The TSA and the EWOT

Steven Horwitz

The op-ed that started all my fun with the TSA has now been published by the Washington Times.  It's just a little application of the economic way of thinking to the issue. 

Comments

[Troy Camplin]

More articles on how the government kills us when it "protects" us need to be published. May I suggest the FDA?

[Daniel Kuehn]

Another application of the EWOT is to think about the censored observed distribution of terrorist attacks (seen vs. unseen, in other words).

You deal primarily with the new proceedures here, but others are going farther than that and talkign about abolishing the TSA (also citing car vs. plane deaths and the unlikelihood of a terrorist attack). This ignores "the unseen" - the terrorist attacks that either were prevented by TSA (of which there were certainly some), or probably more importantly, the terrorist attacks that were never even attempted because the terrorists did not expect to be able to get past the TSA. An EWOT would incorporate the fact that the distribution of deaths from terrorist attacks on planes is an extremely selected distribution, and you can't make good inferences about the TSA without taking that selection mechanism into account.

This of course does not amount to a defense of the new proceedures or any other proceedures - nor would I want to make such a defense. But it is another example of the EWOT applied to the TSA.

[Daniel Kuehn]

Troy -
And this is the problem with a lot of critiques of the FDA too - not that they are wrong on any given point but that they take only the "seen" into account and not the "unseen". Because we have an FDA testing regime we have product development that is much more careful so that it can conform to that testing regime than it otherwise would have been. You (presumably) want to use the counter-factual of "what would our current pharmaceuticals do if they could sell any drug they want". But the actual counter-factual should be "what would the sort of pharmaceuticals that would exist without the FDA around do if they could sell any drug they want". Ceteris paribus is a dangerous concept if it strays too far away from the margin (and even on the margin its of less uses than a lot of people think).

Again - this clearly does not amount to a defense of the FDA, nor do I intend it to be.

[Dave Prychitko]

I agree with Daniel's first comment.

[Troy Camplin]

All one really has to do is compare drugs approved elsewhere that have not received approval by the FDA and compare the lives extended by that drug over the time period in question to see the number of unnecessary deaths caused by the FDA. It's significant.

One could also look at the number of years more the FDA requires of testing than other countries require for another measure.

This is all beside the point that businesses don't like to kill their customers -- something FDA supporters seem to fail to take into consideration. Either way, I think the above two considerations would be enough to roundly criticize the FDA.

[Daniel Kuehn]

Troy -
So you're still looking at an approval process elsewhere, right? (you say "that have been approved elsewhere"). That will just tell you whether the FDA is relatively more or less welfare enhancing than the European drug approval authority. It doesn't tell you anything about the value or lack of value of the existence of the FDA.

My position is not to say that the FDA is good to have (I would wager it is, but I can't empirically make that case). It's to say that trying to answer that question is considerably harder than you are suggesting it is.

re: "One could also look at the number of years more the FDA requires of testing than other countries require for another measure."

This is probably more fruitful if there is relative equivalence between drug testing standards and a lot of variation in required testing periods. You could probably figure out a workable empirical test from that. But it would still be pretty tenuous, and only a local impact. You still couldn't claim anything about the FDA more broadly.

re: "This is all beside the point that businesses don't like to kill their customers"

Oh that's cute.

You're probably right - they don't kill their customers. But they'll certainly screw them over if they don't expect to be found out, or even if they expect to be found out but not be found legally liable.

I agree - you certainly provide more than enough to criticize the FDA and I would join you in the criticism that you lay out. But it amounts to a critique, not a refutation.

[Daniel Kuehn]

*by "still a local impact" I mean it's a local treatment effect, not "local" in terms of geography. i.e. - it'll tell you what the impact would be of extending or reducing testing periods. It will not tell you whether having a testing period is good or not.

[Troy Camplin]

Did I say anything about it being a refutation of the FDA? One could have just as easily read into my initial comment that I was criticizing it. Much like the TSA is criticized, not necessarily refuted, by Steve.

Refutation may yet come about from such a suggested study, of course.

But since you bring up refutation, if we are able to prove that even with shady companies, fewer people would be dead as soon as they died if it weren't for the presence of the FDA, that might come closer to a refutation, would it not?

[Daniel Kuehn]

Troy -
First, I was just intending to identify exactly what such an analysis could prove - I wasn't trying to make a claim about your claims.

But speaking of your claims, you did suggest that this would assess "the number of unnecessary deaths" caused by the FDA - and I still think that's not true. Or put it this way - we should be more specific about the way in which it is true. It would show you the number of unnecessary deaths relative to some other drug approval organization. It would not show you the number of unnecessary deaths relative to having no FDA. The FDA could do worse than a European regime, but much better than some other counter-factual (for example, better than both a communist drug approval regime and a free-market drug approval regime).

And as you'll note, I also didn't suggest Steve was "refuting" the TSA - I noted his purposes was more circumscribed.

re: "But since you bring up refutation, if we are able to prove that even with shady companies, fewer people would be dead as soon as they died if it weren't for the presence of the FDA, that might come closer to a refutation, would it not?"

If you proved that then that would be a refutation, sure. I have no idea how you could prove that though.

[Troy Camplin]

One can do analyses that indicate a trend in one direction or another, but you seem to imply that you would settle for nothing more than what is impossible: experimental proof, or something equivalent (a system without an FDA, a system with one). Comparative studies can be highly suggestive.

[Cheap Louis Vuitton]

They hold a great expectation that their next generation will explore a full career potential and lead a more prosperous and successful life than their own

[Replica Christian Louboutin Shoes]

I love that we can sit in the silence together and know that silence is ok!