« Terrorism-- Posner's Comment | Main | Response to Comments on Terrorism--Posner »



Feed You can follow this conversation by subscribing to the comment feed for this post.

Gabriel M.

As with any trade-off, it depends on the individual preferences of those in question. Who are we to say that what is needed is more security and less privacy?

Fighting terrorism is as efficient as possible given the constraint of private liberties. If it's inefficient than it can reform but under the same constraint.

I'm all for profiling. I think it's one of those improvements that don't require further degradation of individual rights.

Other than that I think Prof. Becker is wrong in imposing his own judgement, erudite as it might be, over the actual choices of those involved. His preference is clear but in no way does it reflect objective facts but rather what one individual thinks the state "must" do.

Matt Rognlie

In my mind, means of combating terrorism fall roughly into four categories:

1. Military -- active engagement of external force, either to induce structural changes that will presumably lower the incidence of terror (Iraq) or to directly confront terrorist operations (Afghanistan).

2. Removing incitements to terror -- this category would include, for instance, a change in Israel policy.

3. Other outer defenses -- in particular intelligence and investigative work; this category encompasses most of what Judge Posner labels "outer defenses."

4. Inner defenses -- essentially the same as Posner's definition.

I think that #3 is most important, #2 is undervalued, #1 is appropriate in rare circumstances, and #4 is necessary but wholly insufficient. I'll elaborate upon these positions in numerical order.

First, military action has a place in combating terrorism, but I feel that it is easily overused and often counterproductive. The war in Afghanistan is an example of appropriate military power: a state-sponsored terrorist apparatus is a clear target widely regarded as legitimate. In such cases, there are still potential negatives, including the incitement of further terror, but it is tremendously important that terrorism be marginalized and not allowed to fester in a particular state.

I am very skeptical of vaguer attempts to shape the world through military means. In particular, I think that an emphasis on democracy conflates correlation and causation. While first-world democracies are not hotbeds of violent radicalism, there is no conclusive evidence that democratic institutions themselves prevent terror. Instead, events suggest that well-established democracy and peace tend to coexist due to shared precursors: tolerance, prosperity, and political stability. The elected ascendancy of Hamas and Hezbollah (which, I should note, would have even higher representation in Lebanon's parliament if the voting system did not disadvantage Shia Muslims) contrasts unfavorably with the relatively moderate, undemocratic regimes in Egypt and Saudi Arabia. When the standard precursors to democracy do not exist, voters may elect unfavorable radicals -- hardly a desirable outcome!

With the benefits so uncertain, the considerable downside of war -- death and further radicalization -- prevails here.

On the second point, removing incentives toward terror need not be outright capitulation, and it's an important tactic. For instance, we hold a very important position vis-a-vis Israel. As the voluntary guarantor of Israel's security, the United States has the right to question its policies and demand more favorable results. In the Lebanon morass, we could (and should) have demanded a more measured initial response. We are under no obligation to support the suffering and subsequent radicalization of an entire nation.

Finally, our other "outer defenses" are critical in fighting terror, as the inner defenses will inevitably be vulnerable. There is no way, for example, that we can design a system preventing trucks laden with explosives from travelling over bridges. Indeed, the current air travel fiasco highlights the weaknesses of inner defenses: despite airports' relatively secure status, we cannot find a foolproof way to stop liquid explosives.

This leads me to disagree with your support of racial profiling. I'm not against all profiling: increased surveillance of young males in general is both prescient and necessary. But we can profile "young males" only because they're too big a group to alienate. It's not as if they will segregate themselves from mainstream society, feeling increasingly isolated as the more extreme among them tend toward violence -- fully one-half of all Americans are, were, or will be "young males."

Young Muslims, on the other hand, form a very small part of the total population. Might racial profiling start a process of alienation that would destroy the generally good relations American Muslims have with society? It's quite risky. Policies in Europe that distinguish or offend Muslims have caused significant friction. Britain's main terror threat is homegrown.

Without cooperation from the Muslim population, our "outer defenses" would significantly weaken. Say that the average terrorist knows five ordinary, non-terrorist Muslims who might catch wind of a plot. If these individuals are friendly toward the government, they'll each tip law enforcement with a certain probability - let's say 80%. If they're not, the probability may decline substantially: perhaps 30%. If these probabilities are independent, in the first case we have a 99.97% chance of identifying potential terrorists, while in the second it's only 83.19% -- a serious failure rate.

This is, of course, an incredibly crude approximation of the dynamics at play. But recent cases indicate that human informants are indeed our most important defenses against terror, and changes in attitude can make an enormous difference in the quality of the intelligence we receive.

Compare this to the temporary and uncertain gain from singling out Muslim men at airports. This might prevent most airport attacks -- but then terrorists would shift their sights to other goals, probably just as destructive. Meanwhile, our capacity to stop these new attacks, predicated on strong human intelligence, might no longer exist.

Maintaining good relations with the American Muslim population is our most important tactic in the fight against terrorism. Although "soft" measures like intelligence don't seem as comforting as metal detectors or baggage screening, they're ultimately the most effective.

Matt Rognlie

I should add a point in support of tactic #2: many observers claim that terrorists will target us regardless of our policies. This is probably true. There will always be some terrorists.

But "some" is the operative word. Terrorism is not a switch: it isn't simply "on" or "off." There are varying intensities, and I suspect that incitements to terror significantly impact these intensities.

Matt Rognlie

I'm afraid that I have one more housekeeping note: in my main comment, I accidentally wrote "incentives toward terror" when I meant "incitements toward terror." Normally this wouldn't be a big deal, but on an economically flavored blog I can't afford to discuss "incentives" inaccurately.


Professor Becker is correct. There is always some tradeoff between liberty and security. Anarchists and fanatical libertarians might not see this, but most people will understand that one cannot expect the same degree of freedom in a time of war that one would expect in a time of peace.

All the liberty in the world is of little value to a corpse. Liberty is only of value to the living--which means that it presupposes an environment of order and security.

Richard Mason

Q: There are seven men waiting in line at your security station: one of Jamaican background, one of Puerto Rican background, one elderly Asian gentleman, and four young Arab men. Given limited resources, which passengers will you subject to close scrutiny?

A: All seven of the men are terrorists. The young Arabs are specifically hoping to attract attention to themselves and divert it from the other three passengers, who happen to be Richard Reid, Jose Padilla, and Abu Bakar Bashir. The non-Arabs are carrying weapons which they can distribute to the Arabs on the other side of the security checkpoint.

As illustrated by this example, profiling does not necessarily yield more safety in return for a little injustice. Instead, profiling itself creates a predictable vulnerability that terrorists can and will exploit. (Suicide bombers in Israel have disguised themselves as Orthodox Jews, as young Israeli hipsters, as pregnant women, etc.) A predictable sampling scheme is easily defeated.

Also as illustrated in this example, the fact that a majority of the terrorists fit the preconceptions of the security guard is not necessarily relevant. A single individual who could expect to pass with reduced scrutiny could be sufficient for the terrorists' plot to succeed.


Richard Mason's argument is not persuasive. Since Al Quada has an easier time recruiting young Muslim men than others--it makes sense to focus attention principally on them. But "principally" does not mean "exclusively."

Profiling should be one of the techniques used, but that does not preclude some measure of random testing as well. Ignoring probabilities is just not rational. Devoting the same level of scrutiny to 90 year old white women as to 20 year old Arab men men will waste resources and increase the danger.


Not so long ago, some people in the USA thought it was OK to go kidnap some people in Africa and and then force them to work for free in the USA. Whenever one group of people gets to make laws that will absolutely never apply to themselves and will instead only apply to other people that they are scared of then really horrible things happen.There are a few people have a lot of courage and compassion who won't do bad things to other people even if they have the chance but, for most people, the only way to prevent them from doing bad things to others is have the possibility that those bad things could also happen to themselves.Just ask most Americans if it would be too harsh to strip search Muslim/Arab men and then also ask these same Americans if it would be too harsh to strip search everyone who wants to travel by air. Most Americans will answer the two questions differently which proves that equal treatment (not allowing racial profiling) is necessary to safeguard against overly harsh treatment.

Dimitri Mitropoulos

Vlad III the Impaler, would have probably had less trouble deterring terrorism than a libertarian society would...


Here Becker says, that Discriminating certain group if group members have higher probability to be terrorists is OK. So here prof. Becker is advocating collective punishment. Just a few posts before he criticized collective punishment. It does not feel self-consistent.


To limit the discomfort and anger caused by profiling, members of the profiled groups should be treated politely and with dignity.An economic solution to this problem would be for the government to pay the members of the profiled group for their time.For a given scrutinizing procedure, the government would pay a rate at which 90% of the profiled group would be willing to undergo the procedure voluntarily. The going rate for a 15 minute interview might be $100 and the going rate for a strip search might be $1000. That way, everyone would be sharing the burden of the increased scrutiny rather than just the profiled group.There are probably better ways of determining the appropriate rate but one possiblity would be to give a small number of randomly selected members of the profiled group a choice of taking the money or the increased scrutiny. If the fraction that took the money was below 90% then the payout could be increased (and if it was above 90% the payout could be decreased).


I think that stopping a terrorist act at the doorstep, as was done with the liquid bombers, is far too late in the formation of a terrorist act to be effective. More effective would be to consider why people become terrorists in the first place, and try to stop them before they even think about destroying Western culture. We can arrest/kill as many terrorists as we can but as long as ideas persist there will be more terrorists. In addition to security measures we should promote understanding of mainstream Islam, unite with other world leaders in rejecting extremist interpretations of the Koran, point out the sections of these documents that call it a sin to commit murder, and figure out what policies we have that piss off terrorists so much. Sweden, Norway, and Switzerland don't get bombed.


Racial profiling is the most efficient strategy to counter terrorism.

It does not need to be ‘strip searching’. It could be as simple as a short questioning session, similar to the ones Americans are subject to when entering some countries and El Al uses on white Americans. A respectful, one minute interview is completely warranted and appropriate. A trained interviewer watches for body language that indicates warning signs.

If terrorists react to this tactic by approaching Jews, hipsters, and pregnant women to be co-conspirators, I’ll bet that the informant rate will jump.

Given the statistics and stakes, if someone if offended, I don’t care.

Justin Donoho

Comparing racial-profiling to slavery is nonsense and an insult to the African-American community.

Furthermore, if there were a serial killer on the loose matching my description (does it matter what my description is?), I would have no problem answering a few extra questions and surviving a few extra pats to help the police.

Richard Mason

Isocrates: Richard Mason's argument is not persuasive. Since Al Quada has an easier time recruiting young Muslim men than others--it makes sense to focus attention principally on them.

The stars of my hypothetical example, Richard Reid, Jose Padilla, and Abu Bakar Bashir, were all young Muslim men. The point is that you cannot necessarily deduce someone's religion from their appearance, especially if they wish to deceive you.

Limes: It could be as simple as a short questioning session, similar to the ones Americans are subject to when entering some countries and El Al uses on white Americans.

El Al interviews all passengers. The most frequently cited success of the El Al interview system is the interception of Anne Marie Murphy, a pregnant Irish girl who had unwittingly been given a bomb by her Jordanian fiance.

Perversely, this incident is trumpeted as a triumph of racial profiling, on the grounds that El Al interviews are "profiling" and perhaps because of the race of the fiance. But it might be more instructive as an example of how ineffective naive racial profiling is likely to be.

If Professor Becker really meant to say that an effective security system must devote increased scrutiny to young Arab men, young Pakistani men, and pregnant Irish women, then I withdraw my objection.

Alexander Ritzmann

Richard Mason is right. Profiling that leads to some sort of “fast lane“ for people who “don’t look like” Muslims would be a additional security risk. Al Qaeda has proven that they are able to adapt. Why risk the detection of a Jihadist at the security check when you can force an 70 year old law professor to carry the weapons/explosives by taking his grandchildren hostage?

I am not talking about political correctness here. Profiling is necessary and working as a tool used by intelligence agencies. But if you want to increase aircraft security, El Al safety procedures have to be the future.

J. Suggs

I have observed that those least likely to be subjected to profiling are the strongest defenders of the strategy. If we look at another serious problem, child molestation, we find that a Nobel Prize winner, Daniel Gajdusek, pleaded guilty to child molestation and spent time in prison. Only about 758 individuals have won the Nobel Prize and I estimate only about 250 of those are still alive (including Gajdusek). Thus, Nobel Prize winners are more likely to be confessed child molesters than the average citizen. Do we want child welfare using its limited resources to investigate these people? Probably not.

Past history can still play a role in security, by looking for improbable events. If the average number of young Arab men on a given flight is less than one, and suddenly five such individuals show up, then two data points, number and ethnicity, should trigger a security response. For all the reasons given by others, just one piece of general information is not so useful.


It would be safer and more ethical to interview all. But it would not be cost-effective.

The pregnant Irish woman was an unwitting accomplice to her Muslim fiancée. Two data points: (1) Muslims with ill intent recognize that they won’t be able to penetrate El Al’s security; (2) Muslims with ill intent are now targeting ‘softer’ security schemes.


I agree with professor Becker to some extent. Sometimes people have to sacrifice some liberty to avoid bigger losses.
9/11 has made terrorism much of a global fear for many countries.Even in China, a relatively "terrorist-attacks-immune" country,we are still worried about the violence activities in the western part, mainly Tibet.terrorists should be eradicated because their attacks cost so many innocent lives.But, as every country is combating terrorism,as everyone is denouncing these terrorists, people should spare some time to think about why terrorism appeared and is becoming a more and more severe issue? No one is born to be a live bomb.They are humans,they also have sympathy for their kind.If not in a desperate situation, they wouldn't have made such a desperate choice.Intense scrutiny can only decrease the possibility of terrorism occurence,but it cannot eradicate it.


Wes brought up an excellent point. There is an article on this out somewhere [I wanna say Slate.com]. Why not pay everyone who gets searched? Give $100 to everyone who is chosen. There is no point in having people volunteer for searches: no one worth searching would volunteer. With the payment, as Wes pointed out, the burden of the search would be spread over to the general population of taxpayers. As a result, it would be much more acceptable to search specific groups more intensively, since they would be compensated. But, as has been said above more than once, this is just treating the symptom. We need to figure out how to treat the disease.


While I can see where the justification for sacrifising our liberties in order to save hundreds of innocent lives comes from, we mustn't forget that thousands died so that we could have them.

You can't just take away and give back freedoms according to the circumstances of the times. It would be nice but it just doesn't work like that.


Can I ask a question?

What would happen if the foiled plot in Britain had not been made known to the public?

Why should the public know everything?

Carl Shulman

Like several commenters, I have considered monetary compensation to members of profiled groups so that the costs of profiling are broadly distributed (as with random searches) while efficiency is increased.

On the other hand, there might be significant difficulties in selecting a payout rate. An arbitrarily set rate is likely to be inefficient. Offering a choice of cash vs search, as suggested by wes, to a random sample of the profiled group offers a workable mechanism for price discovery. However, we would then face a tradeoff in determining the rate of acceptance on which to base payment (90th percentile?). If there is high variability in reservation prices among law-abiding travellers, then selecting a price that induces 90% acceptance will confer a sizable surplus to many travellers. In addition to increasing the cost of compensation directly, this will provide an incentive for more frequent flights among the profiled group, which will raise costs of both compensation and searches/interviews. (Incidentally, this would also cause difficulties for the price discovery system, as the shift in the mix of profiled travellers towards those with low reservation prices would lower the 90th percentile reservation price in a random sample.)

Alexander Ritzmann suggests that 'fast-tracking' non-profiled groups would provide a large opening for terrorist groups to employ such individuals for attacks. Such an opportunity would be created by searching *only* profiled groups, but the costs of recruitment and supply of terrorist volunteers vary substantially across groups. Involuntary recruits are much more likely to fail than voluntary ones, as those who are unaware that explosives/poisons have been placed in their belongings will not avoid scrutiny and may discover an oddity, while recruitment by blackmail is difficult when participation is likely to result in death. Increasing the level of scrutiny for low-cost recruits at the expense of reduced scrutiny for high-cost recruits should therefore increase the expected cost of launching a successful attack for terrorist groups.

David Bartlett

The chilling personal conclusion that I always seem to come to when hearing discussions about this broad topics, is that a Terrorist based philosophy of attack is so difficult to stop. I would even venture to say impossible to stop.
Lets say for instance that checks at airports are tailored (whether politely or impolitely) towards certain groups. Theoretically, what is stopping the terrorist groups from canvassing for apparent "neutrals" and paying them enough to carry out an attack for them. One cannot deny that sooner or later an individual can be bought at the correct price.
The nature of a Terrorist based approach, especially on the grand scale that is beginning to become apparent, is that it tends towards being infitely unstoppable.
This was just one small example of how terrorism may adapt. Lets not even mention the fact that many of terrorism main golas are achieved passively in between so called "attacks".

What I think is more likely to be a route to any resemblance of world peace is the accpetance of these glaring facts.....and therefore the slow and careful adaptation of approaches in policy and culture so that the very basis for a terrorist stance is eroded away.

This is of course assuming that world peace is a desired result for any involved party and that those involved are equally bent on continuing towards global intergration, as opposed to one that sees certain groups sticking to their own business.

I get the feeling that these assumptions won't come close to holding really....if one is to truly analyse the motives and ideas behind much of what is going on.

A tough one.....


"Pay to search­" is very stupid idea. I think everyone with basic economical literacy will understand this. Now real terrorists trying to act without causing suspicion and innocent people act naturally. Search compensation gives some innocent people (those with lower marginal cost of their time) motivation to attract suspicion. When government offered bounty for killed rats (or other vermin species, I do not remember exactly) people just started breeding rats for money. Familiar outcome will be observed as soon as "bounty for suspicion" will be announced.

Now I want to explain something to people saying about symptoms and diseases. I'll give you an example. Currently people do not earn according to their contribution to humanity's manufacturing might. Many work hard and get barely enough to survive. Few have billions and do not work at all. We can fight the symptoms by social insurance, charity etc. But if we try to fight the disease we attempt to kill ourselves. Injustice and violence are bad, but these are necessary conditions caused for social, economical, historical lives. Energy conversation law is not good for mankind, but we must not fight the basic law itself, we must find ways to prevent energy loses, create better engines etc.

People, I know, this subject is very touchy, but please, express your economic thoughts, not just your feelings, it's Becker's blog.

The comments to this entry are closed.

Become a Fan

May 2014

Sun Mon Tue Wed Thu Fri Sat
        1 2 3
4 5 6 7 8 9 10
11 12 13 14 15 16 17
18 19 20 21 22 23 24
25 26 27 28 29 30 31