You may have read something about how, according to Census data from 2011, cities in the US are growing at a faster rate than the suburbs. Kaid Benfield‘s post on the subject is pretty representative of what popped up on dozens of urbanist blogs.
But hold on a minute. According to a number of other writers – including Aaron Renn at Urbanophile and Chris Briem – the data support no such conclusion. Renn cites a map of population change in metro areas across the country that is color-coded to reflect population change in the core cities of those metros. The majority of metros reflect a decline in population in their core cities; Renn argues that because population in most “core cities” didn’t actually increase, suburbanization is still increasing. Most Americans still prefer “detached, single-family housing,” he argues in comments – and the best place to find that is in suburbia.
It sounds logical enough on the surface. But I’ve got some issues with that analysis.
It starts with the terminology. The map Renn cites defines “cores” as cities with over 100,000 people. Here in the Tampa Bay area, we have three of those: Tampa, St. Petersburg and Clearwater. All of these cities are fairly large, area-wise, and all have some neighborhoods with urban characteristics, as well as some that would be more recognizable as suburban.
So if Tampa, for example, has both urban-style neighborhoods (multifamily housing, denser population, garages facing alleys) and suburban-style neighborhoods (single-family housing, garages fronting streets, fewer or no sidewalks), what are we to make of the fact that its population increased? We can’t tell anything, at least not about whether people are revealing a preference for urban or suburban living, unless we know which specific neighborhoods gained residents.
It’s quite easy to conflate the ideas of a “core city” and the urban core. A core city includes the entire city, as long as it meets a certain minimum population. The urban core, on the other hand, is generally considered to be the most recognizably urban part of a city. It is by definition smaller than the core city; it is also the best part of the city to examine if we want to understand how population changes reflect Americans’ preferences for urban vs. suburban living.
You have to really know a city to know where its most urban neighborhoods are. And if you want to make claims about Americans’ preferences for urban vs suburban lifestyles, you’ve got to have this information for pretty much all cities with more than 100,000 people. You might be able to figure it out by mining Census data, but it wouldn’t be easy. Certainly not as easy as that map implies.
Renn at the Urbanophile says that the fact that these “core cities” do not show significant growth demonstrates that the urban renaissance in this country has been overstated, and that people do and will continue to prefer the detached single-family housing options that are more readily available in the suburbs. I’m willing to entertain the first suggestion – that Americans’ enthusiasm for cities isn’t really as strong as we’ve come to imagine it to be – but the second just doesn’t follow. In fact, in most of the cities I’ve been to, detached single-family housing is by far the most plentiful option.
This is why I am more and more convinced that the Atlas of Suburbanism is a good idea. The suburbs don’t start at the city limits, at least not in any meaningful way – suburbia is a way of life more than a neighborhood or an address.
But all this bickering may very well be meaningless, at least in this specific case. Briem argues that the Census data tells us nothing new, because there wasn’t actually any real new data released. Still, it’s a case study in not drawing overly broad conclusions from a too-thin data set. Be sure you know what your data is actually telling you before you hit the “post” button.