Friday, July 25, 2008

Racial demographics of CNN/Money's 25 best cities

++Addition++The racial percentages have been adjusted slightly thanks to astute commentary. Please see the explanation at the end of this post.


CNN and Money magazine just announced the 100 best small cities in the US to live in. With populations of 50,000-300,000, most cities in major metropolitan areas that border the major city that anchors the broader MSA are up for consideration. One of the factors that figures into the rankings is a "racial diversity index". Higher scores indicate more racial diversity, or more bluntly, a proportionately smaller white population.

The reason I point this out is that the other measurable variables clearly add or detract from the value of living in an area. You'd expect a higher percentage of days free of cloud cover to, ceteris paribus, make a place more desirable than a lower percentage would. Conversely, you'd expect longer average commute times to make a place less desirable than it would be if commutes were shorter. But do most people scouting out a new community to move to feel that greater racial diversity is a benefit that will influence their decisions, as is assumed by CNN/Money?

Here is the racial composition of the top 25 (nationally), excluding Hunter Mill and Sully, both in VA, for which I couldn't get information on (the demographics are similar across all 100, so just looking at the top quarter is a personal time saver). I estimated that 61% of Hispanics choose "white" as their race and came up with the non-Hispanic white percentage from that:

Race% best (national)
White76.2 (66.4)
Black5.7 (12.8)
Hispanic6.5 (14.8)
Asian8.8 (4.4)
Nat Am0.5 (1.0)
Pacific Islander0.0 (0.2)
Other2.5 (?)
2+ races2.3 (1.6)

Yeah, the cities are whiter than the nation as a whole, but not really by that much. Of course, rural areas are whiter than more settled areas are, including the suburbs, so the white population is at a 'disadvantage' at the city level compared to state-level comparisons.

To answer the previously stated question, the short answer is no. Should a factor that is inversely correlated with "best living" in the minds of most people be embedded in the "best living" score? Doesn't this just distort the value provided by factors that actually measure desirable qualities like average scholastic test scores compared to those of the state as a whole or the percentage of the population that is married?

But to the extent that racial diversity does factor in, Asians are the preferred filler. The Asian number is even more impressive because the cities tend toward the center of the country due to cost-of-living and residence affordability.

* The Census doesn't breakdown non-Hispanic whites at the city level, but Peter noted in the comments that my previous estimate of 85% of Hispanics considering themselves white may have been too high. His intuition is correct, according to this article from the NYT. Half of Hispanics chose "white" as their race in the 2000 Census, with a little over 40% choosing "other" (only 2% chose "black"). Further, virtually all of the "other" category is comprised of Hispanics (97%). So I counted the entire 2.5% "other" as Hispanic, and the additional 4.0% of the total as white Hispanic to arrive at the non-Hispanic white percentage (listed simply as "white" in the table).


Sleep said...

More than 90% of these cities seem to be suburbs of much larger metropolitan areas rather than free-standing cities of their own.

The few towns that Im familiar with on that list are all high-income areas that would be far out of reach for me. I cant say if this is true of the whole list or just the ones I know.

That's about all I have to say. Sorry I have nothing outrageously useful to add.

Peter said...

I estimated that 85% of Hispanics choose "white" as their race, since the Census doesn't breakdown non-Hispanic whites at the city level, and came up with the non-Hispanic white percentage from that

My initial thought was that 85% was way too high, but considering that the survey is covering mostly affluent suburbs it probably is a reasonable estimate. Hispanics in communities of that sort most likely are more Caucasian-looking than those in cities.

John S. Bolton said...

There could be a severe restriction of range effect here. Municipalities with very high or very low diversity tend to be larger or smaller, while this ranking is of mid-sized ones. My township is on this list, but it actually consists of several dozen school districts each with their own high school, and no busing across district lines. The rest are probably unitary school districts with several high schools though. CA seems to have especially low representation, while MN is very high. In any case, there are close to zero places above 50,000 population today which are good for young couples starting families, but who don't want their children integrated to minority dysfunctionality; when, 40 years ago, there would have been at least dozens. This is the cornering effect, which explains the restrictionist support, and perhaps the referenda against quotas and special tolerance for illegals.

John S. Bolton said...

Correction: I said several dozen school districts when I meant one dozen.

agnostic said...

One argument in favor of including a measure of diversity is that Blacks and Hispanics cannot live in flourishing communities on their own -- look at Latin America and Africa.

They are attracted to the most thriving regions because all of that infrastructure can support a dependent population.

Audacious Epigone said...


Yes, that's what most of them are, although some are stand-alone cities in their own right (Olathe, in my neck of the woods, is one of them).


This NYT article reports that only half of Hispanics consider themselves white, with most of the other half choosing "other" (only about 1 in 50 consider themselves black). It also shows that nearly all (97%) of "others" in the 2000 Census were Hispanic (I was unaware of that). Assuming the entire 2.5% of other is Hispanic leads to an estimate of about 60% of Hispanics considering themselves white--the 85% figure is too high. I've refigured the percentages accordingly. Quite perspicacious of you, thanks!


For what it's worth, the two cities I'm familiar with each have only one school district.

The geographic distribution is interesting--not surprisingly, the Midwest and upper Midwest are heavily represented. What the US Census dubs the "West North Central" (KS, IA, MN, MO, NE, ND, SD) is probably the best region in the country for a middle class family to be.


It's like being rewarded for a 'handicap' then? "This is a pretty nice place to live, and it really seems nice given how functional it is with the demographic base comprising it". Except the high Asian figure makes that difficult to gauge.

Peter said...

It also shows that nearly all (97%) of "others" in the 2000 Census were Hispanic (I was unaware of that).

Wow, that's surprising. I figured that most "others" would be Hispanic, but not 97%. There certainly are plenty of groups that would be likely to classify themselves as "other," for instance Middle Easterners, Persians, Armenians, half-Asians, Brazilians, etc.

agnostic said...

Not so much as a handicap principle thing, just that, if you don't have time to crunch a bunch of numbers about test scores in the schools, number of hair salons per capita, etc., you can just ask "Is it nice enough that dependent groups have migrated there for support?"

Superdestroyer said...

Hunter Mill Virginia is really Reston Virginia and Sully virginia is really Centerville and Chantilly Virginia.

Resto Virginia has been involved in a racial controversy is that the local high school South Lake has a higher minority percentage than suround schools. Enough people with schools have moved out of the community that the Fairfax County School board wants to redraw boundries. The school board did the boundry adjusting in such a way that the rich whites in McLean were excluded buy middle class whites will be transferred into the district.

Audacious Epigone said...


Thanks for pointing that out. Looks like Reston is about representative of the other 25 but a little more black (9.1%) and Asian (9.6%). Centerville/Chantilly epitomize the kind of diversity that is rarely in conflict with a higher than average quality of life, with over 15% of the population being Asian. Both areas are upper-middle class professional.