Saturday, April 06, 2013

A world without macroeconomists?


Miles Kimball, myself, and some other folks were having a discussion on Twitter the other day about the question of "What would happen if there were no economists?" So we decided to do some blog posts on the topic.

Of course, the definition of "economist" is fuzzy; if you make a chart of past GDP growth rates on your home computer just for fun, does that make you an "economist"? So I'll assume that "economist" means "academic economists, or economists working at government research institutes." In other words, paid econ researchers.

Anyway, the question is much too hard! There are a LOT of kinds of economists, and they do lots of different things. Tax economists study tax policy. Labor economists study individuals as they make their way through the economy. Decision theorists study how individuals make decisions. Applied game theorists study how people bid on auctions, and lots of other stuff like that. Financial economists study financial markets. Econometricians do statistics. Development economists study how countries get rich. Trade economists study patterns of trade. Industrial organization economists study...um...Well, anyway, there are a lot of different types of economist. The question of what would happen if they all vanished is fun far-out speculation, but I'll leave that task to somebody else.

Instead I thought I'd just stick to the type of economist everyone hears about: Macroeconomists. What if universities, and the Fed, stopped hiring people to do macroeconomics?

Well, for one thing, universities probably wouldn't save any money. The demand for econ profs is based mainly on the demand for undergrad econ classes, which probably wouldn't much change if one subfield of econ research were abandoned. Econ classes would just focus more on micro and on subfields of micro, and would be taught by micro researchers. As for the U.S. government, it would save a tiny bit of money, but not much, since macroeconomists have tiny research budgets.

Would we lose the ability to forecast the macroeconomy accurately? It's tempting to conclude "No," since macro models of the type overwhelmingly used, created, and studied by modern macro researchers (DSGE models) haven't really proven better at forecasting than the non-structural spreadsheet-type models used by most private firms, or than consensus individual forecasts (actually one DSGE model has performed slightly better in recent years, but this might just be data mining and publication bias). It's telling that private firms don't hire people to make DSGE models, but do hire people to make forecasts with much simpler tools.

However, here's an interesting thing about research, and about science: Past discovery is no guide to future discovery. Chemists were basically a joke for centuries before they stumbled on a few key principles, and rapidly turned into the most reliable discovery-factories in all of science. Biologists had an even longer history of uselessness before they became incredibly useful thanks to new technologies. So someday, macroeconomists might learn how to forecast the economy extremely well. We really just don't know. A breakthrough in forecasting power would yield huge payoffs to society.

OK, what about policy? Macroeconomists will gladly tell you that modern models are not a lot of use in forecasting - that their main use is in giving policy recommendations, conditional on your assumptions (i.e. "If for whatever reason you believe that the economy works this way, here's what you should think you can do with policy.") But as I've often griped, I don't think they really do this particularly well...it's too hard to choose which one to use, even if you know your own general priors about how the economy works. And people have silly priors anyway. Not to mention that even if you choose a model, its output may be incredibly hard to interpret and use. And when the world seems to hit on a consensus policy that seems to work well - the Fed using interest rates to lean against fluctuations in inflation and GDP, for example - the models seem to follow the policy rather than the other way around.

So does this mean that macro research is useless for policymaking? No! Not at all!! Because here's an interesting thing about policymaking: No matter who advises the policymakers, policy is going to get made. That includes economic policy. So if there were no academic and Fed macroeconomists around to advise policymakers, who would policymakers listen to on economic matters?

My guess: Some very dangerous people. 

For all the talk of academic macro being politicized, it's much less politicized than the macroeconomic discussion outside of the research community. My own experience is that most macroeconomists are pretty apolitical, and research supports that...but even if my sample is biased, macro's interventionist and laissez-faire schools are pretty close to each other ideologically, compared to, say A) armchair-theorizing politicians, B) TV commentators, C) the denizens of internet forums. It really is a jungle out there. You have David Stockman. You have Ron Paul and his followers. You have David Graeber and his followers. And worse. You have "Austrians" who think all of economics can be deduced from some vague derp. You have Marxists who think - well, I'm not sure, because they tend to denounce and vilify you if you even ask them what they mean, but it sounds nuts. In short you have a cavalcade of vast unending wackitude, often with a proven track record of wrecking economies and societies.

So it's possible to see macroeconomists as doing plenty of good, simply by sitting there not being absolute wackaloons. A million DSGE models from which it is impossible to select sounds a lot better to me than three or four totally nutcase worldviews, the selection of any one of which is likely to cause human tragedy on a vast scale. (Note: This idea, of macroeconomists as a vaccine against macro-lunacy, was first suggested to me by Justin Wolfers.)

(Update: In the comments, Robert Waldmann writes:
A world without macroeconomists wouldn't be a world without economists who talk to journalists about the macroeconomy. The risk is that the voice of the economics profession will be trade theorists, economic historians, behavioral economists and finance economists such as Krugman, DeLong, C Romer, Goolsbee, Fama and Cochrane. 
I think Waldmann is right, and the danger from having zero dedicated macroeconomic researchers is not as great as I'm making it out to be; other economists could still hold back the crazies, they'd just have a harder time doing it with no empirical research or macro theory to back them up. So the crazies might not immediately take over, but they might gain more clout and respect.)

And actually, there's a huge area of macroeconomic research that I haven't even mentioned yet: macro empirics. Empirical macro is fundamentally constrained by the limitations of time-series data ("history only happens once", i.e. ergodicity and stationarity are near-impossible to verify) and cross-country comparisons. But within those limits, macro empiricists can tell us a lot, just by quantifying things and noting seeming regularities. That's enormously helpful both for policy and for forecasting. You need controlled experiments to create a reliable predictive theory of the world, but there's a huge amount you can know without controlled experiments, just by watching the world go by and keeping careful track of what you see. And macro empiricists can certainly do the latter. It's thanks to them we know things like Okun's Law, which tells us that fast growth is consistently associated with low unemployment. It's also thanks to them that we know that investment is much more sensitive to the business cycle than consumption. Or that slightly less than half of people seem to be "hand-to-mouth" consumers who don't obey the Permanent Income Hypothesis. Etc. Without macroeconomists, we just wouldn't know these things, and reasonable people would argue about them.

Also, I didn't even mention spinoff products. Macro empiricists, struggling to wring some insight out of terrible data, often come up with truly remarkable innovations that may be useful in many disciplines. I'd put Chris Sims' work with structural vector autoregressions in this category.

So, in short, from macroeconomists we get at least the following benefits:

1. Quantification of macroeconomic phenomena.

2. Observation of correlations between macroeconomic variables (and between macroeconomic and microeconomic variables).

3. The creation of "spinoffs" like sVARs.

4. The unknown potential for big breakthroughs in forecasting methods.

5. An "anchor" that keeps policy away from the dangerous extremes urged on us by various politically-motivated fringe groups.

So without macroeconomists, it seems to me that we would lose quite a bit, actually. And since macroeconomists are very cheap, all in all, it seems we should keep them around.

That doesn't mean we bloggers have to stop complaining about them, of course. ;-)

88 comments:

  1. "I sometimes think that the general public would know more about economics if the economists did not exist" - Francis Cripps - 1979, The Guardian, 27 Feb.

    ReplyDelete
  2. The chance of a society producing macroeconomists approaches 1 as the complexity of economic relations approaches ∞.

    The chance of a society possessing macroeconomists producing a model with predictive power greater than extrapolating regression lines is uncertain, but is probably inversely correlated to global GDP growth levels.

    ReplyDelete
  3. John S7:27 PM

    What exactly is wrong with Ron Paul? He gives too much airtime to Salerno and other anti-frac reserve people, but he also invited Free Banker Larry White to testify and give an opposing opinion. To my knowledge, he doesn't want to outlaw frac reserve banking.

    http://www.youtube.com/watch?v=jVm3Yzjq8zE

    In terms of policy, he has indicated many times that he doesn't want to immediately abolish the Fed, as that would indeed be far too disruptive. As far as policy, the main economic bill he pushed for was the "Free Competition in Currency Act" which primarily sought to end capital gains and sales taxes on gold and silver as money.

    http://www.freebanking.org/2011/09/13/the-free-competition-in-currency-act-of-2011/

    Is there any harm with removing the tax barriers against competing currencies? Or suggesting that a central bank isn't necessary?

    ReplyDelete
    Replies
    1. http://noahpinionblog.blogspot.com/2011/07/now-ron-paul-wants-to-debase-currency.html

      Delete
    2. John S10:45 PM

      Yes, I had forgotten about the Rothbard-inspired idea of debt repudiation. But is it really crazier than Steve Keen's debt jubilee idea? And others (Skidelsky, Yglesias) have written favorably about public debt forgiveness w.r.t. the Eurozone, yet I don't think you consider them crazy like you do Ron Paul.

      http://www.project-syndicate.org/commentary/down-with-debt-weight
      http://www.slate.com/blogs/moneybox/2012/05/02/the_full_argentina.html

      Also, your concerns about hyperinflation are unfounded because his first priority would be to set up a Free Banking system in which the money supply would not depend on central bank independence. He doesn't support a centrally-managed gold standard; he believes that gold would likely serve as the monetary base in a Fed-less system, but he would leave it up to the free market to decide the issue.

      From Wikipedia: "I wouldn't exactly go back on the gold standard but I would legalize the constitution where gold and silver should and could be legal tender" - Ron Paul

      More on Free Banking and gold: http://www.freebanking.org/2012/02/11/more-on-the-gold-standard-with-regret/

      Ron Paul's ideas were radical, but I believe they were more sophisticated than you give him credit for.

      Delete
    3. From Wikipedia: "I wouldn't exactly go back on the gold standard but I would legalize the constitution where gold and silver should and could be legal tender" - Ron Paul

      --------------

      What is the empirical content of "gold and silver should and could be legal tender"?
      If legal tender consists of dollar bills, along with gold coins how does ANYTHING change?
      Normal people will use dollar bills like they do today. Nutcases will hoard their precious gold coins and fob off their dollar bills onto suckers (Gresham's law and all that). How is this world any different from today's world, except that the gold coins being hoarded are stamped "US Min" rather than "Franklin Mint"?

      Delete
  4. A world without macroeconomists wouldn't be a world without economists who talk to journalists about the macroeconomy. The risk is that the voice of the economics profession will be trade theorists, economic historians, behavioral economists and finance economists such as Krugman, DeLong, C Romer, Goolsbee, Fama and Cochrane.

    I actually think that macroeconomists have very little influence on macroeconomics as perceived by policy makers and the public. Summers is influential. He is also a total heretic (he didn't leave the field, the field left him). Blanchard and Woodford are influential macroeconomists and reasonable people.

    The world has chosen not to benefit from the insights of the most influential academic macroeconomists Robert Lucas and Edward Prescott. I don't think you think this is a bad thing.

    On consumption and the PIH the data suggest that somewhat less than half of consumption is by hand to mouth consumers. They don't suggest anything about the rest. The rest of fluctuations in consumption can't be predicted by predicting current GDP. Also radioactive decay can't be predicted by predicting current GDP. The PIH is not a good theory of radiactive decay.

    The (representative agent with an additively separable utility function and no liquidity constraint version of the PIH) glass is not much more than half full. I am not aware of evidence that it isn't empty.

    ReplyDelete
    Replies
    1. I think you overlook the indirect effect. The non-macro economists still get much of their ideas from macro. So if there were no macroeconomists, those non-macro voices that you think has more influence on journalists and policy would be more in the dark and (even) less useful. So even though I agree that macroeconomists have little (direct) influence, they do have a lot of indirect influence.

      Delete
    2. As to the false charge, that "The world has chosen not to benefit from the insights of the most influential academic macroeconomists Robert Lucas "

      Do you mean that idiot who, when President of the AAE, said there would be no more Great Depressions, like the one we are in?

      Noah, there is a second reason for having macroeconomists.

      Every one in a while they will provide a keen insight into how the world actually is working. For example, using my theories on information, Karl Smith just wrote in Forbes:

      With apologies to the less wonkish, China is using physical capital as a loss leader in order to grow cities that will produce network effects will in turn foster the human capital that really makes a country rich.

      Now, Miles never taught you about information being the principal factor or production, but here you have in one sentence pretty much a total and complete explanation why there are no jobs and why Obama The Fool has created no jobs. He was a fool for relying on Larry Summers et al., who really had no idea what they were doing.

      It matters not whether China figured it out by accident or design. What matters is that it works.

      That is the problem with macroeconomics. It favors mental models over thought and experimentation.

      In the US we have millions of idiots against industrial planning---it is opposed to Hayek and Rand. China says, BS, we can afford the losses because we get the added value of information and network effects.

      Delete
  5. I think you go to easy on the macroeconomists who say their models are good for policy analysis even if they aren't good for forecasting. To be useful for policy analysis a model has to approximate reality. There are two known ways to see if this is true of a model. One is to treat it as a null hypothesis and test it. Macroeconomists are absolutely unwilling to do this. The other is to use it to make out of sample forecasts and compare them to data.

    Poor forecasting ability means there is no evidence that the model is of any value or interest whatsoever.

    I will restate the argument you quote without affecting its validity at all "Voodoo priests will gladly tell you that modern models are not a lot of use in forecasting - that their main use is in giving policy recommendations, conditional on your assumptions (i.e. 'If for whatever reason you believe that the economy works this way, here's what you should think you can do with policy.')"

    An argument which begins "if you believe" can't possibly justify well my salary for example.

    Also no one believes in DSGE models. The full argument is "Friedman methodology .. even if you don't believe the economy works the we assume when making models, you can't be sure our advice is wrong, and Lucas critique if you don't believe the economy works the way they assume when making models then their success in fitting data doesn't imply you should pay attention to their policy evaluation."

    I consider both arguments valid. I know I don't know I "can't be sure" and that empirical results don't imply. So I know I know nothing useful.

    In short neither I (the macroeconomist) nor you (the policy maker) have a clue. So pay me a salary (oddly they do).

    ReplyDelete
    Replies
    1. I think you go to easy on the macroeconomists who say their models are good for policy analysis even if they aren't good for forecasting. To be useful for policy analysis a model has to approximate reality. There are two known ways to see if this is true of a model. One is to treat it as a null hypothesis and test it. Macroeconomists are absolutely unwilling to do this. The other is to use it to make out of sample forecasts and compare them to data.

      Poor forecasting ability means there is no evidence that the model is of any value or interest whatsoever.


      But I agree with this...how am I going easy on them?

      Delete
    2. Robert, as you well know, the word "good" is ill-defined and, frankly, subjective. First, limiting the set of infinite possibilities to a range, even a wide one, is of some value. Second, no one bothers to discuss the forecasting horizon. Next minute? Next day? Next week? Next year? When people make such vague statements, I would appreciate if they bother to specify what they have in mind.

      Delete
  6. John S10:56 PM

    Some good comments by John Papola on Arnold Kling's blog:

    "More than ever I am convinced that the professional study of economics peaked in the classical liberal era with John Stuart Mill and rapidly became a mostly destructive force in society thereafter... This seems to be the result of the mathematization and hyper faux-specialization of the “economics profession”, combined with the fact that most employment opportunities for economists are in academia (where the real world is irrelevant) or government directly (same problem, worse incentives)."

    Kling summarizes: "[The economics profession] is a self-perpetuating, in-bred, smug, narrow guild. I do not know what to do about it."

    http://www.arnoldkling.com/blog/john-papola-on-the-state-of-economics/

    ReplyDelete
    Replies
    1. That's a good comment?

      I'd love to see an overgeneralised nihilistic one.

      Delete
  7. "What if universities, and the Fed, stopped hiring people to do macroeconomics?

    Well, for one thing, universities probably wouldn't save any money. The demand for econ profs is based mainly on the demand for undergrad econ classes, which probably wouldn't much change if one subfield of econ research were abandoned. "

    Not if online education takes off...then we'd just need one macroeconomist to run the site.

    ReplyDelete
  8. Argosy Jones3:59 AM

    "Poor forecasting ability means there is no evidence that the model is of any value or interest whatsoever."
    What Robert said.

    Screw 'em.

    We'll figure things out by trial and error just as quick without bogus macro models as we will with them.

    So we'll try some stupid crazy ideas. We do that anyway. We're not exactly batting a thousand since about 2003, when Lucas said, "the central problem of depression prevention has been solved".

    By all means keep some captive specimen macro researchers at every university. Just keep 'em off the Sunday talk shows and the CEA until they get a model that can make a decent forecast of the things people actually care about and test it for a couple of decades in the privacy of their seminar rooms.

    Empirical work is ok.

    ReplyDelete
  9. Barry7:01 AM

    Two queries:

    1. What about self-styled non-partisan research institutions?

    2. Where can I get data on the numbers of economists you discuss (with or without those at 1.)?

    ReplyDelete
  10. Dear Noah,

    I share your assessment of the current field of econowannabes in the Street. But these people's problem is precisely that they wannabe macroeconomists, that they want to understand and manipulate the economy for greater good. You cannot possibly believe that macroeconomists are apolitical or do not have ethical and professional commitments to benefitting society through their work.

    Macroeconomists were responsible for the soviet transition to markets which has been a demonstrated failure. Starving farmers and eventually manufacturers on the other hand did a handy job of on-the-ground reform in China. Macroeconomists at LSE were responsible for the planning disaster in India that ensued from Independence until the end of the license raj. Macroeconomists broadly supported failures of development all over the world, while say the Green Revolution fed zillions of people by tinkering with technologies -- by messing around *inside* the production function.

    What, would you say without using counterfactuals of "disasters prevented" would you say the large-magnitude contributions have been of tinkering with interest rates and fiscal spending?

    Macroeconomists have done amazing work which has benefitted our understanding of the economy enormously. Hyperinflation is by all signs a solved problem. We understand growth comes from productivity, even if we don't understand TFP that well.

    But your thesis that macroeconomists are dispassionate experts who keep an insane world from destroying itself with its hack ersatz theories, is self righteous, and precisely the attitude that's engendered some of the economic tragedies of the last century.

    Keep up the good work though; I love this blog.

    ReplyDelete
    Replies
    1. G. Peterson: "Starving farmers and eventually manufacturers on the other hand did a handy job of on-the-ground reform in China."

      John M. Keynes: "Practical men, who believe themselves to be quite exempt from any intellectual influence, are usually the slaves of some defunct economist."

      Delete
    2. CA:

      The Chinese story I referenced is well documented. See Victor Nee's recent book Capitalism From Below, and Coase's recent book on the same matter. It has also been well known and documented by a great deal of development economists for more than a decade.

      I'd agree that I were the slave of some defunct economist, and exempt from intellectual influence, if I weren't a transparently aspiring academic myself, and if the ideas I cited weren't empirically verified.

      Delete
    3. Well, I just wonder where these practical men got their ideas from. I have a feeling they are not re-inventing the wheel. And neither are you, regardless of whether you are an academic or not. In fact, being an academic makes it all the more likely that you are the slave of some defunct economist of the past.

      Delete
    4. In broad strokes you're correct -- largely people's beliefs always take their roots from ideas that have been espoused before them. New ideas are merely recombinations of old ideas. I don't think I'm reinventing the wheel, and I'm not sure how my point about a bad track-record of macroeconomic engineering turned into a debate about whether I'm original, or whether Chinese farmers took their inspiration for a system of property rights from books or not.

      I'm certainly not anyone's slave just because I rely on a network of ideas in history in order to create my own.

      Delete
    5. Oh, come one Graham, it's a metaphor! Anyway, here is Keyne's whole quote:

      "The ideas of economists and political philosophers, both when they are right and when they are wrong, are more powerful than is commonly understood. Indeed the world is ruled by little else. Practical men, who believe themselves to be quite exempt from any intellectual influence, are usually the slaves of some defunct economist. Madmen in authority, who hear voices in the air, are distilling their frenzy from some academic scribbler of a few years back."

      Delete
    6. It's a good quote. I agree.

      Delete
  11. "However, here's an interesting thing about research, and about science: Past discovery is no guide to future discovery. Chemists were basically a joke for centuries before they stumbled on a few key principles, and rapidly turned into the most reliable discovery-factories in all of science. Biologists had an even longer history of uselessness before they became incredibly useful thanks to new technologies. So someday, macroeconomists might learn how to forecast the economy extremely well. We really just don't know. A breakthrough in forecasting power would yield huge payoffs to society."

    Can't one say the same about homeopathy ;-)

    ReplyDelete
  12. Noah, I was a little worried about this series when you mentioned it on Twitter but I think you started at the right place. Macroeconomists are constantly wrestling with counterfactual states of the world...so why not put them inside the exercise.

    Oddly enough I spent Friday afternoon in a room full of economists (many of them macroeconomists) talking about how we can be more effective in our work. Our topics ranged from narrow issues of forecast presentation to the broad issues of attacking groupthink...almost all of it was motivated by improving our analysis and forecasting of the macroeconomy in support of monetary policy.

    Our staff analysis is just one input in the policy process and as you suggested, I think those inputs would be far narrower if the macroeconomists disappeared. The plural of anecdotes is not data and policy makers often often have a set of contacts with anecdotes. Those contacts may well be right, but it seems valuable to have a broader perspective too.

    One particular weakness (and strength) of macroeconomists is the need for a story. We are the only subfield in which general equilibrium is central, so our stories get tricky. We spend a lot of time honing the staff story of the macroeconomy and buffing up all the pieces when we know (and admit) they can't all possibly be right. In academic macro, I like to see the multitude of models and stories. I just think we should assign some probabilities and learn to live with the fact that there is more than one possible story. I think in general macroeconomists (in all there hiding places) have a tough time with dissenting views and it is a real disservice to the profession.

    One last thing on forecasting: it is futile but necessary. Monetary policy acts with a lag so policy makers need a sense of where we are headed. I don't know many economists who like to forecast...even when you're right it can be miserable and you're rarely right. But it is important to keep trying.

    (Disclaimer: I am a macro forecaster who does applied micro research, so I wouldn't totally disappear even in this counterfactual world.)

    ReplyDelete
  13. Isn't it true that a lot of the whackadoodles who get airtime get it due to credentials earned, in prior years, in academic research? So in that sense, macroeconomics departments are dangerous because, rather than inoculating the public discourse against whackadoodles, they provide a breeding ground for them.

    ReplyDelete
    Replies
    1. Not really. The true wackadoodles are not trained macroeconomists. Of course, there are some trained macroeconomists who say silly stuff, but it's not nearly as wacky as the stuff, say, David Stockman says.

      Delete
    2. Anonymous12:57 AM

      The true wackadoodles are not trained macroeconomists.

      Yeah, sure.

      Hubbard. Mankiw. Summers. Feldstein.

      No "wackadoodles" there, you bet.

      But nice of you to add some evidence to John S comment, above.

      Delete
  14. I like this post but I think macroeconomist's main problem is how they see themselves: objective, value free scientists. The problem is that the values are hidden: perfect competition is the ideal, competition in general is good, efficiency is good, etc etc.

    But everyone is biased, really. Could it be that through their 'moderate' opinions, macroeconomists don't ask the really difficult question? The answer isn't always 'in the middle.'

    By the way, you speak as if getting rid of macroeconomists would produce disaster. But how much worse can it get? Austerity is completely f***ing Europe, to the extent that fascism is on the rise in Greece; global poverty is still massive; there was an enormous financial crisis 5 years ago that the world has yet to recover from, and the biggest recent success stories - East Asian countries - didn't design their policies with even a nod to what mainstream economic theory tells us.

    Personally, I'd be happy to replace economists with biologists and physicists. They might start slow, but their empiricism and superior mathematical skills would produce better theories very quickly.

    ReplyDelete
    Replies
    1. Anonymous9:25 AM

      Bored physicists (read: Walras or Jevons) is the reason a lot of economics is as screwed up as it is today.

      Economists don't need better mathematical skills. It would make things worse, and feed the conceit that you can actually precisely and analytically model intractable problems.

      What economics needs is people good with computer science; okay with uncertainty; strong with statistics; and comfortable without precise answers in a rigid framework.

      Sure if you got rid of macroeconomists the world might not immediately collapse but, today, if you got rid of all macroeconomic KNOWLEDGE we'd be back on the gold standard at 10x the austerity believing that deficits are always bad.

      Delete
    2. I agree with your first statement, but it's odd how inconsistent it is with the your last statement. I think the formalization of economics in models and math was good, but we economists tend to push most good things past the point of diminishing returns so it can be a trap too (physics envy taken too far). Grand theories or stories which run well through our machinery are prized over odd bits of info and patterns on the economy.

      As an example, one of the advisors for my research seminar in grad school (an amazingly creative macroeconomist) told us to pick questions we could answer, not ones that were interesting. It's the best advice for young economists who want a job someday but it's totally demotivating. Happily I work where we get tough, impossible to answer macro questions every day. I can't answer them in a tenure track way, but I do my best.

      Btw physicists and biologists would need whole new tool kits to replace macroeconomists ... randomized experiments in macro are rare and often considered immoral. Social sciences rock!

      Delete
    3. I was responding to Unlearningecon, not Ashok who I agree with on a better treatment of uncertainty.

      Delete
    4. "As an example, one of the advisors for my research seminar in grad school (an amazingly creative macroeconomist) told us to pick questions we could answer, not ones that were interesting. It's the best advice for young economists who want a job someday but it's totally demotivating. Happily I work where we get tough, impossible to answer macro questions every day. I can't answer them in a tenure track way, but I do my best."

      I think this is the most important advice there is. To a fairly uneducated (compared to all you econ PhDs!) person like me, the work in economics that really seems the most useful is the quiet stuff we don't hear about. The staff economists on government committees working with uncertain numbers and statistics over "brilliant models".

      Questions you can answer with a certain level of uncertainty offer such rich insights for policymakers.

      As for different tool kits needed in economics from physics and biology, I suppose skeptics want economics to be theoretically falsifiable. And a lot of the times it just doesn't seem to be the case. Economists guard every theory under an assumption that by definition can't happen.

      This is why a move to computational work and a focus on econometrics has done so much good for the discipline.

      Perhaps Economists need to be incentivized (perhaps by a central bank/government) to find specific, but useful/key, patterns of the macroeconomy. Rather than genius models that require years of real analysis or whatever the hell else people use. I also support greatly expanding economics budgets (which are about a 70k on average per project, all costs included, according to the ECB) to support a powerful empirical/computational engine.

      Delete
    5. @both of you, in a muddled-together sort of way:

      I almost see econometrics as a different discipline to economic theory. It has its own problems, obviously, but those are mostly common to all statistical work in social science (inferring causation, scarce data, various 'tricks' that may obscure more than they solve). Overall econometricians are far more careful about their assumptions and which techniques they choose, and have, IMO, evolved more as time has gone on. Much of economic theory, on the other hand, is based on stuff from almost 150 years ago.

      "I agree with your first statement, but it's odd how inconsistent it is with the your last statement"

      Not at all. I actually find physicists are far more willing than economists to concede that their science involves value judgments and they always should be concerned with people when doing their technical work. Just because I think economists should use mathematics, doesn't mean I think theories can be value free.

      "Bored physicists (read: Walras or Jevons) is the reason a lot of economics is as screwed up as it is today."

      Perhaps, but it's a question of using the right mathematics. Physics has moved on a lot since the days of these guys and now has more appropriate tools for modelling complex systems.

      "Sure if you got rid of macroeconomists the world might not immediately collapse but, today, if you got rid of all macroeconomic KNOWLEDGE we'd be back on the gold standard at 10x the austerity believing that deficits are always bad."

      Well, a lot of people right now believe the latter. Accountants would actually suffice to tell people this is not so.

      Even with the sparse knowledge available in economics, some theories are either trivial or falsified. Normal scientists would have allowed the field to progress beyond this.

      Delete
    6. No, Unlearningecon. I disagree with almost everything you said. While I can see deficits and blind spots in current (macro) economic inquiry, I actually see most of the necessary pieces for progress laying around. I would call all the economists I know scientists...in the sense of people trying to understand how the world (in our case, the economy and some non-economy) works. Every time I present my research to the top macroeconomist at my work he asks about my identification ... every time, first question. It's not just econometricians who are careful.

      I think more about the difference between psychologists and economists since we looking at more the same behavior than physicists or biologists. Economists focus in on central tendencies and average behavior ... the typical storyline ... and psychologists are often checking out all the perturbations ... the malleability of the story. Both have strengths. There actually has been much progress is bringing the insights of psychology to economics and forcing us to rethink market mechanisms. Sure it's still something of a mess. Oh wait, that's kind of what the process of scientific discovery looks like...a mess.

      Delete
    7. The problem is that the values are hidden: perfect competition is the ideal, competition in general is good, efficiency is good, etc etc.

      Okay, "competition in general is good" is a result, not a value.
      "Perfect competition is the ideal", well, it can be if you like, but it's an unachievable ideal. (And if you have it as an ideal, how do you expect suppliers in markets which have high fixed costs and low SRMC to actually achieve a rate of return on investment and thus keep investing?)

      "Efficiency is good" is 'hidden' in every Econ 101 text book. For heaven's sake, if economists are hiding their value that efficiency is good, then what would it take to not hide something? Billboard ads on every major motorway? Are you going to fund these?

      So, two of the things you list are not values of economists in general, and one is about as unhidden as it can be.

      Could it be that through their 'moderate' opinions, macroeconomists don't ask the really difficult question? The answer isn't always 'in the middle.'

      Well, I suppose it could be, how about you tell me what the really difficult question is?
      But, it strikes me that macroeconomists do ask some fairly difficult questions, such as "what's the economy going to do next year?" Or "how can we justify aggregating demand and consumption"? even if they happen to omit whatever "the really difficult question" is. I think the real problem in macroeconomics is in coming up with answers.

      And I think we can acquit macroeconomists of always answering "in the middle". Whatever you think of the Lucas critique, or money neutrality, regardless of whether you think they're totally foolish ideas or if you think they're so fantastic you always send Robert Lucas flowers on his birthday, they're not ideas that are "in the middle".

      But how much worse can it get? Austerity is completely f***ing Europe,

      In defence of macroeconomists, everyone, even Keynesians, say that government debt should at least fall in bad times, but instead Europe was letting government debt rise in good times and bad.

      global poverty is still massive;

      And falling significantly, see http://web.worldbank.org/WBSITE/EXTERNAL/NEWS/0,,contentMDK:23130032~pagePK:64257043~piPK:437376~theSitePK:4607,00.html. To quote:
      "An estimated 1.29 billion people in 2008 lived below $1.25 a day, equivalent to 22 percent of the population of the developing world. By contrast, in 1981, 1.94 billion people were living in extreme poverty."

      there was an enormous financial crisis 5 years ago that the world has yet to recover from,

      That's a rather Western-centric view. See http://www.indexmundi.com/g/g.aspx?c=xx&v=67, world GDP per capita is now going up again.

      and the biggest recent success stories - East Asian countries - didn't design their policies with even a nod to what mainstream economic theory tells us.

      A surprising claim. China and Vietnam took off once they moved away from Communism, not being Communist is a very mainstream economic policy.
      And if we look at the Heritage Index's of Economic Freedom, the most successful East Asian countries, Japan, South Korea, Taiwan, all score 70 or above on property rights, and 80 or above on Business Freedom. Overall Taiwan is 20th in the index, Japan 24th and South Korea 34th. (China and Vietnam are lower, and much poorer, the question is whether their growth rates will slow once they've used up the boost from their past moves away from Communism).

      Delete
    8. The value judgments mostly stem from 'efficiency is good.' For example:

      "Okay, "competition in general is good" is a result, not a value."

      It is a value as a combination of 'competition causes efficiency' and 'efficiency is good.' The perfectly competitive ideal is based on the same judgment.

      ""Efficiency is good" is 'hidden' in every Econ 101 text book. For heaven's sake, if economists are hiding their value that efficiency is good, then what would it take to not hide something? Billboard ads on every major motorway? Are you going to fund these? "

      IDK if you're trolling or what but I can tell you my textbook does not explicitly state that efficiency is 'good'; it only assumes it throughout, then talks about how best to maximise this efficiency. Economists regularly assert their discipline is value free.

      The rest of your comment is almost like a parody of 'history according to economists (or libertarians):

      "A surprising claim. China and Vietnam took off once they moved away from Communism, not being Communist is a very mainstream economic policy. "

      How rigorous. Meanwhile in the real world, things are more simple than 'not being communist.' What China did was the exact opposite of what economists recommended in eg Russia (a disaster): they preserved existing institutions and 'built' capitalism on top of it. For example the existing state quotas were preserved but peasants were allowed to produce and sell grain on top of this. IDK about Vietnam

      "And if we look at the Heritage Index's of Economic Freedom, the most successful East Asian countries, Japan, South Korea, Taiwan, all score 70 or above on property rights, and 80 or above on Business Freedom. "

      Oh dear. 'Economic freedom' indices. You mean the ones that are vaguely defined and basically just support 'free market' prejudices?

      http://www.slate.com/blogs/moneybox/2013/03/29/mercatus_freedom_map_and_the_fallacies_of_libertarianism.html

      Again: meanwhile, in the real world, S. Korea and Japan heavily supported infant industries. In fact in S. Korea state officials actually worked inside firms. The whole thing was incredibly corrupt. Furthermore, after the collapse of BW SE Asian countries kept their own version of the system with capital controls, fixed exchange rates and the rest.

      This was until their late 90s liberalisation they had a massive financial crisis that undid a lot of their work beforehand. So even if that Heritage index has some accuracy, it does not account for their past industrialisation.

      "That's a rather Western-centric view. See http://www.indexmundi.com/g/g.aspx?c=xx&v=67, world GDP per capita is now going up again. "

      It is, mostly driven by a select few countries like China. However that doesn't discount that the financial crisis was a disaster for almost everyone and still is for many.

      "And falling significantly, see http://web.worldbank.org/WBSITE/EXTERNAL/NEWS/0,,contentMDK:23130032~pagePK:64257043~piPK:437376~theSitePK:4607,00.html. To quote:
      "An estimated 1.29 billion people in 2008 lived below $1.25 a day, equivalent to 22 percent of the population of the developing world. By contrast, in 1981, 1.94 billion people were living in extreme poverty."

      Extreme poverty may be falling, but 80% of the world still live on less than $10 a day. No, these things are not 'outside' capitalism; they were originally born as a direct result of it:

      http://en.wikipedia.org/wiki/Late_Victorian_Holocausts

      I'm not sure why economists seem to think Africans are so inept that poverty is their 'natural state' but I can tell it's not; it was largely engineered.

      Delete
    9. Unlearning Econ: You say an awful lot, and my response is over the character limit.
      So, some key points, yes, you noticed hat your textbook assumes that effiency is good. Hardly hidden isn't it? All writers assume some knowledge on the part of their readers, including a basic knowledge of humanity. If you want to question whether efficiency is good, or any other such basic assumption, as an Econ 101 student, do it in class.

      In terms of "a select few countries", these countries include the extremely populous countries of China, Indonesia, Bangladesh, Brazil, South Korea, along with other countries such as Ecuador and Equatorial Guinea. (these being countries in which real GDP capita has grown between 2007 and 2010) (Source data IMF). And these are simply the countries that have published economic data up to 2010, meanwhile say India grew by 10% between 2007 and 2009.

      Meanwhile in the United States, GDP per capita, at US$42,219 in 2010, was back to those dark disastrous days of 2004. In the United Kingdom, the diaster has resulted in GDP per capita back at 2003 levels.
      Loosing your job is often a disaster, and what's happening in countries like Greece is a disaster, but your statements about "a disaster for almost everyone" are ridiculously overblown.

      For someone who recommends replacing economists with biologists and physicists, partly because of their empiricism, you take a remarkably anti-empirical view of the world.

      How rigorous. Meanwhile in the real world, things are more simple than 'not being communist.'

      I presume you meant to type "more complex". But it still is true that "not being economist" is a very mainstream idea. Even you yourself don't dispute it, the debate is now between lassiz-faire capitalism versus whether there is some role for government intervention.

      You criticise Economic Freedom Indexes, but point to a link to a very different index, that's only about the USA. Nor do you recommend a better measure (and I note that industrial policy, or "the rest" are very vague ideas too).

      A lot of your criticisms of economic results are very historical - economists are trying to learn from Russia's privatisation experience, and why that failed while Eastern Europe has been doing so well - that was 20 years ago. You point to the 19th century European colonisalism time, that was over 100 years ago. You point to South Korea and Japan, and there's a lively debate in economics about to what extent their wealth was due to industrial policy versus the basics of respect for property rights. Plenty of countries have tried industrial policy without East Asian successes. See http://eyasusolomon.blogspot.co.uk/2011/07/asias-development-miracle-and-africas.html for a discussion. Economics is hard, testing is done in reality, and it's very hard to sort out what works and what isn't. The events you point to are part of the history that economics is trying to incorporate. That global poverty is falling doesn't prove that economists are now getting better at this, but it's not data in support of your claim that economics is doing terribly.

      Finally I notice you've ignored my question about what the really difficult question is.

      Delete
    10. "perfect competition is the ideal"

      please stop writing this (you often do) it's giving me a stomach ulcer.

      Perfect competition describes a market in which individual producers can sell as much as they want at the prevailing price and their sales have no influence on that price. There is nothing "ideal" about that. There is no reason to think we'd be better off if all markets looked like that, because many (most) of the products and services we want to consume are inherently incompatible with perfect competition. Why, for example, would we want a restaurants to produce food that is a perfect substitute for each others?

      I think you mean that economists tend to advocate making markets "more competitive". For my sake, please change your phrasing.

      And, as has been pointed out to you before, macroeconomics tend not to build models in which markets are perfectly competitive.

      oh god now I've re-read the rest of your comment and there' too much objectionable about it, I have to stop

      Delete
    11. Claudia,

      "Economists focus in on central tendencies and average behavior ... the typical storyline ... and psychologists are often checking out all the perturbations ... the malleability of the story."

      You seem to be suggesting that economic behaviour is, for the most part, on average approximately that of models, and psychologists only model 'deviations' from this. But there is good reason to doubt this narrative: basic psychological concepts like copying, social norms etc. may drive behaviour as whole, not just be 'deivations' from rationality.

      Tracey,

      (2) I say value judgments are hidden because economists do not draw attention to them. See, for example, Tyler Cowen:

      "Economic analysis is itself value-free"

      http://www.nytimes.com/2013/03/17/business/the-egalitarian-tradition-of-economics.html?pagewanted=all

      (2) I'm not sure why you'd say something like "GDP is only what it was in 2004" when it's obvious that growth is necessary to sustain employment. Aggregate GDP figures don't tell us how most people are doing, and in, for example, the UK, unemployment is at its highest since the 1980s, while median incomes have declined in real terms:

      http://www.guardian.co.uk/money/2011/nov/23/uk-household-earnings-fall

      And I was actually primarily referring to the financial crisis' impact on the very poorest in the world:

      http://www.guardian.co.uk/politics/2009/mar/27/global-crisis-impact-poor

      (3) Yes, that index link was completely the wrong one. Here is the one on Heritage:

      http://www.dollarsandsense.org/archives/2005/0305miller.html

      Among the errors are: looking at changes in spending rather than levels; discounting industrial policy; looking at headline rather than effective tax rates; concluding that black markets lower economic freedom (wouldn't these be the epitome of a 'free market'?)

      (4) As for that comparison between SSA and SEA notes, there are large political differences between the two areas: to put it glibly, much of Africa has far less political and social cohesion. Even so - and, again as that article notes - their period with industrial policy outperformed the previous 'Washington Consensus' period. The recent growth can probably attributed to improved political circumstances and a commodity boom. The fact is that Africa has not really industrialised so it's difficult to say that their growth will be sustained:

      http://www.foreignpolicy.com/articles/2013/01/04/the_myth_of_africa_s_rise?page=0,0

      It might also be something Noah noted recently: basically, countries have to take it in turns to use export driven growth.

      (5) I just didn't see the part about the 'really difficult questions.' What I meant was asking questions about capitalism itself. Economic analysis tends to take capitalism as a given and, aside from some vague references to how the USSR failed, does not look far beyond that.

      Luis,

      I agree that a perfectly competitive market would, in reality, be absurd. However, I'm not just throwing that statement around for no reason. Here is a quote from a textbook:

      "First we show how a perfect market economy could under certain conditions lead to ‘social efficiency.’ … [we then] show how markets in practice fail to meet social goals. These failures provide the major arguments in favour of government intervention in a market economy."

      in other words: because, in practice, markets fail to adhere to perfectly competitive assumptions, there is a role for government. Theoretically, at least, perfect competition is the best possible outcome and the need for intervention only arises because the real world fails to meet that. This doesn't mean economists envisage a perfectly competitive world, but within economic theory, there is a judgement that, because it maximise efficiency, perfect competition would be desirable.


      Delete
  15. Noah, you're just wrong. There are macroeconomists and models with excellent predictive power, they just aren't in the mainstream and therefore are routinely ignored. How do you account for Warren Mosler's call in 1996 that the first major demand shock in the eurozone would break the currency when member nations began having financing difficulties?

    How do you account for James Galbraith's use of the sectoral balances model to forecast the dotcom bubble three years beforehand?

    How do you account for 2005-2006 forecasts of a financial crash and recession by Wynne Godley, Nouriel Roubini, Dean Baker, Robert Schiller and Steve Keen?

    How do you account for heterodox models which have batted 1.000 over the last five years on interest rates, inflation, austerity, bond yields and deficits?

    It's patently absurd to suggest macroeconomics cannot reliably provide forecasting when it very clearly can, if one doesn't refuse to open their eyes.

    ReplyDelete
    Replies
    1. Let's say there are a 1,000 models. Let's take five years. By luck, they all have a 50% chance of accurate forecasts.

      By the end of five years, you have about 31 models that work "reliably".

      Let me know how many you have next year.

      By the way, some of the models you're talking about really are cool and noteworthy achievements. On the other hand, simply predicting a crisis or two means nothing.

      Delete
    2. Ashok:

      It's not just a case of luck. What we are talking about is people who get the *causal mechanics* right. That is not just luck.

      Delete
    3. How do you account for heterodox models which have batted 1.000 over the last five years on interest rates, inflation, austerity, bond yields and deficits?

      About the same way I account for stage magicians who strap their assistant to a wheel, set the wheel spinning, and throw knives at the wheel and never hit the assistant. No one is *that* good at either knife throwing or modelling.

      The measurement error in measuring historic inflation and historic deficits alone is so large that I find it deeply implausible that anyone could get 1.000. You'd have to not only have perfectly modelled the underlying economy, but also perfectly modelled the error process of statistics departments in measuring the underlying economy, and the error process of accountants in measuring government spending and revenue.
      See http://mgiakoumis.com/2012/10/22/revision-of-historical-general-government-deficit-and-debt-data/
      And http://www.philadelphiafed.org/research-and-data/publications/research-rap/2011/inflation-targeting-and-revisions-to-inflation-data.pdf

      Delete
    4. Tracy,

      I think the 'batted 1,000' was a rhetorical flourish, and Ben wasn't claiming that anybody has predicted key variables as accurately as you infer.

      Anyway, if you discount correct predictions on the basis of luck, how does the science move forward?

      Delete
  16. Actually, we have a natural experiment in a world without macroeconomists: Germany. I'm only half joking.

    ReplyDelete
    Replies
    1. hee hee. Back last June, after the summit where the EU agreed to "break the cycle of dependency between bank and state debt" - which was perceived as a big defeat for Merkel, there was a huge public argument in the german economics establishment about what the effects of that would be. Public letters signed by dozens of Professors, accusing each other on blogs of being nationalists / deluded / a disgrace to the profession.

      Commenting on the argument, a leading member of Merkel's CDU (Lammert, Parliamentary President) commented acidly (my translation):

      "Of all possible approaches to the crisis handling in the recent period, the gathering of expert opinion has been the least useful... were the political decision-makers to rely on them, they would simply display their inability to make a decision".

      But it was interesting, as a non-economist, to see just how little influence the economics profession actually has, at the moment, in germany. Because in fact, there is a mainstream economic opinion in germany, as articulated by the Council of Economic Experts. And it isn't the policy Merkel is following.

      http://www.voxeu.org/article/european-redemption-pact

      Delete
  17. One thing is clear from this thread, and the mountain of comments that flood in every time Noah does a reasonably technical macroeconomic argument -- people love jocking macroeconomists -- even if only to senselessly bash them. Like I said above, this seems to reinforce my statement that macoreconomists and in-the-street hacks suffer the same problem; they all want the keys to the temple advising governments.

    ReplyDelete
  18. bjdubbs3:29 PM

    What about the bond vigilantes? Sure they're irrelevant today, but most of the time people like Soros and Bill Gross and Buffett and other market participants have more sway than bloggers and commentators. And most of them are Minskyists. In the words of Paul Tudor Jones, “The accumulation and then the repayment of debt basically drives every economic cycle that there is.”

    ReplyDelete
  19. I think other social sciences--especially political science and sociology--might pick up some of the slack from the missing macroeconomists. That would bring good and bad, but it would be critical to counteract the crazies.

    ReplyDelete
  20. Anonymous9:33 PM

    Wanted to point out this little gem:

    "So if there were no academic and Fed macroeconomists around to advise policymakers, who would policymakers listen to on economic matters?

    [lists various "crazy people" who might make up economic policy]

    In short you have a cavalcade of vast unending wackitude, often with a proven track record of wrecking economies and societies.

    So it's possible to see macroeconomists as doing plenty of good, simply by sitting there not being absolute wackaloons."

    I just want you all to mediate on the idea of macroeconomists preventing economic advice from people "often with a proven track record of wrecking economies and societies."

    Just let it sink in.

    ReplyDelete
    Replies
    1. Anonymous,

      LSE economists to India midcentury. American economists to Russia 1990s. European economists to Greece.

      Just let it sink in.

      Delete
    2. Anonymous11:40 PM

      You see my point exactly.

      I just want you all to mediate on the idea of macroeconomists "preventing" economic advice from people "often with a proven track record of wrecking economies and societies."

      Fixed.

      Delete
  21. Oops, almost forgot to mention the cavalcade of wackaloons who openly championed central planning for decades in the 20th century.

    ReplyDelete
  22. You've had an extremely solid string of posts recently. First year of professorship not too challenging eh? :]

    ReplyDelete
    Replies
    1. No, it's just that I'm so amazing that I can crank out an awesome post in a 30-minute snack break. :P

      Delete
  23. Anonymous8:32 AM

    I guess you might have asked the question one hundred years ago:

    "Why do we need engineers at university studying electrical system modelling?"

    The answer is that if we had gotten rid of those studying electrical distribution systems, we would not today be able to model continental electrical systems with millions of electrical flows, pressures, storage etc etc.

    So, if we got rid of the macroeconomists from universities, then the goal of actually being able to model (and hence control) economies would never be achieved.

    The fact is, that early electrical models were clunky and unreliable, and subject to bitter fights (eg CP Steinmetz vs Edison) just like today's macroeconomics, but they were overcome by dint of time and slow progress.

    Keep at it fellas.

    ReplyDelete
  24. I'm reminded of the Hitchhikers Guide to the Galaxy, where the galaxy is governed on the principle that "Anyone who is capable of getting themselves made President should on no account be allowed to do the job". Possibly we need real macroeconomic policy advice to come from anonymous people who have no interest in giving policy advice, while Zaphod Bernanke gives policy pronunciations generated by a specialized macro Eliza.

    ReplyDelete
  25. Anonymous4:43 PM

    "So I'll assume that "economist" means "academic economists, or economists working at government research institutes."

    So, although long ago I obtained my PhD from Michigan in economics, because I work in the private sector advising clients who run large firms or manage multi-billion-dollar investment portfolios, I'm not a real economist? Time to step down from the ivory tower, Noah.

    ReplyDelete
  26. stupid cupid9:05 PM

    margie is dead in is prime...

    ReplyDelete
  27. Bill Ellis10:26 PM

    Noah says..."So someday, macroeconomists might learn how to forecast the economy extremely well."

    I wonder if there might be some sorta "observer effect" that would apply here...

    Can the economy be extremely well forecast with out effecting the outcome ? Won't knowledge of the particulars of the forecast cause people (competitors ) to change their strategies ?

    Not to say that there is not a lot of room for improvement until we have to worry about that.

    ReplyDelete
    Replies
    1. Bill Ellis10:30 PM

      And if it could be forecast extremely well does that mean that it could be commanded extremely well ?

      Delete
  28. Noah,
    as an empirical macroeconomist I would like to thank you for the support. However, when I read that without empirical macroeconomists we wouldn't know

    "that slightly less than half of people seem to be "hand-to-mouth" consumers who don't obey the Permanent Income Hypothesis"

    I can't help thinking that without theoretical macroeconomists (e.g. Friedman) we wouldn't have a permanent income hypothesis to begin with. Other things we wouldn't have are the following:

    A) We would still be picking points on the Phillips curve and trying to curtail cost-push inflation with price controls (a la Nixon following Tobin's rather than Friedman's advice). Because of Friedman's and Lucas' work we recognize the role of monetary policy.
    B) We would not know why governments pursue hyperinflation (e.g. Sargent's work).
    C) We would not know that the use of nominal anchors can reduce the sacrifice ratio (e.g. Kydland and Prescott's work).
    D) We would not be aware of the problem of dynamic efficiency and how its relation to the design of pension systems (e.g. Diamond's work).
    E) We would now know that economies have a tendency to converge to a steady state with constant capital per effective worker and rental rate (e.g. Solow's work). The Marxian theory of capital over-accumulation with ever-falling profit rates would be alive and kicking.
    F) We would not know why middle-income jobs are disappearing and the impact of skilled-biased technological change on rising income inequality (too many contributions to list).

    The list can go on and on, but I hope you are catching my drift. The idea that theoretical macroeconomics is limited to a bunch of DSGE models trying to explain the business cycle, or that the biggest issue in macroeconomics is the size of the government spending multiplier is IMHO inaccurate.

    ReplyDelete
    Replies
    1. I can't help thinking that without theoretical macroeconomists (e.g. Friedman) we wouldn't have a permanent income hypothesis to begin with.

      True!

      A) We would still be picking points on the Phillips curve and trying to curtail cost-push inflation with price controls (a la Nixon following Tobin's rather than Friedman's advice). Because of Friedman's and Lucas' work we recognize the role of monetary policy.

      Ha. I doubt this heavily. Phillips curve failed as a policy tool, then people got attracted to monetarism.

      B) We would not know why governments pursue hyperinflation (e.g. Sargent's work).

      Oh, I'm pretty sure we still don't know that...

      C) We would not know that the use of nominal anchors can reduce the sacrifice ratio (e.g. Kydland and Prescott's work).

      I'm not even sure that's true...

      D) We would not be aware of the problem of dynamic efficiency and how its relation to the design of pension systems (e.g. Diamond's work).

      Agree...but doesn't that originally come from the Samuelson OLG model?

      F) We would not know why middle-income jobs are disappearing and the impact of skilled-biased technological change on rising income inequality (too many contributions to list).

      Hate to burst your bubble, but that's probably not even right...I'm seeing a number of empirical studies contradicting that hypothesis.

      The list can go on and on, but I hope you are catching my drift.

      Well, I definitely see that you and I have slightly different conceptions of the meaning of the verb "to know"... ;-)

      Just because a theory or hypothesis is plausible and seems to generally fit certain stylized facts doesn't mean we "know why" those things happen...

      The idea that theoretical macroeconomics is limited to a bunch of DSGE models trying to explain the business cycle, or that the biggest issue in macroeconomics is the size of the government spending multiplier is IMHO inaccurate.

      Well, I think it's no coincidence that most of the contributions you mention come from before DSGE came and devoured the field... ;-)

      Delete
    2. Noah,

      Yes, people turned to monetarism after Nixon's attempts failed, resulting in higher inflation and higher unemployment. But, due to monetarism, at least they had a viable paradigm to turn to. Is this not important? In fact, without Keynes, who I consider the first macroeconomist, there wouldn't even be the concept of a Phillips curve to debate. We also wouldn't have the concepts of money demand and liquidity preference, and so on.

      Eighty years ago we used models where people mechanistically consumed fixed portions of their current income and where policy-makers could climb up and down a Phillips curve. Today we recognize that people's current consumption decision are influenced by future income, credit constraints, the unceraintly of the path of future income(e.g. under bounded rationality), etc. We are discussing how menu costs, coordination problems, and other factors influence the slope of a dynamic Phillips curve. People who expect yes and no answers may see this as meaningless. But those with a training in scientific research should understand that the change in the nature of the discussion does constitute progress. It allows us to map the conditions under which one outcome is more likely than another. This is significant not only theoretically but also practically. Why do central banks target nominal rather than real variables? Why have they been given relatively more independence? Why do they target interest rates rather than monetary aggregates? These things didn't happen only as a result of practical experience. There is theory behind them.

      Regarding D, according to Wikipedia, Samuelson was in turn inspiried by Irving Fisher. But I am not sure what this proves.

      As far as your objections to B, C, and F go, can you provide more substance? At least as far as F is concerned, are there convincing rebuttals to the work by Acemoglu, Caselli, Autor, Krussel et al, and others? And this is a relatively recent theory. Anyway, I don't agree that DSGE is the cookie monster. :-)

      Delete
    3. Here's your skill-biased technological change (or lack thereof): http://marginalrevolution.com/marginalrevolution/2013/04/trade-vs-technology-in-terms-of-their-labor-market-effects.html?utm_source=twitterfeed&utm_medium=twitter

      Yes, people turned to monetarism after Nixon's attempts failed, resulting in higher inflation and higher unemployment. But, due to monetarism, at least they had a viable paradigm to turn to. Is this not important?

      Well yes, but this is kind of what I was saying in my blog post, about providing an alternative to the crazies ("non-viable" alternatives).

      People who expect yes and no answers may see this as meaningless. But those with a training in scientific research should understand that the change in the nature of the discussion does constitute progress.

      This is kind of what I meant in my blog post about the "option value" of macro research...someday we may find models that work.

      It allows us to map the conditions under which one outcome is more likely than another.

      Well no, I don't really think this is true. Not yet.

      Why do central banks target nominal rather than real variables? Why have they been given relatively more independence? Why do they target interest rates rather than monetary aggregates? These things didn't happen only as a result of practical experience. There is theory behind them.

      I kind of think practical experience was by far more important than theory.

      Also, don't central banks target real output gaps???

      Regarding D, according to Wikipedia, Samuelson was in turn inspiried by Irving Fisher. But I am not sure what this proves.

      Nothing, I was just curious.

      Delete
    4. Remember, there's a difference between predictive theories and Just-So stories. With a Just-So story, you find a story that seems plausible (to you, if not to others) and which seems consistent with past stylized facts, and then you say "We know" why those stylized facts happened. If things change and your story breaks down, you make up a new story. This is not science, nor is it real knowledge, it's just the illusion of knowledge. It's something we do to satisfy ourselves that we understand things that we don't understand...or to suit our political preferences...

      Delete
    5. Noah:

      On that logic, the theory of evolution is merely a Just So story.

      Delete
  29. "Well yes, but this is kind of what I was saying in my blog post, about providing an alternative to the crazies ("non-viable" alternatives)."

    But then this constitutes recognition that established macroeconomic theories do work to at least an extent. Otherwise why not go with the crazies? You can't say what you said, and then say that maybe some day we will find models that work. You need to define what you mean by "work".

    The paper you cite states
    "However, the occupational structure of employment of these labor markets has polarized within each sector, as employment shifted from routine clerical and production occupations to more highly skilled managerial or professional occupations, as well as to lower skilled manual and service occupations." Moreover, it finds the impact of technological change to be more significant in the 1980s, less now. It certainly does not find no impact. And it is examining the impact on employment, not on income inquality.

    Finally, yes some central banks do tarket output gaps. However, not all do. And those who do, do so in conjuction with a nominal target, e.g. the rate of inflation.

    ReplyDelete
    Replies
    1. See also Card & DiNardo:
      http://davidcard.berkeley.edu/papers/skill-tech-change.pdf

      But then this constitutes recognition that established macroeconomic theories do work to at least an extent. Otherwise why not go with the crazies? You can't say what you said, and then say that maybe some day we will find models that work. You need to define what you mean by "work".

      If "work" = "do not, when applied to policy, obviously wreck the shit out of the economy", then sure, modern macro theory works better than crazy stuff! That's my point. But in terms of "theory explaining reality", then no, they don't work...

      Delete
    2. There's all these recent papers casting lots of empirical doubt on the SBTC theory, yet people still cling to SBTC, which I can only assume reflects some combination of A) they've gotten used to thinking it was true for quite a while, B) it fits their political preferences, and C) it seems intuitively plausible.

      Delete
    3. Well, when on one hand you have a ton of both theoretical and empirical papers documenting the transformation of the labor market during the SBTC and on one hand you have the Card&DiNaro (and maybe a few more) simply raising a few questions what do you expect people to believe? Read Acemoglou (1999), read Caselli (1999), read Autor's report for the Center for American Progress (hardly a conservative institution here: http://noahpinionblog.blogspot.com/2013/04/a-world-without-macroeconomists.html#comment-form
      read even the first paper you cited by Autor, Dorn, and Hanson here: http://economics.mit.edu/files/8763
      All of them are bursting with empirical evidence!

      Delete
    4. Oops, Autor's paper for the center for progress is of course not on your blog, but here:
      http://www.americanprogress.org/wp-content/uploads/issues/2010/12/pdf/autordugganpaper.pdf

      Delete
    5. ton of both theoretical and empirical papers

      Why are a lot of theoretical papers more persuasive than one? that's just the weight of a whole lot of respected people repeating the mantra that the story looks plausible.

      simply raising a few questions

      Translation: "providing evidence that contradicts my priors"

      read Autor's report for the Center for American Progress

      The proposal for modifying Disability Insurance?

      Delete
    6. Darn, bad idea to comment in-between classes! Here is, I hope, the correct link.
      http://economics.mit.edu/files/5554
      It is on labor-market polarization, a theme that Autor also explores in the paper cited in marginal revolution.

      A lot of theoretical papers are more pursuasive to the extent that the outcome is invariant to the framework (and therefore the assumptions) that the models makes. In this case, the impact of SBTC is similar regardless of whether one starts with a typical neoclassical model (Walrasian auctioneer, workers get paid their marginal product) as in Krusell et al (2000) or a disaggregate model with costly search and monopoly power as in Acemoglu (1999). Obviously, this strengthens the hypothesis.

      "Translation: "providing evidence that contradicts my priors""
      No, it means that statements like "gee whiz, but the wage of engineers relative to social scientists has not increased" or "gee whiz, but the rise in wage inequality slowed down in the 1990s despite the continuing progress in IT" do not stack up well against a much bigger volume of empirical examinations that use more detailed and disagreggate data. Card and Dinardo admit as much when they "acknowledge that these predictions may be oversimplistic." And they also admit that the versions of SBTC they examine (and there are more) in fact "are consistent with some of the changes that have occurred over the past 3 decades..." When priors are based on a much bigger load of empirical evidence, then yes, simply raising questions based on eyeballing a few charts simply will not do.

      Delete
    7. Oh, labor market polarization is definitely real, as is the rising education premium (at least, during the 80s). But SBTC is a hypothesis to explain that fact. Education premia might have risen because of free trade, and increased supply of uneducated workers living in other countries.

      A lot of theoretical papers are more pursuasive to the extent that the outcome is invariant to the framework (and therefore the assumptions) that the models makes...Obviously, this strengthens the hypothesis.

      I don't think it really does...

      a much bigger volume of empirical examinations that use more detailed and disagreggate data...a much bigger load of empirical evidence

      Really? Where is this load? Do they cite it? Where can I find it?

      Delete
    8. "But SBTC is a hypothesis to explain that fact. Education premia might have risen because of free trade, and increased supply of uneducated workers living in other countries."

      But it is not either one or the other! Read the paper by Autor et al that you first cited. By looking at local labor market they decompose the two effects. This also answers you last question. But to add to it, the theoretical papers I cited, Caselli (1999), Acemoglu (1999), Krusell et al (2000) also present empirical evidence. Then you have the work by Autor on polarization, which also supports SBTC as a source of rising inequality. On the other hand, the questions that Card and Dinardo (C&D) raise can be answered quite easily. You fail to acknowledge that the paper by C&D was written in 2002, before a number of empirical studies by Autor and others (interestingly, C&D do not cite the papers by Caselli and Acemoglu). Even Krugman seems to think that the evidence is there:
      http://krugman.blogs.nytimes.com/2011/03/06/autor-autor/
      So yes, the load is there for anyone who is more interested in finding it than finding excuses to bitch and moan.

      Delete
    9. By the way, I am not sure how Autor would describe himself but Katz and Krueger (currently advising Obama) are not macroeconomists, but rather labor economists. So I suppose your criticism of macroeconomics extends to the entire discipline.

      Delete
    10. OK I'll look over these papers and write a post on SBTC...

      Delete
    11. Now you are talking baby! Criticism is necessary, but there is a lot of exciting research going on also. Let's talk about some of that once in a while.

      Delete
    12. Of course, I don't want to do too much of that, because I'd critique people's methodologies (like in a seminar), and then they'd think I was a jerk and I'd get a very bad reputation... ;-)

      Look what happened when I critiqued that Acemoglu paper on entrepreneurship and risk!

      Delete
    13. Ha ha, maybe you are of a more critical nature than I am.

      Delete
  30. Anonymous4:06 PM

    Give us your call about Thatcher legacy.

    ReplyDelete
  31. Question: When a blogging macroeconomist posts a long defense of his professional commitments and collects everyone else who's ever commented on the macroeconomy into a bin of crazies, doesn't it kind of set up a dichotomy where anyone who opposes his view goes right into the crazy bin?

    I have to at least hand it to Noah for setting up such a slick dichotomy whereby he can argue chiefly from authority and swat away any criticism before it's even made. I played right into it.

    Regards from the Crazy Bin

    ReplyDelete
  32. Hi Noah,

    I've been following your blog for a while now, I really enjoy it. I think you often ask the right questions about what economists can do and how to improve their work, and you often end up by saying that their work should be more empirically based (like here).

    I think this is very true. But I often feel an urge to tell you (so I'm doing it here at last!)that there is a type of researcher who focuses a lot on what really happened in the past: economic historians. Some of them have proper economic training, and even teach economics, but they also try to draw lessons from the past in the most balanced way possible (they are afraid of generalisations - which can be a problem in itself).

    The problem is that their work (like mine for example) can't provide one big macro model all at once, one that would explain everything. But I think a careful study of past economic policies and their results can help. There is a trade-off between the level of detail that can be reached in analysing one or two instances of an economic phenomenon and the amount of generalisation that can be done by undertaking a study of say 30 countries over 50 years (economic historians tend to do the former, while empirical economists tend to do the latter).

    However I think both types of researchers have a lot to learn from one other, and more cooperation between the two fields would be highly beneficial. In the past economists used to work a lot more on economic history (think about Keynes and Hayek who studied the Great Depression, or Friedman and Schwartz who wrote an entire monetary history of the US). Maybe the trend should be revived?

    Best,
    Natacha

    ReplyDelete
  33. Jamie1:54 PM

    I am not an economist, and I haven’t commented on your blog before, but I like this post. You are asking good questions.

    My professional background involved helping businesses and government departments solve a variety of problems. As a result, I see the macro-economics profession in the same way that I’d view a failing organisation full of good people who have nevertheless lost sight of their main purpose.

    People in failing organisations often focus on “insider” questions. These are questions which are of interest only to insiders and which can be answered only by insiders. They act as a comfort blanket. In macroeconomics, you see this when, for example, economists debate with each other whether one sort of macroeconomic model is better than another type of model. Insider questions rarely get at the root causes of failure and mostly just create factionalism.

    One of the reasons that people like me can help failing organisations is that we ask “outsider” questions. For example, what services do you provide; who are your customers; are your services useful to your customers; how do you measure the effectiveness of your services; how could you improve your services; what would constitute progress? Outsider questions are more effective questions because they are more fundamental and more challenging.

    "What would happen if there were no economists" and “what is the purpose/benefit of the academic macro-economics profession” are outsider questions so it’s refreshing to see you ask them. I have three comments though.

    Firstly, if you ask most non-economists about the purpose of the academic macro-economics profession, they are likely to say something about educating the rest of us effectively in how the macro-economy works. I notice that is missing from your answer. Why is this? Why do you think the general state of public economic literacy is so low? Why do academic economists show such little interest in helping to do anything about this?

    Secondly, I doubt that many non-economists would agree that macro-economists constitute “an anchor that keeps policy away from the dangerous extremes urged on us by various politically-motivated fringe groups”. Indeed, part of the problem is the inability of the economics profession to present a single face to the rest of us. It is difficult for non-economists to tell the difference between empirical economists and ideologically motivated economists. Pretty much every macro-economist says “I am an empirical economist. My friends who agree with me are empirical economists. All of the people who disagree with me are ideologically motivated extremists”.

    Thirdly, I don’t know what something like “the creation of spinoffs like sVARs” means. You are answering an outsider question with an insider answer. That’s not allowed. Your answers will be useful only if they are meaningful to outsiders and expressed in plain language.

    Nevertheless, you should continue to ask provocative questions. You mention chemists in your post. Chemists became useful only when they stopped pretending they were superheroes who could turn lead into gold and instead just tried to understand and explain how stuff works.

    ReplyDelete