and labor productivity (Edo, 2019).

What do these findings suggest for policy making?

It is important to remember that immigration has broader social and economic effects, one of them is that it can help to offset negative trends. Immigrants are mostly of working age, which implies that in age declining societies, immigration can counteract this tendency (Orrenius et al., 2020), promoting economic growth, as they replenish one of the production factors: the labor force.

For instance, in the U.S., first-generation migrants have contributed more than natives to the increase of the labor force in the period 2005 to 2015. Also, projections for the next decades show that first- and second-generation immigrants will be the ones that offset the decline in U.S. born workers with U.S. born parents (Orrenius et al., 2020).

Additionally, high-skilled workers’ immigration can have wider positive externalities. Data from the U.S. shows that high-skilled immigrants, who are concentrated in STEM (science, technology, engineering, and math) occupations, contribute positively to patenting and innovation (Edo, 2019).

As shown, there are at least two channels through which immigration can positively contribute to economic growth. Trying to reconcile these broader implications with the previous findings on labor markets, it seems that the key question for policy making is whether there is complementarity rather than substitutability between the natives’ and immigrants’ skills. With higher complementarity, the potential costs for some natives are mitigated and the society can enjoy larger benefits from immigration.

Therefore, policies that are only oriented to a reduction in the size of the immigration influx without any consideration of the skill-composition of immigrants (“one-size-fits-all” solution, such as completely limiting free movement), would most likely yield sub-optimal results. While this kind of measures could be beneficial for some workers, they will likely have broader (negative) effects on economic growth.

Moreover, these policies could be completely ineffective if firms change their production techniques in response to the labor force’s shrinkage, as shown by Clemens et al. (2018), who studied the decision to exclude half a million Mexican bracero seasonal workers from the U.S. While the policy change objective was to improve the native’s wage and employment outcomes, the labor force’s shrinkage did not have the expected results, as firms adapted their production technique by substituting labor for capital.

In conclusion, the evidence presented supports skill-selective policies, that is looking for complementarity between citizens and new workers by addressing labor and skills shortages. Moreover, besides regulations regarding entry, the policy mix should include “mitigating policies” where needed. While market rigidities have proven to protect low-skilled workers from potential adverse effects, measures without a broader impact on the economy could also be successful. For instance, according to Somerville & Sumption (2008), training programs for the native workers could help them make the best use of their competitive advantages (e.g., native language), thus reducing the degree of substitutability with the new immigrants.

firstsecondthird