When I refer to cities, I am talking about more than just the central-city municipalities that carry the name of their city-region. For example, when people think of Toronto in its global context, they do not normally think only of the jurisdiction of the municipal corporation of the city of Toronto—they think of the wider, growing city-region that includes surrounding municipalities like Mississauga.

While examples of populous city-regions comprised of one municipality certainly exist, our maps are typically dotted with city-regions composed of dozens of municipalities.

And that is the dilemma. If we make a central-city municipality—and perhaps also its surrounding suburbs—more autonomous, this only reifies the existing boundaries. These boundaries are invariably arbitrary, outdated and irrelevant—and will become more so all the time. So, for example, what we normally think of as the city of Montreal or Vancouver could not be self-governing because there is no institution to make the central-city municipality and its surrounding suburban municipalities self-governing together.

In order to focus on what is really important to us—the economic and social well-being of a city—we must focus on the city-region as a whole. And determining territorial boundaries for that purpose is not simple.

Urban analysts need to pay greater attention to the problems posed by boundaries. At the level of the sovereign state, boundaries can be and are accepted as constant, except in the most unusual of circumstances. In urban government, boundaries are always contested, except in the most unusual of circumstances.

This basic fact dramatically limits the potential of urban governments to replace other governments as the main decision-making authorities for city-regions. And these, no doubt, are issues that Canadians will be grappling with in cities from coast to coast for many years to come.

I think that the term “self-governing” warrants some explanation. It obviously would include any scheme that establishes a city as a sovereign state with membership in the United Nations. Singapore is frequently cited as an example of a self-governing city-state, but as an island its boundaries are easier to draw, making its historical and geographic experience difficult for other cities to replicate. Moreover, even Singapore is not without its problems as a model for a self-governing city (for instance, the city has expanded economically into neighboring Malaysia and Indonesia).

A self-governing city could also be a constituent unit of a federation. Almost by definition, the sub-national jurisdictions within federal states possess considerable autonomous authority over a wide range of functions that are especially important to the governance of cities. Hamburg and Berlin—each a Land within the Federal Republic of Germany—are classic and oft-cited examples of city-states within a federal system. Brussels and Madrid are more recently established examples. Yet as these cities undergo growth and transformation, establishing boundaries regularly poses significant problems in their governance.

There is sometimes a focus on what is often referred to as “metropolitan government”—that is, a structure superimposed within a city-region on local governments, and designed to provide a wide range of common local-government services and regulatory functions for the entire area. Some suggest that metropolitan government should form a distinct level of government, along with national, provincial and local governments.

This seems impractical as a general rule. Although various cooperative arrangements among municipalities within cities are obviously desirable and necessary, these work better when they have a governance structure and a focus on one function in particular, such as public transit.

Multi-functional metropolitan governments are not feasible for the world’s largest and most important city-regions because boundaries will vary by function. For example, including Hamilton with Toronto for regional public transit (Metrolinx) makes sense, but appropriate boundaries for police and social services are likely to be quite different.

The point is that city-regions cannot be self-governing entities—except in rare instances—as either sovereign states, units of federations, or even as multi-functional metropolitan governments.

I am sympathetic to the issues raised by city advocates who are concerned about the oppressive regulatory regimes imposed by higher orders of government. However, inflated rhetoric about more autonomy for cities as a remedy to the challenges of unresponsive higher orders of government is ultimately unhelpful. We should have a conversation about how to improve cities in a complex federation like Canada, without the distraction of debates on municipal self-government.

To find out more on the problems of boundaries and their relation to city governance and public policy, check out Andrew’s latest and most provocative book, “The Limits of Boundaries: Why City Regions Cannot be Self-governing” (Montreal and Kingston: McGill-Queen’s University Press, 2008).