In my book I suggest that houses make poor habitat for democracy. The big argument here, I think, is that houses tend to prioritize control over private space. In so doing they generally encourage detachment from public life. Houses suggest we can cut ourselves and our families off from those around us rather than investing in ways of figuring out how to live together. The “single-family detached” house reifies the notion that the only ties that matter are those of kinship, cutting off our obligations to the broader social world. As I describe in my book, and as further detailed in the work of historians like Frykman & Löfgren (see their excellent book Culture Builders) and John Gillis, this was a strategic move on the part of a rising urban middle-class in the late 19th and early 20th centuries. Housing was meant to serve as both a stage for demonstrating moral worth and a sanctuary from the moral claims of others (the latter rendered obsolete by the extension of market governance).

In a broader sense, most houses are surrounded by other houses. That’s a legacy of zoning, of course. But it tends to insure that even when residents leave their housing, they don’t have many public places to go, and to the extent they bump into anyone, their encounters tend to be with people much like themselves.* As described by a variety of political theorists (from Susan Bickford to Thad Williamson to the wonderful Iris Marion Young), who we regularly run into influences who we think we need to take into account in our politics. In much the same way that getting our news through social media tends to provide us only a very partial picture of the world, surrounding ourselves with similar people does much the same, and tends to lead us toward intolerance of difference.

So too many houses is concerning for collective projects like democracy. On a related note, there is a very strong correlation (r=0.82) at the state-level between the proportion of structures made up of detached houses and the proportion of votes that went to Donald Trump. Here’s the scatterplot:

Housing data come from Census (American Fact Finder) 2010-2014 ACS 5-year estimates. Vote data comes from USelectionatlas.org (thanks Dave Leip), and of course is subject to change – especially as recounts get under way! And yes, the correlation is still quite strong if you remove Washington DC (r=0.74) or drop mobile homes (r=0.76), bivariate analyses I’ve already tried (all are defensible variations on the theme). I haven’t run the analyses yet, but at the county level, you’d likely get similar results, if not stronger. More urbanized counties voted heavily for Clinton, as you can see mapped out here (and let’s not forget that she won the popular vote).

So did the USA build its way to Donald Trump? There’s a correlation and a plausible connection, fitting with an alarming streak of intolerance running through the Trump campaign. But there’s also Vermont and New Mexico (one of my home states!), clear outliers from the pattern. Other stories might be told as well and surely will be in the months to come, but this one’s definitely worth keeping in mind.**

*-that said, suburbs are getting more diverse! See, for example, Christopher Niedt’s book.

**-for instance the historical association between racism and zoning in the USA is also worth noting, with the persistence of racism possibly playing a role in both the way zoning has configured the housing stock and the 2016 vote (see, for instance, Dorceta Taylor’s work on some of these connections)…