We’re very lucky to have three very good papers that zero in on some of the key economic issues facing immigration policy. But before turning to the papers, I would like to make a few general observations to put the papers in context and to underline the importance of taking a hard look at immigration policy at the current time.
Canada has one of highest levels of immigration of any country in the world. The annual level of immigration has averaged 233 thousand per year since 2000, which in per capita terms is three times the officially reported 921 thousand per year level in the United States. Around a fifth of our population is foreign born, almost 60 per cent higher than in the United States.
A major shift in immigration policy was introduced by the Mulroney government in the mid-1980s, ending a ten-year decline in immigration which had brought annual immigration levels down to 84 thousand in 1985. The new Tory policy raised immigration to 152 thousand in 1987 and to above 250 thousand in 1992 and 1993. While the Liberal Government, which took power in 1993, proposed a 1-per-cent-of-population target for immigration, it allowed the numbers to slip to 173 thousand in 1998.
In the mid-1990s, there was a shift to an emphasis on education instead of specific occupational skills in demand and the concept of “absorptive capacity” was scrapped. By emphasizing education rather than specific job skills, immigration policy was trying to produce a more skilled and flexible labour force, not to fill particular occupational needs. And, subsequently, immigration rose to above 250 thousand by 2001 and has remained in the 220-230 thousand immigrants per year range representing around .75 per cent of our population of about 30 million. In addition, since the mid-1980s immigration has not been reduced in response to slowdowns and rising unemployment like it was in the past.
Before the last election, then Immigration Minister Volpe announced the Government’s intention to increase immigration by up to 100,000 per year to 320,000, but he only raised the target for this year to 255,000. While the new Immigration Minister Monte Solberg recently suggested that 300,000 was “too high,” all Prime Minister Stephen Harper would say when questioned following his May 12 announcement on immigration was that “Canada needs more immigrants” without giving any numbers. The Prime Minister’s vagueness is understandable given that immigration has become an extremely politically sensitive issue with the media waiting to pounce on any evidence that the Tories are less pro-immigration than the Liberals.
There are many aspects to immigration: economic; social; political, and even security. While most politicians justify their support for higher immigration in terms of economics, which to me should mean that it will make all Canadians on average better off and not just the immigrants, the real reason for their support is that more immigration is good politics.
Immigrants have traditionally voted Liberal. That’s why Conservatives under Mulroney tried to win away some of that support by adopting more liberal immigration policies. Any of you who have been involved in party politics must know that immigrants play an even greater role behind the scenes in constituency associations and candidate selection than in elections. As a consequence, many MPs, particularly from urban areas, knowing the votes and support to be gained from the immigrant community, spend most of their time on mundane immigration matters rather than weighty affairs of state.
But enough about politics. Today we’re not here today to discuss the politics of immigration, but the economics. It is also perhaps more important as in no other policy area is so much bad economics used to justify good politics.
Now I turn to the economics of immigration in theory. Since this is supposed to be a session on policy, I hope you’ll forgive me for oversimplifying.
As you all know, immigration is a source of population growth and economic development. But while all economists would agree that immigration raises GDP, there’s certainly no consensus that immigration increases per-capita income except in the case of increasing returns to scale technology, which is unlikely to be a factor with relatively free trade. To the contrary, the simplest economic growth model tells us that, with homogeneous labour, constant returns to scale and after capital has had time to adjust, immigration will not raise the per capita income of the already resident population at all.
It’s obviously more complicated than that. In the short-run, immigration means that more workers have to work with the same capital stock. This lowers wages and raises profits according to economic theory. As is well explained in George Borjas’s (1999) analysis of U.S. immigration, Heaven’s Gate, there are gainers and losers from immigration. Businesses gain and workers lose. But not necessarily all classes of workers lose. That depends on the skills of the immigrants, the nature of the production function, and their consumption patterns. If immigrants are mainly unskilled as in the United States, the wages of unskilled labour is likely to be depressed. This issue has given rise to a lively debate in the United States between George Borjas (2003) and David Card. (2005). And the wages of skilled labour will either decrease or increase depending on whether skilled and unskilled labour is a substitute or complement. Skilled labour also gains from their purchases of cheaper labour-intensive goods produced by the unskilled immigrants. On the other hand, in Canada, where the immigrants are supposed to be more skilled, the depressing impact on wages may be greater for skilled labour, but that is only if the immigrants indeed take up skilled employment, which is far from always the case.
The assumption of homogeneous labour is, of course, very restrictive. Immigrant labour can be more skilled or less skilled than the existing population, and hence more or less productive. If immigrants are more skilled and productive, and earn more, immigration can make the already-resident population better off through their net fiscal contribution to the goods and services offered to all residents by Government. On the other hand, if immigrants earn less, the already-resident population will be worse off because they will have to finance the progressive social programs enjoyed by the immigrants.
A growing stream of studies based on the 2001 Census coming out of Statistics Canada has revealed troubling trendsin recent years Their key finding is that there has been a dramatic deterioration in the economic performance of immigrants the source countries have shifted from industrialized countries to the Third World. What exactly do these studies show?
A convenient survey of the studies, some of which were carried out by Abdurrahman Aydemir, is provided by Garnett Picot and Arthur Sweetman (2005). It reports that the cohort that came to Canada in the early 1990s and that were aged 16 to 64 and worked full-time and full-year only earned roughly 70 per cent of comparable Canadians after 6 to 10 years in Canada . And the cohort that came to Canada in the late 1990s only earned around 60 per cent of comparable Canadians in 2001. The last immigrant cohort to eliminate the earnings gap with other Canadians is the one that entered in the late 1970s and then it was only after 16 to 20 years.
Analysis series released by Statistics Canada for the 2001 Census show the underlying deterioration in the labour market performance of recent immigrants aged 25 to 44.Their employment rate has fallen to 65.8 per cent in 2001 from 75.7 per cent in 1981, while that of non-immigrants has risen to 81.8 per cent from 74.6 per cent, driven by the burgeoning labour force participation of women. Correspondingly, the unemployment rate for new immigrants has doubled to 12.1 per cent, while that of non-immigrants has remained roughly stable around 6.4 per cent. A worrisome question that can’t be ignored is: how much worse would this deterioration have been if the economy had slid into a recession?
Even more troubling because of the large numbers of new immigrants who are not able to work full-year, full-time is the increase in the low-income rate among new immigrants who have been in the country five years or less. This is the proportion of a group falling below Statistics Canada’s Low-Income Cutoff, a generally used measure of poverty. It rose to 35.8 per cent in 2000 from 24.6 per cent in 1980. This reflects the relatively high numbers of new immigrants who work part time, or are not earning money because they are unemployed or not in the labour force as well as the fact that new immigrants earn less when employed full-year, full-time. In contrast, the low-income rate among non-immigrants actually fell to 14.3 per cent from 17.2 per cent over this same period. The low-income rate for new immigrants in 2000 was thus a striking 2.5 times that of non-immigrants. The rising low-income rate among immigrants, who according to a study by Hou Feng and Garnett Picot (2004) have been increasingly settling in “enclaves,” accounts for more than the total increase in low-income rate in Toronto, Vancouver and Montreal.
If these trends continue and are confirmed by the 2006 census, it will be increasingly difficult for proponents of continued high immigration to claim that more immigration is going to make already-resident Canadians better off. The maintenance of the Canadian welfare state will require an ever increasing tax load on the already-resident population to preserve social programs and to prevent growing inequality. For instance, Herbert Grubel (2005, p.19) has estimated crudely, but conservatively, that the cohort of immigrants who entered the country in 1990 received net fiscal transfers of $6,294 each or $1.36 billion in total. And applying this to the 2.9 million immigrants that entered the country between 1990 and 2003, he estimated the total net fiscal transfers to recent immigrants would be $18.3 billion per year
If immigration is going to make already-resident Canadians better off, the immigrants selected have to be those who will do well economically after they take up residence in Canada. Two of the papers today shed important light on this important issue.
The first by Abdurrahman Ayedemir explores the relative performance of different immigrant visa category holders using new data from the Longitudinal Survey of Immigrants to Canada (LSIC). Not surprisingly, those admitted on the point system “are the most educated and possess the highest language skills among all visa categories.” The results of Ayedemir’s study are somewhat encouraging in that they show that those selected for their skills have higher weekly earnings than family class and refugee class immigrants, even if they don’t have stronger labour force participation or employment outcomes. It is discouraging though to hear that years of schooling, which is almost all foreign education, has a zero or negative impact on participation and employment outcomes and only a small positive effect on earnings. As Aydemir says, this would constitute a reversal of the results of earlier studies.
I’m not so sure about the significance of Aydemir’s finding that 90 per cent of the schooling differential between skilled worker and family class immigrants is due to within region differences and less than 10 per cent due to differences in the national origins. I agree that it’s a good thing to be able to pick out the most educated from a country that has a relatively low level of education. But, on the other hand, I wonder if it might not be better to draw our immigrants from countries with higher levels of education because that would ensure that the subsequent family class immigration of relatives of the economic class immigrants had a higher skill level.
While Aydemir’s contention that 3 out of 4 members of the 2000-2001 immigrant cohort were based on skill requirements and assessed by the point system exaggerates the selectivity of immigration as it counts the spouse and children of the Principal Applicant, it is reassuring to hear that “the point system has an indirect effect on spouses’ skill distribution...due to assortative matching.” A problem with only accessing the Principal Applicant who is most likely to be male is that, regardless of their education level, the female spouses often come from countries where it is not customary for woman to participate in the labour force. This explains their lower level of participation once they arrive in Canada and contributes to lower family incomes.
Aydemir’s observation that the weight on education has increased over time by raising the pass mark and lowering the points for a given level of education makes one wonder how much worse the performance of economic class immigrants would have been if the system hadn’t been tightened. It’s definitely a good thing that since the 2000-01 cohort was screened the points for a high school diploma have been lowered to 5 and for a university degree raised to 20 to 22 points out of maximum of 25.
Aydemir’s finding that speaking is the only language factor that has a positive and significant impact of labour market outcomes is interesting. I wonder if this has any thing to do with the fact that this is the only skill that the immigration officer is exposed to in the interview and has a chance to assess. It’s hard to believe that reading and writing skills aren’t important for skilled immigrants, particularly professionals. And, as noted by Aydemir, this has been confirmed in other studies.
But I guess that Aydemir’s paper shows that point system works to a certain extent in identifying skilled immigrants at least on paper. However, it leaves open the question of whether the system could actually be tightened to generate skilled immigrants who would actually outperform existing Canadian residents. If the skilled class doesn’t actually do this, it’s hard to see how it can really be counted as being so skilled. The market is telling us that for one reason or another the economic class immigrants aren’t as skilled as their credentials would lead us to believe. If not, discrimination must be rampant in Canada.
To fully understand what is going on, we need much harder data on the real skill and educational levels of immigrants, including language skills. It is important to make sure that the skill and educational levels of immigrants are indeed what they claim. Just because somebody says they’re a university graduate and has a paper diploma doesn’t mean they actually graduated from a university. It is also necessary to evaluate the quality of educational institutions in the Third World. If the quality of particular educational institutions is low, why should their graduates be given the same points as graduates of the world’s leading universities? And what about language skills. Can an Immigration Officer really do a very good job of assessing language ability in a brief interview?
This is where the paper by Bonikowska, Green and Riddell comes in. It is by far the most innovative and sheds some very important light on the question of why the recent cohort of immigrants is performing so poorly in spite of their reported educational and skills levels. By using data on cognitive skills, including literacy, numeracy and problem-solving skills, from the International Adult Literacy and Skills Survey (IALSS), they are able to reject the hypothesis that lower earnings of male immigrants for given educational and experience levels reflects discrimination and explain the lower earnings in terms of lower levels of cognitive skills.
The 45-percentage point differences in average skill levels between immigrants with no Canadian education and the native born is striking, particularly given the higher level of educational attainment of the immigrants (Table 2). And this is only for men. If I may be permitted to speculate, the results for women could very well be worse given the lesser priority put on educating woman in the Third World countries from which most of our immigrants come.
While it’s possible to get some information from the regressions, it would be interesting to see a simple cross tabulation of education level and cognitive skill scores for the different classes of immigrants, with the group that got some education in Canada broken out separately, and for the native born. It would also be interesting to see the cognitive skills scores broken down by country of origin for immigrants as well as immigrant class. This would help us to identify where there is a discrepancy between cognitive skills scores and educational levels.
Bonikowska, Green and Riddell’s finding that if immigrants had the same average skill level as the native born, the earnings differential would be eliminated for those with a high school education and halved for those with university is very telling.
The relatively higher skills level of those with some Canadian education is also important. It may be telling us something about the quality of education in many of the countries where immigrants receive their education. It is worth noting that there is only one university from outside the industrialized world on the list of the top-100 universities compiled by the Institute of Higher Education at the Shanghai Jiao Tong University. The State University of Moscow. In contrast, Canada has 4. And there are only 16 universities from the Third World among the top-500, whereas Canada has 23. We need to know how many Third World immigrants actually attended any of these elite institutions in their countries or abroad for that matter. We already know that most Canadian students graduate from the elite Canadian universities since in effect almost all the larger Canadian universities are classified as elite. And the same applies to immigrants who get their university education after landing in Canada, which might explain whey they tend to do as well as native born Canadians.
Now for the paper by Guillemette and Robson. It addresses the all too oft-cited argument that immigration is essential on demographic grounds because of the ageing of the Canadian population. This concern about ageing and population decline stems from the fact that in Canada fertility rates at 1.5 have fallen below replacement rates. Can immigration really offset population aging? I’m sure everybody here knows that answer even if they don’t want to say it. We can leave that to Guillemette and Robson who don’t mind voicing politically unpopular opinions.
In their paper, they effectively present immigration scenarios that show that reasonable levels of immigration can not do much to prevent the aging of the Canadian population. This supports the points made by Roderic Beaujot (1999) in his various studies on immigration and demographics. It is also consistent with the findings of the Demographic Review (1994) which also showed that even absurdly high inflows with 50 per cent under age 15 were not enough to produce continued population growth with a fertility rate of only 1.7. And with those of Denton, Feaver and Spencer (1997) who estimated that the difference between 200,000 and 500,000 immigrants per year would only lower the ratio of those over 65 from 24.8 to 21.3.
On the other hand, Maxime Fougère, Simon Harvey, Marcel Mérette and François Poitras (2003, p7.) contend that “maintaining the recent trend in migration in the future world would significantly contribute to reduce the rise in the elderly dependent ratio.” But this is from a base case with zero immigration and the actual reduction is only to around 40 per cent from 50 per cent by 2045. So it doesn’t really contradict the Guillemette and Robson or the other earlier studies.
There is also a political economy reason why immigration can’t offset the effects of aging that needs to be mentioned. Since immigration is not really for economic reasons, there will always be political pressure for family class immigrants to creep up as a share of the total. As they average around 5 years older, they will have a much smaller impact in mitigating the impact of an aging population. And the political appeal of family reunification makes it impossible to weigh immigration too much in favour of economic class.
Given the strength of the consensus among economists and demographers that immigration can not prevent the ageing of the Canadian population, it is very humbling for our profession to be ignored by politicians who keep arguing that we need more immigration to offset the effect of our aging population. Don’t they read their briefing notes? Or do they just choose to disregard the truth?
In addition, there is a question of the extent to which dependency ratios are overstated. Dependency rates are supposed to measure the number of those who are too old to work relative to the working population. But over time health has improved. And life expectancy for males in Canada has increased from 68 in 1960 to 76 in 2000 and for females from 74 to 82. This is an 8-year increase in life expectancy yet people are still considered too old to work after 65. Part of the solution to the growing costs of pensions resulting from the ageing of the population must be raising the retirement age as is being done under the U.S. Social Security System.
There are also other innovative approaches that are being adopted in Europe where the problem of population ageing is much more severe than in Canada. These include part-time work by the elderly to meet labour shortages.
A typical statement is we need more immigrants to pay the pensions and health care of aging baby boomers? How likely is this? What if the immigrants don’t earn enough to be net fiscal contributors to the government’s finances. Or what if they eventually vote to cut pensions and increase other forms of social benefits?
Two of the three papers we heard today provide evidence that our current approach to immigration policy is not working. One provided hard evidence that the skill levels of immigrants are not exactly what they seem on paper. And the other showed why immigration is not a viable strategy to prevent population ageing. The third paper gives us an interesting look at immigrant selection.
We need more such serious research to find out why recent cohorts of immigrants are performing so poorly so that we can improve our selection procedures and criteria.
We also need to carry out more research on the fiscal costs of recent cohorts of immigration. Then even if we still decide to continue to pursue an immigration policy directed at political, social and humanitarian objectives, it will be with a much better understanding of the economic costs associated with such a program.
In this regard, it is also important to examine the impact of immigrants on wages. And not just on aggregate wages. Are the large numbers of highly skilled people coming in depressing wages of skilled workers like engineers, scientists and computer technicians? Or is the real impact on unskilled occupations such as the earnings of taxi drivers for example? There has been a real debate in the United States over this issue with dueling economists Borjas (2003) versus Card (2005), but so far nothing comparable has begun in Canada.
And after we have better information on the costs and benefits of current immigration policy, including its distributional impacts, we need seriously to consider the politically incorrect question of whether it really makes economic sense for Canada to continue to take in year after year such large numbers immigrants who may not really be as skilled as we want.
Beaujot, Roderic (1999) Immigration and Canadian Demographics: State of the Research.
Borjas, George J. (1999) Heaven’s Door: Immigration Policy and the American Economy (Princeton University Press)
Borjas, George J. (2003) “The Labor Demand Curve Is Downward Sloping: Reexamining the Impact of Immigration on the Labor Market,” Quarterly Journal of Economics.
Card, David (2005) “Is the New Immigration Really So Bad?” NBER Working Paper 11547, August.
Denton, F., C. Feaver and B. Spencer (1997) “Immigration, Labour Force and the Age Structure of the Population,” A paper presented at the Canadian Employment Research Conference on Immigration and the Economy.
Fougère, Maxime, Simon Harvey, Marcel Mérette and François Poitras (2003) “Ageing Populations and Immigration in Canada: An Analysis with a Regional CGE Overlapping Generations Model,” Paper presented to the 4th Annual Research Conference on Social Security, Antwerp, Belgium, May 5-7.
Grubel, Herbert (2005) "Immigration and the Welfare State in Canada: Growing Conflicts, Constructive Solutions," Fraser Institute Public Policy Sources, Number 84, September.
Health and Welfare Canada (1994) Charting Canada’s Future: A Report of the Demographic Review.
Hou, Feng and Garnett Picot (2004) "Visible Minority Neighbourhoods in Toronto, Montréal and Vancouver," Canadian Social Trends, No. 72, spring, pp.8-13.
Picot, Garnett and Arthur Sweetman (2005) “The Deteriorating Economic Welfare of Immigrants and Possible Causes: Update 2005,” Statistics Canada, Catalogue no. 11F0019MIE2005262.