The plenary talk is another story. It’s Krugman, of course, and his talk is posted here. To his credit, Krugman understands that he may not be up to the task:
It’s a great honor to be asked to give this talk, especially because I’m arguably not qualified to do so. I am, after all, not a Keynes scholar, nor any kind of serious intellectual historian. Nor have I spent most of my career doing macroeconomics. Until the late 1990s my contributions to that field were limited to international issues; although I kept up with macro research, I avoided getting into the frontline theoretical and empirical disputes. So what am I doing here?
Now, if you thought that Krugman was going to be humble, there are two clues in this first paragraph that he won’t be. One is the use of “arguably” in the first sentence. The second is in the third sentence, where he claims to have “kept up with macro research.” We all know that Krugman has not kept up, let alone understood what happened in the 1970s.
Of course it’s all downhill from here. Krugman thinks he was invited, (i) because he knows what a liquidity trap is, and (ii):
The other reason I’m here, I’d guess, is that these days I’m a very noisy, annoying public intellectual, which means among other things that I probably have a better sense than most technically competent economists of the arguments that actually drive political discourse and policy.
What’s that “technically competent” about? Is Krugman telling us he is not technically competent? Surely not. He passed the MIT prelims, was granted a PhD from the same institution, published some papers, and was granted a Nobel Prize in Economics. I think he means that he is a member of the set of technically competent economists, but he has some special qualifications, which are apparently being noisy and annoying. I can also attest that he can produce several blog posts per day, two New York Times columns per week, and can use FRED. I’m a little puzzled as to what insight this gives him into the world of policymaking relative to say, the set of residents of Manhattan, most of whom could also easily be trained to use FRED and blog for the New York Times. However, maybe Krugman wrote this in a hurry, so we should a least give him a chance. Maybe he has something new to say?
There is some talk about the General Theory. Here’s a puzzler.
Chapter 12 is a wonderful read, and a very useful check on the common tendency of economists to assume that markets are sensible and rational. But what I’m always looking for in economics is intuition pumps – ways to think about an economic situation that let you get beyond wordplay and prejudice, that seem to grant some deeper insight. And quasi-equilibrium stories are powerful intuition pumps, in a way that deep thoughts about fundamental uncertainty are not. The trick, always, is not to take your equilibrium stories too seriously, to understand that they’re aids to insight, not Truths; given that, I don’t believe that there’s anything wrong with using equilibrium analysis.
Intuition pumps? Going beyond prejudice and wordplay? What is the man trying to get across?
There’s more in here. We are reintroduced to the Keynesian Cross and IS/LM, which Krugman feels are sadly neglected, by graduate students:
But you can see right away part of our problem: who teaches the Samuelson cross these days? In particular, who teaches it in graduate school? It’s regarded as too crude, too old-fashioned to be even worth mentioning.
Actually, I like to talk about the Keynesian Cross in intermediate macro. It takes about five minutes and, indeed, most students have never seen it. One student said that he saw it in the appendix to his book in a high school AP macro course, i.e. the Keynesian Cross was the deep model that was deemed too difficult for the average student. What I use Keynesian Cross for, though, is something that Krugman would not approve of. It's a vehicle for showing the students what an Old Keynesian sweeps under the rug. To "crude, too old-fashioned," we can add "wrong."
Section 4 of the talk is titled "The Strange Death of Keynesian Policy." Where do people get these ideas? Would they like a world where macroeconomists thought about nothing but Keynes? When did Keynesian policy ever die? Here is what Krugman has in mind.
It [central banking] worked in part because the political insulation of central banks also gave them more than a bit of intellectual insulation, too. If we’re living in a Dark Age of macroeconomics, central banks have been its monasteries, hoarding and studying the ancient texts lost to the rest of the world. Even as the real business cycle people took over the professional journals, to the point where it became very hard to publish models in which monetary policy, let alone fiscal policy, matters, the research departments of the Fed system continued to study counter-cyclical policy in a relatively realistic way.
Now, in how many ways can I show you that this paragraph is out to lunch in a severe way?
1. Central banks are intellectually insulated. A marked development in central banks since about 1970, in the United States and elsewhwere, is that central bank economists have become increasingly interactive with academics, and have become much more active participants in serious macroeconomic research. At one time, a central banking job was perceived to be a dead end for a macroeconomic researcher. Now, it is common for people to move between academia and central banking (as I have done myself), central banks have active programs for visiting academics, and they run conferences where high-level research is presented and debated. Further, the standards for publication in many central bank research departments are as high as they are in serious economics departments.
2. Central bankers are busy studying ancient texts, presumably the General Theory and such things. See (1) above. These people are as much on the frontier as anyone.
3. It is hard to publish models in which monetary policy matters. Here are some that slipped through, and they are just by one guy (and coauthors) - one of those dark age economists:
a) Expectations and the Neutrality of Money
b) Money and Interest in a Cash-in-Advance Economy
c) Liquidity and Interest Rates
d) Interest Rates and Inflation
e) Menu Costs and Phillips Curves
There is more whining in the talk about the "intellectual regression of too much of our profession," etc., but you get the idea.
I'm wondering how much the organizers of this conference paid for "Mr. Keynes and the Moderns?" Did Krugman get his usual $80,000 (or whatever) speaking fee, or what? If so, they didn't get much for their money, other than recycled blog pieces. Now, the key question is: What would Keynes think, if he were at this conference? He would be the guy with the mustache in the corner, muttering: "Who invited that obnoxious journalist?"
It is pathetic that Krugman is still harping on about outmoded models, along with the other mummified Keynesian de Long.ReplyDelete
But we have to concede the point that these old models were on to something - market economies suffer periodically severe coordination failures. Krugman and other gravediggers of 1936 lack the technical apparatus to move the story on to more convincing models of coordination failure, but who does have a good model of this failure ?
This is more or less just Krugman's blog posts strung together into a speech, rather boring. I'm not sure how anybody can think the Keynesian Cross is useful. It's just a silly model. I remember when I first encountered the Keynesian Cross. I found it in a textbook at the library when I was 14. I thought it was stupid then and I think it's stupid now. It's amazing to me a Nobel Prize winner in economics (and I have no doubts he richly deserve the prize for his work, it was good work) thinks every problem in macro can be solved by moving a few curves around on a graph.ReplyDelete
Also, something you missed which I thought was especially odd is his claim that central banks are "monasteries" of this dark age of macroeconomics. The researchers seem pretty diverse to me, but the major policy makers in central banking seem more or less Keynesian (New and Old). Perhaps I'm mistaken, but was the rationale for the massive quantitative easing derived from some real business cycle model? Somehow I doubt it.
You might also find this interview with Krugman "interesting": http://thebrowser.com/interviews/paul-krugman-on-inspiration-liberal-economistReplyDelete
There is a large coordination failure literature, which he does not mention of course. Look at Roger Farmer's work to start.
Yes, the monastery stuff was another piece of nonsense. He knows little about macro or what macroeconomists actually do.
Yes, I saw that.
I find the constant Krugman bashing in this blog fairly annoying. It would be more interesting to see what interpretation can Stephen Williamson make of current events and what policy recommendations would flow from those interpretations.ReplyDelete
As it is, to the unitiated reader, SW's blog is reduced to bitter rants about Krugman Thoma et al, often descending to ad hominem attacks. It does not however have much to say on other issues.
The contrast with bloggers of a broadly similar political bent such as Greg Mankiw is fairly stark. One can agree or disagree with him, but he is using economic analysis to understand reality and coming up with insights.
I am sure that SW's research is very interesting and he has valid insights, it's just that one would not be able to tell from the blog.
I agree with the last anonymous. Please give the Krugman bashing a rest. At this point it's saying far more about you than him.ReplyDelete
"It would be more interesting to see what interpretation can Stephen Williamson make of current events and what policy recommendations would flow from those interpretations."ReplyDelete
Somebody has not been paying attention.
Exactly. You're not reading if you think I spend all my time on Krugman. However, as I've said before, I think someone has to do it. The man is an embarrassment to the profession, and people should know that. The comment about Mankiw is interesting. I hope you understand that Mankiw is careful because he is interested in working in Washington. Thus, to the extent he takes anyone on, he does not do it directly. Now, here's a question. What makes you think that Mankiw and I have a "similar political bent?"ReplyDelete
Krugman's post-2000 superficial writings appeal to the average person who does not have much background in economic thought. It is so obvious that if the government spends more money and creates new jobs, this adds to the national income and we're all better off thanks to the multiplier. No need to enquire any further. This is good enough for the average reader of NYT or Thoma's blog. Seems even to be good enough for Thoma, sad to say.ReplyDelete
anonymous texas adjunct
I don't like the Krugman bashing much either, but it's fair game: Krugman bashes modern macroeconomists on a daily basis, so it's only fair to see those people push back. While it'd be nice to see less bashing from both sides, at this point it's pretty clear that this is not going to happen. We've switched to a bad equilibrium in which both sides are condescending and find justification in the other's attitude. It's clear neither side will be the bigger man.ReplyDelete
I think we understand the blogosphere discussion now, and what motivates those people. The interesting thing in this context was to see what Krugman would do in an academic forum. The typical plenary talk at a conference is someone who is working at the research frontier in a field, who distills his or her research for a general audience of academics. Krugman is supposed to be talking about Keynes but, as he recognizes, he is not an expert in the field. So what will he say. I was pretty curious about it, but it was a let-down, as he just regurgitated his blog. A lazy mind at work apparently. What do you think Woodford thought of the talk? Most of us can credibly dissociate from what Krugman says, but Krugman wants to be a member of Woodford's club, I think.
I don't know what "side" you think I'm on. I'm just interested in honesty and good science. But we won't get that if we let dishonesty and bad ideas fly by.
I demand that you make this blog fit my interests rather than yours. I hate this blog so much that I feel the need to constantly read it and spend time writing comments telling you that you are too mean to the arbiters of truth, justice, and the Keynesian cross.ReplyDelete
Should I note that this is sarcasm? It seems so much like earlier comments that I suspect the answer is yes.
"Can't we all just get along?"ReplyDelete
"I demand that you make this blog fit my interests rather than yours. I hate this blog so much that I feel the need to constantly read it and spend time writing comments telling you that you are too mean to the arbiters of truth, justice, and the Keynesian cross."ReplyDelete
Your ideas intrigue me, and I wish to subscribe to your newsletter. In the event that you have no newsletter, I want a pony.
Your blog looks like whining without taking on the substance of his comments. A fun read, though, if somewhat pathetic.ReplyDelete
He correctly point out some embarrassing fallacies in John Cochrane’s writings that dates back to the "Keynesian cross". It pretty basic, just says that once output is endogenous, you can't say it’s a matter of accounting that G reduces C one to one. He also points out some very basic things with endogenous output and interest rate determination to pull the pants down on Niall Ferguson. Both pretty legitimate points and both date back pretty much to Keynes, so seem pretty nice for the occasion.
You seem pretty occupied with semantics. BW, it does ring true that much of the more interesting business cycle stuff is done at the Fed these days, because it stopped being fashionable in certain academic circles and journals, although of course he goes a bit over the top by bringing up pictures of monasteries. Oh, and the Lucas papers you link to seems mostly about why discretionary monetary policy is not very important or useful, pretty much underlining the point he's making......
I think Krugman criticism is a great service if it is on the substance and if it clearly explains why what he says is wrong.ReplyDelete
Of course those are important ifs. We really do need the explanation of why what we says is wrong. That really would be a great service.
June 20 12:37: fair enough, John Cochrane did appear to misrepresent neoclassical theory by using Ricardian equivalence argument incorrectly. Niall Ferguson is not an economist - who the hell cares what he thinks?ReplyDelete
Krugman shows no familiarity or interest with the ferment of research going on in macro today...such as Angeletos and Lao, Gabaix, Carvalho, Acemoglu, Gilchrist, Gertler, Brunnermeir, Mendoza, Gillman and many others.
He is coasting, relying on memories of macro research from the 1980s. He has an enormous responsibility as econ writer at the NYT. He should be reading the literature and making it accessible to the lay reader, not using shopworn old theories as a pivot for his partisan politicking.
"it does ring true that much of the more interesting business cycle stuff is done at the Fed these days, because it stopped being fashionable in certain academic circles and journals, although of course he goes a bit over the top by bringing up pictures of monasteries. Oh, and the Lucas papers you link to seems mostly about why discretionary monetary policy is not very important or useful, pretty much underlining the point he's making......"ReplyDelete
What do you mean, "ring true." It's not true. I think he's saying that what is interesting is work on New Keynesian models. Now, aside from the point that it would help if he were more specific, suppose we accept that as true. Then, (i) work on NK models is being done both inside and outside Feds, and (ii) that stuff is having no trouble getting into journals, whether top, middle, or mediocre. Of course there is plenty of interesting work that is not about NK models as well.
On Lucas: Those papers are all ones where "money matters." It doesn't matter in the way Krugman thinks it does, but so what? Is that supposed to be a foregone conclusion? I get my work published. Randy Wright gets his work published. Our work is all about how money matters. There is plenty of work on how fiscal policy matters as well, if Krugman would care to look.
I teach the Keynesian Cross and IS/LM in my graduate course. I think it is essential to know that model for several reasons. i) one needs to know the history of the field. ii) one needs to know IS/LM's weakenesses if one uses other approaches in one's research. iii) it allows one to understand and communicate with many policymakers/researchers across the world.ReplyDelete
PS., I also teach DSGE models and use them in my own research
Steve's Krugman-bashing was responsible for the Vancouver riots.ReplyDelete
I was actually in Canada when that happened, though in Kingston Ontario, not Vancouver. Canadians seemed thoroughly embarrassed by the whole thing, but the precedents go a long way back:ReplyDelete
Apparently hockey can bring out the Canadian dark side.
If I was allocated an order of magnitude more time, I would probably do that. As it is, time is so scarce that most of us can't waste it on the history of thought, and why some old stuff is wrong. It works well to start with basic representative agent neoclassical growth models, and then start adding things - heterogeneity, monetary frictions, incomplete markets, private information, monopolistic competition, sticky prices. It works, as that is the way most of these things evolved. Woodford, for example, is a direct descendant of Prescott, with Calvo and Dixit-Stiglitz thrown in. There's little of Keynes in basic Woodford, other than the language.
what is it in is/lm that is "wrong"? its true that methods in woodford are descendant of prescott etc, but thats just a question of language. the basic lesson, at a substantive level, of the simple woodford model is the same as in is/lm with a dynamic phillips curve added bit with more details. output demand determined, interest rate cuts increase output, etc.ReplyDelete
"what is it in is/lm that is "wrong"?"ReplyDelete
Practically everything, including no role for expectations, confusing real and nominal interest rates, and a totally-meaningless AD/AS diagram. A better question is what is in ISLM that is actually useful.
The way IS/LM was put together in the old days, there was a separate compartmentalized theory for each sector. We had a separate theory of consumption to give us the consumption function, and investment theory to give us the investment function, a money demand theory to give us the money demand function, etc. Then, the 1970s come along, and we start thinking about how all these things are related, in important ways. For example, the consumers who are supplying the labor are also consuming and holding money, so all those decisions have to be related. Further, we start thinking as well about how forward-looking decisions matter for macroeconomic behavior, and the static IS/LM model starts looking pretty lame. Fortunately some smart people gave us the tools to think about all of these interrelated and dynamic decisions. Woodford absorbed all that stuff and synthesized it, along with other people, into New Keynesian theory. He was brilliant at marketing the ideas, in part because he sold those ideas in a language that made them look like Old Keynesian ideas, which they are actually not. In any case, Woodford understands, as Krugman doesn't, that once you acquire a dominant technology you abandon the old one. When you have a Model T, you can throw away your buggy whip.ReplyDelete
Actually, I like the horse/car analogy a lot. Some people are just afraid to drive, but if you can coax them into the car and get them going, they like it. Further, we don't have to explain why the car dominates the horse to everyone to whom we teach driving.ReplyDelete
I like Bob King's quote: ISLM has as much chance of being a useful macroeconomic framework as the Pinto does of being the sporty car of the 1990s. Paraphrasing, of course.ReplyDelete
I don't know how easy this is to do in blogspot, but maybe you could somehow tag your posts "Krugman" or "non-Krugman" so readers can more easily filter. And I certainly agree Krugman is worth critiquing, as the most popularly read econ writer, but precisely because he has such readership there are plenty of other econ bloggers who devote considerable attention to criticizing him. Granted, a Bob Murphy is not a Scott Sumner is not a Steve Landsburg is not a Casey Mulligan is not a Steve Williamson. But you shouldn't think to yourself "If I don't do it, who will?"ReplyDelete
As a layman I found the proposed Marshallian/Walrasian distinction in macro to be interesting. What do you think of it?
On the Marshallian/Walrasian distinction:ReplyDelete
1. Friedman thought of himself as a Marshallian, as that is what he understood. Every economist has to have a theory, otherwise we cannot make sense out of the world. Friedman was brilliant, but very non-technical, and his theories were very simple.
2. DeLong misunderstands the people he is calling "Walrasian," in part because his theories are simple, on the order of Friedman's. On the senior end of the macro profession, people like Lucas, Sargent, and Prescott, are really just high-tech Marshallians, in the sense DeLong is using the word. They are intensely interested in the data and the interplay between the data and theory. It's not like they lost anything in terms of "realism" by doing sophisticated theory. In fact I think they would argue that what they do is more realistic.
Before I started reading this blog, I couldn't stand Krugman and was frustrated that so few commentators discussed economics from a "modern macro" viewpoint. Since reading this blog, I have become much more sympathetic to Krugman. He makes a lot of sound points and is prepared to back up his claims with evidence. I still don't agree with him on many issues, and I find his penchant for hyperbole and ad hominem attacks tiresome, but I find him well worth reading and he often convinces me to think about things in a different way.ReplyDelete
By contrast, I find this blog makes me seriously question the whole modern-macro paradigm, particularly the freshwater variety. I find this blog frequently falls into a smug, patronizing posture, which confuses internal consistency with truth and mathematics with science. Just because a model is based on rationality and general equilibrium does not imply it is an accurate representation of reality, or that it provides a better guide to policy makers than a model such as IS/LM.
Now, maybe this just says more about Krugman's rhetorical powers vs. Steven's, but I can't help thinking that people inside the modern-macro bubble are just convincing themselves and each other they are being "scientific" and "technical" and are so much smarter than people like Keynes (who I believe had a degree in mathematics from Cambridge) and Friedman despite the fact that they have a very hard time applying their theories to the real world (e.g., Lucas recently appealed to Friedman's explanation of the Depression, apparently modern macro wasn't relevant!). And then they become defensive when people like Krugman point out the glaring deficiencies in their approach. Maybe a little more humility on both sides would be beneficial.
"so much smarter than people like Keynes (who I believe had a degree in mathematics from Cambridge) and Friedman"ReplyDelete
We may not be smarter, but we know a lot more than they did -- we have more data and better tools. Just because Keynes said something doesn't make him right, and just because Friedman may have been right about one thing doesn't mean he was right about anything else.
And Krugman is never right, only arrogant. At least Prescott is right some of the time.
Seems like there are 2 issues re: Krugman.ReplyDelete
1) Is Krugman right to argue we should abandon DSGE methodology and go back to the
Hicksian ISLM ?
Ridiculous on the face. ISLM will never again be a vehicle for research. Its a cheap way of illustrating coordination failure via sticky wages/prices. But we don't know if that is the real coordination problem, so policy based on that assumption will be problematic.
2) Is he right to argue we need a large boost to government spending to close the estimated large output gap ?
"We may not be smarter, but we know a lot more than they did -- we have more data and better tools"ReplyDelete
Well Krugman would argue that we know less, because a lot of important stuff has been forgotten and is no longer being taught.
As for better tools, I assume you mean either mathematical techniques or computers. In both cases I would question if they are the right tools for the job. The mathematics is only necessary if you think a dynamic general equilibrium framework with rational agents yields a bettet insight into the economy. I'm not sure that it does. And when it comes to computers, it's a case of garbage in, garbage out. Computers can be useful tools, but I don't think they have revolutionized economics to date. When someone writes a program that can accutately predict the behavior of the economy, and if those predictions are different from what we would expect based on traditional macro, then I will concede that superior tools have invalidated the old theories.
"1) Is Krugman right to argue we should abandon DSGE methodology and go back to the Hicksian ISLM ?"ReplyDelete
I don't think he wants to abandon DSGE, he just wants to acknowledge the limitations of that apprach whilst rediscovering some of the insights of IS/LM, rather than focussing on the flaws in IS/LM and treating DSGE as scientific truth.
krugmans latest paper on the great recession uses a modern (small scale) dsge model. so its hard to arge he wants to go back to is/lm. i think the only point hes making is that you can learn more from is/lm than the blabber from john cochrane or the rbc crowd. better stil to read new keynesian stuff.ReplyDelete
Anonymous 8:15, you probably shouldn't be forming strong opinions based on what you read in blogs. While blogs are nice, they complement actual study in economics, not substitute it. If you really want to know what modern macroeconomics is you need to spend some time learning the tools and reading the literature and talking to macroeconomists. It is unfortunate that this takes long years of dedication to the subject, but then again "there is no Royal Road to geometry".ReplyDelete
Anonymous said "As for better tools, I assume you mean either mathematical techniques or computers. In both cases I would question if they are the right tools for the job."ReplyDelete
As a mathematician, I would say they are exactly the right tools. However, these tools could be used in new and non-traditional ways.
In my opinion macroeconomists should reexamine the foundations of their models Exactly how is aggregation justified? How is using a representative agent justified? The idea that ordinary people maximize the expected present value of utility over an infinite time horizon, and solve a Bellman equation to do so, strikes me as unbelievable.
Steve, I sent you an email with an agent-based model of a service economy driven by random transactions. Did you ever look at it? The model included fiscal and monetary policy, default rates, propensities to buy and sell. Rather than writing equations, the model is embodied as a computer program that allows the user to experiment with policy changes and chart the economic response. Inspiration came from Krugman’s discussions of the baby-sitting co-op model.
If Krugman is an "embarrassment" to the profession, the profession itself is an embarrassment.ReplyDelete
He is rated by many economists as their favorite economist (yes one survey isn't definitive but it captures the zeitgeist). Now clearly if all he does is spew nonsense twice a week for the last 10 years, he wouldn't be invited to keynotes and be rated highly.
The more plausible explanation is simply that Steve gains no traction for his ideas (even though some should be more prominent in the market place of ideas imho). He seems to be reacting badly by lashing out at others and hiding behind the veil of being provocative for stimulating discussions. Personally, I don't mind if he lashes out at Krumgan who does the same to others.
And please drop the, "I just wanna do science" shtick. You can't do science when you have a lab as unpredictable as human affairs. Outside the bubble, nobody considers you guys scientists. Smart academics doing difficult work yes. Scientists, no.
Ricardo Caballero from MIT - who might just be smarter than both Krugman and Williamson and even de Long and Andolfatto (gasps from the crowd) says:ReplyDelete
"What does concern me about my discipline, however, is that its current core
—by which I mainly mean the so-called dynamic stochastic general equilibrium
approach—has become so mesmerized with its own internal logic that it has begun
to confuse the precision it has achieved about its own world with the precision
that it has about the real one."
I'm sorry to say that, most of the criticisms above are just hitting the straw man, and too boring to reply. They are just stereotypes of Krugman as the speaker of reality, and non-Krugman (maybe rbc crowd, Lucas or Williamson) as bad guys playing around mathematics tool on nonsense modelsReplyDelete
Go read any rbc paper: their goal is to have better models to replicate data, maybe unemployment volatility, income distribution, or lifetime consumption/ saving. Any of these is nonsense or just showing off mathematic technique? If you have no idea of what rbc is doing, congratulation, you are hitting squarely the straw man!
To anon promoting econjwatch.org:ReplyDelete
Dictator Stalin voted third most popular Russian, Reuters
who cares this kind of poll?
Interesting that SW has no comment on PK's substantive comments refuting Cochrane, etc. but has lots to say about Also chooses not to answe the comments of some of the posters above on substantive issues. his comments about central bank operations, etc. SW would make a fine politician - he chooses not to answer difficult questions or not address their true intent, all with a straight face.ReplyDelete
"And Krugman is never right, only arrogant. At least Prescott is right some of the time."ReplyDelete
Keep in mind, Stephen too, that Krugman writes about a lot more than just macro (and I'm not convinced he's wrong on macro). He writes a lot on healthcare (where Stephen seems to agree with him as a supporter of a large government role), trade, political economy, taxes, and more that people who disagree with him on macro might agree with, and find useful.
"I'm not convinced he's wrong on macro"ReplyDelete
People who work on macro universally consider him wrong. Their opinion matters much more than yours.
Anyone that ever took a macroeconomics phd class knows it is a joke. Maybe the reason you are so aggressive to people that criticize it is because deep inside you know that and you are afraid of losing your job hahah.ReplyDelete
Most models don't even have (place for) money, those who do don't have unemployment, or credit, or government, rarely you see a model of open economy. These models consistently fail to adhere to data (even when "properly" calibrated) and the "advances" are nothing more than ways to distort the model so it fits better, with hardly any commitment to the real world or theoretical common sense (see, e.g. Jermann & Quadrini (AER t.b.r.)) where they "explain" the effect of the crisis on job losses by the fall in probability of lenders to take the capital from bankrupt companies (in a model without bankruptcy).
The truth is that at best case scenario DSGE models are good at predicting variables (but hardly better than pure econometric models), but a Phd macro student hardly understand better how the world economy functions and why things happen the way they do than an economics undergrad.
Really, anyone who takes a PhD macro class knows it is a joke? I just finished the first year of a PhD program and have my macro qualifying exam in one week. I came into the PhD program wanting to do micro and industrial organization and having a strong dislike for macro (based on my undergrad courses where we learned IS/LM among other things). But now, having finished the first year, macro seems to be the most mathematically rigorous, interesting, and seemingly realistic of the 3 sequences (micro, macro, and econometrics).ReplyDelete
Most of the people who criticize macro either cannot cope with the math, are not trained economists, or are not up to date on the literature. If you can handle the math and rigor of the current cutting edge macro research, you would be less likely to criticize its robustness.
There do seem to be some important features of our economy not yet incorporated into the DSGE models but that means there is plenty of room for innovative research. For these supposed PhD students who say macro is a joke, feel free to provide a better framework, it would make a great dissertation and should be pretty easy since you think the current models are so wrong.
“People who work on macro universally consider him wrong.”ReplyDelete
I’d like to see the survey. Woodford seems to largely agree with him; so does Blanchard, and in 2003 six other Nobel Prize winning economists, along with 450 others signed a letter against the 2003 Bush tax cuts stating, "To be effective, a stimulus plan should rely on immediate but temporary spending and tax measures to expand demand, and it should also rely on immediate but temporary incentives for investment.", ( http://berkeley.edu/news/media/releases/2003/02/nyt_economists.pdf ). That sounds pretty Krugman/Keynesian to me. Look through the signatures; I bet you’ll find some macroeconomists at top universities.
Keep in mind too, if your definition of macroeconomist is someone who has made a career in DSGEs, then it shouldn’t be too surprising that the vast majority of them consider wrong someone who’s a strong critic of DSGEs.
This blog has ruined my life. I will now proceed to write 5,000 words explaining why I do not understand opportunity cost.ReplyDelete
Williamson writes Krugman is never right on macro. Serlin returns serve with I'm not convinced.ReplyDelete
So: what metric are these guys proposing to test if Krugman is "right" ?
I would think that modern macro would say you should propose a macro model, and test its predictions against the data on multiple dimensions. I know what Prescott's model is, and what it predicts for many variables: what is Krugman's model and how well does it do compared to Prescott's ? Serlin ? Serlin ? (in my best Ferris Bueller voice....)
“For these supposed PhD students who say macro is a joke, feel free to provide a better framework”ReplyDelete
You have to realize it’s not just the model, it’s the, hopefully intelligent, interpretation of the model. You could have a great model, but a terrible interpretation of the model to reality, like a completely literal interpretation which acts as though reality is exactly equivalent to the model, all people really do know all publicly available information, they all really do have complete expertise in every area, and have all the time in the world to utilize that expertise to do intricate analyses of everything, and then have perfect self-discipline to apply those analyses, thus we can expect perfect expectations effects like exist in the model…
An intelligent analysis would attempt to modify what the model spits out given the unrealisticness of the assumptions, to interpret it in light of that.
Richard it is obvious that you just don't like that the DSGE model provides intuitive results counter to your dogma. Thus, you create these fanatical interpretations of a model you don't even know, clearly based on statements like "[agents] really do have complete expertise in every area" and "[agents] have all the time in the world to utilize that expertise to do intricate analyses of everything."ReplyDelete
What are you talking about? People should be required to learn the DSGE model if they are going to critique it.
1- I do not know of any respected macroeconomist who has the same naive interpretation of their models that you just described. Seriously, what was the last time you saw a talk by a macroeconomist? Go to talks and speak with these people then come back and tell us about their interpretations.
You have shown time and again in this blog that you do not have a clear understanding of these models, how to interpret them and their scope and limitations. Which is fine. I don't understand many things about many fields either. The difference is that I know that I don't know.
2- EVERY model is, by definition, an abstraction of reality. Models are not realistic and they are not supposed to be. Get over it.
3- It takes a model to beat a model. You don't like the implications of models in macro? Fine! What is your model?
one of many anon readers
Alas poor Richard. He joins the list of people like Greg Ransom and Noah something or other who come to this blog clueless about modern macro, but certain they know what's wrong with it, and get their asses handed to them on a plate. They dont learn. This isn't like the Thoma blog...it's not a community of kumbaya-singing Keynesians who know government can fix all our oowies, and its not the blog of his Royal Smugness deLong where nothing interesting has apparently happened in macro since Hicks published in 1937.ReplyDelete
Go read Thomas Sargent's interview in The Region to get a clue as to what's going on in macro
yes it is boring to repeat and reply those criticism. But it is interesting to understand why some people love to keep criticizing along those lines: math is bad (so we need to go back to old ancient writing), rationality is not a reliable assumption (so we need to assume people crazy), etcReplyDelete
Krugman is so smart that he knows by heart what his audience love to listen (and willing to pay $80k for that). For me just too stupid to understand, that why I am no hope to win a Nobel.......
A lot of the comments defending modern macro don't seem to understand the criticism. It is not that using mathematics is bad or that making unrealistic assumptions like rationality implies the model is wrong. It is that the traditional macro model seems to offer an adequate, if not superior description of reality. The onus is thus on the DGSE people to explain why their counter-intuitive interpretation of reality is the correct one.ReplyDelete
When they bother to do so, the argument seems to be that:
1) their model uses advanced mathematics and is therefore superior (and anyone who doubts the usefulness of such techniques obviously doesn't understand them, even if like Keynes, they are highly competent mathematicians).
2) their model is more logically consistent and more consistent with microeconomics (a claim which is irrelevant if this produces results contrary to reality).
3) there is a certain interpretation of statistical evidence that validates their theory (as if this is on a par with experimental evidence in the physical sciences).
Now, these arguments are just not good enough, so if you are so confident your view of the economy is the right one you should be able to come up with better reasons why you are right and the critics are wrong.
These debates keep coming up and are remarkably circular. The gist seems to be one side says "modern macro models are too stylized to be helpful in understanding the world, my old model was more intuitive and realistic and I still believe it" which elicits the response of a) "you don't know the modern models or the math", b) "our models are based on optimization behavior and are therefore superior" and c) "our models statistically fit the world better".ReplyDelete
In some sense, the traditionalists are right: most modern macro models are still extremely stylized. The need to formally model all decisions in the model means many important elements are systematically stripped out. Most likely, many of these elements are quantitatively important. Trying to build bigger models, a la Smets and Wouters, is also problematic given the vast number of parameters and shocks involved, along with the number of implicit modeling choices made.
Serlin also has a valid point when he attacks the all-knowing agents of modern macro models. Many results in modern macro hinge on this assumption. For example, the amazing ability of price-level targeting to fix everything in NK models depends on agents' ability to understand what PLT means for future dynamics in a way that stretches credibility. Of course, this point is well-recognized in modern macro and there's a lot of work on incorporating information constraints in models (Woodford, Reis, etc...) and on quantifying the empirical size of these rigidities (Gorodnichenko). But at the end of the day, most models don't incorporate these elements and likely won't for a long time. So we'll continue to have the same arguments on this and many other missing components of macro models.
More broadly though, I think a lot of it has to do with people's perception of the objective function. For traditionalists, it seems clear that if output is well below trend levels, then policy needs to work to offset that gap in whatever way necessary. hence the calls for QE3, higher inflation targets, new stimulus etc... Those working with modern macro models tend to have a welfare criterion in mind, where there generally is not such a mechanical relationship between policies that close a traditional measure of the output gap and those that enhance welfare.
Of course, welfare is what we care about at the end of the day, so the modern macro view has the right objective. But given the shortcuts taken by modern macro models to make them tractable, how reliable are these welfare results? I don't know, but even as someone who works actively with these models, I still put significant weight on my priors when evaluating policy implications of my and other models. Indeed, I suspect that this is an optimal response to the lack of clear evidence on the relative validity of different macro models. Most macroeconomists have strong priors which evolve only gradually in light of a preponderance of evidence. Take SW, when was the last time that research meaningfully changed your prior about something policy-related?
"when was the last time that research meaningfully changed your prior about something policy-related?"ReplyDelete
I've changed my thoughts about capital taxation -- it now seems clear that the zero-tax case is true only in a (nearly) measure zero set of possible economies, those that do not have binding incentive-compatibility constraints (due to either limited commitment or asymmetric information). Capital taxation appears to be necessary for efficiency; I don't care about equality.
And Serlin is hopeless guys, there is no point in engaging him.
"It is that the traditional macro model seems to offer an adequate, if not superior description of reality."ReplyDelete
Based on what? Statistical fit? Whoopee, Smets-Wouters just shot your argument to shit. It fits quite well, and even forecasts better than a VAR (corrected statistically for the number of parameters). Go back and try to find another argument for why we should ignore the tautological point that aggregate C must be equal to individual c added up across people.
"Based on what? Statistical fit? Whoopee, Smets-Wouters just shot your argument to shit. It fits quite well, and even forecasts better than a VAR (corrected statistically for the number of parameters)"ReplyDelete
No, based on observing the real world. One of the deficiences of modern macro is the use of statistical techiniques to validate the model, as if they prove it to be true. I don't find such evidence any more convincing than the Phillips curve. Econometrics and calibration are not a substitute for controlled experiment. In the absence of experimental data we have to make predictions about the real world to judge theories. Eg Friedman predicted the Phillips curve would break down.
"One of the deficiences of modern macro is the use of statistical techiniques to validate the model, as if they prove it to be true."ReplyDelete
Nobody thinks models are true. Didn't you learn anything in your classes, assuming you actually took any? And what are predictions? They are forecasts, perhaps forecasts conditioned on policy changes, but forecasts nevertheless.
The assertion was that the "old timey" model represents the real world better. I asked what the standard for this assertion was. If it is just statistical fit, then the assertion is false; that does not imply that I think statistical fit is the appropriate standard. But if you're going to claim something, you better make sure you're right.
Anon 7:03 AM wrote, "It fits quite well, and even forecasts better than a VAR (corrected statistically for the number of parameters)."ReplyDelete
One doesn't correct forecasts for numbers of parameters. One does correct in-sample fit for numbers of parameters -- see your favorite information criterion.
But out-of-sample forecast statistics are sufficient statistics for forecasting. That is, model complexity doesn't matter, conditional on out-of-sample forecast stats.
Can you cite the Smets-Wouters paper that you are talking about?
VARs don't necessarily give good forecasts. Even if Smets-Wouters beats VARs, I strongly doubt that it beats professional forecasters.
What do they use, if not VARs? Dartboards?
Chris is right, the consensus forecast of professional forecasters is typically much better than any time series model. See the JME piece by Ang, Bekaert and Wei (2007). Why do they do better? One reason might be the notorious "add" factor: they run their econometric model then adjust the forecasts to reflect what they consider to be special factors or conditions not included in the model. Because, as you may recall, the real world is sometimes a tad more complicated than our models and, on occasion, this matters.ReplyDelete
Anon 12:10 asked, "What do they use, if not VARs? Dartboards?"ReplyDelete
I am not a professional forecaster but my understanding is that forecasters use equations tailored to each of the sectors of the economy that they are trying to forecast.
Generally, forecasters use a lot more data and information than can be fit into a VAR and they use it judiciously, with a lot of structure (restrictions) and judgement.
As "C" correctly notes, pros then add various adjustments to their results or methods, depending on the circumstances. For example, they know when Japanese industrial capacity will come back online in various sectors. They know what the weather was and when it is important. They know when there has been fighting in an oil exporter or a surprising election result in a major country. All these things are important but not easily formally incorporated into models.
VAR /= statistical modeling.
VARs don't forecast well. That is why people like Bob Litterman popularized Bayesian VARs, which do a bit better but still don't do nearly as well as a good model run by pros.
They only use the dartboards for forecasting exchange rates.
I don’t have any dogma. I just want to better understand the point of view of Stephen, people who agree with him on macro, and his criticisms of Krugman.ReplyDelete
I’m also not saying everyone interprets modern macro models to reality literally, or overly literally. I’m just saying this could be a problem with at least some given some of the things I’ve read.
I have a background in economics far greater than the average well educated person (BA in econ, MBA, ABD in finance). It would be nice if these points of view, things like we should be raising rates now, not doing QE, could at least be explained clearly and intuitively to me, let alone well-educated laypeople with no degree in economics.
I have not looked at this in a few days. Excellent discussion here, people. I have just one thing to add, which is for Richard: I don't think you are really interested in understanding. Based on about a year of reading your comments, I think you're an immovable object.ReplyDelete
Anonymous 7:03am said: "Go back and try to find another argument for why we should ignore the tautological point that aggregate C must be equal to individual c added up across people."ReplyDelete
In his "Some International Evidence on Output-Inflation Tradeoffs" 1973 AER paper, Lucas had some difficulty with aggregation issues. In that paper, 'y_t(z)' is the log of output in market 'z,' and 'y_t' is the log of aggregate output. For some reason, Lucas and his reviewers (and many who subsequently taught/cited/etc this paper) failed to recognize that the sum of logs is not equal to the log of a sum.
"I think you're an immovable object."ReplyDelete
Just not true. You just haven't persuaded me of much. But I've changed or modified my views many times as I've learned economics. At first I was a lot more pro-market as a young'n. Then I learned in classes about externalities, income and substitution effects, backward bending labor supply curve, natural monopoly, etc., and I've made a lot of other modifications as I've learned more.
I also bring up a lot of criticisms to see what your response is, to test your arguments, and to learn.
Perhaps the only way understand your arguments is to spend years full time studying your macro, which means just about the entire general public won't ever understand it, but at least for lots of other economics, and science and technology, the basic ideas and intuitions can be explained to well educated laypeople.
If you want to convince people not in your specialty you have to be able to defend your ideas against criticisms and explain them in a clear intuitive way.
Partly, though, I do feel like a student who asks questions without making the effort to first do the reading. But, it's that or nothing. I have vanishing little time these days I can spend on this, but still a frustrating desire to understand why you say a lot of the things you do.ReplyDelete
"I don't think you are really interested in understanding. Based on about a year of reading your comments, I think you're an immovable object."ReplyDelete
Steven, comments like this are why I say you are smug and patronizing. As far as I have seen Richard has made entirely reasonable points and has simply found your responses unconvincing. You seem to have the attitude that what you know is fact and anyone who disagrees needs to be educated. Maybe that was how you were taught economics, I don't know.
I think the advantage of Krugman's blog is that he is consciously directing his arguments at the layman, which forces him to actually explain his point of view in a way that makes sense to people in terms of the real world. Most of your explanations seem to rely on arguments from authority (Lucas said it, so it must be true!) or implying that because your views are based on "technically rigorous" research it is beyond the wit of people who haven't got a PhD in economics. Well, I do have a PhD in economics, and I'm not convinced. Maybe you could try making an argument that doesn't assume everyone but you is ignorant. Just a thought. Also, just maybe people like Richard and Krugman have something to teach you about economics if you give them a fair hearing.
I do admit to aiming higher than Krugman, as I want to speak to people who know some economics, or a lot, in addition to people who know very little of the field. I do of course make arguments based on other people's ideas, but I also think for myself, and don't blindly follow Lucas or anyone else. The problem with Krugman, from my point of view is that what he writes appears to make sense, even to you, who should know better. What he's selling is politics masquerading as economics, and I think he is actually harming the science.ReplyDelete
I do admit to aiming higher than Krugman ....ReplyDelete
My, don't we fancy ourselves?
Krugman writes in a way that enables people with very different backgrounds to understand him. Some, like me, have studied economics and then gone on to work for a living. Having retired I’m interested in knowing what progress has been made in the field. Lots of economists have things to tell me, but Krugman is by far the best communicator.
Other Krugman readers, like Matthew Yglesias, have never taken courses in the subject, but are quite well-read. He can see that a lot of what purports to be economic science is actually philosophy – and sloppy philosophy at that. I mention him because I noticed that you were sneering at him recently.
Actually, if you really do think that Krugman is “harming the science” you might get somewhere by arguing that case on a level that makes sense to Yglesias. But to do so you will have to address questions like: What makes you think economics is a science? How should this ‘science’ settle empirical questions? In what way is Krugman harming that process?
Being a fairly regular reader I don’t really expect you to rise to the challenge. You’ll probably say that you have better things to do and resume your diatribes, which are easy to write. But I’ll be delighted if I’m wrong.
I’m at about zero time I can spend on this, so I may comment very little for a long while, if at all. But I’ll leave you with this.ReplyDelete
I don’t think I disagree with you on most things, especially when I can get at and understand what you really think. I know little about DSGEs, but I think they are probably useful when interpreted well. And I think that even Krugman actually agrees with you on many things, if you can really get down to understanding each other, especially with things outside of macro, like healthcare, long term growth, taxes.
But here are some of the big specific things you have not been able to really persuade me of, or make me understand your point of view:
1) That we should raise rates now, that higher inflation wouldn’t benefit the economy by repairing balance sheets, allowing real short term rates to be cut, and lowering sticky real wages.
2) That QE cannot work to lower longer rates no matter how large. If the government can raise the price of cars by printing $200 billion and buying cars, doubling demand if it wanted to, why can’t it print enough money and buy enough bonds to raise their prices (thus lowering their interest)? Seems like they’re shifting out the demand curve, which will raise price.
3) You say Japan’s lost decade or two is different from the US situation. I think it would be helpful to understand your economics if you wrote what you think Japan should have done or still should do.
4) As far as I can tell you think the US now should just wait this out, providing a good level of unemployment benefits to help people weather it, and try to work on getting the financial system in order and better regulated? Is that correct?
5) You seem really convinced the reasons for the current unemployment are predominantly, or completely, structural, but I haven’t seen you provide much evidence, or address the evidence that critics say refutes this view.
Keep in mind, I’m not saying I’m sure you’re wrong on these things, but I haven’t found convincing explanation and evidence that you’re right, and I’m not the only one. You haven’t been paying attention if you haven’t seen there’s lots of economists at top universities and with Nobel Prizes that need convincing and explaining too.
Kevin: "My, don't we fancy ourselves?"ReplyDelete
Steve "fancies himself" because he writes for people with a higher level of economic education than those for whom Krugman writes?
Kevin: "Krugman writes in a way that enables people with very different backgrounds to understand him."
Sometimes this is true. Most of the time, however, Paul writes sermons to please people who already agree with his politics.
Kevin: "Some, like me, have studied economics and then gone on to work for a living."
Congratulations. The rest of us are on the dole, I guess.
Kevin: "Other Krugman readers, like Matthew Yglesias, have never taken courses in the subject, but are quite well-read. He can see that a lot of what purports to be economic science is actually philosophy – and sloppy philosophy at that."
Yglesias has never studied economics but he can see right through the consensus of tens of thousands of people who have studied and tested their ideas against each other for many decades. How embarrassing for us.
Kevin: "Actually, if you really do think that Krugman is “harming the science” you might get somewhere by arguing that case on a level that makes sense to Yglesias."
Do you think that Yglesias is interested in learning about economics or having his pre-suppositions confirmed?
If it is the former, maybe he should actually take some classes.
If the latter, I doubt that anything that Steve has to say will convince him.
Kevin: "But to do so you will have to address questions like: What makes you think economics is a science?"
Let's go to the dictionary:
1.The intellectual and practical activity encompassing the systematic study of the structure and behavior of the physical and natural world through observation and experiment
- the world of science and technology
2.A particular area of this
- veterinary science
- the agricultural sciences
3.A systematically organized body of knowledge on a particular subject
- the science of criminology
4.Knowledge of any kind
Whatever the failings of economics -- and I am actually more of a critic than you would guess -- it is pretty clearly a science under definitions 3 and 4.
Kevin: "You’ll probably say that you have better things to do and resume your diatribes, which are easy to write. But I’ll be delighted if I’m wrong."
Steve probably does have better things to do, but my research is not as good so my opportunity cost is lower. If a response from Steve would "delight" you, I can hope that my response is at least satisfactory.
All the best, Chris
"Whatever the failings of economics -- and I am actually more of a critic than you would guess -- it is pretty clearly a science under definitions 3 and 4."ReplyDelete
The problem is most people understand science to mean definitions 1 and 2, and that is what most economists seem to imply when they talk about economics as a science. If economics is just science in the sense of systematized knowledge then it has no more authority than criminology, politics, or philosophy--areas where most people acknowledge that different views may be equally valid. If science is "knowledge of any kind", then economics is on a par with theology and literature.
This is why using advanced mathematics doesn't necessarily help make economics more rigorous. Unlike say theoretical physics, which is based on physical laws and uses mathematics as a tool to analyze complex phenomena, economics is based on assumptions which are flaky at best, such as rationality and utility maximization. Having a mathematical model based on the laws of gravity might be lead one to accept its implications. Building a logically consistent model that is based on incorrect assumptions does not imply that the model's implications are sound. Thus traditional macro is just as rigorous and scientific as modern macro.
Anon 4:03: "The problem is most people understand science to mean definitions 1 and 2, and that is what most economists seem to imply when they talk about economics as a science."ReplyDelete
I disagree. We all know that economists don't study the physical world. They study reolutions to scarcity, decision-making, markets, etc.
It is a given that using advanced mathematics doesn't necessarily make economics more rigorous. And I've long argued that economists often mistake insights about mathematical models for insights about the world. That is a problem. But I am not the only one who recognizes it. Go read Ricardo Caballero's recent interview in the MN Fed's journal. See the section on "pretense of knowledge." Caballero is Chair of the Econ Dept at MIT, so he knows economic research.
But math is necessary to be more rigorous than ordinary reasoning without math. Math formalizes arguments and makes clear their assumptions and limitations. It can also be used to quantify how important the effects are.
Economics is systematized knowledge.
What puzzles me is that some people know -- or should know -- that many, many thousands of people have spent their lives studying economics, yet they think that any economic knowledge must either be obvious to any literate person or false.
"I disagree. We all know that economists don't study the physical world."ReplyDelete
Well, OK, if you want to quibble replace physical/natural with human (although I'm pretty sure humans are part of the natural world).
The point is there's a big difference between "proper" science like physics or chemistry and pseudoscience like criminology, let alone "knowledge of any kind". Economics is often portrayed as being science in the sense of physics, and it really isn't.
And nobody is arguing against math--IS/LM uses math--I'm just making the point that more advanced math isn't necessary or better. What matters is whether the implications of the model match reality. Building a complex model on dubious foundations proves nothing. Nor do arguments from authority, i.e. invoking people who are experts in economics.
Disappointing that a Canadian should be as parochial as the Americans. Neither neoclassical or Krugman/Hicks Keynesianism has survived the Great Recession. The only school that correctly called it were the post-Keynesians of the Minsky/Keen and Lavoie-Godley variety. One would have thought that Williamson would be aware of his fellow Canadian Lavoie, but apparently not.ReplyDelete
I have dipped into Post-Keynesianism once in a while, and I have never learned anything from it. In spite of spending a lot of time in Canada and among Canadian economists, I have never heard of or met Professor Lavoie. What's his take? I'm curious.ReplyDelete
The Godley Lavoie work appears to be an attempt to build on Tobin's famous 1969 JMCB paper on a general equilibrium monetary theory. They attempt to integrate flow-of-funds ideas into the income-expenditure ISLM framework. A model of economy-wide balance sheet transactions between broadly defined sectors. Ad hoc microfoundations are used much as Keynesian-stye models have always done.ReplyDelete
Anon 11:59: "The point is there's a big difference between "proper" science like physics or chemistry and pseudoscience like criminology, let alone "knowledge of any kind"."ReplyDelete
Now, you are moving the goalposts again.
The first question--posed by Mr. Donoghue-- was "What makes you think economics is a science?"
We went to the dictionary and it certainly meets two of the four definitions.
Then you said that definition doesn't count because people don't know what science actually means. So, you get to think that it isn't a science because you don't like the definition of science. OK. Fine.
Now you say "there's a big difference between "proper" science like physics... and pseudoscience like criminology, let alone "knowledge of any kind"."
But you don't differentiate physics from criminology. What makes them different?
Let me point out some ways that they are similar and one important way that they are different.
Similarity: Physics has some pretty exact models of simple phenomenon and a lot of harder questions that it can't answer. Criminology might have approximate answers to complex phenonmenon. They both attempt to systematically build knowledge, using mathematical models to do so.
Difference: Physics can use more experiments. Criminologists are much, much more limited in the types of experiments that they can do.
Anon 11: 59: "Economics is often portrayed as being science in the sense of physics, and it really isn't."
You need to define your terms.
Economics is a systematic attempt to evaluate evidence and build knowledge. It is not a study of "natural" or "physical" phenomenon in the usual sense.
If you don't like the progress that economics has made compared to physics, that is fine, but it like saying that golfers are better at golf than basketball players are at basketball. They are still both sports.
Anon 11:59: "I'm just making the point that more advanced math isn't necessary or better."
If I ever find someone who makes this claim, I will be sure to relate your skepticism along with my own.
Funny that you would think that one would be parochial for not being interested in what Godley Lavoie does. I read Tobin's 1969 paper in 1977, and found it very interesting. Then a few years later I learned some other things, and it became clear to me that I would not want to go back to Tobin's framework and use it for doing economics. Basically, I found some better tools. Last week I bought a new turntable, and I love it. It's an old technology, but for some things I think analog sounds better than digital. I would not trade the computer I have now for the Toshiba 286 laptop I had in 1989, though. Maybe you think that is parochial.
Chris, it’s certainly possible that when Stephen Williamson charges Paul Krugman with “harming the science” he actually means that Krugman is doing damage of some sort to a “systematically organized body of knowledge” in some way or other. If that’s what he means, life’s too short for speculating about how one would prove such a vague charge.ReplyDelete
I don't know if you're being deliberately disingenuous, but I am not "moving the goalposts" at all. Since the Scientific Revolution, the term science has been understood as a special type of knowledge about the world, based on empirical evidence and subject to test via controlled experiment. It is this sense that essentially corresponds to dictionary definition 1. There is an older definition, where science implies a sort of reliable knowledge on a particular subject, and was essentially interchangeable with philosophy. Indeed, the word science is based on the latin word for knowledge, as opposed to art which means skill.
Now, when people like Steven talk about economics as science, it's pretty clear they mean it in the former sense, as a sort of dispassionate enterprise of collecting empirical evidence and formulating and testing hypotheses. The difference between this type of science and organized systems of knowledge (such as criminology, politics, sociology, or astrology) is that the former is regarded as much more objective and reliable. It is considered gauche to question established laws of science and even to question less reliable claims made by scientists (e.g., climate change). The idea is that the scientific method leads us to an undeniable truth about the world, leaving little room for debate. By contrast, everybody fells entitled to their own opinion on subjects like criminology, even though some opinions are clearly better informed than others. I'll continue below since this comment is getting kinda long.
"Difference: Physics can use more experiments. Criminologists are much, much more limited in the types of experiments that they can do."ReplyDelete
My point is that experiment is not an optional extra when it comes to science, it's a crucial ingredient. In the words of Richard Feynman:
The principle of science, the definition, almost, is the following: The test of all knowledge is experiment. Experiment is the sole judge of scientific 'truth'.
The difference between natural science and economics/criminology is that theories in natural science are based on reliable data about measurable quantities (e.g., voltage, gravity) and are tested via repeatable, controlled experiment. This is the source of the confidence to which society places in scientific explanation and prediction. By contrast, economic theories are based on questionable evidence, often based on unobservable quantities (e.g., utility) and are typically tested only using one-off, uncontrolled experiments with the aid of statistical techniques like econometrics. This is why economic theories are less reliable than those in natural science, and why different economists disagree in their explanations of economic phenomena.
"If I ever find someone who makes this claim, I will be sure to relate your skepticism along with my own."
You could start with Stephen Williamson and his fellow travelers. He repeatedly criticizes traditional macro on the basis that modern macro is more rigorous/sophisticated or uses better tools, by which he seems to mean it uses advanced math.
"...Krugman is doing damage of some sort to a “systematically organized body of knowledge” in some way or other. If that’s what he means, life’s too short for speculating about how one would prove such a vague charge."ReplyDelete
That does not need "proof," as it's painfully obvious. He's on a mission to destroy the credibility of a branch of the economics profession. Of course, this has its funny side too, as Krugman is quite confused about what it is he is up against. The evil beast in his mind bears no resemblance to what macroeconomists do today, or even to what they did in 1975. See this:
On another matter, it's wrong to hold up the natural sciences as some kind of ideal. I was just reading an article in the NYT (Sunday edition I think, I could not find the reference) about how researchers have a hard time publishing replications of experiments that come up with different results. For example, in physics, it's not like there is a theory, the experiment is run, the experiment confirms the theory, and we are done. Experiments are only as good as the human beings designing them. Experiments can be badly designed, experimenters can distort results, and some even make up the data. In experimentation with new drugs, the profit motive can get in the way of good science. Thus, some of the problems you find in economics are not unique to us. You are not going to find purity in the natural sciences.
"He's on a mission to destroy the credibility of a branch of the economics profession."ReplyDelete
If you mean the Cochrane-Fama-Trichet branch I've no doubt about that. However that's not an attack on a body of knowledge. If you think he has it in for the New Monetarists you are surely mistaken.
"He's on a mission to destroy the credibility of a branch of the economics profession."ReplyDelete
No, he thinks a large part of the profession is wrong and is on a mission to make macro more credible by making it more plausible. If anyone is persuing economic science, it's Krugman.
"Experiments are only as good as the human beings designing them. Experiments can be badly designed, experimenters can distort results, and some even make up the data. In experimentation with new drugs, the profit motive can get in the way of good science."
Good point. Why even bother having experiments?
Anon 3:45: "My point is that experiment is not an optional extra when it comes to science, it's a crucial ingredient."ReplyDelete
I am a big fan of Richard Feynman. Very interesting guy. Obviously, he was a brilliant physicist. But you are taking his words out of context.
First, Feyman was talking about the sciences that study the natural world. He was not using the broader definition of meanings 3 and 4. As I have said before, I don't think that economics is the study of the natural world.
Second, there are lots of types of experiments. There are double-blind, controlled studies, such as those used in physics and medicine. Economists actually do some of these, in experimental laboratories. I've actually been a research assistant -- in a past life -- on such projects.
But there are other ways to experiment. There are natural experiments, in which one takes advantage of unusual changes in an environment to draw conclusions about causality. For example, labor economists have used the Vietnam-era draft lottery to study the effects of college education and military service.
There are also experiments that tease out causation and effects from data. These fall under the topic of statistics and econometrics.
In the present context, are we to believe that we cannot hope learn about a topic without being able to conduct controlled experiments? Really? We cannot learn anything?
That response seems more than a bit nihilistic to me.
Anon 3:45: "[Steve] repeatedly criticizes traditional macro on the basis that modern macro is more rigorous/sophisticated or uses better tools, by which he seems to mean it uses advanced math."
No, he doesn't. Perhaps I should let Steve defend himself but I think that he means that modern macro is better because it works from the ground up, starting with the decisions of individuals and firms, taking expectations more seriously and assuming that people are not systematically stupid.
Older macro used more questionable assumptions, such as the idea that people simply consumed a constant proportion of their income or could be easily and systematically fooled by policy choices.
I doubt that Steve is claiming that more complex math makes it better.
(But it is true that the math keeps getting more complex.)
Anon 3:45: "This is why economic theories are less reliable than those in natural science, and why different economists disagree in their explanations of economic phenomena."ReplyDelete
Yes, economists do disagree in their explanations of many economic phenomena. But they also agree -- to a very wide extent -- about other economic phenomena.
Physicists agree about some things and disagree about more complex things.
Do we know more about both physics and economics than we did 50 or 100 years ago? Yes.
Do we know as much as we'd like to know? No, and we never will.
To say that "economic theories are less reliable" is like saying that golfers are better at golf -- because they hit the ball every time -- than baseball players are at baseball. (Even the best baseball players miss the ball most of the time.)
It just doesn't make any sense to compare them.
"Feyman was talking about the sciences that study the natural world. He was not using the broader definition of meanings 3 and 4. As I have said before, I don't think that economics is the study of the natural world."ReplyDelete
The point is that Feynman was talking about science in the proper sense that economics aspires to. Are you seriouly saying that economics is science in the same way as criminology and astrology? I don't think studying the natural word is the key difference between economics and physics. Rather, it is the lack of proper scientific method in economics.
"There are natural experiments, in which one takes advantage of unusual changes in an environment to draw conclusions about causality."
I'm aware of this approach and alluded to it in my previous comment. My point is that such "experiments" are no substitute for controlled, reproducible experiment. Natural experiments are rarely decisive in the way controlled experiments can be.
"There are also experiments that tease out causation and effects from data. These fall under the topic of statistics and econometrics."
Again, I mentioned such methods before. Econometrics is very clever but rarely decisive. There's invariably some reason why results may be misleading.
"are we to believe that we cannot hope learn about a topic without being able to conduct controlled experiments? Really? We cannot learn anything?"
You can learn things without elavating them to the status of science. Utility is an interesting concept but it is not scientific. We can learn Ricardian equivalence but it doesn't make it fact.
"modern macro is better because it works from the ground up, starting with the decisions of individuals and firms, taking expectations more seriously and assuming that people are not systematically stupid."
It would be better if we knew how individuals make decisions, but we do not.
"I doubt that Steve is claiming that more complex math makes it better."
He seems to imply this with all the talk about "tools", but maybe I've misunderstood him.
"To say that "economic theories are less reliable" is like saying that golfers are better at golf -- because they hit the ball every time -- than baseball players are at baseball. (Even the best baseball players miss the ball most of the time.)"ReplyDelete
If economics doesn't strive to be regarded as science, you are right. But I think it does, in which case both econimist and physicists are playing the same game, but the economists are losing, sadly.
"However that's not an attack on a body of knowledge. If you think he has it in for the New Monetarists you are surely mistaken."ReplyDelete
Yes, I don't think he knows we exist. However, he definitely has it in for the older generation of macroeconomists, particularly Lucas and Prescott. It's gotten worse over time to the point where the critique (i.e. "dark age of macro," etc.) becomes anti-math and anti-optimization.
"If anyone is persuing economic science, it's Krugman."
I guess you think science is done in blog pieces. This is journalism, not science.
anon @3:45 am. I suppose that you do not consider astrophysics or paleontology a science since you can't do controlled and repeatable experiments there as well. Correct?ReplyDelete
"He repeatedly criticizes traditional macro on the basis that modern macro is more rigorous/sophisticated or uses better tools, by which he seems to mean it uses advanced math."ReplyDelete
There is nothing very advanced about the math used in most graduate economics programs in the United States. The math you need to get through is considerably less than what an undergraduate math major knows. When I say we have better tools, it's not just the math, it's the whole modeling approach, principally the idea that you want to put together analytical tools that allow you to analyze policy correctly.
"I suppose that you do not consider astrophysics or paleontology a science since you can't do controlled and repeatable experiments there as well. Correct?"ReplyDelete
Astrophysics is part of physics, so is clearly science. Plus, observations using telescopes are quite close to controlled experiments. I do not consider palentology science any more than archeology is.
"There is nothing very advanced about the math used in most graduate economics programs in the United States"ReplyDelete
So you don't think there's a big dIfference between the math required for modern macro vs. traditional macro?
"When I say we have better tools, it's not just the math, it's the whole modeling approach, principally the idea that you want to put together analytical tools that allow you to analyze policy correctly"
I'm not sure what this is supposed to mean. I guess you're talking about the Lucas critique, but that doesn't require different tools. It's more of a caveat than a methodology.
"Astrophysics is part of physics, so is clearly science." By the same token, economics is a part of a bigger field called "social sciences" that can sometimes use controlled experiments, e.g. in psychology. Hence it clearly makes it a science too, according to your logic.ReplyDelete
"observations using telescopes are quite close to controlled experiments."
How is observing a universe using telescopes any closer to a controlled experiment than observing an economy? In both cases you cannot control their conditions and cannot repeat them. (Unless you know how to create and recreate e.g. a particular supernova explosion all over again. I'd be curious to know....)
"So you don't think there's a big dIfference between the math required for modern macro vs. traditional macro?"ReplyDelete
When I was an undergraduate, someone taught me macro using Branson's book. I was supposed to be in a "math section" of intermediate macro, which seemed to require knowing calculus and algebra at the introductory level. Cramer's rule was very helpful. Later on, I learned some statistics, probability theory, and time series analysis, and that was helpful in econometrics class. The revolution in macro required knowing some dynamic optimization, but I had some of that anyway from the math department and from math econ. With some exposure to serious microeconomics (some GE helps) and a little bit of dynamic programming and continuous time optimization, getting used to modern macro was not so hard. Relative to old-time macro it's just the dynamics, but that's critical.
I wouldn't call the Lucas critique a caveat. That sounds like it's on the order of checking for typos, which is way off.
Anon 9:42 AM "I don't think studying the natural word is the key difference between economics and physics. Rather, it is the lack of proper scientific method in economics."ReplyDelete
You have not described any firm difference between physics and economics except the subject matter.
You found a quote from Richard Feynman saying that experiments are crucial to science. But experiments are not just controlled experiments, they include many methods, as some of the commenters have pointed out.
Physicists do experiments. So do economists. So do geologists. So do psychologists.
Not every idea in economics can be tested dispositively with experiments, controlled or otherwise. Not every theory in physics can be tested with experiments.
You say that economics "aspires" to be like physics. But it doesn't. It is economics. Who says otherwise?
Economics doesn't "aspire" to be science. It is science.
You say that economics lacks "scientific method." What does that mean?
I don't intend to be obnoxious but what is the basis for your opinion about methods of economic research? Are you an economist? A grad student?
(I think that I will bow out of this discussion.)
"By the same token, economics is a part of a bigger field called "social sciences" that can sometimes use controlled experiments, e.g. in psychology. Hence it clearly makes it a science too, according to your logic."ReplyDelete
No, I do not accept any social science uses serious controlled experiments. Psychology is largely hokum. Economics is by far the most credible of the social sciences.
"How is observing a universe using telescopes any closer to a controlled experiment than observing an economy? In both cases you cannot control their conditions and cannot repeat them."
I think because at the level of the universe it's fairly clear what your observing, whereas in economics everything is interelated. Plus, different people can take independent observations of the universe (rathet than relying on official data) and the same phenoma happen repeatedly, so you can observe several supernovas, for example. It's not as good as experiment, but it's pretty rigorous. Plus, it all fits in with experiments in smaller scale physics.
So, to summarize your views, a science is anything that you consider a science. Controlled and repeatable experiments are necessary if you want to discredit economics, but are not required for astronomy because there is something else that is "pretty rigorous". Sometimes they do not count because you think the experiments are "hokum". No wonder that those fools here that stick to lousy dictionary definitions cannot understand you.ReplyDelete
"You have not described any firm difference between physics and economics except the subject matter. "ReplyDelete
Yes I have. Physics consits on laws of nature based on experimental evidence which leaves very little room for doubt. Economics consits of informed opinion which can't be rigorously tested.
"You say that economics "aspires" to be like physics. But it doesn't. It is economics. Who says otherwise?
Economics doesn't "aspire" to be science. It is science. "
No, I have established that economics is not science in the true sense. Many economics texbooks and economists like Steven like to frame economics as science in the sense that economic theories are meant to be based on hard-headed examination of the evidence, so that economic analysis and advice is viewed as respectable and reliable.
"You say that economics lacks "scientific method." What does that mean?"
Scientists propose hypotheses based on empirical evidence and test them experimentally. If the experiments disprove the hypothesis, it is discarded. Economists base hypothesis on fantastical assumptions, then argue about the empirical evidence.
"I don't intend to be obnoxious but what is the basis for your opinion about methods of economic research? Are you an economist? A grad student?"
I fail to see how this is relevant. You may as well ask if I'm Jewish. The truth is the truth, no mattet who speaks it. Obviously I have some training in economics, and I am interested in science, economics and math.
"(I think that I will bow out of this discussion.)"
I guess I win then!
Anonymous at 3:01pmReplyDelete
Look 99% of people understand what science is, so I don't know why the concept is so difficult for the people who read this blog. In any case, I don't seek to discredit economics by pointing out the rather obvious fact that it is not science. History is not science, nor is law, nor is engineering, nor is math. All of these are important subjects, but they all have their limitations. If economists learn to accept the limitations of their subject then we might have more respectful debates about policy.
The doubts you express about the natural sciences by way of the NYT piece you reference border on the fatuous. Prior to reading it, did you think that no physicist or chemist made experimental errors (some accidental, and some by design)? Part of the "gold standard" in these fields is the replication of experimental results; in 1989, the cold fusion claims of Stanley Pons and Martin Fleischmann were quickly found to be terribly faulty through failure to replicate their results.
There is no such equivalent in economics, which is why, as per the sense in which anon 3:26am and many others define it, economics is not a science. But, so what? Is it all that important to claim it is? The power, relevance, rigor, and fun in doing it all still remain even if we agree to ditch the "science" label.
Yes, there is an equivalent: Every regression and every computer simulation should be replicable.ReplyDelete
And, anon 3:26am and anon 3:45am (probably the same person) have no coherent definition of a science, they keep changing it as they see fit.
Don't exaggerate. Of course it's all a matter of how we define science, which apparently we're having trouble with here. I just get tired of hearing, in particular from natural scientists, about their natural superiority. Dig into any field, and you will find disagreement, bad work, and cases where it is hard to resolve critical issues. Economics has some unique difficulties due to poor data, experimentation that is far too costly to carry out in most circumstances, and the fact that changes in policy change behavior. That does not mean that economics is not a science. There is theory. There is data. We use the former to make sense out of the latter. Science.
Anon 6:20pm wrote: "Every regression and every computer simulation should be replicable."ReplyDelete
Yes, but this is besides the point. Joe reports in the AER that beta-hat on 'X' is 0.45, and using his data for the US, others can also get this result. Using a longer US data set, Sally and her co-authors report in the QJE that beta-hat is 0.61, and this result can also be replicated by others using Sally's data. None of these acts of replication establishes that the population value of beta is either 0.45 or 0.61; and assume it turns out that the economic significance of the difference between these two values is non-trivial. If it makes you feel better, go ahead and call it "science." But it's a very different creature from what's done in the natural sciences; and hey, that's ok.
This is much like what happens in medical science, don't you think? You can get very different results across different clinical trials. Life is pretty difficult for those people too. Seems you're going to have a hard time drawing these sharp distinctions between us and what is done in other fields.ReplyDelete
Steve @8:16pm: I agree with your medical "science" analogy. Lots of more so pure biologists and chemists are pretty snooty towards medical research qua science (and not only because it is not necessary for every union member to hold a PhD). Ask your biologist & chemist friends about the prestige distinction between NSF and NIH funding.ReplyDelete
"anon 3:26am and anon 3:45am (probably the same person) have no coherent definition of a science, they keep changing it as they see fit."ReplyDelete
That's not true, I have the same definition of science as most people since the Scientific Revolution: A systematized body of knowledge about the world around us, based on empirical evidence and capable of making testable predictions about the phenomena it purports to explain. As Feynman points out, it is crucial that scientific theories are subject to controlled experiments so that false hypotheses may be refuted.
Now, there are some areas of science that match this definition better than others, but that doesn't imply that "anything goes" and we can call anything science. At least not with a straight face.
"There is theory. There is data. We use the former to make sense out of the latter. Science."
Hogwash! If you don't follow the Scientific Method, you don't have science. At best, you have pseudoscience.
"This is much like what happens in medical science, don't you think?"ReplyDelete
It's far from obvious that medical science is science per se. It is based on science such as chemistry and biology, and they seem to make a serious effort to approximate controlled experiments (they don't just ask people how they feel, what drugs they've taken lately, and run a regression) so it has a stronger claim than economics. But just because it's got "science" in the name doesn't make it science (political science, anyone?).
So, if the argument is economics is science because it's like medicine, then the conclusion is flawed because the premises are flawed. Economics is less scientific than medicine, and medicine is arguably not true science.
Fair enough, one could argue that there is spectrum, with physics at one extreme, medicine in the middle, and say, sociology at the other end. But economics would be closer to sociology, and nobody takes sociology seriously as science.
I see. Science good. Other stuff bad.ReplyDelete
"I see. Science good. Other stuff bad."ReplyDelete
No, science good, pseudoscience bad. The problem is your insistence on claiming economics is science when it clearly is not. This way you paint those who disagree with you as cranks and luddites when in fact their views are equally valid.
Exactly. You want to call stuff you don't like pseudoscience. A sociologist has as much claim to being a scientist as anyone else. You may not like the way he/she uses statistics, and may wonder why his or her methods are not better. You may think a clinical experiment was done badly, and that the results are being misinterpreted. These people are all doing science, as are we. Now, as to what I agree and disagree with, that's about science. I do not like bad science, and when I express a view about something I don't like this isn't similar to liking peanuts and not liking spinach. I'm actually disagreeing with the scientific approach, such as it is. And yes, there are cranks and luddites, and one is called Krugman.ReplyDelete
“[Krugman] definitely has it in for the older generation of macroeconomists, particularly Lucas and Prescott.”ReplyDelete
Of course. For example, he claims that Prescott boasted (circa 1982) that his students never heard Keynes’s name. If that’s actually true Prescott certainly deserves criticism. (If it’s not true then of course Krugman owes him a public apology.) Ignorance should never be a matter of pride, whether economics is a science or not. Even if Prescott asserts as a fact that Keynes was just wrong, it’s important to discuss where he went wrong and how we found out. It’s not as if he was a minor figure. You’ll have a hard time finding a biology student who hasn’t heard of Lamarck, even though his main claim to fame is a discredited theory.
Oh, so now we judge economics and Krugman by gossip?ReplyDelete
Al Roth on 'Is Economics Science?":ReplyDelete
"You want to call stuff you don't like pseudoscience."ReplyDelete
No, I just want to accept reality and call things what they are. I like music, that doesn't make it science.
"A sociologist has as much claim to being a scientist as anyone else."
Really? What about an astrologer?
"I do not like bad science, and when I express a view about something I don't like this isn't similar to liking peanuts and not liking spinach. I'm actually disagreeing with the scientific approach, such as it is. And yes, there are cranks and luddites, and one is called Krugman."
You haven't given any evidence as to why Krugman's science is inferior to yours. You just don't like his approach, you disagree with his analysis. It is like disliking spinach, or voting Democrat.
Ironically, Krugman is more focused on explaining the real world, is thus arguably more scientific in his approach.
"Ironically, Krugman is more focused on explaining the real world, is thus arguably more scientific in his approach."ReplyDelete
And here it is -- the claim that the rest of us are not trying to understand the real world. Arrogant asshole.
"Physics consits on laws of nature based on experimental evidence which leaves very little room for doubt."ReplyDelete
Wrong. There is lots of doubt about lots of things in physics.
"I fail to see how this is relevant. You may as well ask if I'm Jewish. The truth is the truth, no mattet who speaks it. Obviously I have some training in economics, and I am interested in science, economics and math."
You are claiming knowledge of the method of studying economics. What is your basis for this claim? It is extremely relevant if you have ever done research in economics. I was correct that you had not.
In fact, your repeated misinterpretations of Stephen's points made it very obvious.
It is true that it is not relevant if you are Jewish. If we were interpreting the Talmud, your religion still wouldn't be directly relevant but whether you had ever studied the Talmud would be.
You have not studied the Talmud. You don't speak Hebrew but you are coming to the rabbinical school to tell them what they are doing wrong. That makes you an idiot.
"And here it is -- the claim that the rest of us are not trying to understand the real world. Arrogant asshole."ReplyDelete
Childish insults aside, my point is that economists like Steven seem more concerned with the logical consistency of their models, whereas Krugman is prepared make ad hoc assumptions in an effort to match reality.
"Wrong. There is lots of doubt about lots of things in physics."ReplyDelete
No, there is a vast amount of physics that is known with near certitude. This knowledge has put men on the moon, created nuclear weapons, given us electronics, airplanes, etc. There is very little in economics which compares with this level of knowledge.
"You are claiming knowledge of the method of studying economics. What is your basis for this claim? It is extremely relevant if you have ever done research in economics. I was correct that you had not"
In fact, I know plenty about economics. One of my teachers warned me that many economists have "science envy", how right he was! But the point is that credentials don't imply wisdom. Even if a parrot was making my argument, it would still be true.
"You have not studied the Talmud. You don't speak Hebrew but you are coming to the rabbinical school to tell them what they are doing wrong. That makes you an idiot."
You will be telling me rabbis are scientists next! They have organized knowledge, after all...
"Of course. For example, he claims that Prescott boasted (circa 1982) that his students never heard Keynes’s name. If that’s actually true Prescott certainly deserves criticism."ReplyDelete
If Ed did not say this, I'm certain it's true. Ed may have on rare occasions taught undergrads at Minnesota, but for the most part he taught PhD students, and he certainly never taught them about Keynes. The question is why you think that is a problem. A student would have to be very misinformed to think that he or she would learn about Keynes by going to Ed's class, as of course Ed's tradition is Arrow-Debreu, Solow, Cass-Koopmans, Brock-Mirman, Ed. Modern macro econompasses so much stuff that you can't cover it all in the first-year macro core, and Keynesian economics is just a small part of it. I think for most macroeconomists, wage and price stickiness is just another friction, alongside private information, limited commitment, and search, for example. We don't somehow consider it central to what we need to know, and once you have the core you can easily pick it up. I'm not offended that Ed does not teach monetary economics, so why should you be offended that he doesn't teach Keynes?
"Even if a parrot was making my argument, it would still be true."ReplyDelete
Would it be true if a ptarmigan was making the argument? What if it was a left-handed ocelot? The other day a voice in the trash can spoke to me. It said, "listenest thou not to they who knowest not."
Seriously, though, the part of physics you feel is "known with certitude" is really easy. Economics is hard. As proof of that, I'm sure you know what happens when physicists try to do economics.ReplyDelete
There’s certainly no reason why Keynes should get a mention in Prescott’s own PhD courses. I should have quoted Krugman verbatim:
“From the 1950s through the 1980s, Keynesian ideas were subjected to an increasingly withering conservative critique, to such an extent that by 1982 Edward Prescott of Carnegie-Mellon University would proudly declare that students at his university never heard Keynes’s name.” (from Peddling Prosperity, 1994)
“When I use a word,” Humpty Dumpty said, in a rather a scornful tone, “it means just what I choose it to mean—neither more nor less.”ReplyDelete
Define science as whatever it is that economists do. Then economics is a science. How about that?
Seriously now, the real question is our ability to do policy based on a view of the world which is disciplined by careful observation of data, as opposed to hunches and preconceptions.
Krugman has one thing going for him: he is extremely smart and he knows how to use FRED. The rest of the profession has something else going for them: they are a lot of people working full time on trying to make sense of usually fairly specific pieces of data, constantly subjecting to each other's judgment and fiercely competing among themselves. I will let you guys decide which will yield better policy ideas.
Or else, we can throw our hands up and say that since we can't learn from the data because we cannot do controlled laboratory experiments, data doesn't matter, scientific inquiry doesn't matter and the pronouncements of a few smart people with FRED is the best we can possibly hope for.
PS.: For those who have a very elevated view of the scientific method, you should read Bruno Latour.
"I'm sure you know what happens when physicists try to do economics."ReplyDelete
What, like this guy:
Robert Joseph Barro (born September 28, 1944) is an American classical macroeconomist and the Paul M. Warburg Professor of Economics at Harvard University. He is among the most influential economists in the world according to IDEAS/RePEc. Barro is considered one of the founders of new classical macroeconomics, alongside Robert Lucas, Jr. and Thomas J. Sargent.
Barro graduated with a B.S. in physics from the California Institute of Technology in 1965, where he learned under Richard Feynman, but realized he "wouldn't be close to the top in those fields".
And seriously, the claim that economics is harder than physics is pretty laughable. I have studied both, and found economics trivially easy by comparison. Newtonian mechanics would correspond to supply and demand and the theory of the firm. Electricity would correspond to the Keynesian cross. If you really think physics is easier maybe you're in the wrong job, because you're either a physics genius or an economics dummy. The only thing difficult about economics is the pretentious level of mathematical formalism. The only people that seem to find economics difficult are professional economists like Cochrane, Fama, and Kocherlakota, who all made well-reported errors in public statements over the past couple of years.
Now, if you mean economics is harder because it is harder to establish scientific laws, then I would agree. I would say it's probably impossible to do so. But claiming economics is harder is no excuse for a lack of achievement. I mean, space exploration, the Manhattan Project, these things were "easy"? Where are the comparable achievements in economics? The euro? Mortgage-backed securities?
"Define science as whatever it is that economists do. Then economics is a science. How about that?"
So now physics is not science, only economics? We are indeed through the looking glass now.
"Seriously now, the real question is our ability to do policy based on a view of the world which is disciplined by careful observation of data, as opposed to hunches and preconceptions."
What, hunches and preconceptions like "individuals are rational utility maximizers" and "prices are flexible"?
OK, so let's define science as whatever it is that physicists do. Then economics is definitely not a science. Hurray! You win!!! Go out and celebrate!ReplyDelete
Now, to something that actually matters: "individuals are rational utility maximizers" and "prices are flexible" may or may not be hunches and preconceptions. So is "the universe has 11 dimensions" and "the big bang was succeeded by a period of inflation". What is not a hunch or a preconception is the data, and the fact that there are a ton of people out there ready to punch you in the head and deny you tenure if your hunches and preconceptions are not compatible with the data. The threshold for compatibility will be different whether you can do controlled experiments or not, the amount of data you have etc., but the general criterion is certainly there within the economics academic community.
Now, you may think the whole peer review system is flawed and that it is really a giant conspiracy to feed Ed Prescott's ego. In this case, please go ahead and help feed Paul Krugman's ego. Whether we are any better off for it is anyone's guess. On the other hand, if there is least a probability that we live in a world where the peer review has something going for it, then by discrediting it Paul Krugman is doing all of us a disservice.
I don't need to play around with the definition of science, the standard one works just fine.
The problem I have--the reason economics should not be regarded as science--is that many models are equally compatible with the data. For example, changes in nominal exchange rates seem to change real exchange rates, but some argue vice versa. The behavior of the price level seems to support RBC models, but Keynesians claim it is consistent with their model. Money supply seems to cause changes in output, but expected changes in output could lead to changes in money. Productivity goes down in recessions, but this could be explained by technological regress or labor hoarding.
My point is that data is inconclusive, and economists should strive for informed opinion rather than science. And opinions carry more weight if they are plausible.
1. Barro is a physicist in the same sense that I am an engineer and a mathematician.ReplyDelete
2. "And seriously, the claim that economics is harder than physics is pretty laughable. I have studied both, and found economics trivially easy by comparison." OK, Mr. or Ms. Genius, (i) Direct us to your published work, and let's evaluate it. (ii) Here's a project for you. Figure out what labor economists have worked out in terms of the theory that explains wages across occupations. Now, go collect some data on the compensation received by economists vs. natural scientists. What does that tell you?
The point is not that you need to play around with the definition. The point is that this is really irrelevant. Names are just this, names. If it makes you happy to be victorious, then be and let us be.ReplyDelete
The relevant question is what makes opinions plausible, specially when you want to do policy based on them. One thing that makes them plausible is to have a strong, competitive community of people working to sort them out, compare it with the data and discuss identification (i.e., how you distinguish between models). To do this effectively it helps to do it full time and to submit yourself to the evaluation of your peers.
Do you have a better idea?
"Barro is a physicist in the same sense that I am an engineer and a mathematician."ReplyDelete
The point is he studied physics to a high level at a top university, was apparently good enough to be awarded a degree, and then went on to be a rather successful economist. Moreover, he apparently did so because he thought he would do better in economics, i.e. that physics was harder. Can you point to any top physicists who got their first degree in economics? Surely, if economics is the harder subject, people should move from economics to physics, not the other way around.
"Direct us to your published work, and let's evaluate it."
I said I had studied the subjects, not that I had published original work. Surely you don't think only published academics understand anything about a particular discipline? Have you published work in math and engineering?
"Now, go collect some data on the compensation received by economists vs. natural scientists. What does that tell you?"
You're suggesting that economics is harder because they get paid more? Are you kidding me?
"The relevant question is what makes opinions plausible"ReplyDelete
The answer is surely that they correspond to the real world. For example, the idea of Ricardian equivalence is pretty hard to swallow. The problem arises when economists insist such ideas are scientific, and thus try to avoid the issue of plausibilty. An even worse example is welfare analysis, which is just ideology dressed up as science.
1. I'm defining "physicist" as someone who has a PhD in the field. Barro does not. When I mentioned that physicists have a hard time with economics, I was thinking about things like this:ReplyDelete
I was talking to Bill Zame (http://www.econ.ucla.edu/zame/) once. He has a PhD in mathematics. I said something like: "You must have found economics easy given your background." He said something like: "Absolutely not. Economics is hard." There are some conceptual ideas in economics, for example how behavior fits with the equilibrium concept to produce predictions about outcomes, that people with training in other fields find hard to get.
2. "Moreover, he apparently did so because he thought he would do better in economics..." It could be he liked it much better, or thought it would pay more. That's why I do economics and not mathematics.
3. "I said I had studied the subjects, not that I had published original work." Ahh. So you are an undergraduate. That makes some of your other comments laughable.
4. "You're suggesting that economics is harder because they get paid more? Are you kidding me?" Not suggesting, or kidding. Across departments in any research university in North America, the economists are paid better than anyone, save people in finance and accounting departments (who are essentially doing economics anyway) and perhaps computer scientists, some engineers, or people in the medical school (who essentially get paid for their clinical work). Labor economists tell us that wages are determined, first, by scarcity, determined essentially by how hard the knowledge of the field is to acquire. That's supply. Second, wages are also determined by how society values the services the field can provide. That's demand. Thus, we are well-paid because it's hard, and people want what we have to give. You learned about that in Econ 101, possibly?
"I'm defining "physicist" as someone who has a PhD in the field."ReplyDelete
So, basically you are saying someone with a PhD in physics is too dumb for economics, but someone who only has an undergrad degree in physics is smart enough to become a great economist? I define a physicist who does economics as someone with a background in physics who learns economics (e.g. a graduate program) and proceeds to do research in economics. I think people find it easier to do this than to move from economics to physics, but maybe you can prove me wrong, if physics is so easy.
Bill Zame: "Absolutely not. Economics is hard"
I like a guy with a sense of humor! Seriously though, economics some ideas that some people find difficult, but by any reasonable standard math and physics are more difficult. I don't know what economics Zame was doing, but I bet he would master the subject quicker than an economist would learn physics.
"It could be he liked it much better, or thought it would pay more."
He said it was because he wouldn't be close to the top in the field. He is at the top in economics, ergo physics harder than economics.
"So you are an undergraduate. That makes some of your other comments laughable."
Credentials don't imply wisdom, but I have plenty of degrees, thank you. I wonder if you're so patronizing towards your students.
"Thus, we are well-paid because it's hard, and people want what we have to give. You learned about that in Econ 101, possibly?"
Perhaps the supply of physicist is greater because they are more respected, so they are prepared to work for less. Or because they find it more interesting, or want to do work that could actually benefit humanity.
If being well paid means economics is hard, then Charlie Sheen must be a genius!
Astrology is hard too. Predicting the future and all that. Indians pay more $$ to astrologers than economists. Clearly, astrology attract the best minds and are richly compensated for the hard, stimulating work they do.ReplyDelete
People who want to do science go to departments that do science. Sellouts, and charlatans end up some other delusional departments.
To all of the big pictures thinkers out there: stop thinking about big methodology issues in the abstract. Focus on explaining actual phenomena and publishing in peer reviewed top 5 economics journals.ReplyDelete
"Perhaps the supply of physicist is greater because they are more respected, so they are prepared to work for less. Or because they find it more interesting, or want to do work that could actually benefit humanity."ReplyDelete
My theory is actually consistent with the data. Yours predicts that we should see doctors being paid less than roofers. If economics was so easy, with all your plentiful education you should have figured that out. Economic Science at work.
"To all of the big pictures thinkers out there: stop thinking about big methodology issues in the abstract. Focus on explaining actual phenomena and publishing in peer reviewed top 5 economics journals"ReplyDelete
Yeah, because clearly being published is more important than being right.
"My theory is actually consistent with the data. Yours predicts that we should see doctors being paid less than roofers"
No, because as ever in economics both our theories are consitent with the data. You just stick with the first one you thought of as if it were a fact. The point doesn't apply to roofers, obviously, I didn't claim it was a universal truth.
Helium baloons are not subject to laws of gravity. That point doesn't apply to anything else than the helium baloon, obviously, I don't claim it is a universial truth. Now we have two theories about helium baloons, both consistent with data. Perhaps physicists should strive for informed opinion rather than science.ReplyDelete
"The point doesn't apply to roofers, obviously, I didn't claim it was a universal truth."ReplyDelete
Not being very scientific are you? I think Kepler is squirming in his grave. Give up and go home. I think we have your number.
"Not being very scientific are you?"ReplyDelete
No, I was discussing economics, not science. Your problem is you are trying to apply scientific laws where they don't exist. Prices are determined by supply and demand, but in each market at each point in time many factors can affect both supply and demand. What is important in one market may be irrelevant in others.
" "The relevant question is what makes opinions plausible"ReplyDelete
The answer is surely that they correspond to the real world. "
Yeah, and saying what corresponds to the real world is easy, trivial, anyone can do, even an anonymous undergraduate commenter. All these discussions about data, identification, and all the rest? Waste of time. You just have to look and see.
By the way, this was sarcasm.
"Yeah, and saying what corresponds to the real world is easy, trivial, anyone can do"ReplyDelete
No, it takes a great economist, like Keynes, Friedman, or Krugman. But this is what economists should strive for, rather than ersatz science, which makes sense to other economists but is at odds with reality. A lot of economics is like science fiction, without the science...
"No, I was discussing economics, not science."ReplyDelete
No, you were attempting to discuss economics. What you have shown me is that you don't know enough economics to hold up your end of the discussion. I'm done.
"What you have shown me is that you don't know enough economics to hold up your end of the discussion."ReplyDelete
You don't accept that people's preferences influence their choice of occuption, and that this will affect the supply of labor and thus wage rates in a particular industry? Some people work as teachers for low wages, is that because being a teacher is easier than being a movie star?
Journals are where social scientists carefully study each other claims. Peer review works - perhaps too slowly but it does work. In contrast blogs are places where anyone can say anything without any serious empirical evidence. Most of the people who write to both Steve and to Krugman are amateurs who spend ten minutes on a problem and then feel happy to pontificate. There is a reason why people spends year working on problems. Understanding the world is a complicated and subtle business. But I guess it's obvious to the non-expert who bases his opinion on various anecdotes and personal experiences. Evidently Tinbergen, Haveelmo, Hansen, Sims, Hechman, Hausman .... were allhacks who wasted hit time on formal analysis because thy didn't have the great fortune to be privy to your anecdotes.ReplyDelete
I am sure this comment will strike many as elitist. It is. The future belongs to the students of the people who publish in the best journals.
People who publish in the best journals, unite!ReplyDelete
RV: Your reference to Bruno Latour is interesting. Do you believe that things like business cycles are social constructions which do not exist outside the minds of those who study them?ReplyDelete
One area where Krugman excels is his recognition of the importance of political factors in macro outcomes. In the austere world of Arrow and Debreu there are no political parties or classes. Yet no serious scientist can look at how events have unfolded over the past few years without taking this into account. Politics dictates how the losses from the financial crash are distributed.ReplyDelete
If we remember, the name of the subject is political economy.
"Journals are where social scientists carefully study each other claims. Peer review works - perhaps too slowly but it does work."ReplyDelete
OK, please share with me the economic laws that have been established by this great Hegelian process of truth discovery. For example, I'm sure economics research can provide reliable answers to the following questions:
1. Are prices mostly sticky or flexible?
2. Are recessions mostly due to changes in productivity, monetary factors, or something else (e.g., politics, confidence, angry deity)?
3. Are economic fluctuations harmful to welfare? I.e., should stabilization policies be attempted?
4. Do individuals form expectations rationally?
6. Is Ricardian equivalence a reasonable approximation of reality?
5. Do individuals seek to maximize utility, and does this correspond to welfare in the true sense (i.e., could they be made better off by forcing them to do something that would reduce subjective utility, like quitting smoking)?
7. Is current US unemployment mostly structural or due to deficient aggregate demand?
My contention is that most "progress" in economics research over the past 50 years has been entirely negative. Friedman, Lucas, and Prescott have demonstrated the deficiencies in the Keynesian model, but the alternative theories they offered fail to hold water. Current economic analysis is based almost entirely on the personal taste and opinions of the economist. Theories are nuanced but not discarded. There has been no progress towards a reliable description of reality, just an "arms race" where debate is conducted using ever most sophisticated math and econometrics, but never reaches a conclusion. In economics, there are no fact, only interpretations.
"I am sure this comment will strike many as elitist. It is. The future belongs to the students of the people who publish in the best journals."
You are part of an "elite" in the same way that Freemasons or Scientologists are. The future belongs to people who are willing to accept reality.
Really, economics is not that hard that you can view yourself as part of an "elite" holding esoteric knowledge. Anyone can learn the basics of economics, and anyone with sufficient math background can learn the current research.
"No, it takes a great economist, like Keynes, Friedman, or Krugman."ReplyDelete
I see, it takes a genius to really see it. Hopefully they are benevolent enough so they won't trick us.
There is a name for this method of access to the truth: prophetic revelation. Nothing could be further from science, whichever way you define it.
I am done with that...
On Latour: I will not elaborate too much because I am not an expert. I brought him up because his field work shows that the practice of science among natural scientists is very far from following the scientific method as we learn in school. So brandishing natural sciences around as the great model of scientific inquiry because they are oh so disciplined means you haven't really had an honest look on how they actually work. Now, you may think this dismisses the whole enterprise and that we should rely on Einstein and Planck for divine revelation. Or you can acknowledge that even working this way, natural scientists have made important contributions and that the whole fixation on the Pure Scientific Method is really a red herring (I am fairly sure Latour holds the latter view).
"Astrophysics is part of physics, so is clearly science."ReplyDelete
"Psychology is largely hokum."ReplyDelete
Great insight. And can you run through for us, Herr Dokter Professor, exactly what disciplines are the acceptable subjects of study and what disciplines we can refer to as "science"? E.g., observational science more generally--science, or "hokum"? Thx
Man, did I really just read through that whole thread?ReplyDelete
Congratulations. I certainly did not read it all.
As you probably guessed, there's only a very small amount of variation so you are not missing much. Basically an anonymous troll repeating the legend, "physics good, economics bad", for 160+ comments, with occasional digs at every other discipline that isn't physics thrown in for good measure.
Yes, I made the mistake of actually trying to reason with him/her/it.ReplyDelete
Ah, but you can't reason with trolls. That's how they get you.ReplyDelete
Have you ever thought of writing a post on Krugman *and* Ron Paul for the ultimate troll meltdown?
You'd probably get twice as many worthless comments. On the other hand, flowers would bloom in comments sections across the econ blogosphere...
Recent NYT article on deficit reduction:ReplyDelete
The lack of definitive answers reflects the reality that economics is not a hard science.
“Reasonable people can sit down and, apart from any political or policy motivations, come up with different answers,” said Robert S. Chirinko, a finance professor at the University of Illinois at Chicago who studies corporate taxation.
Dr. Chirinko said his own studies had found that tax increases caused real but modest reductions in corporate investment. Studies by the Harvard economist Martin Feldstein, a noted conservative, find larger reductions.
“The problem is that economists really can’t run experiments like in science class,” Dr. Chirinko said. “I keep coming up with the same answer, so I have some confidence. On the other hand, Martin Feldstein is a great economist.”