Race, roads, or regulation? Canadian and American cities

 When asked why Canadian cities are so built-up compared to American ones, Martin Wisse of the Netherlands suggests, “No white flight.” It’s an appealing answer, given America’s massive urban ghettoes, but it’s almost certainly wrong.

 

Consider the differences between Seattle and Vancouver. For an American city, Seattle is very white and disproportionately childless. The city, however, splatters across the horizon, with few high-rises outside the central business district, while Vancouver is relatively compact and its proliferating residential towers reach to the sky. All this despite geography in Seattle that should encourage density!

The Seattle-Vancouver comparison alone should make you doubt the validity of the “white flight” hypothesis. There are other reasons, however, to reject it. Until 1990 none of Canada’s metropolitan areas declined in population — the country didn’t see shifts in economic geography like what turned cities like Buffalo, Detroit, New Orleans, Flint, Akron, Pittsburgh, Dayton, and Syracuse into what they are today. (The 1990s recession changed that, but the places that shrank tended to small, and located in Atlantic Canada or northern Ontario and Quebec.)

More importantly, white flight can’t explain why the nonblack sections of American cities look rather different from their Canadian counterparts. Nor can white flight (by itself) explain why American suburbs strongly resist residential high-rises, whereas Canadian ones (Montreal excepted) take to them with relative zeal. For example, Chicago’s North Side does not look like Toronto, save for a belt along the lake. Nor does Boston’s central and northern urban area — Back Bay, South End, North End, Southie, Brookline, Allston, Brighton, Cambridge, Somerville, Chelsea, Watertown, Arlington, Belmont, Charlestown, and East Boston — look much like East York or Scarborough. It is true, of course, that the great American ghetto is an American phenomenon, created by America’s strange racial caste system, but racial tensions don’t explain why Houston looks very different from Calgary.

So if isn’t race, then what is it? Another possible culprit is roads. American cities have a lot more freeway-miles per capita than their Canadian counterparts. The logic is not that freeways destroyed the neighborhoods that they cut through. I’m sorry, but speaking as someone who grew up in neighborhoods that had roads driven through or beside them, that argument is just stupid. (Yes, this applies to Toronto.) The plausible logic, rather, is that inner city neighborhoods were always relatively undesirable, but without subsidized freeways it would have harder to leave them for lower-density suburbs.

I am sympathetic to this argument, but I have some doubts. (Not least that my wife and I prefer dense inner city neighborhoods to the suburbs.) It is almost certainly true that American-style sprawl requires massive investment in freeways. Fewer freeways, more density. More density, more high-rises. The problem? If density the key, why does Los Angeles look so different from Toronto despite being denser? The 11.8 million people who live in Los Angeles’s urbanized area (that is, the area defined by the lights visible from space at night, minus fringe areas with a population density below 1,000 people per mile) are packed in at 7,068 per square mile, whereas their 4.6 million Torontonian counterparts live at 6,835 per square mile. Perhaps it is the freeway net that allows Los Angeles to splatter its density more uniformly than Toronto, but that feels ad hoc as an explanation, especially since Toronto’s suburbs are actually very densely-populated compared to Boston, Chicago, or New York.

The other possible explanation is regulation. In this view, roads may have contributed to American sprawl, but zoning is what really keeps it in place. For this view to be correct, Canadian zoning has to be less strict than American zoning. Are Canadian cities more likely to approve density, allow smaller lots and narrower streets, require fewer parking spaces, and permit higher buildings? This is not hard to believe, considering the far greater control that provincial governments exercise over municipalities in Canada, but I do not know the answer.

 

In point of fact, the “roads” and “regulations” hypotheses may be hard to disentangle. If a lack of roads drives up land values (by limiting the effective supply of land), then there will be more political pressure to allow dense development. After all, people need places to live. On the other hand, if it is easy to get out to the burbs (via all those freeways) then there will be less political pressure on inner-city areas to permit taller buildings. Cambridge, for example, saw a high-rise boom in the 1970s, which the city deliberately stepped on. As a result, the Central Square subway stop (where we live) is stupidly surrounded by two-story buildings instead of 20-story towers. Our building was built in the early 1990s, and was effectively capped at five floors. Without Boston’s two ring roads (I-95 aka “Route 128” and I-495), the pressure to build up in places like Cambridge would have been harder to resist.

 

In fact, however, it is possible to disentangle between “roads” and “regulations,” by testing for the effect of zoning changes on land values. Of course, it’s never that easy — endogeneity is always a problem. Ed Glaeser has a paper suggesting that zoning is determinative in shaping American cities in California and the East Coast north of Virginia. He has a particularly-detailed study of Boston that supports the zoning hypothesis: market forces (given the existing road net) would make Greater Boston rather denser were the regulations less restrictive. (Newton would look like Brookline; Brookline would look like Brooklyn; and Boston itself would be filled with high-rise towers.) On the other hand, his same work implies that in most of America it is the road net what produced sprawl, and not race or land-use regulation.

In short, most American cities sprawl because the federal government built roads out to the horizon. The exceptions are along the coasts, where U.S. cities would look more like Canadian ones in the absence of tight zoning laws. The increasing segregation of immigrants appears to be an effect of the above, rather than a cause. Of course, the great American ghetto is a striking and horrible thing, but we would have had them with or without freeways and zoning.

Comments

[bodzin]

And then the question is, why the tighter (more conservative) zoning in the U.S. than in Canada?

Mike Lewyn says http://www.planetizen.com/node/43255 the difference is from lower crime and pricier fuel in Canada.

Pat Condon, who happens to live in BC, says http://www.planetizen.com/node/132 it's all about freeways.

Having lived much of my life in Canada, more of it in the USA and now living in Santiago, the difference I see between the USA and Canada is mostly that Canada has a more centralized government. This means that urban schools are the same as rural and suburban ones, reducing urban flight and disinvestment. Basically, there's no reason NOT to live in the city. It also means Canadians may be better at coming up with long-term regional plans and sticking to them, rather than making a bunch of ad-hoc decisions, neighborhood by neighborhood. While everyone wants to encourage democracy and public input, it has gone way overboard in the USA, to the point of ballot referendums on individual development projects in places like California and Colorado.

PS I bet you didn't know I was also a land-use guy. I used to work at the Congress for the New Urbanism, where they study this stuff a lot.

[Noel Maurer]

I did not know that!

Glaeser's work suggests that Condon is correct overall, but Lewyn has a point for the big coastal cities in the Northeast and California. Ironically, he's probably wrong about Atlanta, the city he uses as an example of zoning-gone-mad. Greater Washington (which Matt Yglesias likes to go on about) would be a much better one, I think.

I completely agree with you about the way community input has gone completely bonkers in most of the U.S. of A.

[bph]

My knee jerk is that the way zoning works in the US does not make the developer pay the full cost of the new development. Instead, municipalities eager to increase the tax base subsidize cheap housing to get the additional housing. Because American property taxes are based on the value of the land + structure, not on the cost of the property to the municipality, it is in the interest of local governments to have a lot of big sprawling estates instead of small compact housing. The latter are just as expensive in schools, police, fire, water and sewer, while generate less taxes for the authorities that provide those services.

I know exactly where that pipe shop and pasta place are. A good friend of mine now lives in a converted factory 4 blocks away, speaking of land use (and 8 blocks away you can see NHL games, just to include some Canadian content).