Let’s consider a series of statements, without considering whether any of them are right or wrong:

1. People should be treated with respect, regardless of their ethnicity.
2. All ethnic groups have the same average IQ after controlling for cultural biases in IQ tests.
3. Women should be treated with respect and accepted in traditionally male-dominated workplaces.
4. There are no biological differences between men and women that affect their interests or abilities.
5. Homosexuals and homosexual relationships should be given the same respect and acceptance as heterosexuals and heterosexual relationships.
6. Children brought up by same-sex parents do as well or better than children brought up by opposite-sex parents.
7. People of all religions should be treated with respect.
8. Muslims are no more likely to commit violence than are people of other religions.
9. Wealthy people and societies should help the poor.
10. Poverty is entirely caused by unfortunate circumstances and unfair institutions and not the choices of the poor themselves.

At this point, you have probably noticed the pattern. Odd-numbered statements are moral claims. They are all about how one should behave. They reflect liberal values. If you agree with claims 1, 3, 5, 7, and 9 then congratulations! You are a decent human being who is capable of living in a liberal society.

Even-numbered statements are empirical claims. Each of them could be proven true or false by a sufficiently well-designed body of research. Even if the research never reaches a definitive conclusion, these statements will be true or false largely independently of anything we can do about them.

Each of these factual statements has some connection to the moral statement immediately preceding it. But what is the connection? Certainly not that you must accept both statements or neither of them!

However, it would be easier to accept each of the moral claims if their corresponding factual claims were true. It’s really easy to respect a member of a group that has nothing wrong or different about it. But I think you’ll agree that a good person will be respectful and accepting of people regardless of their group identities, especially if that group has problems or differences.

But we face a problem of trying to hold together a diverse liberal society and not everyone has an easy time adopting liberal values. There are two elements of this problem: First, we want to convince as many people as possible to adopt liberal values. Second, we often want to identify and exclude people who don’t hold those values from our voluntary associations.

Let’s pause for a moment and introspect about how we feel about the factual claims above. Suppose one of your Facebook friends posted an article arguing that different racial groups have different IQ distributions because of natural selection pressures. How would you feel about the person sharing this information? Would you be angry? Would you conclude that they are maybe harboring some secret resentment towards people of other ethnicities? What if they explicitly said otherwise? Would you trust them?

Liberal values are dominant in our society and everyone knows they are. Therefore people who do not hold these values are likely to still pretend to hold them. So if someone claims they are not bigoted, it’s cheap talk. “I’m not a bigot,” is something a bigot would say.

So we need signals to sort out the bigots from the non-bigots. One signal our society has seemed to settle on is that non-bigots express their non-bigotry by professing to believe in the even-numbered factual claims above. This is not such a good signal since it is easy for bigots to mimic. Therefore we’ve fallen into what game theorists call a “pooling equilibrium” where most people profess to believe these factual claims, and anyone who deviates from them is labeled a bigot. In other words, certain factual beliefs have become repugnant and expressing one of these repugnant beliefs has become a quick way to be labeled a bigot.

This creates some difficulties. For one thing, it makes researchers shy away from studying these topics. Charles Murray became a pariah by writing a few paragraphs about race in an 800-page book on IQ research. If you were a researcher and you saw what happened to Murray, would you take that into account when deciding what to study? I would.

But if these factual claims are true, it would be nice to know that for sure. We could be more confident in the decisions we make based on them. And, as is often the case in research, we could learn about other related topics by studying them.

The second and more important difficulty is, what if one of these factual claims turns out to be false? As I said above, whether these are true or false is largely out of our control. It would be very convenient if they were all true. It would make it easy to hold liberal values and to convince others to adopt them. But if one of the factual claims I listed (or another factual claim that fits the same pattern) is false, that could have dangerous consequences. Some of these beliefs are inconsequential. But others are very consequential.

For instance, suppose we were to design an anti-poverty program on the assumption that claim 10 is true (“Poverty is entirely caused by unfortunate circumstances and unfair institutions and not the choices of the poor themselves.”). If claim 10 were false, then our program would likely fail, resulting in more people in poverty. And if some policy wonk came along and said, “this program is failing because it fails to account for poor people’s bad choices,” we’d say he just secretly hates the poor and ignore his recommendations. We would dismiss someone, even an expert, for professing a repugnant belief.

Or what if a corporate diversity program was designed on the assumption that claim 4 was true (“There are no biological differences between men and women that affect their interests or abilities.”)? If claim 4 were false, then the program might shoot for the wrong target (perfect equality), it might misdiagnose the failure to hit that target (blaming bias and sexism for all male-female differences), and it might fail to embrace actually helpful diversity initiatives (not accounting for women’s differences in designing corporate policy). How would we react if someone were to argue that claim 4 is false while strenuously and repeatedly affirming his belief in claim 3? Would we label him a sexist and have him fired? He deviated from the pooling equilibrium and expressed a repugnant belief, so that’s exactly what we would do.

Even if claims 2, 4, 6, 8, and 10 (and all others fitting this pattern) turn out to be true, do we really want to base our social order on enforced belief? People are likely to get resentful and angry if told to conform or be punished. If someone tells you to “pay no attention to that man behind the curtain!” won’t you be especially curious what’s behind there? We might actually push people into rejecting our enforced beliefs and our liberal values by creating an environment of stifling orthodoxy.

If you haven’t already guessed my position on this, here it is: No factual claim should be repugnant. We should construct arguments for liberalism and toleration on moral grounds, not tie them to factual claims that could turn out not to be true. And we should assume good faith when arguing about these topics. If someone claims not to be a bigot, and the only evidence for their bigotry is that they express an unpopular belief on a sensitive topic, I say we give them a pass. Yes, this might mean letting a few actual bigots under our radar. But who cares? If an actual bigot goes their whole life pretending to be tolerant, so much so that they actually behave like a tolerant person, then their bigotry isn’t hurting anyone. And if their bigotry really does hurt someone, we can call them on it when that happens. We should find other ways of sorting out bigots; costly signals that don’t lead to a pooling equilibrium. Because what we’re doing now is neither sorting out the bigots nor promoting liberal values. What we’re doing now is constantly getting into pointless fights over who is and isn’t a bigot. That needs to stop.