First, let me start by defining xenophobia. In my understanding, xenophobia is an irrational fear of those who are seen as foreign or different in some way. Disliking something that just so happens to be different while having good reasons for doing so (e.g. a westerner disliking Islam on accounta it's not actually true, or on accounta it's really shitty to women, or some other similar thing) and responding rationally to that dislike (e.g. arguing against Islam in public fora while still working hard to ensure that the civil rights of Muslims are respected) is not xenophobic. On the other hand, hating something because it's different and/or for other irrational reasons (e.g. disliking Muslims because their crazy vengeful patriarchal asshole of a deity is slightly different from your crazy vengeful patriarchal asshole of a deity), especially when one responds irrationally to that dislike (e.g. wanting to turn the Middle East into a sea of radioactive glass), is xenophobic.
My contention in regard to profiling of "Muslim-looking" people is that it's an irrational response to fears of Islamic terrorism, and thus is xenophobic. There are a few reasons that I say it's irrational:
- First of all, terrorists are a tiny minority even of Muslim or "Muslim-looking" airplane passengers, so to react with fear to all "Muslim-looking" airplane passengers is a carelessly broad-brush response to the small number of them who are actually terrorists.
- Second, focusing on Muslims concentrates all of the fear of terrorism onto a group of people who, while they do seem to constitute the majority of modern airline terrorists, are not the only terrorists in the world (cf. Wade Michael Page, Anders Breivik, Eric Rudolph, Timothy McVeigh, et al.). Thus we irrationally neglect other terrorist threats when we focus only on Islamic terrorism.
- Third, profiling "Muslim-looking" people imposes disproportionately high costs on these already-marginalized groups in an attempt to mitigate a threat that kills even fewer people every year than lightning does.
- And, fourth, profiling of this type doesn't even fucking work. In fact, while profiling of Muslim-looking people would certainly be costly, both financially and in terms of the harm it does to the innocent individuals who are treated like criminals, it may actually make us more vulnerable to terrorist attacks rather than less.
I've noticed that people don't tend to believe this last point, because, sure, profiling Muslim-looking people sounds pretty good, doesn't it? We know that currently the people who blow up/hijack airplanes are more likely to be Muslims than not, so you subject all of the probable Muslims to heightened scrutiny and give everybody else a lesser screening, in order to concentrate your terrorism detection resources on the group which is most likely to contain terrorists. Therefore you catch more of the terrorists and everybody wins. Yay! But there are several problems with this.
Sam Harris debated Bruce Schneier on this subject recently. Bruce is a genuine security expert, unlike the rest of us plebs who just tend to go on our gut feelings on these issues, and unfortunately Sam, despite his expertise in other areas, turned out to be just another gut-feeling pleb on this issue. In their debate, Sam argued in favor of this particular form of profiling, while Bruce argued against. I summarize below the problems with eyeball profiling that Bruce and other such experts have described, and also the better alternatives that they tend to recommend.
The first problem is that it's not clear how you identify a Muslim just by looking. All it takes to become a Muslim is to say the shahadah in Arabic in front of witnesses. And then you're a Muslim. You don't have to be of Arabic descent. You don't have to have an Arabic accent. You don't have to mark your body or have a certain haircut or wear a beard or carry a Koran or dress in a particular way. And, in fact, if you are intending to commit an act of terrorism in a western country, you don't want to do any of these things, because you don't want to stand out, and you know that everybody's all freaked out about Muslims and so if you're visibly Muslim they will be paying more attention to you even if there isn't an explicit profiling program in place. So you get a western haircut and shave your beard, you train away your accent, and you dress in jeans and a T-shirt and sunglasses and nobody looks at you twice.
So "looking Muslim" isn't actually necessarily a terribly good proxy for "being Muslim", and it's especially not a very good proxy for "being one of the Muslims who wants to blow up an airplane". Thus we don't help our airport screeners out very much if we just tell them to concentrate on all the people who "look Muslim". So what ends up happening instead is that, lacking any actual useful guidance, they pick on people for (a) being of Arabic descent (or at least, for being of what a bunch of ill-educated screeners think looks like Arabic descent) or (b) being comfortable displaying their Muslim religious beliefs in their attire (or, being comfortable displaying Sikh beliefs in their attire in front of screeners who don't know the difference). In other words, our screening becomes racist *and* it's more likely to pick on Muslims and others who have no harmful intentions than it is to pick on Muslims who *do* have harmful intentions.
The second problem is that if you declare that you're going to concentrate your attention on people who "look Muslim" or "look terroristy" in some way, then you thereby imply that certain other people will be subjected to lesser scrutiny. Groups that people often get outraged about the ridiculousness of screening intensively include old people, disabled people, children, women, and blond, blue-eyed white people. So then what's going to happen is that the terrorist leaders are going to make sure that when they recruit people for martyrdom operations, they recruit those who match your "low-scrutiny" profile rather than your "high-scrutiny" profile. So in fact instead of increasing your security by focusing on the people who you believe are more likely to be terrorists, what you do by creating "high-scrutiny" and "low-scrutiny" classes is create a giant hole in your security with a big flashing neon sign pointing to it that says, "Hey, terrorists, this is how you should look if you want to waltz right through our security." And then the terrorists waltz right through your security and you're left sitting around with your thumb up your butt wondering what happened.
The third problem is that even if the terrorists don't change their strategies, and even if we assume that every single terrorist is a member of your "Muslim-looking person" class, terrorists are still extremely rare within that class. There are about 700 million airline passengers in the U.S. every year. Muslims make up around 1% of the population of the country, so assuming they are similarly represented in air travel, that means that about 7 million Muslims fly every year. I'm having trouble finding numbers on attempted airline attacks, but let's be generous and say that starting with the 19 hijackers back in September 2001 there have been at most a hundred or so terrorists who attempted attacks which involved passing through airport security. I'm pretty sure this is a massive overestimate, but, as I said, I'm being generous. So that's about 10 terrorists every year, and we'll assume for the sake of argument that they're all Muslim and all "looked Muslim". This means that at most a little over one in a million Muslims attempting to board an airplane in the United States is an actual terrorist. And of course that means that almost certainly much less than one in a million "Muslim-looking" people will be a terrorist, seeing as how we'll end up sweeping up all kinds of Sikhs and Jews and non-Muslim Mexicans and Italians and Eastern Europeans and Jamaicans and so forth into our "Muslim-looking" net as well.
Now assume for the sake of argument that we're subjecting all these "Muslim-looking" folks to a "heightened screening" procedure which has a negligible rate of false negatives and produces false positives only 1% of the time. Sounds pretty good, right? That means that every year we will most likely catch all ten terrorists for the year. However, we'll also end up flagging as terrorists approximately 70,000 non-terrorists who just happen to "look Muslim", and end up putting them through the humiliation, travel delays, possible jail time, lost wages, and other expenses (to them and to our government) required for them to prove to the satisfaction of our judicial system that they are not terrorists. In other words, even under the assumption that this profiling idea actually works to detect terrorists (which it doesn't, as I explained above), in order to stop those ten terrorists, we've just terrorized 70,000 perfectly innocent people for the crime of being of the wrong ethnic background, having the wrong accent, or dressing in the wrong way. In other words, we've just made our society a hell of a lot more racist in its impacts on people and done an awful lot of damage to relations between those communities and the majority ethnic and religious groups, even though we thought we were doing a good thing. In fact, we've probably pissed off some people (not even necessarily the ones we profiled) so much that we've created a few more terrorists in the process, thus setting ourselves up for next year's round of attacks.
And fourth and finally, there are of course the false negatives to worry about too. What if the next white supremacist jackhole does decide that he's going to blow up a plane? Oops, our profiling on "Muslim-looking" people just missed him, and now there's another flaming crater filled with plane fragments somewhere. Even if only one out of the 693 million non-Muslim passengers is a terrorist, that's all it takes to kill a lot of people, and we just ignored that guy because he fit very nicely into our low-scrutiny class.
So, for all these reasons (Muslims who don't "look Muslim", the explicit creation of a loophole for terrorists to sneak through, the substantial cost associated with flagging a large number of false positives, the non-negligible possibility of false negatives, and the incredibly discriminatory effects of the system), profiling by "looking Muslim" or any other such lazy heuristic is just an incredibly bad idea.
And, in fact, it's a bad idea with such racist effects that we shouldn't do it even if it was a good idea. Most people who aren't racist jagoffs recognize that police profiling of African Americans as supposedly more likely to be criminals is part of the problem in our society rather than part of the solution to our problems, so it's not clear why anyone thinks that it's okay to do it to Muslims just because that's about terrorism instead of drug crimes. After all, not only is the average person a hell of a lot more likely to be killed by lightning than by a terrorist attack, we're also a hell of a lot more likely to be killed in a drug-related crime than by lightning. So by this rationale, if profiling worked and we considered it to be an acceptable tactic, then we should be applying it first to people involved in the illegal drug industry, and worry about all that trivial terrorism nonsense later. And this is of course ignoring the fact that profiling doesn't actually work.
What does actually work is a combination of several things. First, you profile based on a large set of of very carefully selected and very specific criteria such as details of behavior at the airport, ticket purchase patterns, and existing law-enforcement information, rather than just on the single very ill-defined criterion of "looking Muslim". Then, if somebody or a group of somebodies gets the right pattern of hits on such a profile, you pull them aside for more rigorous scrutiny. This kind of precisely specified multifactor profiling reduces false positive rates to much more manageable levels, and supposedly when done well it has pretty acceptable false negative rates as well. But of course, you can't completely trust that, so you also single out random people for more rigorous scrutiny too, including the little old ladies and the toddlers in wheelchairs and the blond, blue-eyed white people and the airport employees. This makes it impossible for the terrorists to be confident that they won't get caught even if they somehow manage to evade your profile.
But of course, what you really want to do is catch the terrorists before they make it to the airport. The thing is, even if a terrorist is sufficiently shady in some way that your airport screening would've detected hir, ze doesn't actually have to get through airport security to do a lot of damage. Ze could, in fact, just blow hirself up right there in the packed crowd waiting to go through the security checkpoint. Or ze could skip out on airports altogether and target a shopping mall, or a movie theater, or a house of worship, or a school, or a military base, or a government building, or a random block of row houses, or a large office or apartment building.
You can't possibly harden all these targets without making everybody's lives an Orwellian nightmare (even if you had the money for it, which we don't), so ultimately you have to rely on getting good, targeted intelligence. Find the violent extremist groups, find the people who supply them with weapons, and watch them, and when they look like they're going to do something, stop them before they get too far. And of course the best way to get this kind of information is to cultivate good relationships in the communities these extremists come from (Muslim and otherwise), so that when people see their neighbors getting all terroristy they report them to the police rather than cheering them on. (Some people propose massive electronic data gathering instead, but that tends to have its own enormous false positive problem. Specific, local intelligence from good informants which then allows you to do precisely targeted electronic surveillance is a hell of a lot better, if you can get it.) But of course if Muslims don't trust the police because they know the police will treat them like suspects just for being Muslim, they're not going to be willing to help out. This means that by alienating Muslims with clumsy profiling tactics, you've just lost what was probably your best chance for actually preventing attacks.
So all of this is why I say that profiling Muslims is xenophobic: it's just a lazy gut-level response to fears of terrorism — these people are other and a small number of them did something bad so I will distrust and discriminate against them all — rather than the result of careful analysis of the issues. Careful analysis says that hassling people for "looking Muslim" is a pretty shitty strategy, and that we need to do something entirely different instead.
As a scientist, I tend to think that the the best way to avoid letting one's gut lead one into xenophobia and other forms of harmful irrationality is to distrust those reptilian-brain instincts when one recognizes them and seek out more careful analyses. Security experts, anthropologists, neuroscientists, political scientists, and others have a lot of useful things to tell us about how to respond to threats from violent extremists, and it would behoove us to actually pay attention to what they have to say, rather than just cruising along blindly on our default tribalism.