The conversation between Burt and Scott about profiling at our airports brings up some mixed feelings on my part.
I think that critics of profiling understate its potential effectiveness. I mean sure, if you target Arabs and only Arabs you miss the next Richard Reid or Johnny Walker Lindh. If you target only people that are obviously Muslim, they will simply shave and dress in business suits (to me, the Muslim wearing the Muslim garb is probably the safest one to fly with).
That being said, by making would-be Muslim terrorists afraid to fly, you make the terrorist’s jobs harder. If you expand the net to target the Richard Reeds, the Nigerians, and mostly the Middle Easterners, you force them to recruit Lindhs. They don’t have an endless supply of would-be bombers. That makes their job difficult.
But the question is… how far do we push? Is the right’s objection to the Freedom Grope merely that WASPs like me must endure it? If not, what are we going to do? How much should we be willing to do to them? We can’t say it’s not much of an imposition. We’re crying bloody murder, after all, when it comes to us. So I have difficulty saying that they shouldn’t object.
And, to what end? The 9/11 attacks were a travesty, to be sure. But it wasn’t the deaths on the plane that were so disastrous. If Richard Reid had blown up a plane, it would have been bad, but it wouldn’t have been another 9/11. To re-create their success, they would need to successfully hijack the plane. There’s little reason to believe that they could. Pilots will let them slit the throats of the flight attendants before opening the doors, which are now steel re-enforced. Passengers won’t take the pre-9/11 advice* to sit quietly and let the experts take care of it.
Truly, there are worse things they can do than blow up a plane. We are really in trouble when they start actually getting creative. The casualty count of blowing up the New Orleans levy would put Katrina to shame because there would have been absolutely no preparation. And there are numerous other targets. They could do a lot more damage by blowing up a bus here and a shopping mall there. Columbine with a tan, times 1000. We don’t want them to start thinking creatively. Cat and mouse at the airport is really small-fry.
So let’s say we make our airports completely safe and they start getting smarter. Do we start profiling Muslims everywhere? Do we start profiling people with African names, or tan skin, everywhere? This is an extraordinary undertaking. Cat and mouse at the airport, even if we’re not playing with the full court press, is far preferable.
This all assumes that Muslims are the terrorists that we primarily have to worry about. On this, I agree with Scott. But that wasn’t the case in 1996. What should we have done then? Said “Oh, well the Murrah building isn’t a plane, so that’s okay”? Considering the casualty count was comparable of what was on a plane, that’s a non-starter. And as someone that objects to the DHS’s psychological profiling of internal right-wing elements, it makes it pretty hard to say that we should come down hard on another group (since I can’t be confused with being a member of said group).
The attempted terrorist attacks since 9/11 at the US have been weak, weak, weak. I’m not going to say that Bush and Cheney (and by extension Obama) have been vindicated, but either we’re succeeding through methods that don’t involve overt profiling**, or the threat isn’t what we thought it was and the 9/11 attacks were an anomaly.
* – I recently worked through Tom Clancy’s novels. In a pre-9/11 book, he states and re-states the conventional wisdom that only an idiot (or a trained badass like John Clark) confronts a terrorist on a plane. Sit down, be quiet, and let the experts do their thing. Everybody knew this.
** – Profiling or no, it’s worth pointing out that next to me (maybe more than me), my Palestinian-American sister-in-law gets the most attention at airports. With the exception of her skin color (to the extent we say that’s an exception), she’s more All-American than I am. She couldn’t change her name fast enough when she got married.
it’s really really hard to conceive of the fans of profiling still being all cool with it if they were the ones being profiled. but as you noted, the perps of the second largest terrorist attack on the us were of a certain skin color , ethnicity and political belief which sounds like enough to create a profile.
The DHS somewhat recently set up a profile for the domestic rightwing terrorist, casting a rather broad description with a lot of intangibles. And the right screamed bloody murder over it. Rightly so, in my opinion. And I would extend the same protection to Muslims.
Profiling, as a security countermeasure, has very limited domain success. In very similar ways, the limitations of it are similar to the limitations of facial recognition software.
Let me talk for a sec about facial recognition software, because it illustrates the point in security countermeasures inside a domain without involving sticky concepts like legal rights.
If I have a set, small population of known good actors, and a target population that is almost entirely that population of known good actors, facial recognition software works pretty well. 200 people on a high security military base daily, with maybe a grand total of 10 visitors a month and thus the software “sees” 6010 instances of verification, of which 6000 need to be mapped to 200 entries in a database and the 10 flagged for additional check. That’s not too hard. I get 12 results, that means I’m checking the 10 guys plus two false positives. Maybe over the course of 1 year I get a false negative (1 failure out of 72,000 or so, that’s pretty damn good), but mis-recognizing a non-base member as a base member doesn’t get the non-base member past the rest of security. Results: good.
Now, do the same thing in an airport. I don’t have a database of “known good” actors. I have the opposite thing: a database of “known bad” actors (which is faulty, but that’s a whole ‘nuther conversation). I have maybe 1 million instances of verification every month (handwave), and I’m comparing them to a tiny fraction of samples. That’s really freaking hard. Moreover, since my sample size is so large, even if I have the same 1 out of 72,000 failure rate I wind up mis-identifying 14 people as terrorists. Shit, I have to evacuate a terminal every time that happens. Results: yeah, uh, not so good.
Profiling works in Israel’s airports. It works there because of a hundred factors, some of which aren’t applicable to the U.S. because of legal reasons, and most of the remaining of which aren’t applicable to the U.S. because of non-legal but practical reasons having to do both with the general homogeneity of the harmless actor population, the general homogeneity of the malefactor population, and the sample sizes and environment in which those sample sizes aggregate.
Israel has two major airports, a fairly homogeneous general population who has all received military training and most importantly relatively small numbers of man-trips via airline travel. In short, it takes relatively few screeners to screen the population of travelers; ergo you can train those relatively few screeners very, very well.
The U.S. has literally hundreds of airports, a wildly non-homogeneous population almost none of whom have military training, and an astronomical number of man-trips per year via airline travel. In short, it takes a humongous number of screeners to even provide cursory screening of the population of travelers; ergo you cannot train the population of screeners well at all.
The malefactors in Israel are all going to exhibit a very high number of common characteristics. Indeed, the psych profile for those malefactors is also pretty common. The malefactors in the U.S. don’t have anything resembling a very high number of common characteristics, and the psych profile is also varied.
In short, a screener in Israel is likely going to be picking out a metallic tennis ball from a bucket of tomatoes when they have a metal detector and someone giving them very high levels of compensation.
A screener in the U.S. is likely going to be looking for… something… in a haystack… it might be a needle, it might be a horsehair, it might be a kitten, but all we know is that it *might* be in there… when they’re wearing metal gloves and they have someone beating them with a dirty stick.
Never mind the fact that it’s just plain illegal. It’s not practical.
I think the question is… practical as compared to what? I would expect it to be less impractical then what we’re doing now, particularly if an augmentation (if we started giving everyone who isn’t an apparent Muslim (Arab, Persian, or Muslim-garbed) a pass, it would be fully counterproductive since Lindh, Reid, and the Underwear bomber could get by (as could the 9/11 bombers with fake ID with Hispanic names). And not likely practical enough to jump the legal hurdles. But, if we view any greater frustrating of potential terrorists as worth great cost, that might be different.
At the root of it, though, not much of what we’re doing is practical except perhaps in the theatrical “making you feel safe” sense. But even there, I think the conspicuous security measures serve as much to remind us of the very negligible threat than to make us feel secure.
The number of airports is an important factor, as I mentioned in a previous thread. Israel has a limited number of airports. In the US, we’d have to guard St. George as closely as LAX because a potential terrorist need only get through the airport with the weakest security and from there can transport to any other airport.
(Along those lines, I wonder if we will not in the future have to go through security all over again for layovers. Doubtful, but not impossible.)
> But, if we view any greater frustrating of potential
> terrorists as worth great cost, that might be
> different.
Full disclosure: Yeah, I don’t.
> At the root of it, though, not much of what we’re
> doing is practical except perhaps in the theatrical
> “making you feel safe” sense.
Yes.
We can abolish the 9 billion/year the TSA is asking for, drop airport security back to pre-9/11 levels, and we’re not going to impact practical security much. Certainly the change is not worth the $9 billion in cheddar. We’d do better just taking that money and air-dropping it over various low-economy countries to reduce economic despair.
Lock the cockpit door (already standard). Don’t let the hijackers in the cabin (now standard).
Nothing else really matters. Confiscating liquids and pocketknives is an absurdity.
Lindh isn’t a terrorist.
No, but he’s an example cited for why profiling wouldn’t work. A white boy who allied himself with the enemy.
In 1986 a pregnant Irish woman named Anne Marie Murphy was planning to take an El Al flight from London to Tel Aviv to meet the parents of her fiance, a Palestinian.
Unbeknownst to her, her fiance had planted explosives in her luggage.
Israeli screeners caught the bomb not because of racial profiling or even behavioral profiling (which she passed, being unaware that the bomb was present), but with baggage screening.
http://www.shabak.gov.il/english/history/affairs/pages/anne-mariemurphycase.aspx
In short: this has been tried already. You can’t depend upon screening the passenger, because the passenger might not know they are a carrier.
This would have worked if they had screened the Irish.
Well, okay. Granted, one should always screen the Irish. They’re undesirable everywhere.
Metal detectors wouldn’t necessarily stop someone from taking a bomb on board. Does that mean that metal detectors are useless? This is a good argument if someone says “All we need to do is profile”, but not an argument as to whether profiling is itself a good idea or a bad one, in conjunction with looking at other things. Most people who favor profiling (keep in mind: I don’t) aren’t suggesting that profiling is all we do.
Metal detectors are largely useless, yes.
If you want to get a stabbing, bludgeoning, or other melee weapon on board you can get or make one that will do the job that will pass through a metal detector just fine.
The same holds true for guns, actually. The actual x-ray screening can be useful, but you need a decent tech at the screen to get the job done and that’s very difficult.
Largely, it’s all deterrent at the airport and not much in the way of actual security. Now, of course there’s some value in deterrence, but not much. Certainly not enough to make several layers of security whose only real practical effect is deterrence to be “more deterrent”.
Put in gaming terms, deterrence doesn’t stack well.
No, but that was his aspiration.
Was it? He chose to join the Taliban, not Al Qaeda.
Someone who has knowledge that they’re about to blow themselves up is all but bound to be really nervous for no apparent reason. Granted that there are people who are afraid of flying or otherwise stressed out, but strange behavior seems to me to be the real clue of someone to watch more closely.
This is a better predictor than anything else, yes. By a couple orders of magnitude.
But again, you need good screeners to get this functionality. And, if you’re the U.S., you need a ton of them. Certainly more than are currently available.
The Russians have notoriously failed as profiling. In their ongoing conflict with Chechen rebels and terrorists, they have repeatedly been attacked by female suicide bombers because they still only profile for males. “Black Widows” have caused hundreds of deaths because the Russians were actively looking away. Yea, profiling works… if you want to tell your enemy exactly what not to look like to evade suspicion.
Yea, profiling works… if you want to tell your enemy exactly what not to look like to evade suspicion.
Which is only an issue if (a) they can easily recruit people outside of the profile and (b) you look only at the groups that you’re targeting.
Profiling is not mutually exclusive with everything else we’re doing. Now you can take the point of view that Patrick seems to*, where it’s all more-or-less pointless. But from a logical point of view, metal detectors plus baggage screening plus random searches plus profiling is likely to be more effective than the first three without the fourth. The notion that there are things the fourth will not catch doesn’t change that.
If you force the terrorist organizations to find American citizens, or stuff their bombs in the stuff of unsuspecting Americans, you’ve made their job more difficult. I simply don’t believe it’s true that “they will find a way.” If they could just find a way when they want to, there would have been more attacks.
* – Lest I come across as too dismissive, my views are not actually that far from his. I still think you do luggage scanning and metal detectors (though I could be convinced otherwise on the latter), but I’m not sure you get into the really intrusive stuff because, while it may actually be more secure than not doing it, the gains are marginal enough that it’s not worth it.
That being said, it does seem that a fair number of people on the right are wanting profiling in return for themselves being left alone. That is more problematic. It’s not clear to me how effective that would be. It would depend on Muslim terrorist cells’ ability to recruit Americans.
However, anyone who is getting irate over what we’re going through right now, simply isn’t in a position to say we should apply such pressure to foreigners, certain minorities, and members of a particular religious organization. Even if it might be effective from a tactical point of view, it’s not worth the other costs. Including foreigners not wanting to come and visit here.
> But from a logical point of view, metal detectors
> plus baggage screening plus random searches
> plus profiling is likely to be more effective than
> the first three without the fourth.
Well, this is where the rubber meets the road in security. Logically, layering security is a great idea (it’s fundamental principle for exactly that reason: defense in depth!)
There are two things one must keep in mind when layering security though: the layers have to actually mean something in a real world scenario, and you gotta pay for all those layers.
Let’s assume we want to protect something. We introduce security measure A. There is a countermeasure that bad guys can use, A’, that gets them past A. So we institute measure B. Measure B does several things: it increases your costs, it catches some subset of the things that A’ enabled the bad guys to do, and it provides incentive for the bad guys to think up D, which ignores both A&B because they essentially test the same thing different ways… or worse, they come up with X, which has nothing to do with either A or B.
If you make “getting a bomb on a plane via the passenger vector” sufficiently difficult (and you can’t, actually, because you can’t do a body cavity search on everybody and that’s where you’re going fast), they just stop trying and put the bomb in the food truck, or they jump a fence and try to put it on the plane, or they bomb the security line… or they don’t bother with bombs at all and they just charter a small plane and ram it into another plane.
Yep. See OP paragraphs 5-6.
I do think that we have differing definitions of “sufficiently difficult.” Or, at least, I see a spectrum. You can’t make it impossible, but you can make it more difficult. The problem is, once you make it more difficult than bombing a levy, you’ve made things worse. Airports may be safe, but nothing else is.
Which brings this back to my original point, which is that at some point you have to say “good enough, and now we take our chances.” I think that point falls well before profiling. Or freedom feel-ups, for that matter. Or banning liquids. Or everything else we’ve concocted since 9/11.
Me? I’d love to be able to keep my shoes on.
Germane to the thread:
http://www.schneier.com/blog/archives/2011/07/organized_crime.html
Trumwill:
Here is my problem, right now the folks that seem to be the ones that are hijacking planes and or trying to bomb them are young, male, not Caucasian, and Muslim. Given this, why not focus on those folks instead of say old Asian women? If this population changes then I would certainly be in favor of changing the focus, but let’s be honest about which groups the terrorist elements are coming from.
List the last 20 incidents that have occurred on American soil that you would classify as terrorism.
How many of those events were executed by someone who was young, male, not Caucasian, and Muslim?
Please show your work.
While we’re at it, are we going to aggressively screen white males who go into banking? After all, they are at the center of pretty much every major banking or finance scandal. What about profiling all white males during those rare times a genuine serial killer services? Evidence shows that most all serial killers are white males.
Part of the problem has to do with how we define “terrorism”. That guy who flew his plane into the IRS building (Joseph Stack, I believe?) was labeled several things, but rarely a terrorist. Somehow, him flying a plane into a building after leaving an anti-government manifesto didn’t quite meet the standards that many seek to use. Yet the guy who shot up the Army base (Hasan) was immediately labeled a terrorist, even though his attack was on a military installation, making it much more akin to a traditional act of war, albeit by an atypical attacker.
If we define terrorism as acts or attempted of mass violence by Muslim fanatics (as many have explicitly advocated changing the definition to), it’s no wonder we end up with only Muslims committing terrorism.
In my view, either these searches are an imposition or they are not. If they are not an imposition, why should we be all outraged at grandma’s getting searched? Or a child? If it’s age in the latter case, do we search Muslim children? If so, then age isn’t the issue. If not, why not? I can pretty easily conceive of ways that could be exploited if we loudly declare that we won’t.
If the searches are an imposition, then we have to be asking ourselves if some sort of burden of proof is required to impose something we consider to be disturbing and a violation. Doing this on the grounds of ethnicity is simply problematic, from a fairness point of view.
There isn’t much to say in favor of our current system, but at least it helps us recognize why someone might consider this intrusive. If we were only doing this to Muslims, I can’t help but believe that a lot of folks who would never have to worry about it would minimize the significance of the intrusion. By saying it can and will happen to everyone, we prevent it from being more intrusive than it otherwise could be.
Which gets to the real point of my post, which is what the endgame is here. If we’re willing to take our chances with a Joseph Stack loon, can we really talk about the comparatively few actual attempts that have occurred since 9/11 (and how they were both stopped regardless) as being worthy of all of this hassle?