Development similarity based on proximity - a case study of urban clusters in Canada

The present analysis of urban clusters (UCs) in Canada deals with two matters of immediate interest: a) investigating the spatial autocorrelation of development levels in towns within such clusters, and b) ascertaining the physical sizes of UCs in Canada (i.e. the spatial extent of the area of strong inter-town development association). The present analysis leads to three general conclusions: •First, development levels of neighbouring towns in UCs of Canada tend to be closely associated, though the intensity of such a development association generally tends to decline as inter-town distances increase. As argued, this spatial association of development rates may be due to the fact that both private investors and migrants consider UCs as integrated functional units, and make their location decisions hierarchically: first, among or between town clusters, and then among or between individual towns in a 'preferred' cluster. •Second, the effect of clustering on urban growth is not uniform. It is stronger in peripheral UCs (specifically in respect to unemployment and income variables), while in centrally-located ones the development levels of neighbouring towns are less interdependent. In general, distances within which inter-town development linkages are sufficiently strong to affect or promote clustering vary with the range practicable for daily commuting, that is, from 20-40 km in the country's core and 60-100 km in its periphery. •Third, the effect of spatial proximity of towns on their functional linkages differs in respect to different development measures. In particular, as found from our analysis of Canada's core areas, only population and housing variables exhibit strong spatial associations, while the effect of spatial factors on employment-related variables – average income and unemployment rate – is weaker. This dissimilarity represents fundamental differences between these two groups of variables. That is, while population and housing variables may be confidently associated with the clustering of residents in socially homogenous areas, the spatial association of employment-related variables may be influenced by inter-urban commuting. Thus, low unemployment in a town may reflect the availability of employment in the larger region rather in the town itself, which is an important caution about the care that needs to be taken in correctly selecting and interpreting indicators of urban functionality and growth potential. An important strategic finding of the present investigation is that local towns appear to follow the path of the central city over time, and local towns adjacent to a wealthy city are likely to perform better than those around a less-prosperous central locality. This result indicates that urban growth may spread across individual towns in both core and peripheral UCs, which has implications for urban and regional development policies and programs at the municipal, provincial and federal levels of government. In particular, the findings of the present analysis thus support the creation and stimulation of UCs in areas where further urban growth is desired. According to this strategy, development resources should be concentrated on selected UCs until they become sufficiently attractive to migrants and private developers. Support of the selected localities should, of course, include a balanced investment in both the housing development and employment-generating sectors. In addition to direct government intervention, various forms of indirect involvement, such as incentives for private investors and tax exemptions can be applied. Then, and based on evidence derived from the application of impact assessment procedures, as soon as the growth of the selected UCs becomes sustainable support may be redirected to other UCs. This hierarchical concentration of resources can then be shifted into more remote areas.