|
Article Excerpt
Abstracts
Using custom tabulated data from the 1996 Canadian Census, this paper examines fixed interval primary, return, and onward migration in Canada at the geographic scale of 40 regions. The disaggregate level of analysis provides additional insight into migration processes in Canada. Results are largely consistent with earlier work, although onward migration becomes increasingly important as the spatial resolution is increased, and the three types of migration have different population redistribution effects than are typically observed when more aggregate spatial scales are used.
Resumes
>. Dans cet article, une anolyse est presentee de la frequence, des patrons spatiaux et les impacts de la migration primaire, de retour et d'etape au Canada pour la periode 1991-1995-1996. Des etudes anterieures indiquent que les composantes principales des flux migratoires, y compris la migration primaire (c-;a-d., la premiere migration), de retour (c-a-d., un retour a un lieu de residence anterieure) et d'etape (c-a-d., les migrations vers une destination subsequente autre que le lieu de residence precedente) differentes toutes en termes de leur intensite, leurs patrons spatiaux, leurs profils typiques et leurs effets nets de redistribution. La migration de retour represente entre 20 et 30% de tous les evenements de migration enregistres. Elle a tendance a etre plus importante vers les regions ayant une croissance elevee et a tendance a augmenter pendant des periodes de declin economique. De plus, la recherche suggere que les migrants de retour ont tendance a avoir un rapport negatif avec l'education, les competences et le revenu.
Bien qu'ayant apporte des resultats importants, plusieurs de ces etudes anterieures ont defini la migration primaire, de retour et d'etape en rapport avec le lieu de naissance (de soi-disant retour de vie) et utilise un filtre spatial relativement agrege (c-a-d., provincial) pour l'analyse. Par consequence, le detail spatial limite et une periode temporelle relativement longue pour mesurer les migrations de retour et d'etape passent a cote potentiellement les details a court terme et a petite echelle. En utilisant des donnes specialement tabulees du Recensement du Canada de 1996, cet article presente une analyse des migrations primaires, de retour et d'etape a des intervalles fixes a l'echelle geographique de 40 regions, avec les migrations etant mesurees pour des periodes fixes de 5 annees pour 1991-1995-1996. L'avantage de la definition des intervalles fixes de 5 annees incluent un intervalle plus courte pour mesurer la migration (un et 4 ans versus a vie), des mesures plus precises de la notion de region 'domicile', une plus grande fiabilite des liens entre les attributs personnels tells que l'education ou le statut dans le marche du travail et la migration, et une comprehension plus importante concernant le concept des retours 'rapides'. Tandis que la plupart des analyses ont etudie la migration inter provinciale, une analyse desagregee de 40 regions fournit une comprehension plus riche des processus de migration au Canada.
L'analyse a produit trois grandes generalisations. D'abord, les resultats appuient de facon generale les conclusions de la recherche anterieure, bien que la migration d'etape devient plus importante avec une resolution spatiale plus fine, refletant la plus grande variete de destinations potentielles. En plus, les trois types de migration possedent des effets de redistribution de population differents en comparaison aux recherches qui utilisent des echelles geographiques plus agregees. Puis, de nombreuses migrations de retour a Toronto, Montreal, ou Vancouver representaient des retours de lieus a proximite, suggerant une migration par etape. Des resultats semblables sont notes pour la migration d'etape, avec les lieus de residence en 1995 etant relativement proche. Enfin, l'analyse a une echelle spatiale plus fine demontre un portrait beaucoup plus complexe de la migration de retour et d'etape et de ses impacts spatiaux en comparaison avec d'autres etudes. Par exemple, le migration de retour a renforce souvent les effets de redistribution des migrations primaires et d'etape, tandis que pour d'autres cas aussi bien la migration de retour ou d'etape furent responsable pour le mouvement de population le plus grand. Pour terminer, des enjeux de la politique sont discutes.
Introduction
It has long been recognized that migration is a repetitive event that is based upon a complex of personal, social, and economic factors affecting each potential migrant differently. In addition, particular attention has been paid to migration events for their ability to affect the demographic and socioeconomic profile of different regions, particularly those that are economically vulnerable. Previous studies indicate, for instance, that the main components of migration flows, including primary (i.e., first time migrations), return (i.e., returns to a previous place of residence), and onward (i.e., migrations to a subsequent destination other than the previous place of residence) migrations all differ in their intensity, spatial patterns, characteristic profiles, and their net redistributive effects. Representing twenty to thirty percent of all recorded migration events, return migration tends to be highest into high-growth regions and tend to increase in importance during periods of economic decline (Newbold 2001). Furthermore, studies suggest that return migrants tend to be negatively selective in terms of education, skills, and income.
While most countries collect data on usual residence five years prior to the Census, Canada (along with Australia), collect information on the place of residence at both one and five-years prior, allowing return migration to be assessed based on a fixed (5-year) interval. Within the context of this paper, it allows return and onward migrations to be defined on the place of residence five-years prior the Census (Bell 1995, 1996; Newbold and Bell 2001) as compared to place of birth (i.e., 'lifetime returns') more commonly noted in the literature (i.e., Long 1988; Newbold and Liaw 1990, 1994; Rogers and Belanger 1990). Moreover, using fixed interval migration measures provides a number of theoretical and empirical advantages over return migrations based on the comparison of place of residence on Census day with the place of birth. These advantages include a shorter interval over which migration is measured (one and four years versus lifetime), more precise measurements of the notion of a 'home' region (Newbold and Bell 2001), greater reliability of the linkage between personal attributes such as education or labour force status and migration, and greater insight into the concept of 'rapid' returns (DaVanzo 1976; Lin et al 1999; Morrison 1971).
The purpose of this paper is to further contribute to the understanding of repeat migration in Canada, based on place of residence five years prior to the Census. Specifically, the purpose of this paper is to examine the frequency, spatial patterns, and impacts of fixed interval return and onward migrations in Canada, employing custom tabulations drawn from the 1996 Canadian Census. While findings will be related to earlier work, differences in the spatial patterns are also likely to emerge, given that these custom tabulations provide a larger sample size (based on the 20 percent of the Canadian population that completed the Census 'long form'), and additional spatial detail with regards to place of residence in 1991 and 1995. Typically, the Public Use Microdata Files (PUMF) only record the CMA of residence at the end of the Census period, with earlier residential locations constrained to the provincial scale. In our case, place of residence in...
|
