More idiocy by the TSA

[PeterNSteinmetz]

And this guy is making money off advising them. I bet the actual sensitivity of these tests, even combined, is just about zero.

https://www.foxnews.com/travel/ever...gers-red-flags-tsa-might-get-you-held-airport

 

[Bob Noel]

for those that can't get the fox link to work:

The airport behaviors that can get you pulled aside for TSA check

Jeff Price, an airport security coordinator-trainer, explained the behaviors, actions and physical traits that mean a traveler may have a higher chance of being pulled aside.

www.dailymail.co.uk

 

[rotorwrench]

I bet the actual sensitivity of these tests, even combined, is just about zero.

Can't speak much for his methods, but the core foundation for his approach has been used by the Israelis to great success for years. The Israeli version was originally one of the methods being looked at before TSA "solution" was passed.

 

[PeterNSteinmetz]

Can't speak much for his methods, but the core foundation for his approach has been used by the Israelis to great success for years. The Israeli version was originally one of the methods being looked at before TSA "solution" was passed.

Do we know that the Israeli approach actually works?

The problem with testing here is similar to that with the TSA. One is attempting to prevent what are very rare events. So rare in fact in the case of the TSA that we are not able to conclude with any statistical confidence that they actually work for something like 50 years. That is if what one wants to do is prevent attacks which destroy commercial airliners leaving US airports by non-crew members.

I do not know the underlying numbers in the case of Israel so can't really comment on the data from that perspective. Certainly the Israelis are very thorough and they believe this works.

Of historical note is that the FAA also tried profiling before the introduction of the metal scanners. But no real data on the efficacy of that either because hijacking had been decreasing since 3 years prior to that starting in 1969.

 

[Jim Logajan]

There isn't any evidence that these highly subjective criteria narrow things down. Indeed, when made public they become easily subverted.
A simple back-test is to examine what was likely observed of the terrorists and hijackers during past incidents. Did the 9-11 hijackers yawn a lot, etc.?

 

[Ron Wanttaja]

Do we know that the Israeli approach actually works?

I believe the last El Al hijacking was in 1970.

There's a difference in what security measures you can implement with the aircraft of a single, state-owned airline with ~50 aircraft vs. an international terminal like JFK....

Ron Wanttaja

 

[rotorwrench]

Do we know that the Israeli approach actually works?

Yes. And on several different levels. Its considered the gold standard in the industry and has been written about a lot. But considering the populace of entire countries would happily go out of their way to blow up an Israeli aircraft the fact they haven’t done so for many decades is a pretty good track record.

The interesting part is after 9/11 when the Israelis offered to teach the US their security system, the end result would have been a system that operated with less than a tenth of the personnel the current TSA takes to operate. However, the main reason it wasn’t adopted was some of the actions are not permitted in the US by law. But final nail was congress couldn’t pass up having 50,000 more government employees.

There's a difference in what security measures you can implement with the aircraft of a single, state-owned airline with ~50 aircraft vs. an international terminal like JFK....

True. But the same process is used wherever EL AL flies to and from regardless of location. And through the ICAO agreement they can use that same system even if its not acceptable in the host country once the passenger is ticketed and at the airport.

 

[PeterNSteinmetz]

Yes. And on several different levels. Its considered the gold standard in the industry and has been written about a lot. But considering the populace of entire countries would happily go out of their way to blow up an Israeli aircraft the fact they haven’t done so for many decades is a pretty good track record.

Maybe, but where is the data and analysis? It may work or it may be that something else is working. There are a lot of potential confounders here so I don't really think we can conclude that their airport security is what is doing the trick.

If people were really determined, I can think of a lot of other ways for people to blow up El Al aircraft, especially in other countries. Perhaps with a bit of imagination you can name a few?

The interesting part is after 9/11 when the Israelis offered to teach the US their security system, the end result would have been a system that operated with less than a tenth of the personnel the current TSA takes to operate. However, the main reason it wasn’t adopted was some of the actions are not permitted in the US by law. But final nail was congress couldn’t pass up having 50,000 more government employees.

The last is a very good point.

Overall, given the high variability in people's psychology, I am highly skeptical that any type of psychological profiling would have the effect of preventing hijackings, particularly of the non-destructive variety.

Indeed, if you look at the timeline of hijackings other than destroying planes by non-crew members in the US, there is a very strong drop in their rate in September 2001. It makes a fairly good case that spending the 8-10 Bn dollars per year on the TSA does decrease garden variety hijackings.

The problem of course is that says little or nothing about terrorist attacks and is probably not worth that much money to prevent the other hijacking types. Indeed, given the costs of the 9/11 attacks, we have probably now spent something like 10X that cost on trying to prevent them just in terms of dollars, leaving alone all the additional negative consequences of the TSA.

 

[rotorwrench]

Maybe, but where is the data and analysis?

The Israelis have it and don’t share it. Matter of fact everyone that was associated with the US training after 9/11 had to sign NDAs with the Israeli govt or something similar if they wanted to participate. Perhaps email EL AL and they'll share it? Or forward it to the Mossad for further action.

f people were really determined, I can think of a lot of other ways for people to blow up El Al aircraft, especially in other countries. Perhaps with a bit of imagination you can name a few?

Well considering EL AL flies to 100+ cities in North America, Europe, Middle East, and Asia, and hasn’t had an incident in 50+ years, I’d say that was pretty good. Since you know of a lot of other ways perhaps you should list them? I can’t think of any.

But just remember, the Israelis took out the Hezbollah head shed with exploding radios and pagers plus took out a Hamas leader in an Iranian military compound. Seems it would be pretty simple for them to keep people from blowing up or hijacking their airplanes.

 

[PeterNSteinmetz]

The Israelis have it and don’t share it. Matter of fact everyone that was associated with the US training after 9/11 had to sign NDAs with the Israeli govt or something similar if they wanted to participate. Perhaps email EL AL and they'll share it? Or forward it to the Mossad for further action.

Pretty typical excuse of government agencies that can’t actually demonstrate their efficacy. “It’s a national secret and we can’t tell you”. The TSA uses this all the time. Sorry, I don’t buy it and I don’t think anyone else should either.

Well considering EL AL flies to 100+ cities in North America, Europe, Middle East, and Asia, and hasn’t had an incident in 50+ years, I’d say that was pretty good. Since you know of a lot of other ways perhaps you should list them? I can’t think of any.

Oh this is easy. Just obtain some surface to air missiles or build them. Alternately attach some bombs at a field with lower security. Sabotage the planes in some other way at other fields. Many many ways.

Like I noted above, the fact that they haven’t had an attack in 50 years is not particularly good evidence that their airport security screening works.

But just remember, the Israelis took out the Hezbollah head shed with exploding radios and pagers plus took out a Hamas leader in an Iranian military compound. Seems it would be pretty simple for them to keep people from blowing up or hijacking their airplanes.

Again, think about it. Executing an attack on a defined target is a much smaller job than defending a very large attack surface. Yes, they are clever in some ways - and arguably violating international law - but they can’t be everywhere all the time.

In the case of the Israelis I suspect other methods are at work. They likely have better moles inside the organizations of their enemies and so can head things off in other ways.

What we are talking about here though is whether psychological profiling can work for air travel.

 

[PeterNSteinmetz]

There is actually a fair academic literature on this question. One can start with scholar.google.com and keywords “psychological profiling to prevent terrorist attacks”

There appears to be considerable skepticism that this can work. See for example https://research-repository.st-andrews.ac.uk/bitstream/handle/10023/3986/380-947-1-PB.pdf . From that article -

“SPOT has met with increasing criticism for not being able to detect terrorists: There have been 23 occasions where terrorists have travelled through SPOT security points and no interceptions have been made using the technique [53]. As Meyer points out: ‘Put bluntly, the program has a 100% failure rate'”


To better understand this I would suggest starting by critically reading the overviews and reviews that are available. The rest cited here is basically anecdote.

 

[rotorwrench]

Pretty typical excuse of government agencies that can’t actually demonstrate their efficacy. “It’s a national secret and we can’t tell you”. The TSA uses this all the time. Sorry, I don’t buy it and I don’t think anyone else should either.

What I always find interesting with our discussions is you are steadfast in that if there is no data or analytics presented, it must be false. Unfortunately, the vast majority of industrial data and analytics is only found in the private/paid sector and not in the public/free sector.

And the simple reason for this is that data and analytics either has value or protects a proprietary process like the Israeli aviation security process. Regardless, it how things roll in the real world of business and if you choose not to accept that, that’s fine too.

 

[PeterNSteinmetz]

What I always find interesting with our discussions is you are steadfast in that if there is no data or analytics presented, it must be false. Unfortunately, the vast majority of industrial data and analytics is only found in the private/paid sector and not in the public/free sector.

And the simple reason for this is that data and analytics either has value or protects a proprietary process like the Israeli aviation security process. Regardless, it how things roll in the real world of business and if you choose not to accept that, that’s fine too.

This strikes me as another one of those potentially non-falsifiable beliefs then.

One can always assert that something is true but the data and analysis can't be shown because it is proprietary or of national security interest. For example, I can say that my incantations down by the airport are what has kept there from being a terrorist attack there in the last 5 years. (That latter is true).

Then you ask me to demonstrate that the incantations are what is responsible and suggest an experiment. And then I say, "oh it is proprietary and I can't do that".

How would you ever falsify my statement? Even in principle?

Is it then reasonable for me to assert that since you can't prove my assertion is false, it must be true?

So the way I look at this is that the reason these sort of issues keep coming up in our discussions is that you are prone to non-falsifiable beliefs

The other thing I will add is that I have no problem with proprietary processes in a free marketplace and with private business processes. Those businesses have to persuade their customers to purchase by one means or another. What I have a big problem with is the government operating that way and forcing people to pay for things or do things they might otherwise choose to based on "secret" or "proprietary" processes. There might be a few cases during war when this is necessary, but certainly not for screening airline passengers.

Additionally, in this case, there is plenty of research out there which strongly suggests that you can't screen out terrorists by using behavioral profiling. Given that research, it is quite reasonable public policy for the taxpayers to demand that efficacy be demonstrated by one means or another.

As you noted, one of the reasons congress adopted the TSA solution to airline security is that it created 50,000 more government employees. I would suggest that the reason that government officials like to use secret processes without demonstrated efficacy is that it pays money to their friends in industry and increases budgets. Typical regulatory capture.

It does sound in any case like we are in agreement though that there is no demonstrated efficacy of using psychological profiling to identify terrorists

 

[rotorwrench]

So the way I look at this is that the reason these sort of issues keep coming up in our discussions is that you are prone to non-falsifiable beliefs

Not at all. The truth is none of the points I bring up are impossible to prove false. They simply require you to purchase the data, or in some cases, you must meet the requirements to access that data.

It's never been the data doesn't exist, it's more I'm not going to spend my money to make a point on a general internet discussion. Now if you are interested in that data I can look into connecting you with the proper data suppliers???

It does sound in any case like we are in agreement though that there is no demonstrated efficacy of using psychological profiling to identify terrorists

Ha. You'll note my original point was psycho profiling was only part of the established process with physical profiling and detailed physical searches rounding off the main factors in the Israeli process.

Regardless, trying to compare any TSA process or the wannabe in the OP article to established high level security processes makes for good entertainment.

 

[PeterNSteinmetz]

True about the TSA for sure and that clown.

And yes, other measures such as you mention may help reduce terrorism, perhaps. But we are talking here about whether behavioral profiling can do so because of the allusion to it working in post #3 above. That post did not mention that other items are needed or qualify the assertion about behavioral profiling in terms of being only part of the package.

I would be interested to hear more about the nature of this data and the cost. I don’t actually imagine it will really prove the point that behavioral profiling can identify terrorists or make a significant contribution to doing so. But I could be persuaded by a general description of the data and methods.

So what is the nature of the data and analytical methods used to prove this?

Now if you want to revise the assertion to - the Israeli approach in total works to reduce terrorism - I might be inclined to credit that more. But I don’t think that is then particularly strong evidence that behavioral profiling works.

 

[PeterNSteinmetz]

One other item I will note here is that whether a belief is falsifiable or not is independent of whether it is in fact true or false.

So applied to this case. If one can name some sort of evidence that could in principle falsify a belief, then it is falsifiable. It is stated above that these beliefs are falsifiable and that the data and analysis exist. So then it should be easy to name the sort of data which could in principle falsify this belief that behavioral profiling can work. Just modify the existing proprietary stuff to illustrate the point Certainly something general can be listed without revealing proprietary information or having to pay for it.

OTOH, I trust it is possible to see how others might begin to think the belief is non-falsifiable if no such evidence can be named generally.

 

[rotorwrench]

Certainly something general can be listed without revealing proprietary information or having to pay for it.

Not really. This data is basically treated as a business asset and is selfishly protected by the developer. And rarely is the core data or analytics shared with the consumer. Instead they receive a report covering their request with results based on that proprietary data or analysis. And all for a fee.

But I can definitely see why the false theory is an important concept for scientific data given most of that type data/analytics must be peer reviewed to be validated. However, there is no such requirement for the data or analytics I refer to.

 

[Jim Logajan]

Wikipedia has a list of notable hijackings going back to 1919:

List of aircraft hijackings - Wikipedia

en.wikipedia.org

Since Sept. 11, 2001 the only notable hijacking/terror attempts that have occurred in the US were by a suicidal man in a Cessna 208 and an off-duty pilot riding in a jump seat of a jet.

I found two hijackings involving El Al.

Other than loners with deranged motivations it isn't clear to me that the few hijackings undertaken by organized terrorists accomplished their goals - and thus appear to mostly apply their efforts elsewhere. After all, for sheer carnage one does not need to plant a suicide bomber on a plane.

 

[rotorwrench]

Other than loners with deranged motivations it isn't clear to me that the few hijackings undertaken by organized terrorists accomplished their goals - and thus appear to mostly apply their efforts elsewhere.

The one thing your list does not show is the number of incidents prevented which is the main part of any mitigation effort. And it's my understanding that number is higher than most think it is to include me. But I think you're correct they've moved on to other areas of interest.

And on the aviation side, drones have become the main threat potential in the last 5 years or so. So mitigative steps are being developed to address those issues.

Hopefully the powers to be will pursue a passive type system like the Israeli screening system vs another clown show attempt like TSA and we don't have Barney Fife drone police running around with anti-drone guns.

 

[PeterNSteinmetz]

Not really. This data is basically treated as a business asset and is selfishly protected by the developer. And rarely is the core data or analytics shared with the consumer. Instead they receive a report covering their request with results based on that proprietary data or analysis. And all for a fee.

That makes me wonder, have you personally seen any of these data or analyses which purport to show either that Israeli behavioral profiling or their system as a whole detects and stops terrorists?