The Counter-Productive Effect of Racial Profiling

In much of the discourse on security and human rights, there is a continuum between the two: less of one means more of the other. If we want to live with more security, we give up certain rights; if we want more rights, we live with greater risk. Yet there is a further perception that certain rights mean more to some people than others. Take for example internment in Northern Ireland. Everyone in theory loses the right to habeas corpus, but in practice this is a greater sacrifice for the Catholic community as the practice was directed at them. Similarly in the West now, we assume that counter-terrorist measures will only be carried out against Muslim communities.

Image courtesy of dannyman © 2010, some rights reserved.
Image courtesy of dannyman © 2010, some rights reserved.

One expression of these perceptions is the use of profiling in airport security. First, the debate is carried out on the assumption that profiling is effective, and the question is whether the cost of human rights is worth the benefits. Secondly, it is generally assumed that the target of profiling is going to be Muslims, and so will not in practice affect the majority of the population, who if anything would gain from the practice with less time-consuming security practices.

The fundamental problem with this practice is, however, that it both undermines civil liberties and undermines security. In order to show this, I’ve developed a series of models. Let us assume for a moment that the only terrorist threat to aviation is from Islamic terrorists (something we know is not true), and that a state uses a policy of profiling to counter this single threat.

The most elementary problem with the practice is that profiling segregates people based upon physical characteristics: mainly race, but also by sex. Yet whether someone is a Muslim is not a question of race (or sex) and at the most basic level profiling cannot even accurately identify the group that it is currently aimed at. Therefore these models will work on the assumption that profiling is aimed at ethnic groups that are perceived to be Muslim, which I’ve designated ”stereotypical Muslim appearance’ (SMA). This includes not only racial groups from the Middle East, but also those from North Africa, Indonesia, et cetera which have high Muslim populations.

Aside from that, there are effectively three models of aviation security. The first is that all passengers go through security equally, with no racial discrimination. The second is that all passengers but a specified group are subject to the exact same procedures as they are in model 1, but that members of the targeted group are subject to extra procedures. The third is that regular passengers are subject to fewer procedures than in model 1, but targeted passengers are subject to the same procedures as model 3. Thematically these models look like this:

Model 1                  N =X                  T=X

Model 2                  N=X                  T>X

Model 3                  N<X                  T>X

N=Chance of member of group deemed non-threatening by transportation authorities being stopped

T= Chance of a person deemed threatening by transportation authorities being stopped

X=Standard chance of being stopped without profiling

So both models 2 and 3 involve some form of racial profiling. Yet, both of them increase the chance of a successful terrorist attack.

Profiling is ultimately based upon probability, so if, say, 90% of al-Qaeda members have Middle-Eastern appearance (and passports) then stopping persons that appear Middle-Eastern is more likely to uncover a member of al-Qaeda.

However this is a fundamentally wrong understanding of how counter-terrorism works. It assumes that terrorists are fixed in their methods, rather than acknowledging that there is a symbiotic relationship between terrorists and counter-terrorists. In short, terrorism is reactive: it looks for holes in the counter-terrorist’s strategy rather than acting solely of its own free will. Terrorist tactics and strategy are dictated largely by the actions of the counter-terrorist.

With this in mind, we should ask ourselves: what is the logical response of an al-Qaeda member to a counter-terrorist strategy of profiling? The answer is very simple: to use an operative who does not fit the racial profile being discriminated against.  In this case, that would be to use someone who does not appear Middle-Eastern to attack an airline.

Thus if we assume that al-Qaeda react rationally, the threat to aviation security will come from persons that do not conform to the traditional archetype of the al-Qaeda member. So how will the profiling system deal with this?

Model 3 is the most obvious liability: a non-Middle-Eastern in appearance al-Qaeda member has a higher chance of successfully carrying out an attack than they would if there was no profiling strategy. In essence then model 3 simply makes it easier for a rational terrorist group to carry out a terrorist attack.

Model 2 is more complicated, yet still potentially poses greater danger than a non-profiling system. Firstly, any policy of profiling will have an impact on security actors, who will naturally regard those who appear Middle-Eastern as more suspicious than their counterparts, which may breed complacency. Yet even aside from this increased possibility of human error, there are additional issues.

Model 2 is based upon the premise that X is not a completely adequate form of security. After all, if X was sufficient there would be no point in increasing it for anyone. If X is sufficient then only models 1 or 3 would be adopted. Therefore adopting model 2 is an acknowledgement that X is not sufficient, and so the security can be bypassed, but it also potentially points out its weaknesses. After all, if one ethnic group is subject to a certain security measure, and even if it is a random procedure only statistically more likely to be applied to them, then this is surely an indication that this area is a weakness when only X is applied?

The result of profiling is, therefore, to give an intelligent terrorist organization increased possibility to carry out a successful terrorist attack. Furthermore, there is ample evidence from history of terrorists exploiting this weakness. The historian Alistair Horne recounts one particularly brutal attack in the Battle for Algiers, at the Milk Bar Café, which was carried out by three female terrorists. They were chosen “for the job because, with their feminine allure and European looks, they could pass where a male terrorist could not.” Even if not the direct response to a policy of profiling, the FLN displayed awareness of the potential for the use of non-traditional bomb carriers: people who would not arouse suspicion.

Therefore, even working from the incorrect assumption that Islamic terrorist is the only threat we face, the adoption of racial profiling as a counter-terrorist strategy is something that is ultimately counter-productive. Opposition to this policy both increases our security and our human rights.

Leave a Reply