Tuesday, March 22, 2011

Power

It makes no sense for people to make so much fuss about the nuclear power plants in Japan. The earthquake and tsunami killed over ten thousand people, and the nuclear plants have not killed a single person. This incident actually makes me feel better about nuclear power; those ancient plants survived a really nasty string of disasters that would not happen in the USA, and modern plants are far safer.

The numbers suggest that the Japanese spent too much money on nuclear safety. If they had spent that money on better tsunami protection, it could have saved hundreds or thousands of lives.

Coal mining kills hundreds of people each year, and the pollution from burning coal kills thousands more. The world would be a much better place if coal plants were replaced by nuclear power plants. Any yet the public paranoia about nuclear stuff makes it impossible to buy insurance for nuclear power plants, which kills investment in them.

Of course, if these numbers are right, building any large power plant now would be a mistake. In one or two decades, solar panels may be so cheap that it will never make any sense to run a power plant during sunny days.

I do not think that the future will have all households making their own power. It will certainly be much easier to live off the grid, with solar and biofuels, but it will still be more efficient to buy your power from a specialized provider. But specialized does not mean centralized. A lot of power is lost in transmission. Instead of big power plants, utilities will, if solar gets cheap enough, scatter solar cells all over the place, so the generation is close to the consumer and a single cloud will not reduce the power too much.

Traditional power plants will mainly provide supply at night and during cloudy days, and to meet demand spikes. For that you need a plant that can be turned on and off quickly, like natural gas. Nuclear power plants only make sense when they are run constantly, providing a steady supply of power.

Basically, In a world of cheap solar panels, nuclear power makes little sense. This is especially true if biofuel production technologies advance, as it seems they will. All issues surrounding nuclear power could become history in a generation.

Monday, March 14, 2011

Secret Fears of the Super-Rich

He found that the rich—especially the inheritors of vast fortunes—have unique sets of worries, and face the added difficulty of knowing that many despise or envy them. "Often the word rich becomes a pejorative," Kenny says. "It rhymes with bitch. I've been in rooms and seen people stand up and say, 'I'm Bob Kenny, and I'm rich.' And then they burst into tears."

The respondents turn out to be a generally dissatisfied lot, whose money has contributed to deep anxieties involving love, work, and family. Indeed, they are frequently dissatisfied even with their sizable fortunes. Most of them still do not consider themselves financially secure; for that, they say, they would require on average one-quarter more wealth than they currently possess.

Such complaints sound, on their face, preposterous. But just as the human body didn't evolve to deal well with today's easy access to abundant fat and sugars, and will crave an extra cheeseburger when it shouldn't, the human mind, apparently, didn't evolve to deal with excess money, and will desire more long after wealth has become a burden rather than a comfort. A vast body of psychological evidence shows that the pleasures of consumption wear off through time and depend heavily on one's frame of reference. 


There seem to be two distinct sets of problems with being rich. One set comes from society. Other people treat you differently, and often worse, if you have lots of money. This is the kind of thing Ayn Rand complained about at length, along with the claim that such attitudes would result in the downfall of civilization. She takes things to an extreme, but it is true that if society does not respect wealth creation then we will get a lot less of it. People who earned a lot of money honestly* should never be ashamed of it, and if they are, that means something is wrong with our culture.

The second set of problems is psychological. Sometimes it can be hard to disentangle this from cultural conditioning, but it seems that money really does not make people happy. The most it can do is solve almost all of the problems that the outside world throws at you. That is not enough to satisfy most people, and it may even lead to dissatisfaction, to the extent that humans seem to have a natural impulse to be problem-solvers:

 Work is what fills most people's days, and it provides the context in which they interact with others. A life of worklessness, however financially comfortable, can easily become one of aimlessness, of estrangement from the world. The fact that most people imagine it would be paradise to never have to work does not make the experience any more pleasant in practice.

The problems of the super-rich are a small subset of a larger social problem. People in rich developed countries live in an environment that is utterly alien to the workings of the neurotypical human brain. The only people who will be truly comfortable with this environment are people whose minds are naturally abnormal, or those who have painstakingly reprogrammed themselves with the right kind of philosophy. Everyone else will be as nervous and confused and neurotic as a poodle in a mansion.

Compared to the miserable existence of most people throughout history, this is a good problem to have, but it is still a problem, and I have no clue how it might be solved on a societal level.

* 'Honestly' here means via any kind of voluntary exchange, anything does not involve politics, government contracts, fraud, or crime.

Friday, March 11, 2011

Chess and Computers

A big question about the future is what computers will do to human labor and economic activity. Here's a good article that looks at what computers have done to chess players. This is probably a good guide to what will happen with expert computers more generally.

Sample quotes:

Chess is an area where educational reform has been extremely rapid and extremely successful.  Chess education today revolves around learning how to learn from the computer, and this change has come within the last ten to fifteen years.  No intermediaries were able to prevent it or slow it down.  Humans now teach themselves how to team with computers, and the leading human players have to be very good at this.  The computers which most successfully team with humans are those which replicate most rapidly.

Chess-playing computers still are not meta-rational.  They do not understand what they do not understand very well, for instance blocked positions and long sequences of repetition.  That is one reason why human-computer teams are so important and so productive. 

Wednesday, March 9, 2011

Education Reform

My dad, a high school teacher, has been saying this kind of thing for years:

 advocates for better education have repeatedly latched on to a depressing litany of fads as the panacea for what ails American education. Believing that they have only a short window of opportunity for change, these reformers push for their ideas to be applied uniformly across the board. "New math," standardized testing, centralization, merit pay, small schools, community control, mayoral control, and dozens of other ideas have ripped through schools, often with disappointment and disillusion not far behind.
Many of these ideas actuallydid have some merit, says Hess, in the sense that they could help some specific students in some specific circumstances. For example, a rigorous focus on a narrow set of tested subjects may be reasonable for schools in chaotic, urban contexts where simply focusing on anything counts as success. But that treatment, like chemotherapy, has powerful side effects that should not be risked on the (relatively) healthy "patients" in more advantaged school districts.


There is no best way to run a school, just as there is no best way to run a restaurant. Different people need different kinds of services. Nobody can design a single plan that works for everyone. Instead of centralized rules, we need a system that gives people freedom to find better ways to teach.

Of course, pure freedom without any responsibility is a recipe for disaster. The output of schools has to be measured somehow. Standardized tests are one option, and are certainly better than nothing, but they still have problems. There is a debate over what exactly schools should do, and most people agree that they should do more than just boost test scores. I think most parents mainly want school to improve their child's future, both in terms of income and quality of life. There should not be too much disagreement on that. 

Therefore, the optimal, incentive-compatible solution would be one that gives money to schools based on the quality of the student's life after school.   If the child gets a good job and is happy and avoids crime and other social ills, the school should be rewarded. If not, the school should get nothing.

Of course, if done wrong, this scheme would result in schools just trying to enroll rich, well-adjusted kids. You have to correct for innate ability and home circumstances to find the improvement that is due to the school.  So at a young age, the government would collect data on children and their parents, measuring things that affect future life outcomes, like parents' income and household situation and the child's health status, IQ, and self-control. Then, these numbers are plugged into a computer model that predicts the child's future income and other measures of quality of life. 

After you get those numbers, you can measure what each school system actually does. If the child's life is better than what the model predicts, the school gets government money. If it is worse, they do not. Private schools and public schools should be treated the same, and parents allowed to choose the school they like the best. This kind of plan would probably create a lot of specialized schools, for example ones that are designed to work with low-IQ children and prepare them for decent jobs.

There are a lot of practical details to be worked out here, and I don't think my readers would be interested in the specifics. But they can be worked out. Whenever you have a problem, a good solution usually starts with thinking about exactly what you want, and then giving people a string incentive to produce those results.

Rural people and politics

An economist recently wrote a book talking about how big cities make people richer and more productive. The key idea is that ideas and efficiency and economic growth come from connecting people, and dense clusters of people make these connections much more likely to happen. Living in big cities also trends to be much more efficient: transportation and search costs are a lot lower, and city dwelling generates less pollution and much less habitat destruction than country dwelling.

The key policy implication is that we should support and encourage life and economic activity in cities in order to make our nation richer. If done right, you get things like the renaissance of New York City. At the very least, we should stop taxing city dwellers in order to subsidize rural areas and lifestyles that tend to trap people in poverty.

Here's a link to a blog post discussing the book.


Yesterday afternoon, I got an e-mail from a "usda.gov" address. "Secretary Vilsack read your blog post 'Why we still need cities' over the weekend, and he has some thoughts and reflections, particularly about the importance of rural America," it said. A call was set for a little later in the day. I think it's safe to say Vilsack didn't like the post. A lightly edited transcript of our discussion about rural America, subsidies and values follows.

Read the whole thing. It is amusing and horrifying, like a train wreck. The secretary of agriculture tries to defend current policy, but just makes himself look like an idiot. The thing that amazed me was that, in his opening comments, he said:

And sometimes people don't realize that 90 percent of the persistent poverty counties are located in rural America. 

which is exactly the point that economists try to make: rural life often leads to poverty and we need to get people out of that situation. If you are trying to defend the merits or rural life, why would you go out of your way point out one of its biggest problems?

This blog post provides good commentary on this issue, I would have written something similar but linking to this is easier.

Tuesday, March 8, 2011

More on Placebos

I just came across this:

Kaptchuk tested the effect of placebo versus no treatment in 80 people with irritable bowel syndrome. Twice a day, 37 people swallowed an inert pill could not be absorbed by the body. The researchers told participants that it could improve symptoms through the placebo effect.
While 35 per cent of the patients who had not received any treatment reported an improvement, 59 per cent of the placebo group felt better. "The placebo was almost twice as effective as the control," says Kaptchuk. "That would be a great result if it was seen in a normal clinical trial of a drug."

I followed the link to the journal article and found exactly what they said to the patients:

Patients were randomized to either open-label placebo pills presented as "placebo pills made of an inert substance, like sugar pills, that have been shown in clinical studies to produce significant improvement in IBS symptoms through mind-body self-healing processes" or no-treatment controls with the same quality of interaction with providers.

The phrase 'mind-body self-healing processes' is suspiciously similar to what you hear thrown around in ads for quack cures. I wonder if this was intentional. Did they copy buzzwords from quack ads, the same way that Forer copied horoscopes for a psychology demonstration? 

This study helps explain why those quack cures are so popular: they actually do work for any condition marked by a subjective experience. With this study you could legally sell sugar pills for IBS relief, as long as the ad mimicked what these doctors told the patients.

Of course, the smart thing to do is skip the middleman and convince yourself that exercise or yoga or meditation or something similar will cause a dramatic boost in your health and make you feel better and more energetic. If you believe it, then it will be true.

Placebos

Recent research, he said, showed that placebos had helped 59% of patients who had been suffering from an upset stomach. Used to treat depression, placebos have the same effect as antidepressants in about a third of cases.
The efficacy of a placebo depends on many factors, according to the report, including the size and colour of a pill.
The more expensive the placebo, the higher the success rate, the study found, and intravenous injections are shown to be more effective than oral medication.


This kind of thing supports assertions that a lot of medicine has more to do with psychology than biochemistry. Going to doctors does not do much for you, compared to things like diet and exercise and avoiding pollution. The main benefit people get from doctor visits is feeling that someone cares about them.

Monday, March 7, 2011

Inflation

After writing my last post on nickels, I felt the need to explain what economists know about inflation and what will probably happen with inflation in our country. There is a lot of rubbish floating around the popular press, and I'd hate for any of my friends or family to be infected by it.

An important lesson from basic economics is that the money supply in the country depends on more than just how much the government prints. The amount that comes off the printing presses is called the money base. That is just a small fraction of the money supply. The rest of the money supply comes from bank lending. The banking system can generate $10 in loans for every $1 in cash in their vaults, and all of those loans end up counted as money. For example, when you take out a mortgage to buy a house, all of the money goes into the bank accounts of the construction workers right away, and then they can start spending it.

Inflation depends on the total money supply, including the money 'made' by banks as they make loans. In other words, banks can cause inflation by excessive lending, and they can cause deflation by lending less than usual. This action of banks is called the money multiplier, because it multiplies the base to produce the actual money supply.

When the economy is functioning normally, the money multiplier is constant. There will always be inflation if the government prints more money. But things are not normal now. The financial crisis caused the money multiplier to fall sharply. The resulting 'credit crunch' is exactly the same thing as 'deflation'. It was right and proper for the government to print lots of money to compensate. If anything, they did not print enough.

The best measure of the total money supply is called 'M3'. It is not perfect but it is the best number we have. It tries to take all the actions of the banking system into account. Inflation hawks rightly cried foul when the Fed stopped publishing M3 in 2006. Here is a graph of the best estimate of M3 over the last few years:


Note that this is not the inflation rate. Inflation is the growth rate of the money supply minus the growth rate in real output. If we make 5% more stuff, and M3 grows by 5%, then there will be no inflation at all, because the ratio of money to stuff stays the same. The country would be 5% richer and prices would be the same.

Still, this graph shows that there was a lot of inflation in 2006 and 2007, while the housing boom was still in progress. The reason that this inflation did not show up in the news was because it did not affect the things used to calculate the Consumer Price Index. It only showed up in housing and real estate costs. But if you wanted to buy a house, this inflation would have really hurt your purchasing power.

When the financial crisis hit, the growth rate of M3 started falling. By the middle of 2009 the growth rate went below zero and the level of M3 started falling. (The graph shows, at each date, the average growth rate for the previous year.) This means deflation. M3 is still falling. The economy has been experiencing deflation for the past couple of years, even though the government has printed a lot of money. That is one of the reasons the economic recovery has been sluggish.

A lot of people are afraid that there will be inflation in the future. It is true that the government has massively expanded the money base. But they have plenty of tools to take that money out of circulation again, shrinking the base back down to normal as the multiplier goes back up to normal. They have demonstrated a willingness to fight inflation, even at the cost of massive unemployment. If they did not inflate in the past two years, when things were bad, they will not start any time soon.

Of course, inflation is not impossible. It might happen. But I think the odds of American hyperinflation, or even of moderately high inflation, are very low. For any time frame you care to mention, the odds that intelligent computers will cause massive economic disruption are greater than the odds that inflation will cause massive economic disruption. There is no reason to be concerned about inflation, or to spend any time or money preparing for it, or to vote for politicians who make noise about it.

Nickel Hoarding

I honestly think that this will work

US five cent coins contain over 7 cents worth of raw material as of this afternoon, mostly copper and of course, nickel. If there is inflation, the prices of metal will increase, and the coin will have 8, 9, 10 cents worth of metal. Pre-1965 dimes contain over $2.42 of metal today, while pre-1965 quarters have over $6 worth of metal. (www.coinflation.com)
If there is deflation, the coins are still currency at face value. They will always be nickels.

Hoarding nickels seems like a good investment for people without a lot of cash and with plenty of time on their hands. Each time you go to the bank, get about $20 worth of nickels. Store them in your house somewhere. There is zero downside risk. If everything is fine and metal prices fall, you can just spend them. If raw materials prices go up, as many people predict, you will win.

The reason this is not more popular is because of transactions costs. You have to lug around the coins. And you could not really get seven cents by selling them for scrap; the costs of separating and extracting the metal would be too high. But history tells us that coin collections tend to keep their value pretty well.

One word of warning: I do not recommend buying silver or gold coins now. I actually recommend selling gold or silver if you have any. And yes, prices have gone up since I said that, but I still think it is a bubble. Prices could easily collapse. But with nickels you do not have that problem.

Human Echolocation

This is awesome. It may ask you to print. Click 'Cancel' if it does. Web pages nowadays are so cluttered with junk that the only way to get a decently formatted article is to click on the 'printer friendly' link.

Daniel Kish has been sightless since he was a year old. Yet he can mountain bike. And navigate the wilderness alone. And recognize a building as far away as 1,000 feet. How? The same way bats can see in the dark.

Kish and a handful of coworkers run a nonprofit organization called World Access for the Blind, headquartered in Kish's home. World Access offers training on how to gracefully interact with one's environment, using echolocation as a primary tool. So far, in the decade it has existed, the organization has introduced more than 500 students to echolocation. Kish is not the first blind person to use echolocation, but he's the only one to meticulously document it, to break it down into its component parts, and to figure out how to teach it. His dream is to help all sight-impaired people see the world as clearly as he does. 

It was never Kish's goal to run a foundation dedicated to the blind. He planned to be a psychologist. But he could not ignore the fact that few blind people enjoyed anything close to his freedom of movement, and he had grown weary of society's attitude toward the blind. "I am belittled, patronized, disrespected, invaded, restricted, and presumed weak, vulnerable, or otherwise incapacitated," he wrote in his journal. It still drives him crazy when he's congratulated for simply crossing the street or preparing dinner.
...
Young people, says Kish, are especially hard-hit. "Most blind kids hear a lot of negative talk. 'Don't do this, don't do that, don't move. No, here, let me help you.' The message you get, if you're blind, is you're intellectually deficient, you're emotionally deficient, you're in all ways deficient." A few sighted people have commented to Kish that they'd rather be dead than blind.
So in 2001 he started World Access for the Blind. One of its missions is to counter every no that blind people hear. Blindness, Kish says, should be understood — by both the blind and the sighted — as nothing more than an inconvenience. "Most of my life," he writes, "I never even thought of myself as blind. In fact, I saw myself as smarter, more agile, stronger, and generally more capable than most other boys my age." 

World Access doesn't turn anyone away for lack of resources. But there are a couple of reasons why the organization hasn't trained more students. The first is Kish's general ethos about how blind children should be raised. "Running into a pole is a drag, but never being allowed to run into a pole is a disaster," he writes. "Pain is part of the price of freedom.

Emphasis mine. That is an important bit of wisdom for anybody:

Friday, March 4, 2011

Cognitive Modification

This is interesting and scary:

Cognitive-bias modification (CBM) is based on the idea that many psychological problems are caused by automatic, unconscious biases in thinking. People suffering from anxiety, for instance, may have what is known as an attentional bias towards threats: they are drawn irresistibly to things they perceive to be dangerous. Similar biases may affect memory and the interpretation of events. For example, if an acquaintance walks past without saying hello, it might mean either that he has ignored you or that he has not seen you. The anxious, according to the theory behind CBM, have a bias towards assuming the former and reacting accordingly.
The goal of CBM is to alter such biases, and doing so has proved surprisingly easy. A common way of debiasing attention is to show someone two words or pictures—one neutral and the other threatening—on a computer screen. In the case of social anxiety these might be a neutral face and a disgusted face. Presented with this choice, an anxious person instinctively focuses on the disgusted visage. The program, however, prods him to complete tasks involving the neutral picture, such as identifying letters that appear in its place on the screen. Repeating the procedure around a thousand times, over a total of two hours, changes the user's tendency to focus on the anxious face. That change is then carried into the wider world.

This is all very speculative, and it will be a while before they get it useful, if they ever do.  But would be great news if this technique works to help people.

The scary part is that this research shows how easy it is to alter people's thinking.  For example, consider a video game where you have to shoot zombies while avoiding shooting civilians.  Anyone who plays the game will learn to focus on the zombies, looking for them while ignoring the unthreatening people.  This is the exact opposite of the treatment presented in the article: it trains people to subconsciously pay more attention to anything that looks or acts threatening.  It will, according to this research, make them more anxious people.

Of course, if we have good research showing that video games have no such effects, then that makes us doubt the effectiveness of the therapy.  As far as I know, all of the research comparing video games to life outcomes shows that they have almost no effect on anything.  But that research does not look at exactly what type of game people are playing.  It may be that some games have good effects and some have bad effects.  Maybe games cause a mix of behavioral changes, some good and some bad.  Perhaps they make people more anxious but increase the ability to focus and solve problems.

What we would need in order to do proper research is a large-scale data collection project that looks at the actual titles of games people buy and play, and then tests them in various ways.  Of course, there is no reason to single out games.  You should look at the movies and TV shows they watch too.

If we ever do get a good, secure, electronic medical records system, then it should be merged with other kinds of data like this.  Grocery stores keep extensive records on the food that people buy, and it would be fairly easy to link that to medical records.  Your Netflix viewing history could be easily merged, as could the records of what you spend with your credit cards.  Then public health researchers could start crunching numbers.  Maybe they would find that people who play a certain video game are more likely to be anxious, or that people who watch certain movies are more likely to suffer from depression, or that one particular food is especially harmful.  The benefits to people from this kind of knowledge would be massive.

Statistical Composition

It is very easy to make mistakes with statistics.  Averages can often produce the wrong impression.  For example, consider a situation where there are two schools ( 1 and 2 )and two groups of people ( X and Y ).

School 1 has 10 students from group X and 100 students from group Y.
Of those, 5 students from group X graduate and 80 students from group Y graduate.

School 2 has 100 students from group X and 10 students from group Y.
Of those, 70 students from group X graduate and all 10 students from group Y graduate.

It should be pretty obvious that, no matter who you are, you should go to School 2.  Group X has a graduation rate of 50% in school 1 and 70% in school 2.  Group Y has a graduation rate of 80% in school 1 and 100% in school 2.  School 2 is simply better for everyone.

But if you just look at the overall numbers, school 2 has 80 people graduate and school 1 has 85 people graduate, with the same number of students. School 1 looks better, even though it is actually worse.

This kind of thing cause all kinds of problems when people throw around statistics on things like schooling and health care.  For example, you may have heard that schools in Wisconsin, with their collective bargaining for teachers, are better.


The NAEP is an annual standardized test given to 4th and 8th graders around the country to measure proficiency in math, science, and reading. Participation is fairly universal; if you've had a 4th or 8th grader in the last few years, you're probably familiar with it. Results are compiled on the NAEP website, broken down by grade, state, subject and ethnicity.
So how does brokeass, dumbass, redneck Texas stack up against progressive unionized Wisconsin?
...
2009 8th Grade Math
White students: Texas 301, Wisconsin 294 (national 294)
Black students: Texas 272, Wisconsin 254 (national 260)
Hispanic students: Texas 277, Wisconsin 268 (national 260)
...(more test score lists)
...
To recap: white students in Texas perform better than white students in Wisconsin, black students in Texas perform better than black students in Wisconsin, Hispanic students in Texas perform better than Hispanic students in Wisconsin.
...
the gap between white students and minority students in Texas was much less than the gap between white and minority students in Wisconsin. In other words, students are better off in Texas schools than in Wisconsin schools - especially minority students.
Conclusion: instead of chanting slogans in Madison, maybe it's time for Wisconsin teachers to take refresher lessons from their non-union counterparts in the Lone Star State.

Tuesday, March 1, 2011

Drug Courts

Finally, government starts doing something smart: drug courts:

Participants undergo intensive treatment instead of prison. Judges receive special training. Rather than simply resolving a case and sentencing the offender, they preside over teams that include prosecutors and defence lawyers, police, treatment and job-training counsellors and case workers
...
A statewide study in Georgia found the two-year recidivism rate among drug-court participants was 7%, compared with 15% for those on probation alone and 29% for drug-users who served time in state prison.
...
In any event, such schemes not only help the participant, but save money. In Georgia a drug-court sentence costs over $10,000 less than a prison sentence ... drug courts produce $2.21 in benefits (reduced crime and costs of incarceration) for every $1 spent; expanding their reach to cover all arrestees would raise the level of benefits to $3.36.

Although technically, this should be classified as 'slightly reduce the amount of stupidity in policy.'  It should be obvious by now that putting druggies in jail is amazingly dumb.