3 Patterns in the incidence of degree holders

View the most recent version.

The overarching purpose of this paper is to better understand the relative contributions of migration and in situ human capital growth to the wide variation that we observe in the shares of degree holders across urban and rural parts of Canada. Before delving into how the incidence of university-degree holders varies across the urban-rural hierarchy, it is useful to explore national trends, if only to provide a benchmark for urban-rural comparisons.

From 1996 to 2001, the share of degree holders in Canada increased significantly. In 1996, the share of the working-age population (age 15+) with degrees was 13.3%, while this share was 15.4% in 2001 (Censuses 1996 and 2001). Over a relatively short span of five years, the share of degree holders increased by 2.2 percentage points or about 16%.

To see more clearly the dynamics of this process, Figure 1 plots the share of degree holders for persons aged 15 through 65 in 1996 and 2001. Both curves roughly take the shape of a wave, albeit one with two crests. The steep back of the wave consists of persons aged 21 to 25, corresponding to years during which students typically complete their undergraduate university degrees. After age 30, the share of persons with degrees generally declines, which reflects the smaller proportion of older generations that completed university.

Comparing the 1996 and 2001 curves, it is evident that for almost all age classes there is a higher share of degree holders in 2001 than in 1996. This reflects the generally higher rates of degree attainment among the younger generations. The only ages where the curves converge are very young ages, when degree attainment is quite rare, and the 41-to-47 age group, where there has been a lull in the growth of degree attainment.

Figure 1
Degree holder share of employment by age, working age population from 15 to 65, 1996 and 2001

As there are differences across time, there are also differences in degree-holder incidence across the urban-rural spectrum (see Figure 2). The urban-rural spectrum encompasses large, medium and small urban areas, as well as rural parts of Canada. Large urban areas include census metropolitan areas (CMAs) with a population above 500,000, while medium urban areas are made up of CMAs with a population from 100,000 to 499,999. Small urban areas take in census agglomerations (CAs) with a population greater than 10,000. Rural is defined as non-CMA/CA census subdivisions.

There is an obvious gradient across the urban-rural hierarchy, regardless of age. The highest rate of degree attainment is in large urban areas, and the lowest rate is associated with rural areas. Moreover, this gradient appears to be set to increase over time. To illustrate, the share of degree holders in large urban areas aged 25 is about three times that of rural areas, while for 60-year-old workers, this ratio falls to around two. This strong 'demographic wave' of degree holders for large- and medium-sized cities suggests that the incidence of degree holders between these city- size classes and rural areas may diverge further.

Figure 2
Degree holder share of employment, by age 15 to 65, by city size, 2001

Together, these figures raise the obvious question of what underlies the significant degree-holder gradient across urban and rural areas. Is it due to higher rates of migration of degree holders to large urban areas, either from elsewhere in Canada or from abroad, or does it derive from higher rates of in situ growth?

This is a difficult question to answer, if only because the rates of degree attainment of urban and rural areas reflect the lamination of degree holders that obtained their degrees and/or moved to a particular location decades ago in the past. With census data that spans five years, we cannot explain the incidence of degree holders per se. That is, we cannot explain the origins of the entire stock of degree holders. However, using census data we can gain a perspective on why we observe such a wide variation in the incidence of degree holders across the urban-rural hierarchy by observing the flow of degree holders.

To do so, we follow two strategies. One is to decompose the growth in the number of degree holders over the 1996-to-2001 period into that which is due to migratory flows (domestic and international) and that which is due to in situ growth of degree holders. To the extent that the lamination of degree holders over this five-year period is representative of past laminations, decomposing growth in this way provides a perspective on the origins of the differences in degree shares of urban and rural areas.

The second strategy involves focusing on persons in their early career stages (ages 20 to 29). It is at these ages that individuals are obtaining their degrees and are most mobile. If we can better understand the sources of growth in these degree holders for urban and rural areas, then we may also gain an understanding of what will drive their future growth.