Bob Murphy sees a contradiction between my 2008 Bayesian commitment on TARP, my 2009 probability adjustment, and my current views.
In 2008, I said that if the economy tanked, “I still won’t be convinced [that TARP is a good idea], but I’ll be less skeptical than I am now.”
In 2009, after the economy tanked, I said, “I now admit to having underestimated the severity of the threat, and
marginally reduce my skepticism about the effectiveness of the bail-out.”
Two days ago, I said, “I’ve been against bail-outs from the beginning. So should have all
economists. It’s reasonable to debate the merits of contracyclical
monetary policy. It’s not reasonable to debate the merits of rewarding
failure on a grand scale.”
I don’t see a contradiction. While I don’t think it was ever reasonable to debate the merits of TARP, it would have been even less reasonable if, after mass hysteria, nothing much happened.
I guess you could say that it’s contradictory to participate in a debate that I consider “unreasonable.” But that’s a figure of speech. I’d also say that “It’s not reasonable to debate the merits of evolution.” However, if people who ought to know better started to question the merits of evolution, it might be reasonable to answer them.
P.S. Bob ridicules my original 2008 position thusly:
A guy says, “You have a fever? Take my special blend of rat poison as
medicine!” Then the guy takes it and his fever goes up. Bryan says,
“Hmm, my a priori hunch was that rat poison wouldn’t help, but man, it
turns out this guy was on his deathbed after all! I’ll need to
reevaluate my initial skepticism.”
Not bad. But you could write a similarly risible paragraph if the guy got better: “Hmm, my a priori hunch was that rat poison wouldn’t help, but the guy survived! I’ll need to
reevaluate my initial skepticism.” When you’re confident, it’s tempting to think that any outcome buttresses your initial view. But that’s mathematically impossible. If A makes you more confident in X, then not-A must make you less confident in X.
Of course, if you know X with absolute certainty, then neither A nor not-A should change your mind. At the same time, though, you’re then obliged to bet on X at infinite odds. So how about it, Bob? Name any absolutely certain non-tautological claim about the economy, and I’ll be happy to wager a fractional penny against everything you’ll ever own – even if I think you’re almost certainly right.
READER COMMENTS
RobertB
May 12 2010 at 12:29pm
A+ P.S.! Elementary concept, but it’s really illuminating to see the superficially convincing violation of it and the rebuttal. (It’s also great to see someone publicly making predictions in advance.)
Mike Rulle
May 12 2010 at 12:31pm
I remember commenting on your original 2008 statement. My recollection is I did not understand your Bayesian prior as it was the opposite of my prior.
My Bayesian prior was that this would create more chaos, not less. The basis for my prior was that California housing—-black hole central for the housing crisis—-was already recovering; i.e., sales were reaching close to record levels at lower prices.
I remember thinking that if things got worse both of our Bayesian priors would have been confirmed—your decreased skepticism on TARP and my increased skepticism. I did not (and do not) know what to make of this!
Bob Murphy
May 12 2010 at 1:28pm
Bryan and I actually totally agree on Bayesian updating and the mathematics of all this. Bryan, what I was objecting to was *not* that you still think TARP failed, but that you criticized current policymakers for not bothering to look at the track record of the first bailout…when your own empirical test came out as a plus for TARP! (So if they *had* been interested in the success or failure of the first bailout, and used Bryan Caplan as a guide, they would have to say that the evidence, if anything, favored TARP.)
“So how about it, Bob? Name any absolutely certain non-tautological claim about the economy, and I’ll be happy to wager a fractional penny against everything you’ll ever own – even if I think you’re almost certainly right.”
I am absolutely certain that Bryan Caplan will write something on economics that I will find ridiculous, at least once before September 30, 2010. I am willing to bet an infinite amount of money on this proposition.
Bob Murphy
May 12 2010 at 1:35pm
Robert B. wrote:
“(It’s also great to see someone publicly making predictions in advance.)”
This will be my last comment on this thread, so as not to beat a dead horse. But I want to make sure you guys understand my position here.
Bryan has done nothing courageous of the kind; he’s like Janet Reno taking “full responsibility” for Waco, and then not resigning. He “went out on a limb” and said that if unemployment breaks 8%, he will reevaluate the effectiveness of TARP.
And then, a long time after that daring public proclamation, he thinks it is so obvious that TARP has been a failure, that he chastises bailout proponents for not looking at the lessons of history.
Doesn’t anyone else see a problem with that, or am I taking crazy pills?
Silas Barta
May 12 2010 at 2:15pm
And then, a long time after that daring public proclamation, he thinks it is so obvious that TARP has been a failure, that he chastises bailout proponents for not looking at the lessons of history.
Doesn’t anyone else see a problem with that, or am I taking crazy pills?
Sure, there is a problem: Bryan had only discussed one datapoint. So if you look at the set of “all empirical evidence Bryan made a prediction about regarding TARP”, you will only find evidence favoring TARP.
The correct conclusion is that Bryan should have proposed more empirical metrics to gauge TARP’s success/failure so that it would be a little bit harder for politicians to twist his words to favor their policies.
(Of course, Bryan probably *did* do that, the discussion is just ignoring it.)
I don’t see the big deal. Your alternative seems to be “No evidence could ever convince me of the effectiveness of TARP.”
John Fast
May 12 2010 at 3:48pm
“No evidence could ever convince me of the effectiveness of TARP.
I’ll repeat what I asked in Bryan’s class a year ago: If someone has a fever and quacks treat him with leeches, which is better evidence for leeching being a good idea, his survival or his death?
I also think we ought to distinguish between two separate issues, which is “How sick is this patient/economy?” and “Is leeching/TARP a good idea?”
I will not argue with people who think the economy was in a bad way, I just don’t think that the bailout helped. I’m perfectly willing to consider the possibility that a bailout for the financial sector would have been helpful if it didn’t have all sorts of earmarks and other garbage attached to it. But swallowing a dose of penicillin wrapped in eight ounces of raw sewage is worse than taking nothing at all.
[Comment edited for unnecessary vulgar language–Econlib Ed.]
Brandon Minster
May 13 2010 at 1:28am
When you say “If A makes you more confident in X, then not-A must make you less confident in X,” what about what the Week 11 notes from my favorite microeconomics course EVER say about “confirmatory bias”? If X is “TARP isn’t helpful,” lots of economic results are ambiguous enough that two people can take a single event as both A and not-A. Is the latest unemployment figure evidence for or against TARP? I think it depends on whether you liked TARP in the first place.
Charlie
May 15 2010 at 1:29am
I just find it ironic that I was reading about a medicine called Warfarin the other day, which was and still is marketed as a rat poison.
Its prescribed to prevent blood clots, though, not fevers.
Charlie
May 15 2010 at 1:33am
“I am absolutely certain that Bryan Caplan will write something on economics that I will find ridiculous, at least once before September 30, 2010.”
So if Bryan stops writing for a few months, he can get Bob Murphy’s net worth plus some future income?
Comments are closed.