2 Review of the literature

View the most recent version.

We start with the assertion that large cities are growing faster in terms of population and employment than are smaller cities and rural areas, with their citizens earning higher incomes and experiencing greater income growth than residents of smaller urban and rural areas (Beckstead and Brown 2006, Polèse and Shearmur 2005). Underlying the growing economic power of cities and their populations are a series of forces, which include changing technologies and falling barriers to trade that have tended to increase the relative demand for highly skilled labour (Yan 2005); rising incomes that enable greater consumption of goods and services (Sanghoon 2005); and, immigration as a source of labour and growth. In combination, these forces may bias growth away from smaller urban or rural areas toward large cities.

Fundamental to the growth of a region is its stock of human capital (see Glaeser 2000). Education and experience are the primary contributors to the accumulation of human capital. From a purely economic perspective, people will invest in an extra year of schooling if they expect that the net present value of their additional earnings from this investment will exceed the forgone income and the out-of-pocket costs (e.g., tuition) of the extra year of education (Fortin and Lemieux 2006). Hence, there are two essential, and countervailing, variables that drive this decision-making process: present value of additional expected earnings; and, costs (opportunity and out-of-pocket). These will tend to vary by location and, as a result, may influence the individual's decision to both invest in more education and to migrate to locations where the returns to such an investment are highest.

Although education and migration are not necessarily exclusionary decisions, ² we can conceptually define the growth of human capital in a region as the sum of two components. The first is in situ growth, whereby growth in human capital is derived from the investments of individuals already situated in a region. The second is through the net effects of migration. That is, human capital can be grown 'within,' or it can migrate to a region as individuals search for new education or employment. Indeed, the intention to enrol in postsecondary can be seen as a human capital issue in which individuals will evaluate their discounted present value oflifetime earnings, net of education costs (Bouchard and Zhao 2000, Fortin and Lemieux 2006). This implies that enrolment will be higher in places with higher returns to a university education. As such, the relative growth rate of human capital across cities and rural regions depends on two factors: decisions of individuals to invest in additional years of education; and, net migratory flows.

Nationally, there is a clear trend toward increased university participation. Full-time university enrolment increased by 29.2% from 1992/93 to 2003/04, although part-time university enrolment declined by 19.4% during the same period. Following the relative expansion of university participation rates over the 1980s, rates stagnated in the 1990s (Fortin 2005), holding steady at approximately 17.5% through much of the 1990s among the 18-to-24-year olds (CAUT 2006). Following 2000, participation rates among the same age group increased to 20.7% by 2003/04.

Despite increased university participation rates since the 1970s (CAUT 2006; Christofides, Cirello and Hoy 2001), several studies have suggested the importance of location relative to degree attainment. Using U.S. data, Blackwell and McLaughlin (1998) note that rural youth aspire to fewer years of education than their urban counterparts, and they typically had lower educational attainment than youth in urban areas. Several authors have recorded similar findings for Canada. Proportionally lower numbers of rural youth complete postsecondary degrees and more drop out of high school. Rural youth tend to aspire to and move into so-called lower-status occupations (fewer professional occupations and more skilled and semi-skilled), they are more likely to report bouts of unemployment, and they are less likely to have full-time, full-year employment (Andres and Looker 2001; Bouchard and Zhao 2000; Christofides, Cirello and Hoy 2001; Finnie, Lascelles and Sweetman 2005; and Looker 2001). Regional differences in university participation are also noted in the literature, with lower participation rates in Quebec, Alberta and Prince Edward Island (Finnie, Laporte and Lascelles 2004).

Historically, high levels of education attainment in large urban areas were seen to result from a higher share of parents with postsecondary education residing in these areas, creating 'built-in' path dependence. However, the presence of a local university also increases participation by increasing the availability of student places and decreasing costs (Frenette 2007). Frenette (2007) noted that the creation of a local university increased local enrolment by 28%, with low-income families benefiting the most, given the expense associated with long-distance residency. Such findings are similar to Card (1993), who found that men who grew up in local labour markets with a nearby college had higher education and earnings than others, with education gains concentrated among those with poorly educated parents—men who would otherwise stop schooling earlier. For those rural youth that do leave their communities to pursue postsecondary education, only a small proportion return, reflective of more restrictive employment or occupational opportunities in their home communities (Looker 2001). Among those that do return, their return makes little difference to the level of human capital in rural areas.

While the financial cost of obtaining a degree may be greater for rural youth, and particularly for those that do not live within commuting distance of a university (Frenette 2007), differences in educational attainment over space are consistently and strongly associated with family advantage (Blackwell and McLaughin 1998). Higher participation rates among youth with parents from high-income families or with high education are consistently noted in the literature, with university participation rates rising as income increases (Drolet 2005; Finnie, Laporte and Lascelles 2004; Finnie, Lascelles and Sweetman 2005; and Looker 2001). Further, Drolet (2005) reports participation rates are more strongly associated with parental education than with income. That is, higher educational levels among parents are associated with increased participation in postsecondary education and a greater likelihood of degree attainment.

Migration may also contribute to the growth of human capital across regions. Specifically, large urban areas potentially have higher shares of degree holders, not just because of higher rates of degree attainment among people who spent their formative years there, but because of net migratory flows. In fact, the migration literature has long-recognized the role of human capital as a motivator for inter-regional migration. Sjaastad's (1962) human capital theory defines migration as an investment in human capital measured by expected future income, balancing the costs of migration against future expected returns measured by lifetime earnings (see also Milne 1991). If benefits exceed costs, then the individual will migrate to the destination that offers the greatest returns. Both benefits and costs could be monetary (e.g., the dollar cost of moving) or psychic (e.g., the psychological costs of moving away from family and friends), meaning that migration is not exogenously determined.

As such, the human capital theory of migration offers a framework within which both income and non-wage effects enter individual migration decisions, with returns to human capital encouraging migration. Individuals are more likely to select destinations with higher income and employment growth rates (Newbold and Liaw 1994, Shaw 1985), while non-wage components capture the amenities and social or cultural attractions of regions. Moreover, the theory explains why migration rates tend to decline with age, with younger individuals having longer periods within which to capture the benefits (expected income) of migration than their older counterparts.

Following human capital theory, large urban centres are often selected because the returns to education are higher. These higher returns can be related to the more efficient matching between workers skills and firm's needs in larger labour markets, raising the productivity of workers and resulting in higher wages (Helsley and Strange 1990, Kim 1989). These higher returns may also relate to the greater demand for specialized skills in larger urban centres. In part, this may be driven by the labour matching effect and the fact that larger markets allow firms to specialize (Kim 1989). To these more standard economic arguments we can add several others that may also play an important role.

• Power couples: Increasingly, degree holders are marrying degree holders, leading to the need to satisfy the career ambitions of both. One solution is movement to a larger labour market. In so doing, the couple satisfy their career ambitions and maximize family income at the same time (Costa and Kahn 2000).
• City lights effect: Persons with more education will tend to earn higher wages. If cities provide more consumption opportunities, highly skilled/highly paid workers may be drawn to these centres.
• Florida argument: Certain larger cities may be very attractive because they are viewed by the so-called 'creative class' (often defined as degree holders) as being places that are open and welcoming to persons with diverse lifestyles (Florida 2002a, 2002b).

Similar factors also hold true for both new immigrants arriving in Canada and for the foreign born making inter-regional migrations within Canada. The foreign born demonstrate a clear preference for residing in or moving to urban areas. Moreover, the presence of ethnic enclaves and communities, particularly in Canada's main immigrant-receiving cities of Toronto, Montréal and Vancouver, reinforces this attraction while aiding the economic integration of new arrivals (Kobrin and Spear 1983, Newbold 1996).

Although the factors associated with educational attainment in Canada are similar for both domestic- and immigrant-born youth, that is, parental education level, income and occupation (Kao and Tienda 1995; Richmond 1986; Richmond and Kalbach 1980; Vernez and Abrahamse 1996; and White and Glick 2000), immigrants resident in Canada tend to enjoy higher educational attainments than the national average (Hou and Balakrishnan 1996). Compared with native-born youth with a similar socioeconomic status and attendance in a public school in the same neighbourhoods, immigrant youth are also more likely to enrol in postsecondary education and to attend college continuously for four years (Vernez and Abrahamse 1996).

Consequently, the literature clearly points to the importance of both in situ and migratory effects as drivers of human capital growth across cities and rural areas. Left largely unexplored is their relative roles and contribution, a question that this paper explores.

2 . Individuals from rural areas will often need to migrate in order to complete a postsecondary education and migration often accompanies the search for post-graduation employment.