More Odds and Ends

I try to maintain a minimum length for my posts of about two pages in Word, 12 point single spaced. Sometimes I have something I feel like talking about that is shorter than that, hence this post. (from my first odds and ends post)

Democracy: A Fermi Estimate.

A commenter responded to my claim that an individual voter in a presidential election had almost no chance of affecting the outcome by citing the case of Florida in 2000, which Bush won by 537 votes, giving him the victory in the election. That got me trying to do a very rough estimate of the chance that a voter in a swing state would change the outcome of the election.

I start with two simplifying assumptions:

In each of the last ten elections there were 10 swing states

Florida in 2000 is the closest any of them came to determining the outcome.

The probability distribution for the outcome should be relatively flat for small variations, so in an election decided by 500 votes the probability that it would have been decided by one vote should be at most one in 500. If the closest swing state election out of a hundred had a .002 probability of being decided by one vote, the probability that the swing state election you are voting in has that probability is one in a hundred. Hence the probability of your vote for your candidate changing the outcome from a loss by one vote to a tie or from a tie to a win is .002x.01 = .00002. If we assume that a tie has a .5 chance of being converted to a win for either candidate, by a coin flip or a second vote, that gives a .00001 chance, one in a hundred thousand, that a vote in a swing state election will change the winner.

Obviously that is a very crude estimate, both because of the simplifying assumptions designed to make the arithmetic simple and save myself the work of getting realistic numbers and because of the lack of any sophisticated statistical modeling. But I think it provides support for my view that the probability of a single vote changing the electoral outcome is tiny, hence that unless you put a very large value on getting your candidate elected you should view voting as something you do because you enjoy doing it, like cheering for your team. If you don’t enjoy doing it but would go to the trouble of voting for five dollars — low for in person voting but not for mail-in — and are in a swing state, you should only bother to vote if the value to you of your candidate winning is of the order of half a million dollars. I can’t think of a corresponding calculation for a non-swing state but suspect the number would be in at least the billions.

Half a million is a lot of money but the election outcome affects a lot of people and you probably value benefits to other people, even if not by as much as you value benefits to yourself. If the value to you of the the benefit you believe the random American gets from your candidate being elected is a dollar that makes the value to you of the benefit to everyone else of his being elected 330 million dollars, so bothering to vote in a swing state election might almost be worth doing. If you believe that electing him is worth ten cents to everyone in the world, not the benefit to them but the value to you of the benefit to them, heavily discounted because nobody is altruistic enough to count benefits to strangers as equivalent to benefits to himself, that comes to eight hundred million so you should probably vote — provided you are in a swing state. I’m not.

On the other hand, you should discount the value for the possibility that you are supporting the wrong candidate. By assumption, about half the voters support his opponent — that is why it is a close election, hence why your vote might matter. Difficult as it may be to make yourself believe it, they could be right.

All of this assumes that you are voting for a candidate because you believe his getting elected is good for the country and the world, not just because you believe it is good for you.

Conjectures Concerning Intelligence and Pulchritude

I once heard about a study that found a correlation between intelligence and beauty amongst women. I do not know if it is true but it is at least plausible. Smart men get rich. Rich men want beautiful wives and can attract them. The beautiful wives of smart husbands produce daughters who are both. To test whether the correlation is real, collect photographs of women from a variety of colleges, have men rank them by pulchritude, and see if the ranking correlates with the ranking of colleges by average SAT score.  If you try it, let me know how it comes out.

As with most statistical tests, one could explain away a result, positive or negative, by introducing non-genetic factors in apparent pulchritude, such as how much a woman can afford to spend on clothes or how much she cares about her appearance, and arguing that they correlate, positively or negatively, with characteristics of colleges that correlate with average SAT. My proposed experiment, like most, would produce evidence, not proof.

I no longer remember, possibly never knew, how the actual study I heard about was done.

I would expect the correlation to be higher and the women more beautiful in polygynous societies and in societies with a high level of income inequality, low average income and enough social mobility to produce a correlation between intelligence and income/status. Testing that conjecture would be harder since it would require objective tests of pulchritude that could compare women of differing ethnicities. One could probably rig the results to support the superiority of your preferred nationality by suitable ethnic selection of male evaluators.

Apparently there is no such correlation amongst men, possibly because women, unlike men, do not put a high value on the looks of potential mates or because they have, or at least in the past had, less control than men over who they marry.

Rainbow Washing the Past

Several of the Doctor Seuss books have been criticized, in some cases withdrawn from publication, for containing what critics have described as racist language, such as the term “chinaman” instead of “Chinese man.” Roald Dahl’s books have been edited by the publisher to replace “enormously fat” with “enormous” and give a female character a higher status  job. The Story of Doctor Dolittle was revised because of, in the words of the author’s son, “certain incidents depicted that, in light of today's sensitivities, were considered by some to be disrespectful to ethnic minorities and, therefore, perhaps inappropriate for today's young reader." The James Bond novels are being revised for reissue with the help of sensitivity readers. The Adventures of Huckleberry Finn has been criticized for containing the “N-word,” in one case edited to remove it “to encourage instructors who would otherwise avoid the book to reintroduce it in their curriculums.”

When an edited version of H.L. Mencken’s diary was published — in deliberate violation of the terms under which he had left it to the library that published it —  various people, including the editor, accused him of racism and antisemitism, largely because he sometimes used ethnic labels for people he referred to, described Lawrence Spivak, for example, as “a young Harvard Jew.” Which he was. As should have been clear from the diary, Mencken was less inclined to judge individuals by race, ethnicity or status than most moderns, probably including his critics:

In the diary Mencken mentions going to visit a friend who lived some distance from the railroad station. The friend sent his chauffeur to pick Mencken up. Mencken found that the (black) chauffeur, although uneducated, was an intelligent man and had a very interesting conversation with him, finding out how the world looked from his point of view. On a second visit Mencken was looking forward to another conversation with the chauffeur. To his disappointment, there was a second guest, a white woman, and in her presence the chauffeur remained silent.

Mencken was a famous, influential, and comfortably well off man. There is no hint in his account of the incident that he felt as though that made him superior to the chauffeur. What mattered was not race, income, education, or status but that the chauffeur was an intelligent person with something interesting to say. (From an earlier post)

So far as I know, nobody has yet tried to publish expurgated versions of Mencken’s writing.

The attempt to eliminate evidence of past writing not currently politically correct is unlikely to succeed in purging the libraries but it might succeed in purging many classrooms, both K-12 and college. What effect will that have of people’s view of the past? The question occurred to me at Baycon, a Bay Area science fiction convention, when a panelist made it clear that she had been surprised, reading “Peter Pan,” to discover that the author did not have modern attitudes on race and gender.

One effect may be to make it easier for people to believe that present-day America is a particularly racist and sexist society when, by any reasonable historical measure, it is precisely the opposite.

Contra TANSTAAFL

Tanstaafl, “There ain’t no such thing as a free lunch,” is a popular slogan with, among others, libertarians. I do not know where it originated but it was popularized by Robert Heinlein, who used it in The Moon is a Harsh Mistress. My father eventually stopped using it on the grounds that it was not true.

Consider consumer surplus, your gain from buying something at a price below what you were willing to pay. If I buy a 2 liter bottle of Coke Zero for two dollars when I would have been willing to pay four, I have gained two dollars of consumer surplus. That is the value to me of the opportunity to buy that product at that price which, unlike the Coke Zero, I did not pay for, hence a free lunch. If I give a talk for a thousand dollars which I would have been willing to give for four hundred that is producer surplus, a gain to me of six hundred dollars due to the existence of someone who valued my talk at a thousand. I did not pay anything to have him exist. A free market society is, in that sense, full of free lunches.

My father replaced Tanstaafl with “Always look a gift horse in the mouth.” Free lunches exist but the claim that something is a free lunch should be viewed with suspicion.