It should be noted that Arab countries engage in ethnic profiling with regard to Arabs or people from non-Arab Muslim countries with great consistency. A Jordanian travelling to Algeria will be checked much more carefully than a person of European descent. It is not a matter of prejudice, but rational security policy.

I take a somewhat different approach to profiling of minorities in the crime context. While I do not accept Posner's argument regarding the alienation of Muslim Americans, since I consider the benefit to outway the likely harm, I think that argument applies forcefully in cases where something like shoplifting is at stake. I'm fine with someone discriminating to save his life, but alienating innocent black Americans further from society is not worth preventing petty theft at Macy's.

I've got an uncontroversial statistic for you:

Number of Terrorist Acts prevented by search
and capture of intending terrorists at airport
security checkpoints from 1920 - 2005:

ZERO (including Sept. 11th)

There is no "current airport terrorism" to
reduce. If the TSA had even a single success
to report you can be sure they would still
be shouting about it.

Given this data about airline security apprehension rates, it is clear to me at least that current racial profiling policies are the result of discriminatory intent rather than efficiency.

But either way, profiling IS discriminatory, the question you are asking is "does the data allow us to conclude that it is efficient to discriminate in this way"

"Limited information about individuals means that group identities is [sic] often useful in gaining information about them."

NONSENSE. Limited information about individuals means that more information about individuals should be acquired. Scenario:

67% of XYZ-Americans eat gumbo. Jack is an XYZ-American. Jack does not eat gumbo. Assuming that Jack eats gumbo does not help us acquire any useful information about Jack. In fact, if we serve gumbo to Jack, he might eat it, because he is willing to tolerate our good-natured hospitality, but he will never return to our house again, because he hates gumbo. Not only did we not learn the truth about Jack (he hates gumbo, unlike our presumption), but also we failed to learn any additional information about Jack (because he now wants nothing to do with us).

A simple solution to this problem is asking Jack whether he likes gumbo before purchasing its ingredients or googling a recipe.

"Similarly, innocent shoppers who are stopped and searched could be compensated for their embarrassment and time."

There is no compensation for loss of dignity. Dignity, for non-economists, at least, can be non-monetizable, or irreducible to numbers. If dignity is priceless to me, there is no compensation. I simply don't want to be profiled.

Scenario: I am blown up in a car, fire roasts off all my flesh, but I get billions of dollars and my skin and health is completely restored by perfect medicine. I am better off, because my health is the same and my wallet is fatter.

Except I had to endure going through being roasted. And going through being roasted is something I never wanted to happen, no matter how much better off I would be in the end. In other words, I am not better off, because no matter how much money you give me, I just didn't want to have my flesh roasted off, ever, under any circumstances. I will be an angry, unhappy, probably suicidal billionaire.

"A further test would be to determine what happens to apprehension rates as the amount spent on airport security increased or decreased."

To avoid the appearance of discrimination, these records must be transparent: open to the public. If accessible to the public, terrorists would have access too. If terrorists know that their distributions of money, arms, etc. is being tracked by age, race, religious affiliation, then they would simply diversify their distributions among enough age groups, races, and putative religious affiliations to make the tracking system cost-ineffective.

Corey- What exactly is your "uncontroversial statistic"? Are you aware that the current racial profiling policy is that any profiling by an airline or security checkpoint screener on the basis of race or ethnicity is specifically prohibited? A number of airlines and individuals have been found to have violated DOT regulations on racial profiling and have incurred civil penalties as part of the punishment. This was put in place my Sec. of Transportation Mineta even prior to September 11th.

I gather your underlying point is that racial/ethnic profiling is de facto discrimination and fails to prevent terrorist acts. If I stated that correctly, I agree completely. However, profiling is not the current policy and therefore your claim about discriminatory intent is rather confusing.

Your assertion about no risk of airport terrorism is flat out wrong. Your only evidence to back that up is some claim that if there was this risk, the TSA would be publicizing it, since they aren't, therefore no risk. As I am intimately involved in airline/airport security, you could not be more misguided.

Hi,
Ethnic profiling is probably useful in examining the bulk properties of a population (for example, medically, one could study rates of heart disease etc as a function of the ethnic group, and draw useful conclusions.). However, the problems with security screening based on ethnic group are two fold:

1. Screening based on ethnicity assumes that certain groups are more likely to commit a particular crime. This is likely enough to be a self defeating assumption, because when certain groups are monitored more closely, the probability of criminal action by members of groups not subject to such scrutiny is underestimated. An example is the attempted terrorist strikes by members of apparently low risk groups like the shoe bomber Richard Reid, or Jose Padilla.

2. As an addendum to (1), criminal enterprises targeted by said screening are liable to exploit the varying emphasis on different groups, by changing their activity to groups with less high visibility.

3. A sufficiently fanatical group bent on committing terrorism is very much deterred by the possibility of being caught before the execution of its mission, but may not be concerned about the
post-facto capture or killing of its members. Thus, there is a fairly strong deterrent effect due to careful screening.

4. However, the effectiveness of rigorous screening based on ethnicity/religion also depends on an ability to discern beforehand, the said characteristics of each individual passing through screening. This, to my mind, is not always possible. The perpetrators of a terrorist act need not neccesarily fit such a narrow profile. For example, the first attempt in 1993 to blow up the World Trade Center was carried out by a non-Middle Eastern Muslim, who would have been hard to distinguish from a very large number of other asians. Similarly, the Oklahoma Bombings were carried out by white americans, (of course they did it for different ideological reasons), and Mr Richard Reid was not Middle Eastern as well.

I would argue that using a narrow categorization like ethnicity would impair the ability to prevent acts of terrorism, given that it is relatively easy for an international terrorist organization like Al Qaeda to use recruits who would not fit the requisite profile. And if one wants to use religious affiliation, it is easy to disguise and hide one's choice of religion.

The whole idea that aQ will used planes again is flawed. They will shift there strategy. On the otherhand running/gunning the cockpit in a 737 might be a possibility in Tijuana.

Mexican airliners often fly with there cabin doors open on flights over Mexicon not bound for US airspace. One of the flights can prang anywhere in Southern CA before an F-16 on hot standby could intercept.

Spending millions of dollars to be PC at airport gates is not reasonable either when the relative probability of an elderly female being a suicidal Jihadist terrorist is almost nil. The chances of winning the Cal Lottery are much greater.

See an anthology of essays on this topic on our site I posted at Roger L. Simon's site:

RLS Link Here

Israeli security takes a whole different approach. Ask the traveler a series of question where they are going instead of insisting on ID (probably counterfeit anyway).

Operatives are trained to give detailed replies on ID checks. But on open questioning it's hard to keep your story straight when you run out of your script.

I think Steve Lansberg has a good suggestion that the airport should compensate individuals being searched. So "profiled" individuals who may be searched more often is paid more often. The rest of us who benefit from the profiling should be willing to pay for it.

Re "v" and Richard Reid

Contrary to public opinion, Reid was not the "village idiot" as he is so often portrayed.

He and another Islamofascist suicide bomber both met up at a radical mosque in London.

The plan was both would be on flights. One over Europe and Reid over the Atlantic. They both had shoe bombs.

The other was later arrested and another shoe bomb was found.

Reid did other surveillance flights for aQ to Israel and later made a detailed report which was later found I believe in Pakistan.

Reid had tried to board the same flight the day before but was gone over with a fine tooth comb and was not allowed to board the flight. The only thing of interest was a lighter. He was wearing he shoebomb that day.

Reid came back the next day and was allowed to board because he was screened so throughly the day before. This time he left his lighter behind and brought matches instead

This is where is fell apart because he deviated from the plan. You see he was not an explosive expert and the fuse was not cut in a manner conducive to being lit with a match. Not for lack of trying the fuse would not light.

Well, lets consider the following:
- searching people is costly, and there are limited resources available to security agencies
- therefore, searching a particular ethnic group more carefully necessarily means not searching another group as thoroughly.
- If we assume that radical groups are aware of this profiling, than we can easily see how they would "beat" such profiling. The use of women suicide bombers in Isreal is such an example.
In other words, the use of profiling based on simple broad categories that are easy to disguise might, in the end, actually make things easier for terrorists. Knowing that an young Moslem man would almost certainly be inspected would certainly prevent terrorists from using people who fit such profile. But it would encourage them to use older women, or males that appear to be caucasian. This is the problem of using ethnic profiles against organized groups: they might be large enough to include those who would easily pass through a system where the probability of being stopped and searched is very unevenly distributed.
This is a different matter than profiling with regards to shoplifting, where the goal is to prevent certain individuals (rather than large organized groups) from doing something.
I think that is the main difference between the use of profiling by insurance agencies and stores and the use of profile by national security agencies: in one case you are trying to stop individuals from doing something, while on the other your are trying to stop an organized group from doing something. As long as the organized group does not care which of its members carries out an attack, simple ethnic profiling would simply make things easier for them.

oops.
I should have read Posner's entry before making my commment, as he made the point I was trying to make much more eloquently and efficiently.
My apologies

I've got an uncontroversial statistic for you:
Number of Terrorist Acts prevented by search
and capture of intending terrorists at airport
security checkpoints from 1920 - 2005:
ZERO (including Sept. 11th)

Not only is that not "uncontroversial," but it's silly. The test is not how many terrorists we catch. The test is how many terrorists we stop. Every would-be terrorist who chooses a tactic other than hijacking because he doesn't think he can make it through the checkpoints is a success for airport security.

You can't see terrorists who are deterred easily, the way you can see terrorists who are arrested, but that in no way means that the number of the former is "zero."

Young Muslim men are responsible for more than 80-85% of all terrorist attacks against the US since 1980, *excluding* 9-11. Of ca 1500 dead in the attacks over 1300 were killed by Muslims, and 'only' some 180 killed by non Muslims, mostly in Oklahoma city. (I am also excluding terrorist attacks in Iraq and Afghanistan).

If we assume that 20% of the population are "young men", young Arab men are about 0.1% of the US population (1.2 million Arabs in Census 2000) and somewhat less of 1% of the world population. The relevant share is probably a weighted average of the two, but let's use the higher number. Let's also be generous, include Persians, us Kurds, turks and other mid-easterners and double the numbers. We still get that young mid-eastern men are approx. 200 times more likely to commit acts of terrorism than everyone one else. As long as the attention paid to this group is not two hundred times higher than the others there is no cause to suspect bigotry.

That "racial profiling never works" stuff is simply an myth. First of all as someone noted because of the nature of crime-prevention we can never know how many more attacks that would have succeeded if we did nothing. But just to give two examples of cases where (probably illegal!) racial profiling against young men from the middle east saved live:

1. As Newsweek reports that the 20th hijacker, the planned fifth hijacker of flight 93, was stopped by, you guessed it, racial profiling. The Saudi al-Qahtani entered Orlando in august 2001, but was turned away by Customs officer Jose Melendez-Perez. This is likely the reason the passengers were able to defeat the terrorists.

As noted by J. Lehman: "Melendez-Perez did this at great personal risk, because his colleagues and his supervisors told him, "You can't do this. This guy is an Arab ethnic. You're racially profiling. You're going to get in real trouble, because it's against Department of Transportation policy to racially profile." He said, "I don't care. This guy's a bad guy. I can see it in his eyes."

2. Ahmed Ressam was stopped on Dec 99 by a U.S. Customs Service from entering the US from Canada. The agent had no real reason to stop him except the fact that he looked and seemed to act suspicious. We now know that he was an al Qaeda with over a hundred pounds of powerful explosives in his car trunk, planning to blow up LAX.

To say that scrutiny only makes it easier for terrorist is overlooking the difficulties to be a successful terrorist. For one thing the pool of willing and able recruits for suicide-bombers is not endless. There are not that many jewish grandmothers or blond Scandinavian businessmen who are lining up to kill themselves in the name of Allah. Women are both less likely to be terrorists and on average less skillful (especially in misogyn Arab culture, Chechens do better). That's why even in Israel, the country where the costs of terrorisms is the lowest, the share of female or non-Arab terrorists is very low.

PS. Not that it should matters, but for the record I am myself a young mid-eastern male. I would not feel safe when traveling with an airline that did not give me extra scrutiny. The fact that I am not a terrorist is private information, and the airlines not knowing it is nothing a rational person should be offended by.

http://msnbc.msn.com/id/3989726
http://www.state.gov/r/pa/ho/pubs/fs/5902.htm
http://www.infoplease.com/ipa/A0001454.html

If this policy of slight profiling were common knowledge then it would or should be overturned for violating the Fourteenth Amendment. If so, it would only add ammunition to the argument, domestically and abroad, that America is racist, anti-Muslim and so forth. The more vigorous people would call it apartheid and draw parallels to alleged Israeli discrimination.

The alternative is that it would be performed but only in an unofficial and unacknowledged manner (or in an intentional manner that is then hidden from the public). In this case, the backlash would be less credible and have less ammunition. However, the government would be unaccountable for actions that directly violate the Fourteenth Amendment. It would hide unconstitutional police and semi-police activities. That is not a very pretty picture - or precedent.

Maybe it would 'work' in the short run and in the long run Muslims and Arabs would shrug off past discrimination as the Irish, Italians, Chinese, Japanese, Jews and so forth have done. But it wouldn't be constitutional and it would fuel wild-eyed crazies, who in turn extrapolate some sort of Zionist-GOP-neocon conspiracy because of a superficially minor incident of discrimination.

This distinction can be made in the terrorist field ...by keeping records on the fractions..... If the fraction were much greater among Moslems searched than among others, this would at least be consistent with an emphasis on efficiency rather than discrimination."

This strikes me as deeply faulty. If there are 100,000 would-be terrorists in the U.S., 95% of them Catholic and 5% left-handed Baha'i women, you're going to strike out an enormous fraction of the time you investigate a Catholic, because there are millions of them, while the 5,000 terrorists who are Baha'i probably represent a very large fraction of all left-handed Baha'i women in the U.S., and perhaps every other search of someone with this profile will bag you a terrorist. So when we compare our apprehension rates with Baha'i versus Catholics, should we conclude we're acting prejudicially against the Catholics and restrict ourselves to other profiling criteria for sniffing out the 95,000 Catholic terrorist sleepers? What if we have no other criterion, and all we know is that 95% of the sleepers are Catholic? It seems like a smart person would save most of his or her scrutiny for Catholics and expect to experience a lot of eye strain before he or she catches a terrorist.

Corey is wrong about no recorded preventions. There is a famous case of a Palestinian who tried to blow up an Israeli airliner by sending his pregnant Irish girlfriend on board with a bomb in her luggage. The El Al screeners in London profiled a young pregnant woman traveling alone as suspicious, gave her a thorough search and found the bomb. Google "Nezar Hindawi".

I am guessing that there are more cases, but all sides have an interest in keeping quiet.

Profiling is not infallible. Like any security device, it makes life more difficult for the would be criminals. The question is, are we willing to give up this device in the name of political corectness?

The appearance of avoiding racial profiling at US aiports has gone to extremes. I read last year that Al Gore was frisked at an airport security station.
Many people don't like Al Gore for good solid reasons, but no one, no one would think Gore is a likely terrorist suspect.