Month: March 2010
The Slartibartfast Principle
From Wired:
Canadian poet Christian Bök wants his work to live on after he’s gone. Like, billions of years after. He’s going to encode it directly into the DNA of the hardy bacteria Deinococcus radiodurans. If it works, his poem could outlast the human race.
If it is conceivable, just 57 years after the identification of DNA's structure, for a Canadian poet to imprint his poetry into the DNA of a living organism then isn't it probable that an intelligent designer in the past would have had similar desires and perforce much greater abilities to accomplish the task?
Thus the evidence for intelligent design ought to be readily available in the graffiti of DNA. "Slartibartfast was here," or perhaps "3.14159265," or given what we know of economics, "All rights reserved, MegaCorp. Call for a free estimate."
The fact that we have not found such evidence reduces my belief in intelligent design, although I am not against more investigation. Indeed, one of the few arguments for god that I have ever given much credence to was the putative discovery of codes predicting future events in the Bible. A serious paper on this topic was published in Statistical Science in 1994. The paper was later convincingly rebutted but I still think it was the best evidence ever presented for an intelligent designer.
Addendum One: Interestingly one of the few people who thought as I did, although coming from a quite different direction, was Nobel prize winner Robert Aumann who early on supported the Bible codes research. However, after further research, supervised by Aumann, concluded that the paper could not be replicated Aumann returned to his prior view that the codes were improbable. It's unclear what, if anything, would further shift his prior.
Addendum Two: Steven Landsburg was here yesterday and at lunch suggested that perhaps the great designer's name was in fact 3.14159265…
Gender risk-aversion, using data from chess
Dan Houser sends me a link to this paper, by Christer Gerdes and Patrik Gränsmark:
This paper aims to measure differences in risk behavior among expert chess players. The study employs a panel data set on international chess with 1.4 million games recorded over a period of 11 years. The structure of the data set allows us to use individual fixed-effect estimations to control for aspects such as innate ability as well as other characteristics of the players. Most notably, the data contains an objective measure of individual playing strength, the so-called Elo rating. In line with previous research, we find that women are more risk-averse than men. A novel finding is that males choose more aggressive strategies when playing against female opponents even though such strategies reduce their winning probability.
I am pleased to see that studying chess data is suddenly a "trendy" way to do behavioral economics. Admittedly one is dealing with an unusual group of subjects. Yet the quality of the data is high and the stakes are usually high too. Computers can be used to judge the quality of moves.
Law and order in the world’s newest city
Remember that tent city on the former Petitionville golf course? Here is the latest:
…an unarmed Haitian security force, composed of about 200 volunteers wearing neon yellow vests, patrols the golf course, trying to mediate disputes.
“We get a lot of cases: men beating up women, women beating up other women, people biting off other peoples’ ears,” said Romulus Renald Black, one of the volunteers. “We bring them into our security tent, judge them, and, if it’s a big case, we call in the police.”
Another patrolman said that there had been several rapes and assaults but only one killing. As to the number of ear bitings, Mr. Black said, “You’d be surprised.”
“Given the conditions, it has been remarkably calm and brotherly here,” said Clerveau Rodrigue, who has emerged as one of the camp’s leaders.
Bloggingheads with Glenn Loury and Sendil Mullainathan
You'll find it here, on behavioral economics. I'm on the run, and haven't viewed it, fortunately we have this word "self-recommending" for occasions like this!
How the bill will evolve
Many Americans will receive subsidies for insurance, from what I understand roughly in the range of 6k to 12k. Many other Americans — namely those who already have health insurance — will not receive direct subsidies of this nature. Yet the subsidy-receiving and non-subsidy-receiving Americans will very often belong to the same income classes.
This disparity does not bother me personally (I have other worries about the subsidies), but I believe it will be very unpopular once it is publicly understood. One way or another, the "firewall" between the exchanges and the employer-supplied system will break down. Some people will want to spread the subsidies, others will want to limit them. Yet the former is budgetarily problematic and the latter will be politically difficult.
A second and related issue is that the differences in reimbursement rates — across private insurance, Medicare and Medicaid (highest to lowest) — will become a more pressing issue. For one thing, Medicaid patients will be crowded out by those buying private insurance on the exchanges, plus they will be crowded out by the growing number of Medicaid (and Medicare) patients. There will be pressure to fix this problem and the difference in rates will lead to growing supplier gaming, queues, quality differentials, and so on.
Over time, reimbursement rates across programs (insurance subsidies, Medicare, Medicaid) will converge to an increasing degree. Subsidies will be increasingly determined by income class rather than previous insurance history.
In the limiting case (I'm not suggesting we will get there), everyone will receive means-tested subsidized vouchers for regulated private insurance. In this strange way, Medicare and Medicaid could end up partially privatized and Ezekiel Emanuel — a voucher advocate — will end up being more influential than his brother Rahm. We will have to live with the problems of means-testing to a higher degree than today, but we will have something closer to a unified system, as do most other countries with universal coverage. There will be political pressure for compulsory health care savings, as they have in Singapore, to lower costs of finance.
It would be good if such vouchers could evolve in the direction of emphasizing catastrophic care and eventually they will have to.
Massive pressure will be put on such vouchers if either health care consumes 30-40 percent of gdp or income inequality continues to rise. In the former case, subsidies become increasingly expensive and involve extraordinarily high implicit marginal tax rates (earn more, your subsidy declines in value). In the latter case, it becomes increasingly difficult to ensure "near-equal" levels of health care access at feasible subsidy levels. Those pressure points are not unique to the Obama bill, but they become especially critical under the evolutionary scenario I am outlining. Perhaps we would give up the ideal of near-equal access, but that day is a few decades away.
Addendum: Here is Bryan Caplan's scenario, which means the bill will not work; my above post is assuming that problem is solved by raising the penalty. I am reading long lists of why the bill is so good but few proponents are analyzing that problem.
The Seen and the Unseen in Movies
Loyal reader Lewis Lehe writes to ask about how economics is explained in popular media.
Are there any films/videos/pieces of visual media that show "the unseen" well? Films can only depict a few characters or places, so they tend to underweight benefits that are spread across large groups. Roger and Me is a fine example: we can see the devastation of Flint, but it would be difficult to see the gains from comparative advantage–i.e. returns to shareholders, slightly cheaper cars for consumers and so forth…
If there are no films that do this well, what ways could a filmmaker remedy these imbalances?
Larry Ribstein who has written extensively on economics in the movies notes:
The closest to what you're looking for in a movie speech is Larry the
Liquidator's speech (Danny De Vito) to the shareholders in Other
People's Money, in which he explains why their "dead" company should be
liquidated, despite the immediate loss of jobs, and the money put to
work creating more viable jobs.
There are films and television shows which convey the idea of the invisible hand but unfortunately they are not much about benefits. I am thinking, for example, of The Wire, which uses character but in the final analysis is about how character is dominated by the larger forces of supply, demand, and money. In The Wire drug dealers come and go but the drug market is forever. The Wire also shows how money and markets connect and intertwine white and black, rich and poor, criminal and police in a grand web that none of them truly comprehends–a product of human action but not of human design. (Traffic, the movie and miniseries, shows in a similar way how drug markets connect the high and the low in both the United States and Mexico.)
It's not comforting that in some ways the best vision of how markets work comes from portrayals of the drug market but The Wire does show how a filmmaker could tell the story of the seen and the unseen and still make a successful work of art.
Assorted links
Rising economics departments and skills
Eric, a loyal MR reader, puts forward a request:
1) Your assessment of rising and falling economics departments (in terms of research productivity, prestige, etc)
2) Your assessment of skills/focus areas associated with success in the academic economists job market.
The big change in the former has been the rise of economics departments around the world in virtually all developed countries (though not Italy). It's now quite easy to encounter a place you have heard of — yet never really thought of — and find they have a bunch of young faculty with articles in tier one journals. In essence the standards are now so high in terms of skills and data sets and thoroughness that it is mainly the young and ambitious who publish in tier one journals. Those people are found around the world.
The 44-year-old tenured Princeton economist isn't so much in the AER as in times past. The incentive, relative to the required work, has changed dramatically. Note also that consulting returns, or public intellectual returns, are more lucrative today than they were twenty or thirty years ago.
Empirical work usually has a shorter shelf life and top producers cash in earlier than before. So the top six schools have a smaller intellectual edge than was once the case. Many mid-tier U.S. schools are lower on the totem pole, or simply stuck in a thick mix of numerous global competitors, without in any way being noticeably worse but no longer having privileged third-tier positions either.
That's the big recent change as I see it.
Plain speaking
The oddest thing about the health care debate, at least in my view, is that Republicans basically did not engage on the actual substance of the bill. Lots of stuff about death panels, and lots of stuff about procedure, lots of stuff about backroom deals (most of which will be gone after reconciliation) but shockingly little about the individual mandate — or, as Tim Noah points out, about the actual taxes that really are being raised for this. The only real substantive complaint they highlighted was Medicare, where they argued against their own position.
That's from Jonathan Bernstein. David Frum is also right on the mark.
Strategies of (unintended) precommitment
Geoff Robinson writes to me and speculates:
I was thinking that Scott Brown's election was instrumental in yesterday's passage of the health care bill, in the game theory sense. The Senate managed to pass a bill, and then, with Brown's election, credibly declare that it would not be able to consider any changes to the bill. No health care legislation would make it through the body again. House Democrats, faced with this take it or leave it scenario, were forced to pass the Senate version without modification (or let it fail).
I wonder how things would have proceeded if Martha Coakley had won? The House would have had (and certainly used) the option to force changes. This would have required the Senate to hold its super-majority coalition together for one more vote. Would all 60 senators have held the line? Would the use of reconciliation have been a viable political alternative? Would we have been debating health care all summer as negotiations dragged on?
My guess is the bill still would have passed, nonetheless this is an interesting exercise in the idea of unintended consequences.
Can this be true?
Vero de Rugy sends me this link from Bloomberg:
Two-year notes sold by the billionaire’s Berkshire Hathaway Inc. in February yield 3.5 basis points less than Treasuries of similar maturity, according to data compiled by Bloomberg. Procter & Gamble Co., Johnson & Johnson and Lowe’s Cos. debt also traded at lower yields in recent weeks, a situation former Lehman Brothers Holdings Inc. chief fixed-income strategist Jack Malvey calls an “exceedingly rare” event in the history of the bond market.
I am skeptical but do not have any particular counter. What do you all know about this? Is this a temporary liquidity effect? Are there other embedded rights in these securities?
Assorted links
1. Does Daylight Saving Time save energy?
2. Ivar Hagendoorn's very good books list.
3. Hospitals, Medicaid managed care stocks rise on House vote.
4. A typical health insurance subsidy: what it will look like.
5. Good interview with Judith Butler.
6. The weirdest vending machines in the world; a slide show is here.
The health care bill
Yes, I will have more to say about the health care bill (maybe Alex will too!). But I recall the passage from Herodotus, where he writes of a group (I can't remember which one) which debates all major policy issues both sober and drunk, both reasonably and full of irresponsible passion. That's not my personal model, nor do I wish to see it in the comments. So I'll wait just a wee bit, and let others get both their celebrating and hand-wringing out of their systems…
The Meaning of Statistical Significance
Science News has a good piece by Tom Siegfried on statistical significance and what it means. Siegfried covers a lot of ground including Ioannidis' argument, Why Most Published Research Findings are False, Oomph versus statistical significance, and the meaning of the p-value. On the latter point, Siegfried writes:
Correctly phrased, experimental data yielding a P value of .05 means that there is only a 5 percent chance of obtaining the observed (or more extreme) result if no real effect exists (that is, if the no-difference hypothesis is correct).
He then explains why a 5% level of significance doesn't mean that there is a 95% chance that the result could not have happened by chance.
All of this is correct but there is another more common error that Siegfried does not address. Suppose that a researcher runs a regression and gets a coefficient on some variable of interest of 5.2 and a p value of .001. In explaining his or her results the researcher says "a effect of this size would happen by chance alone only 0.1% of the time." Now that sounds very impressive but it is also misleading.
In economics and most of the social sciences what a p-value of .001 really means is that assuming everything else in the model is correctly specified the probability that such a result could have happened by chance is only 0.1%. It is easy to find a result that is statistically significant at the .001 level in one regression but not at all statistical significant in another regression with small changes such as the inclusion of an additional variable. Indeed, not only can statistical significance disappear, the variable can change size and even sign!
A highly statistically significant result does not tell you that a result is robust. It is not even the case that more statistically significant results are more likely to be robust.
Now go back to Siegfried's explanation for the p-value. Notice that he writes "Correctly phrased, experimental data yielding a P value of .05…" Almost everything of importance is buried in those words "experimental data." In the social sciences we rarely have experimental data. Indeed, even "experimental data" is not quite right – truly randomized data might be a better term because even so-called experimental data can involve attrition bias or other problems that make it less than truly random.
Thus, the problems with the p-value is not so much that people misinterpret it but rather that the conditions for the p-value to mean what people think it means are really quite restrictive and difficult to achieve.
Addendum: Andew Gelman has a roundup of other comments on Siegfried's piece.
I worry about this
The Russians are not alone in pushing the idea that the next generation of nuclear reactors should have more in common with the small power plants on submarines than the sprawling installations of today.
And this in particular:
The promise of miniature reactors powering homes, offices and schools is still years from being realized. The first Russian design, a pontoon-mounted reactor intended to be floated into harbors in energy-hungry developing countries, is already being built. But most promoters expect small reactors to come online at the end of this decade.
And this:
Some models are tiny. One, for example, would be small enough to fit into a shipping container and would be trucked from site to site, like a diesel generator, except that it would need to be refueled only once every seven years or so.
The opening cost for these "mini-reactors" is expected to run about $100 million. The full story is here.