Racial Profiling Is Necessary To Defeat Terrorism

It only makes mathematical sense. Think about it - our biggest terrorist threat nowadays arises from Islamic extremism. Obviously not every Arab-looking individual is a terrorist, but it is (indisputably) mathematically more likely that he/she is dangerous than an 80-year old grandmother. Given the fact that the police and security forces have only limited resources, of course it makes sense to allocated these resources to the greatest threats. Using probability to help sniff out terrorist threats should be a key weapon in the fight against terrorism.

That is the basic contours argument often stated by racial profiling supporters, anyway. We hear it all the time (and here), and it does seem to have a persuasive sheen to it, superficially at least. In an age where we are more likely to trade liberty for security, it carries a good deal of weight and is thus an important argument to address.

Because of that, I'd like to get your opinion on just what is the best argument against racial profiling as the calls for such methods inevitably increase. I've identified a few:

1.    The "libertarian" argument. People should not be placed in groups simply because they share some of the same superficial physical characteristics as other dangerous individuals. Individuals should be treated as individuals on their own merits, even if it is "true" that Arab men are more likely to be terrorists.
2.    The "constitutional" argument. Without question, racial profiling contravenes America's constitutional traditions. Everyone is entitled to the equal protection of the laws, and when an assumption of guilt is placed upon members of a particular discreet and insular minority, it is a clear violation of equal protection.
3.    The "losing the war" argument. By resorting to racial profiling techniques, the terrorists have already won an important victory. We prize liberty and freedom; they want to destroy this. By reducing our own liberty, we help the terrorists accomplish their ultimate goals.
4.    The "slippery slope" argument. Giving the government greater police power to stop and search innocent individuals on the basis of little or no suspicion opens the door for even greater police abuses down the road. Recent actions by the British police only add to the suspicion that the police will eventually abuse their expanded powers.
5.    The "futility" argument. Racial profiling, though perhaps compelling in theory, is not practically effective. Focusing on a large group (say, all Arab men) to help find only one or two individual terrorists is a waste of police resources. There's nothing to say that those few individuals won't still slip through. Stops based upon individualized suspicion are probably more effective.
6.    The "moral" argument. Simply put, it is immoral to profile people based upon their race without individualized suspicion. Think the "golden rule" - this is not how we would want to be treated. A further scientific-sounding "rational" argument isn't really even necessary, for the same reason that such a "rational" argument is unnecessary when criticizing slavery - it is simply objectively wrong.
7.    The "costs and benefits" argument. You think your mathematical pro-profiling argument is good? Well how about this - the costs of racial profiling are high in terms of police resources, lost productivity, and public resentment. The benefits, on the other hand, are small - you are still very unlikely to catch clever terrorists. Because the calculated costs here outweigh any potential benefits, profiling is not a public good.
8.    The "unintended consequences" argument. By placing an assumption of guilt upon all members of a group with no reasonable suspicion at all simply based upon unchangeable racial characteristics, you seriously undermine (understandably) that group's attitude towards the government. Getting help from the Arab-American community may be crucial in fighting Islamic terrorism, and if profiling causes fewer people in that community to trust and help the police, profiling will have seriously backfired.

This is not meant to encapsulate all possible arguments against profiling, and clearly there's some overlap here. With that said, I would like to hear from you about which of these arguments (or any other argument) you believe is the best response to the "mathematical" pro-profiling argument presented above. I'm not saying that there must be a single "best" argument here - but if you had to choose just one, and you considered the question both from theoretical "rightness" and based upon which argument might have the most practical political traction, which would you choose?