The Zero-Sum Society

A big trend of the 21st Century is more economic activity being crammed into smaller geographic portions of Western nations, at the same time as those areas become unaffordable to more and more people. Cause and effect probably go in both directions there. But one of the consequences of this trend seems to be that areas with more expensive property are also becoming more left-wing. You could see this in the 2016 US election:

Map_US_Home_Values_by_State__955W2016_movvalue

This also seems to have been the trend in last week’s UK general election, where Labour made bigger gains in areas with more expensive property:

 

 

While one can understand why rich people would find cosmopolitan liberalism like Hillary’s more attractive than Trump’s rude nationalism , the trend seems to be for young educated people who don’t themselves have all that much money (but live in expensive places) to be attracted to relatively unreconstructed Marxists like Bernie Sanders or Jeremy Corbyn. It strikes me that Marx was wrong for lots of reasons, but the biggest is because classes under capitalism are not in fact engaged in a zero-sum competition for resources, but in positive-sum economic exchange. Capitalism has been such a success by allowing for that positive sum process of trade and innovation to accelerate and progress. But if the supply of a good considered necessary to a dignified life (like owner occupied housing) becomes almost perfectly inelastic in spite of increasing demand, then the society becomes much more zero-sum after all, with what is good for Peter being bad for Paul. In this way, left wing politics become more analytically correct as the society constrains and limits the ability of capitalism to give people what they say they want- a dignified home and means of providing for themselves.

In the US, it’s still possible for people with some education to shoo the fly out of the bottle, as it were- there are enough jobs in places where it’s still quite cheap to live that to a significant degree living in a place where you “can’t afford to start a family” is a matter of choice. But just as, for the California Democratic party, making housing steadily more expensive over the last sixty years has turned out to be a very good plan, so too conservatives should probably prioritize Affordable Family Formation over tax cuts if they want more people to be conservative in the future.

21 thoughts on “The Zero-Sum Society

    1. It’s median home value from the 2010-2014 Census. I believe it’s drawn from the American Community Survey’s questions, so this is what people say their home is worth. My sense is that higher turnover areas are in areas with higher values (and people are more likely to sell their houses if they’re going up in price) so home values tend to be below median market prices, though they obviously tend to go up and down with them. I’m not sure what the Census does to make sure the values are accurate (ie, tax assessments are probably pretty accurate overall, but I’m not sure if just asking someone what their house is worth always gets you a realistic answer.)

      Like

  1. Also, California is full of crackerbox houses on small lots. The kind of two story brick house on a one acre lot that is common in the suburbs of say Chicago or Philadelphia is rare in California.

    Liked by 1 person

  2. The counties in MI with the largest Bernie vote as a share of total primary vote (Democrat and Republican) were Washtenaw (U of M, obviously), Marquette (university town), Ingham (Michigan State), Isabella (another college town), Kalamazoo (another college town), Benzie (a rural county, not sure what’s there, but near Grand Traverse and Leelanau), Genesee (Flint and immediate outskirts), Grand Traverse (well-educated resort town), Leelanau (same deal as Grand Traverse, but less densely populated and more educated), and Alger (rural UP).

    The counties in WI with the largest Bernie vote as a share of total primary vote (Democrat and Republican) were Menominee (Native Americans), Dane (Madison),
    Bayfield (traditionally heavily White rural Dem), Ashland (same deal as Bayfield), Eau Claire (apparently a college town), Portage (apparently a college town), La Crosse (apparently a college town), Iowa (near Madison, very White, not a college town), Vernon (organic farms, not a college town), and Rock (Janesville).

    Basically, rural northern Whites, strongly resembling the early Republican coalition.

    Liked by 1 person

      1. Yes; the counties with a high Black population would be higher up the list were this the 2016 election with Bernie as the Dem nominee.

        Like

  3. I think that there are problems with your graph of percent of vote received by Trump vs log of median home value. The major problem is that the numbers on the x-axis are ridiculous. The logarithms of the home values range from about 11.5 to about 13. When I undo the logarithms to get the actual values, they range from $10^11.5 to $10^13 ($316,227,766,016 to $10,000,000,000,000). Unless you are using Zimbabwean dollars, these numbers don’t make sense.

    Here’s a reality check: The highest home value on the top map is Hawaii ($504,500) and the lowest is West Virginia ($100,200). The log (base 10) of 504,500 is 5.7 and the log of 100,200 is 5.0.

    But why use logarithms in the first place? The median home values across the country only vary by a factor of five, so the log function isn’t really all that helpful. Why not just use the actual home values?

    Also, the “percent of vote” numbers you use are not actually percentages, they are decimal fractions (i.e. percentages divided by 100).

    Note: I posted a version of this comment on Sailer’s blog, but he hasn’t approved it yet.

    Like

    1. BTW, from an anedoctal perspective, it’s not unusual to use “log” for “natural log”, especially among people who write statistical models. As for the log-scale, it’s a modeling choice that I assume Toad thought represented the relationship more accurately — other choices are possible, of course. I’d have used the logit of the vote share, for example (since it’s bounded between 0 and 1), but I suppose in this case, with such a noisy relationship and so few data points (oh the social sciences!), it wouldn’t make a big difference.

      Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s