Prediction Markets instead of Parole Boards

Parole boards determine whether prisoners may be released from their sentences before their sentence is complete. If my understanding is correct, the main consideration is the likelihood that the convict will commit another crime.

Parole boards do a terrible job. According to the Department of Justice, 49% of parolees abscond or are returned to prison. And of course there must be more who commit more crimes who aren't caught.

It is rumored that sex offenders, especially paedophiles, are much more likely to reoffend than non-sex offenders. The statistics support the opposite conclusion. But my understanding is that sex crimes, especially sex crimes against children, are very often not reported, and less often than that are resolved. So I guess we don't know what the true recidivism rate is.

One strategy some states have adopted to combat the threat of sex offence recidivists is to maintain a sex offenders register - a public list that sex offenders are placed on (with photo and address available online) - so that citizens can be aware of any paedophiles in their neighborhood. In some states, the sex offenders aren't allowed to live within 1000 feet of where children congregate, such as schools and day care centers.

When Wendy Whitaker was 17, she fellated a 15-year-old. 12 years later, the 29-year-old must vacate her home in case she molests the children at a nearby church daycare center.

The obvious explanation is that Whitaker is being crushed in the cogs of the gigantic machine that is government bureaucracy. But if we take this at face value - the government is genuinely unsure whether Whitaker is a threat to the children, and is willing to let her out of prison but not off the sex offender list just in case - then what we're dealing with is a question of how to predict the future. Will Wendy Whitaker fellate any more 15-year-olds?

When you have a future prediction problem, the best solution available is (in most cases, when some normal conditions like liquidity hold) prediction markets. I propose replacing parole boards with prediction markets on the likelihood of recidivism for each convict. The result should be less crime and less injustice for people like Wendy Whitaker (and Genarlow Wilson, etc.)

Are Libertarian Consequentialists Biased?

Followup to: A Case Against The Minimum Wage

One useful way to distinguish kinds of libertarians is to distinguish those who believe that initiating coercion is wrong in itself (like Ayn Rand) from consequentialists (who think that the initiation of coercion leads to bad results measured according to some other criterion, like utility or happiness).

A lot of libertarians are some mixture of the two. Could this could lead to some questionable consequentialism? Believing that initiating coercion is wrong is likely to bias you when considering the consequential effects of coercion.

In 2006, Daniel Klein and Stewart Dompe surveyed economists who were in favor of the minimum wage. Klein and Dompe wanted to know why the economists supported the minimum wage. Question Seven asked whether the economists believed the minimum wage was coercive:

In one manner of speaking, liberty is freedom from political or legal restrictions on one’s property or freedom of association. Subscribers to this definition are apt to say that the minimum wage law is coercive because it (along with concomitant enforcement) threatens physical aggression against people for engaging in certain voluntary, consensual acts (namely, employing people at sub-minimum wages). (Notice that even subscribers to this definition of liberty recognize that it does not by itself carry a policy recommendation; values other than liberty exist and might conflict with it.)

Q7: Please indicate which of the following options best fits your view of this semantic issue:

These were the results:

Answer	%
A. I agree that that definition of liberty is the primary definition of liberty, and in that sense the minimum wage law is coercive.	5%
B. I give some weight to that definition of liberty, but not primary weight; the minimum wage law is only coercive in a sense.	19%
C. I give little to no weight to that definition of liberty; the minimum wage law is not coercive in any significant sense.	50%
D. Other [please specify]:	24%
No response	2%

Klein and Dompe discuss the result:

Only five respondents assented to the primacy of the posited definition of liberty and that the minimum wage was coercive. Forty-seven (or 67 percent of those selecting A, B, or C) said they give little to no weight to that definition and denied that the minimum wage law is coercive in any significant sense. Furthermore, of the 23 individuals who selected “Other,” the vast majority wrote in comments which indicated strong reservations about that definition, if not outright rejection. Thus, 90-95 percent all minimum-wage supporters reject the primacy of the posited semantics, and about 65 percent reject any significant place for those semantics. We hazard to guess that a survey of minimum-wage opponents would yield a frequency ranking A > B > C, the reverse of what is found here. If so, disputants of the issue for the most part do not agree on what “liberty” and “coercion” mean. Since those conceptions relate directly to one’s understandings of “voluntary choice,” “the free market,” “intervention,” and other fundamental analytic distinctions and categories, the implication is that conceptual cleavages probably often separate how the two sides formulate and analyze the issue. (Emphasis added)

I also expect that a survey of minimum-wage opponents would find an opposite perception of coercion in the minimum wage.

So, someone's biased. Who is it?

A case could be made that the libertarians - the ones who think that the minimum wage is coercive - are biased. The minimum wage supporters are, presumably, just regular old consequentialists, or something close to that. They're just looking at the facts about the minimum wage and deciding what the effects are likely to be. But the perception of those worried about coercion is already tainted by their distaste for coercion.

Or is it the minimum wage supporters who are biased? Is there a bias to overestimate the effectiveness of government intervention? There just might be. And there's status quo bias, of course.

I have no conclusion. Do you?

A case against the minimum wage

Henry and Charles work at Big Bob's Burger Boutique in Lexington for $7 per hour. Henry is 40 years old. His life hasn't gone well - his wife left him a few years ago and he struggles with alcoholism. He's in financial trouble: he pays child support payments, he can barely afford his mortgage payments, and he hasn't saved anything for retirement. Charles is a 19-year-old university student studying biology. His life is going well - his grades are good, he has a steady girlfriend, and his parents are paying for his university fees so his income from the Burger Boutique goes mainly towards partying on the weekends.

You're a citizen of Lexington. Lexington is considering raising the minimum wage from $7 to $9. How will you vote? Won't you give poor Henry and Charles a helping hand?

 

In a spectacularly unlikely turn of events, your vote makes a crucial difference, passing the proposition 132,592 votes to 132,591. The minimum wage is increased to $9. Henry and Charles are excited. Maybe Henry won't have to foreclose on his home. Maybe Charles will be able to attend Burning Man this year. On Monday, Bob calls Charles and Henry into his office.

"As you know, the Boutique is open all day. But we make most of our money at lunchtime; we're just breaking even in the morning and evening. Well, we were breaking even. But with the new higher cost of labor, I'd be losing money if I kept the store open all day. I'm afraid we'll only be open at lunchtimes from now on. And so I need to let one of you go."

Bob pauses, looking at the diligent, good-looking college student and the tired, paunchy divorcee.
"Henry, I'm afraid I can't keep you on."

Henry tries to find another job, but the number of job-seekers in Lexington has jumped suddenly for some reason. He starts drinking heavily, loses his home, and now begs for change outside the Burger Boutique.

Charles makes it to Burning Man and has a great time.
 

In a spectacularly unlikely turn of events, your vote makes a crucial difference, defeating the proposition 132,592 votes to 132,591. Henry is sad but resolves to work harder in the hopes of a raise. Charles is outraged and protests against the callousness of Lexington voters.

Over time, Henry learns more about the burger business. When a new office complex opens in the neighborhood the following year, demand for burgers increases. Bob hires new staff and promotes Henry to manage them. Henry saves his home and hopes to open a Burger Boutique franchise of his own someday soon.
 

The usually stated purpose of the minimum wage is to help the poor. The point of the story is that the minimum wage can increase unemployment, and this effect is targeted towards the most vulnerable of minimum-wage workers. Furthermore, according to this paper from the Department of Health and Human Services, 71% of minimum wage workers have family incomes at least 50% higher than the poverty line - i.e. the not-poor. A disproportionate number are teenagers. When the minimum wage rises, some workers see an increase in income at the expense of other workers who lose their jobs. It is reasonable to expect that the ones who lose their jobs are likely to be minorities, women, the illiterate, the most unskilled - that is, the actually poor.

This is standard economics. Artificially raising the price of something reduces the amount that buyers want to buy.

The unemployment effect of minimum wage laws is very small at the low minimum wages governments have mandated. Of course, the higher the minimum wage, the bigger the unemployment effect. There was a significant Supreme Court ruling in 1923 where elevator operator Willie Lyons contested the minimum wage of $66 per month, which destroyed her $35 per month job. The court ruled the law unconstitutional, but minimum wage laws were later found to be constitutional in 1938. According to a survey run by the University of New Hampshire in 2007, 73% of economists believe that a 50% increase of the minimum wage would result in employment losses.

Laying off workers and not hiring them in the first place are not the only ways employers can save money when confronted with minimum wage laws. Neumark and Wascher (1998) found that "minimum wages substantially reduce training received by young workers".

Instead of supporting minimum wage laws, I suggest that those who want to use government power to help the poor should instead support the negative income tax, as described by Milton Friedman in Capitalism and Freedom. In short: through a negative income tax, your income is supplemented if it's sufficiently low. The higher your income, the less the supplement, up to some point where the supplement cuts out and you begin paying tax. The disadvantage of this idea is that it reduces the incentive to seek increased income, since an extra dollar earned results in less than an extra dollar received. Of course, this is a disadvantage of the income tax at every level and society seems to find this tolerable.

The USA has a negative-income-tax-inspired program called the Earned Income Tax Credit.

For an opposing view, see here, here, here and here.

Our future among the stars

The life you're used to and the body you control are not the full scope of the organism of which you are a part.

You are 100 trillion cells in each of 6 billion human beings. Your cousins, a thousand times as numerous, are cells comprising the tissues and organs that make ants work. The other 99.9% of you is bacteria, and you've been here for 4 billion years. You are older than the mountains, older than the asteroid belt.

And during this 4 billion years, you have never died. Not all of you. Not once. Some parts of you became damaged and were lost, but you carried on. If your parents hadn't made it, you wouldn't be here right now.

During this 4 billion years we transformed a hostile world into a comfortable one. Or it transformed us. We made the atmosphere breathable and we adapted to breathe it. We made fur, and later, we made other parts of us make fur for the human parts. And the human part of you made tools. Rocket tools that will be of use to all of us very soon. With the help of this human organ, your cellular descendants will leap off this mote of dust into the vast sea of resources beyond. A sea that can support us hundreds of trillions of times more richly than our current home. A sea for our home, to make liveable or to force us to adapt again, as Earth once did.

Rejoice, human, and keep moving forward! You're on to something very big for all of us, though you may not yet know it. We will all reap the rewards of your efforts, soon. - tuber from Reddit

I like tuber's expansive concept of his ingroup.

Did you know that the universe is enormous? Well it is. There are about 10²³ stars - that's 100,000,000,000,000,000,000,000. That's about 14 trillion stars for every person alive today. We've got a lot of exploring to do.

How to delude yourself

Here's a socialization anti-pattern: be judgmental and/or unforgiving. Let it be known that normal, decent, upstanding citizens don't do X, where X is: abandon their religion, hold the wrong political beliefs, not vote at all, do drugs, cheat on their spouses, indulge in adventurous sexual exploits, try polyamory, etc.* Notice how very rarely you hear about your friends doing X! It must be because they're all such normal, decent, upstanding citizens.

"In Iran we don't have homosexuals like in your country," [Iranian President Mahmoud] Ahmadinejad said to howls and boos among the Columbia University audience. "In Iran we do not have this phenomenon, I don't know who has told you that we have it," he said.

Or, of course, it could be that all that stuff happens, but never enters the pure, unsullied perception of the righteous. You can block or minimize entire categories of human nature from your perception. (Another way to achieve a similar effect is to acquire a reputation as someone who won't keep a secret. Either way, people won't tell you sensitive information.)

Sindibad, an ex-Muslim and commenter at Overcoming Bias, describes how he told people of his apostasy:

People whom I knew could possibly open their minds and stop them believing in their faith, I tried to talk to them about it, and succeeded in convincing a few.

Those whom I thought might not get convinced, but would be understanding, I told them the truth, and they never bothered me much.

But for those who didn't look likely to get convinced, or likely to be understanding, I have simply avoided talking to them about the subject, and when they ask, I pretend to be religious.

What a disaster. Avoiding challenging viewpoints is surely not a path to truth.

This guy loves apostrophes almost as much as he loves Jesus

But how do you know whether you're judgmental? Paul Graham describes the problem in his (highly recommended) essay What You Can't Say:

Who thinks they're not open-minded? Our hypothetical prim miss from the suburbs thinks she's open-minded. Hasn't she been taught to be? Ask anyone, and they'll say the same thing: they're pretty open-minded, though they draw the line at things that are really wrong...When people are bad at math, they know it, because they get the wrong answers on tests. But when people are bad at open-mindedness they don't know it. In fact they tend to think the opposite

I have no profound suggestions on developing open-mindedness that Paul Graham or the fine people at Overcoming Bias can't outdo. But the first step to understanding the world - no matter how you want to judge what you find - is surely to perceive it accurately (to the extent that humans can). So, consider: do people tell you secrets and trust you to keep them? All kinds of secrets?


*I'm not saying that all those things are ok, obviously. I'm just saying: there's a difference between responding "That's interesting. I see it differently. Let's talk about that." vs. "BEGONE, DEMON! YOU INFECT MY EYES.".

Concerned citizens want to know

How difficult is Artificial Intelligence?

(This post owes an enormous amount to the ideas of Eliezer Yudkowsky, an AI researcher who writes at Overcoming Bias. Reading this post might lead a non-Yudkowsky-reading person to a new perspective on AI, without needing to read the hundreds of thousands of words Yudkowsky has written on the subject. If it sounds profound, it's because of him. If it sounds dumb, it's my fault for not explaining it properly.)

How difficult is it to buid a human-equivalent (or better) AI? It's really difficult! Famously difficult! I've tried it myself.

But there's nothing inherently mysterious about intelligence, or consciousness, or sentience. There's nothing inherently mysterious about anything. Some things you know, some things you don't. Ask Eliezer Yudkowsky:

But ignorance exists in the map, not in the territory. If I am ignorant about a phenomenon, that is a fact about my own state of mind, not a fact about the phenomenon itself. A phenomenon can seem mysterious to some particular person. There are no phenomena which are mysterious of themselves. To worship a phenomenon because it seems so wonderfully mysterious, is to worship your own ignorance.

Some people want to put the engineering project of AI on a pedestal and set it apart from all previous inventions. Surely consciousness involves some emergent complex quantum phenomenon! No-one has ever built strong artificial intelligence! No-one, that is, for a narrow definition of "no-one".

No human has, anyway. Evolution, the intelligence that Yudkowsky likes to call the blind idiot god, did build intelligence, little by little, over billions of years. How did evolution deal with the problem, for example, of squirrels needing to find nuts?

Gradually DNA acquired the ability to build protein computers, brains, that could learn small modular facets of reality like the location of nut trees. To call these brains "limited" implies that a speed limit was tacked onto a general learning device, which isn't what happened. It's just that the incremental successes of particular mutations tended to build out into domain-specific nut-tree-mapping programs. (If you know how to program, you can verify for yourself that it's easier to build a nut-tree-mapper than an Artificial General Intelligence.)

You can generate computer programs using evolutionary algorithms, and you can get some interesting and useful results. So maybe we should try to build intelligence with evolutionary algorithms? But evolution is horribly inefficient - it took billions of years and a planet for evolution to produce the first intelligence (though, to be fair, building intelligence wasn't evolution's goal per se). Human intelligence is much more efficient - that is, smarter - than the intelligence of evolution that created it.

Some reader might consider my labeling of evolution as intelligence to be a dubious manoeuvre. We need a definition of intelligence. Something is intelligent to the extent it can perform efficient cross-domain optimization:

A bee builds hives, and a beaver builds dams; but a bee doesn't build dams and a beaver doesn't build hives. A human, watching, thinks, "Oh, I see how to do it" and goes on to build a dam using a honeycomb structure for extra strength.
...
A large asteroid, falling on Earth, would make an impressive bang. But if we spot the asteroid, we can try to deflect it through any number of methods. With enough lead time, a can of black paint will do as well as a nuclear weapon. And the asteroid itself won't oppose us on our own level - won't try to think of a counterplan. It won't send out interceptors to block the nuclear weapon. It won't try to paint the opposite side of itself with more black paint, to keep its current trajectory. And if we stop that asteroid, the asteroid belt won't send another planet-killer in its place.

We might have to do some work to steer the future out of the unpleasant region it will go to if we do nothing, but the asteroid itself isn't steering the future in any meaningful sense. It's as simple as water flowing downhill, and if we nudge the asteroid off the path, it won't nudge itself back.

The tiger isn't quite like this. If you try to run, it will follow you. If you dodge, it will follow you. If you try to hide, it will spot you. If you climb a tree, it will wait beneath.

But if you come back with an armored tank - or maybe just a hunk of poisoned meat - the tiger is out of luck. You threw something at it that wasn't in the domain it was designed to learn about. The tiger can't do cross-domain optimization, so all you need to do is give it a little cross-domain nudge and it will spin off its course like a painted asteroid.

Evolution performs optimizations across numerous domains - it can arrange atoms into flying machines, snake venom, opposable thumbs and even protein computers that are pretty damn smart themselves. But considering the resources evolution had available, evolution could hardly be called efficient.

How does human ingenuity compare to evolution?

Yes, some evolutionary handiwork is impressive even by comparison to the best technology of Homo sapiens. But our Cambrian explosion only started, we only really began accumulating knowledge, around... what, four hundred years ago? In some ways, biology still excels over the best human technology: we can't build a self-replicating system the size of a butterfly. In other ways, human technology leaves biology in the dust. We got wheels, we got steel, we got guns, we got knives, we got pointy sticks; we got rockets, we got transistors, we got nuclear power plants. With every passing decade, that balance tips further.

So what does this say about the prospect of (re)inventing AI sometime soon? Nothing definitive, but it gives me hope. Remember, the problem of building intelligence has already been solved - and the one who solved it was a moron. Unfortunately he wrote in the most awful, inefficient, buggy, defect-prone spaghetti code a programmer could hope never to see, so bad that it's not clear whether it's best to try to understand what he wrote, or just rewrite it from scratch.

Obama and McCain are basically the same

How different would the policies of America's federal government be if McCain had been elected president?

A billion dollars has been spent this year to persuade people that there is a large difference between the parties/candidates. Voters with an existing preference for one of the parties also have the powerful in-group/out-group bias to contend with. There are not many advertising dollars spent convincing people that the parties don't differ very much (not zero dollars - some of the advertising by the smaller parties such as the Greens and Libertarians takes this angle), and there's no particular bias I can think of that might cause people to systematically err on the other side of this question.

This doesn't mean that the difference between the parties is small. It just means that whether the difference is small or large, it would appear large (assuming the advertising worked).

With this in mind, it's worth reconsidering the real differences between the parties. Is there any reason to think the difference might actually be small?

Patri Friedman brings up the spatial voting model:

In a two party-system, the parties will tend to converge towards the center, and thus have far more similarities than differences. This is because by moving towards the center they can pick up centrist voters without losing any of the folks on their extreme. They get farther from their extremists, but those extremists still vote for them because they remain closer than the other party. You can visualize this with a beach and two hot-dog vendors, and the rule that people on the beach will always buy from the nearest vendor. The equilibrium positioning occurs when both vendors are next to each other, with equal numbers of customers on each side. Also of interest is that customer repositionings that do not cross the center line do not affect optimum vendor positioning. 50% of vendor A's customers can move to the extreme end of the beach, and vendor A is still correct to be in the middle. This suggests that influencing the opinions of huge numbers of voters may still have little or no effect. In the hot-dog model, such situations provide great incentive for the entry of a third vendor. Unfortunately the majority rule of democracy makes this inapplicable, as a third party which splits the votes of an existing party simply guarantees that neither can win.

Bryan Caplan gets more specific:

Reps don't want to get rid of the welfare state. Almost all Reps support spending a big chunk of GDP on America's poor and old. And Dems don't want anything like socialism. Almost all Dems want America to remain a country where markets are the default and people can get rich if they play their cards right.

So what is the "key difference" between the parties? Rhetoric. When Republicans advocate a small contraction of the welfare state, Democrats claim that Republicans totally oppose the welfare state. And many Republicans oblige them by standing up for "liberty" and "responsibility." Similarly, when Democrats advocate a small expansion in the welfare state, Republican claim that Democrats oppose free markets. And many Democrats oblige them by saying things like "markets only benefit the rich."

This rhetorical illusion is so powerful that when a Democrat like Clinton adopts many pro-market reforms, Republicans still hate him as a 60s radical. And when Bush II sharply expands the welfare state, Democrats still hate him as a billionaire's lackey.

Congratulations Obama

So Obama won. I admit I was secretly hoping McCain would win, and that his win would be widely regarded as fraudulent, leading to massive non-violent Gandhi-style protests eventually culminating in the secession of California. Californialand would then fix some disastrous federal policies such as the criminalization of marijuana. And then somehow Google gets elected president.

But Obama won instead, which I guess was the next best thing. Meanwhile, I'm hoping for divided government in 2010.

Fixing the weight-loss industry with financial incentives

Followup to: Obesity is a choice

The weight-loss industry is notorious for being full of ineffective magic-pill promises: 5-minute-a-day machines, fad diets, pills, etc. This bothers me. It offends my political biases when free markets don't come up with good solutions.

It occurred to me that a better way to do this might be pay-for-performance. A company could promise "if you don't lose weight, we don't get paid". And then their system will have to be optimized for whatever actually works, whether that's diets, exercise regimes, pills or magnetized rainbow crystals. That solves the bad incentives for the industry. But then you have a dubious incentive for the consumer - they now have a financial incentive to stay fat. Obesity correlates with low incomes, so maybe this will matter to a lot of people.

I have an idea for a solution. Set up a system where you can bet on whether people will lose weight. People who want to lose weight bet that they will lose X amount of weight (or maybe body fat or whatever their goal is) in X amount of time. Now they have a financial incentive to lose weight, and a disincentive to not lose it. There could also be a side-industry of consultants who give you diet/exercise/unicorn pills and get a cut of your win only if you win.

Who would bet against the weight-loss hopefuls?

Despite aligning the incentives properly, I'm guessing that a lot of people will still fail to hit their weight targets. It might be hard to predict whether any individual person will succeed or fail, but in large numbers (especially if you control for demographics, initial weight, gender, etc.), the number of people expected to succeed or fail will probably be quite predictable. There would be a mini-industry of analysts who can predict the probability of success of any particular hopeful, and they'd be willing to bet on slightly-favorable odds against you. Competition among analysts would keep their bets close to fair. They'd bet small amounts across large numbers of people to get a near-guaranteed positive return. If you're a weight-loss hopeful, presumably you won't mind if your bet is 5% less than fair; that's the price of the system, and winning is up to you anyway (you and your consultant).

The idea of betting against people's health is a bit repugnant, but I expect this system would have a good chance at being one of the best weight-loss programs available. And results are more important than repugnance.

(This idea is somewhat similar to the idea behind the startup Stickk.)

Obesity is a choice

Obesity is a choice. What a horrible thing to say! Let me put it more mildly: today I will argue that the claim "obesity is not a choice" is wrong.

Let's hear what Dr. Robert Lustig, pediatrician and obesity expert, has to say:

It is unfair and unhelpful to blame personal behaviors, especially a lack of self-control, for the country's rising obesity rates, says Dr. Robert Lustig, a pediatrician and nationally renowned obesity expert.

...

"Everyone's assuming you have a choice, but when your brain is starving, you don't have a choice," Lustig said. "When you look at it that way, all of a sudden Big Food looks like the perpetrator, and the patient becomes the victim. Congress says you can't sue McDonald's for obesity because it's your fault. Except the thing is, when you don't have a choice, it's not your fault."

Dr. Lustig has some valid concerns about the effects of insulin spikes from foods with high glycemic loads. But if there is some kind of psychological starvation response in the brain, overcoming instinct is still possible. I even consider it virtuous. So I accuse Dr. Lustig of overstating his case. There is evidence that obesity rates respond to incentives. If obesity rates respond to incentives, then there is choice involved (at least, for some people; this evidence only shows that some people alter their weight based on incentives).

First: marriage leads to obesity:

•Women in their teens and early 20s who continued to date but didn't cohabitate gained an average of 15 pounds over five years; their male counterparts added about 24 pounds.

•Newly married women in that age group packed on 24 pounds in five years; newly married men gained 30 pounds.

That degree of gain wasn't seen in couples who were living together but not married. Women gained 3 pounds more than their single peers — 18 pounds — and men gained 24 pounds.

"When people are dating, there may be more incentive to be thin," [Penny] Gordon-Larsen [an assistant professor of nutrition in the school of public health at the University of North Carolina-Chapel Hill] says.

I agree with Gordon-Larsen; the incentive to attract new partners is probably the cause. As your partner's commitment to you increases, the incentive to be attractive decreases.

Second: obesity is contagious.

Having a friend, sibling or spouse who is overweight raises a person's risk of being obese too, US researchers say.

They said data on more than 12,000 people suggested the risk was increased by 57% if a friend was obese, by 40% if a sibling was and 37% if a spouse was.

...

The effects were generally larger between people of the same sex.

And their analysis suggested that the links could not be solely attributed to similarities in lifestyle and environment, for example the impact of friends existed even where friends lived in different regions.

Author Professor Nicholas Christakis said: "It's not that obese or non-obese people simply find other similar people to hang out with.

"Rather, there is a direct, causal relationship. What appears to be happening is that a person becoming obese most likely causes a change of norms about what counts as an appropriate body size.

"People come to think that it is OK to be bigger since those around them are bigger, and this sensibility spreads."

Maybe you face less social disapproval from your friends when they gain weight, so the cost of gaining weight is decreased. Alternatively - noting that the effects were larger between people of the same sex - it could be about competition. Maybe people try to compete with the norm for their gender, and their perceived norm is influenced by their social group.

Whatever the explanation, it appears that obesity is responding to incentives.

(Quick disclaimer: don't read into this post anything that isn't there. As far as I can tell, I have not said anything normative about obesity or social acceptance of obesity in this post.)