Analytical Studies Branch Research Paper Series

11F0019MIE

International Mobility: Patterns of Exit and Return of Canadians,
1982 to 2003

PDF version

International Mobility: Patterns of Exit and Return of Canadians, 1982 to 2003

by Ross Finnie

Executive summary

Through the late 1990s, many thought that Canada's best and brightest were leaving the country in unprecedented numbers. Using the Longitudinal Administrative Database (LAD), constructed from individuals' tax records, this study sheds new light on the extent and nature of the flows of Canadians to other countries and the patterns of return over the period 1982 to 2003, asking who leaves and who returns. It finds that absolute numbers of leavers remain small (0.1% in any given year, i.e., one tenth of 1.0%). However, with the exception of those at the highest income levels, few return.

The LAD is a 20% random sample of all Canadian tax filers constructed from Canada Revenue Agency tax files. It follows individuals longitudinally and matches them into family units on an annual basis, providing individual- and family-level information on incomes, taxes, and basic demographic characteristics.

The departure model uses the following sets of variables: basic demographic characteristics (age, family type, province/region of residence, language, and area size of residence); income (an indicator of having received Employment Insurance [EI] income is included); economic conditions faced by the individual (i.e., provincial and U.S. unemployment rate); a series of calendar year dummy variables (to capture time trends and influences which operate at a national level, such as policy changes); finally, a set identifying recent immigrants and the number of years since immigration.

In the return model, the dependent variable is the probability of returning to Canada in any given calendar year. Unlike the departure model, this uses a hazard model approach, whereby only individuals who are observed to first leave the country are included, and they are tracked in a precise year-by-year fashion from the year of their departure. Individuals' characteristics as of the point of departure are included as regressors, thus identifying the relationship between the rate of return and these attributes.

Raw data show that annual rates of departure from Canada in the 1982-to-2003 period are low, ranging from 0.045% to 0.133%. Departure rates follow the economic cycle to a significant degree: with a stronger economy, there are fewer leavers. But this is by no means a perfect correlation. Other factors are at play as well: the U.S. economy, the Canada–United States Free Trade Agreement and North American Free Trade Agreement, and immigration trends.

Only a minority of those who leave ever return. Over the period covered, about 2.5% returned after one year, and about 15% returned within five years. There was a substantial increase in returns since 2000, however, mirroring what was happening on the departure side. Return rates are (like departure rates) significantly higher for those at higher-income levels, suggesting such individuals are generally more mobile.

The main departure model results show that the rates of leaving the country are fairly average for the youngest individuals (aged 18 to 24), then rise (aged 25 to 34), and decline thereafter.

Couples with no children have low departure rates, a result which contrasts with interprovincial mobility, where having children and being married are generally related to lower mobility rates. By way of contrast, single parents, and single mothers in particular, have relatively high departure rates.

People living in Atlantic Canada are considerably less likely to leave, those in British Columbia (including the territories) have the highest rates, while those in Ontario and the Prairies (including Alberta) are in the middle rank.

Francophones in Quebec have the lowest rates of all Canadians, but Quebec Anglophones have a much higher rate of departure than all other groups. Francophones outside of Quebec have somewhat higher departure rates than English speakers in the province/region in which they live.

Individuals living in larger cities are about twice as likely to leave as rural dwellers, and substantially more likely than those in smaller cities.

Men receiving EI are less likely to leave than other men, but such an effect does not hold for women. The provincial unemployment rate has a significantly negative effect on leaving, while he higher the rate relative to the United States, the greater the likelihood of leaving.

The higher the individual's income, the greater the probability of leaving, especially at the highest income levels ($60,000 to $100,000 and $100,000 and more). To the degree that income levels capture the exit of highly skilled workers, rates of departure are greater for the most talented workers. That said, their numbers are small, as relatively few have incomes at these levels, and the vast majority of leavers are in the lower-income categories.

The immigrant effects are very strong. In a male immigrant's landing year, his chances of leaving the country are 10 times greater than those of a non-immigrant Canadian with similar characteristics. Rates rise over the early years in the country, beginning to decline 6 years following immigration — although even those who have been in Canada as long as 16 years (or more) have more than double the rate of leaving the country in any given year than non-immigrants.

The adjusted trends show that once controlling for other factors, rates have not moved as dramatically over time as the raw data suggest. Otherwise put, some of the swings in the 1990s and since were linked to factors explained by the variables included in the models.

Individuals are more likely to return after having been away two years than just one, but then the rate of return declines. The rates reach a maximum of 4.4% in year 2, then drop to 3.7%, 2.5%, and 2.0% over the following three years. After five years, 15.1% of those who had left had come back.

Those 65 and older are the least likely to return, the two younger groups (18 to 24 and 25 to 34) are the most likely, and the other groups are in between. There is no clear pattern by family status, except that lone parents have the lowest rates of return.

Provinces with lower rates of departure have higher rates of return. Francophone Quebecers are the most likely, and Anglophone Quebecers the least likely to return. Similarly, those in rural areas and smaller cities are more likely to return than those from larger cities.

Those who had EI before departing do not behave differently than others. A higher provincial unemployment rate means a lower rate of return, but the effect is not statistically significant. The same is true for the relative Canada–United States unemployment rate.

Return rates of high-income individuals are significantly higher than those with lower income levels. They are more evidently mobile — in the case of returns as well as departures.

Among recent immigrants, those who leave within the first few years of arriving in Canada are less than half as likely to come back as non-immigrant Canadians, and while immigrant behaviour gradually moves towards that of non-immigrants with the number of years they had been in Canada before leaving, their lower rates of return persist even after having been in Canada more than a dozen years.

When the calendar year variables are plotted, rates declined through 1990, remained flat through the 1990s, then rose after 2000. Thus, while rates of leaving the country generally rose through most of the 1990s, return rates remained flat. When departure rates fell significantly after 2000, return rates rose. Whatever was attracting Canadians abroad through the 1990s, return rates held steady, while the more recent changes which appear to have been working in the opposite direction since 2000 hold for both departures and returns.

You need to use the free Adobe Reader to view PDF documents. To view (open) these files, simply click on the link. To download (save) them, right-click on the link. Note that if you are using Internet Explorer or AOL, PDF documents sometimes do not open properly. See Troubleshooting PDFs. PDF documents may not be accessible by some devices. For more information, visit the Adobe website or contact us for assistance.