1 Introduction

View the most recent version.

There is a strong, positive and consistent relationship between the population size of a metropolitan area and the share ofits population that has a university degree. In Canada, the degree-holder share of the rural population was 7% in 2001. This share rose to about 10% for small urban areas with a population from 10,000 to 99,999, and rose further to just under 15% for medium-sized urban areas with a population from 100,000 to 499,999. For large metropolitan areas with a population greater than 500,000, their degree-holder share was on average just above 20% (Census 2001).

Human capital, which is often measured by degree holders ¹ , has been an important source of growth for cities and a driver of differences in wage levels. Evidence from the United States, for example, clearly shows a positive association between initial levels of human capital and long- run growth (Glaeser, Scheinkman and Shleifer 1995; Glaeser and Saiz 2003; Shapiro 2005). Variation in human capital is also a major determinant of wage differences across cities in France (Gobillion, Duranton and Theiry 2006), the United States (Rosenthal and Strange 2006, Yankow 2006) and Canada (Beckstead et al. 2008). The reason that large cities have higher wage levels is less because their large and industrially diverse economies provide firms with a productive advantage, but more because their populations tend to have invested more in their human capital through formal education.

Given the large and systematic variation in the population share of degree holders across the urban-rural spectrum, as well as its implications for economic welfare, a natural question to ask is what led to these patterns. It can be argued that growth in human capital in any particular location can be decomposed into two processes—net migration and in situ growth—that work alone or in combination, and can result in one locale having a higher share of human capital than another.

Variation in human capital growth across cities may be associated with net migratory flows (domestic and foreign), where particular cities are able to consistently attract more educated workers than they lose through migration. Faggian and McCann (2006), for example, note that a recent study conducted by the U.K. Treasury and Department for Trade and Industry found that there were 40% more university graduates employed than were educated in London; conversely, areas in the economically weak north of the U.K. employ fewer graduates than are educated there: this suggests that large scale inter-regional migration flows are capable of shifting human capital geographically. In the Canadian context, domestic migrants are also attracted to Canada's largest urban centres because of higher returns and greater opportunities. At the same time immigrants predominately settle in the same centres and increasingly arrive in Canada with high levels of human capital and advanced degrees.

Most of the recent work that emphasises the role of human capital as a driver of growth and, therefore, as an important tool for economic development (e.g., Florida 2002a, 2002b) has assumed migratory flows underlie much of the variation in human capital across geographic space. However, incentives to invest in human capital can vary significantly from place to place, which can result in significant differences across the urban-rural hierarchy.

Recent literature on educational attainment, for example, notes greater degree attainment in large urban areas. In part, this represents proximity and availability of postsecondary education (Card 1993, Frenette 2002, 2003, 2007), increased returns to education in large urban areas (Bouchard and Zhao 2000, Fortin and Lemieux 2006), historically higher levels of degree holders in urban areas, family expectations and income, and a greater diversity of employment opportunities. In contrast, several studies have shown that rural youth tend to aspire and move into so-called lower-status occupations and they are less likely to attend university (Andres and Looker 2001; Bouchard and Zhao 2000; Christofides, Cirello and Hoy 2001; Finnie, Lascelles and Sweetman 2005; and Looker 2001).

Although the literature clearly points to the importance of in situ and migratory effects as drivers of human capital growth across cities and rural areas, the quantitative impact of their relative roles is left largely unexplored. The purpose of this paper is to understand the geographic concentration and growth of human capital across Canada's urban-rural hierarchy. Should research focus on those factors that attract human capital to a city or region, or should the emphasis be placed on those local forces that influence the decisions of individuals to invest in education? Using data from the 1996 and 2001 Censuses, this paper tests the relative importance of net migratory flows (domestic and foreign) and in situ growth of human capital. In so doing, we aim to provide an empirical perspective on the relative importance of these two sources of human capital growth across the urban-rural spectrum.

The remainder of the paper is organized as follows. Chapter 2 provides a review of the human capital, migration and education literatures. Chapter 3 briefly reviews national trends in the share of degree holders, and Chapter 4 provides an accounting framework for measuring the components—migratory and in situ —of degree-holder growth. Chapter 5 evaluates differences in the growth of degree holders across the urban-rural hierarchy. Chapter 6 discusses the measurement and correlates of in situ growth through an analysis of degree attainment by province and place of residence. Chapter 7 concludes the paper.

1. Degree attainment is used as the measure of human capital in this paper. This is consistent with the literature in which this paper is situated, which relies on the strong association between human capital and higher levels of education. Human capital can also result from experience and other types of educational investment, which are not accounted for in this analysis.