The Economics of Immigration and Roxham Road, Part 3: Empirical Evidence on Immigration
Both proponents and opponents of immigration tend to greatly exaggerate their cases
In the most recent two issues of this newsletter, I covered the basic economic theory of immigration, followed by a brief history of immigration in Canada. This week, I want to write on the empirical evidence of immigration to get an idea of whether the fears and praises of immigration are justified.
For this post, my primary source is a paper from the Canadian Journal of Economics by Edo et al. (2020), which is from a special issue devoted entirely to immigration. This paper in particular is a very thorough and well-researched literature review which I highly recommend you read. The authors conclude that overall, economic studies find the impact of immigration on average wages and employment of native-born workers is either zero or slightly positive in the medium to long term. But in the short run, the impact of unexpected migration is negative.
These results are especially important for countries where immigration comprises a significant proportion of their labour markets. For example, Edo et al. (2020) note that countries like Germany and the U.S. have immigrant populations making up over 15% of their labour forces, while it is around 25% for Canada, Australia, and New Zealand.
Before moving on, I invite you to consider a paid subscription to this newsletter so I can more easily afford the time and resources to conduct research for it. While I do very much enjoy writing for its own sake, this is also a source of income for me, so any contributions you can provide will be much appreciated.
Wages and Employment Levels
Theoretical Models
In the first part of this series when I covered the basic theory, I assumed for simplicity that all workers are the same, so there is one set of labour demand and supply curves for each country. However, more realistic theoretical models divide workers by skill level, an implication of which is the effects of immigration on workers in the host country depend on their skill levels relative to the skill levels of the new immigrants.
More specifically, Edo et al. (2020) report that immigration will reduce the wages of existing workers who compete directly with the new immigrants, which is consistent with the basic theory. However, wages will increase for native-born workers whose skills complement those of the new migrants, since they can effectively work together and thus increase their productivity. In short, immigration will have a persistent effect on the structure of wages across skill groups.
However, these negative wage effects might not be so significant if the immigrants have different language skills, as well as different quantitative and relational skills. In other words, different skills can enable new immigrants and native-born workers to still work together for the benefit of everyone. For example, an applied study cited by the authors shows that for the U.S., lower-skilled immigrants specialize in manual-intensive jobs for which they have comparative advantages, while native-born workers of similar educations pursue jobs more intensive in communication tasks. Therefore, immigration tends to push native-born workers of comparable education into more cognitively oriented and communication-intensive jobs that are relatively better paid and more suited for their skills.
Simulations and Natural Experiments
Edo et al. (2020) cite simulation studies which have two key findings: first, the average effect of immigration on the wages of native-born workers is zero or slightly positive, in the long run. This result depends on the degree of substitution between the two types of workers, such that if they are imperfect substitutes, then there is a slightly positive effect of immigration on wages.
Second, the skill composition matters with respect to the long-run impact of wages on domestic workers, because there will be differing increases in labour supply for different groups of workers. This means some native born workers will be winners from immigration, while others will lose. For countries like Canada, Australia, France, Germany, and Switzerland, where immigration since the 1990s was disproportionately high-skilled, immigration has therefore led to reduced wage inequality between high-skilled and low-skilled workers. On the other hand, since immigration to the U.S. has tended to involve more low-skilled workers, wage inequality has worsened there. It is therefore not surprising that lower-skilled workers in the U.S. have worried the most about the security of their jobs.
There have also been natural experiments, where large influxes of immigrants have occurred quickly and (sometimes) unexpectedly, so researchers can observe labour markets in the host country both before and after immigration occurred. One such study was by Card (1990), who focused on the Mariel boatlift that occured in 1980 when Cuban leader Fidel Castro allowed Cubans to emigrate from the Mariel port. More than 100,000 of them went to Miami due to its proximity to Cuba, increasing Miami’s labour force by 7% quickly. Most of these immigrants were low-skilled, with 60% of them lacking high school diplomas. Nonetheless, Card (1990) found the average wage and employment levels of non-Cubans were unaffected by the immigration.
However, Borjas (2017) reappraised the Mariel evidence and found the migration did reduce the wages of non-Hispanic male high school dropouts, which is not surprising since we expect negative wage and employment effects on native-born workers with similar education levels.
Economic Integration of New Immigrants
The next concern is how well new immigrants intergrate economically in their host countries. According to Edo et al. (2020), it does not start well as they initially earn lower wages than comparable native-born workers. However, the wage gap does fall over the years after migration as their wages rise more rapidly over time compared to those of domestic workers — known as “economic assimilation”. One reason for the lower entry earnings could be due to the increase in labour supply upon entry to the country, but it could also be related to labour demand, for example due to racial and ethnic discrimination.
Immigrant Characteristics Driving Successful Integration
Language skills also appear to be essential for positive assimilation of immigrants. For example, the authors cite Canadian evidence that immigrants who lack language skills for their host country achieve little-to-no gains from their occupational or educational skills. Therefore, language skills to some degree determine/facilitate the productivity of their other skills. Indeed, according to one study cited by the authors, just attending a language course full time for six months leads, on average to an 11.3% wage increase for the new immigrant. To be clear, the result is not so much due to the actual language skills learned, but the soft skills and information acquired during the training sessions.
Then there is the age of the immigrant upon arrival to their host country: young children are much more likely to assimilate successfully than those who arrive after their mid-30s. Furthermore, there is evidence that people who are admitted for economic reasons have much superior labour market outcomes than those who are admitted for family unification reasons, or refugees/humanitarian migrants/asylum seekers. Perhaps surprisingly, there is not much difference in success between the latter two groups.
Government Policies and Operations
Edo et al. (2020) cite evidence that Canada, Australia, and the U.S. have all made improvements with policy reforms, and their screening policies are increasingly similar, so we might expect integration of immigrants to be rather similar across these countries. On the contrary, it seems there is a quantitatively important U.S. performance advantage to positive immigrant self-selection, meaning they go where they think they have the best opportunities.
In that respect, one study cited by the authors focuses on Canada where high-skilled immigrants initially arrive to the country on the basis of the points system, and then quickly move to the U.S. It is found these immigrants are more highly educated than people from the same country who go directly to the U.S. rather than making a detour through Canada; they also have better labour outcomes in the U.S. than those who go directly there. This two-step process affects a high proportion of young highly-skilled immigrants to Canada.
As for refugee policy, one cited study from France shows longer durations in “limbo” while they await the adjudication of their asylum claims, are associated with reduced socioeconomic integration even though they are ultimately deemed to be refugees. But a caveat is this result applies chiefly to those with low levels of education; ones with university degrees do not appear to experience these negative effects.
Then there is a Finnish study where there appear to be substantial efficiency gains in integration when the government targets new immigrants with active labour market programs, rather than lumping them into programs for the established population.
Finally, with respect to non-labour market integration in the U.K. and Canada specifically, Edo et al. (2020) cite evidence that life satisfaction scores for immigrants from a large and diverse set of source countries are seen to mimic those of their hosts, rather than the people in their home countries.
Fiscal Impact of Immigration
Another concern of many people is immigrants will choose their new homes based on how much they can take from the system while giving as little as possible back to it. This creates an adverse selection mechanism whereby net beneficiaries of the system are attracted to it like a magnet, while net contributors to it are repelled — thus the name “magnet hypothesis”.
Tests of this hypothesis find different results given the country examined, reflecting differences in social welfare systems. However, Edo et al. (2020) cite a Canadian study where immigrants are found to use social assistance more than unemployment insurance, and there are substantially higher transfers to families with children,
In another strand of research, researchers investigate the fiscal impact of immigration by using a static accounting framework, the purpose of which is to compare benefits derived by immigrants from the public sector with their contributions to compulsary levies. Overall, immigrants are found to be fiscally neutral, meaning they get back only what they give to the system.
Immigration, Attitudes, and Political Preferences
According to Edo et al. (2020), immigrants might have positive and negative effects on the people who are already in their host country — known as “externalities” — since migration affects the cultural, racial, religious, and ethnic composition of that country.
Given what has been covered so far in this post, we expect low-skilled workers should not be opposed to high-skilled immigration, and vice versa, if we ignore preferences for cultural amenities. However, this is not what Edo et al. (2020) find in the literature; they instead find people do not only care about labour market concerns; they will still oppose immigration if they feel they did not get out of life what they feel they deserve. To put it another way, it is to some degree about differences in cultural values: more educated people tend to be less racist and place a greater value on cultural diversity than their counterparts. They are also more likely to believe immigration generates benefits for the host country as a whole.
More generally, voters tend to strongly oppose refugees who migrate for economic reasons, but they are more accepting of asylum seekers who are victims of torture or who are vulnerable in other ways — such as being disabled or having no surviving family members. Regardless, native-born residents tend to prefer asylum seekers who are more employable, all else held equal.
Immigration, Trade, and Productivity
Edo et al. (2020) list four reasons why immigrants might increase trade between their host country and their original country of residence:
Reduce transactions costs associated with language, culture, and local knowledge
Knowledge diffusion
Facilitating networks of trust which replace or facilitate markets
Exercise their preferences for purchases
On that note, the authors cite a Canadian study where a 10% rise in immigration is associated with a 1% increase in exports to the immigrant’s home country, while imports from that country rise 3%. Therefore, economic immigrants tend to have a greater influence on trade than other migrants, although immigrants in specialized “entrepreneur/business” classes have relatively little effect.
Particularly in the case of high-skilled workers — such as those in STEM occupations — immigration can also influence productivity and wages via their contributions to human capital formation and innovation. Immigrant inventors and scientists also bring with them their knowledge and technologies, for example, the Jewish scientists who left Nazi Germany for the U.S. during WWII.
Finally, immigrants affect productivity and economic performance by bringing new knowledge and skills that complement domestic ones.
Summary Remarks
As Edo et al. (2020) conclude, different countries have their own experiences with immigration due to the complexity of the interactions between the composition of immigration flows, receiving country institutions, and other factors. However, some overarching patterns include the following: extreme concerns about the positive and negative effects of immigration are unwarranted. It is, however clear that government policies and practices regarding the selection and integration of new migrants affect labour markets, as well as fiscal and social/cultural outcomes. If governments and immigration officials design systems properly, then everyone can benefit from them.
Thank-you for reading to the end. Next week, I will conclude this series by combining everything written so far to provide an analysis of what Canada has done right and wrong with its immigration policy in recent years, particularly as it relates to Roxham Road, as well as what the Canadian government should do going forward.