Les recherches indiquent que l’accroissement de l’immigration n’a pas d’incidence (positive ou négative) sur l’emploi et les salaires, mais ces travaux restent à approfondir.
In recent months there has been much policy debate about whether to increase Canada’s immigration levels. On Tuesday, federal Immigration Minister John McCallum announced the government would increase the base line intake of immigrants next year to 300,000, arguing it “lays the foundation for future growth.” Meanwhile, the federal government’s Economic Advisory Council on growth recently recommended that the government increase its annual immigration levels to 450,000. The new Century Initiative, formed by leading Canadian voices from the private sector and nongovernmental organizations, is advocating for Canada to reach a population of 100 million by the year 2100. Given these initiatives, this seems like an ideal time to ask what we know from existing research in economics on the likely economic impact of increasing Canada’s immigration levels.
There is a considerable literature on this subject in economics. While that literature includes some excellent Canadian studies, much of it focuses on the US and Europe. Since our immigration situation is so different from those in both those regions, not all the conclusions can be applied to Canada. Thus, one of our goals here is to present a list of what we do not currently know for Canada but would need to know to create effective policy. That list suggests some important gaps.
As a starting point, we need to ask what Canada’s goal is for immigration policy. We believe that the ultimate goal for all policy is to create a just and prosperous society in which citizens have substantial opportunities to pursue their goals. Instrumental to that larger goal is the objective of increasing per capita income. It is important, at the outset, to emphasize the per capita part of this goal: simply maximizing GDP is not helpful if it does not improve GDP per person (and, ultimately, the well-being of individual citizens). Pro-immigration statements often put too much emphasis on total GDP without showing that there is a link between increasing the total size of the economy and individual well-being. Simply noting that India has a larger GDP (and, of course, much larger population) highlights the point that setting out to maximize GDP or population is not an obviously worthwhile pursuit.
Pro-immigration statements often put too much emphasis on total GDP, without showing that there is a link between increasing the total size of the economy and individual well-being.
Our ultimate interest in a just society also means that we are concerned not just with average well-being but also with how any gains from economic growth are distributed. It is now commonly argued that Brexit and support for Donald Trump is ultimately grounded in a lack of attention to sharing the benefits of globalization.
For most people, their main source of income is employment. That means that assessing the impact of increased immigration on the well-being of current Canadian residents should focus on the impacts on wages and employment. A sizeable body of economic studies has emerged in recent decades on exactly this question.
The seminal (and most widely cited) study in this literature is a study of the impact of the Mariel Boatlift on the Miami labour market by David Card, a professor at the University of California at Berkeley. In May through September of 1980, Fidel Castro opened the doors, allowing Cubans to leave the country. Over 125,000 took the opportunity. The vast majority of these “Marielistas” went to Miami, increasing that city’s population by 7 percent in a matter of months. To put this in perspective, it would be the equivalent of a decade’s worth of Canada’s permanent immigration arriving in a five-month period. Card asked what this influx did to wage and employment rates by comparing outcomes in Miami with those in comparable US cities that did not receive an extra inflow. The answer was surprising: initially wages and employment rates for those already living in Miami before the boatlift declined, but about three years after the huge influx, wages and employment rates in Miami showed no effect, relative to comparable cities.
The conclusion that immigration has a limited effect on wages and employment rates in receiving economies just a few years after an influx has been confirmed in other countries and settings. There are studies that show negative effects on wages within skill groups defined by education and experience (including a well-cited study for Canada), but even in those studies, the estimated effects are not large. Our reading of the sizeable existing literature is that increased immigration inflows have limited impacts on wages and employment rates in the medium to long run. We are not alone in this conclusion. A recent report from a panel of experts for the National Academy of Sciences, Engineering and Medicine in the US concludes, based on an examination of the existing literature, “When measured over a period of 10 years or more, the impact of immigrants on the wages of natives overall is very small.” Their conclusion on employment impacts is the same.
The result that immigration has small impacts on wages and employment rates is one that is commonly accepted among most economists.
The result that immigration has small impacts on wages and employment rates is one that is commonly accepted among most economists, especially those who study immigration, but it does not seems to be well known by non-economists. Some commentators argue that immigrants have large negative impacts – stealing jobs and reducing wages. That, indeed, is one source of anti-immigration sentiment in the US and Europe (and, likely, Canada). The claim of substantial negative impacts is simply not supported by the evidence. On the other hand, neither are claims of substantial positive impacts. Immigration cannot be relied upon as a source of higher per capita incomes – at least, sources of incomes that show up in the wages and employment prospects of Canadians. In some ways, this is liberating. It means that we can think about other goals for immigration policy – in terms of yielding a more diverse, dynamic society or offering opportunities to people from other parts of the world – without concern for significant negative effects on Canadian workers. But it also implies that we should not view immigration as a source of greater prosperity.
These are statements about average impacts in the labour force. But do these small effects on average hide large losses for some groups and gains for others? The US and European literature implies that there is some distributional impact. Previous immigrants and low-skilled native- born workers tend to experience negative impacts, while more-educated residents benefit from increased immigration though, again, neither of these effects is large. One paper using Canadian data suggests that the impacts could be the opposite for Canada.
This is the first point at which we encounter the question of the applicability of results from other countries. Canada’s point system implies that our immigrant inflow is more skilled than those of most other countries. Estimates of immigration impacts in the US, for example, are typically viewed as being driven by the large illegal, and generally low-skilled inflow to that country. It is not clear that increased immigration to Canada, to the extent that it is selected, would also have negative impacts on wages for previous immigrants and less skilled workers. More Canadian research is needed to understand the impact of our immigration across skill groups.
There is some evidence that more-skilled immigration inflows have positive effects on wages and employment for less-skilled workers but these impacts are, again, small. This, in itself, may seem surprising, given that proponents of expanded immigration argue that a benefit of immigration is that it fills gaps in the skill structure, to the benefit of all. But underlying this claim is a false depiction of how economies work. They are not machines that stop working if one cog is missing. Market prices change alongside shifts in society, and companies and consumers ultimately respond to those price changes. The economy, which is an organic and flexible entity, adjusts itself. This is not to deny that greater skills are good for an economy, just that claims of extraordinary growth based on filling gaps are not accurate.
With the impending retirement of the Baby Boom, some have argued for immigration as a means of rebalancing Canada’s age distribution. The results on this topic, however, are definitive: immigration is not a means to substantially alter Canada’s age structure and impending increase in the dependency ratio. Inflows of immigrants are simply too varied in their age structure. They do not consist just of young workers but include older family members as well.
Immigration, it is also argued, yields positive benefits through innovation and firm creation. There is some international evidence of a positive association between immigration and innovation, but the magnitude of the effect is unclear and the studies involved tend to either have a very localized focus (e.g., the impact of Einstein on the research output of those around him) or to not have clearly addressed the question of separating correlation from causation (are immigrants creators of innovation or do they move to it)? The evidence on firm creation by immigrants is also in its infancy. The work that does exist indicates that immigrants are more likely to start firms than the native-born, but that their firms are smaller in size.
The now discredited immigrant investor program was not successful partly because its developers leapt into immigration policy changes without a good understanding of the consequences.
Two words of caution are appropriate in this area. First, the now discredited immigrant investor program was not successful partly because its developers leapt into immigration policy changes without a good understanding of the consequences. The program thus appears to have had limited impacts on per capita incomes but sizeable impacts on house prices in Vancouver and Toronto. Second, it is important not to get distracted by individual stories of successful immigrant entrepreneurs. They certainly do exist, but that is not really relevant. The relevant questions are, would we admit enough new entrepreneurs to raise per capita well being by increasing the total number of immigrants? Would we be able to select who will be successful entrepreneurs at time of application? The research discussed earlier indicates that increased immigration levels do not substantially improve wage and employment outcomes in the medium term – whether immigration has its effects through bringing in labour, skills, or entrepreneurs. Here, too, more Canadian research is needed.
A final, particularly Canadian question related to immigration is that of geography. Since the introduction of the selection system in the late 1960s, most immigrants have settled in Canada’s major cities – the Greater Toronto Area, Vancouver and the Lower Mainland, and Montreal. Some have argued that immigration might be better for the economy – and might not add to congestion in our major cities – if the provinces were more involved in the selection process and could admit applicants suitable for their local economies. We have some experience relating to this view. Quebec has had its own selection system for many years, and Manitoba has had an active Provincial Nominee Program (PNP) since 1998. Most other provinces have also introduced PNPs.
For immigration to work toward a goal of spreading people and skills more widely across Canada, we need three conditions. First, we need to be able to get immigrants to the less urban regions. Second, they would have to stay there, at least long enough to have an impact. Third, their presence would have to actually have an impact on the local economy, increasing wages and employment opportunities there. PNPs have been successful in meeting the first of these goals. We can use our experience so far with PNPs to examine the two other conditions as well. Some studies of PNP immigrants indicate they are initially more successful than other immigrants – perhaps because having a job offer is often required in PNPs – and that, while many leave their initial province of arrival within three years, a substantial number remain. But immigrant incomes grow more slowly than those of other Canadians, so that incomes of PNP and non-PNP immigrants wind up converging after 5-7 years. And the more educated newcomers are more likely to move, reducing the value of PNPs for increasing the human capital stock in areas outside the big cities.
On net, increasing the number of PNP immigrants may affect areas outside the largest cities, but there hasn’t been enough study of longer-term outcomes to be sure. Bringing in a lot of immigrants through this route may have some of the desired effects, but it might instead just generate large, delayed inflows into our major cities. More research is needed before we go too far in this direction.
Poverty is rising among immigrants at a time when poverty rates for native-born Canadians have been declining.
Some of the information we have on the experience of newly arrived immigrants in recent decades is also relevant for thinking about the impact of expanding immigration levels. Newly arrived immigrants make significantly less money than native-born workers, particularly once we condition on education and experience. To some extent they do catch up with time in Canada, but recent cohorts of entrants have fallen further behind and are not catching up as fast. As a consequence, poverty is rising among immigrants at a time when poverty rates for native-born Canadians have been declining (see Garnett Picot and Feng Hou’s analysis in the IRPP’s Income Inequality book). Before 2000 social assistance receipt among immigrants was generally below that of the native-born, but recently it has consistently been higher. The earnings and employment experience of immigrants relative to the otherwise comparable native-born is an area where there is considerable Canadian evidence.
These trends imply that newly arrived immigrants are a net drag on government budgets: they pay less in taxes on average and make average or slightly above average use of government services and benefits. Second-generation immigrants do well, however, which may offset this net drag to some extent, but the initial impact of a large increase in immigration should be expected to be an increase in taxes, a decrease in services, an increase in deficits, or some combination of the three. The exact extent of any such effect needs more study for Canada.
Overall, our reading of what we know so far leads us to five main conclusions. First, there is no evidence that substantially increased immigration will lead to increased growth of a type that shows up in positive impacts on wages and employment of those already here. Canadians have generally favourable views toward immigration, and most agree with the statement that “immigration is good for the economy.” What a large body of evidence indicates is that immigration is neither good nor bad for the economy, as assessed by its impact on the living standards of current citizens. Second, there is also no evidence of dire consequences from immigration. This opens the possibility of using immigration policy to meet other goals without having substantial concern about the economic consequences.
One of those goals might be diversity. Canada has been successful in its attempts to accommodate diversity. However, it is worth noting that there is a large body of research literature in political science, sociology and economics that concludes that ethnic, cultural or racial diversity has negative impacts on the ability of some democracies to function well (for example, by lowering interpersonal trust) and the willingness of the more fortunate to support taxation and social policies that help the less fortunate. So far Canada appears to have avoided these negative consequences, but we should be cautious about making the kinds of substantial increases to immigration, especially to our big cities, that could disrupt that success.
Third, fiscal impacts of substantial increases in immigration are likely to be negative, at least in the short run.
Fourth, immigration is not a solution to demographic problems such as population ageing.
Fifth, there are some areas in which we need more information, especially about the Canadian context. These include the distributional impacts of our immigration, the impacts on innovation, and the long term consequences of trying to target immigration outside the biggest cities.
We believe there should be a major research and policy investigation to examine the aspects of immigration about which we do not have good Canadian evidence, before any major changes are made. Fortunately, Canada has excellent immigration data, as well as well-trained social scientists who could use that data, in order to generate the needed research. We believe (and we think the current federal government believes) in evidence-based policy. Immigration is an area where there is a need to generate that evidence and the real potential to do so.
Do you have something to say about the article you just read? Be part of the Policy Options discussion, and send in your own submission. Here is a link on how to do it. | Souhaitez-vous réagir à cet article ? Joignez-vous aux débats d’Options politiques et soumettez-nous votre texte en suivant ces directives.