Tuesday, August 23, 2011

Common Sense and Macroeconomics

This WSJ article by Stephen Moore was ridiculed by Paul Krugman and David Glasner as anti-intellectual, and Noah Smith used the article in an attempt to ridicule Ed Prescott.

Moore's basic argument is that Keynesian economics does not satisfy the rules of common sense, whatever that is, and he gives some specific examples. Moore's discussion is somewhat confused. In particular, he seems to think that modern macroeconomics and Keynesian economics are synonymous. He also argues that, in analyzing unemployment insurance we should only be thinking about the incentive effects, and apparently not about why the government should be supplying the insurance, the role of UI as a transfer, or how the financing of the transfer could matter. However, it's not surprising to me that Moore is confused, and I have some sympathy for the poor guy.

Indeed, one could get the idea from reading the blogosphere that modern macro is dominated by the legacy of Keynes, when the truth is that most practicing macroeconomic researchers are not spending their time thinking about multipliers and the paradox of thrift. There are indeed ways in which Old Keynesian economics - basic Keynesian cross and IS/LM - defies common sense, i.e. the common sense that comes from standard microeconomics.

What is common sense anyway? It has to be something most of us possess, clearly. Don't put your hand in a fire. Don't cross the street without looking both ways. The latter piece of information came to me along with: Don't run out between parked cars. There were no parked cars where I lived. Everyone parked in the driveway or the garage. Therefore, that did not apply to me. Looking both ways before crossing the street must also not apply to me. When I was 8 years old, I crossed a highway without looking and was run over by a car. I have looked carefully ever since.

When I first took economics, here is what common sense was for me. It was the tail end of the hippy era and I thought that people in suits were greedy bad people out to steal from the poor. Rent control was a good thing as it took money from greedy landlords and gave it to poor apartment-renters. Wage and price controls seemed like a good way to control inflation, though I did not quite understand what it was that was inflating. And so on.

In microeconomics, my professor convinced me that a simple supply/demand apparatus determining equilibrium prices was a useful way to think about price determination, and he showed us how this apparatus could be used to make sense out of actual prices and quantities that I could observe. Greedy people could actually be getting rich and performing a useful social function. Rent controls could actually harm everyone - rich and poor - in the long run. Wage and price controls could screw up the allocation of resources in a severe way. There is no free lunch. Everything comes at a cost.

Then I took macroeconomics which, in terms of the machinery I had built up in taking micro, made little sense. If competitive equilibrium with market-clearing prices was so useful in microeconomics, why couldn't we do that in macro? What is this paradox of thrift? What's this nonsense about the multiplier free lunch? Why did I have to un-learn all the micro I had learned in order to do macro? It took another 6 or 8 years and many readings and re-readings of work by Lucas, Wallace, Prescott, and others, before everything finally made sense.

There is nothing easy, natural, obvious, or common-sensical about Keynesian economics. If you try to do it properly, or try to turn it into a story, you come up with more questions than you can answer. Why would we think that firms would shut their doors or would-be workers would remain unemployed because of an unwillingness to lower their prices and wages, respectively? Why would a firm, faced with an increase in customers, increase its output and not its price? Why would many economists, faced with phenomena that appear very un-Keynesian, and with plenty of sophisticated tools available, revert to the rustiest old contraption in the shop?

27 comments:

  1. When asked why they aren't hiring, firms say they have no demand. When asked why they aren't spending, households say they have no jobs, or worry they might not. Sounds like a classic Keynesian demand-driven recession to me. What are the "very un-Keynesian phenomena" you had in mind?

    ReplyDelete
  2. “Why would we think that firms would shut their doors or would-be workers would remain unemployed because of an unwillingness to lower their prices and wages, respectively?”

    One reason why I would be reluctant to tell the boss that we should cut prices, in order to shift some excess inventory, is that I can predict what the boss will say pretty accurately: Wimps always propose cutting prices; get up off your arse and find some customers. It’s his prerogative to decide when it’s time to cut prices, just as it’s a naval captain’s prerogative to order the crew to strike the colours. For a somewhat similar reason I wouldn’t try to get a job by approaching an employer and saying: Whatever you’re paying that guy over there, I’ll do just as good a job for a few cents less. People usually don’t respond very well to offers like that. Quite often they think: If he’s so keen to stab my employee in the back, how long before he does it to me?

    Ideally I would like an economic theory which recognizes that we are social animals. Serious microfoundations would explain why we behave differently from chimps, who resemble us closely in some ways, and dogs who resemble us in others. But economics is a long way from being able to reconcile itself with biology. Meanwhile, microeconomics has inadequate macrofoundations and vice versa.

    ReplyDelete
  3. "It took another 6 or 8 years and many readings and re-readings of work by Lucas, Wallace, Prescott, and others, before everything finally made sense."

    Five thousand years of readings could not make the "General Theory" make sense, as can be ascertained in half an hour by reading even a portion of Hazlitt's line-by-line refutation.

    Your story of George and Martha ignores a number of important facts about survival and psychology. Would these facts go away if I was to read Wallace and Prescott for 6 years?

    ReplyDelete
  4. "What are the "very un-Keynesian phenomena" you had in mind?"

    Where are the sticky wages and prices?

    Kevin,

    Some individuals have a lot of power. I am not the same as a dog. Where do you go with that?

    Paul,

    "Your story of George and Martha ignores a number of important facts about survival and psychology."

    Such as? What do you have in mind that is so important to the problem at hand?

    ReplyDelete
  5. Where are the sticky wages and prices?

    In my local branch of Tescos. The price of bread has only changed once in the last 12 months. Do you have some sort of measure of price stickiness which is supposed to cause Keynesian economists to re-think their ideas? I'm pretty sure they know that prices change now and then.

    I am not the same as a dog.

    I didn't mean to suggest that you are.

    ReplyDelete
  6. PA: "Your story of George and Martha ignores a number of important facts about survival and psychology."

    SW: "Such as? What do you have in mind that is so important to the problem at hand?"

    Here are a few:

    It ignores the fact that George is less likely to want to grow peanuts if he knows the government is going to supply him with other means to obtain cranberries.

    It ignores the fact that, since he eats nothing else, George must get out of bed at some point to avoid starving, and so there is no need for the government to go into debt to get him out of bed.

    It ignores the interest burden on the taxpayer.

    It ignores the fact that over time the peanut debt burden on taxpayers may be unpayable in the peanuts that they are able to grow.

    It assumes that monitoring millions of people is as easy as monitoring one person.

    It assumes that the cost of monitoring George and forcing him out of bed is zero.

    It posits a government that can force people out of bed to dig holes, and assumes that the cost in terms of loss of liberty is zero.

    It ignores the fact that other people, not George, will pay for the cranberries supplied by the government.

    It assumes that a price must be negotiated the night before, not agreed after George gets out of bed.

    It starts off using money prices, but then invents bonds denominated in peanuts.

    It doesn't allow the possibility that George might want to take a turn of growing cranberries himself to relieve his boredom with growing peanuts. (i.e. it tries to solve a problem using the government when the private sector has potentially more effective solutions).

    By solving his problem for him, it takes away the incentive for George to solve his own problem.

    It disempowers and de-educates George by attempting to solve his problem for him.

    ReplyDelete
  7. Kevin,

    "In my local branch of Tescos. The price of bread has only changed once in the last 12 months."

    No, that's the standard answer. What I want you to do is to show me how, in the current state of the world, price and wage stickiness at particular establishments is related to low output and unemployment at those specific establishments. For example, housing construction is in the toilet. But I don't think you want to suggest that the prices of houses have failed to fall, or that roofers are unemployed because their wages are sticky. Maybe you have something else in mind.

    Paul,

    The story includes all the key elements that Old Keynesians argue are important. Prices are sticky - it's difficult to change them. There is lack of Ricardian equivalence, i.e. people fail to take account of their future tax liabilities. People are failing to make Pareto-improving trades because they got the prices wrong. If you don't like the story, then you don't like Keynesian economics, which is the whole point.

    ReplyDelete
  8. "The story includes all the key elements that Old Keynesians argue are important. Prices are sticky - it's difficult to change them. There is lack of Ricardian equivalence, i.e. people fail to take account of their future tax liabilities. People are failing to make Pareto-improving trades because they got the prices wrong. If you don't like the story, then you don't like Keynesian economics, which is the whole point."

    So the point of the article was that people who don't like that story don't like Keynesian economics?

    I thought the point was that these stories, although valid, don't explain recessions?

    If so, then this leaves open the question of whether you use this story as an explanatory vehicle in educating people about real world economics.

    If you do use it as an explanatory vehicle, either to others or to yourself, it is important to explain, and keep in mind, the many ways in which the story is not representative of the real world.

    If on the other hand it's used simply to explain the way Old Keynesians think, then I think it's a good story.

    ReplyDelete
  9. What I want you to do is to show me how, in the current state of the world, price and wage stickiness at particular establishments is related to low output and unemployment at those specific establishments.

    Who says it is? A theory which predicts what happens at particular establishments is micro, not macro. If someone presents me with a Walrasian model I'm not going to say "show me a brewery turning down business because the marginal cost of supplying one more keg of beer exceeds the current market price." I'm fairly sure that no such brewery exists, but I wouldn't reject a macro model based on a price-taker assumption on those grounds.

    ReplyDelete
  10. Heyyo...My post was not ridiculing Ed Prescott! My post was pointing out that neoclassical econ is actually dominant in the field, and that Moore would like it! :P

    (Besides, I've already ridiculed Ed Prescott in PLENTY of posts. Why beat a dead horse? Heehee.)

    ReplyDelete
  11. "Who says it is?"

    Mike Woodford.

    Noah,

    I don't know. Sure looked like ridicule to me. Maybe a poor choice of words on your part? Anyway, what I'm thinking is that I should withdraw the offer of admission. Based on the information I have now, I don't think you would cut it. You certainly don't have Prescott nailed down. What you are calling assumptions is what's in the calibration. You may not like how he does it, but he was pretty careful about choosing parameters based on the growth and micro evidence. He didn't just assume elasticities. Add your misunderstanding there to your notion (from the "pulpification" piece) that an "implication" could include markets clearing and such. When we discuss "implications" of models, what we mean is things the models imply about what we should observe in the data. Heehee.

    ReplyDelete
  12. Fair enough, I've read almost nothing by Woodford, but if you've got a compelling argument against Woodfordian economics he should address it.

    ReplyDelete
  13. I found this graphical evidence of sticky wages mighty persuasive:
    http://www.themoneyillusion.com/?p=10518

    Bryan Caplan links to more evidence of downward nominal wage rigidity.

    ReplyDelete
  14. anonymous,

    I'm not sure where the persuasion is. In the post you link to, I'm not told what that graph is about (e.g. what's the variable on the horizontal axis?) or where the data comes from. I'm looking for something more serious.

    ReplyDelete
  15. Here's a link to a previous post on sticky prices:

    http://newmonetarism.blogspot.com/2011/05/sticky-prices-and-keynesian-narrative.html

    ReplyDelete
  16. "One reason why I would be reluctant to tell the boss that we should cut prices, in order to shift some excess inventory, is that I can predict what the boss will say pretty accurately: Wimps always propose cutting prices; get up off your arse and find some customers."

    But here's the rub -- why do you think you can predict this response? He will only do that if he thinks it will make him personally better off. In other words, he won't change prices precisely because it is not optimal to do so.

    ReplyDelete
  17. Wow, I can't believe I actually checked this comment thread to see if someone replied to my comment. I am WAY too bored.

    Esteemed Professor Williamson, the Frisch elasticity looks a whole lot smaller in the micro data, micro data to which apparently Edward C. Prescott did not give much credence...when you have two conflicting sources of evidence for the values of a parameter and they are so much in conflict, then either you assume that one source has the right of it (as Dr. Prescott did), or you do robustness checks and report those in your paper (as Dr. Prescott did not). Struck me as slightly fishy, since Prescott was trying to explain why Europeans work so few hours, and the micro data is all about the intensive margin, but whatever, he is the Nobelist and I am the Nobody.

    Regarding "implications", semantics aside, you can look at the data to see if markets clear (they do sometimes) and if the law of one price holds (it does most of the time), and you can test whether people behave as price takers (which they often do not), and so you can see how appropriate the general equilibrium framework is when analyzing a given market. If general equilibrium had zero implications, then data would not be able to tell us when GE is and is not appropriate for modeling (i.e. GE is not simply a mathematical technique). If GE had no implications we would be able to use GE to model any situation that we can model with game theory. But we cannot do that.

    Also, I believe you have misused the common internet phrase "heehee". Usually the phrase is meant to indicate that the writer has amused himself with the sentence or paragraph immediately written before the "heehee". However, your pre-"heehee" paragraph was written in an irritated tone, leading the reader to believe (honestly, if perhaps incorrectly) that you were writing out of irritation instead of out of a desire to amuse yourself. Hence the "heehee" seemed strained, or forced. I understand that members of older generations may not understand the slang used by denizens of the "internet," so I am just trying to bridge the generation gap.

    As for the revocation of the offer of admission, I admit I am not extremely disappointed, as I always seriously doubted that you had the authority to make transfer admission decisions at Wash U. Thus I never got my hopes up, and I fortuitously avoided being emotionally devastated when you withdrew the originally dubious offer.

    Finally, note that if I were to use "heehee" right here, it would also be out of place, as my previous paragraph employed a tone of "laconic sarcasm" rather than a tone of "snarky taunting".

    Peace out, homez! :)

    ReplyDelete
  18. Esteemed Noah,

    1. No, you haven't been reading your Prescott. Read "Theory Ahead of Business Cycle Measurement," and Richard Rogerson's related work, which is particularly helpful on the elasticity issue. That work is all about how your reconcile measured labor supply elasticities at the individual level and the aggregate level, and intensive vs. extensive margins. Get someone in your department to help you with references.

    2. "you can look at the data to see if markets clear" Wrong. "you can test whether people behave as price takers" Wrong again. "If GE had no implications we would be able to use GE to model any situation that we can model with game theory. But we cannot do that." Nonsense.

    3. No, I used "heehee" correctly. Not sure where you read "irritation" into it. I was in fact amusing myself, just as I am now.

    Yes, it has been a slice. Go read your Prescott, and get some work done on that dissertation too.

    ReplyDelete
  19. A bit hard on Noah there Stephen...

    ReplyDelete
  20. Did I just read a professional economist write that the two statements:

    "you can look at data to see if markets clear" and "you can test whether people behave as price takers" are wrong......?????

    Wow.

    ReplyDelete
  21. "Wow."

    That should not startle you. You cannot look at a "market" and say it does or does not conform to the competitive paradigm. Maybe Noah thinks that you can observe shortages or inventory accumulation and it looks like the market is not clearing, but those things are consistent with optimization and equilibrium.

    On testing price-taking, maybe Noah thinks you can see firms selling the same stuff at different prices, so that looks like non-competitive behavior. But in Arrow-Debreu we can index goods by location and date, which permits any amount of dispersion in prices you could want.

    Noah mentions game theory. Arrow Debreu places no restrictions on what we should see in the data. Game theory the same. Of course, once we are into the particulars of specific models there are things I can do with a game-theoretic construct that I cannot do with the competitive paradigm, and vice-versa. Game theory allows me to think about problems where strategic considerations are important. Competitive theory is better at giving me sharp predictions, as game-theoretic models often have multiple equilibria.

    ReplyDelete
  22. I'm not an economist, so the NBER paper is paywalled for me, but I was presuming Krugman & Sumner's reporting was accurate. Sumner calls it a "pay increase dispersion graph". The horizontal axis is marked adjrate1_growth, suggesting that it indicates how much wages have grown. There is a glaring discontinuity at zero, which is reported as indicating no change in the nominal wage. Krugman describes the graph as showing downward nominal wage rigidity. The abstract describes the paper as measuring (among other things) "the probability that an individual will experience a nominal wage change". Could be that I'm wrong, it was misreported by everyone I read and something completely different is being measured, and New Classical economics (is what what you do?) is completely vindicated from Sumner's charge. If that is the case I would be interested in hearing just what is going on. I found your sticky prices post interesting (I actually can't recall anyone else giving the comparative stickiness of different goods), so a similar explanation of the reported wage analysis would be greatly appreciated.

    Noah, Sumner discussed how Prescott's macro data can be reconciled with micro data here:
    http://www.themoneyillusion.com/?p=4692

    ReplyDelete
  23. I see. It's an interesting observation, but it's a long way from the observation to the conclusion that, if we increase government spending by $1, then GDP increases by $2, say, or that monetary policy actions have a big effect on real GDP. It is possible, for example, for prices to appear sticky, when money is actually neutral. See this paper:

    http://economics.sas.upenn.edu/system/files/10-034.pdf

    With regard to wages, labor contracts are complicated. Observed wages can perform many roles, in addition to the way we think about a wage as being a price that adjusts to clear the market. Wages provide incentives, and they may be smooth over time for an individual due to implicit insurance arrangements between firms and workers. Some of the wage adjustments you are seeing in the picture are there for incentive reasons - e.g. upward adjustments as rewards for good performance. In the downward direction, there is always termination for really bad behavior, and those people drop out of the sample. There is a lot of mobility in the labor market, particularly for high-ability people, so observing wages at a particular establishment may give you a biased picture of what is going on. There are other ways to compensate workers than with wage payments. Further that picture could change a lot if the inflation rate changes. For example, if you get a persistent deflation, people will start to think of nominal wage cuts as normal.

    On top of all that, the key macroeconomic policy question is: what happens when policy changes? To evaluate that, you need a good model that takes account of how behavior changes with the policy change.

    Conclusion: That picture may seem suggestive, but there is a lot of careful economics that has to be done before I can jump from that to a policy conclusion.

    ReplyDelete
  24. "When I was 8 years old, I crossed a highway without looking and was run over by a car. I have looked carefully ever since."

    How long before a commenter blames this event for your criticism of Krugman?

    ReplyDelete
  25. Yes, dropped on my head at an early age too.

    ReplyDelete
  26. Noah calls Ed "Dr Prescott" in an attempt to mock him. Sadly for Noah, there will be no "Dr Noah Whatever", since he'll never be graduated given his poor understanding of economics.

    ReplyDelete