TSA blows a billion bucks on unscientific "behavioral detection" program, reinvents phrenology

I hate the comparisons to Israeli airports. I am using wikipedia for reference here. Feel free to generate your own numbers, but I would request you show your work.

Israel has 12 airports total. That’s 5 international airports and 7 national airports. They are separate airports. The state of New York has 13 airports. And we mix international airports with national airports. A quick and likely inaccurate count suggests there are around 400 airports in the contiguous US.

Israel’s top 6 busiest airports serviced 15,414,419 passengers between 2011 and 2012. In 2012, Atlanta served 45,798,809 passengers. And that’s just ONE airport. The 6 busiest in the US handled 193,346,532 people in 2012.

In short, I am arguing that what Israel does works only because of the scale of their operations. I very strongly doubt we could scale that to what happens in the US and have it work as effectively.

Israel has a very easy profile – “Not Israeli”. There is no single profile and no combination of profiles that could reasonably cover the broad range of supposed threats to the US. To even cover the last 10 years worth of attacks on aircraft, you basically end up with a profile of “male and under 35 years of age”.

The other problem with profiling is it is very susceptible to a “clean skins” attack. Like everything else the TSA does, it ends up giving the appearance of security, but not really doing anything else.

5 Likes

I am in Israel, and everyone gets the once-over, everywhere. Taking the train? Bags get x-rayed and you get metal detected. Going to a shopping center, hotel, parking garage? Say hello to the security guy, who looks in your trunk no matter who you are, and may look in your bags if he wants to. I got driven onto a college campus by a professor with the right sticker on his window, but the security there looked like the security at a train station. Taken together with the difference in scale mentioned above, I’m willing to accept that Israelis are better at this than we in the US ever will be.

Edited to add: at the airport in Israel, the screening feels much more low-key (to me, at least) compared to the US. Keep your shoes and belts, for example. They have these things that look like 3-D x-ray machines for your bags, which you keep with you until they have been 'rayed. One or two people look you in the eye while asking why you were in the country and where you are going.

3 Likes

It’s very difficult to compare the US to Israel.

As others have pointed out, the sheer number of travelers just can’t be compared.

But the nature of air travel there is different, too. Israel, being roughly the size of New Jersey, has very little domestic air travel. The vast majority of air travelers are going in and out of a single airport (TLV) on international flights, where they’re already subject to customs and immigration inspection, and where some extra time in the security line for a personal interview (and EVERY passenger in and out of Israel gets interviewed before getting on the plane) isn’t that big a deal. Can you imagine the uproar from the US business travel community if every single passenger was forced to answer a bunch of probing questions at the airport before passing through security? Not to mention the cost…

The staffing is vastly different. Israel has mandatory co-ed military service after high school, but women generally don’t serve in combat units, so they have thousands of bright young women to perform all those interviews, many of whom will go on to top colleges afterwards. How many TSA BDOs do you think attended (or could get into) a top college?

And of course civil rights are treated differently. In the US it is (and IMO should be) illegal for TSA to pull people out of line for additional questioning just because they have an Arab or Muslim name. I’m proud that the US protects minorities as effectively as we do, but there’s no question that adding profiling to the security toolkit makes Israeli airport security’s job a little easier. That’s not to say that profiling = effective security (there have been non-Arab terrorists who have attacked Israel). But the US’s extreme aversion to it means that the TSA is forced to jump through hoops to make sure they aren’t accused of it, which can lead to expensive, ineffective programs like the TSA’s ridiculous BD program.

1 Like

Not when you’re trying to get an answer about a matter of fact.

When it comes to “real world knowledge” human beings are quite adept at rationalizing or completely ignoring failures while paying special attention to successes. This leads to confirmation bias. If you already think these practices work then, as far as you’ll be able to tell, they will.

When you step back and apply some skepticism many times you find no clear signs of efficacy.

That you personally believe these methods are effective does nothing to demonstrate that they actually are effective. For that you may have to rely on those lab coats. Sorry, it’s just how evidence-based reasoning works.

(Someone already pointed out that the “citations” already brought into the conversation to demonstrate the effectiveness of these methods are from an advertisement.)

3 Likes

Except in this case the effectiveness of the methods is quite demonstrable. I’m not presenting an anti-science position by any means but pointing out that rather than calling the situation a result of confirmation bias, look at the numbers of attacks thwarted vs a quantifiable threat impact analysis of loss of life prevented.

I didn’t say the situation was a result of confirmation bias; I said that confirmation bias frequently leads to excessive confidence in unproven means and that may be the case when you rely on “real world experience” and “common sense” and “intuition” instead of proper empirical study.

You haven’t presented any evidence; you apparently think I should accept your claims on the basis of you saying so. I don’t. If you want to provide some actual evidence feel free but I’m not going to hold my breath.

2 Likes

It all strikes me as an extension of the sort of woo-woo Jon Ronson wrote about in The Men Who Stare at Goats (please disregard the Clooney film), in which the interrogation techniques at Abu Grahib and Guantanamo Bay are linked to deranged magical thinking entrenched in the US military.

You could bankroll a lot of seasons of “Lie To Me” for that price

Summary with key results right at the top. Good thing you didn’t hold your breath.

Yeah, good thing. A news article that doesn’t cite sources is not exactly proof positive. I’m looking for some kind of quantitative (preferably unbiased) study that demonstrates its effectiveness, not a short news blurb that prefaces its claims with phrases like “some say”.

Perhaps the most bizarre aspect is that the headline seems to mention some incident of terrorism in Detroit foiled by Israeli airport security but this isn’t so much as mentioned in the article itself.

I thought you were sending us a link to some kind of quantified evidence that profiling practices in particular provided results. Instead, we get a fluff piece that says: “This success is due to a sophisticated system that combines intelligence reporting, profiling, and state-of-the-art technology for detecting weapons and explosives.”

There is no doubt that Israel has a good record for airline security. To what extent this is based on their ability to identify dangerous people by their behaviour is completely up in the air. Based on a large swath of studies that show that people basically can’t do that - including trained people - I would hazard a guess that very little of it is based on the ‘profiling’ part and much more is the intelligence and technology part.

Remember, what the TSA is trying to do is determine who to screen based on observing behaviour.

2 Likes

If you like you can read this paper (PDF) with a formal risk assessment of US vs Israeli methods. You can follow the list of references starting on p 137. Hope that helps satisfy your curiosity.

To what extent is also based on self reporting. I have to wonder how accurate that is, when looking at the NSA 's claims of intercepted terrorist plots, they can’t seem to be substantiated.

1 Like

The study looks at the overall effectiveness of the Israeli system but doesn’t seem to say much about the effectiveness of behavioral profiling techniques in particular. It’s especially telling that in the conclusion the author does not cite behavioral profiling techniques as among the advantages of the Israeli system over the US system:

Two primary reasons account for the difference in system effectiveness of the RBS and Israeli models with respect to the current U.S. model. Perhaps the most critical impact stems from how alarms are treated during the screening process. Unlike the current U.S. process, which largely resolves the individual alarm, the other two approaches treat every alarm as a systems alarm…The other element that contributes to differences in system effectiveness is the time allotted to complete primary X-ray screening.

There may well be some more useful studies among this study’s references but I don’t have enough time right now to take a look.

In short, no. That study does not satisfy my curiosity because it has precisely zero bearing on the question that is actually being discussed: are behavioral profiling techniques effective. (I’m wondering if you even read the conclusion before you linked to it.) Given that your initial comment was so blustery and dismissive of skeptics and that, when pressed, it took you three comments to provide anything that even looked like it might be evidence for your claims, and that, when examined, that evidence didn’t actually support your claims you’re not a particularly credible commentator from my perspective at this point.

That’s accurate. The security assessment at Israeli airports is simultaneously less intrusive than the American one, as well as more effective, I am willing to wager. Usually you are interviewed by a seated staffer, and observed by a standing one. The seated staffer asks some innocuous questions about your name, intent of visit, length of visit, &c and the other one just watches you for dissembling. If this does not go well, you will get a second screening in a secured area that is VERY THOROUGH.

In addition, they look for strange patterns. I have an Anglo name, but my wife and kids have Semitic names. They are totally fascinated by this and carry my passport away every time we travel to Israel. Do I speak Hebrew? (They always ask in Hebrew first). Am I Jewish? Am I dati? The last is a loaded question, as “dati” is a term of art amongst observant Jews.

You might take umbrage at privacy and freedoms and all that, but the other sense they convey well is “Fuck your privacy and freedoms. You are getting in our 500mph flying gas tank, answer these questions.”

1 Like

This is Homeland Security we’re talking about. Evidence is irrelevant from the get-go.

I look forward to our new “security scans”:

source

3 Likes

I was reading a Cool Tools article, and was inspired to build an attachment for a dremel router that will grind off offensive bumps from NSA and DHA skulls. Heading over to kickstarter now…

4 Likes

Makes it a whole lot smarter just to not go around pissing people off, eh?

I think that you are going to need a larger device – you only need to remove one bump from those pinheads…