The Dilemma of Counterterrorism Policy

Any institution delegated with the task of preventing terrorism has a dilemma: they can either do their best to prevent terrorism, or they can do their best to make sure they’re not blamed for any terrorist attacks. I’ve talked about this dilemma for a while now, and it’s nice to see some research results that demonstrate its effects.

A. Peter McGraw, Alexander Todorov, and Howard Kunreuther, “A Policy Maker’s Dilemma: Preventing Terrorism or Preventing Blame,” Organizational Behavior and Human Decision Processes, 115 (May 2011): 25-34.

Abstract: Although anti-terrorism policy should be based on a normative treatment of risk that incorporates likelihoods of attack, policy makers’ anti-terror decisions may be influenced by the blame they expect from failing to prevent attacks. We show that people’s anti-terror budget priorities before a perceived attack and blame judgments after a perceived attack are associated with the attack’s severity and how upsetting it is but largely independent of its likelihood. We also show that anti-terror budget priorities are influenced by directly highlighting the likelihood of the attack, but because of outcome biases, highlighting the attack’s prior likelihood has no influence on judgments of blame, severity, or emotion after an attack is perceived to have occurred. Thus, because of accountability effects, we propose policy makers face a dilemma: prevent terrorism using normative methods that incorporate the likelihood of attack or prevent blame by preventing terrorist attacks the public find most blameworthy.

Think about this with respect to the TSA. Are they doing their best to mitigate terrorism, or are they doing their best to ensure that if there’s a terrorist attack the public doesn’t blame the TSA for missing it?

Tags: DHS, homeland security, incentives, terrorism, TSA

Posted on August 19, 2011 at 8:55 AM • 22 Comments

Comments

Clive Robinson • August 19, 2011 9:52 AM

Glib response number 1,

So wasting other peoples money in huge amounts to CYA is a normative behavior for a person paid from the same purse.

Paul Dittrich • August 19, 2011 10:27 AM

I will be very surprised if any regular reader of this blog casts a vote for “doing their best to mitigate terrorism”

IMO, the TSA is just one of the most visible examples of a government agency which places a much higher priority on appearing to solve a problem rather than actually solving the problem.

rdm • August 19, 2011 10:35 AM

It’s hard for me to understand how anyone can anticipate where blame will fall after an event. But I can imagine that annoying people with your presence will deter some people from complaining that you were not there.

Johnston • August 19, 2011 10:46 AM

There are lots of “interests” involved. Michael Chertoff for example is getting (more) rich from full body scanners.

Someone (almost surely) stole billions in cash in Iraq, a war that was presented as anti-terrorism. That doesn’t even include the huge bonanza for the military-industrial complex in general and subsequent bankrupting of the states (the icing on the cake).

Some of them are motivated primarily by money. Some are motivated by the expansion of their political power. Many fewer I think are motivated by the fear of potential blame. That’s not why the people who make the rules got into politics in the first place.

Anonymous 1 • August 19, 2011 11:04 AM

From what I hear it’s almost as if the TSA is trying their best to give people reasons to want to commit terrorism.

Johnston: Maybe not so much politicians (although even they want re-election) but for bureaucrats not being blamed may be the most important thing (of course even they’d like a larger department, but just not getting fired if something goes wrong would matter a lot to them).

Dan Gillmor • August 19, 2011 11:15 AM

The next time there’s an attack involving airplanes, TSA will surely get blamed in any case. Among other reasons, the press will suddenly discover that security theater has been the actual policy.

Don • August 19, 2011 11:48 AM

When it comes to politics, you mean to claim that perception can be more important than reality?

Doug • August 19, 2011 12:07 PM

In another context, if I leave my sleeping baby in the car to go into a convenience store I’ll be arrested if a cop comes by. Not because it is more dangerous to the baby than many common activities, but because the cop would be vilified if he walked on without finding me.

cakmpls • August 19, 2011 12:39 PM

That is, I presume, a rhetorical question.

Steve H • August 19, 2011 12:45 PM

In the cold war we made bombs bigger than the other guy, and created deterance. This was a false theory.
Now we remove privacy for security to “protect people”. This is a false theory.
When does perceived threat become a threat?
I have noticed that the biggest PR boost for government is better the the correct method of doing something.
Bomb in a Radio = all laptops scanned
bomb in shoes = all shoes removed
bomb in underwear = full body scanners
It shows a perceived solution, but you can still fake a pilot ID and walk thru.

paul • August 19, 2011 1:21 PM

Of course, averting blame only works for relatively small terrorist events (or ones that can be portrayed as relatively small in the media). Any time you have a huge smoking hole in the ground, your antiterrorism organization is going to get blamed regardless of what they did, unless they have a huge PR and disinformation campaign ready to go, and preferably an irrelevant new target to distract people’s attention.

Richard Steven Hack • August 19, 2011 1:39 PM

“We show that people’s anti-terror budget priorities before a perceived attack and blame judgments after a perceived attack are associated with the attack’s severity and how upsetting it is but largely independent of its likelihood.”

Well, that’s fairly obvious. Once an attack HAS occurred, the probability of that attack is unity. Of course, the probability of ANOTHER attack of that type cannot be predicted at all, other than “>0<1” – which is no prediction at all.

So people obsess over what happened and not what might or might not happen.

It’s tunnel vision caused by fear.

Nothing strange there.

The real issue is how the politicians and alleged “experts” react. And that is almost always intended to do two things: 1) remove blame from themselves, and 2) take advantage of the situation to increase their power.

And thus we have the TSA.

Nothing strange there either, if you understand the nature of government.

So we needed a research study to prove this? Really? I guess so. What that proves is how ignorant of how the real world works the entire population of said world is.

Oh, wait, nothing strange there either.

Pre Post Confusion • August 19, 2011 3:35 PM

I think this same question confronts software testers: minimize the likelihood of defects or minimize the likelihood of being blamed for defects which remain.

Have a look at this story.

They ran a test for every year from 2002 to 2009, but did not also test 2010. Seems like a reasonable test plan to me. But the news report speaks of “embarrassing errors”, and says a “crucial test was omitted.” Reader comments include “lack of competence,” and “doesn’t know how to cross a T or dot an I.”

The emotional stresses of being a tester probably outweigh the intellectual difficulties when deciding what to test and what not to test, even outside government bureaucracies.

Paso • August 19, 2011 3:39 PM

Sad thing is that blame avoidance is often opposed to fixing problems. So CYA behavior is normal and often gets in the way of actually helping people for a LOT of things, not just terrorism policy.

In fact, a lot of the problems in politics can be traced to an effect like this, not to mention many in private industry, etc.

jerry pournelle • August 20, 2011 12:13 AM

Both, I’d think. The problem for TSA is that there is nothing they can do that will stop a determined intelligent person willing to get killed doing it from bringing down an airplane. Assume a dedicated TSA officer competent enough to know this: what can be done other than do Kabuki, security theater, to prepare for the storm that comes when you fail? Relaxing the drill isn’t likely to make everyone less safe than going the full drill, but if the disaster happens after relaxing the drill the condemnation will be worse.

What in the world would you do, given that you can’t “discriminate” or “profile”?

Jerry Pournelle

Richard Steven Hack • August 20, 2011 2:40 AM

It’s called “telling the truth”, Jerry. That’s what one does.

Clive Robinson • August 20, 2011 3:02 AM

If you think about it the “fear of being blaimed” actually causes the bureaucrats from actualy making any worthwhile changes.

If you go through the reasoning it goes something like this,

1, Assume future terrorist attacks are a certainty.
2, Assume no matter how many attacks are detected or stoped a successfull attack will happen.

This is the old “terrorist only have to be lucky once, we have to be lucky every time” observation.

Now having decided that at some point you are going to fail how do you mitigate the problem.

As noted above by others one way is to “fight the last battle over and over”. That is you stay out of the prediction business and only work on what is known from the past. This is guarenteed to fail even if your adversary is not intelligent so how do you mitigate it. Simple spend all your resources as quickly as possible and blaim those above you for failing to provide the resources.

This works because if you go into the prediction business it does not matter how many times you get it right you know from 1&2 above at some point you are going to fail. When you do fail you know their is going to be an enquiry and you know it’s going to be an 4r5e kicking contest. The easiest way to have your butt kicked is to be identified by those running the enquiry as wasting resources. That is they say you wasted resources on prediction defences XY&Z which ment resources where not available to stop attack W. As those running the enquiry are part of the same club that provides the resources you know darn well that they will fight tooth and claw to find reasons of why they cannot be blaimed.

The result is you end up with a stalemate, any resources provided will only be spent on “known attacks” by the agency, and the money will keep flowing in because those providing it to the agency know what the game is and dare not be seen to be stinting on the supply of resources.

So logicaly the only way our defenses will improve in any meaningfull way is if a terrorist attacks in a new way…

Clive Robinson • August 20, 2011 4:15 AM

@ Jerry Pournelle : What in the world would you do, given that you can’t “discriminate” or “profile”?

That is a problem that needs explaining to those holding the purse strings, against those just itching to grab the big bucks that goes with profiling and discrimination, before the purse string holders get seduced down that path.

The problem is that profiling is a discriminatory process that works for some things like goods inwards testing of effectivly inanimate objects that can be measured, and only appears to work for others, and can be gamed by an intelligent adversary.

An example of such a discriminatory process is “credit profiling”, from a businesses position the balance sheet clearly shows it reduces losses to fraud when implemented so it works from that asspect.

However because the businesses can not see on the balence sheet, the profit they don’t make from the business credit profiling rejected, they assume incorrectly it is a good stratagem to follow.

Worse the businesses don’t see that the criminals learn to avoid the profiling so fraud starts to rise again. The reason is the profilers see it first and adjust their models, and thus exclude even greater numbers causing even more lost business.

Credit profiling provably gives rise to discrimination as even greater numbers cannot get needed credit, which eventually has what Richard Steven Hack referes to as “Blowback”.

The thing about this form of blowback is it comes from two directions, those discriminated against who can be identifed by the same profiling, and those who pass profiling but see it as an injustice, because it “favours the haves over the have nots”.

The prime example of this blowback through history has been “slavery”, it only realy stops in one of two ways, violently when the salves revolt or when enough of the slave owning classes recognise and convince their brethreen it’s wrong, and this might or might not be a peacfull revolution.

So there is historical evedence that discrimination eventual gives rise to it being cast aside often violently, the only question is how long it takes.

And that is the problem the more subtal the discrimination the longer it takes to get cast aside, and in that time unprincipled people profit by it.

Thus discrimination is a way to ensure “you and yours” ar in the “haves” group and can profit from the “have not’s” in various ways.

This permiates into such things as the social fabric like imigration, health insurance and taxation which as those in the US have seen can get extreamly fraught. And in other countries the discriminatory behaviour has caused riots, insurrection and out right revoult and the consiquent death and destruction that entails.

hope • August 20, 2011 5:21 AM

“And that is the problem the more subtal the discrimination the longer it takes to get cast aside, and in that time unprincipled people profit by it.

Thus discrimination is a way to ensure “you and yours” ar in the “haves” group and can profit from the “have not’s” in various ways
”
Wouldn’t there always be discrimnation thought?, there is bound to always be there in some form, maybe just the jones next door,all what not.

I’m guessing normal that would be a probem, but tecnology will change that,say the rich want a comfortable life style, they will need energy,health,food,water.. all off it can when used effecincy to make a large population live gets compressed into high energy(chain reaction/fire) that cuts both ways.

If the haves allow they nots to increase, it will still reach a tipping point but not based human group dinimics but tecnology, if the haves press down the nots to stop them riseing then its the human aspect not the tecnology that rebalances it.

Say for imgration you allow any one from any country to enter, but the can’t live within 4 blocks of another person from there old country, the counrty balance or force imgrants to assimalte to the culture as much as possiable, if damage start to increase lowing the influx to keep the check inplace should make it last longer, and give the country the ablilty to react to other country.

As the way the worlds going(Spock), unless there is another world to use the same system, that slow creeps will crash, rinse repeat.

worlds instead of counrtys.

MarkH • August 20, 2011 8:32 AM

@”Jerry Pournelle”, who wrote “What in the world would you do, given that you can’t ‘discriminate’ or ‘profile’?”

This is a right-wing chestnut whose patent falsity does not prevent its constant repetition. Security personnel are permitted to practice, and often do practice, many sorts of discrimination and profiling. Behavioral profiling is widely recognized as a potent security tool.

Profiling by ethnicity or apparent religious affiliation may be discouraged or prohibited — but these are in any case weak security practices. At Ben Gurion airport, is a person whose features appear Scandinavian or East Asian exempt from careful security screening?

Plainly, there are many who feel deeply frustrated that we don’t defend our security by techniques like Robert A. Heinlein’s racially-selective death ray. If Bruce has tried to teach one thing to the world during the past decade, it is that the way to effective security measures is not to follow our visceral reactions and simplistic answers — but rather to transcend them.

Anonymous 1 • August 20, 2011 12:43 PM

Doug:

In another context, if I leave my sleeping baby in the car to go into a convenience store I’ll be arrested if a cop comes by. Not because it is more dangerous to the baby than many common activities, but because the cop would be vilified if he walked on without finding me.

How is the cop meant to know you’re only leaving the baby for a short period of time? For all he knows that baby was already there for half an hour and you were planning to spend an hour at the shops (babies actually have died because people did that).

Steve H:

In the cold war we made bombs bigger than the other guy, and created deterance. This was a false theory.

Never mind that the US and USSR never actually directly fought each other, something they would have done if neither side had the bomb (and the bombs were only made as big as they needed to be, at first that was really big because targeting wasn’t very good but as missiles got more accurate warhead size actually reduced).

The obvious inability of the leadership to remain safe in a nuclear war (even if the leaders survive they won’t have much of a country left to run) is a very powerful thing and has likely saved hundreds of millions of lives (not that proxy wars are good, but they’re a lot better than the world war we’d have got otherwise).

jerry pournelle:

What in the world would you do, given that you can’t “discriminate” or “profile”?

I suspect much of the concern over Caucasian Muslims actually stems from people thinking that racial profiling works (of course some of it also comes from them tending to be recent converts).

Of course getting over the idea that racial profiling is effective would probably help there, if people could stop being racists of course.

Me • August 22, 2011 9:07 AM

“The next time there’s an attack involving airplanes, TSA will surely get blamed in any case. Among other reasons, the press will suddenly discover that security theater has been the actual policy.”

Trouble is, that the solution will be “bigger (more intrusive)” security theater.

Schneier on Security

The Dilemma of Counterterrorism Policy

Comments

Leave a comment Cancel reply