Asymmetrical Information

Here's the really tough question at the core, though:

If almost-all of the people you're after really do fit an ethnic or religious profile (Muslim, "brownish"), should you deliberately not concentrate on those groups, to avoid the appearance of racial prejudice?

Racism, justly despised, is, after all, wrong because it ascribes to an entire group attributes not shared by that group. But if the vast majority of terrorists really are brown, muslim people, is it racism to focus on brown, muslim people, while consciously and actively aware that most brown, muslim people are thoroughly innocent?

That's the hard question, and if more attacks succeed, it will only become harder.

The problem I have with this post is that it written by someone in no danger of being profiled. It's all well and good to say "profile me, I have nothing to hide" when you know damn well that you won't be singled out anyway.

Read Schneier on Secrity blog. His points on these kinds of things are 1) unselective screening is just a waste of time, since there are so few terrorists even in the sample of "likely" suspects; 2) if you let people refuse searches and leave (as they are doing in the subway now), then the terrorist can just leave if spotted and try again some other time; 3) if there is a simple category (young men travelling alone with backpacks), the terrorists will adapt to that -- they'll carry luggage and have plane tickets peeking out of their pockets. And finally, 4) if you secure subways, they can just move to some other target. There are an unlimited number of places where the public gathers.

If you are going to screen, he advocates letting screeners use their judgement, but making sure they are trained to spot suspicious behavior in general. He sites this as the Israeli approach to security. But he thinks that agressive intelligence to find the source of these operations is the way to go, not trying to protect public infrastructure.

It's been awhile since I looked at 4th Amendment law, but random searches don't have to have any articulated suspicion or cause. There are lots of people who will scream bloody murder anyways.

Non-random searches require reasonable suspicion, and I don't know that just adding "a Muslim-looking appearance" (coupled with the fact that the main terrorists we are looking for are indeed all of Muslim appearance) reaches the point of reasonable suspicion.

Another way to constitutionally search is to have the searchee's consent, so the laws pertaining to riding the MTA could be changed to state that the rider always consents to being searched as a condition for riding. This would be legal, but would probably raise an outcry. At least in my liberal community, it would.

It's all well and good to say "profile me, I have nothing to hide" when you know damn well that you won't be singled out anyway.

I don't object to being searched. I object to being searched for no rational reason. If the police were looking for Klan terrorists, I'd be a candidate for profiling -- I'm white, male, and grew up in the South. But they're looking for Muslim terrorists, virtually none of whom are white. Arab women are more likely to be terrorists than white men are. Randomly searching me and skipping the Arab man (or woman) sitting next to me is a foolish waste of resources.

Every search has a cost, to the policeman's time and the target's dignity. If the police conduct a thousand searches, a thousand people's dignity is violated. If they use profiling, a thousand people's dignity is violated; if they don't use profiling, a thousand people's dignity is violated. The difference is that if they use profiling they're about a hundred times more likely to actually catch a terrorist.

Out of curiousity, would you object to profiling if it was doing on the basis of citizenship or national origin instead of race? You object to profiling Arabs over whites; what about profiling people born in Saudi Arabia versus people born in Vermont?

The brown skin is more important than Muslim. And here in the southwest everyone looks like a 'terrorist'. Which stokes rampant rightists anti-immigrant behavior.

I am wondering if they find something they are not looking for, but is illegal (like drugs), can they arrest you?

Philosophically, I'm against random searches. But I think we (the public) have grown used to them, so don't protest too much. The question is, do they work, or are they something we do just to make ourselves feel better?

Non-random searches require reasonable suspicion

I believe that all they need to do is stick a sign on the door saying "passengers are subject to search". You consent to any search you may be subjected to by the simple act of stepping onto the subway train (which is, let's remember, usually government property).

Dan--we do, in fact, profile people based on national origin already, just not on the subway. Some countries are visa-waiver countries; for others, we make it very difficult to get here. And there are now more stringent rules keeping track of young men (there we go) from certain countries--obvious ones like Saudi, Pakistan, Yemen, etc. So here's an interesting thing to remember: we already do some of this. Just not in the subway, mainly because it would be impossible without passing a law requiring a national ID.

Willie B - I am saying explicitly that I don't mind being searched. I'd much rather be searched myself than have a granny searched. But you (and another commenter) raise a fair objection: It's easy for me to be pro-search because I don't mind being searched for my own reasons. I will never be carrying drugs, or anything else I'd even be embarrassed about. What if your stash of pornography is pulled out of your bag in front of everyone in the subway? It's legal, but should be covered in your right to privacy.

Can' remember exactly where I read it, but some Israeli contrasted the US and Israeli approaches to screening as follows. "The Americans are searching for the weapon, whereas we are looking for the terrorist."

While young men are relatively privileged by their age and sex, if we also singled those who look Muslim, or light brown, we would be targeting groups that are already in a tight spot in America. I can't support it.

Contributor A,

I sympathize with your dilemma. Let me facilitate your recovery from left-wing "thought" with a question:

Why do you have no problem with stopping men?

Answer: because they're more likely to be malefactors.

So, since God knows Muslim men are much more likely to be terrorists than men in general, why do you have a problem with stopping Muslim men, or those likely to be such?

Answer: In extending the policy from men (a group of wretched refuse with no pressure group defending them) to Muslim men you've now crossed a liberal tripwire from a despised group (white men, many of them heterosexual) to a non-despised group (terrorists).

So you either need to abandon either your willingness to focus on men, or your reluctance to focus on Muslim men, to be intellectually consistent.

Abandon the PC rubbish - racial and ethnic profiling, like age and gender profiling, are basically sensible police work, namely, focusing on those most likely to be involved in crimes. Let's stop pretending that black teenagers and blue-haired grannies are equally likely to commit drive-bys, and that soccer moms and twenty-something Muslim men are equally likely to commit terrorism.

Sorry for the heat here, but this kind of nonsense winds me up. We're in a war, and can't afford this kind of woolly-mindedness.

I just noticed this:

What if your stash of pornography is pulled out of your bag in front of everyone in the subway? It's legal, but should be covered in your right to privacy.

I won't even dignify this with a comment if you retract it now.

Hm, Occam. I'm not even sure what you'd like me to retract, nor am I likely to retract anything under threat anyway, so have at it.

I seem to recall the folks at Hit and Run noting that you could be arrested for other non-terror violations if discovered in one of these random searches. Which is bull, as far as I'm concerned. Whole thing is bull, millions of people ride the subway every day, you'll never stop an attack, it's just the appearance of safety. A lot like seatbelts on airplanes.

So here's an interesting thing to remember: we already do some of this. Just not in the subway, mainly because it would be impossible without passing a law requiring a national ID.

I'm aware that we already do it in some contexts, but I think you're mistaken about the feasibility of doing it on the subway. A foreign accent is a good indicator, for example, since few people who did not grow up speaking English are able to shed the accent of their native language.

But a national ID card would certainly help.

"I don't know of any other recent female terrorists."

Actually, in Sri Lanka, a lot of the suicide bombers have been women. A woman blew herself up in the assassination of former Indian Prime Minister Rajiv Gandhi in 1992. The LTTE, which is the organization she was from, had large numbers of female combatants, and in fact the most likely profile of an LTTE suicide bomber (they were the people who really began to use suicide bombings in a big way) was that she was young and female.

wrong blog address in the comment above ;)

Occam's Beard:

Wow. In an effort to be civil, I'll actually set aside my gut-level personal opposition to profiling while responding to you. Because you're still, uhm, making some stretches here.

Why do you have no problem with stopping men?
Answer: because they're more likely to be malefactors.

Right. I also (most of the time) happen to know a man when I see one.

So, since God knows Muslim men are much more likely to be terrorists than men in general, why do you have a problem with stopping Muslim men, or those likely to be such?
Answer: In extending the policy from men...to Muslim men you've now crossed a liberal tripwire from a despised group...to a non-despised group (terrorists).

Thanks for keeping the "liberals love terrorists" vibe going. Real helpful, real productive. Also, again, I tend to know men when I see them (it helps to be one, maybe). Height, build, Adam's apple, facial structure, etc. Unfortunately, not all Muslims are walking around in helpful "I Love Allah" t-shirts.

So you either need to abandon either your willingness to focus on men, or your reluctance to focus on Muslim men, to be intellectually consistent.

Well, as a hypothetical proponent of profiling, I'd rather encourage a willingness to focus on behavior, as someone said earlier. But here's the thing: let's say I'm willing to focus on Muslim men. How the hell am I supposed to recognize them with any accuracy? I don't want to bog the thread down with personal history, but I'll summarize the points I'd like to illustrate.

(1) Most people's ability to sync up appearance with ethnic background is atrocious.

(2) People from the Mediterranean basin (by which I mean southern Europe, northern Africa, and the Middle East) have a broad tendency to look (a) alike, and (b) like lots of other peoples not from the Mediterranean basin.

(3) Not all Arabs are Muslim. Not all (I believe not most?) Muslims are Arab.

So an effective program in which we taught our security officers exactly what sort of physical appearances they ought to keep an eye out for would either: (a) have to be so specific to focus on those people who just look obviously Arabic and nothing else, which would be stupid, since that wouldn't even cover huge amounts of Arabs, or (b) have to be so broad that pretty much everyone who wasn't obviously from northern Europe, eastern Asia, or (maybe, if we were feeling nice) sub-Saharan Africa would be covered. That's a lot of people. An unhelpfully huge percentage of U.S. population, as a matter of fact. The net's a tad wide.

So I think "racial" profiling would be ineffective, and its ineptitude would inevitably be tied to racist abuses. Like the Portuguese-American guy who got the crap beaten out of him for being "Iraqi" last fall in San Diego, except with cops and guns. (Like the Brazilian man who was killed in London. At least there the police have the excuse that he ran.)

Contributor A,

I was referring to your equation of the embarrassment of someone having his masturbatory materials exposed in public to the slightly more weighty issue of the potential death and dismemberment of his fellow citizens.

In matters that are literally of life and death, privacy drops off the bottom of the list as a priority. If that's not manifestly clear, it's time to get the moral compass recalibrated.

Ask the families of the fifty-six dead in London if they would have minded the small intrusion on civil liberties resultant from the searching of random passengers on the underground trains and the bus where they died!

My son and his girl-friend travel the rails and roads of London every day of their working week, and they wouldn't object, so why should anyone else?

One problem I see with racial and ethnic profiling, is that you have to be able, with a fleeting glance, determine what "race" or "ethnicity" a person is. Take for example the shooting death of innocent Brazillian Charles de Menezes in the London subway. If you look at his picture, and you didn't know he was Brazillian, what would you describe him as?
http://www.abc.net.au/news/newsitems/200507/s1422086.htm
If you believe he "looks" like an Arab, is that enough to act on? And if the emphasis is looking only at "Arab-looking" people, who's going to be watching out for the next Timothy McVeigh, Ted Kaczynski or Eric Rudolph, or are we at the point where DOMESTIC terror isn't as "terrifying" as that which is imported?

--Cobra

Right. I also (most of the time) happen to know a man when I see one.

Irrelevant, and in fact, not always obvious. This argument on recognition fails, as we'll see below.

Thanks for keeping the "liberals love terrorists" vibe going. Real helpful, real productive. Also, again, I tend to know men when I see them (it helps to be one, maybe).

It's not that liberals love terrorists so much, it's just that they love to feel good about themselves, as long as the consequences land on someone else. Deep down, they don't figure that they themselves will be blown up – it's an abstraction to them - so they can strike a noble pose at no cost to themselves. And that is immoral.

To my point about white males, let's face it, there's no other group on earth that you could possibly single out for opprobrium, denigration, and discrimination than white males. None. It's open season. You know this is true.

Searching men generally is no problem unless they belong to a "victim" group, in which case, mirabile dictu, they become pure as the driven snow, and the whole matter becomes an outrage of the first water. That miracle, which rivals the transubstantiation in magnitude, results from liberal thoughtlets, and the inability/unwillingness of liberals to finally draw the line on political correctness and admit publicly that certain groups are disproportionately responsible for certain problems (in this case, Muslim men and terrorism).

I'm confronting you with your own hypocrisy; you're happy to conduct gender profiling, and age profiling, but not racial or ethnic profiling. To extend your argument, you can't always tell how old someone is, anymore than you can always tell to which ethnic group someone belongs. Perfection of recognition is not the standard, and in any case is beside the point. Appearance is one aspect of a constellation of factors that police should be able to take into account. See below for more on this.

[…] Unfortunately, not all Muslims are walking around in helpful "I Love Allah" t-shirts.

So…your problem is with recognition, i.e., the implementation of profiling, and not with the profiling itself? You'd have no problem with stopping and questioning every single person who was identifiably Muslim, say every one wearing an "I Love Allah" t-shirt? Your argument fails if you agree with this.

By your argument, we should never imprison anyone; we might make a mistake. If you can't do something with 100% accuracy, you shouldn't do it at all?

Well, as a hypothetical proponent of profiling, I'd rather encourage a willingness to focus on behavior, as someone said earlier.

Unfortunately, the behavior we're talking about is blowing up innocent people, and once that behavior has been evinced, it's kinda late to start wondering whether we should have talked to someone. If by behavior you mean furtiveness, see below.

But here's the thing: let's say I'm willing to focus on Muslim men. How the hell am I supposed to recognize them with any accuracy?

Optometry may help with this. Do you mean to tell me that you could look at the London bombing suspects and not have a teeny weeny suspicion that they might, just might, be Muslims? And you'd be happy to watch their backpacks for them while they just step out of the subway car for a moment? Please. You'd be profiling them like a champ. Admit it.

Seriously, though, you've missed the point entirely, which is this. I want the cops to have the ability to stop and question anyone whom they consider a likely candidate without having an army of activists jumping on them if most of those people are Muslims, or whatever group is most connected with the crime the police are trying to prevent.

That's it.

Will the police stop every person of that group? No, nor should they. But if someone's appearance, circumstances, demeanor, etc. make the hair stand on a police officer's neck, I want him to be able to question that person without the ACLU or some other earnest lefty group filing suit because XX % of the people stopped belonged to some group. So I don't view approval of profiling as an imperative, but rather as an enabling doctrine.

Here's the critical question for you: would you want the police to stop and question people like the London bombing suspects – obviously Middle Eastern in descent, carrying backpacks around the subway – or would you rather they focused on my wife, who's in London right now? If the former, you agree with me.

And if the emphasis is looking only at "Arab-looking" people, who's going to be watching out for the next Timothy McVeigh

Why is it that so many people seem to think that "racial profiling" equates to "white people never, ever get searched"?

Nobody is saying that white people shouldn't be searched. We're saying that if you're going to search anybody, you should prioritize your efforts based on your knowledge of what kinds of people are most likely to be terrorists.

are we at the point where DOMESTIC terror isn't as "terrifying" as that which is imported

Domestic terrorists have primarily targeted government installations (militias), abortion clinics (right-to-lifers) and construction/research sites (environmental wackos). Those are different profiles. Blowing up users of mass transportation is primarily a Muslim terrorist activity.

Hypothetical scenario: let's say you know for a fact that one person in a group of 100 is a Muslim terrorist carrying a bomb. The group of 100 contains 99 white guys from Boise and one Arab guy. You only have time to search three people. Do you pick the Arab and two random white guys, or do you pick three random guys (and, 97% of the time, end up with three white guys)? This is a worthwhile question, because statistically speaking it is almost guaranteed that the Arab is your man.

A strong case can be made that it isn't worth searching anybody at all -- you invade a lot of people's privacy and are unlikely to find any terrorists at all. But once you've decided that you're going to search, it makes the most sense to search the people most likely to not be innocent commuters.

it's just the appearance of safety. A lot like seatbelts on airplanes.

Put that one back on the shelf, Timothy. Aircraft designers do not add weight to aircraft for psychological aesthetics. Seatbelts in airplanes do serve a useful purpose in the case of severe turbulence and, in the extreme, non-fatal mishaps at takeoff and landing.

Mike:

Ask the families of the fifty-six dead in London if they would have minded the small intrusion on civil liberties resultant from the searching of random passengers on the underground trains and the bus where they died!

First, the UK in general has shown a long history of tolerating civil liberty invasions for the sake of feeling safer, but as a State-side resident who has observed US airport screening procedures turn into a waste of resources that (with rare exceptions) convince me not one whit of their efficacy, I'm not sure random bag searches are going to do much.

Second, that misses the broader issue: the shell of civil society is a very fragile thing. If elements within or capable of entering your society have determined that Blowing Stuff Up Near Groups of People is an acceptable protest option, you aren't going to be "safer" in the sense you describe barring bunker or police state. Random bag searches may rarely 'get lucky' (as any shotgun approach sometimes does) but in the meantime, how many miscreants slipped through the system elsewhere because the cops were preoccupied with verifying that a diaper bag does, in fact, contain diapers?

Occam’s Beard:

This is unfortunate; I tried to narrow the scope of your first post in my response, so that it wouldn’t turn into a sprawling mess, but now you’re inviting me to do just that. I’d rather have kept it to the points I responded to, but let’s go [Edit: it turned out way too long, so I’m going to ignore some of the things you said. If you actually are inclined, I’d email them to you, but I consider them too off-topic of what I want to discuss, let alone what the thread is about, to justify clogging up so much space]:

Searching men generally is no problem unless they belong to a "victim" group, in which case, mirabile dictu, they become pure as the driven snow, and the whole matter becomes an outrage of the first water. That miracle, which rivals the transubstantiation in magnitude, results from liberal thoughtlets, and the inability/unwillingness of liberals to finally draw the line on political correctness and admit publicly that certain groups are disproportionately responsible for certain problems (in this case, Muslim men and terrorism).

Now I’ll concede that there may be some people out there who are accurately depicted by what you have written. I don’t know any of them. Here’s my basic objection to the potential misuses of profiling: if you ask a police officer what made him suspicious of a given suspect and his response is “Well, I was told to keep an eye out for men, since men commit the vast majority of crimes,” that’s pretty dumb; not so much if his response is “Well, he was keeping his head down and moving erratically and kept clutching at his backpack nervously and made a clear attempt to steer as clear of me as possible, plus he was a man and…” It’s simply a fact that a lot of the former (swap “black” for “men”) happens under the guise of the latter in everyday police work around the country.

I'm confronting you with your own hypocrisy; you're happy to conduct gender profiling, and age profiling, but not racial or ethnic profiling.

Mirabile visu! I have been confronted with my own hypocrisy and see myself as though for the first time! Actually, I’ll concede that there might have been some inconsistency in the position I took, since I was quickly trying to formulate a position that had some of the characteristics of my own, but wasn’t. Namely, I was attempting to have a discussion in which the efficacy or legitimacy of profiling was admitted, and I merely pointed out what I thought was a particularly huge flaw in the idea given the current people we’d like to be profiling. I’m sorry if saying so in my first paragraph was insufficiently clear. I actually don’t like the idea of profiling (appearance), period, at least not in the way that people tend to mean it and it tends to play out.

Appearance is one aspect of a constellation of factors that police should be able to take into account.

No argument here (either from real me or hypothetical me).

So…your problem is with recognition, i.e., the implementation of profiling, and not with the profiling itself? You'd have no problem with stopping and questioning every single person who was identifiably Muslim, say every one wearing an "I Love Allah" t-shirt? Your argument fails if you agree with this.

Well, for the aforementioned hypothetical purposes, my problem was recognition. One of my actual problems with profiling is recognition. Where’d you get the idea that I’d have no problem with stopping every single person who was identifiably Muslim? When did anyone support stopping every single person who fit any broad category? Isn’t the idea to provide a pool from which to either (a) draw random subjects for search, as a deterrent, or (b) provide another tool among many which can be used to determine whether or not a person is worthy of suspicion? My point was that Muslims aren’t particularly recognizable as such, and I fear what happens when the directive to keep an eye out for Muslims gets misused.

Unfortunately, the behavior we're talking about is blowing up innocent people, and once that behavior has been evinced, it's kinda late to start wondering whether we should have talked to someone. If by behavior you mean furtiveness, see below.

I did (mean furtiveness).

Optometry may help with this. Do you mean to tell me that you could look at the London bombing suspects and not have a teeny weeny suspicion that they might, just might, be Muslims? And you'd be happy to watch their backpacks for them while they just step out of the subway car for a moment? Please. You'd be profiling them like a champ. Admit it.

Well, if anyone asked me to watch their backpacks for them while they stepped out of the subway car for a moment, I’d be yelling for the cops. Paranoid urban liberal, I guess. That’s just not what you do. I don’t take food from strangers on the street pursuant to my assessment of their suspiciousness; I just don’t. And here’s a link to the four men the British police IDed as responsible for the bombing:

http://news.bbc.co.uk/1/hi/uk/4678837.stm

Scrolling down the page, if I’m being honest about “what if I passed him on the street”: the first man I’d probably assume was Arab, the second man I might guess was mestizo, the third I’m not sure (the picture’s not so great, but I might go Arab, or maybe Indian). The fourth man looks like a guy from Jamaica. Just kidding; I’d assume he was African-American (unless I was in England). So we’re a very very shaky 50% on my pegging those men as “Muslim.” (And remember, Muslim doesn’t mean Arab.)

But if someone's appearance, circumstances, demeanor, etc. make the hair stand on a police officer's neck, I want him to be able to question that person without the ACLU or some other earnest lefty group filing suit because XX % of the people stopped belonged to some group. So I don't view approval of profiling as an imperative, but rather as an enabling doctrine.

This is a much more reasonably worded articulation of your position than anything in the post which I initially responded to. And it still doesn’t change my point, which is that attempts to profile based on so-called “Muslim appearance” are likely to fail (to me) because the parameters are either going to be laughably narrow or laughably broad. Appearance ought to have its part in a network of things a policeman might be looking for; the network is what ought to give justifiable cause for a search. Appearance alone is not sufficient (and I know you’re not arguing that it is), but in practice a judgment about appearance can be (and often is) used as the foundation for a network of observations that never would have happened (and never should have happened) without that first spark. And American policemen (and their British counterparts) can successfully ID a black man as a black man pretty well; I simply don’t believe they can do the same for Muslims. For reasons I raised in my first post, which you didn’t respond to. I never claimed (racial) profiling ought to be perfect, I merely claimed that in this case especially its imperfections rendered it useless.

Anony,

Arguably the terrorists have shifted to ground transport because air transport has become more difficult to target. If so, that's a success.

A more valid argument would be that increasing security in one aspect may merely shift terrorism to other arenas, without affecting the overall incidence of terrorism.

Random searches aren't intended necessarily to turn up bombs, but to raise the ante on those proposing to use them. Recall the millennium bomber; he wasn't captured because of a search, but because he got the vapors at the prospect of facing a customs agent and split. (Customs initially suspected him of smuggling drugs.)

As if to prove the point, as I write this, NBC News is reporting that one hairball in the Egyptian bombings blew himself up (prematurely) because he saw a checkpoint ahead.

So it's important not to take too mechanistic a perspective on terrorism prevention. Absolute prevention is impossible, but that doesn't vitiate the utility of trying to make it more difficult.

Q,

I appreciate your thoughtful and reasoned response, and apologize if my exasperation with the anti-profiling argument made me intemperate. Plus, I respect anybody who knows his Latin! Sorry for unduly broadening the thread (mea culpa? Couldn't resist.) Let me try to refocus it.

Namely, I was attempting to have a discussion in which the efficacy or legitimacy of profiling was admitted, and I merely pointed out what I thought was a particularly huge flaw in the idea given the current people we’d like to be profiling. I’m sorry if saying so in my first paragraph was insufficiently clear. I actually don’t like the idea of profiling (appearance), period, at least not in the way that people tend to mean it and it tends to play out.

Q, I don't like it either. Even less, however, do I like behavior that makes such profiling seem less unacceptable than it otherwise would be.

But if someone's appearance, circumstances, demeanor, etc. make the hair stand on a police officer's neck, I want him to be able to question that person without the ACLU or some other earnest lefty group filing suit because XX % of the people stopped belonged to some group. So I don't view approval of profiling as an imperative, but rather as an enabling doctrine.

This is a much more reasonably worded articulation of your position than anything in the post which I initially responded to. […] Appearance ought to have its part in a network of things a policeman might be looking for; the network is what ought to give justifiable cause for a search. Appearance alone is not sufficient (and I know you’re not arguing that it is), but in practice a judgment about appearance can be (and often is) used as the foundation for a network of observations that never would have happened (and never should have happened) without that first spark.

Fair enough. Surprisingly, I think we fundamentally agree. I don't want people targeted simply because they're thought to be Muslims. What I want is the converse: I don't want them excluded from questioning because they're thought to be Muslims, and the police don't want to screw up their ethnic statistics and wind up in court (the classic comparison of proportion of group X stopped vs. proportion of the population, a wholly specious argument in my view).

Keeping a little closer eye on potential terrorists than on blue-haired grannies, I think is reasonable (sensible, even); if they act hinky, have a word with them and judge their demeanor. If they throw off high voltage, then investigate them further. But if that investigation turns up, e.g., a bomb, I don't want the evidence excluded because a higher proportion of the same group have been the subject of police attention than their proportion in the population.

Deal?

Hmm, so we generally agree it's a complete waste of time to search anyone but men who aren't obviously under 15 or over 40. Can we also exclude men 15-40 travelling with others outside this category? The consensus seems to be yes.

We disagree that it would be wrong to further exclude lily-white xanthochroi, African-Americans, and east Asians?

I honestly don't get it.

That being said, random searches are almost totally useless, and will cost the city a billion dollars a year in additional wages and billions more in lost economic output caused by delays. It's a sham supported by law enforcement because, while it won't catch any terrorists, it will catch a few people with illegal drugs or weapons, excuse them of the burden of developing effective anti-terrorism methods, and give them extra funding. Always beware of handing the police the authority to do things they ordinary may not do when that authority will not help to accomplish the goal that motivated the grant of authority in the first place.

Let's recognize that almost nobody, especially not the NYPD, is interested in developing effective anti-terrorism tools, like undercover officers who can penetrate the communities where terrorists dwell, or networks of informants maintained through tried-and-true methods like cash payments or the threat of arrest for minor crimes. Easier just to throw money and untrained manpower at a problem, because it gets you more despotic power and a bigger budget while you can claim you're "doing something, dammit." Anyone else here trust the NYPD with another blank check?

So what happens when terrorists rent several apartments and blow up several buildings in spectacular, coordinated fashion?

Do you want to discuss the police doing random searches of apartments in the name of safety? And arresting you for anything else they happen to find that they don't like? Perhaps putting you on a list if they find magazines not to their liking?

We're almost there.

We have got to look at the incentive system here... Willing suicide bombers, as noted above, are a small needle in a large haystack of predominantly young mulsim males. And invariably, these guys end up looking like normal everyday people than guys with towels on their heads.

So what motivates these guys? Is it the 72 virgins they will be promised in heaven? OK, how about some opposite motivation... If they have families here, toss their families out. That actually may work better in the UK than the US, but it seems to me that it might provide the right counterbalance. What kind of a coward ditches his family for 72 virgins?

To me, the simple argument against profiling is this: As soon as the terrorists learn what the profile is (and they will, because of all the big mouths who love to show off what they know) they will begin recruiting people who don't fit the profile. They will be able to do it, too: find a blond, blue-eyed 25-year-old woman who either believes in their cause or who has such low self-esteem that she'll do anything for a man who shows her some attention. Or a blond, blue-eyed 45-year-old woman who has such low self-esteem. Or a blue-haired granny who believes in their cause to the point that she's willing to blow herself up (she's already lived a full life, see, and better she should go than one of those nice young men who has his whole life ahead of him).

By not using racial profiling we keep the terrorists from latching onto a group that has publicly been "exempted" from searches for their recruiting.

My 3-year-old daughter was "flagged" for a detailed search on our recent airplane journey. It probably took three times as long to "process" her as it would have an adult of any stripe - the wand freaked her out, the pat-down both tickled and scared her, getting her to assume the appropriate positions was a four-handed exercise and I only have two, both of which were occupied with the baby I was also carrying...

IOW, "random" searches are laughable. OTOH, keeping secret your security service's criteria for non-random search selections seems like a good idea to me, for Occam's reasons: while the search itself may not turn up anything, the fact of a targeted search may push a nervous potential bomber over the behavioral edge from a little jittery to observably weird. Of course the criteria would have to be reviewed by somebody with non-loose lips and good judgment, to ensure that they weren't stupid and that civil liberties were being protected to the extent possible under the circumstances. I don't think the ACLU would fit the bill, not as a comment on their intelligence or ability to keep a secret, but because their brief makes it too likely that they'd regard ANY targeted search as a breach of civil liberties.

Also, ISTM that it's not just the number of incidents of terrorism that count: it's the efficacy of the incidents, as London has clearly demonstrated. If terrorists can be forced to expend lots of resources, human and otherwise, only to cause relatively minimal damage and loss of life, that's a success too. Not a perfect one, but better than nothing.

Quarterican: "the first man I’d probably assume was Arab, the second man I might guess was mestizo, the third I’m not sure (the picture’s not so great, but I might go Arab, or maybe Indian). The fourth man looks like a guy from Jamaica. Just kidding; I’d assume he was African-American (unless I was in England). So we’re a very very shaky 50% on my pegging those men as “Muslim.” (And remember, Muslim doesn’t mean Arab.)"

#2 looks Latin American to me, too. #3 looks like he could be related to one of my Jewish friends. OK, his whole family do look pretty much like Arabs. And #4 is a black who converted to Muslim...

The worse problem with profiling by appearance is that much can be changed. That black guy couldn't appear white, but with peroxide, makeup, and clothing he probably could pass for a black grandmother.

I suspect that the real reason profiling works for the Israelis is that security is a very important job in Israel. They've got smart, observant, hard-working people doing it. In the USA, it's an alternative to working at McDonald's.

AT, I wouldn't be so quick to judge the NYPD as uninterested in effective antiterrorism. You should read the article on that in the July 25 New Yorker. The seriousness with which the department has approached antiterrorism to this point (exceeding, in some cases, that of the federal agencies) makes its ridiculous random-search policy all the more puzzling.

Quite simply, random searches are all but useless. Show-of-force efforts like the NYPD's Hercules teams are much more effective, as are intense efforts at intelligence gathering and analysis. But those don't have quite the PR impact.

Teri,

I have to disagree with you. Although there may be an occasional twisted exception, there just aren't that many blond, blue-eyed whiteys (whether grandmothers or otherwise) who want to detonate themselves for The Cause. Terrorists trying to recruit these kinds of people will find themselves with substantially fewer suicide bombers (which is a good thing).

Beyond that, terrorists who try to recruit outside their normal candidate pool (young non-white Muslim males) are much more likely to be turned over to the police by tipsters. It's hard to troll for depressed teenage Scandinavian suicide bombers without making the teenager's family, friends, and neighbors suspicious about those Muslim "darkies" who keep hanging out with Sven. And Sven himself is much more likely to turn his "friends" over to the cops once he realizes his friends want him to kill himself and a bunch of other people too.

In short, profiling would be a very good thing if it forced terrorists to turn to demographics that are both "low percentage" in terms of recruiting success and also exceptionally risky in terms of exposing the entire terrorist cell to the police.

plenty of chechen women have been suicide bombers - the recent school hostage situation in Beslan (?), for instance

hmm.. no .... that siege in russia where the russian special forces used gas and killed half the hostages - that is what I had in mind - definitely women suicide bombers involved there

Sorry for not reading each and every comment in detail, but the string has gotten rather long for intense review given my comment is rather brief.

It us my understanding that "profiling" of terrorists is in the end ineffective because the terrorists, once they know what the "profile" is will find and recruit folks that do not fit the profile, and those folks will slip through the profile-driven non-random searches.

In the natural world (aka in Physics), the fastest way to find something about which you know little is a truly random search, aka a "random walk".

Would a reasonable application of these types of techniques help avoid civil liberties issues while actually measurably reducing risk?

Guys,

I think we've (and the authorities) have kind of missed the point.

"Random" search to my mind is a misnomer, doing a true random search is pointless. The proportion of miscreants is too low to hope to catch them that way. It's like throwing rocks into a lake to catch fish.

The rationale for a "random search" seems to me to provide cover for a police officer to use his instincts without having a tete a tete with an ACLU lawyer. I believe experienced police officers (much like NCOs) can probably take a pretty good read on people pretty quickly. If such an officer observes a swarthy type carrying a backpack and acting furtively I want him to be able to question that guy without having to worry about being accused of anything untoward simply because the guy was swarthy. If you want to call that a "random search" to pacify the ACLU, so be it, but it's not, really. It's a way to allow the police to perform effective police work, namely, by giving them carte blanche to question people – of whatever race – who act suspiciously, without their needing to hire a lawyer.

Regarding the subway searches, Flex Your Rights Foundation has an excellent Citizen's Guide to Refusing New York Subway Searches detailing how to reject a potential search while entering the subway. Thought you might be interested.

To me, the simple argument against profiling is this: As soon as the terrorists learn what the profile is (and they will, because of all the big mouths who love to show off what they know) they will begin recruiting people who don't fit the profile.

Al Qaeda can't just run out, look at the job postings at PsychosForHire.com, and hire a dozen blonde-haired blue-eyes women willing to blow themselves up on subways for fun and profit.

They have to recruit people predisposed to do this sort of thing, virtually none of whom (as it turns out) are native-born Americans or foreign-born Caucasians/Chinese/Japanese/etc. Their available talent pool consists almost exclusively of people whose ethnicity is that of predominantly-Muslim nations.

Think about it. If Al Qaeda could switch from "Arab" to "Caucasian-American" that easily, why'd they use Arabs for the 9/11 attacks? They knew, after all, that middle eastern Arabs attract attention and suspicion from both passengers and the FBI. Using a bunch of white guys from Boise would have carried a lot less risk of discovery. Yet they didn't do it. Why do you think that is?

Jack,

In addition to Dan's post, you can look exactly two posts above your original comment to find a rebuttal of the theory that terrorists will simply recruit "outside the profile." Bottom line: recruiting outside the profile has a much lower probability of success and is extremely risky.

Two distinct issues here.

First is the principle of profiling, by which I mean taking (or, more accurately, not wilfully ignoring) cognizance of the relative probabilities of a certain gender/age/race/ethnic group being involved in criminal activity. I believe that the propriety of this was the question before us.

The second is the efficacy of the practice of such profiling. That doesn't really bear on the principle, and as such, is irrelevant. If profiling in some circumstance were not efficacious, there'd be no motivation to employ it, so the issue would be moot.

So for discussing whether profiling is proper or not, let's assume a circumstance in which it was efficacious, i.e., in this case, that the terrorists are Muslims and Muslims only, and that the police can recognize them when they see them. Should police take cognizance of that recognition, or not? That's the question.

This is discussed here:

Maybe it's wrong to discriminate against a people just because they occasionally blow up a plane, but I'm going to try to talk you into it. Say, for example, you're an Arab at the airport. First of all, welcome to one of my country's beautiful airports, suspicious traveller from afar. You'll notice that security searches everyone in the airport including you. That means you have to wait in line for three to four hours while they go through the luggage of all the old ladies ahead of you for no good reason other than try to not hurt you and and your people's feelings. Now imagine how nice it would be if they ignored everyone else and went straight for your luggage. You just saved four hours of standing in line, and all you have to do to get on the plane is not carry a bomb.

It doesn't seem right to say that profiling doesn't or wouldn't work. Logically, if the object is to interdict Muslim extremist bomb carriers before they strike, focusing on muslim youths, particularly males, makes more sense than searching every 5th or 9th person, regardless of who they are. It won't make interdiction significanlty more effective, but it ups the odds, one would think. If the prospect of being stopped and searched forces Islamofacsist planners to quit using dark-complected, Middle-Eastern appearing young males as dynamite mules, the policy will have worked far better than a random policy or no policy at all. The questions of fairness and civil liberties remain, regardless of profiling's utility. The questions will remain viable unitl we have more attacks in the U.S. Given enough attacks and a high enough body count, the demand to profile, among many other presently unavailable options, will be overwhelming.

Dan writes:

>>>Domestic terrorists have primarily targeted government installations (militias), abortion clinics (right-to-lifers) and construction/research sites (environmental wackos). Those are different profiles. Blowing up users of mass transportation is primarily a Muslim terrorist activity."

The suicide bombers in Iraq don't limit themselves to mass transportation.

>>>"If almost-all of the people you're after really do fit an ethnic or religious profile (Muslim, "brownish"), should you deliberately not concentrate on those groups, to avoid the appearance of racial prejudice?..."
and
>>>"Hypothetical scenario: let's say you know for a fact that one person in a group of 100 is a Muslim terrorist carrying a bomb. The group of 100 contains 99 white guys from Boise and one Arab guy. You only have time to search three people. Do you pick the Arab and two random white guys, or do you pick three random guys (and, 97% of the time, end up with three white guys)? This is a worthwhile question, because statistically speaking it is almost guaranteed that the Arab is your man."

You have to determine who exactly is "white". Again, these comments make "profiling" look as though every Muslim or Arab (there is a distinction, since the largest ethnic Muslim population in the world is Indonesian) has exaggerated, stereotypical physical features that are easily identifiable on sight. Look what happened in NYC the other day where 5 Sikh men were zip-tied and placed on their knees at gun point. The Sikhs, FYI, are the sworn enemies of radical muslims, but since they have "brownish skin" they ended up gaffled at the hands of the NYPD.

"...When a Gray Line Bus Tour supervisor reported that five men carrying stuffed backpacks had boarded a double-decker tourist bus, a police captain summoned the city's heavily armed Emergency Services Unit, Deputy Police Commissioner Paul J. Browne said. Police ordered riders to get off with their hands in the air, patted them down and searched their bags while 51st Street between Eighth and Broadway was cordoned off...
...Officers briefly handcuffed the five South Asian men who had aroused the suspicion of bus company employees. They did not have backpacks, Browne said. "After we determined that no one on the bus constituted a threat, everybody was released," he said. "The responding officers showed the appropriate caution in light of the information and the situation."
--http://www.washingtonpost.com/wp-dyn/content/article/2005/07/25/AR2005072501603.html

Again, common sense must apply here. If the standard is, "Hey, that guy looks like a Muslim...grab him", we're in for a hellacious time here in America.

I have a question for those so amped up on profiling. Would you apply this scrutiny upon "brownish" people when purchasing firearms, particularly the semi-automatic assaulty weapon variety? Wouldn't that be as sound a policy as screening buses for bombers, as we learned from snipers, school shootings, and mad postal clerks how "terrorizing" a determined person with a gun can be.

--Cobra

Occam -- First, let's be clear about this much: I don't object to profiled search targeting. It's a tricky line to walk and must be used carefully, but it shouldn't be discarded from the available tools. IMO random searching, however, fits the "fishing by throwing rocks into the lake" analogy someone cited above.

Continuing on to your post:

Arguably the terrorists have shifted to ground transport because air transport has become more difficult to target. If so, that's a success.

This is missing a lot of context. Materially, I could duplicate 9/11 using a plastic knife carefully sheathed inside an expansion bay of an otherwise-normal laptop. Ground security would never have a clue, even if my ticket were randomly targeted for a full-body strip-search. The difference is that now, as Richard Reid discovered, if the method doesn't succeed in under thirty seconds I will have every passenger on the airplane taking a swing at me. But that is because of a shift in passenger attitudes, not security measures, and doesn't really speak to the topic at hand.

A more valid argument would be that increasing security in one aspect may merely shift terrorism to other arenas, without affecting the overall incidence of terrorism.

Then make it, if you so wish, but again I see that as merely tangential.

Random searches aren't intended necessarily to turn up bombs, but to raise the ante on those proposing to use them.

The irony went right past you: if a method isn't intended to necessarily turn up bombs, it probably won't. To truly 'raise the ante,' the method has to have sufficient probability of success in order to discourage the behavior in question. We haven't even eliminated traffic speeding using means more stringent than pure randomness; the unabomber avoided arrest for years using pretty low-tech gear and surveiliance, and even then was only accosted on a tip from a relation IIRC. Against real-world evidence like that, random searching is just spittle on Mrs. O'Leary's burning barn.

Your next two paragraphs involve checkpoint-style scenarios that are neither comparable nor analogous to random searching. Unless you completely fail to see why, there's no need to waste time parsing them, so moving on:

So it's important not to take too mechanistic a perspective on terrorism prevention. Absolute prevention is impossible, but that doesn't vitiate the utility of trying to make it more difficult.

The question is not absolute prevention, which is impossible (barring, as I stated in another post, bunker or police state). The question is whether the proposed means of making it more difficult has any real utility to an end. In the case of genuinely random searches in a venue like e.g. a subway, I say they do not, and that the resources wasted on searching would be better spent elsewhere.

Wow, Cobra, thirty minutes into evening prime time and nobody is taking the bait? You might be getting a reputation. Have you considered changing your screen name and stylistic signature?

The suicide bombers in Iraq don't limit themselves to mass transportation

We're discussing searching passengers on mass transportation. Exactly what sorts of terrorists attack which other targets isn't relevant to that discussion. The point is that Al Qaeda targets mass transportation and militias/pro-lifers/environmentalists don't.

You have to determine who exactly is "white".

No, you don't, because the goal here is "make sure you search the Arabs and Pakistanis", not "make sure you avoid searching any whites, Sikhs or hispanics".

Again, common sense must apply here.

Precisely. America isn't fighting a bunch of white suicide bombers, we're fighting a bunch of Arab suicide bombers. Ergo common sense tells us not to waste time searching the ten white guys while the ten Arab guys board the subway car.

If the standard is, "Hey, that guy looks like a Muslim...grab him", we're in for a hellacious time here in America.

Nobody is suggesting we "grab" everyone who "looks like a Muslim". This is about the best tactics for conducting a *search* of passengers. Either don't do the search at all or do it intelligently; blindly searching one out of every hundred Swedes along with one out of every hundred Saudis is just dumb

Dan writes:

>>>No, you don't, because the goal here is "make sure you search the Arabs and Pakistanis", not "make sure you avoid searching any whites, Sikhs or hispanics"."

That might be a little more difficult in the future with folks like THESE running around:

>>>An unholy alliance:
Aryan Nation leader reaches out to al Qaeda"

"You say they're terrorists, I say they're freedom fighters. And I want to instill the same jihadic feeling in our peoples' heart, in the Aryan race, that they have for their father, who they call Allah...
"...(August)Kreis wants to make common cause with al Qaeda because, he says, they share the same enemies: Jews and the American government.
...The terms they use may be different: White supremacists call them ZOG, the Zionist Occupation Government, while al Qaeda calls them the Jews and Crusaders."

http://64.233.161.104/search?q=cache:rYgC1XHYMzMJ:www.cnn.com/2005/US/03/29/schuster.column/++Al+Qaeda,+Aryan+Nations&hl=en&ie=UTF-8

Kreis' picture from the article would suggest to me, sans beret, that this joker wouldn't get a second glance from profiling types here dead-set on nabbing a "swarthy" middle-easterner. I would also question what would prompt people to look at John Walker Lindh or Jose Padilla with any more interest than the average taxi cab driver or hot dog vendor in NYC.

Profiling simply isn't that easy.

Anony-mouse, I'm not baiting with my gun sale statement. I'm just curious what the logic is for intensely screening "brownish" people when riding subways and buses, but not when purchasing firearms. There seems to be a disconnect here.

--Cobra

anony,

Your next two paragraphs involve checkpoint-style scenarios that are neither comparable nor analogous to random searching. Unless you completely fail to see why, there's no need to waste time parsing them...

In what other context than a checkpoint does the issue of randon searching even arise?

BTW, the "fishing by throwing rocks in a lake" simile was mine. Truly random searching is moronic.

First, let's be clear about this much: I don't object to profiled search targeting. It's a tricky line to walk and must be used carefully, but it shouldn't be discarded from the available tools.

anony, then we agree. It is a tricky line, and I'm not happy about using it, but I agree in the circumstances it's a tool we can't abjure.

When purchasing firearms, you've got to present ID and get a "Brady check" against a government database. Although far from perfect, that does generally fulfill the goal of the system, which is only to prevent known criminals from walking into a store and buying a gun.

You don't get anything like this precision by looking at skin color. (Unfortunately, you don't get anything like the accuracy of the Brady check from the TSA's no-fly list, either.)

Al Qaeda can't just run out, look at the job postings at PsychosForHire.com, and hire a dozen blonde-haired blue-eyes women willing to blow themselves up on subways for fun and profit.

They have to recruit people predisposed to do this sort of thing, virtually none of whom (as it turns out) are native-born Americans or foreign-born Caucasians/Chinese/Japanese/etc. Their available talent pool consists almost exclusively of people whose ethnicity is that of predominantly-Muslim nations.

The problem here is that physical appearance doesn't provide that good of an indicator of who is Muslim. The largest Muslim country in the word is Idonesia, and the people there don't look anything like the 9/11 hijackers. There are plenty of Chinese and Filipino muslims, too and some of them Abu Sayed, for example, are inclinded toward terrorist activities. There are also here in America, a large number of African American muslems. None of these people would be stopped by profiling men who look "Middle Eastern" or what a white person's view of what a Muslem should look like. Furthemore, to there are quite a lot of phsical similarities between Indians and Pakistanis, yet Pakistanis are very likely to be Muslem, while Indians are much less likely to be so, although there are Indian muslems. To many Americans, Sikhs, who are not muslems, appear to be muslem, too.
Racial profiling certainly can be useful, but it woudn't have caught The Shoe Bomber, Jose Badilla, John Walker Lindh or any other domestic terrorist in the United States.

Occam's Beard:

What I was trying to say (two days ago now) was that whether or not you thought profiling was that the principle, were it in fact legitimate, is liable to be infected by the perceived efficacy of the practice; i.e., some dude's idea of what a Muslim looks like is probably going to target a lot of non-Muslims and miss lots of Muslims, even lots of Arabs. This was the point of my "either the net is too wide or too narrow" idea; if you restrict yourself to men who look like the first London bomber from that link I posted, you're missing a huge percentage of the people you'd ideally like to be profiling. If you open it up more broadly (say, to encompass the general appearance of all four London bombers) you're dealing with what is, practically, too huge a population to effectively be profiling (in my opinion; someone who's done security work might be able to correct me here. Would you rather be able to focus on a few select groups or rule out a few select groups [i.e., everyone who looks Norwegian or Chinese {not to say that there aren't Chinese Muslims or, presumably, Norwegian Muslims}].)

This also goes to what Dan was saying; believe it or not, people in my experience are *remarkably* bad at doing the sort of visual discriminating you're talking about. They see, to a large extent, what other prejudices tell them they want to see. E.g., I and my father (who look about the same) have been mistaken for all of the following (primarily by members of the relevant groups):

Indian (North American, I presume), Indian (northern, I presume), light-skinned African American, Hispanic (by which different people have meant different things, some of which apply), Iranian, Arabic, Ashkenazi, North African, not to mention Greek, Italian, French, Spanish (small bingo), and Portuguese (major bingo).

If we're not very good at this sort of thing, then doesn't it seem that attempting to engage in decision making based upon it doesn't have a great outlook? And, as I've said, going the other direction just captures too small a group. To make an analogy: there are people whose appearance makes me highly certain that they're of Japanese origin. I'm also aware that huge numbers of, probably most, Japanese don't carry the facial stereotypes that allow me to do that. Increased familiarity, of course, breeds subtler perceptions, but...

In what other context than a checkpoint does the issue of randon searching even arise?

In the one we are presently discussing. AFAICT from Contributor A's post, unless NYPD are planning to set up airport-style security checkpoints at all subway station entrances, the proposed random searching will be exactly that: random. A pair of cops stands near the entrance or walks down the platform, points randomly, and says "May I have a look inside your bookbag please?"

One would hope they DO use a bit of common sense in selecting search targets, just as cops on the beat normally do when deciding whether or not to stop and investigate a person or group of persons, but that's easy enough to get around. Suppose you look like the very embodiment of the Muslim Arab stereotype -- how likely are you to get pulled for searching while wearing an Armani suit and clasping a Gucci briefcase?

Terrorists are overwhelmingly young males (with a few young females added to the mix in Israel and the Palestinian territories, which just goes to show you how grim that situation is. I don't know of any other recent female terrorists.)

I don't know if you count 1991 as "recent," but Riajiv Ghandi was assassinated by a female Tamil suicide bomber.

(with a few young females added to the mix in Israel and the Palestinian territories, which just goes to show you how grim that situation is. I don't know of any other recent female terrorists.)

Chechnya as well.

Dan and DRB, I am sure that the families of the people at Jonestown, Waco and the Raielians wish that you were correct.

"Please don't screen grandma over there, with her grandbaby. The gaggle of fourteen-year-old girls? Leave them to gossip about who's doing it at their school. Rabbi Weinstein?"

Kewl, so if I'm a young Muslim male terrorist, you want to make it utterly easy for me to bomb a subway car by slipping my bomb into a package carried by the eighty-year-old Jewish woman, or the 22-year--old Swedish beauty, or one of those gaggle of 14-year-old girls. Excellent! Thanks for clearing the path!

Comments are Closed.