Believe those who are seeking the truth. Doubt those who find it. Andre Gide

Monday, September 20, 2010

Tyler Cowen's Happy Face Inflation Policy

Stopped in at my local Starbuck's yesterday (after a brutal soccer game that left me half paralyzed) and picked up a copy of the NY Times. In it, I came across this article by Tyler Cowen: Can the Fed Offer a Reason to Cheer? The article constitutes a nice summary of what many people think is wrong with the economy and what should be done about (at least, as far as monetary policy is concerned).

Here is the basic idea, as I understand it. The future  is dark and uncertain; there is no sound basis for making long-horizon forecasts. This opens a door for psychology. Communities are now prone to psychologically-inspired waves of optimism and pessimism ("animal spirits," to use Keynes' colorful phrase). Since current investment demand primarily runs off of long-horizon forecasts (relating to the expected return to future capital), these animal spirits lead to large fluctuations in "aggregate demand" via investment spending. Things are even worse in a monetary economy, as the existence of cash facilitates a psychologically-inspired "flight to safety" that, while perhaps individually rational, contributes to a socially irrational contraction in aggregate demand. The contraction in aggregate demand means lower sales volumes, so firms lay off workers, who are now unemployed. As unemployed workers generate no wage income, they cut back on purchases, which contributes further to lower demand. Firms are induced to cut prices, which leads to deflation. As debt is nominally denominated (not indexed to inflation), the deflation raises the real debt burden of indebted households and firms. Debt renegotiation is possible, but the process is costly and slow. The upshot is a wave of bankruptcies that contributes to the depressed business climate.

The "social problem," in a nutshell, concerns the dynamics of mass psychology in a free market system.  In a free market, individuals are (by definition) free to make their own economic decisions. Naturally, they do so without full regard of how their individual decisions might influence aggregate outcomes (an individual, after all, is tiny relative to the community). People value their economic liberty for a good reason, but liberty comes at a cost. The cost is that individuals may not always coordinate their decisions in a socially desirable manner (for example, in noncooperative games, Nash equilibria are generically suboptimal).

Well, if that is the problem, then the solution is straightforward; in principle, at least. Unlike individuals, the government is a large player. If we are collectively depressed for no good reason, then why not have the government bring on some good cheer? Insufficient demand is easily rectified by more government spending and a loose monetary policy. Inflation (some inflation, at least) is good. It gets people to spend their money (it would otherwise lose purchasing power). As people spend, sales volumes rise, profit margins rise, and firms are induced to employ more workers.

It is a seductive argument. And on the surface, it is hard to see what, if anything, is wrong with it. At least one or two prominent bloggers  are apt to call you "stupid" for not seeing it as a self-evident truth. Let's not be bullied by these over-inflated egos and try to think this through ourselves.

The first thing we should realize is that the argument may have some merit. It may even be completely true (in the sense that the described theoretical forces provide a good approximation for quantitatively important forces that actually operate in the economy). On the other hand, it may not be the whole story. Heck, it may even be false. (I realize that this is not an appealing message for those who find comfort in religion, but I am a social scientist, not a preacher).

Let's start with psychology. There is no question that people get emotional and that emotions can sometimes color decision-making. Accepting this does not, however, lead immediately to the conclusion that emotional decision-making is individually or collectively irrational. If I see a bear on my running trail, I freak out and run away (this actually happened to me, and looking back, I think I acted in a perfectly rational manner!). And from a Darwinian perspective, it is hard to see how a propensity for collective irrationality has led to our flourishing modern day civilization (although, I have to admit that wars are crazy and that Collapse is always a possibility).

True, market optimism appears to wax and wane, but so what? It is possible, even if one does not find it entirely plausible, that these undulations constitute, at least in part, waves of rational optimism and rational pessimism. I have a simple model here that formalizes this view. In that model, an increase in government spending, even in a liquidity trap scenario, is not the correct policy. (The model replicates many of the key properties of a standard New Keynesian model, so it would be hard to discount the model on the basis of its predictions).

The "deficit of optimism" hypothesis espoused by Cowen and others has other potential shortcomings. Among other things, it tends to ignore what transpired just prior to the collapse in confidence. An "Austrian" view is that an artificially low interest rate (via Fed policy earlier in the decade) created an unsustainable over build in capital. The present depression is more like a coming to your senses after a bout of irrational optimism. An alternative hypothesis that generates an observationally equivalent rational expectations outcome can be found in this (unduly neglected) paper by Joseph Zeira: Informational Overshooting, Booms and Crashes.

I wonder, as well, what direct evidence supports the notion of depressed "animal spirits?" In this post, I took a look at the short and long-horizon forecasts (for real GDP growth) in the Philadelphia Survey of Professional Forecasters. Call me stupid if you want, but I don't see long-horizon forecasts jumping around all over the place. Indeed, the long-horizon forecast displays a remarkable stability even through the worst parts of the financial crisis. It is interesting to note that these forecasts imply a decline in what an econometrician would measure as "potential GDP." But if the prior boom was simply a "bubble" (as many in this camp are inclined to argue) is there any reason to want a return to that "false" level of potential? You can't have your cake and eat it too: either we had a bubble and potential is now lower, or there was no bubble and potential remains unchanged.

But what about the high rate of unemployment? Does  this not constitute evidence of a "deficit of optimism?" Is it not evidence of sticky wages and an economy wide lack of demand? I have spoken elsewhere about the sticky wage hypothesis; see here. Moreover, I am not so sure about the "negative aggregate demand shock hypothesis" affecting the labor market when I look at data like this. There's lots of interesting stuff happening at the sectoral level that might account for a lot of what we are experiencing; see also this post by Steve Williamson.

I have argued elsewhere that a good case can be made for redistributive policies that help individuals smooth their living standards across different contingencies. Public works sound like a good idea if they can be justified in NPV terms (though, these interventions are typically sectoral in nature). Extending UI through a deep recession seems like a sensible idea--but please don't be surprised if this leads to higher unemployment rates and extended unemployment durations--and then argue that this is evidence of insufficient demand!

I am, however, deeply skeptical of Tyler Cowen's proposed remedy of higher inflation; or, more precisely, the happy expectation of higher inflation. Don't get me wrong--I think that an unexpected disinflation (or deflation) is likely to be harmful in the short-run (because debt is nominally denominated). The disinflation we have been witnessing has, in my view, little to do with a psychologically driven short-fall in aggregate demand and has a lot to do with an entirely rational increase in the world demand for U.S. government money/debt. Keeping inflation expectations targeted at 2 or 3 percent (or 1 or 0, for that matter) seems entirely appropriate. But there is little reason, in my view, to expect such a policy to "cure" unemployment.

These are, of course, my own views and not necessarily those of the Fed (my current employer). Indeed, as Cowen points out, my current boss appears to have expressed a very different opinion in the past; see here. I don't agree with Bernanke on this point (I visited the BOJ in 2002 and came to different conclusions), and I feel privileged to work at an institution that encourages different ways of thinking about things. Having said this, I have to disagree with Mr. Cowen's portrayal of Chairman Bernanke's current position:
In failing to push harder for monetary expansion, is Mr. Bernanke a wise and prudent guardian of the limited discretionary powers of the Fed? Or is he acting like a too-hesitant bureaucrat, afraid to fail and take the blame when he should be gunning for success?
(I wonder what the Ron Paul supporters must be thinking about that first statement?). I think that the Chairman has made his views pretty clear here: Deflation: Making Sure "It" Doesn't Happen Here. In a nutshell, I do not think that he views Japan as a relevant scenario for the U.S.; see also his Jackson Hole speech:
A rather different type of policy option, which has been proposed by a number of economists, would have the Committee increase its medium-term inflation goals above levels consistent with price stability. I see no support for this option on the FOMC. Conceivably, such a step might make sense in a situation in which a prolonged period of deflation had greatly weakened the confidence of the public in the ability of the central bank to achieve price stability, so that drastic measures were required to shift expectations. Also, in such a situation, higher inflation for a time, by compensating for the prior period of deflation, could help return the price level to what was expected by people who signed long-term contracts, such as debt contracts, before the deflation began.
However, such a strategy is inappropriate for the United States in current circumstances. Inflation expectations appear reasonably well-anchored, and both inflation expectations and actual inflation remain within a range consistent with price stability. In this context, raising the inflation objective would likely entail much greater costs than benefits. Inflation would be higher and probably more volatile under such a policy, undermining confidence and the ability of firms and households to make longer-term plans, while squandering the Fed's hard-won inflation credibility. Inflation expectations would also likely become significantly less stable, and risk premiums in asset markets--including inflation risk premiums--would rise. The combination of increased uncertainty for households and businesses, higher risk premiums in financial markets, and the potential for destabilizing movements in commodity and currency markets would likely overwhelm any benefits arising from this strategy. 

In short, I see no inconsistency here in terms of his earlier policy recommendations for Japan (which I think were wrong) and the policy that he is currently advocating.

Note: I apologize  if my thoughts above appear a little scattered. But I was hit rather severely on the head  with a soccer ball yesterday!

28 comments:

  1. David,

    For evidence on declining nominal economic expectations see here and here.

    ReplyDelete
  2. David:

    Thank you for the links; they are very interesting (looks like you have a great blog by the way -- I am now an official follower).

    With respect to your Great Aggregate Demand Crash, I don't think anyone denies that there was a crash. The question, in my view, is what caused the crash? The policy implications are likely to be different if was psychologically inspired or whether it was a the byproduct of a rational readjustment of expectations.

    I think that your third point on your second link is extremely unfair. The expansion in the size of the Fed's balance sheet has been enormous, and the recent decision to maintain its size (as MBS runs off) is not what most people would call contractionary. The Fed has made it very clear that it wants to avoid deflation. Maybe you think that the Fed should be even more aggressive, but really, as I discuss in my post, are you really so sure that this is the magic remedy? Really? Can you be so sure?

    ReplyDelete
  3. David, thanks for the kind words regarding my blog. I too have your blog in my reader. There are few Fed insiders blogging so it is always a treat to see what you have to say.

    I am not sure I like the "optimism deficit" vs. "rational readjustment of expectations" characterization. Isn't it possible that the rational readustments of expectations could be occuring because the market has come to view current monetary policy as being systematically too tight? Yes, there may be a rational readjustment of expectations due to the realization that there are some serious structural problems (which the Fed can't fix), but why should that be the only thing driving the rational readustment? What can't the rational readjustment also be driven by a belief that the Fed will not intervene to stabilize inflation expectations?

    I don't mean to sound unfair, but being passive and allowing inflation and AD expectations to deteriorate means the Fed is also allowing current AD to weaken. All the increased demand for safe assets (e.g. treasuries and money) amounts to a negative AD shock that is not being offset by Fed easing.

    So how could the Fed do better? It could announce an explicit nominal target and say it will do whatever is necessary to hit that target. (I prefer a nominal income or NGDP target) Monetary policy was able to generate the 1933-1936 recovery which came in a far worse economic environment than today so why not now? Yes, it took radical steps like devaluing the dollar and not sterilizing gold inflows from Europe, but such radicalism helped shake off the deflation expectations.

    To be clear, I am not saying the Fed can solve all our problems. I do believe, however, it can do more to stabilize the nominal macro environment which would make it easier for the painful structural adjustments to occur.

    ReplyDelete
  4. David (I mean A, not B):

    A few points:

    1. There are games where the Nash Equilibrium is not efficient (Prisoners' Dilemma). And there are games where there are two Nash Equilibria, and one is better than the second (Stag Hunt). Both are examples of "coordination failures", but in a very different way. In PD, you would really need to change the payoff matrix in some way to get people to the efficient equilibrium. You need to change the game, in other words. But in SH you don't. Since there is nothing to rule out either equilibrium, a happy face, or sunspot, could get you from one to the other.

    The current coordination failure is a bit more like the second than the first.

    It's not quite that simple. The distinction between those two types of coordination failure is too stark. People disparage the old "stability analysis" of game-theoretic equilibria (you know, or maybe you don't, where we used to do a cobweb around the two reaction functions to see if it converged to Nash), but I think that sort of stability analysis was basically correct, in a game where people are learning where the Nash equilibrium is, because the payoff matrix is not common knowledge.

    Locally the current equilibrium is PD. Globally, it is SH. You need a small change in monetary policy, to change the payoff matrix temporarily, to get us out of the local PD and jump people to the good equilibrium in SH.

    2. "I wonder, as well, what direct evidence supports the notion of depressed "animal spirits?""

    Does the pile of little bricks under my chair from 2 years ago count? I was certainly scared. So were lots of people I spoke to and heard about. Read Garth Turner's blog, for example, and search out all the old posts about eating squirrels.

    ReplyDelete
  5. David, As I was reading your post I thought you were advocating a sort of new classical or RBC model where nominal shocks don't have real effects because wage and price stickiness are no big deal. Then toward the end you said deflation could be bad because it would increase debt defaults. Is that the only reason you think nominal shocks might have business cycle effects? Are you saying that the 50% fall in NGDP between 1929 and 1933 would not have created a recession unless it created a financial crisis? (Assume the 50% fall had been completely caused by tight money, rather than an endogenous response to the banking crisis.)

    It is true that some nominal shocks that seem to lead to recessions are associated with fiancial crises. These include 1930-33 and 2008-09. But there are also plenty of severe nominal shocks that lead to sharp recessions without any financial crisis. These include 1920-21, 1929-30, 1937-38, 1981-82, etc. In fact, in the US the normal recession is associated with a nominal shock, but no financial crisis. Yes, there are a few "supply-side" recessions (like 1974) with no nominal shock, but they are fairly uncommon.

    So I'm wondering if you think any recession resulting from a nominal shock is not caused by wage and price stickiness, but rather results from debt defaults. Note that debt defaults don't directly increase the real MC of production, but sticky nominal wages can.

    Finally, I was surprised to see you reply to David by suggesting that monetary policy must be expansionary because the Fed balance sheet had become much larger. As you know, the standard monetarist argument that more money is expansionary assumes the money is non-interest beearing. The Fed currently pays interest on ERs at a rate above market levels (i.e above T-bill yields) Almost all the increase in the base has been ERs. Obviously if the Fed had intended this move to be expansionary, they wouldn't have paid banks to hold on to the money.

    Indeed in all of American history, they never paid interest on reserves until October 2008.

    ReplyDelete
  6. David B:

    OK, I can see where you're coming from. I think our only point of disagreement might be in terms of the likely quantitative effects of an even more aggressive easing at this point.

    By the way, I have been advocating an explicit inflation target while here at the Fed. If I understand correctly (and I haven't quite figured Americans out yet), there is widespread support in the Fed for an explicit inflation target. Evidently, the "dual mandate" somehow gets in the way of this. I suppose the fear is that an explicit inflation target will be seen as abandoning concern for unemployment...and this is politically unacceptable, especially in the current environment.

    Nick:

    Thanks for reminding me of that distinction. I sometimes wonder whether "coordination failure" might not be better modeled as landing outside the NE payoffs, if you know what I mean? In the stag game, the economy seems, in my view, to be perfectly coordinated on a high or low level!

    As for animal spirits, I am reminded of a business analyst who, in the depths of the crisis, recommended to his viewers his top two positions: cash and fetal.

    There is, however, something that one might call "rational fear." It's hard for me to disentangle emotional vs. rational responses, but maybe you're better at it.

    Scott:

    Please don't think me stupid, but I am not entirely sure what you mean by a "nominal shock." I presume you mean an exogenous increase in money demand, or something like that. This type of shock makes no sense to me. Can you please be more specific? (Btw, I wanted to sign up for your blog, but could not figure out how to do so).

    With respect to the interest on reserves, it is tiny. I highly doubt (though I could be wrong) that setting that rate to zero would have any quantitatively important effect. The problem in the economy has little, if anything, to do with monetary policy (in my humble opinion!). It is the real side that has to get fixed. This was the view I formed when I visited Japan as well. Maybe I should write a post on it. Many thanks for your comments!

    ReplyDelete
  7. David,

    This seems related to this post and the previous one: http://mises.org/daily/4724

    ReplyDelete
  8. Interesting discussion. The notion that the real side of the economy is the problem means different things to different people. For some, it means wait and be patient. For others it means, pull out all stops to accelerate the process.

    But I disagree that monetary is not the problem. The US fed should have raised overnight rates late last year or earlier this year. It should have clearly deployed policy to send a clear message that the end-of-the-earth event of late 2008 was over and behind us.

    P.S.

    David A: First, thanks for sharing.

    Second, I often wonder if public agency employees in the USA have much more freedom to discuss policy than public agency employees in Canada. Canadian line ministries, for example, are horribly politicized and economist-civil servants rarely exploit what the discipline has to offer. Or is the USA perhaps worse and it is simply the privately-owned, arms-length-regulated US federal reserve where you can have this kind of public discussion?

    ReplyDelete
  9. David: your post inspired my "sort of" response. Really just a longer-winded version of my comment above: http://worthwhile.typepad.com/worthwhile_canadian_initi/2010/09/stag-hunt-and-macroeconomics.html

    I like the "cash and fetal"!

    I remember in the 87 stock market crash, though I had little skin in the game, I spent all day listening to all-news radio, trying to feel the fear, and reading old books about the 1930's, trying to create an existential panic in myself. I figured it was part of my education as an economist.

    ReplyDelete
  10. Prof J: Thanks for the interesting link. I agree with the parts I understand; but I do not understand much of what he is saving. I think this reflects the weakness in my knowledge of Austrian economics.

    Westslope: Yes, I should have been more explicit what I meant by the real side being the problem. Yes, perhaps the Fed should have raised rates as you suggested. It is my impression, however, that doing so would have been exceedingly difficult from a political perspective (with unemployment so high). Finally, I do not know the answer to your last question!

    Nick: I liked your post on the Stag Hunt. It is a delightful question, isn't it? "Do We Live in a Tinkerbell Economy?" I'll use it the next time I teach!

    ReplyDelete
  11. David, I agree with Scott Sumner that current monetary policy is too tight. I agree with you that interest on reserves is tiny. I believe the main problem is Fed's exit strategy. Markets believe there is significant risk that the Fed will tighten too early. For example, Sweden has raised interest rates because of the asset bubble risk, even though Riksbank expects that inflation and output gap will remain below target.

    ReplyDelete
  12. 123: Well, I understand and respect your opinion, although I am less sure myself (one way or the other). The "exit strategy risk" is a possibility. Obviously, when you double the side of your balance sheet, stressing an exit strategy seemed like a good idea at the time (and having a contingent exit strategy is still a good idea). The concern as of late, however, has clearly shifted to continued disinflation. The Fed has made it plain, in my view, that it will do everything it can to avoid a persistent inflation below target. To me, this sounds like a loose enough monetary policy!

    ReplyDelete
  13. David,

    Shostak can be like that. Robert Murphy is better in that he doesn't assume he's writing for people who "already know this stuff." By the way, I think Mises institute likes St. Louis Fed. At least, that's where all their data comes from.

    I don't understand anyone who says monetary policy is too tight, though. I guess I'm in your camp. But is there another way to look at it? I look at two (three) things: the expansion of securities holdings (like MBS & bonds), and the target fed funds rate (also the discount window rate). If the first is growing rapidly, and the second (third) are low, are we not ina loose monetary policy regime?

    ReplyDelete
  14. David: two comments, two questions.
    1. Austrian business cycle theory is, as far as I know, about malinvestment rather than overinvestment (in response to your use of the term "over build").
    2. Indicators of animal spirits: various confidence surverys such as the University of Michigan's 'Consumer Sentiment' index (etc).

    You mention that you think Bernanke's (many) policy prescriptions for Japan were wrong; what would you have prescribed?

    How would you measure whether monpol is "loose" or "tight"? I'm aware of several measures, but am interested in your personal view.

    ReplyDelete
  15. David, There are many ways of identifying nominal shocks, but my preference would be a change in either the supply of money or money velocity (closely related to money demand) which causes NGDP growth to sharply diverge from the previous expected path. For instance, prior to 2008 NGDP had been rising about 5% a year for decades. Then after 2008:2, it fell 3% over 12 months, or about 8% below trend. I view that as a negative nominal shock, partly because I favoring targeting NGDP growth rates. Someone who favors targeting inflation, and who thinks the recession was caused by real factors, might be less concerned by that slowdown in NGDP growth. The reason I prefer NGDP rather than inflation as an indicator of nominal shocks, is that prices can rise due to easy money (more AD) or some sort of adverse supply shock. Yet the macroeconomic effects are quite different. When people like Cowen call for more inflation, they really mean more AD, or NGDP, as higher inflation resulting from an adverse supply shock would not boost real GDP at all. So why ask for more inflation, when it is more nominal spending that you really want?

    I suppose this all sounds too Keynesian to you, but I assure you I am a Chicago-trained anti-Keynesian. Sorry I can't help you with my blog. The address is Themoneyillusion.com. I'm told people can get it on their RSS feed, but I have no idea how those things work.

    ReplyDelete
  16. David, I think that having a contingent exit strategy is a great idea. Indeed, when interest rates are at a zero bound, a good contingent exit strategy is the most important thing in monetary policy. I believe that current disinflation was caused by poorly specified contingencies in the exit strategy. In Jackson Hole Bernanke clearly recognized the problem:
    " An alternative communication strategy is for the central bank to explicitly tie its future actions to specific developments in the economy. For example, in March 2001, the Bank of Japan committed to maintaining its policy rate at zero until Japanese consumer prices stabilized or exhibited a year-on-year increase. A potential drawback of using the FOMC's post-meeting statement to influence market expectations is that, at least without a more comprehensive framework in place, it may be difficult to convey the Committee's policy intentions with sufficient precision and conditionality. The Committee will continue to actively review its communication strategy, with the goal of communicating its outlook and policy intentions as clearly as possible."

    It is not enough that the Fed will do everything it can to avoid a persistent inflation below target. The Fed should have committed to return inflation to target in a short defined time frame, until the Fed makes that commitment, monetary policy is too tight.

    ReplyDelete
  17. David,
    here is the address of Scot's blog for your reader:
    http://www.themoneyillusion.com/?feed=rss2

    ReplyDelete
  18. Prof J: It seems clear enough that there exists no consensus on the definition of tight/loose monetary policy. In my view, one should define these terms when they are employed. They then simply become labels, and we can move onto the more substantive issue of what optimal policy should look like.

    Anon: [1] I thought malinvestment was just old-fashioned lingo for socially-inefficient investment. [2] No, survey data does not identify animal spirits; the measured expectations could be based on fundamentals, not psychology. [3] I would have prescribed a long-term inflation target of 0% and I would have urged restructuring in the banking sector. Japan's problems are structural and inflation is not the cure for structural problems. [4] Good question. I'm more accustomed to thinking about optimal policy, rather than what constitutes tight/loose policy. I will have to think about your question (thanks for asking).

    Scott: In a model I posted earlier on this blog site, an "information shock" that led to a rational downward revision over the future return to capital investment led to a jump in the demand for real demand for money (and a corresponding downward jump in the equilibrium price-level). If I simulated this model and showed you the data, you would have identified this event with a negative AD shock. And yet, the recession caused by this is real (in my model). Can you refer me to one of your posts or papers that can help me see how you go about identifying AD shocks? Thanks!

    123 So if there is no explicit inflation target (or no commitment to it), you would say that policy is always either too tight or too loose depending on whether inflation is below or above target, correct?

    ReplyDelete
  19. David,

    Okay, I can get behind leaving behind labels and getting to the meat of the matter.

    Malinvestment, in the Misesian (Austrian) sense, generally means socially inefficient. It also means unsustainable in the sense that too much of one good is being produced at the expense of not enough of another good. I think your use of the term overbuilding is apt here. But the key is one thing was overbuilt while other things were starved for resources.

    ReplyDelete
  20. To clarify the discussion on "malinvestment", let me pose a question:

    Consider Aiyagari (1994, 1995), in which precautionary saving due to incomplete markets and idiosyncratic shocks leads to an inefficient (relative to the complete markets outcome) level of aggregate capital stock. Is this level of investment considered "malinvestment"?

    In other words, how does the term "malinvestment" differ from the neoclassical definition of inefficiency (which just says that any real allocation that differs from the one a social planner would choose under full information is inefficient)?

    ReplyDelete
  21. Joe:

    I was waiting for Prof J to answer you, but I think he's gone from here. Let me try to explain, though I'm no expert.

    Malinvestment literally translates as "bad investment." I presume this means bad in the sense that people are somehow fooled into overinvesting along an unsustainable path. It is unsustainable because at some point, individuals realize the truth about something. It is at this point that the market crashes, with large amounts of capital now laying idle. Exactly how all this fits together logically, I'm not sure, but I'm curious to find out.

    ReplyDelete
  22. David, Thanks for the reply. I think your comment mixes up two issues, which I'd like to disentangle. One is how we define AD shocks. That is completely arbitrary. And the other is how various monetary policy counterfactuals affect the path of employment. In that area I am simply relying on mainstream new Keynesian/monetarist models that say an increase in money demand will tend to reduce employment.

    So let's take an example. You have some sort of real information shock that will cause a recession regardless of what happens to AD. I have no problem with that, although I might question the empirical relevance. To me the real issue is what happens under various monetary policy reaction counterfactuals:

    1. A constant fed funds rate
    2. A constant monetary base
    3. An inflation target
    4. A NGDP target.

    Like many monetarists, I find it convenient to define AD as a rectagular hyperbola, i.e. a given level of NGDP. So that's how I identify AD shocks. But nothing important hangs on that assumption. So let's consider two counterfactuals, stable money and stable NGDP. Suppose first that the Fed targets NGDP growth at a constant 5%. M offsets fluctuations in V. The economy is hit by a real shock and unemployment rises from 5% to 6.5%. So far I am trying to replicate your view. Now assume a different counterfactual. Assume the economy is hit by the same real shock, but this time the Fed doesn't offset the fall in V. I.e. the shock raises demand for base money, just as you assumed. (Perhaps because interest rates fall.) Because in this second hypothetical the Fed does not offset this increase in base money, NGDP falls 3%, instead of rising 5%. My claim is that in this case the very same real shock that would cause a recession in any case, will now cause a much deeper one--perhaps 10% unemployment. Jobs will be lost due to lower AD and NGDP, which were not directly impacted by your real shock. Indeed I think this distinction is an implication of virtually every new Keynesian and monetarist model that assumes wage and price stickiness. I think all those models say that if the Fed doesn't allow the real shock to depress NGDP, unemployment will rise by less than if it does. And one final point, these numbers illustrate pretty much what I think happened in this recession, a real shock that was magnified by the Fed's inability or unwillingness to prevent it from causing money hoarding and depressing NGDP.

    ReplyDelete
  23. David and Joe,

    Not gone, I assure you. I follow very few blogs, and this be one of 'em. But, I do actually have to do my own research (shock of all shocks, I know).

    Anyhow, David basically has the right idea here. Malinvestment = bad in the sense that entrepreneurs make calculation errors and invest in the wrong industries. In our particular recent experience in the U.S. and many other countries (like Spain, for example) the wrong industry was housing.

    It is quite easy to confuse malinvestment with overinvestment because on one level they are similar. Malinvestment means investing in the wrong thing - but this is the same as too much. The reason the word malinvestment is preferred is to indicate that it is not a general level of investment problem, but a misdirection into the wrong industries away from better ones.

    The logic is actually rather complicated, and it relies on the time-structure of production. What I would like to do, rather than get into the full-blown Austrian style, is link this to Long and Plosser (1988, I think), where they had two kinds of goods: intermediate and final. One may think of malinvestment as building too many intermediate goods at the expense of final goods.

    The reason for this entrepreneurial errors have to do with monopoly control of money, but that's another story.

    ReplyDelete
  24. "So if there is no explicit inflation target (or no commitment to it), you would say that policy is always either too tight or too loose depending on whether inflation is below or above target, correct?"
    No. In every case I would say that the current policy is too tight if future expected inflation is too low. For example, current 5 year inflation expectations according to TIPS are equal to 1.24%, and monetary policy is too tight. If these expectations are too high, monetary policy is too loose. Explicit target and commitment is just a tool that reduces excess macro volatility.

    With dual mandate output gap forecasts should also be considered when assessing tightness of monetary policy, here is a good example by Svensson:
    http://www.riksbank.com/upload/Dokument_riksbank/Kat_publicerat/Tal/2010/100504e.pdf

    ReplyDelete
  25. Scott:

    I see. So you simply define AD as PY. Then, an "AD shock" is whatever moves PY.

    I am tempted to say that this is bad language. In a general equilibrium model, any shock is going to have an effect on PY. Ergo, every shock is, by definition, an AD shock.

    On the other hand, it is not necessarily bad language if the language is widely understood and not easily misinterpreted. So perhaps the fault lies with me, for not being familiar with this language.

    ReplyDelete
  26. Scott: I suppose what I said above is not exactly true. There conceivable are shocks that might leave PY unchanged in some model specifications. Still, I'm not entirely sure that this AD shock language is useful. The appropriate response to an AD disturbance my very well depend on the nature of that shock. So we should identify the nature of the shock itself; and not be satisfied with labelling it as something that causes AD (as you have defined it) to change.

    ReplyDelete