A signaling theory of class x politics interaction


55


Scott Alexander

The media, most recently The Economist and Scientific American, have been publicizing a surprising statistical finding: in the current economic climate, when more Americans than ever are poor, support for policies that redistribute wealth to the poor are at their lowest levels ever. This new-found antipathy towards aid to the poor concentrates in people who are near but not yet on the lowest rung of the social ladder. The Economist adds some related statistics: those who earn slightly more than the minimum wage are most against raising the minimum wage, and support for welfare in an area decreases as the percentage of welfare recipients in the area rises.

Both articles explain the paradoxical findings by appealing to something called "last place aversion", an observed tendency for people to overvalue not being in last place. For example, in laboratory experiments where everyone gets randomly determined amounts of money, most people are willing to help those with less money than themselves gain cash - except the person with the second to lowest amount of money, who tends to try to thwart the person in last place even if it means enriching those who already have the most.

"Last place aversion" is interesting, and certainly deserves at least a footnote in the catalogue of cognitive biases and heuristics, but I find it an unsatisfying explanation for the observations about US attitudes toward wealth redistribution. For one thing, the entire point of last place aversion is that it only affects those in last place, but in a massive country like the United States, everyone can find someone worse off than themselves (with one exception). For another, redistributive policies usually stop short of making those who need government handouts wealthier than those who do not; subsidizing more homeless shelters doesn't risk giving the homeless a nicer house than your own. Finally, many of the policies people oppose, like taxing the rich, don't directly translate to helping those in last place.

I propose a different mechanism, one based on ... wait for it ... signaling.

In a previous post, I discussed multi-level signaling and counter-signaling, where each level tries to differentiate itself from the level beneath it. For example, the nouveau riche differentiate themselves from the middle class by buying ostentatious bling, and the nobility (who are at no risk of being mistaken for the middle class) differentiate themselves from the nouveau riche by not buying ostentatious bling.

The very poor have one strong incentive to support redistribution of wealth: they need the money. They also have a second, subtler incentive: most redistributive policies come packaged with a philosophy that the poor are not personally responsible for the poverty, but are at least partially the victims of the rest of society. Therefore, these policies inflate both their pocketbook and their ego.

The lower middle class gain what status they have by not being the very poor; effective status signaling for a lower middle class person is that which proves that she is certainly not poor. One effective method is to hold opinions contrary to those of the poor: that redistribution of wealth is evil and that the poor deserve their poverty. This ideology celebrates the superiority of the lower middle class over the poor by emphasizing the biggest difference between the lower middle class and the very poor: self-reliance. By asserting this ideology, a lower middle class person can prove her lower middle class status.

The upper middle class gain what status they have by not being the lower middle class; effective status signaling for an upper middle class person is that which proves that she is certainly not lower middle class. One effective way is to hold opinions contrary to those of the lower middle class: that really the poor and lower middle class are the same sort of people, but some of them got lucky and some of them got unlucky. The only people who can comfortably say "Deep down there's really no difference between myself and a poor person" are people confident that no one will actually mistake them for a poor person after they say this.

As a thought experiment, imagine your reactions to the following figures:

1. A bearded grizzled man in ripped jeans, smelling slightly of alcohol, ranting about how the government needs to give more free benefits to the poor.

2. A bearded grizzled man in ripped jeans, smelling slightly of alcohol, ranting about how the poor are lazy and he worked hard to get where he is today.

3. A well-dressed, stylish man in a business suit, ranting about how the government needs to give more free benefits to the poor.

4. A well-dressed, stylish man in a business suit, ranting about how the poor are lazy and he worked hard to get where he is today.

My gut reactions are (1, lazy guy who wants free money) (2, honorable working class salt-of-the-earth) (3, compassionate guy with good intentions) (4, insensitive guy who doesn't realize his privilege). If these are relatively common reactions, these would suffice to explain the signaling patterns in these demographics.

If this were true, it would explain the unusual trends cited in the first paragraph. An area where welfare became more common would see support for welfare drop, as it became more and more necessary for people to signal that they themselves were not welfare recipients. Support for minimum wage would be lowest among people who earn just slightly more than minimum wage, and who need to signal that they are not minimum wage earners. And since upper middle class people tend to favor redistribution as a status signal and lower middle class people tend to oppose it, a recession that drives more people into the lower middle class would cause a drop in support for redistributive policies.