We based our initial analyses on the principle that concerns for internal validity ought to precede those for external validity. Without first gaining a confident understanding of the relationships between hypothesized independent and dependent variables, issues of where they may be generalizable to are probably moot.

Consequently we initially focused on metropolitan areas (that also provide some external validation in North America). Though we recognize that the ultimate development of coherent systematic evidence for sound policy decisions will require similar study in Canada and America’s diverse places.