Creative workers’ segregation and why city-level rankings annoy me

Richard Florida ran a series on the Atlantic Cities on various types of economic segregation in cities, and while it’s a good idea, I don’t really understand what he’s doing, and ranking indexes makes my eyebrows do that pinched thing they do when I am not happy. They find that LA is #1 in segregation of creatives. And while Florida (thankfully) doesn’t go there, somebody will: this is MORE PROOF that sprawl is bad. Because LA = must be sprawled = must be bad.

I will spare you the theory about why economic segregation makes sense, and that there are few reasons to a priori assume that economic segregation is a problem. Suffice it to say that there are good reasons for industries and their employees to cluster.

So here’s the post on creative class segregation. They use a dissimilarity index. DIs tend to be very good at showing segregation between two groups: they were designed back in the day to show residential segregation between binary race categories: white and black. We find with DIs that aggregating them causes them to behave oddly, as does having multiple categories. Those things can be dealt with, and it may be that Florida and his coauthors have done so.

Still, I am looking at the actual index numbers with my head tilted and thinking: so what? LA’s number #1 here, and the DI is 0.344, with 1 being perfect segregation. What does 0.344 even indicate? The range in the DI for the top 10 goes from 0.344 to 0.281. When we are looking at that small a range, exploding those differences by setting up rankings to represent what are, in fact, differences in hundredths in an index grossly distorts the concepts being measured. The income segregation numbers, shown in one of Florida’s prior columns, go up to 0.70. THAT is enough to get my attention. But while the map is nice, all I see here is meh: places that still have some hard industry left and have creatives come up an intense navy blue, one of my favorite colors no doubt, but awfully intense-looking for what are, in reality, tiny segregation numbers.

Can anybody who studies economic geography tell me why the rest of us should care about these DIs? Are they good news? Bad news? Because I kind of think they are “whatever” news.

Howlers in Vishaan Chakrabarti’s NYT op-ed on cities

In general, I’m all for planners and architects and their “Cities are awesome with awesomesauce” messaging. Cities are wonderful. Cities are sexy! Cities are infinitely many more times better than suburbs! Suburbs are just full of fat jobless losers bringing all the cool people down and ruining the environment. Suburbs, in fact, are ruining America.

Ok? Now that I’ve joined the zeitgeist and proven that I’m not one of those Joel Kotkin types that all the cool urban kids revile, can we talk about numbers?

Vishaan Chakrabarti is an associate professor of architecture at Columbia, so I don’t feel bad picking on him, but several of my students have circulated his NYT Op-Ed and it makes me rather sad. His recent Op-Ed notes that cities are growing (most are) and now full of those fab millenials, families, and retirees alike. There are many howlers in the article, but this one is a serious problem:

A staggering 90 percent of our gross domestic product and 86 percent of our jobs are generated in 3 percent of the continental United States, namely our cities.

Where did these numbers come from? The Bureau of Economic Analysis releases GDP numbers by metro region, not by incorporated areas (cities) as far as I know. Where did he get this number? I have no idea because he doesn’t tell us.

I’m pretty sure that’s not his research, and it’s not ok to throw around those numbers without telling us who actually produced the information, even in an Op-Ed. I think it’s a misquote of a Brookings report that came out of a few years ago or from Ed Glaeser’s book, but I could be wrong. You say “According to a recent report from researchers at….” and move to your point. Even if that is Chakrabarti’s own research, his book, or his team’s research, he should tell us where the info comes from.

It IS true that MSAs are the engines. You can download the data from BEA and show that: Total US GDP is 15,566,077 (millions of chained $$$$2005); the amount of that which comes from MSAs is $12,206,566 (millions of chained $$$2005). So that’s some nice trillions there. But that percentage is about 80 percent, rounding up. Cities are part of metro regions, so how cities can be producing 90 percent of GDP while regions are producing only 80 percent is confusing. I don’t get it. Am I missing something? I am using real GDP adjusted for inflation. Anybody seen the 90 percent number?

I’m too lazy to look up the jobs numbers, but I suspect there is a similar problem there (e.g. conflating metro regional numbers with “cities”.) But at least you could get employment numbers by zip code rather than by MSA.

But I have an intellectual problem of traducing suburbs and then incorporating their portion of GDP into the argument for why cities are so much better than suburbs.

Chakrabarti throws around a lot of assertions about what millennials want, how cities are centers of support for marriage equality, etc etc….and I’m pretty sure that information comes from Pew and other sources. Given how poorly Chakrabarti presents the economic growth numbers, I’m not buying these assertions either.

Howlers and failing to discuss his sources notwithstanding, Chakrabarti wants to know why are we subsidizing those terrible suburbs?

Because most voters live in suburbs. That’s why:

Total population 2010 by incorporated area to MSA

Screenshot 4 21 14 3 09 AM

And while cities are growing, suburbs are growing, too, and in many instances, at a much faster rate than central cities, even still:

Share of metropolitan growth 2000 to 2010

Screenshot 4 21 14 3 07 AM

I’m not sure I handled the Chicago pop loss properly in these calculations (It’s 3:10 am, by my excuse, and I’d welcome somebody to check the numbers), but you get the idea here. This is all metro area growth, and while most cities did grow–and Chakrabarti is right, that’s good news–their suburbs did as well or better than they did.

Now, we can argue causation. Subsidies might prompt people to move to suburbs, which might explain these numbers. Or it might explain why it’s so difficult to shift those subsidies. Or it might suggest a political economy of people who get the subsidies they want. But if we are just counting voters, there’s your answer. It also doesn’t help that downtown voters’ participation rates are lower than those in suburbs, generally. (See Jeff Seller’s new book).

I dunno. In general, urban op-eds make me happy. And in a world where people seem to believe whatever the hell they want to believe (climate change is university conspiracy designed to control us; Obama is a Muslim after our guns; General Motors ruined our jolly streetcars),the sloppy thinking in this Op-Ed maybe just advances the cause we all have in helping people understand that cities are important, because cities actually are important, and not unrelatedly, those of us who study cities want people to realize how important cities are and, by extension, how important we ourselves are, for being big-a-time experty-experts on them.

The other part me of just thinks this type of stuff makes urbanists look innumerate or like con artists. Check the numbers. Then double-check them. Then explain why, in polycentric post-city metro regions, we need cities to stand out. There’s an argument to be made for strong, amenity-rich centers even in major metro regions, and it has been made. Just not here.