What is wrong with racial/religious profiling?

[daishi5]

If there are only 20 potential terrorists who want to attack our air travel system, then attempting to catch them at all with security at the airport is a moronic waste of time and money. We'd be far better off using those resources to actually track down and capture those 20 guys instead.

You suggest the flawed assumption in the paper is that the terrorist organization has a "very large" pool of potential attackers, yet your counter-example is at the opposite end of the spectrum. A more reasonable middle ground would seem to be a larger, but limited, pool of attackers than the organization wishes to use in an actual attack. For a relatively sophisticated operation like 9/11, that would seem to be a more realistic situation.

Your assumption of 90% accuracy of the profile seems pretty high to me, but unless it's 100%, the flaw remains. Terrorists can test to figure out which 10% of their applicant pool doesn't match the profile, even if the profile itself is secret. That 10% is the number of people they'd use for the next 9/11. Simple.

The counter-argument you make, that catching 9 out of 10 9/11 operations is better than catching 1 or 2 of them sounds good, but you're assuming that the limiting factor in terrorist operations is willing terrorists. I doubt the terrorists conducting the 9/11 attack were all Al-Qaeda had to offer. Instead, it seems like money, planning time, and the ability to keep a secret when too many people know it seems like it probably played a bigger role as to why 9/11 didn't happen with a few dozen planes instead of a handful. The other thing to consider is that terrorism isn't a military operation, "success" in a terrorist's terms isn't about actual damage inflicted, but psychological damage. And for that, succeeding at all is usually enough. I'm not sure 9/11 would have had THAT much of a bigger effect if the number of hijacked planes had doubled, but it would have been much more difficult for Al-Qaeda to pull off.

I used 20 only because they used the number 20. I made the assumption that the 20 are the only willing participants because I assume that terrorist organizations are going to use their full potential, if they have been holding back and have a lot more resources they aren't using then that assumption is meaningless. However the number 20 is completely arbitrary, lets just call it X, and let X stand for the complete and total number of willing bombers. I made the 90% assumption because that is the assumption they gave, not my own guess.

But, the point is, in game theory the penny game they reference is a game of equal strategy. For every strategy, the opponent has a strategy that can defeat it, and that strategy is no more expensive. Neither of these assumptions fit the argument I am making. Middle eastern terrorist groups are highly unlikely to have a supply of elderly asian females willing to be suicide bombers. If we don't scan elderly asian females, yes Al-Qaeda can get an elderly asian female past us, but only if they actually have one of these people willing to do it. Now, they can attempt to recruit one, but this brings us back to the limitations you described. Money and secrecy, those recruitments go well outside their normal operations, so this will cost more, it will be less secret, and they will have very few willing recruits, and I bet few of them will be actually capable of success. (Most would make dumb mistakes, suicide bomber and intelligence are not two traits I would expect to have high correlation.)

Right now they have a large pool of middle easter and african males which are their normal operatives, if we scan those they become useless. They need a new pool of recruits, but that pool is more expensive. It won't stop the attacks, but it makes it harder and reduces the number.

And again, this is academic, I believe it would work, but it would be a deep betrayal of American values in my eyes and I don't support it.

 

[fskimospy]

For arguments sake, let me revise my original statement and narrow it down to costs involving convenience, soldiers' lives, or money.

I do not agree with you that statistical and adaptive profiling in airport security lanes infringes on our Constitutional rights.

Implementation of screening at national entry points and security screening lanes does not equate to internment camps and fascist cops on every corner telling the dark-skinned folk to present their papers. I think that you're presenting an exaggerated list of nonsense in order to elicit an emotional response from the readers here, with hopes of garnering emotion-based support, and not to prove any real point.

You're better than that Rain...

So no amount of inconvenience should be spared if it saves a single life? We should have all of America spend billions of man hours each year sitting in security checkpoints? We should spend limitless piles of cash on security if it saves just one life? Again, I don't think you've thought this through.

All security is a cost/benefit analysis, and there's always a point at which additional security does not yield enough saved lives to be worth it.

 

[Rainsford]

For arguments sake, let me revise my original statement and narrow it down to costs involving convenience, soldiers' lives, or money.

I do not agree with you that statistical and adaptive profiling in airport security lanes infringes on our Constitutional rights.

Implementation of screening at national entry points and security screening lanes does not equate to internment camps and fascist cops on every corner telling the dark-skinned folk to present their papers. I think that you're presenting an exaggerated list of nonsense in order to elicit an emotional response from the readers here, with hopes of garnering emotion-based support, and not to prove any real point.

You're better than that Rain...

Of course I am

And I know you DON'T support the measures I was suggesting. I was indeed intentionally exaggerating my list to make a point, but not the one you're thinking. It wasn't some ham-fisted ploy to garner an emotional response from the folks here, what I was REALLY trying to do is illustrate that there MUST be limits to how far we're willing to go to fight terrorism. And while I suspect most people on both sides agree on that, the rhetoric doesn't often support it. Too often arguments against anti-terrorism measures are dismissed with some variation of the phrase "what, you don't want to stop terrorist attacks?"...like you just did with my comment.

Fighting terrorism must be treated as a cost/benefit analysis, just like anything else. And the truth is that terrorism is incredibly rare. Should we actively try to stop it? Of course. But the argument that something might prevent a terrorist attack doesn't really justify very much, since terrorism honestly isn't that big a threat to begin with. In my opinion, instituting a policy that separates honest Americans into "good guys" and "bad guys" is going to negatively affect us much more than the occasional ineffective terrorist.

 

[woolfe9999]

Of course I am

And I know you DON'T support the measures I was suggesting. I was indeed intentionally exaggerating my list to make a point, but not the one you're thinking. It wasn't some ham-fisted ploy to garner an emotional response from the folks here, what I was REALLY trying to do is illustrate that there MUST be limits to how far we're willing to go to fight terrorism. And while I suspect most people on both sides agree on that, the rhetoric doesn't often support it. Too often arguments against anti-terrorism measures are dismissed with some variation of the phrase "what, you don't want to stop terrorist attacks?"...like you just did with my comment.

Fighting terrorism must be treated as a cost/benefit analysis, just like anything else. And the truth is that terrorism is incredibly rare. Should we actively try to stop it? Of course. But the argument that something might prevent a terrorist attack doesn't really justify very much, since terrorism honestly isn't that big a threat to begin with. In my opinion, instituting a policy that separates honest Americans into "good guys" and "bad guys" is going to negatively affect us much more than the occasional ineffective terrorist.

Obviously I agree with your key point here, that fighting terrorism is a cost-benefit equation just like everything else. I would add the following to the mix - the damage done by a successful terrorist attack is higher than the number of deaths involved. Terrorist attacks cause widespread fear, even panic, can collapse economic markets, and even damage U.S. prestige. Deaths caused by care accidents and cancer don't do this. This is not to suggest that it isn't still a cost-benefit, because it is. However, might I suggest that it alters the equation a bit, such that we might actually be willing to endure a little more dollar cost and inconvenience per death prevented than we would in preventing deaths from more common causes. There has to be a limit, of course, but perhaps the equation should reflect that in this case there is damage beyond just the actual death toll.

- wolf

 

[Rainsford]

I used 20 only because they used the number 20. I made the assumption that the 20 are the only willing participants because I assume that terrorist organizations are going to use their full potential, if they have been holding back and have a lot more resources they aren't using then that assumption is meaningless. However the number 20 is completely arbitrary, lets just call it X, and let X stand for the complete and total number of willing bombers. I made the 90% assumption because that is the assumption they gave, not my own guess.

But, the point is, in game theory the penny game they reference is a game of equal strategy. For every strategy, the opponent has a strategy that can defeat it, and that strategy is no more expensive. Neither of these assumptions fit the argument I am making. Middle eastern terrorist groups are highly unlikely to have a supply of elderly asian females willing to be suicide bombers. If we don't scan elderly asian females, yes Al-Qaeda can get an elderly asian female past us, but only if they actually have one of these people willing to do it. Now, they can attempt to recruit one, but this brings us back to the limitations you described. Money and secrecy, those recruitments go well outside their normal operations, so this will cost more, it will be less secret, and they will have very few willing recruits, and I bet few of them will be actually capable of success. (Most would make dumb mistakes, suicide bomber and intelligence are not two traits I would expect to have high correlation.)

Right now they have a large pool of middle easter and african males which are their normal operatives, if we scan those they become useless. They need a new pool of recruits, but that pool is more expensive. It won't stop the attacks, but it makes it harder and reduces the number.

And again, this is academic, I believe it would work, but it would be a deep betrayal of American values in my eyes and I don't support it.

No, I know what you're saying...I think this is a fascinating discussion topic from a technical perspective, whatever the ethical implications are.

I guess my point is that the ratio of terrorist attacks compared to the overall size of the recruiting pool is already pretty low...which tells me that the limiting factor in frequency and size of attacks might not be the number of potential terrorists out there. If not, then limiting the size of the "good" recruit pool isn't going to really do much to limit the number of operations they can carry out. From their perspective, it doesn't matter if they have 100 or 10 guys who can do the job if they only need 5.

 

[daishi5]

No, I know what you're saying...I think this is a fascinating discussion topic from a technical perspective, whatever the ethical implications are.

I guess my point is that the ratio of terrorist attacks compared to the overall size of the recruiting pool is already pretty low...which tells me that the limiting factor in frequency and size of attacks might not be the number of potential terrorists out there. If not, then limiting the size of the "good" recruit pool isn't going to really do much to limit the number of operations they can carry out. From their perspective, it doesn't matter if they have 100 or 10 guys who can do the job if they only need 5.

Well, my argument is highly dependent on the axiom that I disputed in the listed paper. That the terrorists have a diverse pool of willing bombers. I am just guessing that their pool of willing bombers is not that big. Airplane terrorists are a completely different deal than normal terrorists. As I understand it, with car bombs they often can detonate it remotely, or have a fall back if the man or woman does not go through with it. Airplane terrorists are different, these people can back out at anytime, all they have to do is not set the bomb off, or not get up and pull out their knife. They also need to gain access to our country, so they have to either avoid the "lists" or the lists must fail (which terrorist groups can't affect.)

I am guessing that a very large majority of their volunteers are males of middle eastern descent. Only a small minority of their volunteers are actually going to be capable of the attacks. Although, they seem to be using people who aren't capable. However, I think this supports my belief that they don't have many volunteers because if they had a decent pool, they wouldn't risk failure on people who can't detonate the bomb. In my mind, profiling won't catch anyone because they will know we are doing it, but it will invalidate a large pool of their willing and able recruits.

And again, if I am wrong about the diversity and depth of their recruit pool, and that would mean the studies assumptions are valid, then the study is right, profiling will be easily turned against us.