Friday, November 27, 2020

Politics makes us stupid: COVID-19 edition

An essay by Ezra Klein at Vox from 2014 says politics makes us stupid. He’s reporting on a study by Harvard Law Professor Dan Kahan. Kahan and his coauthors first present a mildly tricky problem about whether, based on the data presented, an intervention made a problem better or worse. In this problem, the subject’s facility with math or statistics predicts whether they get the problem right.

But in a politically charged version of the test, using exactly the same numbers, numeracy stopped being a good predictor of who would get the problem right. Instead, the person’s ideology predicted how they would answer the question. Higher-than-average math skills didn’t help participants when the data showed a result out of line with their political tribe (and if you think that numeracy is a trait exclusive to a particular political view, then you might already be too far gone in your tribalism).

Here’s Klein:

Being better at math didn’t just fail to help partisans converge on the right answer. It actually drove them further apart. Partisans with weak math skills were 25 percentage points likelier to get the answer right when it fit their ideology. Partisans with strong math skills were 45 percentage points likelier to get the answer right when it fit their ideology. The smarter the person is, the dumber politics can make them.

This story might suggest that people have a more or less stable and consistent ideology (gun control: bad) that they take great pains to avoid betraying. Motivated numeracy bias is a way to preserve stable ideological commitments that define their identity. 

But it turns out that even this is too optimistic a view about people.

Most people don’t really even have stable political beliefs. In Neither Liberal Nor Conservative, political scientists Kinder and Kalmoe present studies showing that the number that do is, at most, 17% of Americans (also, “stable” here refers to over the course of only about a year). Rather, most people are political innocents. They don’t support political leaders or parties because their beliefs line up with those leaders’ policies. Instead, their political beliefs line up with those leaders’ and parties’ policies because they support those leaders or parties. 

This is significant because we very quickly saw in 2020 the politicization of COVID-19. Whether or not you support locking down, opening up, wearing masks, shaking hands, etc. is better predicted by your political affiliation than anything else (and probably least of all by how familiar you are with epidemiological models or data).

By now, of course, we know that lockdown is a "liberal" position and back to normal is a "conservative" position. Is there any way we could have confidently predicted this ex ante? Maybe. But imagine a Twin Covid America thought experiment where Trump came out decisively in favor of strong lockdowns. Then, of course, Republicans there would be mustering arguments about negative externalities to justify mask mandates, restrictions on movement and curfews to promote public health, safety and order. And Democrats in Twin Covid America, of course, would be mostly against these policies because, among other things, wealthy Americans have a much easier time staying in place at home than poorer working class people with low levels of savings. Having a high-paying job tends to correlate with being able to work from home and having people with low-paying jobs deliver their Grubhub orders and groceries. Wealthy families are more likely to have a parent who can work fewer or more flexible hours or leave their job to help with their children’s distance education. Meanwhile, children in lower-income households are much more likely to face challenges that make distance education ineffective (indeed, here in the non-twin US we are beginning to see evidence of a widening learning gap). Good progressives in Twin Covid America would want to keep things as open as possible, rather than impose a general lockdown, in the name of social justice, and would write op-eds (like this one) about the likely unintended consequences of mask mandates.

The fact that the opposite happened here, and different considerations featured in the two sides’ positions, is actually pretty arbitrary. It probably doesn’t reflect a genuine, hard-won, intellectually honest working out of how to apply underlying values and principles to a particular social problem. Again, only a small minority of people even have a coherent ideology that determines their political views. And so-called “liberal” or “conservative” values are so generic and indeterminate that someone can always offer a post hoc “rationalization” of why the positions shook out the way they did. 

Kyle Swan

Philosophy Department

Sacramento State


  1. Kyle,

    I'm sympathetic to much of these studies, but what of this possibility? There are many things that liberals and conservatives agree upon, such as being against murder, stealing, rape, physical assault, etc. Couldn't it be possible that hundreds of thousands of deaths and some earlier projections of over 1 million deaths from the pandemic would be enough to unite both parties in taking wearing a mask seriously generally speaking? It's just that the lies of Donald Trump regarding the efficacy of masks and the pandemic itself, the right wing media echo chamber with low truth rates (contrasted with the general reliability of the mainstream "liberal" media), and studies showing that republicans are much more susceptible to fake news than democrats (Pennycook 2019, Tarlach 2019) has led Republicans to be against mask mandates. Democrats will be in favor of mask mandates regardless of leadership just like for other issues such as murder, and republicans would have been in favor of them absent of unique circumstances that they're more susceptible to than Dems like the era of post truth, inter alia.

  2. Sure I think that's possible! Is there something I wrote that made you think otherwise? However, despite liberal and conservative agreement on the badness of murder, assault, etc. we get obviously tribalistic responses concerning how to reduce, police, punish, etc. these offenses.

  3. Kyle,

    Thanks for this.

    Let me wonder out loud here how the more pessimistic approach (articulated in this piece) might handle so-called political independents.

    In other words, in a nation like the current USA, where two parties dominate almost every elected office (and perhaps most of the 'appointed' offices), should we expect so-called 'independents'

    By 'independents' I mean, not just those who merely think of themselves as 'independent thinkers' (I mean, everyone thinks that--like sheep!), but those who have formally declared themselves to be neither Democrat nor Republican, or indeed any party at all--you can and must (!) choose whether to do this, I just learned first-hand, whenever you renew your driver's license at the CA DMV online.

    I suppose it's possible that such 'independents' would be the litmus test for...truth and rationality, right? Since they are formally no longer beholden to the numeracy-corrupting influences of political affiliation?

    If not them, then perhaps someone else?