Monday, March 20, 2006

Kent Budge asks the perennial question: why do bad CEOs get paid so much? I've always been stumped by this one too, but reading Tim Harford in Slate today, I found the most interesting hypothesis I've read in a long while.

Economists tend to prefer the insights of Edward Lazear, who suggested that young employees are often encouraged to perform by the promise of a cushy cruise toward retirement. Lazear realized that it is hard to measure and reward performance directly, but underpaying the young and overpaying the old tends to encourage effort. As long as there is some risk that shirkers will be sacked, workers will beaver away knowing that if they can hold onto their jobs, it will all be worthwhile.

Lazear also suggested that the boss's fat-cat salary is designed to motivate the lean and hungry young junior executives beneath him and has very little connection to his own performance. (Much is explained.)

A heaping contradiction

The paradox of heap (or Sorites Paradox) starts with the following premises.

(1) A million grains of sand make a heap.
(2) If N grains of sand make a heap, then so do N-1 grains of sand.
(3) A single grain of sand does not make a heap.

Many people believe this leads to paradox because they intuitively affirm (1), (2), and (3) despite the fact that one can easily deduce a contradiction from (1), (2), and (3). (Consider the minimum N for which N grains of sand make a heap.) I don't think much of this paradox because I think premise (2) is obviously false.

"Do you agree that a million grains of sand make a heap?"
"Do you agree that one grain of sand makes a heap?"
"Do you agree that a 500,000 grains of sand make a heap?"
"Do you agree that a 250,000 grains of sand make a heap?"
"Do you agree that a 125,000 grains of sand make a heap?"
"Do you agree that a 62,500 grains of sand make a heap?"
"Well, uh, no; at least that's not how I'd use the word 'heap.'"
"Do you agree that a 93,750 grains of sand make a heap?"
"I don't like where this is going. I'm not sure."
"So you don't agree?"
"Well, I partially agree."
"But you don't fully agree?"
"No, I don't fully agree."
"Do you fully agree that 125,000 grains of sand make a heap?"
"Do you fully agree that 109,375 grains of sand make a heap?"
"Do you fully agree that 101,563 grains of sand make a heap?"
"Uh, yes."
"Do you fully agree that 97,656 grains of sand make a heap?"
"I don't know."
"So you don't fully agree?"
"I don't know!"
"Do you knowingly fully agree that 101,562 grains of sand make a heap?"
"Do you knowingly fully agree that 99,609 grains of sand make a heap?"
"Uh, yes."
"Do you knowingly fully agree that 98,632 grains of sand make a heap?"
"No, I don't know"
"Do you knowingly fully agree that 99,121 grains of sand make a heap?"
"Do you knowingly fully agree that 99,364 grains of sand make a heap?"
"Do you knowingly fully agree that 99,487 grains of sand make a heap?"
"Look, I know where this is going. Let's just say that I knowingly fully agree that 99,600 grains of sand make a heap, but I don't knowingly fully agree that 99,599 grains of sand make a heap."
"Are you ready to renounce premise (2)?"

Sunday, March 19, 2006

Now that's what I call health care reform!

Saturday, March 18, 2006

The specter for biowarfare seemingly gets scarier all the time.
Into a relatively innocuous bacterium responsible for a low-mortality pneumonia, Legionella pneumophila, Popov and his researchers spliced mammalian DNA that expressed fragments of myelin protein, the electrically insulating fatty layer that sheathes our neurons. In test animals, the pneumonia infection came and went, but the myelin fragments borne by the recombinant Legionella goaded the animals' immune systems to read their own natural myelin as pathogenic and to attack it. Brain damage, paralysis, and nearly 100 percent mortality resulted: Popov had created a biological weapon that in effect triggered rapid multiple sclerosis. (Popov's claims can be corroborated: in recent years, scientists researching treatments for MS have employed similar methods on test animals with similar results.)
Wisely, Technology Review got Allison Macfarlane to make the case for not getting too excited about the about their above article.
For example, data on the infectiousness of an agent varies widely, depending on the agent. Because of limited experience with anthrax, susceptibility data have often been extrapolated from animal trials that have little bearing on human response to agents. In the case of smallpox, with which scientists had much experience in the 20th century, some factors remain uncertain, such as the transmission rate.

In the models of bioweapons attacks, the ability to weaponize an agent and disperse it effectively is estimated in part from open-air trials done by the U.S. Army between the 1940s and 1960s. These trials used live simulants of agents on major U.S. cities, but the behavior of a real bioweapon agent in such a situation remains uncertain. Williams's article doesn't describe in any detail the ability of terrorists to weaponize any of the theorized agents. Yet making effective bioweapons would take a tremendous amount of work.

While a state-sponsored program might have the means to do that work, terrorist groups probably don't. With so much uncertainty surrounding the outcome of a bioweapons attack, it does not make sense to plan extensive biodefense programs when more-certain threats, particularly those involving nuclear weapons, require attention.
Macfarlane is right about these uncertainties, but I can't fully endorse his conclusion. In the long term (I'd guess thirty years tops), biotechnology and nanotechnology will solve the weaponization "problem." The economic motivation for this advance is clear: farmers have been fighting a biowar against pests for a long time. They will gladly pay for better weapons. We can't wait until then to start researching countermeasures against the use of these weapons on humans.

We can already easily engineer diseases capable of defeating the individual human immune system. It still probably takes a lot of infrastructure to overwhelm the collective immune system of the human species (i.e., a lot of us survive), but that won't be true forever. The collective natural immune systems of the developed nations will ultimately be augmented by artificial ones. (The undeveloped nations will not probably not survive the next century without a lot of help.) Alas, grand artificial immune systems are very much public goods, and politicians are usually myopic or backward-looking. I predict artificial immune systems will be underfunded until a catastrophe makes their necessity manifest.
A while back, I criticized Mark Steyn for being overly pessimistic about European Demographics. I've just now discovered this article making a similar point more quantitatively.
First, Western nations aren't the only ones with falling birthrates. The Muslim world is seriously declining as well. Iran, Turkey, Algeria, Morocco, Tunisia, Albania, Lebanon, and Malaysia are all below the 2.1 replacement line, while Bangladesh, Egypt, Indonesia, and the Muslim parts of India are close behind and falling rapidly. A few Muslim nations do indeed have high fertility, but the common denominator is not Islam itself, as Steyn implies, but a lack of modernization. Many non-Muslim countries that also haven't fully modernized have high rates as well, such as Laos, Uganda, and Paraguay.
Europe, on the other hand, is admittedly a trickier case. Native fertility is indeed low, while Muslim growth rates and levels of extremism have remained high. Over the next 50 years, Europe projects to lose about 100 million people, while European Muslims will double their numbers to about 20% of the total European population. If Turkey joins the EU, Muslim numbers will rise even further.

But will this bring Sharia law, as Steyn fears? I don't think so. Even under the most high-growth projection (which is by no means certain), Muslims will remain a minority on the Continent. Their radicals may want Sharia law, but they won't get it at the ballot box.

Friday, March 03, 2006

Wednesday, March 01, 2006