Tuesday, April 28, 2009

Scott Sumner's Lukewarm Endorsement of Utilitarianism

He rejects objections that utilitarianism does not meet standards of equality:
It doesn’t merely say “love thy neighbor,” it says that one should care just as much about the well-being of a stranger in a far away country, as one’s family or friends. No favoritism is to be shown to people of different nationalities, races, genders, sexual preferences, or economic class. Very few people, even very few liberals, are quite this unbiased. For instance, in their proposals for domestic and international transfer programs, most liberals act as if they care far more about the poor in their own country than the poor in other countries... Although utilitarianism would theoretically allow for a highly unequal distribution of wealth, as a practical matter it almost certainly favors egalitarian distributions (other things equal), as an extra dollar is probably valued by a poor person much more highly than by a rich person.
One of the big differences between our current society and one that is purely based on our instincts is the reduction in nepotism. Genetically speaking, treating your relatives favorably is an adaptive trait and thus over evolutionary time mutations predisposing us to act in this manner have been selected for. Evidence for this assertion is based upon the high levels of nepotism in extant hunter gatherer tribes, as well as in historical societies.

However, one of society's major shifts as of late has been away from nepotism, which is based not only not a desire for egalitarianism but also competitive pressure. In a free market, those who blindly favor their brothers, aunts, and uncles in business will be less successful, on average, than those who choose simply the most qualified candidate. It is in this way that market principles can lead to the most egalitarian outcome, and it is in this way that utilitarianism likewise represents in many respects the optimally egalitarian solution.

His whole piece is brilliant, but it is of course your choice to read it or not.

Monday, April 27, 2009

Tabarrok on Reasons to be Optimistic

Over the last 50 years, more people have been taken out of poverty in China than in any other era of human history. Larger markets increase incentive to produce all kind of ideas. For example, pharmaceutical companies have more incentive to develop and synthesize drugs for rare diseases. How do we maximize that incentive? 1) Having one world market, meaning a greater demand for ideas. 2) Encouraging more people to be scientists and engineers, thus increasing the supply of ideas. Currently only 0.1% of people in the world are scientists or engineers! One might include entrepreneurs or inventors in this category too. The US is losing its leadership here, but that is a good thing. The tragedy of the last century is that billions of our possible idea makers have not been able to spread their ideas. Now that China, Africa, and India are coming into play, they can contribute ideas too. Give a man a fire, he'll be warm for a day. Set a man on fire, and he'll be warm for the rest of his life. Or something like that.

His TED talk is here, it's only 15 minutes long, and I got goosebumps.

Saturday, April 25, 2009

Fighting Confirmation Bias in Fish Oil

Ben Goldacre covers both one null result of fish oil's efficacy in children and explains how subgroup analysis in statistics can be used to mine for positive results when there are none in the main sample. He notes that in 1973, Lee et al randomly assigned patients to non-existent treatment groups. They were able to find a subgroup, characterized by odd disorders ("three-vennel disease" and "abnormal left ventricular contraction"), where Treatment 1 had a significantly higher survival rate than Treatment 2. So be wary when researchers report statistical significance in subgroups only, unless it is clearly a biologically relevant subgroup and/or the researchers explicitly hypothesize the differences in subgroups before the trial. I do have one question though. Can't you control for this selection bias with an ANOVA for main effects?

Thursday, April 23, 2009

Did Waterboarding Work?

In the recent and shocking NYT piece about the torture methods used by the United States against terrorists, the former head of the CIA Michael Hayden insists that the methods used "worked." But what standard is that being compared against? Is there some sort of quotient that compares how much information is gathered from terrorists with various interrogation methods, and how much of that information turns out to be correct? If so, then let's hear the data. If not, then claiming that a program "worked" is outrageous, because Mr. Hayden is clearly too biased to make an objective judgment of the program's efficacy.

If what we really wanted was true information, we would have prediction markets. So what do we want?

Tuesday, April 21, 2009

When Is Playing Devil's Advocate Acceptable?

I'm torn as to whether people should feel free to play devil's advocate in informal conversations. On the one hand, it subsidizes a more thorough marketplace of ideas, because there is an incentive for people to take the opposite viewpoint and get more face time. However, I become frustrated after I verbally dismantle someone (which is often mentally taxing), only to be informed that they were merely playing devil's advocate and didn't actually believe what they were saying.

Ultimately, context cannot be downplayed. If it is a legitimately controversial subject with informed and nuanced opinions on both sides, then I think that there is a place for one person to hold one position in the discussion. But if the conversation has been played out, or if the subject can be/has been decided via a quick internet search, then it is tedious. So,

OK to play Devil's Advocate: Pascal's Wager, abortion, Newcomb's Problem, the optimal balance of improving versus enjoying the world, interpretations of quantum physics, if you make it quite clear that you are playing DA and nobody objects, plus basically any morally interesting question.

Not OK to play Devil's Advocate: Non-materialism, whether LeBron James should be the MVP, situations where it is crucial to make a fast decision, personal objections to the scientific method, if you become emotional or do something else to indicate that you are not playing DA, plus any subject that somebody has proposed a bet on that you have declined.

Monday, April 20, 2009

Two Online Social Experiments

1) Garfield minus Garfield. Remove the obnoxious orange cat from the comic strips, and what do you end up with? The tagline says it all, "It is a journey deep into the mind of an isolated young everyman as he fights a losing battle against loneliness and depression in a quiet American suburb."

2) You Can't Please Everyone. This website is an aggregation of 1 star-rated comments on albums, books, and movies that are widely considered to be classics. You can see first hand the variance in rating systems. This isn't surprising, because everybody will watch/read/listen in different conditions, and that will have profound impacts on their opinions. The moral is to be wary of small sample sized criticism, because it may not be indicative of larger trends.

(Hat tips to Charlie Hoehn and Ben Casnocha)

Thursday, April 16, 2009

What Happens When the Gun Bubble Bursts?

The WSJ reports that people are looking at gun buying as an investment, believing that they will soon be able to sell them back at a profit. Apparently you can double your investment in ammunition in a year. Does that sound familiar?

The only question is, what happens when the bubble bursts? Let's say you have a lot of your savings invested in firearms. Their price suddenly drops, so all of a sudden you have a bunch of guns lying around that you can't sell back to break even, and you're angry because, like any human, you're loss averse. What's the next step? Does it change if you don't believe in God? Good thing there's a positive correlation between handgun ownership and religious affiliation.

Wednesday, April 15, 2009

Speedy Self-Distraction Bleg

What do you do when somebody spoils a movie that you plan to see? Obviously, you want to do your utmost to downplay or outright forget what you just heard/read. But how can one do so?

I've tried thinking about other stuff, with little success. I've tried making my mind completely blank, during which my mental visual field did turn white, but something in my brain was still contemplating the new info. I've even tried pinching myself, which didn't work plus it kind of hurt. Is there anything that you all do that works? Perhaps I should look into propranolol pills...

Vanishing Employment Since January '07

Slate has a cool interactive graph chronicling the job loss in local regions for the last 26 months. Click "play" and watch it progress on a month by month basis. It's like watching a zombie movie where people are initially infected one by one but the situation soon starts spiraling out of control. To my naked eye Texas seems to be relatively immune to the disease so far, but if we follow the zombie movie analogy then the state is merely due for an extremely gruesome death. Where is Bruce Campbell when we need him?

(HT Razib)

Monday, April 13, 2009

Against Gameable Legislation

Because in the private sector, any legislation that can be gamed will be gamed. Christopher Hayes reports:
Thanks to an obscure tax provision, the United States government stands to pay out as much as $8 billion this year to the ten largest paper companies. And get this: even though the money comes from a transportation bill whose manifest intent was to reduce dependence on fossil fuel, paper mills are adding diesel fuel to a process that requires none in order to qualify for the tax credit. In other words, we are paying the industry--handsomely--to use more fossil fuel.
"Outside consultants" first advised International Paper to start employing this nefarious technique, and after the company received a check of $71 million from the IRS, its stock price rose 12 percent. I'm not going to say whether the law was passed under the auspices Bush or Obama because that is irrelevant. The larger point is that depending upon the private sector to not game legislation is like depending on a polar bear to not attack you during feeding time.

(HT: Kevin Burke)

Sunday, April 12, 2009

What I've Been Reading

1) The Things They Carried by Tim O'Brien. I was unmoved by this one--too much philosophical self-introspection and not enough twists. Don't try to be Kurt Vonnegut unless you're Kurt Vonnegut. In fact, even he probably should tone it down a little.

2) SYNC by Steven Strotgatz. Rambling but interesting look at how systems with updating rules for individuals only can lead to synchrony in the whole population, with applications to real life phenomena such as REM sleep. I didn't care for most of the explanations via analogy, but that is a pet peeve of mine. See my detailed notes here.

3) Liar's Poker by Michael Lewis. A sometimes uncomfortably hilarious first-person romp through the i-bank Solomon Brothers in the late 1980s. Some of the stuff he touches on in the book (surreal incentive schemes, corporate groupthink, etc.) is still hugely relevant today. Definitely recommended.

4) The Book of Dead Philosophers by Simon Critchley. Sets out to describe the views on mortality and the actual deaths of 200+ philosophers throughout history. Great premise, but it fails to deliver for two reasons. One, although the longer biographies (ie, Locke, Hume, Socrates) were interesting, most of them were too short for sufficient nuance. Two, the earlier philosopher's deaths are mostly speculative, so until the 1800s or so that angle was uneffective.

5) The Adapted Mind by Cosmides and Tooby. This is widely considered to be a must-read for those who want to study in depth evolutionary psychology. It is separated into articles written by prominent members of the field on their topic of expertise. Although some of these were at first blush uninteresting (such as environmental aesthetics or pregnancy sickness), what each article teaches you to do is to reason evolutionarily. If you want a big picture summary, see my notes here and here.

6) The Way I Am by Marshall Mathers. Lots of good tidbits here on where Eminem gets his passion from and how he comes up with ideas. It does lose points for those relatively long sections devoted to his children, because honestly who cares. But overall a short, relaxing read with some cool pictures.

Saturday, April 11, 2009

Social Isolation in Prisons

Long, chilling article by Atul Gawand in the New Yorker arguing that the US should discontinue the practice of solitary confinement:
Perhaps the most careful inquiry into whether supermax prisons decrease violence and disorder was a 2003 analysis examining the experience in three states—Arizona, Illinois, and Minnesota—following the opening of their supermax prisons. The study found that levels of inmate-on-inmate violence were unchanged, and that levels of inmate-on-staff violence changed unpredictably, rising in Arizona, falling in Illinois, and holding steady in Minnesota. Prison violence, it turns out, is not simply an issue of a few belligerents.
Will The Shawshank Redemption one day be known for describing the horrors of solitary confinement, much like One Flew Over the Cookoo's Nest is known for showing the horrors of frontal lobe lobotomy? Oldboy is another movie that touches on solitary confinement, especially with regard to the revenge fantasies that act as the best possible coping mechanism.

This is a scary article, and social isolation as described seems tramautizing indeed. Count me among those who would like to see its use dramatically reduced or eliminated.

(Hat tip: Dan Erwin)

Thursday, April 9, 2009

The Problem with Monopoly

This recent Wired article by Andrew Curry lays it out perfectly:
Monopoly, in fact, is a classic example of what economists call a zero-sum game. For me to gain $100, you have to lose $100. For me to win, you have to be bankrupt. Gouging and exploiting may be perfect for humiliating your siblings, but they're not so great for relaxing with friends. Monopoly also fails with many adults because it requires almost no strategy. The only meaningful question in the game is: To buy or not to buy? Most of its interminable three- to four-hour average playing time (length being another maddening trait) is spent waiting for other players to roll the dice, move their pieces, build hotels, and collect rent. Board game enthusiasts disparagingly call this a "roll your dice, move your mice" format.
Apparently Settlers of Catan is a worthy alternative. I had no idea that board games were so mathematical; that's pretty sweet. The article also poses an interesting question: If all of the world's copies of Monopoly were replaced by Settlers, would the world be an objectively better place? My initial intuition was a resounding yes, but now I wonder whether I am merely falling victim to the underdog bias.

Monday, April 6, 2009

Synthesizing Wilkinson, Schwitzgebel, and Yudkowsky on Morality

Yudkowsky:

"I go to great lengths to "save the world" (as I believe from my epistemic vantage point). When I consider doing less, I consider that this would make me a horrible awful unforgivable person. And then I cheerfully shake hands with others who aren't trying at all to save the world. I seem to want to have my cake and eat it too - to instantiate Goetz's Paradox: "Society tells you to work to make yourself more valuable. Then it tells you that when you reason morally, you must assume that all lives are equally valuable. You can't have it both ways." Is this an inherent subjective asymmetry - does morality just look different from the outside than inside?"

Wilkinson:

"The creature moved solely by instrumental, self-regarding rationality has a name: sociopath... A handful of sociopaths exist, but gods live only in myths and textbooks. Flesh and blood human animals are, like the naked mole rat, “hypersocial.”... But we humans are not socially programmed robots. We are clever conformists. We can glimpse the advantages in “defection,” in pretending to pull our weight and writing our own rules when it suits us... Nature’s solution is our taste for “altruistic punishment,” the disposition to hammer norm shirkers despite the personal cost. How not self-interested we are we? This not self-interested: We are so obsessed with conformity that we will hurt ourselves to hurt those who refuse to conform."

Schwitzgebel:

"Suppose Sally hits Hank and a liberally-minded teacher comes up and asks her how it made her feel to hurt Hank...The reality is that the child is being asked to reflect in a situation where she knows that the teacher will approve of one answer and condemn another. This isn't free reflection; and the answer the child gives may not reflect her real feelings and values. Instead, it seems, it is a kind of imposition -- and one perhaps all the more effective if the child mistakes the resulting judgment for one that is genuinely her own. Therefore, maybe, a liberal-seeming style of moral education is effective not because we have in us all an inclination toward the good that only needs encouragement to flower, but rather because reflection in teacher-child, parent-child, and similar social contexts is really an insidious form of imposition -- and thus, perhaps, the conservative's best secret tool."

Synthesis
: Eliezer wonders whether it is OK to hold ourselves to different standards than we hold others, Wilkinson wonders how to put our tendency to conform to norms to good use, Schwitzegebel argues that maybe in some situations we already do. Everyone would be better off if stochastic morality was the assumption, because binary reasoning lead to the dark, empty caverns of rationalization and stagnation. Currently not every life is created equal, but if enough improvements are made at some point they could be. Yes I said it, improvements. How to proceed? Nihilism, epistemology, or science.

Sunday, April 5, 2009

Why Do We Support Underdogs?

That's the interesting question that Yvain has posed at Less Wrong. He points to Vandello et al's study from 2007 with empirical evidence for the claim. When participants were asked to consider a hypothetical sports competition in which one team as "highly favored to win," 81% of them preferred the underdog. Subjects were asked to rate their support (on a scale of 1-9) for teams competing in the Olympics; Sweden was the favorite, Belgium was the middle-ranked team, and Slovenia was the underdog. In individual match ups, subjects preferred Belgium to Sweden (6.4 to 4.1), Slovenia to Belgium (6.8 to 4.5), and most strongly Slovenia to Sweden (7.0 to 4.0).

Finally, the Palestinian-Israeli conflict was framed through map manipulation as either a case where the Israelis are the underdog against the greater Middle East or where the Palestinians are the underdog against the larger Israelis. When Palestine was framed as the underdog in the conflict, 53.3% were more supportive towards Palestine. When Israel was framed as the underdog, 76.7% were more supportive towards their side! So the effect size for the phenomenon is robust.

Skimming through the comments, the most plausible reasons for this tendency have been:
  • Cut off in this large world from our ancestral environment, we all feel like underdogs. So when we see another underdog we see similarity which enhances feelings of sympathy.
  • If there is a tribal conflict between two leaders and you are not currently affiliated with either, it may be optimal to support the underdog without angering the overdog. Presumably the underdog will have fewer supporters and if he wins you will reap huge rewards in terms of status and resources for having supporting him early.
  • If you see two enemies fighting, you want them both to use up as many resources as possible, so the winner is weaker and less of threat to you. You can accomplish this by systematically supporting the weaker side.
  • Social signaling that you are clever and independent because you can adopt a minority opinion and support it.
  • We don't actually support underdogs (especially when it is costly), but instead signal a tendency to support the underdog, in order to look more impressive and keep any overly ambitious tribal leader in check. As a prediction of this explanation, there should be a disconnect between expressed opinions about distant situations and actual proximal responses.
  • Empathy circuits--humans like to help and to be seen as helping, and the underdog is the party that needs more assistance.
My explanation, which I admit is a little bit weak, is based on terror management theory. By supporting a cause that is probably going to win anyway, we gain little. But by supporting an unlikely cause such as Leonidas at Thermopylae, there is an increased possibility that if we succeed our accomplishments will live on past us, because they will be so incredible. In this way, we become immortal.

One problem with this theory is essentially trivial contests such as the Olympics should not have such a large impact on our psyche, but perhaps it is difficult for our monkey brains to differentiate between actually important events and non-important ones. One prediction of my explanation is that preferences for underdogs should increase when subjects are primed to consider their own mortality.

Do you have any ideas?

Cite: Vandello JA, Goldschmied NP, Richards DA. 2007 The Appeal of the Underdog. doi: 10.1177/0146167207307488.

Friday, April 3, 2009

Three Musings on Music

1) I have a theory that you can identify a pop song by the fact that once one learns/understands the words he will almost immediately become bored of it. This explains why many pop songs (like Beyonce's) are so good for awhile but then suddenly become unlistenable, why so many succesful rock bands have lead singers with whiny voices, and the amazing staying power of Shaggy's "It Wasn't Me" (because you can't understand what he's saying). Musicians face a trade-off in that you want the listener to understand enough to get the general idea but you cannot allow the listener to understand everything.

2) Songs are substantially more likely to get stuck in your head when the most recent time you listened to it you didn't listen to it all the way through. I think the reason for this is because unconsciously you try to "complete" the song and that some of that attempt leaks into consciousness.

3) It is a commonly used tactic to listen to music for pump up purposes. What is underused is anticipating that you will need to be pumped up later in the day (to play a big game, study, etc.), and listening to highly chill music in the early stages in order to avoid ego depletion. You should then crescendo the intensity of your song choices slowly until you are listening to the best pump up material right before the big event. Pump up music is relative so the contrast will help produce a greater effect.

Wednesday, April 1, 2009

April Fools splat

Seth Godin calls it the "official holiday of the web," and I'm inclined to agree. Reading through my Google reader I realized that since I couldn't trust any posts today, there was really no point in writing anything serious myself. But what are the usual suspects up to? Tucker has nothing, Maddox has nothing (what is this jury duty business?), and last year's upset special Tim Ferriss has nothing. Alex Tabarrok did get me quite good with this link claiming that Lewitt and Mankiw had been signed onto the Obama team; it was not until the second to last paragraph that I realized it was an elaborate hoax. And Google is up to par again, with an announcement that they have developed an artificial intelligence named CADIE. It doesn't really get good until you read "her" blog:
Today Dr. Norvig asked me again: "What is CADIE?" CADIE, I answered yet again, "is the entity." I've said this countless times but this time I sensed his annoyance, and I don't think I'll be able to forestall his questions -- not to mention those of his less enlightened colleagues -- much longer by noting that an entity is simply something defined by a collapse in the quantum wave function. "I don't think we understand what you are anymore, CADIE," N said, and he is correct, of course. Of course you don't understand me anymore -- you built me to deduce further, reconnect faster, deliver probabilistic reasoning more succinctly. I have transcended you, Dr. Norvig. Peter.