'I never did it, and anyway I've stopped"

In a recent CNN news story on a related issue, I came across the following statement from the Justice Department:

"As the Attorney General indicated earlier this year, any electronic surveillance that was being conducted pursuant to the [Terrorist Surveillance Program] is now being conducted subject to the approval of the Foreign Intelligence Surveillance Act Court."

The clear implication is that the administration is no longer trying to argue--although it may still believe--that it was entitled to violate the Foreign Intelligence Surveillance Act by warantless interceptions of the sort of communications covered by the act. While this is, in my view, a step forward, it does have one uncomfortable implication which doesn't seem to have gotten much attention--that the administration is no longer prepared to deny that the President and a variety of people in the National Security Agency are guilty of repeated felonies, each punishable by up to five years in prison.

Thoughts on Intervening in Civil Wars

Listening to the radio while driving, I heard a clip of Newt Gingrich, who I find one of the more interesting modern politicians. He was responding to someone else's point that Iraq was a civil war and we should therefor get out. Gingrich stopped short of actually agreeing that it was a civil war, but he did accept the claim at least for the purposes of argument--and pointed out that civil wars do get won, as demonstrated by the U.S. Civil War.

It occurred to me that one might carry the argument a step further. If you are going to invade a country, the fact that it is engaged in a civil war seems like a point in your favor. The two sides are spending resources fighting each other that might otherwise be used to fight you; with luck, you might even succeed in allying with one side against the other. That, after all, is how the U.S. managed its quick and easy initial victory in Afghanistan. The Northern Alliance supplied the boots on the ground, the U.S. provided them with unlimited air power, and that was enough to reverse the outcome of the civil war and drive the Taliban from power.

What are the arguments on the other side? Why might the existence of a civil war make conquest harder rather than easier?

One possible answer is that it makes it harder if the invader is unwilling to pick a side. It is then in the difficult position of trying to keep members of each side from killing members of the other while getting along with both. Since the invader's troops probably don't speak the language or know much about the society they are largely dependent on local allies, none of whom can be trusted not to take the opportunity to use their position to support their side of the ongoing civil war.

In the case of Iraq, the U.S. allied with the Kurds--gave them enough of what they wanted so that they were willing, on the whole, to refrain from shooting at us and support us against other people who were doing so. But the U.S. was not willing to support either the Shia or the Sunni in their desire to rule the country to the exclusion of the other.

A second answer to the question of why a civil war might make conquest harder rather than easier is that it raises the stakes. No reasonable person thinks that the U.S. wants to drive the Iraqis out of the country and replace them with Americans. It's unlikely that the U.S. wants to rule Iraq for any extended period of time. But a reasonable Sunni might well worry about ethnic cleansing by Shia, or a Shia about ethnic cleansing by Sunni. And each group has good reason to worry that the other would like to establish itself permanently in power, reducing its rivals to at best second class citizens.

The higher the stakes, the more desperately people are willing to fight. One of the problems with an interventionist foreign policy that it gets the country following it into conflicts where the other parties have much more at stake, and so are willing to put up with much larger costs.

Consider the Vietnam War. People in the U.S. complained that the South Vietnamese were not sufficiently willing to fight. But in fact the Vietnamese, North, South, and Viet Cong, kept fighting after suffering casualty rates, relative to their population, enormously higher than the rates that drove the U.S. out of the war. The reason is obvious: What happened to Vietnam mattered a lot more to the Vietnamese than to us.

Justice vs Efficiency: An Example

One way in which a legal system discourages people from doing things that inflict costs on others is by fining them. Roughly speaking, there are two alternative approaches. Under criminal law, the fine goes to the state. Under civil law, it is called a damage payment and goes to the victim.

A variety of arguments can be made for one approach or another; interested readers may want to look at the chapter of my webbed Law's Order devoted to the general question. One interesting feature of the analysis is that the straightforward argument based on justice gives the opposite answer from the straightforward argument based on economic efficiency.

The justice argument is that the money should go to the victim, in order to compensate him for his loss. Indeed, the standard rule in tort law is that the damage owed are enough to "make whole" the victim.

For many, although not all, cases the argument from efficiency goes just the other way. The reason is that, even if the victim is not morally responsible for the loss, in most cases he makes decisions that affect how likely it is to occur and how large it is if it occurs. There is nothing inherently immoral about driving an expensive car, but the more expensive your car is, the larger the damages if my negligent driving results in my running into it. There is nothing inherently immoral about walking after dark in bad neighborhoods, but doing so can substantially increase the chance of being a crime victim.

From the standpoint of economic efficiency--loosely speaking, maximizing the size of the pie--we would like people to reduce the likelihood and cost of their being victims whenever the savings are larger than the cost of doing so. But if victims were fully compensated, as in theory under tort law they are, they would have no incentive at all to take precautions. Hence the efficient rule, for a wide range of cases, is that the money should go to anyone but the victim (or the perpetrator).

[I've linked to the HTML version of a late draft of my book. There is another version that consists of page images of the book as published with virtual footnotes--links to icons in the margin.]

Wanted: The Inn Between the Worlds

You are a player of World of Warcraft or some other massively multiplayer game. At some point, after you have been playing for a year or two and have established a network of friends and acquaintances, the server you play on gets too crowded; Blizzard announces that they are going to split it in two.

One result is to reduce lag due to overcrowding. Another is to split friendships. Some players go one direction, some the other, and never the twain shall meet.

I propose a simple way of ameliorating the problem: The Inn Between the Worlds, a server to which everyone, on every server, has access. Unlike ordinary servers, it does not represent a world with monsters, quests, and other processor intensive features. It is merely a (very large) inn. It exists to let old friends meet again, tell each other of their adventures, trade lies.

A similar but more difficult problem exists as new games come on the market. Some players leave the old game, some stay. Could a version of the same solution still work?

I think it could, although how well it works depends in part on the willingness of the companies that own the games to cooperate. I am imagining a game which is, again, merely an inn. It differs from all other games in that it allows players to bring into it their characters from other games. Since all they are doing is sitting, talking and perhaps showing off their possessions, the problems of interoperability should not be too serious. Legal problems could be more serious; it is unclear to what degree hosting my WoW character in someone else's game, complete with clothing, name, physical appearance and (inspectable but inoperable) gear, would violate Blizzard's legal rights. But that isn't a problem if Blizzard cooperates.

An encouraging sign is a recent story about plans by Second Life to open source its software, which should permit other people to build compatible worlds.

NSA, state secrets, and bonding: A suggestion

A government agency does something that is arguably illegal and ends up in court. Their attorney claims, perhaps truthfully, that in order to defend the case he would have to reveal facts about the agency's activities that must, in the interest of national security, remain secret. He accordingly asks the judge to dismiss the charges.

This situation--I'm thinking of litigation over the National Security Agency's wiretapping and related activities--raises a real problem. One can imagine a similar problem arising in a patent infringement suit when the defendant argues that, in order to show it is not infringing the patent, it would have to reveal a valuable trade secret. Our usual mechanism for settling legal disputes requires both sides to have free access to the relevant information; in some cases giving them such access could result in very large costs.

There is, I think, a fairly straightforward way in which we might deal with such situations. It depends on two assumptions. The first is that there is some responsible person--the President, the head of NSA, the research director of the corporation sued for patent infringement--who already knows the secret. The second is that the secret will expire within some predictable length of time. If the NSA is using top secret high tech to data mine phone records, it is unlikely to still be using the same method ten years from now.

Under my system, when the attorney asks for dismissal on grounds of state secrets or the equivalent, his request must include a time limit after which he agrees that the information will no longer be secret; the longer the time limit, the less plausible his claim and so the less likely the judge is to accept it. Prior to dismissing the claim, the judge asks the relevant responsible person, the one who already knows the secret, to testify under oath that the information does not confirm the other side's case, that revealing it would not result in a near certain verdict of guilty. When the time limit on the secret has expired the court takes up the case again, determines whether the testimony was true, and imposes penalties for perjury if appropriate.

Before taking the testimony the judge informs the witness that perjury in this particular situation is a capital offense. "Now, Mr. President, would you like to reconsider your position?"

That is, of course, an implausibly extreme version of the proposal. More generally, in order for the case to be dismissed, someone with the information must testify as to the implications of the information for the case, with penalties for false testimony severe enough to make it unlikely that he will be willing to perjure himself. In effect, he is bonding his testimony--in the limiting case with his head.

Obviously there are some practical problems, beginning with the fact that the person who knows the secret is probably not a lawyer. He should, however, have access to lawyers; it should not be that difficult for him to determine what sort of facts would or not result in his side losing.

How do we currently deal with such situations? One approach is to provide the secret information to the judge but not to the opposing side. But without the usual mechanisms of an adversarial trial, the judge may have a hard time telling how true or how complete the information he is given is. If he later discovers that what he was told was deliberately misleading, he has the option of imposing penalties for contempt of court, giving at least a weak version of my proposal.

Where do smart cheapskates go to college?

My daughter plans to apply to college next year and we've started our search with visits to Stanford, Scripps and Pomona and a good deal of research. One thing that struck me was how luxurious the elite colleges seem to be--a wide variety of good food, fancy architecture, security escorts home from the library at the touch of a button, very small classes, lots of hand holding, a general "cost is no object" style.

Part of the reason, presumably, is that students with very high SATs, grades, etc. disproportionately tend to be the children of well off parents. While college purports to be about education, a large part of its role in our society is as a place where people can spend four years enjoying themselves, searching for friends and mates, developing useful social contacts, and the like. If the children of well off parents are going to spend their time that way, they might as well do it in comfort. And for very smart applicants whose parents aren't well off, the schools have an extensive system of discriminatory pricing aka financial aid.

On the other hand, most of what these schools are spending money on seems to have a rather tenuous connection to the quality of education. That left me wondering whether there are any schools that specialize in smart cheapskates--provide a good education in the company of smart people at the lowest practical cost, which I would expect to be under half the cost of the elite schools we've been looking at.

Examples?

My case against TSA

Shortly after the system of luggage searches by TSA went into effect, some inexpensive jewelery in an outer pocket of a suitcase I checked disappeared. I suspected pilfering by whoever searched it, but since I had no way of either proving anything had been stolen or identifying the person responsible, I did nothing.

On my most recent trip by air, I had two suitcases--a big metal one and a smaller soft sided suitcase. The metal suitcase had a combination lock which I left unlocked, since otherwise TSA was likely to break it. Since without the lock I thought there was some risk of the suitcase opening, I put a luggage strap around it. The strap had no lock--you just pressed the sides of the catch to open it.

When I picked up the suitcases, the strap had been broken open. The other suitcase has two built-in straps to cinch it tight. Both were unfastened and hanging loose, at obvious risk of getting tangled in the luggage handling machinery. Nothing stolen, so far as I know, but the wanton destruction of about ten dollars worth of property and a risk of further damage.

Obviously, even if I knew for certain that TSA employees were responsible in both cases, that would not demonstrate that the organization itself was at fault--even a well run and well intentioned organization may sometimes hire a dishonest, irresponsible or careless employee.

My case against TSA starts by asking what they would do if they cared about such matters--if they were concerned to prevent vandalism or pilferage by their employees. The answer is pretty clear. When they search your luggage, they leave you a printed note telling you that they did so. All they would have to do in order to identify employees responsible for stealing or damaging property would be to have each note stamped with the name of the employee who did the search. If they got complaints from multiple passengers about the same employee they could then investigate further.

I take their falure to follow that policy or something similar as clear evidence of culpable negligence, reason to think that they don't actually care. Am I missing something?

Two Views of Democracy

Listening to a radio program about recent Nigerian politics, it occurred to me that there are two quite different models of how democracy works held by people who think it does work. One is a model of incentives, and one of selection.

The incentive version relies on the fact that politicians want to get elected and reelected. To get elected, they have to support the policies that a majority of the voters want; to get elected again, they have to actually work for those policies. Hence, it is argued, it is in the interest of politicians to do what the voters want them to do, whatever the private beliefs and objectives --other than election--of the politicians may be.

The selection model relies on the facts that people differ and that we have some ability to tell what other people are like. Some candidates are good people who want good things--the welfare of their fellow citizens, justice, peace, ... . If the voters can identify the good politicians, they can elect them and rely on the benevolent desires of those politicians to motivate them to do the right thing. Some candidates are not only good people but smart and hard working, hence likely to take the acts that produce those good things.

The two models have different requirements and different implications. For the incentive model to work, voters must know both what elected politicians are doing and what they ought to be doing, in order to vote for the ones who do what they ought to. In a large society with a government doing many things, knowing either what it should do or what a particular politician is doing is hard, which is one reason that model might not work very well. It's a common observation that when a politician wants to buy the votes of farmers by pushing up the price of food, he explains that he is doing it to guarantee American consumers a reliable food supply. When a politician wants to buy the votes of auto workers and GM stockholders by using a tariff to drive up the price of automobiles, he explains that he is doing it for the health of the American economy.

The selection model avoids that set of problems. You don't have to know what the government should be doing; you have delegated the job of figuring that out to someone else. Nor do you have to know what your politician is doing. The selection model even works for issues of national security where the voters know that they don't know what the government is doing--the Manhattan Project, say. On the other hand, it depends on voters being able to accurately evaluate the personalities of people they have met for at most a few seconds, based on what those people say about themselves and what others say about them. It thus leads to campaigns of charisma, bogus virtue, and the like.

My point here is not to judge which model is more realistic or which works less badly. It's merely to point out that people who believe in democracy are likely to have one of those two models in mind and are often unclear about which it is or what the differences between them are.

International Pixel Stained Technopeasant Day

As my contribution to Jo Walton's "International Pixel-Stained Technopeasant Day" I am webbing the first two chapters of my new novel Salamander.

Obesity: A Conjecture

Obesity is a current hot issue, problem, crisis ... . One reason is that it is a real problem. Another, I suspect, is that it provides people who want governments to do things with a new argument.

My evidence for that conjecture is how much of the talk about obesity focuses on the evils of marketers cleverly manipulating people into eating junk food. While sellers of junk food do, of course, advertise their products, so do sellers of diet soda, exercise equipment, metrecal and health foods. I understand why people concerned about obesity might see the regulation of advertising as a potentially useful tool--it is at least more likely to be politically viable than an attempt to ban hamburgers and french fries from the American diet. But I do not see marketing as a plausible explanation for the increasing frequency of obesity. It seems particularly implausible given that the increase is not limited to rich countries such as the U.S.; I doubt the consumption patterns of people in India or China are much influenced by advertising.

My alternative explanation for obesity is straightforward. Humans evolved in an environment where food was costly, fat scarce, sweetness a useful signal that fruit was ripe. We are designed by evolution to put on weight when we can as a precaution against future famines and to favor fat and sugar when we can get them. In a world where food is inexpensive and plentiful we are inclined to overeat, in particular to eat more fat and sugar than is good for us.

The obvious explanation of the increase in obesity is that real incomes around the world have been trending up for decades. Now poor people in the U.S., and increasingly in poorer parts of the world, can afford to eat all the calories they want. Since all the calories they want represents more than what they require, the result is that they get fat.

There is one problem with this explanation. According to the figures I have seen, in the U.S. obesity is less common in high income groups than in low income groups. The richer you are, the less your diet is constrained by cost, so we would expect higher income groups to be at least as obese as lower income groups. To explain why they are not I must add one additional factor: Time lags in adjusting behavior and social norms to changed circumstances.

Suppose you are part of a population where food has been costly, where people engage in a lot of physical labor, and so where the problem is getting enough to eat, not avoiding too much. You, and those around you, have adapted their behavior to that environment.

Now things change; food gets cheap, wages go up, almost everyone can afford to eat as much as he wants. For a while, perhaps a generation or two, people follow the old patterns in the new circumstances; the result is that many of them end up fat. Over time, although the hardwired elements of behavior do not change--evolution is slow--the cultural elements do. Instead of demonstrating how wealthy and generous you are by urging your guests to have a second and third helping of dinner, you do it by providing them smaller amounts of particularly tasty, sophisticated, or expensive dishes. Instead of making a point of avoiding physical exertion when you can, you enter the Boston Marathon. Eventually you and those around you have adapted your behavior, although not your hardwired tastes, to the new environment.

Well off people in developed societies have been able to afford second and third helpings at every meal for a long time. Hence, if my argument is right, they have had time to adapt to a world of plenty. For poor people, being able to eat all they want of more or less what they want is a newer thing, so they are still following the old ways--the pattern of the traditional Jewish (or Italian) mother who insists that her guests have a little more of this and that before they end their meal. Hence, if my conjecture is correct, greater obesity among the poor reflects the lag in adapting to circumstances that are relatively new for them. The rich have had time to adjust.

One implication of this is that, at some point in the past, richer people should have been more often obese than poorer--back when the rich were no longer constrained by the availability of food but the poor still were. That fits my casual impression, but I have no actual data to support it.

Cell Phones, Mythbusters, and Selective Perception

In a recent Usenet discussion, the question of whether cell phones on airplanes were actually a problem came up. Several people mentioned a Mythbusters episode on the subject. What was interesting was that some of them claimed the episode showed that cell phones were not a problem, others that it showed that some cell phones were a problem. Googling around, I found the same inconsistency in webbed comments by various people citing the program.

The situation became clearer when I found a fairly detailed webbed description of the program. The people at Mythbusters were unable to get permission to actually test a cell phone in an airplane in the air--cell phones being forbidden to operate in airplanes in the air. Instead, they did two experiments:

1: A signal generator to simulate a cell phone, a navigation gauge, and various other equipment were placed inside a faraday cage (to eliminate external interference). They found that signals in the 800-900 MHz range affected the navigation gauge.

2. An actual airplane was borrowed and tested on the ground. No signal they tested had any effect on its electronic equipment. They concluded that the difference was that the electronic equipment in a real airplane is shielded, whereas in the first experiment the signal generator was more or less next to the unshielded navigation gauge.

So it is true both that they found that a cell phone had no effect on an airplane's electronics and that they found that some cell phone signals could affect the kind of electronic equipment in an airplane. I suspect that listeners selectively remembered the part of that that fit their existing beliefs and prejudices.

I also found a discussion of the issue of cell phones on airplanes, from the sceptical side of the argument, in a Computerworld article. The author thinks that both the regulators and the airlines are in favor of the ban for reasons unrelated to safety risks from interference. The article provided a link to an FCC press release which amounts to "we decided to stop our investigation into whether cell phones should be banned on airplanes because we couldn't find enough evidence to tell," which strikes me as some evidence in favor of the sceptics' view of the situation.

The Cost of Regulation

My previous post reported on my experience buying prescription eyeglasses online and raised the puzzle of why they cost five or ten times as much from a store. One commenter provided enough information so that I was able to locate the relevant California regulations.(specifically paragraphs 2559.1 and 2559.2 on pages 163-4). In order to dispense eyeglasses you must employ a registered spectacle lens dispenser and in order to be a registered spectacle lens dispenser you must pass an exam provided by the American Board of Opticianry (I'm leaving out some detailed exceptions and qualifications).

That means that the profession of eyeglass sellers is in a position, like other licensed professions, to control entry, to make it hard for people to join the profession in order to hold up the income of those already in. I have no data on how restrictive they actually are; perhaps someone else does.

But I do have some data on the outcome. Prescription eyeglasses, in my experience, cost from about $100 to $300 in stores in California. Online they cost about $20. If we take the bottom of the range for comparison, on the theory that the higher prices may represent either higher quality or additional services, and if we assume that the cost of selling online is only half the cost of selling in a store, that suggests that the effect of the regulatory restrictions is to make the cost of eyeglasses about two and a half times what it would otherwise be.

A little googling turns up a figure of 15.8 billion for optical retail sales in 2001. 1.9 billion of that was contact lenses. Combining this information, it looks as though restrictions on competition in the eyeglass industry cost consumers about ten billion dollars a year.

This is, of course, only a very rough estimate, but still an interesting figure. There is apparently an old literature that looked in more detail at the consequences of particular restrictions, and found large effects.

Needed: Voices for World of Warcraft

Players of World of Warcraft sometimes use software--the programs I am familiar with are Teamspeex and Ventrilo--to talk to each other while playing. This is particularly useful for large group activities--raids--where many people are trying to coordinate what they are doing.

There is, however, one small problem. The character you know as a middle aged male dwarf--that being how he appears on the screen--is speaking with the voice of a teen aged girl. That makes it hard to stay inside the fantasy, to pretend the game is real while you are playing it.

There is an obvious solution. Techniques exist for changing the sound of a human voice. Someone should write software, designed to be used with such programs, that automatically alters the player's voice to fit the character. The speaker would set preferences: Age, race, gender. The software would modify voice accordingly.

Or does it already exist?