Shostack + Friends Blog Archive


New Security Measures: Effective, Non-intrusive

smoking-clown.jpgOr not.

The BBC reports that “10,000 bags misplaced at airports,” and a “Boy boards [a] plane without tickets (sic).”

Meanwhile, here at home, we have a program that engages in behavioral profiling in some airports. How effective is it? The New York Times reports in “Faces, Too, Are Searched at U.S. Airports:”

In nine months — a period in which about seven million people have flown out of Dulles — several hundred people have been referred for intense screening, and about 50 have been turned over to the police for follow-up questioning, said John F. Lenihan, the transportation agency’s security director at Dulles.

Of those, half a dozen have faced charges or other law enforcement follow-up…

So lets see. Six ouf of “several hundred.” I make “several” to be from 2 to 5. So that’s between .6% and 3% of the people these dedicated officers choose to interview end up being arrest-worthy. Now, estimates are there are roughly 10-15 million illegal immigrants. 10 million out of 300 million? 3%. So we would do no worse, and perhaps better, pulling people off the street at random, and violating their right to be left alone.

The evil clown shown here is not an airline security official, but SmokingClown4 by FaceIt, from Flickr.

6 comments on "New Security Measures: Effective, Non-intrusive"

  • Nick says:

    It seems that this is a pilot program. One has to assume that the effectiveness will increase as the program matures.
    Even still, approximately 1 of every 10,000 passengers in Dulles was being pulled aside for additional screening, a minor inconvienience to be sure. I am most concerned that the screeners will be incfluenced by racial or ethnic biases but I would hope that with proper training and oversight, this can be minimized.
    I’d rather have a person using human intuition that can adapt to a changing environment than an inflexible automated system. Of course, finding the balance between technology and programs like this would be ideal. This program or one similar to it,used to augment the relatively static system in place today, seems like a good thing to me.

  • Adam says:

    I agree with you that humans are a useful part of the security screening, I don’t agree that the fact that it’s 1 in 10,000 passengers makes the intrusion minor. As a reductio, what if we were beating 1 in 10,000? Would that be “a minor inconvienience?”
    We have substantial, constitutional restrictions on the ability of the police to randomly stop and interview people, because the authors of the constitution were well aware that the intrusion is not minor.

  • Nick says:

    I understand the argument against random intrusions but there are a few important points regarding this program that are being missed.
    The first is that the screeners are not police, they’re TSA employees with no legal authority. All they can do is stop someone from proceeding into the terminal and call the police who are assigned to the airport if necessary.
    Second, certainly, 1 in 10,000 beatings would not be acceptable but this is a bad analogy. Each person who is stopped is not physically assaulted and no long-term harm is done to them. A more thorough screening only takes 10-15 minutes which I do not consider too much to ask 100 passengers a month to adhere to (estimate based on a “few hundred” as stated in the article and that the program has been going for about 6 months).
    Third, these are not completely random selections. The training that these officials receive and the actions of the individuals they are stopping gives the screeners a strong belief that something is wrong.
    I’d prefer that some passengers be selected by qualified, trained officers who observe that something is amiss than the current system where 1 in 10,000 tickets are randomly tagged at check-in for additional screening. While I think entirely random screening also adds value, I think a balance should be achieved and if I were the one setting the scale, selection with human interaction would be the majority over completely random selections.
    This is very obviously a slippery slope and there is risk of misuse of this power. I expect that there will be a large amount of oversight over this program but in my opinion, if run properly, this type of human interaction is going to add value while having only a small impact.

  • Adam says:

    Your first point about TSA officers having no legal authority is irrelevant. People fear them. Second, the law is not only concerned with long-term harm. Assault and battery is generally not a long term harm. Miranda, Terry, and a slew of Supreme Court decisions limit the authority of police to stop and question. Finally, and third, yes they are random. That the police think they have a reason doesn’t mean they do. Such was the point of my calculations in the main post.
    PS: This administration agrees to no oversight of anything. I don’t see why you are optimistic.

  • Nicko says:

    Adam, you are wrong; they are not random. You may not like the basis on which people are chosen; whether they meet some “reasonable cause” threshold is open to debate but even if they fail to do so that does not make them random.
    The constitution and the supreme court’s interpretation of it rightly limit the authority of police to stop and search but they absolutely do not prohibit it. Since you raised an absurd “reductio”, let’s look at the other extreme. If instead of two TSA guys who have had four days in the classroom and three training days in the field, we had two trained psychologists with PhD theses in behavioural profiling, and extra searches were carried out if both of them had suspicions, would you be OK with that? I believe that case law indicates that that would probably be sufficient for arrest and forcible search, rather than merely choosing to perform a level of screening that the TSA are already authorised to perform.
    IANASCJ [*], but it seems that the constitution has historically been interpreted such that the greater the intrusion against the individual the higher the threshold of care before allowing the intrusion. Thus naturally for lesser intrusions we may allow a lower threshold. I agree that the there needs to be some open discussion as to both the level of training and the level of oversight for this sort of operation, but I certainly do not think it should be dismissed out of hand.
    [*] I Am Not A Supreme Court Judge 🙂

  • Adam says:

    You’re right, they’re not random. They do, however, seem hard to distinguish from random.

Comments are closed.