Lucas and Fama

Over at Econlog I have a new post discussing the legacy of my favorite economist—Milton Friedman.  Here I’d like to discuss how my view of macro has been shaped by Robert Lucas and Eugene Fama, two other famous Chicago economists.  In some ways these are odd choices. I focus on demand-side models of the business circle, and Fama seems to think AD doesn’t matter.  And I never even took a class from Fama. I did take several classes from Lucas, and he was my PhD advisor.  But I never really bought into his “microfoundations” approach to macro.  My “musical chairs” model of unemployment is about as far from Lucas as you can get.  He’d be horrified.  But great ideas can show up in all sorts of surprising and unexpected places.  And even though I think that Lucas and Fama overestimate the flexibility of labor markets, I believe they are right on the mark in some other important areas.  Here’s what I learned from Lucas:

1.  The long run is now:  I used to have a lazy belief that the term “in the long run” meant something like “in the future,” and short run meant the present, or very near future.  Lucas taught me that the long run is (also) right now.  The term ‘long run’ refers to situations where factor X effects factor Y with a long delay.  It has nothing to do with present and future.  And yet I often hear even professional economists talk as if the term “in the long run” meant in the future.

Here’s one example.  The short run effect of monetary policy is called the liquidity effect.  A change in the money supply causes interest rates to move in the opposite direction, in the short run.  The long run effects of monetary policy are called the income effect, the price level effect and the expected inflation (Fisher) effect.  These three effects all cause interest rates to move in the same direction as the money supply.   If you erroneously thought that “short run” meant “now” you would interpret any current change in interest rates through the lens of the liquidity effect.  That is, you’d assume falling interest rates reflect an easier monetary policy, and vice versa.  Lucas taught me that what’s going on right now is equally likely to be the long run effect of policies that happened earlier.  Thus falling interest rates might just as well be the long run effect of an earlier tight money policy.  How can you tell which is which?  Ben Bernanke says you look at NGDP growth and inflation.  And what happened to those two indicators in 2008-09?

2.  Think regimes, not discretionary policy choices:  It’s tempting to debate monetary policy without reference to a policy regime. People debate “what should the Fed do now,” as if the question makes sense in the absence of a clear policy framework.  Now in fairness there are some policies that are so obviously bad, under such a wide range of plausible policy regimes, that it’s tempting to just come out and say “money is too tight” (i.e. 1931) or “money is too easy (i.e. 1979).  But in most cases the discussion pits one plausible policy against another, with no agreed upon destination.  Unfortunately people play lip service to the need for clear policy rules, but in practice they don’t buy into this approach because they don’t trust policymakers to have the right policy rule.  Thus the Fed minutes show a confusing debate over how best to get to the right destination, among policymakers who don’t even agree on the correct destination.

And people may have good reason for not trusting the policymakers to adopt the correct rule, look at the BOJ in the 1993-2012 period, and the ECB in recent years, both clearly aiming at the wrong policy target.

3.  Rational expectations:  It makes no sense to have a model of the economy that assumes X, but which is filled with people who believe “not X.”  As Bennett McCallum pointed out this idea is better called “consistent expectations,” as the term “rational” has all sorts of connotations that have nothing to do with the public’s expectations being consistent with the model of the public’s behavior.  Rational expectations also underlies the “Lucas Critique,” the idea that a statistical relationship that is true under one policy regime, may not hold up if the policy regime is altered to take advantage of that statistical relationship.

4.  The “voluntarily unemployed” might still be miserable:  If someone loses a job as an accountant and turns down an open position at McDonalds, they might be considered “voluntarily unemployed.”  That makes me agree with Lucas that the term is pretty much useless.  I’d add that I don’t find either side of the “deserving/undeserving poor” debate to be making useful arguments, for similar reasons. I focus on how policy affects outcomes, and leave to God the question of who is or isn’t deserving of more out of life.

Fama is famous for the efficient market hypothesis, which underlies several important pieces of market monetarism, or at least my peculiar version of MM:

1.  No wait and see:  When a new policy initiative aimed at boosting AD is announced, it makes no sense to “wait and see” if it will work. The market reaction immediately tells us the expected impact of the policy, and anything different that occurs will reflect unexpected shocks that violate the “ceteris paribus” assumption.  If the Fed was expected to cut rates by either a 1/4% or 1/2%, and the fed funds futures market assigns probabilities to each outcome, then the actual response of TIPS spreads and stock prices to the policy announcement tells us almost all we will ever learn about the policy.  If we had a NGDP futures market we could even do better, but the markets we do have give us a pretty good idea of the impact of unexpected policy announcements.

2.  Target the futures price of the policy goal:  It’s silly to have intermediate targets such as the fed funds rate, or the exchange rate. Simply adjust the monetary base as necessary to keep the futures price of the policy goal variable on target.  But of course first you need to create a futures market for the policy goal variable (preferably NGDP.)

3.  Bubbles?  No such thing:  Or more precisely the bubble theory is completely vacuous–it doesn’t help us to better understand the world around us.  When house prices soared in the early 2000s in Canada, the US, Australia, New Zealand and Britain, people cried bubble. That did not help me to understand why in Australia they later soared even higher, in the US they plunged lower, and in the other three they mostly moved sideways.  When Bitcoin soared from $1 to $30, people cried bubble.  That told me nothing useful about why prices later rose to $1000 and then fell to $600.  When Robert Shiller said stocks were a bubble in 1996, it told me nothing useful about the future path of stock prices.  And I could go on and on.

4.  Ignore the financial system:  I know what you are thinking.  ”Wait a minute—surely Fama never said to ignore the financial system, that’s what he spent his whole life studying.”  But he did understand that money (MOA) and credit are completely different entities.  He suggested that the Fed could stabilize the price level (and by implication NGDP) simply by controlling the currency stock.  In a sense he was talking about control of the monetary base, but in those days the base was almost entirely currency and required reserves, and he thought reserve requirements were a silly regulatory policy that could be dispensed with.  So control of the stock of currency was all you need in order to control the value of currency.  And since the value of currency is (by definition) the inverse of the price level, that’s all you needed to control the price level, or any other nominal aggregate.  Monetary theory without banking and finance—the 2008 financial crisis shows us how important it is to divorce these concepts.  Economists who see them as being linked got 2008 all wrong.  They thought the real problem was banking distress, when in fact the real problem was nominal (GDP.)

MMTers should not read Fama’s currency paper; it would give them a heart attack.


Tags:

 
 
 

75 Responses to “Lucas and Fama”

  1. Gravatar of John Hall John Hall
    18. March 2014 at 05:57

    This post might be improved by adding some references to papers by Lucas and Fama related to those points.

  2. Gravatar of Ilya Ilya
    18. March 2014 at 06:12

    Scott,

    Do you think one day you might give us a reading list of such papers that have influenced you? Perhaps what you would include in a graduate syllabus. I’ve been meaning to ask you if you have a graduate macro/money syllabus that you might share.

    Ilya

  3. Gravatar of Don Don
    18. March 2014 at 06:47

    “If we had a NGDP futures market” begs the question, why isn’t there an NGDP future’s market? It seems there is a market for everything in either New York or Las Vegas. Surely there is value in having such a market. It would be the perfect hedge. Is it because NGDP is not easily measured?

  4. Gravatar of TravisV TravisV
    18. March 2014 at 07:03

    I have seen the future and it is this blog post!

    Mark it down, in 5 years this post will look brilliant, in 10 years even MORE BRILLIANT, etc. etc…..

  5. Gravatar of benjamin cole benjamin cole
    18. March 2014 at 08:09

    Excellent blogging.
    So…Fama says no bubbles and EMT says no bubbles…why is the Fed so concerned about bubbles…as they sound silly being concerned about inflation?

  6. Gravatar of Ilya Ilya
    18. March 2014 at 08:10

    How about a bitcoin futures market?

    Here’s one… https://icbit.se/

    Another… https://www.predictious.com

  7. Gravatar of Major_Freedom Major_Freedom
    18. March 2014 at 09:04

    1. So long and variable lags are alive and well on this blog, contrary to the claims of some of your readers who told me otherwise.

    2. It’s more useful to think in terms of cui bono. Can the larger, more politically connected banks easily gobble up smaller banks if NGDP remains stable through recessions? How does this weigh against the benefit of “Fed friends” in Congress whose economic role is to continue the Fed charter and require something politicially self motivated in return? Price targeting wasn’t enacted because of economist consensus. It was enacted because it was believed banks in the cartel could more efficiently inflate together at roughly the same pace, and the gradual increase in reserves upon which price targeting is based, was believed to be the most effective means, so as to avoid individual fractional reserve banks from expanding too much relative to other banks.

    3. The notion that an assumption of consistent expectations about the future economy is required in order to have any model of the economy at all, should have been rejected for that very reason. Every model has its detractors. Yes, even the model that every model has its ditractors, has its ditractors. Rational expectations is not realistic. Indeed, it is quite easy to learn that disagreements are not merely a burden. They are fundamental to the existence of exchanges. Trade takes place because of disagreements over the values of the things exchanged. I think $5.00 is worth less than a hamburger, and someone else thinks the hamburger is worth less than the $5.00. That is why we think of trading. This same principle is true for expectations. If for whatever reason everyone had the same expectations, then the market would collapse. Everyone would try to create the exact same portfolio, which of course is impossible to do.

    4. Similar to “Kill them all, let God sort them out.” Only instead it is “Initiate violence against people in THIS way instead, and let God sort everything else out.”

  8. Gravatar of Doug M Doug M
    18. March 2014 at 09:29

    Don,
    Why isn’t there a NGDP futures market?

    The futures exchanges create new contracts when a few participant sense that there will be sufficient “two-way flow” to sustain the market.

    Who has an interest in “Selling NGDP?” i.e. want to buy a contract that will rise in price if NGDP falls? I suppose that there are some cyclical companies that will see there revenue fall if NGDP falls, and they could use these contracts to smooth their “top line?” Is this an efficient way to smooth the top line? Will the auditors approve of this hedge? If the auditors insist that a derivative contract have a different accounting treatment from management’s intents it will kill the usefulness of the contract.

    Who is the natural buyer of NGDP? Who needs to hedge against NGDP rising faster than expected? Is rising NGDP really their risk? Or is it rising prices or rising interest rates?

    There are some people who have an economic interest in hedging inflation risk, hence there is a market in inflation-linked bonds.

    Just because there may be value in the price signals of the NGDP futures market, doesn’t mean that there is an economic interest for anyone to trade that market.

    There is one more problem in the NGDP futures market. If the fed is actively pegging the price of NGDP futures as Sumner suggests that they should be doing, who is going to trade the contract? Or how do you structure the contract so that someone can still hedge their risks, or profitably speculate in the contract?

  9. Gravatar of Dan W. Dan W.
    18. March 2014 at 09:43

    Scott,

    You write: “the bubble theory is completely vacuous–it doesn’t help us to better understand the world around us”

    Really? I think bubbles tell us a lot about the world in which we live. They remind us that humans are generally short-sighted and greedy. They remind us that there is a natural desire to have something for nothing. They remind us that the financial world is unpredictable and subject to rapid, whimsical, changes in the expectations of market participants.

    What you appear to be saying is that bubble theory does not lend itself to predictions. And you are correct. But just because the scientific understanding of bubbles is lacking does not mean they do not exist. Rather, it may simply mean that we do not understand them. Big Difference!

    Is it your thesis that large, widespread, malinvestment does not occur? Is it your argument that that any spending is rational spending? If so that is an amazing assertion. Why then bother with economic science at all if there is no wrong answer?

  10. Gravatar of ssumner ssumner
    18. March 2014 at 11:07

    John, I added one link. Some of the ideas I heard in face to face discussion. Others are well known (ratex, the EMH.) Fama’s anti-bubble position is repeated in almost all his interviews.

    Ilya, I only teach undergrads. Under FAQs (right column), I list some readings I like in question 21.

    Don, Classic externality problem. NGDP futures offer huge gains to society, at a tiny cost, but small gains to individual traders. Bill Gates should set one up.

    Thanks Travis.

    Dan, I already know that people want something for nothing. I want something for nothing! I don’t need bubble experts to tell me that.

    The rest of your comment makes no sense. I was talking about price bubbles, not malinvestment. Malinvestment and malconsumption can be produced by bad public policies (which is a policy issue) or private mistakes (which is not a policy issue.)

  11. Gravatar of Nick Rowe Nick Rowe
    18. March 2014 at 11:12

    Canadian house prices (mostly) kept rising too, after a small blip down. They are now more than 20% above the 2008 level.

    Here is our version of the Case-Shiller index:

    http://www.housepriceindex.ca/Default.aspx

    It is tempting to argue that the US had a negative bubble.

    Good post.

  12. Gravatar of Greg Hill Greg Hill
    18. March 2014 at 11:42

    Scott,

    You write, “It makes no sense to have a model of the economy that assumes X, but which is filled with people who believe ‘not X.’ As Bennett McCallum pointed out this idea is better called ‘consistent expectations.’”

    Imagine a model that assumes agents with different expectations, e.g., “the bulls and the bears,” or competing views of the macroeconomy derived from RBC, on the one hand, or some version of New Keynesian thinking, on the other. If agents in such a model know that there are different, even conflicting, expectations held market participants, is that enough to qualify as “consistent expectations”?

  13. Gravatar of benjamin cole benjamin cole
    18. March 2014 at 11:52

    Nick Rowe’s comment is worth expanding upon…a “negative bubble” in housing prices…imagine if the Fed concerned itself with negative bubbles rather than inflation…

  14. Gravatar of Al Al
    18. March 2014 at 12:01

    One problem with the “No Wait And See” argument is “The Long Run Is Now” argument. If expectations have been conditioned by past behavior, then a shift in policy action may be interpreted as a less significant shift in policy regime. So for a given point estimate of policy efficacy, real time observers may struggle to distinguish lagged expectation shifts from ineffectiveness.

  15. Gravatar of Doug M Doug M
    18. March 2014 at 13:00

    “Don, Classic externality problem. NGDP futures offer huge gains to society, at a tiny cost, but small gains to individual traders. Bill Gates should set one up.”

    Markets only work if there are gains for the individual traders! Setting up the market is useless if no one wants to trade. Or as Yogi might said, if the the fans don’t want to come, you can’t stop them.

  16. Gravatar of Major_Freedom Major_Freedom
    18. March 2014 at 13:08

    Price bubbles are not bubbles.

    Bubbles are composed of real factors. Oftentimes malinvestment is caused by (greater than free market) monetary expansion, so that is why we tend to see positive correlations between production booms and prices. The bubbles aren’t the historical price trends. If for example the Fed did not start to tighten up its OMOs in 2005-2006, if it instead accelerated its monetary expansion through 2006 and beyond, such that the historical price trend of houses did not fall the way it did, then that wouldn’t mean the housing bubble that was well formed by 2005 somehow would not have existed. The housing market was what it was in 2006. No amount of subsequent funny money could have changed that fact.

    Never reason from a price change.

  17. Gravatar of Vaidas Urba Vaidas Urba
    18. March 2014 at 14:25

    Scott,

    Here is Fama on bond risk premium: “The regressions confirm growing evidence that expected term premiums in bond returns vary through time”. I am not sure if you agree.

    BTW I have a new guest post at Insecurity Analyst blog (inspired by the discussion at Nick Rowe’s blog), where I discuss how the ECB is trying to get away with not doing QE:
    http://insecurityanalyst.blogspot.com/2014/03/draghi-is-guiding-interest-rates-yellen_18.html

  18. Gravatar of Dan W. Dan W.
    18. March 2014 at 14:37

    Scott,

    OK, no price bubbles. But what about credit bubbles? Can NGDPLT always balance fluctuations in credit demand? That appears to be the assertion but how might that play out in the real world when demand for credit might collapse very quickly, for whatever reason.

  19. Gravatar of TravisV TravisV
    18. March 2014 at 14:52

    Major_Freedom,

    Let’s face it: you’re here primarily because Scott Sumner is a tremendously brilliant and insightful thinker.

    His view of the way the world works isn’t very different from your view, if at all.

    Even if you find monetary expansion morally unacceptable, Sumner’s work is still indispensable for understanding macroeconomics and the Great Depression, right?

    Consider this post on the “root causes” of the Great Depression:
    http://www.themoneyillusion.com/?p=4286

    No matter what your ideology is, it’s unmistakable that it’s a work of analytical genius!

  20. Gravatar of Major_Freedom Major_Freedom
    18. March 2014 at 16:02

    TravisV:

    I can’t square your assessment of why you believe I’m here, with my memory of correcting Sumner on many different occasions.

    I also can’t square your assessment of Sumner’s explanation of the cause of the Great Depression, with what I know to be analytical/economic flaws in the AD theory. At some level I think Sumner knows it’s problematic, because what he does with the AD theory is what he often does when presented with a contradiction or flaw in his beliefs: he doubles down, deepens the confusions and declares the flaws a virtue. For example, he said this in that post you linked to:

    “Although I have argued that monetary tightening was the proximate cause of the onset of the Great Contraction, and may also have contributed to the 1929 stock market crash, I do not believe that the Depression had any single ‘root’ cause; in the sense of a cause deeply embedded in preceding events.”

    Here he takes a gap in his theory (no explanation of any cause for what could have made so many people suddenly hoard so much money and destroy so much money through defaulting on debt created by credit expansion), and rather than being honest and admit his theory is among other things incomplete, he doubles down and presents that gap as some sort of virtue, and says there is no prior cause, as if that gap is a feature and not a bug.

    I’ve read many explanations for the Great Depression, and Sumner’s is not original. It is a derivative of Keynesianism interpreted purely nominally.

    Any explanation of any economic phenomena, including the Great Depression, cannot ignore real factors and then be claimed as satisfactory, or the best there is, or the most useful.

    I suspect that the reason Sumner doesn’t seriously address prior causes to AD is that at a subconscious level, he is afraid to learn the possibility that the prior cause is the existence, not just insufficient inflation from, the Federal Reserve. By stopping at AD and nothing prior, he can then pretend to blame the Fed for the Depression but not in a way that would imply it must be abolished, in that the cause is the existence of socialist management of money, but rather in a way that sanctions it, in that the cause is not enough activity from the socialist managers.

    Without the Fed, Sumner’s entire intellectual investment would be exposed as a malinvestment. You know that sayong about the hardest thing to do is teach someone something that they have a personal incentive not to understand it, haven’t you? Sumner attacks Fed attackers because it’s akin to an attack on him.

    If only he learned economics from an individual action foundation. Then he wouldn’t be caught in the embarrassing situation of being free market in food, clothing, transportation, medicine, energy, housing, real estate, water, etc, etc, etc…..but socialist with money.

    He went to Chicago. You can’t graduate from that school if you write all your papers on why the Fed should be abolished. Lucas would have refused to be his supervisor. He is a product of a school started on the investment of politically connected financiers and friends of central banking.

    You can fawn all you want. I know better.

  21. Gravatar of Major_Freedom Major_Freedom
    18. March 2014 at 16:06

    Austrian theory includes both real and nominal factors in the cause for the Depression. It does need to stop arbitrarily at point X out of fear of being exposed as inconsistent. Purely nominal explanations such as MM are just as flawed as purely real explanations like RBC.

  22. Gravatar of benjamin cole benjamin cole
    18. March 2014 at 16:50

    Major Freedom:
    Tell the truth: are you wearing jodpuhrs and boots right now? Do you have a mustache? Dueling scars? When you hear martial music do you get involuntary arm spasms?

  23. Gravatar of Tom Brown Tom Brown
    18. March 2014 at 16:55

    benjamin cole, “jodhpurs” Lol, I swear I learn a new word from you just about every week.

  24. Gravatar of Tom Brown Tom Brown
    18. March 2014 at 16:59

    benjamin, it’s one of these:

    http://tinyurl.com/n4ujw88

    http://tinyurl.com/lde4kd7

  25. Gravatar of Ben B Ben B
    18. March 2014 at 17:27

    Travis V,

    Tell the truth; you have a Scott Sumner shrine in your basement, don’t you?

    Or at least you have a shirt with his face on it, right?

  26. Gravatar of Jason Jason
    18. March 2014 at 17:29

    I was inspired by this post to have a go at teasing out the various components of the interest rates you mention in the Lucas point #1. Here’s a graph:

    http://twitter.com/infotranecon/status/446095168546934784

  27. Gravatar of Major_Freedom Major_Freedom
    18. March 2014 at 17:33

    Benjamin:

    I don’t know what you’re talking about, but I suspect the demeanour you had in writing it is closely related to why your knowledge in economics is lacking.

  28. Gravatar of Dustin Dustin
    18. March 2014 at 17:50

    Seems Lucas (1): No Wait and See is quite similar to Fama (1): EMH

  29. Gravatar of Dustin Dustin
    18. March 2014 at 17:51

    ^^ Lucas (1): Long Run is Now

  30. Gravatar of Dustin Dustin
    18. March 2014 at 18:07

    Scott,

    Off Topic: John Williams, SF FRB, indicates that as the economy shifts from extreme times to more normal times, forward guidance should be more vague and qualitative because, essentially, managing policy for a healthy economy by one or two numbers is very difficult.

    http://economix.blogs.nytimes.com/2014/03/12/qa-confidence-about-the-feds-course/?_php=true&_type=blogs&_php=true&_type=blogs&_r=1

    Maybe so when using the wrong one or two numbers! Perhaps less focus is on monetary policy as the economy recovers, and therefore a policy regime change to NGDPLT may be less of a concern for policy makers…. but this reasoning seems to indicate that in a healthy economy, as much as ever, NGDPLT makes a lot of sense as it enables very certain forward guidance via a single number. NGDPLT makes managing policy for a healthy economy quite easy.

  31. Gravatar of ssumner ssumner
    18. March 2014 at 18:12

    Nick, That can’t be right. Everyone knows there were a bubble in 2008. A crash was inevitable. You need to recheck those figures.

    Seriously, great post on exogenous money.

    Greg, I have no problem with models containing a diversity of opinion. But I don’t want models where the consensus view says not X, if the model says X. That sort of model will not provide good policy advice (in most cases.)

    Al, That’s why I mentioned that you’d want to look at what the futures markets were predicting. You want to see how markets react to the unexpected part of the policy announcement.

    Doug, Nope, you can force them to trade with small subsidies.

    Vaidas, Thanks for the link. I’ll defer to your expertise, and Fama’s, on the question of bond term premiums.

    Dan, We subsidize borrowing, so there is probably too much borrowing going on. I wouldn’t call that a bubble, just a market distortion. Like growing too much wheat because of price supports.

    Thanks Jason, but how did you get the liquidity effect? Those effects tend to be small and transitory.

    Dustin, I see them as different, what is the connection you see?

  32. Gravatar of Dan W. Dan W.
    18. March 2014 at 18:14

    I was hoping for a more intelligent discussion.

    The claim of NGDPLT is that it is both prudent and practical to maintain a predetermined level of NGDP growth. Sounds simple enough but the real world is not so simple.

    Consider a period when economic growth is aided by an super-expansion of credit. The value of assets is increasing as is the amount borrowed against those assets. Unless NGDPLT means credit can expand exponentially for infinity the credit expansion supercycle will stop and then retrench. It must retrench for a simple reason: The peak value of assets was not based on real or nominal incomes*. As such the problem cannot be fixed by “tweaking” nominal incomes. The distortion in asset values is too large and monetary stimulus will be slow and mostly ineffective at fixing the debtor’s pain.

    What should happen is bankruptcy wherein the creditor assumes new terms for the debt. But whether there is bankruptcy or not nominal income will decline from the previous, supercycle, levels. In response to the awareness of the bad economic decisions that they made people will change their borrowing and spending habits. There will be an economic slowdown.

    In such an environment it begs the question why one would presume monetary policy should make up the difference. The difference to what? If the economic growth of the past years was based on faulty economics does it make sense to make a return to that level of economic activity the target? But that is not all. The more absurd assertion is that Monetary policy should be used to make it seem that that bad debt never happened in the first place (the LT in NGDPLT). How might this be done without introducing even greater distortions in the economy is a great mystery. Just trust the math.

    *The peak value of assets was based on the fraudulent concept that ever increasing amounts of equity can be extracted from assets that are ever increasing in value. Otherwise called an economic perpetual motion machine. And these are just as imaginary as are the mechanical ones.

  33. Gravatar of Michael Byrnes Michael Byrnes
    18. March 2014 at 18:16

    Dan W wrote:

    “But what about credit bubbles? Can NGDPLT always balance fluctuations in credit demand? That appears to be the assertion but how might that play out in the real world when demand for credit might collapse very quickly, for whatever reason.”

    Malinvestment is the suboptimal – even wasteful – use of scare resources, right?
    To the extent that there has been malinvestment, stuff that people want (and are eager to pay for) cannot be produced in sufficent quantity to meet demand. Instead you have a shortage, and prices of those goods rise, while you maybe have an oversupply of stuff that people don’t want to buy (say, houses in suburban Arizona), and prices of those goods fall.

    The good news, is that this kind of problem can be self-correcting. The high prices of the undersupplied goods will drive efforts to produce more of those goods, while the low prices (and perhaps glut) of the overproduced goods will be a disincentive for forther production.

    However, an adverse demand shock will stop this correction from happening. NGDPLT, by stabilizing nominal income expectations, will allow the economy to sort out its malinvestments.

  34. Gravatar of ssumner ssumner
    18. March 2014 at 18:18

    Dustin, I agree about Williams.

    Vaidas, Regarding your post, do you think the ECB’s forward guidance is clear enough to provide a fair test of the Woodfordian model?

  35. Gravatar of Dan W. Dan W.
    18. March 2014 at 18:18

    Scott,

    The first line of my last comment had to do with the chit-chat that preceded your post.

  36. Gravatar of ssumner ssumner
    18. March 2014 at 18:24

    Dan, You said:

    “I was hoping for a more intelligent discussion.”

    If so, then don’t post the sort of comment that follows this statement. It’s clear that you don’t know what NGDPLT is. Nominal income does not fall after a credit “bubble” burst if you have NGDPLT. In addition, credit bubbles are far less likely to occur in the first place, as they tend to be associated with highly unstable NGDP.

    Later you talk as if I favor targeting RGDP. What’s that all about?

  37. Gravatar of Oliver Oliver
    18. March 2014 at 19:12

    Dan,

    I’m not trying to make any excuses for market monetarists, but couldn’t a period of NGDP growth aided by a super-expansion of credit be considered “above target level growth” and necessitate tighter monetary policy by the central bank? I don’t think Sumner ever claimed that NGDPLT would be able to fix the mistakes of previous policy regimes.

    I think the real challenges facing market monetarists are the questions of how to determine the desired NGDP growth level and how to deal with changes to the desired level of growth . I personally find NGDPLT highly appealing, but it’d take a lot to convince me that developing nations should target 5-6% nominal growth now, or that they should continue targeting their currently high growth rates long into the future.

    Jason Smith has a nice post on falling NGDP (trend) growth rates in the US here:

    http://informationtransfereconomics.blogspot.com.au/2014/03/whats-up-with-ngdp.html

  38. Gravatar of Major_Freedom Major_Freedom
    18. March 2014 at 19:43

    Michael Byrnes:

    “The good news, is that this kind of problem can be self-correcting. The high prices of the undersupplied goods will drive efforts to produce more of those goods, while the low prices (and perhaps glut) of the overproduced goods will be a disincentive for forther production.”

    “However, an adverse demand shock will stop this correction from happening. NGDPLT, by stabilizing nominal income expectations, will allow the economy to sort out its malinvestments.”

    But NGDPLT, because it requires the same non-market currency issuance as price targeting, would put pressure on relative prices that caused the same malinvestment.

    An “adverse”, market driven “demand shock” is the only way to allow unmolested corrections to take place. The problem is not that underproduced goods should have had exactly offsetting spending associated with them to offset the overproduced goods which had decreased spending. It’s not optimal to have exactly offsetting spending forced by government, because that is the very cause of the problem of partial over and underproduction that you see.

    ———————

    Sumner:

    Dan’s point is not that credit bubbles will cause declines in NGDP in NDPLT regimes. His point is that a credit expansion fueled boom makes a bust inevitable, and cannot be stopped by “tweaking” aggregate incomes through inflation. Credit collapses ultimately because of the real factors that limit it from being accelerated ad infinitum.

    You have to integrate real side factors into your understanding of markets in order to “get” this. As long as you focus only on money and spending, you’ll miss the bigger picture.

  39. Gravatar of Dustin Dustin
    18. March 2014 at 20:03

    Scott.

    RE: Lucas (1) and Fama (1). They seem to both be describing how markets incorporate expectatations into today’s decisions. Expected increased future cash flows leads to increased spending today.

    In your Lucas example, a rate reduction increases our expectation of future cash flow (income), so we spend more today.

    Fama’s EMH generally asserts something very similar, that a boost in an security’s expected cash flow generation (total return) increases our willingness to pay for the asset today.

  40. Gravatar of TravisV TravisV
    18. March 2014 at 20:27

    Ben B.,

    I do have a beautiful shrine to Scott Sumner right here. And I kneel and bow to it at least five times a day!!!

  41. Gravatar of Prakash Prakash
    18. March 2014 at 22:52

    Prof. Sumner, You’ve mentioned that even if an explicit rules based policy is adopted, people may expect it to be violated. That statement, combined with a comment you had made in one of the replies a couple of posts ago, has me a little confused. That statement was “monetary policy has nothing to do with productivity”. Ideal monetary policy targets nominal median wage growth level a few percentage points above the growth of productivity, right? Anything higher acts as a capital on tax and anything lower runs into money illusion.

    My point is that even if there is an explicit policy and the economy is suffering, then people believe that the regime will change. This weakens the credibility of the policy. So, there will have to be a policy that targets productivity or it will necessarily run into the problem of lower credibility.

    Major Freedom, Monetary theory need not be based on the necessity of a fed. bitcoin is written on the basis of a monetary theory, that a commodity that gets scarcer is a good money. It is possible for future electronic currencies to be written using any kind of rules. Rules that require interfacing with the real world like NGDP targeting, will continue to require trusted authorities to distinguish between genuine sales and round tripping. However, there can be purely insystem rules like targeting nominal transaction costs to grow at a certain rate. There can also be crypto currencies which have their own inbuilt bond markets, by locking and unlocking different amounts of money at a given time.

    Which brings me to a point where I hope more of the community would find a commonality. Crypto-currencies are the ultimate in monetary policy credibility, atleast monetary base wise. You need a 51% majority in bitcoin to create a “fork”. There can be crypto currencies where this majority number can be increased to higher levels as well, using a combination of proof of work and proof of stake. If even the most credible of central authorities are subject to political pull, then the entire project of popularizing NGDP futures markets is bound to have a limited impact. Political pressure will force undershooting or overshooting of the target dependent on the whims of the day.

    Future monetary policy theorists are probably better off designing a good crypto currency. And creating crypto currencies requires much less investment when compared to a NGDP futures market.

  42. Gravatar of Prakash Prakash
    18. March 2014 at 22:54

    Errata – anything higher acts as a *tax on capital*

  43. Gravatar of benjamin cole benjamin cole
    19. March 2014 at 00:37

    Tom Brown:
    I got my jodhpurs on, Wagner on the LP, and re-reading Mises.
    Funny, I just got a call from the ECB. They want to know if I can join them!

  44. Gravatar of Vaidas Urba Vaidas Urba
    19. March 2014 at 00:54

    Scott:
    “Regarding your post, do you think the ECB’s forward guidance is clear enough to provide a fair test of the Woodfordian model?”

    Here is what Woodford wrote:
    “I shall further argue that the idea that balance-sheet policies can be used as a substitute for forward guidance, and therefore excuse a central bank from any need to make commitments about future policy, is mistaken. Much of the effect of balance-sheet policies seems to have resulted from their being taken as a signal about likely future policy; and to extent that this is true, explicit forward guidance will surely allow the desired effect to be obtained more reliably”

    If the ECB gets away with not doing QE, this would strongly support Woodford’s model. Maybe the current forward guidance is not very clear, but the ECB is doing a lot to clarify it, and we are likely to get additional clarifications before the ECB tries QE.
    On the other hand, if the ECB is forced to do QE, we will have to conclude that Woodford’s practical policy prescriptions have failed.

  45. Gravatar of ssumner ssumner
    19. March 2014 at 04:45

    Dustin, Yes, they both do relate to expectations, but thery are still separate points.

    Pakesh, Yo usaid:

    You’ve mentioned that even if an explicit rules based policy is adopted, people may expect it to be violated. That statement, combined with a comment you had made in one of the replies a couple of posts ago, has me a little confused. That statement was “monetary policy has nothing to do with productivity”. Ideal monetary policy targets nominal median wage growth level a few percentage points above the growth of productivity, right? Anything higher acts as a capital on tax and anything lower runs into money illusion.”

    This is a common misconception. The size of the excess tax on capital depends on the rate of NGDP growth, not the rate of inflation. That’s because the optimal tax on capital is zero, so any tax paid is excess. And returns on capital are correlated with NGDP growth, not inflation.

    Vaidas, Perhaps, but any forward guidance they do offer is likely to fall short of the NGDPLT, or even PLT guidance, that he would call for.

  46. Gravatar of ssumner ssumner
    19. March 2014 at 04:46

    Vaidas, Quick follow up. Perhaps it would weaken the argument for guidance alone with IT.

  47. Gravatar of TravisV TravisV
    19. March 2014 at 06:20

    Prof. Sumner,

    I think Brad DeLong attacked you in this new post…..

    http://equitablegrowth.org/2014/03/19/2297/in-which-i-try-and-fail-to-understand-the-current-state-of-right-wing-monetary-economics-wednesday-focus-march-19-2014

  48. Gravatar of Willy2 Willy2
    19. March 2014 at 07:05

    Bubbles do exist. There’re a number of good and well established gauges that say whether in one or the other market there’s a bubble or not.

    But that doesn’t tell when a such bubble is going to burst.

  49. Gravatar of Vaidas Urba Vaidas Urba
    19. March 2014 at 07:14

    Scott,

    Woodford says QE works through signaling only (on the liability side of balance sheet), and any benefits from signaling are easier to achieve with explicit guidance. The ECB is averse to large balance sheet, and is currently trying to improve forward guidance. If they are forced to do QE, this means that guidance implicit in QE works better, contra Woodford. Yes, they are not doing NGDPLT, but they won’t be doing it when they,start QE too.

  50. Gravatar of Doug M Doug M
    19. March 2014 at 08:40

    “Doug, Nope, you can force them to trade with small subsidies.”

    Explain this one.

    Trade happens when the transaction makes both parties better off. Who are you going to subsidize such that parties that had no interest in trading now do?

    How do you introduce such a subsidy in such a way that the subsidy itself doesn’t massively distort the market prices?

    You what the Fed to base its monetary policy off of the price signals from this market. You do realize that this would be literally reasoning from a price change.

  51. Gravatar of Daniel Daniel
    19. March 2014 at 08:42

    I see a lot of idiots (hey there, Willy2) being adamant about the existence of “bubbles”.

    When asked upon the usefulness of such a concept, all they can say is that the price will come down … someday. Which is a rather long-winded way of saying it doesn’t have any.

  52. Gravatar of Willy2 Willy2
    19. March 2014 at 09:47

    It seems “Daniel” is a fan of the “Efficient Market Hypothesis”. Don’t worry, it’s only a “Hypothesis”.

    Take e.g. the housing bubble in the US. Housing simply became too expensive because:
    - Home prices more than doubled between 1997 & 2006.
    - Short term interest rates went up from 1% uo to 5% between 2003 & 2007. Mortgage payments for those subprime mortgages were directly tied to that particular rate.
    - Oil prices went up sevenfold between 2001 & mid 2008
    - while at the same time median household income went down from 2000 up to say 2005 and then median income rose again up to very late 2007. But this peak of 2007 was slightly lower than the peak in 2000.
    These are a good set of reason why homeprices were bound to tank one day. It was only unpredictable when those prices were going to tank.

    Home prices in Los Angelos, Miami were at 9 to 10 times average annual income. In Sydney today, home prices are at 8 to 10 times annual average income.

    And any honest realtor will tell you that a house at 4 or more times annual income is expensive & 6 times or more annual income is bubble.
    Home prices WILL come down even more here in the US because median household income has dropped well below the peaks of both 2000 & 2008 and continues to go down.

    Based on this kind of metrics it’s very clear & easy to spot bubbles. One only has to know at what metric one should use.

  53. Gravatar of Daniel Daniel
    19. March 2014 at 09:48

    So if something is too expensive for your tastes, it’s a “bubble”.

  54. Gravatar of Dan W. Dan W.
    19. March 2014 at 10:50

    Bubbles? I like Herb Stein’s definition:
    “If something cannot go on forever, it will stop”

    This statement is even more elegant in its shortened form:
    “Trends that can’t continue, won’t.”

    A corollary is Thatcher’s law:
    “Eventually you run out of other people’s money”

  55. Gravatar of Mark A. Sadowski Mark A. Sadowski
    19. March 2014 at 11:41

    Scott,
    Off Topic.

    On Monday the Federal Reserve Economic Database (FRED) rolled out the *new* Fred Graphs:

    http://fredqa.stlouisfed.org/2013/11/20/introducing-the-latest-and-greatest-fred-graphs/

    I knew this was coming but at the bottom they had the following claim:

    “FAQs

    What happens to my saved graphs?

    Your saved graphs will still function and will be automatically converted to the new graph format. All settings should be retained. If anything looks strange, please let us know…”

    Consequently I wasn’t worried about my 700+ FRED graphs.

    So what a shock when Monday came around and absolutely none of my graphs looked right, my “Saved Graphs” page was now in 48 separate pages with 15 giant tiles in no particular order.

    I contacted the St. Louis Fed about this matter and some progress has been made. Now the Saved Graphs page is again in alphabetical order and has the option of a “list” that makes the giant tiles disappear. I still have to go through 48 separate pages hunting for graphs but at least it is feasible to find them even if it is somewhat laborious.

    And the appearance of most of my graphs has been restored to their original settings. But some of them are still badly “messed up” for lack of a better phrase. For example, here’s what one of my graphs looked like before Monday:

    http://thefaintofheart.files.wordpress.com/2013/06/sadowski_5.png

    Here’s what the same graph looks like as of today:

    http://research.stlouisfed.org/fred2/graph/?graph_id=123169&category_id=

    Note that the blue and red line consisted of four and five separate lines respectively which were spliced together. This was possible because you could set the endpoints of each individual line separately from the others. This is no longer possible and I have been told by the people at FRED that problem relates to the new “sliding scale” feature found at the bottom of the graphs. (Previous functionality was evidently sacrificed for the new feature, one which personally I never knew that I even needed.)

    The FRED webprogrammers are currently working on these problems, and I have no wish to antagonize them as they do appear to be trying to accomodate me, but I am concerned about the affected FRED graphs on my account, as well as the ability to make similar graphs in the future.

    (continued)

  56. Gravatar of Mark A. Sadowski Mark A. Sadowski
    19. March 2014 at 11:42

    (continued)

    On the other hand I have noticed very little concern expressed in the econblogosphere and this leaves me somewhat puzzled. The only evidence I can find that anyone finds the current situation troubling in any way are the following two posts:

    http://newarthurianeconomics.blogspot.com/2014/03/the-new-fred.html

    http://beforeitsnews.com/economics-and-politics/2014/03/fred-screwed-up-2462878.html

    If you have a FRED account and you haven’t been there since before Monday you may want to check to see if everything is satisfactory. If not please be sure that you politely email FRED, informing them of your concerns, otherwise the situation will go unnoticed until there is far less pressure on them to do anything about it.

  57. Gravatar of Willy2 Willy2
    19. March 2014 at 11:43

    My tastes DO NOT determine whether or not there’s a bubble or not. One simply has to look at good/well established gauges. And no, I did not invent those gauges. It took me a considerable amount of time to learn what those gauges are and when and at what level those gauges reach the territory called “Bubble-land”.

    But bubbles can exist for a long, long time before they burst.

  58. Gravatar of Daniel Daniel
    19. March 2014 at 13:43

    Willy2 & Dan W,

    I take it you’re not familiar with something called “supply and demand”.

    Also – you’re apparently too dense (but it’s ok, you’re not alone in this) to grasp what I’m trying to say – saying that “someday the price will go down” HAS NO PRACTICAL USEFULNESS.

    Can any of you predict a fall in price ahead of time ? What’s that, you can’t ? I thought so. So of what use was your illusory ability to identify “bubbles” (also known as “confirmation bias”) ?

    Merely saying that “X is over-priced” is like saying the Sun rises in the east. Yes, someday the price will fall. So what ? You didn’t predict it, so shut up.

  59. Gravatar of Willy2 Willy2
    19. March 2014 at 14:00

    Agree & disagree. Predicting the precise moment when e.g. home prices fall is very difficult.

    But predicting when the stockmarket falls or crashes is actually remarkably simple. E.g. in mid 2007 a number of credit market gauges (e.g. declining short term rates) started to show a divergence with the overall stockmarket(s) and that was before the market top in november of 2007.

  60. Gravatar of Daniel Daniel
    19. March 2014 at 14:11

    Wow. Are you so lacking in self-awareness that you can’t tell it’s just monday morning quarterbacking ? You predicted a change in the stockmarket AFTER IT HAPPENED – and you expect me to be impressed by that ?

    How’s about you show me a consistent record of successful predictions ? Because such a thing would definitely make me change my mind about the EMH.

    Until then, I’d advise you to stick to statements which convey actual information (as opposed to mood affiliation). Saying “X is a bubble” is USELESS unless you can tell when the price will fall.

  61. Gravatar of Dan W. Dan W.
    19. March 2014 at 14:21

    Daniel,

    You shouldn’t be so polite. It gets in the way of your argument. By the way, the historical record of the distinguished academic Economist Guild is that it is rather inept at making predictions. How often do you tell them to “shut-up”?

    Do you dispute the empirical observation that people can be foolish? If not then why do you dispute the empirical observation that large numbers of people can behave foolishly? It is a simple observation that people, groups, institutions and even governments can fool themselves into using too small of a discount rate, if not a negative discount rate, to justify financial decisions. Such behavior is often foolish.

    NGDPLT is a tautology that assumes away all the flaws, warts, friction and inefficiencies inherent in the economy. It then promises that in this purified world it is the program that will ensure prosperity. Well, you can make all the assumptions you want but you can’t change the world we have. Until NGDPLT theorists make accommodations for the real world and explain how to make their program sufficiently robust to handle the foolishness of economic actors, both large and small, and to do this without reaping great harm on other market participants, their thesis will remain to many just an interesting academic exercise.

  62. Gravatar of Dan W. Dan W.
    19. March 2014 at 14:24

    Daniel,
    —-
    You shouldn’t be so polite. It gets in the way of your argument. By the way, the historical record of the distinguished academic Economist Guild is that it is rather inept at making predictions. How often do you tell them to “shut-up”?

    Do you dispute the empirical observation that people can be foolish? If not then why do you dispute the empirical observation that large numbers of people can behave foolishly? It is a simple observation that people, groups, institutions and even governments can fool themselves into using too small of a discount rate, if not a negative discount rate, to justify financial decisions. Such behavior is often foolish.

    NGDPLT is a tautology that assumes away all the flaws, warts, friction and inefficiencies inherent in the economy. It then promises that in this purified world it is the program that will ensure prosperity. Well, you can make all the assumptions you want but you can’t change the world we have. Until NGDPLT theorists make accommodations for the real world and explain how to make their program sufficiently robust to handle the foolishness of economic actors, both large and small, and to do this without reaping great harm on other market participants, their thesis will remain to many just an interesting academic exercise.

  63. Gravatar of Daniel Daniel
    19. March 2014 at 14:38

    By the way, the historical record of the distinguished academic Economist Guild is that it is rather inept at making predictions.

    That’s the whole point of the EMH. But then again, you have a proven track record (around here, at least) of saying dumb things – so I wouldn’t expect you to understand what exactly the EMH implies.

    You’ve also proven yourself too dense to understand what “sticky wages” implies – so I don’t see any point in trying to explain the advantages on NGDP targeting, since you wouldn’t understand them.

    And no, I don’t see any need to be polite with pompous idiots. Ya know, like people who use arguments along the lines of “humans act” (as if it means anything).

  64. Gravatar of Michael Byrnes Michael Byrnes
    19. March 2014 at 15:25

    Dan W. wrote:

    “NGDPLT is a tautology that assumes away all the flaws, warts, friction and inefficiencies inherent in the economy.”

    It assumes nothing of the kind. An economy can have all sorts of non-monetary problems that will have non-monetary consequences.

    NGDPLT assumes that it’s a bad idea to add monetary problems on top of non-monetary ones.

  65. Gravatar of Daniel Daniel
    19. March 2014 at 15:36

    Dan W. spends a lot of time arguing that wages aren’t sticky. And since stickiness doesn’t exist, you can’t have “monetary problems”.

    How exactly can you reason with someone so out of touch with reality ?

  66. Gravatar of Jason Jason
    19. March 2014 at 15:45

    Actually, the liquidity effect is almost totally invisible on its own (especially before the 1990s); it had to be teased out. I assumed a quantity theory of money P = P(MB) ~ MB^b and a log linear interest rate r = r(MB, NGDP) = c log NGDP/MB + k consistent with long run neutrality (i.e. homogeneous of degree zero in NGDP, MB). I separated the individual effects via the following decomposition:

    Liquidity effect
    r(MB + dMB, NGDP) – r(MB, NGDP)

    Combined income/price level and liquidity effect
    r(MB + dMB, a*NGDP) – r(MB, NGDP)

    Income/price level effect alone
    r(MB + dMB, a*NGDP) – r(MB + dMB, NGDP)

    where a = P(MB + dMB)/P(MB), i.e. the effect of monetary expansion on the price level. The parameters b, c and k above were chosen to fit to the CPI and 10-year rate data.

    I put a picture of this model in terms of diagrams here:

    http://twitter.com/infotranecon/status/446430331936071680

    The key is making the effect of the changes in the monetary base consistent in the two markets (essentially having both the interest rate and price level models be pretty good fits to the data guarantees it will be). There is an update as well — the Fisher effect disappears when you look at the 3-month (short run) interest rate:

    http://informationtransfereconomics.blogspot.com/2014/03/the-fisher-effect.html

  67. Gravatar of ssumner ssumner
    19. March 2014 at 17:10

    Vaidas, Well I would be able to do highly effective guidance, without QE, if they put me in charge of the ECB. But I see your point.

    Doug, You said:

    “You what the Fed to base its monetary policy off of the price signals from this market. You do realize that this would be literally reasoning from a price change.”

    No, reasoning from a price change is when you make inferences about quantity from a price movement. In this case you’d be making inferences about prices from a price change.

    As far as the subsidy, you simply pay an above average interest rate on money deposited in margin accounts. I’ve published several papers that discuss the idea. It is very easy to do. No different from what companies do with prediction markets.

    Mark, I always hate it when people change things–it creates nothing but problems.

    Dan, Some of your comments are so bizarre it’s hard to tell if you are actually aware of what NGDPLT is.

    But I do think you have proved that the answer to this question:

    “Do you dispute the empirical observation that people can be foolish?”

    Is: “No, I can’t dispute that anymore.”

  68. Gravatar of Willy2 Willy2
    20. March 2014 at 00:53

    @Daniel: Bubbles do exist & will burst.

    See e.g. this: The Hong Kong property bubble (home prices >10 times average annual income) is about to implode:

    http://www.reuters.com/article/2014/03/19/hongkong-property-chinese-idUSL3N0MG1B220140319

  69. Gravatar of Daniel Daniel
    20. March 2014 at 02:22

    Willy2

    Nature hasn’t gifted me with patience. And when dealing with idiots, I’m even less patient.

    So I will only say this – if you’re so sure about your prediction, how’s about you put your money where your mouth is. Go massively short on Hong Kong housing – and if you’re right, you’ll be a rich man.

    Otherwise, you’re just another moron who thinks he can see the future in tea leaves.

  70. Gravatar of Willy2 Willy2
    20. March 2014 at 05:31

    Thankfully there’re A LOT OF tealeaves that can be read and all tealeaves combined give an excellent picture of what’s going to happen.

    BTW1: chinese investors have bought also real estate in Australia. Expect that bubble to pop as well in the near future.

    BTW2: the S&P 500 is a bubble as well.

  71. Gravatar of Willy2 Willy2
    20. March 2014 at 05:33

    BTW3: There’s a good wat to short “Hong Kong”.

  72. Gravatar of Daniel Daniel
    20. March 2014 at 06:06

    Yeah, I get it – everything is a bubble. Let me know when your insights make you rich. Won’t be holding my breath.

  73. Gravatar of Cory Cory
    20. March 2014 at 08:11

    Seems to me that this post provides the vehicle through which the GOP could get everything they really want.

    “So control of the stock of currency was all you need in order to control the value of currency. And since the value of currency is (by definition) the inverse of the price level, that’s all you needed to control the price level, or any other nominal aggregate.”

    I’m a mere ingenue but if the GOP would relieve itself of its affinity for hard money I don’t see why this statement does not provide the opportunity for the GOP to propose eliminating the income tax without worrying about the Deficit.

    When the Treasury Deficit spends because there is no longer an income tax, it issues a treasury security that contracts the money stock just like a tax would. There is no net increase in M, it seems to me. And, eliminating the income tax would be good for the supply side so no need to worry about Real GDP collapsing, it seems to me.

  74. Gravatar of J J
    20. March 2014 at 08:27

    Professor Sumner,

    With regard to rational expectations:

    I can have a model with a good state of the world (technology shock) and a bad state of the world. The data generating process in my model gives a probability of 0.6 to the good state next year and 0.4 to the bad state. I can have some agents who think the probability of the good state is 0.5 and others who think the probability is 0.3. Whether this is useful depends on what I want to study and what type of policy advice I want to extract. Of course, the central bank should not presume to be smarter than the market, so we don’t ever know what probability distribution the data generating process follows. Yet, a model with heterogeneous beliefs (and how far those beliefs are from reality) can help explain why some get rich in the capital markets and others do not. It can also help explain how incomplete markets affects outcomes.

  75. Gravatar of ssumner ssumner
    21. March 2014 at 04:58

    Cory, I strongly favor abolishing the income tax.

    J, I agree that models with heterogenous beliefs are useful, but I see them as being consistent with ratex. What you need to avoid is public policies that only work if the public stupidly doesn’t understand the policy. Those are likely to fail at some point, when the public catches on.

Leave a Reply