Tipping Point

The Economist explains why blockbusters get better ratings than niche products:

Although you might expect people who seek out obscure products to derive more pleasure from their discoveries than those who simply trudge off to see the occasional blockbuster, the opposite is true. ... blockbusters get better ratings from the people who have watched them than more obscure ones do. ...

A disproportionate share of the audience for a hit was made up of people who consumed few products of that type.  A lot of the people who read a bestselling novel, for example, do not read much other fiction. By contrast, the audience for an obscure novel is largely composed of people who read a lot. That means the least popular books are judged by people who have the highest standards, while the most popular are judged by people who literally do not know any better. An American who read just one book this year was disproportionately likely to have read "The Lost Symbol", by Dan Brown. He almost certainly liked it.

... As a media product moves from the pool of frequent consumers into the ocean of occasional consumers, the prevailing attitude to it—what Hollywood folk call word of mouth—can become less critical. The hit is carried along by a wave of ill-informed goodwill.

I've seen connoisseurs of all kinds of things complain about how the masses uncritically accept whatever is popular, but I'd never seen a systematic explanation of why that would usually be the case. 

It is clear that, in order to make a big hit, you have to make something that everyone will like a little, rather than something that some people like a lot.  Hollywood, with its business model of pandering to the masses and then driving things with marketing, has known this for years.

'Self Knowledge' Review Part 2

Earlier, I talked about a hundred-year-old sex education book that I got at a yard sale.  This is a continuation of that review.  I will now discuss how he got the science of heredity wrong.

We now know that all genetic information is unchanged by anything that happens in the life of the organism*.  Changes in a species come through mutation and natural selection.  If the carriers of a gene have more fertile offspring, then the gene will become more common.  No individual can do anything to change the genes that he or she will pass onto the next generation.  Good heredity, or eugenics, means finding a mate with the best genes.

But 100 years ago, this was not common knowledge.  Many people had a Lararckian idea of inheritance.  They thought that things that happened in an individual's life were commonly passed on to offspring.  This belief was often mystical.  The author believed that children would inherit most of the physical, mental, and moral attributes of their parents.

His working definition of 'heredity', which is never stated in the book, is 'anything that happens to a child before the child is old enough to make decisions'.  His definition of 'eugenics' is 'anything that improves heredity'.  This is a much broader definition than the 'selective breeding' definition that we use.  He talks about eugenics as something that couples can start to do even after they get married, and gives advice on what to do.

Here are some of the things that he puts in the category of 'eugenics':
1) Making sure that the father has not been using alcohol, tobacco, or drugs for a good length of time before conception
2) Making sure that the couple does not have sex unless they would welcome a child
3) Giving the mother good food and a healthy environment during pregnancy
4) Ensuring that the infant receives proper love, care, and attention

Basically, anything that we would call prenatal care or early childhood care is included in his definitions of heredity and eugenics.  And the book is full of good advice that matches modern research.

He got the mechanism wrong.  We now know that the developing minds of children are heavily influenced by the environment in early life.  A good home environment will generally produce good kids, while a bad environment will put the child at a huge disadvantage.  Kids, even ones too young to talk, can tell when they are not wanted, and that knowledge can destroy them.

But he knew nothing about the cognitive abilities of infants.  He saw the connection between the behavior of parents and later success in life, and assumed that the mechanism was either mystical or biological.  He thought that he was being scientific, but without a good understanding of causality and basic facts, most of what he said ended up being folk wisdom and superstition.

Sometimes folk wisdom is confirmed by scientific research.  Sometimes it is not.  While much of what he said was good and useful, some of it was bizarre, wrong, and possibly dangerous.  I will cover that in a later review.  It is always useful to read something like this with a modern eye.  By spotting his mistakes and seeing where they came from, you are less likely to believe similar semi-scientific stuff produced by modern writers.

You do see hints that he was moving beyond the culture of the time.  When discussing heredity, he made a passing mention of Jews, mentioning how they cared about good heredity.  The implication was that they were a superior race of people.  This was in a culture of chronic antisemitism.

Jews do have, on average, IQ's of about 5 points above average.  I don't want to digress into a discussion of IQ as a valid measuring instrument, but it is correlated with success in life, and it is influenced both by genes and early childhood development.

And here I stumble across one of the most politically incorrect themes in science.  It is almost impossible to have an intelligent conversation about human biodiversity.  Liberals hate admitting that one group of people might better or worse than average in any way, and conservatives hate admitting that genetics might have a big impact on how people act, and their success in life.

The facts show that people are different for a variety of reasons, some of them genetic.  The mind is affected by natural selection just like any other part of the body, and different groups of people have lived in very different environments.  It shouldn't be too hard to talk about these differences, while still holding firm to a belief that all humans are valuable and deserve basic human rights and an equal chance to succeed.

But it is hard, and one of the reasons it is hard is because eugenics and talk of racial differences has a very bad connotation.  People have used it as an excuse to commit some very nasty crimes.  I'll mention that in a later review of this book.

* There are a few transgenerational epigenetic effects, but these don't seem to have much impact on heredity.

Financial Literacy Fail

One of the questions on this week's homework in my Econ class was:

Your uncle repays a $100 loan from Tenth National Bank by writing a $100 check from his TNB checking account. Use T-accounts to show the effect of this transaction on your uncle and TNB. Has your uncle's wealth changed? Explain.

A disturbingly large number of the students said that the uncle's wealth goes down by $100.  Even the A students seemed to think that paying off a loan makes you poorer.  This is truly frightening.  We have a generation of people who do not seem to understand that a loan is negative wealth.  When I mentioned it to my office mate, he replied that this is a symptom of a flawed education system, and I agree.

In future classes, I will probably spend more time on basic financial literacy.  I designed my curriculum this semester to focus on public policy issues and help the students be informed voters.  But it is probably more important for them and the country as a whole to make sure that they understand basic facts about money.  Teaching this stuff is not technically the purpose of the course, and I shouldn't have to do it, but if I don't do it, then nobody will.

How People Really Make Decisions

Most people seem to live their lives according to the the following rule set:

1) Do whatever the happiest-looking people around you are doing.  If that makes you happy, keep doing it.  If not, proceed to Step 2.
2) Do something else at random. If that makes you happy, keep doing it.  If not, return to Step 1.

Note that this process will usually end up producing a situation where it appears as if everyone has carefully examined all of the possible options and chosen the one that makes them happiest.  So even though the 'utility-maximizer' model of economic agents has nothing to do with how people think, it is a good first approximation of the world in most situations.

Violence and Sexual Coercion

As a result of discussion in yesterday's workshop, I ended up coming across this survey of Sexual Coercion among University Students in the United States and Sweden.  It is not an economic study, but it still brings up some useful data and is worth discussing.

They surveyed college students in each country, when meant, bluntly, that they were sampling rich White kids in each place.  So we can assume that the demographics are roughly identical, and that we are measuring cultural differences among people of similar socioeconomic status.  Surveys are notoriously unreliable, which is why economists don't like using them, but let's just assume that the unreliability is the same in both places and the differences are real.

The survey shows that, in the United States, about 50% of males and 70% of females report that they had been subjected to nonphysical coercion (constant arguments, lying, intoxication) to engage in sexual activities.  In Sweden, the numbers were about 20% for males and 40% for females.  Basically, Swedes of both genders are 30% less likely to be the victim of sexual pressure, and women in each country are 20% more likely to be the victims.

The survey really should have asked if people did these things, and how many people they did them to.  That way we would know if the increased victimization rates were the result of more people doing these things, or of the perpetrators doing them to more people. But the following analysis holds in either case:

Economic theory says that people will use a tool if the benefits are higher then the costs.  If a tool is used more often, it could be because the costs are lower, or because the benefits are higher.  Given that there are no legal sanctions for nonviolent coercion in relationships, the costs of using coercion are either psychic or social.  Either you feel bad about doing it, or word of your actions gets out and leads to ostracism.  The benefit of this behavior is, presumably, a higher chance of having sex.  Even if the costs were exactly the same among two different groups of people, you would see differences in the activity if one group places a greater value on sex, or if these actions were more effective at obtaining sex.

So we have four theories that might explain the observed differences:

1) Swedes feel worse about using coercion.
2) Swedes are more likely to socially punish people who use coercion.
3) Americans desire sex more.
4) Americans are more willing to give in to pressure to have sex.

Their survey cannot tell us which of these four theories is correct, but a properly designed survey could.  You could test theory 1 by asking people how they felt about their actions.  You could test theory 2 by asking how they would respond to a friend who used these tactics.  You could test theory 3 by asking people about their level of sexual desire.  You could test theory 4 by asking about how people respond to pressure.

But they don't do this.  They just speculate and say that 'more research is needed' without making any useful suggestions for how to do that research.  They seem to be working mainly in the framework of Theory 1, talking about how socialization and education make people less willing to use coercion and force.  Theories 3 and 4 are not mentioned at all.

There is actually a fifth theory that could explain the results.  Sexual coercion is only used when one partner wants sex and the other does not.  So the differences could come entirely from sorting mechanisms.  To see what I mean, imagine two societies with identical people and identical costs and benefits of sexual coercion.  In each society, half the men and women want sex, and half do not.  In one society, people openly announce their sexual desire and use that to choose partners.  In this case, all the men and women who want sex pair up with each other, as do the nonsexual ones.  There is no sexual coercion in any relationship.  In the second society, people do not announce their sexual desire and partner assignment is random.  In this case, 25% of the relationships will be between two sexual people, 25% between two nonsexual people, and 50% of the relationships will have one person who desires sex and one who does not.  In this society, we will observe much more sexual coercion, even though the people are identical. 

The survey also uncovers interesting data about physical violence.  In the US, approximately 28% of men and 35% of women report using physical violence ( throwing things, roughly shoving, hitting) against their partners.  In Sweden, the numbers are about 15% for men and 25% for women.  So Swedes of both genders are about 10% less likely to initiate violence, and women in each country are 10% more likely then men to initiate violence against their partners.

So it seems at first glance, that women are more likely to use physical violence and men are more likely to use nonviolent coercion.  But the question about nonviolent coercion only related to sex.  They should have also asked how likely people were to use nonviolent coercion for nonsexual reasons.  That way we could separate the type of coercion used from the desire for sex.  They do ask how often men use physical violence to obtain some kind of sexual contact, and the number is about 16% for US men and 4% for Swedish men.  Women almost never use physical force to obtain sexual contact, even though they seem more likely to use physical force in other contexts.

This all shows why scientists must be guided by theory and must be very careful about the questions they ask.  This survey tells us very little about the differences between men and women, or why Sweden is different than the USA.  The only thing it really tells us is that Sweden is a much better place to be in a relationship or to go out on a date.  But without knowing exactly why, we really don't have a useful guide to making things better here.

Celebrity Chef School Lunch

The Financial Times reports that a British chef (Jamie Oliver, or 'The Naked Chef') accidentally created a random experiment in school lunches.  One school system implemented his recommendations for healthier school lunches, and other similar ones did not.  Preliminary results are that the new food reduced absences and increased test scores.  In other words, healthy foods had an immediate and significant positive effect on school outcomes.

The columnist makes another important point.  We need to be running more experiments like this:

Obviously that discovery is important in its own right. Jamie Oliver was correct to emphasise the importance of feeding schoolchildren good food. But the whole episode matters for another reason. Too often, critical scrutiny of what works and what doesn't in our society has been replaced by a pure emotional response. ...
Surely what counts is that a new idea was tried out on a respectable scale, and now we have a chance to figure out whether it worked. What astonishes me is that it took a television company and a celebrity chef to carry out a proper policy experiment.

But really, we already knew that healthy food is good for you.  The big question, that any responsible parent has dealt with, is: "How do you get kids in the habit of eating healthy when we are surrounded by a culture of obesity?"  When they are really young, you can easily control the menu by fiat, but when they get to the preteen years you have to instill some self-control.  A perfectly healthy school menu does nothing if the kids ignore it and smuggle in junk for lunch.

It seems to me that the best way would be to constantly tell them that unhealthy foods are childish, and that healthy foods are the mark of grown-ups, and that anyone who insists on eating unhealthy stuff is a baby.  That would work especially well on that age group: "If you want to be treated like an adult, then you should eat like an adult."

But that strategy only works of parents do not eat junk.  Kids are the ultimate hypocrisy detectors.   They copy what we actually do, not what we say.

Freedom of Expression

The CATO institute recently published a good article showing that the pain of seeing people do things that offend you is not a good reason to impose any controls on of free speech.  It argues that it is dangerous to pass any laws that prevent things like flag burning.

This is a good snarky quote, showing how flag burning might actually be good for people who love the flag:

Venerators' revenge may take many lawful forms, not the least of which is simply the satisfaction (profound at times) of thinking that the  desecrator is a twit, and of feeling oneself morally superior to him.

I definitely agree that the Republican party's perennial attempts to introduce a 'flag burning amendment' is wrong and dangerous.  It would set a horribly bad precedent of attacking our basic First Amendment rights, the ones that have been so important in keeping our society free and healthy.

I also think that the stance is politically damaging for them.  All intelligent social conservatives know that the Constitution is far more sacred than the flag, and so should be treated with more respect.  The Republicans are showing their willingness to mess with our founding document for the sake of political expedience.