Why You Should Be Skeptical of Statistics on City vs. Suburban Population Growth
A few weeks ago the U.S. Census Bureau released new population data that gave urbanists reason to cheer. The figures showed faster growth rates in city centers than in suburban areas for 27 of the 51 largest metro areas in the United States between July 2010 and July 2011. (Our Nate Berg mapped the changes here.) It looked like the first time cities had outgrown their suburbs in many decades, prompting the Wall Street Journal to declare "shifting attitudes about urban living."
Maybe the times are a-changing, maybe they're not, says Columbia University planning professor David King. He took a closer look at the numbers and determined it was too quick to call the contest. Writing at his blog, Getting From There to Here, King explained that growth rates tell only part of the story. While central cities won the relative population match-up, they got clobbered in the absolute one.
Take the shifts in metropolitan Atlanta. According to the Census, the downtown area grew at 2.4 percent while the suburbs grew at only 1.3 percent — a clear relative gain for the city. But the suburbs are much more populated to begin with, King reminds us. That means only 10,135 more people settled in the city, while 63,226 more settled in suburbia. In absolute terms, just 14 percent of metro Atlanta growth occurred downtown.
King's reservations may dampen an urbanist spirit or two, but his point isn't that cities aren't catching up to suburbs — just that these particular numbers don't declare a victor. Unlike many writers who have taken the suburban side in this ongoing debate, King comes at things with an impartial hope that stronger evidence will inform better planning. We asked him to expand on his points a bit for Atlantic Cities readers.
"There's a lot of people who like to live in cities — I certainly count myself as one of them," King says. "I would like to see more discussion on: are we happy with the magnitude of the effects we're seeing."
Why is the absolute number more important here than the growth rate?
Most of what was actually reported was essentially that the central cities and the suburbs are growing at about the same rate. That seems reasonable. But again, in absolute terms, because we've been suburbanizing for so long, it's really is a very small effect on the overall city. When we're talking about the success stories of downtown revivals, we're talking about a couple thousand people. We're just not talking about big numbers. Not to say this is bad but it's certainly not enough to say that there's been a sea change — a substantial change in the way people approach cities.
Considering how long it's been since cities outgrew suburbs, do you think these shifts are significant enough to report, even taking into account some of these methodological problems?
I think it's good to report on trends as they're coming. I do worry that we often will lose the forest from the trees. If you think back to what would be, broadly speaking, U.S. urban policy for the last 40 years. In that time, we've certainly spent a lot of money on roads, but we've spent a lot of money on transit and we've spent a lot of money on economic development in cities. We've spent a lot of money on trying to revitalize cities. I think it's reasonable for us to feel good that we at least seem to have arrested the decline of central cities. That I think is an unambiguous good. But we have to look at how much it costs us to do that. Are we getting the results we want?
As you point out, there are many metro areas on the Journal list with small growth-rate differences — like .1 percent — that might not be statistically significant. What does that mean for the broader conclusion?
Let's say all those .1s or .2s are not statistically significant. Would you still say it's important that we've arrested the decline, or would you say it's not worth talking about? I think it's worth talking about. One thing that I think needs some rigorous study is, there's no obvious pattern that you see. Are these cities that are growing faster than the suburbs the ones we would expect to be growing faster than the suburbs because we've been doing things like investing in transit? Because we've been doing things like economic development in the central city? Because we've been relaxing the zoning code? Whatever it might be. I think there's a lot we can learn, but just looking through the table the Wall Street Journal provided, there's no clear sense that one particular set of policies or one broad trend is showing up.
It strikes me that this is the type of debate that could continue forever.
One of the reasons it's frustrating that you just hear about city versus suburbs is there's so much heterogeneity of suburbs, that it's really not fair to treat all suburbs as the same. Some suburbs are dense. Some are old streetcar suburbs. Some have been trying, through transit investment and investment in main streets downtown, to create walkable denser communities. This has been happening throughout the country.
What we need to do is stop looking at these crude city-versus-suburb divides and we need to start looking at where is the growth actually happening. Is it happening in places where we'd expect — are people voting with their feet, so to speak, to go to these denser places? Is it actually the case that people want more urban existence? I think we'd probably find evidence that's the case. That does not mean the overwhelming majority of Americans want that. One of the issues is how big of a market is there for the urbanist ideal. We just don't know.
Despite your concerns about the latest Census figures, you seem hopeful that there's a shift to the city.
I like cities. I would like to see more people enjoy them. Certainly I hope we can say definitively that these things are happening. I worry that if we say definitely these things are happening, when in 5, 10, 15, 20 years we've continued to decentralize, we haven't wasted a bunch of time trying to come up with solutions to many of the problems we think are the core issues. Such as what do we do about the environment? What do we do about equity? How do we ensure that people have access to jobs and health care and education?
Is the fear that there's something we'd do given this data, and the early conclusions, that could spoil things in the long run?
This all gets very speculative. If these trends are actually happening, then we should put them into high-quality research and actually make a case. Looking at the global warming context, having solid science doesn't necessarily help anything. But certainly it would go a long way towards having strong evidence that we do need to change the zoning code or we do need to focus development efforts here or we do need to rethink how we're doing transit. That's one of the reasons we should think about the magnitude of these effects that are going on.