1. Steve responds here. He writes: "You are not among those calling me an ideologue Daniel and the post was really aimed at, or at least frustrated with, those who who HAVE called me (and others) those names even as their own policies keep failing and failing. You've argued in good faith for the most part... That said, I ain't apologizing for that post."
OK, well I know I haven't called Steve an ideologue. I've probably called some specific Austrians ideologues before, but I'm guessing I had good reason to. Correct me if I'm wrong but I don't think Steve's argument was "the people that call me ideologues are ideologues themselves!" his argument was "We know monetary and fiscal stimulus failed and anyone who continues to argue otherwise is an ideologue, a dogmatist, guilty of state idolatry, etc.". If that was his logic, it still seems pretty (1.) presumptuous in his claims about stimulus, and (2.) condescending in its accusations. But perhaps we should take the qualifier that he was only refering to people who accuse him of being an ideologue at face value. There's nothing wrong with the alternative logic he originally provided in the post of course. In fact it's better than the logic he provides here - believing something despite evidence to the contrary is being an ideologue. I just think Steve is wrong about there being evidence to the contrary.
2. Jeff Friedman, editor of Critical Review, emailed me yesterday to aprove of my response to Steve here. He asked: "Can we boil it down and say that in a world of immense complexity, where controlled experimentation is usually impossible and the number of possible relevant variables uncountable, one can never know whether, in the absence of a policy that has failed to achieve the political promises made for it, things would have been better--or would have been worse?"
In the strictest sense, I think the answer is "of course". But I don't think we can stop there. After all, everything we do or fail to do is pregnant with uncertain implications. There is no safe default position that we can run to when we recognize the sort of uncertainty we're dealing with. So while I agree with Jeff that "we can never know", my response question is "well can we get a decent sense of it?" and I think the answer is clearly "yes". This leads me to,
3. Bob Murphy's comment on my response post. He writes: "I've found another overconfident guy, Daniel, who just last week declared:
"[W]e already know what isn’t working: the economic policy of the past two years — and the millions of Americans who should have jobs, but don’t."
Source.
(I know what you will say, chill out, that's still funny.)"
The source, to spoil it for those who haven't clicked through, is Paul Krugman. It could have been funny if Bob found a line where Krugman calls huge swaths of other people ideologues with absolutely no basis (just like Steve does). That would be funny, although the intended irony would be blunted by the fact that I've always been happy to admit that Krugman does this all the time. What I find odd about Bob's post is that I've never expressed concern with confidence in a particular theory. I'm not one of those people who thinks we're lost in the wilderness and that macroeconometrics is impossible and it's all a rouse (ex post story-telling, in Russ Roberts' words). I've always thought that nihilism was misplaced. I recognize the empirical problems and think that understanding those problems is precisely what helps us navigate towards answers that we can be confident in (if not 100% sure of). Krugman has done a good job, in my opinion, in defending his confidence. He has done a much worse job on his interpersonal skills. I've found Steve's defense of his confidence lacking - which is precisely why I posed the question "how do you know what you know?" Because the confidence only seems to be based on "Obama and Romer and Bernstein said we'd be better and we're not better" and that - to me at least - doesn't seem to be the best way to approach answering the question.
The more I see people arguing about historical data which doesn’t fit their theories (doesn’t matter if its Keynesians, Monetarists or even Austrians), the more I understand Mises why he rejected that kind of reasoning so thoroughly. Yet I am absolutely sure if unemployment would have stayed at 8% and lower, Keynesians would have cried victory! Although the same analysis of “Can we boil it down and say that in a world of immense complexity, where controlled experimentation is usually impossible and the number of possible relevant variables uncountable, one can never know whether, in the absence of a policy that has failed to achieve the political promises made for it, things would have been better--or would have been worse?” would apply with the same impact as it now does when it seems that the enacted policies have “failed”.
ReplyDeleteDaniel,
ReplyDeleteYou are fairly dogmatic in your defense of Keynesianism and Paul Krugman.
I realize you are probably sick of me saying this Daniel, but your double standards continue to astound me.
ReplyDelete(1) Steve Horwitz declares that the policies of the last two years have clearly failed, meaning we shouldn't have done them. You come back and say very sarcastically, "How can you know this Steve? The best minds in our science are working day and night on this! So far only two papers in the last 50 years have controlled for the IV issues properly, and even there we're really not sure. If you can be that sure, tell me, and I'll be the star of my grad program!"
(2) Krugman declares the policies of the last two years have clearly failed, meaning we should have done more. You say, "Yeah Krugman has a lot of evidence on his side, I got no problem with his claim."
You don't see any discrepancy between (1) and (2)?
No double standard whatsoever Bob.
ReplyDeleteI've told you what I think is a good argument and what I think is a bad argument. IV is a solution to the endogeneity bias, but for those that don't use those methods (or provide other solutions to those methods), we still know what direction endogeneity biases the estimate in. In addition to these empirical approaches I've talked about, we also have calibration of models that seem to predict what we see fairly well.
I've caveated the hell out of multiplier estimates. I've criticized poor studies relying on interstate comparisons that produce positive estimates. I've laid out pretty clearly what I think makes for a good estimate and what makes for a bad estimate, and while it's very hard to put a coherent picture together we do have results. Those seem to suggest that Krugman is standing on firmer ground than Steve. Krugman also gets points because the positive-multiplier model he presents is able to predict a lot of quirky things that Steves implicit model has a harder time predicting.
I have no interest at all in having a double-standard on this. I don't have a career of papers and statements anchoring me to one position. I just want to get a sense of what argument makes the most sense.
From what seems to me to be a good methodology we're in a dicey place but Krugman seems to do better than Horwitz. Both oughta tone down the ideologue accusations, but that's another point entirely.
Now - either show me how that's not the case given my prefered methodology or give me a reason to abandon my prefered methodology. But I'm not holding a double standard here.
Steve had a post pointing to the Romer-Bernstein forecast and arguing that that was sufficient for calling it a "failure". It's simply not convincing by what I think are reasonable standards of a convincing argument. I'm not the one that decided to make that argument - Steve is. So either tell me how I'm misapplying my standard to him or tell me why my standard is wrong.
Do I make different judgements than others? Yes. But I don't put my finger on the scales and I come into things presuming that most of you guys don't either. Giving Krugman a leg up would be completely meaningless.
Most people are saying deficits are the terrible, terrible problem facing us. Even a lot of otherwise liberal people have absorbed this paralysis over the deficit and assume it is THE problem. It's not like I'm embracing a popular position here.