Reply to Tyler Cowen, part 2

A few weeks back Tyler Cowen made this observation about my approach to monetary crisis:

Recently Scott Sumner visited us and I pondered the following.

Let’s say that at the peak of a financial crisis, the central bank announces a firm intention to target a path or a level of nominal GDP, as Scott suggests.  If everyone is scrambling for liquidity, and panic is present or recent, and M2 is falling, I wonder if the central bank’s announcement will be much heeded.  The announcement simply isn’t very focal, relative to the panic.  A similar announcement, however, is more likely to work in calmer times, as the recent QEII announcement has boosted equity markets about seventeen percent.  But for the pronouncement to focus people on the more positive path, perhaps their expectations have to be somewhat close to that path, or open to that path, to begin with.

(Aside: there is always a way to commit to a higher NGDP path through currency inflation, a’la Zimbabwe.  But can the central bank get everyone to expect that the broader monetary aggregates will expand?)

The question is when literal talk, from the central bank, will be interpreted literally.

I’ve already taken one stab at addressing Tyler’s post, today I’ll take a couple more.  BTW, you should read his entire post, as my excerpt doesn’t present the full argument.

Part 1.  The root problem isn’t the Fed, it’s macroeconomists.

Most people think of my blog as a critique of the Fed, and in one sense they are right.  But if anyone was crazy enough to read the entire nearly 2000 page blog, they’d notice another theme—a critique of the economics profession.  In previous posts I observed that the Fed policy was usually close to the consensus of macroeconomists.  That’s probably why the consensus of economists almost never blames the Fed for recessions, in real time.  Instead, they say “most recessions are caused by drops in aggregate demand, but this one’s different because of blah blah blah.”  After a few decades pass by, macroeconomists look at the time series data on output, prices and NGDP, and come to the conclusion that it wasn’t different after all.  Most recessions (with the notable expectation of 1974) are now at least partly blamed on a demand shortfall.   The Great Depression is obviously the best example of this sort of re-evaluation, but I am quite confident that future generations of macroeconomists will feel the same way about the big decline in NGDP during 2008-09.  They’ll scratch they heads and ask why economists weren’t demanding easier money in the second half of 2008.

So on one level my blog is a critique of Fed policy, suggesting they should have behaved differently in 2008.  But at a deeper level I am trying to grab the economics profession by the shoulder and shake them up:  “You guys need to think about monetary policy in a different way, the current approach is what caused this crisis.”

You might ask why focus on economists, do they have that much influence?  After all, economists believe in free trade but almost all countries have trade barriers.  But monetary policy is different.  Over time the policy apparatus has been increasingly turned over to economists, as non-economists find it hard to critique, or even to understand, concepts like the Taylor Rule.  (Do you think Pelosi and Barney Frank knew IOR was contractionary when Congress gave Bernanke that tool in late 2008?)  The best example of this phenomenon is the widespread adoption of inflation targeting by central banks all over the world after the debacle of the Great Inflation.  Another example is the Taylor Principle.

Obviously I can’t claim that central banks precisely implement the preferred policy of macroeconomists, after all they don’t even agree among themselves.  John Taylor thought money was too easy in 2003 and Paul Krugman did not.  But my hypothesis does fit the current debacle fairly well.  During the second half of 2008 there was relatively little criticism of the Fed for being too tight.  Indeed at the time I couldn’t find any, although later I did discover a few other like-minded critics, mostly quasi-monetarists like myself.

So here’s one reply to Tyler Cowen’s argument.  If the Fed promises to do what a majority of economists think it should do, then the markets will find that policy credible—even in the midst of a financial panic.  The root cause of the problem in 2008 wasn’t Ben Bernanke’s passivity, it was that the economics profession did not see any problem with the Fed deviating from what I had assumed was their implicit policy—targeting the forecast.

Recall that almost everyone agreed that AD was likely to fall much more sharply than was desirable.  So the problem was not that economists thought that the economy did not need more AD.  They most assuredly thought it did, at least by October 2008.  And the problem was not that interest rates were stuck at zero, they were 2% during the first week of October, and 1.5% throughout most of the remainder of the month.  No, the economics profession knew the economy needed more AD and they knew the Fed had the ability to ease policy further.  Why didn’t they march on Washington and demand easier money?  I can’t say.  I tossed around many theories but in the end I can’t think of anything more satisfactory than the “boo-boo” theory.  Macroeconomists focused on the banking problem, and simply took their eye off the ball.  They can’t walk and chew gum at the same time.  They think that stabbing someone in the gut who also has pneumonia doesn’t make them worse off, because “the real problem is pneumonia.”  And they thought that the Fed letting NGDP fall sharply during a banking panic wouldn’t make things worse because “the real problem was the banking panic.”

Part 2.  Was the crisis really an exogenous shock?

Tyler’s comment uses the fairly widely held assumption that the crash of late 2008 was an exogenous shock, and the issue the Fed faced was what to do about it.  I have two problems with this view.  First, I believe there were two quite distinct shocks in late 2008, one of which was actually caused by the Fed.  Second, I believe that even the other shock, (the banking panic) was partly caused by the Fed’s tight money policy (although not entirely.)

The commenter Cameron has been sending me a lot of quite interesting information recently, and I plan to do a post discussing his work in the near future.  He has uncovered evidence that the stock market was being strongly impacted by monetary policy decisions during late 2008.  For instance, check out his post here first, and then here.  For now I’ll just repeat that while the banking crisis during September 2008 did depress stock prices, the severe stock market crash occurred in early October, during a period of little financial news but strong indications that NGDP was falling sharply and that the Fed did not intend to do anything about it.

I can’t prove this hypothesis, but I believe the most likely explanation for the October crash is that as of September 30, 2008, markets thought the Fed would not allow NGDP to fall at the fastest rate since 1938, and that as of October 10 2008 they realized that they were wrong.  Yes, I know that here I am committing the cardinal sin of attributing my beliefs to the market.  Actually, I am saying the markets were a bit ahead of me because if the market hadn’t crashed, then even on October 10th I probably wouldn’t have realized quite how severe the recession was becoming.  (I suffer from data lags, but the stock market as a whole sees all macro data in real time, even if individual traders do not.)

To conclude:

1.  If the economic consensus had been that the Fed should target NGDP, or even do price level targeting during a financial crisis, the Fed would have done exactly that.

2.  If the consensus was that the Fed should do level targeting in a crisis, and the Fed in fact did level targeting in a crisis, then the markets would have believed the Fed would carry through with level targeting in a crisis.

3.  With level targeting, monetary policy is able to greatly cushion the blow of a financial shock, even if monetary policy is completely impotent in the short run in a purely technical sense.  Thus even if the Fed is unable to raise current M2 or lower current short term rates, level targeting will greatly reduce the fall in NGDP, and thus make the recession much less severe.  There are two reasons for this, one general and one specific to the 2008 crisis.  The general reason is that current AD is strongly affected by future expected AD, and asset prices are one important transmission mechanism linking the two.  And the 2008-specific reason is that because level targeting would have made the asset price crash much smaller, it would also have made the financial crisis much smaller.  Contrary to the conventional wisdom it wasn’t all about sub-prime mortgages, when NGDP expectations plunged in the second half of 2008 the crisis got several times worse, even though the sub-prime fiasco was already fully priced into the markets by mid-2008.

Thus when Tyler Cowen asks whether an aggressive Fed policy would have been credible in the midst of a crisis, my response is that if the Fed had the right policy regime, the crisis they would have been in the midst of would have been much smaller.

I’m trying to change the conventional wisdom on what the Fed should do in the crisis, much as Friedman and Schwartz changed the conventional wisdom about the Depression.  Ideas matter, and this time the Fed did take steps to prevent a repeat of the Great Depression.  With the lessons learned from this crisis we’ll eventually have a new conventional wisdom, and the macroeconomic fallout from the next financial panic will be even smaller.

Does a more stable macroeconomic environment cause the financial sector to take bigger risks?  Yes, but the solution to that problem is not to create recessions, but rather to better regulate the financial industry.  Creating more recessions doesn’t just encourage banks to take fewer risks; it also makes it much more likely that any given level of risk will result in a financial crisis.  As we’ve just seen.  With stable 5% expected NGDP growth the recent sub-prime crisis would have been a footnote in the history books, roughly like the 1980s S&L crisis.  If you don’t believe me go back and read the news from late 2007 and early 2008, when the sub-prime fiasco was well understood.  Only when the Fed let NGDP fall sharply in the second half of 2008 did the banking problem turn into the extraordinary, huge, gigantic, vast, enormous, mammoth, tremendous, titanic, humongous, immense, colossal, gargantuan, stupendous financial crisis of late 2008.



37 Responses to “Reply to Tyler Cowen, part 2”

  1. Gravatar of Doc Merlin Doc Merlin
    2. January 2011 at 16:27

    “or even do price level targeting during a financial crisis”

    We effectively did, early in the crash (but not late), this is what caused the NGDP shortfall in the first place. The FED saw PPI spiking and decided that cost-push inflation was on the horizon. Then they then quickly raised rates, and we saw oil and commodity prices immediately crash, and then housing prices crash. This lead to the banking crisis as NGDP had a shortfall.

    In short, targeting prices is BAD, you need to target long range income expectations.

  2. Gravatar of Benjamin Cole Benjamin Cole
    2. January 2011 at 16:52

    Another fascinating commentary from Scott Sumner.

    As to why did economists stand by when the Fed should have stimulated: Roughly speaking, I think the “monetarists” are “right wingers.”

    Right-wingers are almost constitutionally incapable of recommending “loose” or growth monetary policies, or aggressive government action (save for military invasions and long-term occupations).

    So the left-wing was braying for deficit stimulus, and the right-wing was trying to ignore the problem (it happened on Bush’s watch after all), or even saying it was not a problem. Government inaction is the conservative option, indeed the conservative default mode, regardless of Milton Friedman’s advice.

    Add to that the idea that some conservatives simply do not want Obama to succeed. They like monetary stimulus when Reagan is President, and detest when Obama is president. I wonder if John Taylor has records of blogs from 2003, when he called for the Fed to tighten (before the 2004 election).

    Oddly enough, Obama’s recent tax cuts seem to have left the right-wing in a satiated opium-like stupor, and they have forgotten all about QE2. Sarah Palin even said, “That ship has sailed. Next issue: Scanners.”

    I suspect now that Bernanke has a clear field for QE2, save perhaps caterwauling by Ron Paul, a certified fruitcake when it comes to all things Fed. Paul has a chair with Fed oversight, one of the worst ideas in modern imes.

  3. Gravatar of scott sumner scott sumner
    2. January 2011 at 16:54

    Doc Merlin, Then why did the Fed call for fiscal stimulus in late 2008, if they thought AD was doing just fine?

    But I do agree with you, price level targeting is far inferior to NGDP targeting, as it can give off ambiguous signals.

    (Odd how my important post gets one comment, and my throwaway post gets 10.)

  4. Gravatar of scott sumner scott sumner
    2. January 2011 at 16:58

    Benjamin, Make that two comments. You make good points. The anti-QE2 people are suddenly all gung ho for tax cuts. I thought the economy was in danger of overheating from all that stimulus?

    If Sarah Palin goes after the scanners, she will be performing a great public service. And if she stops talking about QE2, that will be two great public services.

  5. Gravatar of Doc Merlin Doc Merlin
    2. January 2011 at 17:21

    “Doc Merlin, Then why did the Fed call for fiscal stimulus in late 2008, if they thought AD was doing just fine?”

    By late 2008 it was already too late, the tightening began in 2007. I am talking about what lead up to the crisis, not what happened once the crisis started. By late 2008, the legal effects (bankrupcies, forclosures, etc) were already underway and expanding the money supply wouldn’t have helped /much/ due to the momentum of the system.

    I know that you agree with this part: This is why the money supply should be market based rather than centrally planned. The fed isn’t forward looking enough to fine tune an economy, or to properly target money. Because of this, when they do monetary policy, they end up being pro-cyclical instead of countercyclical.

  6. Gravatar of Markus Markus
    2. January 2011 at 17:37


    I have a couple of questions about NGDP futures targeting. For the sake of fairness I should mention that I tend to Austrian Economics, although I disagree with them on many issues. I’ve been reading your blog occasionally for a while and I am starting to like your ideas. For the purpose of this discussion please regard me as a freshman who has only a vage idea of economics. So in case my questions sound silly, keep that in mind, please 🙂 So here are my questions:

    1. You are saying that RGDP follows NGPD. Do you have any evidence for that? Couldn’t it be that NGPD followed RGDP in 2008? After all, you are saying yourself that the (real) recession began in December 2007, but it wasn’t before fall 2008 that we saw NGPD drop. So which caused which to fall?

    2. You are saying that monetary policy was too tight in 2008 measured by the market demand for a looser monetary policy in order to keep NGDP growing. What is the reason markets needed loose monetary policy in 2008? Sometimes they don’t, right? It it just something that sometimes occurs without a specific reason or could this need reflect underlying structural problems that should be adressed?

    3. You are saying the problems are monetary. How do you feel about the trade deficit? Do you think an economy that needs an input of aprox. 3% of GPD worth of goods without selling something in exchange to stay alive is fundamentally healthy? Is it wrong to assume that if the US had more exports, those $400 billion would come back and could be used to pay wages and the like instead of monetary stimulus? Wouldn’t reimporting those dollars by means of exporting goods have the same effect as monetary expansion, but be a much more honest way?

    4. You are calling for a specific NGDP growth target one year in advance. What if something unexpected (terrorist attack, natural desaster) happens within that year that makes the calculation obsolet? Would you change the monetary policy before the year expired? Whould would be the criterion for such an action?

    5. You are saying that if we overshoot (end up to have 6%) we will aim for 4% the following year. What if we overshoot for a couple of years? Will we than aim for -x%? How would that affect markets?

    6. You want to exclude North Koreans because they might vote as a bloc. What makes you think other people wouldn’t? After all, not even very well paid investment bankers seem to be able to evaluate the quality of their own debtors and therefore rely on basically three ‘private’ rating agencies. Do you think it is unlikely that there would be two or three NGDP agencies that make predictions that everybody would follow? Wouldn’t that just shift the power from the FOMC to those agencies? (Given the track record of rating agencies in 2007/2008 I am not sure if I want to entrust them with monetary policy as well)

    Okay, I guess that’s it for the moment! Looking forward to your answers!

  7. Gravatar of derivative_economist derivative_economist
    2. January 2011 at 19:12

    Professor Sumner,

    I’ve given considerable (although casual) thought to the objectives of NGDP targeting and the merits that your strategy of the fed pledging (and meeting) a rate of nominal growth. Here’s the one (substantive?) objection that I’ve come up with.

    If the Fed were to pursue a policy such as this it seems to me it would paper over structural asymmetries and thus create unstable economic trends. For an example separate from the financial crisis, take the GM/Chrystler bailout. Assuming the federal government had allowed the firms to fail, as I’m sure you believe they should have. In the current policy environment, nominal GDP takes an obvious hit. But in a world of NGDP targeting through a combination of aggressive monetary policy and market trust in the Fed’s commitment to its growth target, the volume of economic activity does not decline–with credit channels well-lubricated and confidence flush, investment and consumption (at least collectively) do not fall in nominal terms.

    But one sector of the economy has taken a serious nominal and real hit. If the economy as a whole is to maintain a steady growth trajectory, other sectors must now absorb unexpected amounts of activity–i.e. demand will shift in an unanticipated fashion to other sectors of the economy, creating frictions that only lay the foundation for other collapses. To be bland, the surge of credit/consumption in areas outside of the domestic auto industry might have (hypothetically) fueled overconsumption of real estate, creating a housing bubble much like we just witnessed.

    I’m not sure if you find this situation plausible, but if this is the type of economic environment NGD targeting would foster, it seems like a very unhealthy one.

  8. Gravatar of Tom Tom
    2. January 2011 at 19:28

    Classical economists Henry Thornton and Walter Bagehot thought that financial crises and recessions were interlinked. The best way to forestall both possibilities was for the central bank, as lender of last resort, to credibly commit in all future crises to expand the monetary base sufficient to offset panic-induced falls both in the money multiplier and the turnover velocity of money. Such base expansion would prevent falls in the price level and the level of real activity, thereby stabilizing nominal income. In short, liquidity provision of such magnitude as to preserve equality between money supply and money demand when the latter was rising– that was the remedy both for crises and recessions.

  9. Gravatar of Doc Merlin Doc Merlin
    2. January 2011 at 19:34


    They turned out to be wrong, as the 70’s showed. Monetary expansion by a central bank doesn’t always fix recessions.

    I would argue (probably alone here) that it can’t even fix demand side recessions. By the time the central bank realizes what is going on, its far too late to fix things.

  10. Gravatar of Bonnie Bonnie
    2. January 2011 at 21:23

    Guess who said:

    “I have already said that the Sub-Treasury will reduce the quantity of money in circulation. This position is strengthened by the recollection, that the revenue is to be collected in specie, so that the mere amount of revenue is not all that is withdrawn, but the amount of paper circulation that the 40 millions would serve as a basis to, is withdrawn; which would be in a sound state at least 100 millions. When 100 millions, or more, of the circulation we now have, shall be withdrawn, who can contemplate, without terror, the distress, ruin, bankruptcy and beggary, that must follow.

    “The man who has purchased any article, say a horse, on credit, at 100 dollars, when there are 200 millions circulating in the country, if the quantity be reduced to 100 millions by the arrival of pay-day, will find the horse but sufficient to pay half the debt; and the other half must either be paid out of his other means, and thereby become a clear loss to him; or go unpaid, and thereby become a clear loss to his creditor. What I have here said of a single case of the purchase of a horse, will hold good in every case of a debt existing at the time a reduction in the quantity of money occurs, by whomsoever, and for whatsoever it may have been contracted. It may be said, that what the debtor loses, the creditor gains by this operation; but on examination this will be found true only to a very limited extent. It is more generally true that all lose by it. The creditor, by losing more of his debts, than he gains by the increased value of those he collects; the debtor by either parting with more of his property to pay his debts, than he received in contracting them; or, by entirely breaking up in his business, and thereby being thrown upon the world in idleness.

    “The general distress thus created, will, to be sure, be temporary, because whatever change may occur in the quantity of money in any community, time will adjust the derangement produced; but while that adjustment is progressing, all suffer more or less, and very many lose every thing that renders life desirable. Why, then, shall we suffer a severe difficulty, even though it be but temporary, unless we receive some equivalent for it?

    “What I have been saying as to the effect produced by a reduction of the quantity of money, relates to the whole country. I now propose to show that it would produce a peculiar and permanent hardship upon the citizens of those States and Territories in which the public lands lie. The Land Offices in those States and Territories, as all know, form the great gulf by which all, or nearly all, the money in them, is swallowed up. When the quantity of money shall be reduced, and consequently every thing under individual control brought down in proportion, the price of those lands, being fixed by law, will remain as now.

    “Of necessity, it will follow that the produce or labor that now raises money sufficient to purchase 80 acres, will then raise but sufficient to purchase 40, or perhaps not that much. And this difficulty and hardship will last as long, in some degree, as any portion of these lands shall remain undisposed of. Knowing, as I well do, the difficulty that poor people now encounter in procuring homes, I hesitate not to say, that when the price of the public lands shall be doubled or trebled; or, which is the same thing, produce and labor cut down to one-half or one-third of their present prices, it will be little less than impossible for them to procure those homes at all.”

    Here’s a link if you’re stumped:

  11. Gravatar of David Pearson David Pearson
    2. January 2011 at 23:01


    How much of the October, 2008 decline was due to the financial crisis aftermath, and how much was due to the Fed not doing enough? As someone very much in the markets at that time, I’d say its very tough to reach a conclusion. However, my best guess is that it was something like 60/40 financial crisis/Fed.

    I actually agree that monetary policy loomed large during October of 2008. The failure of the successful October 3 TARP vote to rally markets was a principal cause of the early October decline. In other words, they (the federal government) took their best shot, and it was not enough. So markets turned their attention to what the Fed would do, and, as you say, they perceived that it would not do enough. The markets expected rate cuts, but they did not believe they would solve the problem, and neither would the Fed’s alphabet soup of liquidity measures. BTW, IMO this analysis less than useful because it is ultimately tautological: “How do we know the Fed is not doing its job? The market declines. Why does the market decline? The Fed’s not doing its job.”

    But the attention on the Fed does not mean that there was no “news flow” impacting markets. Rumors swirled around the potential failure of TBTF banks (Morgan Stanley, Citi, State Street, Merrill, and even Goldman after the Buffet injection). For instance, BofA preannounced horrible 3q earnings on Oct. 6, as well as a huge dividend cut, and the market fell steeply that day. Obviously, many European institutions were also in question. The Lehman CDS settlement of 10/24 was thought to be a potentially catastrophic event. Several large hedge funds were thought to be failing (through redemptions) and liquidating enormous stock portfolios. On the macro front, the ISM readings were plunging, and the October Consumer Confidence index fell 23 points to 38 — its lowest reading on record. Some of these events/rumors occurred after the decline in equities, but then, of course, equity markets discount the future.

  12. Gravatar of gnikivar gnikivar
    3. January 2011 at 00:08

    Professor Sumner,

    As far as the monetary policy / financial crisis view is concerned, I’m pretty firmly on the monetary side. The TBTF banks all ended up paying back the loans suggesting that it was a liquidity problem and not a solvency problem. Basically, something that hadn’t happened in over 60 years, a sudden drop in NGDP happened, and banks didn’t have good plan Bs. I think a lot of people find it hard to believe that a housing boom as big as the last one could pop without causing massive problems. However, current asset prices are in part a function of future epected asset prices which are in turn partly determined by NGDP. A sudden and sustained drop of housing prices could simply be more and more investors assuming that we have permananetly moved towards a slow NGDP regime.

    I hope this isn’t a dumb question, but there are certain elements of the NGDP futures targetting that I don’t fully understand. Suppose that NGDP comes in a little bit less than epected, and the Fed has to pay the owners of NGDP futures money. That presumably is monetary stimulus because people may spend that money. You usually tie NGDP targeting with asset price rises. However, I’m not clear on the mechanism that requires higher asset prices. (I’m not an economist, so please keep any eplanations simple)

  13. Gravatar of scott sumner scott sumner
    3. January 2011 at 07:16

    Doc Merlin: You said;

    “By late 2008 it was already too late, the tightening began in 2007. I am talking about what lead up to the crisis, not what happened once the crisis started. By late 2008, the legal effects (bankrupcies, forclosures, etc) were already underway and expanding the money supply wouldn’t have helped /much/ due to the momentum of the system.
    I know that you agree with this part: This is why the money supply should be market based rather than centrally planned.”

    I do agree about the market determining the money supply, but I don’t agree about 2007. Even in September 2008 the Fed could have prevented a severe recession if it had been much more stimulative. Market NGDP forecasts were weakening somewhat, but were far from catastrophic. The mistakes of 2007 were far too small to cause a severe recession.

    Markus, You asked:

    “1. You are saying that RGDP follows NGDP. Do you have any evidence for that? Couldn’t it be that NGPD followed RGDP in 2008? After all, you are saying yourself that the (real) recession began in December 2007, but it wasn’t before fall 2008 that we saw NGPD drop. So which caused which to fall?”

    They tend to move together, so timing can’t answer the question. In a sense, much of macro ever since David Hume has been an attempt to show that nominal shocks have real effects due to wage and price stickiness. There is too much evidence to summarize here, you might start with Friedman and Schwartz’s Monetary History of the US.

    2. We know the economy needed easier money because NGDP expectations were falling. A stable monetary policy keeps NGDP expectations stable, structural problems or not.

    3. The trade deficit was even bigger when the economy was booming. So it didn’t cause the recession. But I do favor a shift toward a high saving fiscal model like Singapore, which would probably reduce the deficit sharply, perhaps producing a surplus. But that’s a separate issue from the recession. We’ve had trade deficits for decades.

    4. No, I would not change the target. If I advocated inflation targeting I would change the target. One argument for NGDP targeting is that you don’t have to change the target when natural disasters occur.

    5. Yes, I would continue “level targeting” even if we got off course for a year or two. This would help cure an overheated economy, or a recessionary economy. So we would actually want somewhat faster or slower growth than normal. And the fact that we were level targeting would reduce the deviations in the first place.

    6. I believe the markets are much better at forecasting than government bureaucrats. The N. Korea comment was really a joke, as they wouldn’t be allowed to participate by their government anyway. In 2008 the markets saw money was way too tight, the Fed didn’t. So I’d rather trust markets than the Fed. You say you are an Austrian–I’d think you like that part of the argument.

    Here’s an alternative take. In 2008 I could have gotten rich selling NGDP futures short. I don’t know about you, but I don’t think it’s easy to get rich. That tells me that NGDP futures prices are likely to be close to the optimal forecast of future NGDP.

    Next time no more than 5 questions!

    Derivative economist, You asked;

    “Assuming the federal government had allowed the firms to fail, as I’m sure you believe they should have. In the current policy environment, nominal GDP takes an obvious hit.”

    Two issues. First, the economy would not take an obvious hit. In fact GM did go bankrupt, I think Chrysler did as well. Big American firms keep producing during bankruptcy. Rather the debts get reorganized.

    But let’s assume there is some shock the greatly reduces output. You still err in assuming NGDP takes an “obvious hit.” The point of NGDP targeting is to make the economy would more like the classical model, decline in one area leads to growth an another. If this cannot occur right away, as you suggest, then prices may rise somewhat and output may fall, leaving NGDP unchanged. I’d still much rather have that outcome then what you’d get with inflation targeting–where something like an oil shock would force the entire rest of the economy into an unneeded recession.

    I also don’t think your analysis could explain something like the housing bubble. That requires an analysis of flaws in our regulatory system. Remember that housing price bubbles in and of themselves are not harmful, only if accompanied by regulatory policies that shovel taxpayer guaranteed funds to banks, who lend it out to sub-prime borrowers. Obviously that sort of system is a complete disaster, with or without NGDP targeting. It’s a pity the finreg didn’t stop sub-prime lending with tax dollars.

    Tom, I agree.

    Doc merlin, You misunderstood Tom, he mentioned keeping prices stable–that’s not what they did in the 1970s, nor were there any financial panics in the 1970s.

    Bonnie, Thanks, I would have guessed Hume.

    David; You asked:

    “How much of the October, 2008 decline was due to the financial crisis aftermath, and how much was due to the Fed not doing enough? As someone very much in the markets at that time, I’d say its very tough to reach a conclusion. However, my best guess is that it was something like 60/40 financial crisis/Fed.”

    There are actually two questions here. How much of the RGDP recession was caused by a fall in NGDP, and how much was caused by the banking crisis. And the second question is how much smaller would the banking crisis have been with an optimal monetary policy.

    I’d guess the recession would be far smaller if NGDP stayed on target (less than 1/3 as big), even with a severe banking crisis. And I’d guess that if NGDP stayed on target, there would have been no severe banking crisis in late 2008, just the milder problems we had in 2007 and early 2008. Combining the two, the recession would have been very mild.

    You said;

    “BTW, IMO this analysis less than useful because it is ultimately tautological: “How do we know the Fed is not doing its job? The market declines. Why does the market decline? The Fed’s not doing its job.””

    Yes, but the very same tautology applies to the gold standard, or Bretton woods. How do we knew the Fed printed too much money under the gold standard? Because the price of gold rose above par. Why did the price of gold rise above par? Because markets believed the Fed printed too much money.

  14. Gravatar of David Pearson David Pearson
    3. January 2011 at 08:23


    You are saying that, like gold, equity prices are driven by perceptions of monetary policy. Therefore, expectations of highly stable NGDP would create extremely stable equity prices (after an initial surge).

    Taking your view to its conclusion, equity volatility would plunge at the individual stock level. At the level of the firm, this means managers have much less incentive to keep a large cushion of equity against the risk of bankruptcy. At the same time, fixed income investors would lever up their portfolios to create excess returns — after all, with little volatility, more leverage is called for in an optimal portfolio. The result would be massive leverage. Ultimately, this would lead to extreme system fragility and risk of a catastrophic event should the Fed’s credibility come into question (i.e., during a future financial crisis).

    I know you probably don’t agree with the above, but I would ask you this: what effect do you think NGDP/Equity targeting would have at the level of the firm/investor? What systemic risks would this behavior create?

  15. Gravatar of David Pearson David Pearson
    3. January 2011 at 08:30

    BTW, the robustness of an ecological system is a function of the heterogeneity of adaptation strategies. In a highly stable system where strategies become homogeneous, the risk of a large shock causing catastrophic extinctions rises steeply. This is not a new idea, so I don’t understand why it is so foreign to economists.

  16. Gravatar of Benjamin Cole Benjamin Cole
    3. January 2011 at 10:51

    I think Scott Sumner slept with this guy’s wife and got caught.

  17. Gravatar of Morgan Warstler Morgan Warstler
    3. January 2011 at 11:27

    Scott previously you’ve told me that if we had targeted NGDP, then you expect we’d have had tighter money 2004-2006.

    How can you square this with telling Doc you don’t agree with 2007?

    Of course, money was too easy 2004-2006. And if it hadn’t been, then there would have been give in the interest rates before we hit the zero bound in 2008.

  18. Gravatar of Dustin Dustin
    3. January 2011 at 11:41

    Benjamin, I’ve learned enough from reading this blog that I can see now that the author of that article is utterly clueless about monetary economics.

    He needs to actually read this blog instead of just the Bio section.

    Who knows? Maybe even somebody as stupid as that author might be able to learn something.

  19. Gravatar of Morgan Warstler Morgan Warstler
    3. January 2011 at 12:04

    Sorry Dustin, Scott gets skewered – and rightfully so.

    Just look at his finance earnings BS.

    If he isn’t going to demand the PURITY of freedom to fail, he’ll never get loose from the power of Gold.

    Scott has resisted at every angle focusing on what can be done to force LIQUIDATION on the housing losers, he doesn’t have the brutal instincts necessary to make policy for the market.

    If Scott’s rally cry was, “eat the losers! and oh btw, target NGDP to 5%.”

    He’d get more right and he’d be more convincing.

  20. Gravatar of CA CA
    3. January 2011 at 12:22

    John Tamny’s critique is pretty slimy, just like Dinesh D’Souza’s article on Obama that also appeared in Forbes.

    You’ll notice that when Scott critiques another’s opinion it’s always the ideas he’s critical of. The good professor never gets personal the way Tamny did. Forbes has become a joke.

  21. Gravatar of CA CA
    3. January 2011 at 12:26

    Scott, a question:

    Had Forbes notified you that this piece was coming out, and have they offered you the space for a rebuttal?

  22. Gravatar of Dustin Dustin
    3. January 2011 at 12:36

    I agree CA. Sumner never has to get personal because he never runs out of ammo. He’s locked and loaded and has the best argument for a better monetary system.

  23. Gravatar of Morgan Warstler Morgan Warstler
    3. January 2011 at 14:32

    slimy = more readers.

    it has nothing to do with the validity of the argument.

    I wish Scott would do more for traffic!

  24. Gravatar of Richard A. Richard A.
    3. January 2011 at 15:09

    John Tamny’s education background:
    Mr. Tamny received a BA in Government from the University of Texas at Austin, and an MBA from Vanderbilt University’s Owen Graduate School of Management. He lives in Washington, D.C.

  25. Gravatar of Alex G Alex G
    3. January 2011 at 21:34

    A credible nominal GDP could in principle damp economic cycles in ‘normal times’ but it is asking way too much to expect NGDP targets to have much stabilising force in the teeth of a true banking panic.

    The problem with macro-economists is that few have much historical sense, and even fewer (Nobel Prize winners included) understand how the modern financial system works.

    To be fair it was natural before the event to assume that a true bank run was impossible in our modern system: it hadn’t happened for over 70 years precisely because deposit insurance, modern central banking and regulation had supposedly abolished the inherent tendency to have periodic bank runs which scarred the the 19th and early 20th century time and time again.

    But the 2008 bank run was a wholesale funding run not a retail bank run, and occurred because of deep changes in the financial system – the growth of the shadow banking system in popular parlance – that means that at least half of the effective “money supply” consists of money like fixed income assets (shadow money). These are largely held by big corporations, sovereign wealth funds and investment managers, and range from short dated commercial paper to all sorts of longer dated paper rendered highly liquid (in normal times) by the existence of the repo market. Those wholesale “deposits” are not insured, are “information sensitive” – the precise condition which used to apply to retail deposits when bank runs were common – and are even more mobile than retail deposits ever were.

    (Better explained here:

    What the sub-prime crisis triggered off was a a chain reaction in which those money like assets became much less capital certain, liquid and money like. In short, the “shadow money stock” fell drastically, an event hardly less deflationary than the sharp fall in conventional money stock that occurred during retail bank runs.

    Hayek and most of his predecessors familiar with the concept of “elastic money” would have understood immediately, yet very few pundits and economists really grasp (even now) this simple and potent idea.


    And as soon as a banking panic violently contracts the effective money supply, no previously credible NGDP target would remain credible without truly enormous (and in this case unprecedented) intervention by the central bank and/or fiscal authorities. Which might in any case fail – or risk failing. And which political reality means cannot easily be repeated.

    At that point your NGDP target simply isn’t “focal, relative to the panic”.

    Truth is there is no quick fix – whether NGDP targeting, a return to gold a la the Tamny rant (obviously a nice fellow if a little handicapped by self doubt), or forcing “LIQUIDATION on the housing losers..”

  26. Gravatar of Tom Grey Tom Grey
    4. January 2011 at 05:28

    This is another important point–the economist consensus is important. The Fed blew it in Oct. 2008. Maybe.
    Let’s see, wasn’t that an election year, when a huge and controversial TARP bailout was given to rich banks who had been making huge bonuses?

    And the majority Dems were in favor, and a majority of the minority Reps were against, but the Rep Pres. Bush was in favor?

    The need or not of TARP, and moral hazard, dominates the Fed’s too tight or too loose in the last quarter discussion. But this chart of Fed changes indicates the Fed was easing pretty fast starting in the end of 2007.

    Professor, do you have a clear graph/chart of what the Fed did do, when, as compared to what you think they should have done at that time?

    Consistent with the more discussed post about finance pay, my own position is that the TARP bailout for the rich was a far worse decision than other Fed actions & inactions in q3-q4 2008.

    Tamny is weak: with the dollar at near all-time lows against gold and nearly every foreign currency I don’t think this is true.

    His anti-Sumner arguments are slimy altho perhaps not so off- target. (Is Scott really such a Keynesian? Maybe I’ve missed the explanations of the misconceptions.)

  27. Gravatar of scott sumner scott sumner
    4. January 2011 at 08:04

    David, You said;

    “You are saying that, like gold, equity prices are driven by perceptions of monetary policy. Therefore, expectations of highly stable NGDP would create extremely stable equity prices (after an initial surge).
    Taking your view to its conclusion, equity volatility would plunge at the individual stock level.”

    Two mistakes here. I am not saying stable NGDP produces stable equities, I am saying NGDP shocks destabilize equities even more than otherwise. But as we saw in 1987, stocks can crash even when NGDP growth is stable. Second, a stable stock market does not imply stable individual stock prices, just as a stable CPI allows oil prices to soar and TV prices to plummet.

    Benjamin, I have a reply.

    Morgan, I don’t see the contradiction. I think money was perhaps a bit too easy in 2004-06, about right in 2007, and more and more too tight in 2008, mostly in the second half.

    If you think Tamny even has a clue about monetary economics, then I am very disappointed in you. Did you read the article of his that I trashed?

    CA, Thanks, I have a reply. No rebuttal or warning was offered. Just a cheap shot.

    Thanks Dustin.

    Richard, Thanks for the info–it’s obvious he’s hasn’t studied any economists since 1930.

    Alex, You said,

    “A credible nominal GDP could in principle damp economic cycles in ‘normal times’ but it is asking way too much to expect NGDP targets to have much stabilising force in the teeth of a true banking panic.”

    This gets things backward, the banking panic was caused by the fall in NGDP. With stable NGDP you have banking problems (like 2007), not banking panics.

    I agree that macroeconomists need to study more history (which is my area of expertise.)

    Tom, Lower interest rates do not indicate monetary easing, that’s a basic mistake I see people make all the time. In December 2007 the Fed cut rates but policy tightened sharply.

    No, I’m not a Keynesian, and Tamny doesn’t know anything about monetary economics.

    I have lots of posts discussing what the Fed should have done, the key was NGDP targeting, level targeting, or at worst price level targeting. They did neither, which was a disaster.

  28. Gravatar of Mattias Mattias
    4. January 2011 at 09:00


    You’ve convinced me that the Fed was behind the curve in 2008. But does that mean that they not only cut interest rates too late and too little, but also that they sold bonds because their (desired) rates were higher than the panicky market would have set if it had been left to it’s own devices? I think Mike Belongia has claimed this but I’ve never seen evidence of it.

  29. Gravatar of Tom Grey Tom Grey
    4. January 2011 at 09:11

    Thanks for answer, Scott.
    However, assuming the Fed would have done NGDP targeting (I’m mostly convinced this is a good goal), what lever(s) do they use for adjustment when it’s high or low.
    You say “policy tightened”, but I don’t think it’s Fed policy that tightened when they were lowering rates throughout 2008.

    But this is where I don’t understand clearly what you mean. For us new(ish) to your blog, briefly repeating your preferred Fed actions, and possibly different gov’t fiscal actions, would be quite educational. (I’d even recommend a brief policy summary you create for future reference. Mine is: all the Big Banks should have gone bankrupt, and the Fed/ gov’t should have given loans to companies needing loans but unable to get them from BigBanks, or local smaller banks.)

    (Also, I half misunderstood Tamny about the dollar being at all time lows against gold, meaning gold is all time high, which is true. Oil, not yet; 1.33 USD to Euro is pretty low; 1.56 USD to GBP is also low.)

    If you keep answering your commenters, you’re sure to keep increasing in traffic! And not remain so obscure … tho as you claim, it should be the argument, not the “name”. But Krugman proves the advantage of having a name (+ Nobel).

  30. Gravatar of Bill Bill
    4. January 2011 at 13:33

    You seem not to know the difference between nominal and real objects. Nobody cares what NGDP is, they care about the real quantity of consumption they can obtain. Nominal consumption, and nominal income, are irrelevant. Under your logic, we’re all fine if real output falls as long as the Fed inflates sufficiently to raise P. That is absurd, and indicates that the students of Bentley University are due a refund.

  31. Gravatar of Dustin Dustin
    4. January 2011 at 15:25

    Scott doesn’t know the difference between nominal and real? Scott thinks we’re all fine during a stagflation?

    Heavens! I want a refund, too, of all the time I spent reading this blog. I had no idea Scott was an economic illiterate.


  32. Gravatar of scott sumner scott sumner
    5. January 2011 at 16:47

    Mattias. In early 2008 they may have sold bonds, but later in the year the main problem was not supplying enough money, plus the interest on reserves.

    Tom Grey, As soon as I get my Nobel I’m sure I’ll have more readers. 🙂

    One key point I keep emphasizing is that low rates don’t mean easy money. Rates are never higher than during hyperinflation, and never lower than during deflation. I prefer to focus on expected NGDP growth rates, others prefer inflation. There is no single tool of monetary policy. You can do monetary stimulus 4 ways:

    1. More money supply from open market purchases
    2. More money supply from discount loans
    3. Less money demand by a lower interest rate on reserves (negative if necessary)
    4. Less money demand by setting a higher NGDP growth target, or doing level targeting, which means promising to catch up for any near term shortfall in NGDP with extra growth later.

    Of course there are many real world complications, but that’s the basics. The Fed should use market expectations (asset prices) to decide when it’s done enough.

    Bill and Dustin, Yes, apparently I’m suffering from money illusion. How ironic!

  33. Gravatar of Alex G Alex G
    6. January 2011 at 19:16

    Scott: Respectfully, that really won’t do. NGDP didn’t suddenly start dropping like a stone of its own accord.

    Banking panics typically occur a short time after a (speculative) boom breaks and the economy goes into recession. But it is the banking panic itself which leads to the violent acceleration of that decline in activity.

    banks, businesses, consumers and investors all take emergency measures to hoard, conserve or generate cash when they become truly concerned about the safety of their money balances, their access to short-term credit and the unknowable extent of counter party risk.

    This behavior is not different in kind from an “ordinary” recession, but it is completely different in speed, intensity and power of contagion.

    NGDP targeting might help generate a quicker recovery once the panic passed – and even mitigate another curious feature of the panic aftermath, which is that inflation worries usually go up in the midst of the deflation, as happened in the 1930s.

    Credible (price) level targeting would probably do much the same trick.

    But the hope that either could prevent the panic itself, and the brutal, compressed decline in GDP that goes with it, is as forlorn as a little boy putting his fist into a head sized hole in the Kariba Dam.

  34. Gravatar of Luis H Arroyo Luis H Arroyo
    7. January 2011 at 02:13

    Markus, I´m not sure if it is an evidence, but the beautiful graph here is a begining to more complicated & mathematical evidence.,GDPC1&scale=Left,Left&range=Max,Max&cosd=1947-01-01,1947-01-01&coed=2010-07-01,2010-07-01&line_color=%230000FF,%23FF0000&link_values=false,false&line_style=Solid,Solid&mark_type=NONE,NONE&mw=4,4&lw=3,3&ost=-99999,-99999&oet=99999,99999&mma=0,0&fml=a,a&fq=Quarterly,Quarterly&fam=avg,avg&fgst=lin,lin&transformation=pc1,pc1&vintage_date=2011-01-06,2011-01-06&revision_date=2011-01-06,2011-01-06

  35. Gravatar of Luis H Arroyo Luis H Arroyo
    7. January 2011 at 02:18

    … I´m not sure, but it seems that RGDP leads the falls, and on the other hand, NGDP leads the ups. That is quite plausible.

  36. Gravatar of Louis F Louis F
    8. January 2011 at 23:53


    First, what was the impact of the huge rise of oil prices in the first half of 2008 on the subsequent economic decline? I’d have to say it was very important. I remember getting very bummed out about $4.50/gallon gasoline. Yes, it cracked in July, but by then the consumer was in full retreat, as were businesses. Add in declining housing prices and it’s no surprise the economy went from wobbly in 2Q08 to free fall in 3Q. By the time the data confirmed the worst, it was too late (go check the sequence of data and revisions).

    Second, although I agree about your observations that monetary policy could have been more aggressive, I think this is too “textbook” and doesn’t have a full appreciation of implementing this in the real world. I think there was a lack of confidence in the Fed at time. Bernanke could have stated that he was goint to get NGDP up to 4%, but who would have beleived him as FNMA and FHLMC and the banking sector was falling apart? For NGDP targeting to work, he had to have a credible plan. How was he going to do it? How were wages suddenly start to rise by 4% with unemployment and joblelss claims rising? How were housing prices suddenly start to rise by 4% as foreclosures were rising? We had 5% inflation in the summer of 2008 and it sucked because it was all in energy and food prices.

    Third, sorry, but in a deleveraging scenario, the losers have to take their hits. They fight to the bitter end, of course; you would, too. None of the banks wanted to mark to market and admit that they were insolvent or at best dangerously undercapitalized; none of the home owners wanted foreclosure. It’s ugly, but better to get on with it–expansive monetary policy can prevent a depression (as it appears it did), but a nasty recession was inevitable in my opinion.

    Overall, society was overleveraged in 2007–too many hedge funds, too much private equity, too much aggressive securities lending, too much housing bought on easy terms, too many credit cards. It was time for Schumpeter’s creative destruction.

  37. Gravatar of scott sumner scott sumner
    10. January 2011 at 19:25

    Alex, Tight money cased the fall in NGDP, which made the financial crisis much worse. The crisis didn’t just happen out of the blue. It was partly caused by bad lending, and partly caused by tight money.

    Luis, Those are nice graphs, very revealing.

    Louis, The oil prices played a modest direct role, but may have played a major indirect role. The direct role was the typical adverse supply shock, which hurt industries like autos.

    The indirect role was that the headline inflation of mid-2008 scared the Fed away from monetary stimulus in the late summer, when it was desperately needed.

    You said;

    “By the time the data confirmed the worst, it was too late (go check the sequence of data and revisions”

    This is why I favor targeting expectations. The markets saw the problem long before the government statisticians, and long before the Fed.

    You said;

    “How were wages suddenly start to rise by 4% with unemployment and joblelss claims rising?”

    I am actually calling for targeting NGDP, not wages. It is possible that wages would have had to rise more slowly to clear the labor market, in which case profits would have risen faster than 5%.

    You seem to have missed my argument for level targeting. It’s about stabilizing expectations to prevent the panic in the first place. It makes no sense to say “Assume there is a panic, why should I believe the Fed can stop it?” I am arguing for policies that would have prevented the panic in the first place. I understand that the Fed wasn’t doing 5% NGDP targeting, level targeting, and hence there might have been some credibility issues, but I am suggesting this is what they should have been doing. But nonetheless, the markets seemed to believe Bernanke would do what was required until early October 2008, long after Lehman failed.

    Even if you believe the Fed couldn’t have stopped the fall in NGDP in late 2008, they could certainly control NGDP expectations for 2010 and 2011. If the Fed had promised to maintain 5% NGDP growth over a longer period (level targeting), then asset prices would not have crashed in late 2008. And this means the banking crisis would have been far milder, as would have the recession.

    You said;

    “Overall, society was overleveraged in 2007”

    What you (and to be fair almost everyone else) miss is that the degree of over-leveraging is strongly correlated with NGDP growth. You simply take it as a given. In your view the crisis happens, and nothing can be done. But what really happened is that the Fed took a small crisis and made it much bigger. Bernanke told the Japanese they were making this mistake, and then made it himself.

Leave a Reply