« Oh, To Be In England | Main | Just You Weight »

October 03, 2005

Do Good Communities Make Liberals?

While studying the connections between social capital and health I stumbled across something rather odd. States with high social capital--strong connections between people and their communities--tend to vote democratic.

Harvard researcher, Ichiro Kawachi, one of the leading lights on social capital and health, has performed several studies that make state-by-state comparisons; and he's shown that, on average, states with higher social capital also have better health outcomes. But as I was peering over some of his charts I couldn't help but notice that states with higher social capital also tended to be "blue" states--they voted for John Kerry in the last presidential election.

Unfortunately, Kawachi reports the results for only 36 states (the others did not have sufficient data to support his study) so my little "finding" here refers only to those states, though they do include all the big ones. That's just one of the limitations, but I still think it's interesting that 6 of the 10 states with the highest social capital voted for Kerry in the 2004 elections. Meanwhile, 8 of the 10 states with the lowest social capital voted for George Bush in '04. Don't believe me? Here's the rank-ordered list....

(In keeping with prevailing media norms, republican-voting states are coded red; democrat-voting ones are blue.)

1 North Dakota
2 Iowa
3 Washington
4 Minnesota
5 New Hampshire
6 Rhode Island
7 Utah
8 Wisconsin
9 Oregon
10 Wyoming
11 Ohio
12 Colorado
13 New Jersey
14 Indiana
15 Michigan
16 Illinois
17 Massachussets
18 Kansas
19 Pennsylvannia
20 California
21 Missouri
22 Arizona
23 Connecticut
24 Texas
25 Kentucky
26 South Carolina
27 Alabama
28 New York
29 Maryland
30 Georgia
31 Arkansas
32 Mississippi
33 North Carolina
34 Tennesse
35 Oklahoma
36 Louisiana

As the list here shows, the relationship is certainly not comprehensive--there's a lot of muddle in the middle--but on the extremes there does appear to be a correlation between low social capital and voting for Bush.

While Kawachi never mentions the voting comparison, in a separate study he offers a plausible explanation in the context of health outcomes. He suggests that high social capital leads to more civic engagement and, in turn, to more investment of resources, money, and concern into the community at large. For Kawachi, that investment is a partial explanation for better health outcomes--places with high social capital care more about the welfare of others.

So I wonder whether--to the extent that democratic voters favor more public investment in the community and republicans less--that Kawachi's explanation fits. Places with higher social capital reflect their preference for community in their voting habits. In other words, good communities foster democratic voting.

Just a thought.

A couple of notes and caveats are in order...

*** Kawachi's measurement of social capital is, in this instance, a shorthand. It's the percent of people responding "yes" when asked whether most people would take advantage of them if given the chance. Most researchers think this is a reasonable, if abbreviated, way to assess social capital.

*** My list is extrapolated (read: eyeballed) from Howard Frumkin et al., Urban Sprawl and Public Health (Island Press: Washington, DC: 2004), p 167.

Posted by Eric de Place | Permalink


TrackBack URL for this entry:

Listed below are links to weblogs that reference Do Good Communities Make Liberals?:


Florida and Glaeser have made careers unpacking this issue, Eric.

According to microecon theory, the coasts (with temperate weather) are attractants, and those who can afford it Tiebout sort and move there. The intellectual capital attracts hi-tech jobs and intellectual spillovers from that capital further attracts those that seek amenities. Amenity seekers (for reasons that I forget) tend to be liberals. But the liberal sorter here is the intellectual capital and amenities. Social capital arises later.

Backing up, those that can afford to sort to good weather are the well-educated ones who tend to seek amenities. These folk tend to be lib'rulls, for some reason.

Posted by: Dan Staley | Oct 3, 2005 10:57:56 AM

Thanks. I'm familiar with Florida's reasoning though I don't find it terribly compelling.

I also don't think the "good weather" explanation is very helpful. It's hard to argue that bastions of liberalism in places like Boston, Minnesota, and even Seattle are due to the lovely climate. And by contrast, the southwest and the Carolinas have (arguably) some of the most agreeable weather in the country but are predominately republican-voting. In any case, there are lots of examples and counter-examples to the weather explanation. I think the liberal politics that Florida talks about have arisen in certain places for a whole huge variety of complicated geopolitical and historical factors, many of which were purely accidental. But I suppose that's a much longer discussion...

Posted by: Eric de Place | Oct 3, 2005 11:11:32 AM

Sure, Eric. These are just the microecon guys' ideas - they aren't perfect explanatory variables, but they do pretty well as a beginning.

Ts and r^2s for average temperature are pretty strong (not 1, of course) for population migration TO an area and pretty good for migration away FROM.

Education attainment tends to be higher where average temps are higher Ts and r^2s for average temperature AND educational attainment are pretty strong (not 1, of course). Urban areas, of course, attract the educated as that's where the good jobs are.

Starting with this, we can begin to tease out knowledge spillovers, strength of social networks, etc. We then build on this when considering the Carolinas, below the Mason-Dixon line - I'd wager we'd see a purpling of the urban areas there in the last 2 decades.


Posted by: Dan Staley | Oct 3, 2005 1:15:23 PM

Great post.

I would strongly agree with your thesis and take it one or two steps further. Not only should the Democratic Party (and other progressive orgs) make building strong communities part of our policy, but also work to develop local communities by building political involvement by being a more open organizations.

There is some more thorough data on social capital in all 50 states here:

Another note, the Republican razzing and the Dem lamenting on our losing the so called "exurbs" or the "10 fastest growing communities" doesn't surprise or scare me. That Progressive messages don't win in communities where most people came from somewhere else, don't have strong ties and will probably move in the next four years isn't bad. Actually, its great.

Posted by: Emmett O'Connell | Oct 3, 2005 1:50:24 PM

Temperature actually seems to have a pretty strong effect here, more so than politics. Take a map and color the first eighteen one color and the last eighteen another. You'll see that the divide isn't red/blue but rather north/south. Maybe cold winters bring people together and make them more reliant on their communities?

Posted by: Eric L | Oct 3, 2005 6:01:33 PM

Thanks for the link, Emmett!

I'm actually unconvinced of a link between political liberalism and social capital. (This, despite the fact that I'd like to believe it's true.)

If you follow the link that Emmett provides, you can find state-by-state social capital rankings for 48 states. From these rankings, the US Southeast is clearly the poorest in social capital. Of the bottom 13 states, ranked by a comprehensive social capital index, 12 are in the Southeast. (Nevada is dead last.) All of these bottom-tier social capital states voted for Bush in 2004.

But once you get out of the southeast, there's no obvious correlation between social capital and propensity to vote for Bush. The Dakotas top the list for social capital in the US; they were both firmly pro-Bush; as were 6 of the top 13 most social-capital-intensive states in the US.

To me this could mean:
1) low social capital produces political conservatism
2) political conservatism produces low social capital
3) the peculiar history & demographics of the US south produce both low levels of social capital and high levels of conservatism.

To me, #3 seems like the clear frontrunner -- with early immigration patterns, coupled with the legacy of slavery, as leading potential causes of both phenomena.

Posted by: Clark Williams-Derry | Oct 4, 2005 10:51:56 AM

Wow. Kawachi does some interesting work.

In *Int'l Journ Epidemiol* 33:4 pp 682+, he (et al.) explores health by association with different categories of social capital and how this emerging field is still finding it's way (a nice way of saying the empiricism isn't there yet). I'll pass on the paper if you want it, sir.

Anyway, my initial reaction to your comment, Eric, would be the social networks that remain after out-migration are very strong, as the low population densities of WY, IA, etc that are not on the coasts attest. As there are few material and intellectual resources remaining (because it all has fled to the cities and coasts), the remaining networks are necessarily strong in order to survive.

Perhaps there are lessons here...

Posted by: Dan Staley | Oct 4, 2005 11:09:36 AM

Good points, Dan. I think you may be on to something.

Yes, please do pass the Kawachi paper on. In addition to his work on social capital's link to health, he's done quite a bit of empirical work on the cross-national and inter-state relationship of income inequality to health. It's fascinating stuff.

Posted by: Eric de Place | Oct 4, 2005 12:01:46 PM

At least for “scholarly” purposes, I’m nervous about the idea of linking social capital and a state’s partisan leanings. Let’s put aside the dubious methodology of using one presidential election as an indicator of a state’s underlying ideological predilections. For me the larger question is why did relatively conservative states such as North Dakota, Iowa, Utah and Wyoming rank so high?

One could brainstorm any number of variables, small population size being just one of them. However, I suspect that a key factor is the relative strength in these states of communitarian social institutions, e.g., the Mormon church in Utah, and various farmer organizations in North Dakota and Iowa.

Why should sustainability advocates pay any attention to “red” forms of communitarianism? For one thing, it’s good politics. Here it is helpful to look at history. The most progressive era in Washington state was made possible by a remarkable alliance of labor unions and farmers during the early part of the 20th Century.

Yes, we live in a very different world today. Nevertheless, I would argue that a sustainability vision has the potential draw together “red” communitarians and “blue” urban cultural creatives. This is why I think it is a tactical mistake to conflate sustainability activism with the Democratic Party and left-of-center organizing.

Posted by: Steven Salmi | Oct 9, 2005 10:51:10 AM

Some geographers and theorists, notably Michael Lind, would point out that the WASPish Yankee settlers of "Greater New England" -- the Northeast and points directly west, most of which were first settled by New Englanders -- were very found of founding "communitarian social institutions" as they marched west across North America. (Portland, Oregon escaped being named Boston, Oregon by a coin toss.) Their descendants, and others who landed in those parts of the country (like the Germans and Scandinavians of the upper Midwest, where I live) largely upheld those traditions. To this day, evidence of this New England influence abound: township governments, municipal home rule, gridded county seats with courthouse squares, tiny liberal arts colleges scattered about, relatively high property taxes for relatively generous school funding, etc.

This thesis strikes me as being too determinist, but it does seem to have some merit in explaining the differences mentioned above. Note also that the top 10 states on the list are among the whitest in the country, and have exhibited fairly low population growth since WW2 -- and thus are relatively untouched by the demographic trends (namely, immigration from outside Europe) that would have disrupted the traditions of the early settlers. The social strife of having a racially mixed population also would reduce the social capital available in a place.

Posted by: payton chung | Oct 24, 2005 6:53:16 PM

Payton has an excellent point. I note that we are attempting to design new city spaces that are very much like the spaces Payton describes above.

Posted by: Dan Staley | Oct 25, 2005 8:10:53 AM

In my experience, having a racially mixed population in a place adds to the social capital rather than reduces it. That's one of the things I loved about living in San Francisco for three years, in the early 1970s. It was the most racially mixed and vibrant city that I had lived in.

Prior to that, I lived in Omaha and Lincoln, Nebraska (where I was born). And one of the things I love about going back to Omaha to visit, is the warm and welcoming African-American community there.

Plus, having come from a bi-racial family, I think it's a beautiful environment when people of different races come together in camaraderie and fellowship.

Posted by: Michelle Parker | Oct 25, 2005 3:38:39 PM