The Other Problem with Noble Lies

They may not be

My previous post discussed whether it was desirable to suppress true information when you believe that it being known will have bad effects. Two examples offered were hiding the implications of evolution for the distribution of abilities by sex or race and distorting the evidence on the implications of climate change to make them look worse. I offered two arguments against such a policy. This post is concerned with a third — that you might be wrong about the effects of the information you want to suppress.

The Weak Case Against Evolutionary Truth

Darwinian evolution produces organisms optimized for reproductive success. The essential difference between male and female is their role in reproduction. While it is possible that the optimal distribution of characteristics is the same for male and female there is no reason to expect it to be, some reason to expect it to be different.

That is obviously true for observable physical characteristics, including ones such as height not obviously linked to reproduction, so very likely true for less observable characteristics such as the distribution of intelligence. That is supported by evidence. While IQ tests have been normed to give the same average for men and women, average performance on different parts of the test is different, as is the distribution of IQ. While it is logically possible that the observed differences are all due to differences in the social environment, there is no good reason to assume so.

It follows that there is no good reason to expect the number of men with the talents necessary for a particular task, writing fiction or researching mathematics, to be the same as the number of women, hence no good reason to attribute the fact that a majority of successful fiction writers are female and majority of top mathematicians are male to discrimination.

Is there a good reason to pretend the opposite? The only argument I can see for doing so — readers may be able to offer others — is that the knowledge that fewer women than men have talents needed to be top mathematicians may discourage some who do from pursuing that career. But a woman interested in mathematics has better evidence than her gender — her actual performance in math classes and contests. If she has the ability to be a top mathematician it will rapidly become obvious that she is more talented in that field than most, male or female.

One cost of suppressing the information is that young adults, male and female, have less relevant information about what careers they are likely to succeed in. Another is that people falsely deduce discrimination from differing outcomes, act to correct imaginary problems.

What about the same issue with regard to different distributions of characteristics by race? The argument for expecting them is weaker, since males of all races have been optimized for success in the male role in reproduction, females for success in the female role. But people of different races as conventionally defined are descended from ancestors adapted to different environments. That fact explains the physical differences by which we identify them — the optimal distribution of physical characteristics is different in different environments — and there is no reason why the same should not be true for mental characteristics. That does not tell us what difference to expect but it does imply that there is no reason to assume there are no differences.

I have seen two different arguments for pretending that there are none. One is that the knowledge that there might be differences, combined with observed differences in outcomes, will cause people to evaluate others on the basis of their race rather than their ability. Doing so is rarely rational — in deciding who to hire, work for, marry one usually has better information about someone than the color of his skin. It may happen any way — most people like to believe in their own superiority, and race is one way of doing so. People have been believing in the superiority of their national, religious, cultural, or other group for that reason since long before the existence of IQ tests or the discovery of evolution and can be expected to continue to do so with or without an accurate account of the implications of evolution.

The other argument I see is the claim that the belief in racial differences is responsible for many of the horrors of the past, including slavery and mass murder. It is probably true that the belief that blacks were incompetent to run their own lives made southern slave owners more comfortable with owning slaves. But the evidence that such a belief was not necessary to support slavery is the existence of slavery in multiple past societies, such as Greece in antiquity, where the slaves were of the same ethnicity as their owners, enslaved for debt or because their city had been conquered by a neighboring city and its inhabitants sold into slavery.

What about mass murders? The most famous modern example is the killing of Jews by the Nazis, based not on the belief that they were inferior but that they were race enemies. The Holdomor, the deliberate mass starvation of Ukrainians by Stalin, was political not ethnic, as was the mass killing by the Khmer Rouge in Cambodia, the worse of the 20^th century democides measured by fraction of the population killed.¹ Other mass murders of the century were mostly associated with wars with no obvious connection to beliefs in racial superiority supported by evolutionary arguments.

A final argument, one I have not seen offered openly but suspect is an important motive for suppressing the relevant implication of evolution, is that admitting that implication weakens the argument for laws against racial (and similarly gender) discrimination by eliminating the main evidence for its existence, differing outcomes. That is true — but to make it explicitly would be to admit wanting to support your preferred policies with arguments you know are false, which brings us back to arguments against dishonesty discussed in my previous post.

What about the other side of the question, the costs of pretending to know that the distribution of characteristics is the same for all races? The most obvious is the interpretation of any outcome differing by race as due to discrimination, making an employer presumptively guilty without evidence and similarly in other contexts such as school performance or college admissions. A downstream consequence, discussed in an earlier post, is that universities compete to get black students with the result that the mathematically talented afro-American who would have done well at RIT, where his fellow students are equally talented and classes designed accordingly, gets into MIT, where the fellow students are even smarter and the classes designed accordingly (Thomas Sowell’s example). The law student who would have done well at SCU among his peers ends up in Stanford instead, learns less as a result and fails the bar exam (Richard Sanders’ example, with evidence).

I conclude that, in the case of both gender and race, pretending to know that the distribution of innate characteristics must be the same, hence all differences in outcomes due to environmental differences, probably discrimination, has bad effects not good ones.

And Climate

The obvious argument against exaggerating the negative effects of climate change is that it leads to policies that make us poorer, substituting more expensive sources of energy for less expensive ones in the false belief that the latter are actually less expensive when all costs are taken account of.

The less obvious point is that a widespread policy of exaggerating the costs of climate change makes it difficult, perhaps impossible, for those following that policy, academics or journalists, to know whether they are correct to do so. All of us, including the professionals in the field, are dependent on second hand information. An economist or agronomist trying to estimate the costs of climate change depends on someone else’s estimate of how much change there will be and with what effects. A climate scientist deciding whether human production of CO₂ is the cause of climate change is dependent on, among others, paleoclimate researchers who produce proxy evidence that the current rate of global warming is unprecedented in the periods for which they have good proxies. An estimate of the cost imposed by an additional ton of CO₂ depends on someone’s estimate of the effect of CO₂ concentration on temperature, someone else’s estimate of the effect of temperature on mortality, someone else’s estimate of the effect of CO₂ concentration on crop yields, someone else’s estimate of the effect of climate change on rainfall and that on crop yield, …  .

All of these estimates depend on multiple judgement calls, since all of the questions are too complicated to be answered with certainty. If all of the people making those judgement calls start with the belief that climate change is a serious problem, conclude that when in doubt they should prefer high estimates of negative effects, low of positive, the result is that each professional is using as an input to his analysis conclusions of other professionals that are all biased in the same direction. Without access to unbiased data he has no way of knowing whether the belief causing him to bias his results, the belief that climate change is a serious problem, is true.

More relevant to that problem. Much more.

---

Past posts, sorted by topic

A search bar for past posts and much of my other writing

---

Of no relevance to this post but of possible interest to my readers:

Substack’s default text to go with the subscribe button is:

Thanks for reading David Friedman’s Substack! Subscribe for free to receive new posts and support my work.

I routinely edit it down to:

Subscribe for free to receive new posts.

I am happy to have people read me but see no reason to thank them for doing so — presumably they read what I write not as a favor to me but because they find it interesting or informative. My Substack is free, so I don’t see in what sense readers are supporting my work.

The default text strikes me as in the same category, stroking strangers so they will like you, as people who call me up to sell me something and start by asking what kind of day I have been having or ads that start out “you deserve” with no reason why you deserve whatever outcome they claim their product delivers.

1. For a detailed account of 20^th century democides see Statistics of Democide: Genocide and Mass Murder Since 1900 By R.J. Rummel.