Almost weekly, someone does the following trick:

Find a bunch of natural-born American citizens. Get them to take the US Citizenship Exam. Snicker while they fail. Write an article about citizens’ ignorance. Title it, "How Dumb Are We?"

This week's show is brought to you by Newsweek. Recently, 38% of US citizens polled failed the exam.

Given that I’m on record saying that uninformed or certain kinds of misinformed citizens have a moral obligation to abstain from voting, you might think I get a big kick out these sorts of things. I don't.

Here is an excerpt from The Ethics of Voting:

People will sometimes cite evidence of voter ignorance and then complain, “Americans are stupid!” Well, yes, some Americans are stupid, but most are quite intelligent and skilled, despite being ignorant about politics (and some other issues, such as world geography). They don’t invest in political knowledge because the investment doesn’t pay.  They have their own lives to lead. They have other valuable things to contribute to society. The political class often lampoons citizens’ lack of knowledge, but citizens often have better things to do (for themselves and for each other) than invest in knowledge that is of no special use to them or to others.

The scholars the Newsweek article cites don't claim that Americans are dumb. They offer structural explanations for American ignorance. The title of the article was likely an editor's decision. But of course the message that gets conveyed and talked about (e.g., by the DJ on the hard rock station I listen to while driving) is that Americans are stupid.

My book will be labeled as objectionably elitist. But as I’ll discuss over the next few weeks, I have an unusually egalitarian and populist view of what it takes to be a good citizen. (I don't think there's anything special or particularly admirable about having political knowledge. I'd love to see some plumbers go around my campus and do a poll showing that most professors can't fix a simple leak.)

There is little logical connection between the US Citizenship Exam and what it takes to be a good voter. Take a look at some sample exam questions here. Only a small number of these questions matter for good voting. Voting well usually requires that you know something about candidate’s likely performance, and so if you can’t identify the current president, you are unlikely to be able to rationally choose between him and his competitor in the next election. However, most of these questions are just trivia and fun-facts. Take this question:

             Who said, “Give me liberty or give me death!”?

I suppose I expect an educated American to be able to answer this question. But knowing this information will not make or break you when it comes to being a good voter. On my theory of voting ethics, citizens should vote for what they justifiedly believe will promote the common good. Being able to identify Patrick Henry’s most famous quotation might help you win Jeopardy, but it won't help you vote well.

Still, even if there is no logical connection between being able to pass the US Citizenship Exam and being a good voter, there may well be a statistical correlation. It might turn out that the people who do best on the exam are also the most likely to qualify as good voters, per my theory of good voting. It might also turn out that people who fail the exam are also the most likely to quality as bad voters per my theory of bad voting. For some evidence of this claim, see Scott Althaus’s work. Althaus shows, statistically, that an "enlightened public" has systematically different political preferences from an unenlightened public.* The citizens who know the least about political trivia and fun-facts have systematically different political preferences from a  hypothetical enlightened public. People's attitudes about politics correlate with their objective political knowledge in ways that cannot be explained by people's demographic factors. I discuss Althaus’s findings at some length in chapter 7 of my book.

**********

* Some explanation of "enlightened preferences": Your enlightened policy preferences are the policies you would prefer if you had maximal political knowledge.  Suppose right now you prefer policies P.  Imagine we keep almost everything about you the same, but boosted your political knowledge.  You now know all there is to know about politics.  Suppose adding knowledge changes your preferences—you now prefer policy set Q instead of P.  These policies are your enlightened preferences.  Your enlightened preferences are what you would prefer if you were fully informed.  (If it acquiring fully information wouldn’t change any of your current preferences, then your current preferences are also your enlightened preferences.)

Althaus has tried to estimate what voters’ enlightened preferences are.  Here is Bryan Caplan's summary of Althaus's method:

1. Administer a survey of policy preferences combined with a test of objective political knowledge.  [Althaus’s data sets are the 1988, 1992, and 1994 American National Election Studies.]
2. Statistically estimate individual’s policy preferences as a function of their objective political knowledge and their demographics—such as income, race, and gender.
3. Simulate what policy preferences would look like if all members of all demographic groups had the maximum level of objective political knowledge.

So, Althaus’s method was to determine, statistically, how policy preferences correlate both with objective political knowledge and with various demographic factors.  Since he had a massive data set, he could determine what effect knowledge has on people’s policy preferences while controlling for biases and attitudes caused by their demographic factors. For instance, he could determine how knowledge affects one’s attitudes about trade policy while controlling for income. Althaus wanted to estimate what an enlightened public—a society that exactly mirrors ours demographically, but in which everyone has complete political knowledge—would prefer.