Dr. Krugman and Mr. Keynes

In earlier research Dr. Krugman pointed out that monetary policy could be effective in a liquidity trap, as long as it was expected to be permanent.  One way of doing that is with inflation targeting (level targeting.)  The Fed should commit to a rising price level path, which could lead to lower real interest rates and higher AD.  Dr. Krugman recently argued that the Fed is too conservative to adopt such a policy today, and things would have to get worse for them to do so.  Of course Ben Bernanke is a student of the Great Depression, and obviously wouldn’t let things get too much worse.  Indeed the so-called “QE policy” instituted in March of this year was a tacit admission by the Fed that things had gotten worse, and that more aggressive steps were necessary.  Since then the outlook has become a tiny bit better, although it’s hard to say exactly why.  (In my view it’s 80% China, 15% QE, and 5% fiscal stimulus.)

What does all this mean?  It means that there is a sort of “Bernanke put” on NGDP.  I don’t know exactly where it is, but I’m pretty sure we were very close to that point last winter.  This also means that any fiscal policy counterfactuals must address the likely monetary policy response.

Every so often Dr. Krugman drinks a secret potion and emerges as the sinister “Mr. Keynes.”  One example occurred recently when he argued that without the massive fiscal stimulus, we would have slid into another Great Depression.  He used a crude Keynesian model, the sort of model that isn’t capable of explaining the level of nominal GDP.  In other words, a model where monetary policy plays no role.  Keynesians think this is justified in a liquidity trap, but even the good Dr. Krugman implied that if things got really bad, the Fed might have taken the sort of drastic steps necessary to get us out of the liquidity trap.

What did the massive fiscal stimulus accomplish?  Who knows?  In my view the Fed would never have allowed a steep deflation.  I think Keynesians overestimate the “multiplier,” but let’s say they’re right, and let’s say that prices would have fallen much more sharply without the fiscal stimulus.  That would have almost certainly caused the Fed to finally wake up and do something effective.  And if they had we might be far ahead of where now, with a budget deficit nearly a trillion dollars smaller.

Just a day later Mr. Keynes showed up again.  This time criticizing Bryan Caplan’s argument that the burden imposed on employers by the health care bill might lead to less employment.  This seems like a reasonable argument, a simple application of supply and demand.  But Mr. Keynes would have none of it.  The General Theory mentions the theoretical possibility that higher wages might not reduce employment; indeed Mr. Keynes once suggested it might even increase employment.  Few economists take this seriously, but Mr. Keynes’ tone of voice implied that Caplan must be pretty dense if he doesn’t accept everything in the GT as gospel.

BTW, Franklin Roosevelt (one of Mr. Keynes’s favorite presidents) tried a high wage policy 5 times.  Each time the policy was adopted during periods of near zero interest rates and very rapid economic growth.  And as I showed in this post, each time the policy brought promising recoveries from the Great Depression to a screeching halt.  I don’t know about you, but I’m not willing to abandon good old supply and demand for a bizarre theory featuring upward slopping AD curves that was tried 5 times, and failed spectacularly each time.

PS.  In the Dr. Krugman post I linked to he hints that the inflation target would need to be higher than 2% to produce robust growth.  That is not accurate.  Because the SRAS is relatively flat when AD has fallen sharply, right now an expected inflation rate of even 2% would imply very robust expected NGDP growth.



37 Responses to “Dr. Krugman and Mr. Keynes”

  1. Gravatar of Lord Lord
    16. July 2009 at 17:28

    It didn’t help employment, but that doesn’t mean it hurt gdp. I doubt it did in fact. It did help productivity. I agree jobs are what are needed though.

  2. Gravatar of Dilip Dilip
    16. July 2009 at 18:40

    So.. when you mention him as Krugman in your posts, you mostly disagree with him. When you call him Paul, you mostly agree with him. When you call him Dr. Krugman, can I take it to mean you are mocking him? 🙂

  3. Gravatar of Dilip Dilip
    16. July 2009 at 18:47

    Scott.. your RSS feed has been hijacked yet again.

  4. Gravatar of Aaron K. Aaron K.
    16. July 2009 at 19:46

    All that showed up on Google Reader for this post was an add for ED drugs. Thought you might like to know.

  5. Gravatar of Alan Alan
    16. July 2009 at 20:50

    I agree with most of the points you’ve raised above, but to be fair to Krugman, I think he was arguing that automatic-stabiliser induced government deficits, rather than discretionary policy, were what has “saved us from a second Great Depression”.

    I had another read of Krugman’s post and I don’t think he’s advocating more discretionary stimulus (at least not in that post; perhaps he does in other posts). I think when he’s arguing for more stimulus he’s referring to being in the form of additional aid to the states (the ‘automatic stabiliser’)

    Like I said, I think your point is valid, and I agree with your general arguments about monetary policy being less expensive and more effective than fiscal policy, in escaping the ‘liquidity trap’.


  6. Gravatar of Current Current
    17. July 2009 at 01:21

    Have you read about Alan Metzler’s interpretation of Keynes? I think it is interesting and it is also a good interpretation of Krugman too.


    If you’re interested.

  7. Gravatar of Current Current
    17. July 2009 at 02:00

    Folks should read Alex in the UK Telegraph today…


  8. Gravatar of Current Current
    17. July 2009 at 02:37

    Bernanke thought he had problems with China.

    Now Chuck Norris has called for greater openness at the Fed…


  9. Gravatar of William William
    17. July 2009 at 03:32

    “This seems like a reasonable argument, a simple application of supply and demand. But Mr. Keynes would have none of it. The General Theory mentions the theoretical possibility that higher wages might not reduce employment; indeed Mr. Keynes once suggested it might even increase employment. Few economists take this seriously, but Mr. Keynes’ tone of voice implied that Caplan must be pretty dense if he doesn’t accept everything in the GT as gospel.”

    Two things came to mind while I read this

    1) The objection against Caplan’s argument was that the actual effects on supply and demand wouldn’t take place until 2013, and it is not reasonable to assert actions in 2013 will make the current recession worse as the current recesion should be over by then

    2) Caplan tried to use Keynes as an authority against Krugman; it appears to me that Krugman was pointing out the discrepancy between what Caplan feels “any sensible Keynesian” should know and what Keynes said.

  10. Gravatar of ssumner ssumner
    17. July 2009 at 03:55

    Lord, Whenever you hurt employment you hurt GDP. The workers losing jobs may be low productivity, but they are not zero productivity. In addition, the data from the Depression that I linked to are output data.

    Dilip, This will be the only time I use the phrase “Dr. Krugman” (a play on Dr. Jekyll and Mr. Hyde) and in this case the views of Dr. Krugman were portrayed as reasonable, while those of Mr. Keynes (his alter ego) were criticized.

    Dilip and Adam, Those ED ads are getting very annoying. I can’t imagine someone would link to my site, find it had been hijacked, and then decided to reward the hijacker by not just buying an ED drug on the spur of the moment, but buying from the exact company that hijacked the blog they were looking for. But I suppose it works or people wouldn’t do it.

    Alan, Good point. But I have a slightly different view

    1. I am 100% certain he does advocate more discretionary stimulus. He has done so many times in his blog.

    2. You may be right about the form of the stimulus. But what bothers me is the implication that if not for the deficits (whatever the cause) we would have had another Great Depression. That is a silly prediction. There is no way Bernanke would have allowed 25% deflation. I think his post also has the implication that those conservatives who opposed the stimulus were wrong. But I agree that this later interpretation is more debatable (if he was simply referring to the automatic stabilizers.

    Current, I’ve read Meltzer’s book and frequently cite it. Keynes is a complex figure and I’m not sure any single interpretation is all right or all wrong.

    I couldn’t open the second link, I’ll try again at work. I disagree with Norris on one thing, I think the income tax was far worse than the Fed. But initially the Fed was worse. I do favor more openness, but don’t know the details of the Ron Paul bill.

  11. Gravatar of Current Current
    17. July 2009 at 04:22

    Scott: “I’ve read Meltzer’s book and frequently cite it.”

    Ah. I haven’t being reading this blog for long enough.

  12. Gravatar of Jim Jim
    17. July 2009 at 06:04


    Sorry, but I’d like to go back to the Cleveland Fed article.

    I brought it up because you have used TIPS as a indicator of expected inflation in your arguments for NGDP targeting. Correct me if I’m wrong, but I seem to remeber you pointing to them as an indicator that the market did not see inflation as a problem at least over the next five years and therefore the Fed’s (or anyone else’s) concern over inflation was unwarranted.

    Now we have a Fed paper that seems to massage the “raw” TIPS data to show that inflation was/is actually higher than what it appears. Therefore, an argument could be made that at least some of the data you were relying on to support your NGDP targeting idea is misleading.

    What really interests me about this is how this data is interpreted or used and what kind of data would be accepted as a reliable measure for NGDP targeting. Thoughts?

  13. Gravatar of Lord Lord
    17. July 2009 at 10:44

    False. You only looked at industrial production and that only for narrow period of the actual increase. Productivity rose and there would have been greater investment for productivity from it. Do not make claims you are unable to substantiate. It only makes your case weaker.

  14. Gravatar of Patrick R. Sullivan Patrick R. Sullivan
    17. July 2009 at 10:53

    Scott, Keith Hennessey is asking for your input, I think:


    ‘Yesterday Senate Minority Leader Mitch McConnell (R-KY) appointed me to be a member of a new Financial Crisis Inquiry Commission. I thank the Leader for the appointment, and will do my best to contribute thoughtful, open-minded, rigorous and responsible analysis and inquiry.

    ‘….In an attempt to become a well-informed member of the Financial Crisis Inquiry Commission, I am seeking input. Please help educate me.

    ‘….I will take help and input from anyone willing to provide it. There are some channels that I know can help me a lot.

    ‘In particular, I would value highly:

    ‘ original writing by individuals with substantive expertise in any of the areas covered by the commission….’

  15. Gravatar of David Pearson David Pearson
    17. July 2009 at 11:59


    Apologies if you’ve addressed this before, but exactly why would the Fed’s 2008/2009 QE be stimulative? The balance sheet expansion was parked as excess reserves dollar for dollar. Can there be a positive multiplier if the reserves were never lent out? Please be specific. Other than ZIRP, has the Fed engaged in any monetary stimulus? Also, how stimulative was ZIRP given that rates for large swaths of the economy have risen since 2007? One could argue ZIRP is less a direct stimulus than an attempt to target velocity by subsidizing bank lending spreads. By all accounts, that attempt has failed.

    You’re arguments seem to contradict themselves. On the one hand, you say only monetary policy can stimulate the economy; on the other, you say that the Fed is making a big mistake by not stimulating the economy (allowing the build-up of excess reserves). Which is it? Is the Fed stimulative or not? And if you believe its not, then is your GDP prediction therefore lower than the Fed’s? Why would the Fed bother to listen to you on NGDP targeting if you both expect the same outcome with or without it?

  16. Gravatar of Bill Woolsey Bill Woolsey
    17. July 2009 at 14:56


    Only monetary policy can stimulate the economy (that is a bit strong in my view, but close.) The Fed isn’t doing it. And it should.

    What is so complicated?

    You seem to have one small confusion. That monetary policy is about increasing loans. Monetary policy is about increasing the quantity of money (the amount of currency, deposits, and the like) beyond the amount people want to hold.

    While this process might result in more bank lending, that isn’t its purpose. The quantity of money is related to the liabilitty side of bank balance sheets. While the banking system must hold more assets to increase its liabililites, there doesn’t have to be an increase in total lending.

    The demand for money is the amount of money people want to hold and has next to nothing to do with what they want to borrow.

    As for what the Fed should do:

    The Fed should stop paying interest on reserves. If the banks still want excess reserves without interest, make the interest negative–charge service fees for holding deposits.

    And, expand base money more. Purchase more T-bills. And then, when those are gone, T-notes, and so on.

    The Fed hasn’t increased base money enough to increase the quanity of money enough, to get nominal income back to target. Paying interest on reserves increases the amount of base money it must create. Reducing the intereset paid to zero or negative decreases the amount of base money the Fed must create.

  17. Gravatar of Scott Sumner on Krugman at catallaxyfiles Scott Sumner on Krugman at catallaxyfiles
    17. July 2009 at 15:25

    […] Sumner, of the Money Illusion, has a great discussion on this point. Every so often Dr. Krugman drinks a secret potion and emerges as the sinister “Mr. […]

  18. Gravatar of David Pearson David Pearson
    17. July 2009 at 17:02


    I disagree. If the banking system holds excess reserves with printed money, there is no stimulus. You say banks need not create a loan asset with the increased liabilities. So what else would they do with the money? They can buy t-bills — which of course is a loan to the Treasury. They can lend it to other banks. They can buy securities. Each of those is a loan, or credit if you prefer. I suppose they could buy commodities or gold, but the seller of that gold would in turn have to either lend it out or return it to the Fed in the form of excess reserves. So I don’t really get your point.

    The creation of excess reserves is not stimulative, and my point is that in the absence of stimulus, Scott’s outlook for GDP growth should be dim indeed.

    The reason this is important is that: 1) the consensus believes the Fed has been stimulative; 2) it has not been; so therefore: 3) all else equal, consensus, and the Fed, is likely to be disappointed by actual growth. If Scott believes that, then why not say so? Why not cast the debate in terms of number of jobs lost, as I said, due to Fed incompetence? Why not single out Bernanke as a failed Chairman who somewhere along the line decided to ignore his own carefully considered, career-making advice!

    By the way, assuming the fiscal stimulus this year is on the order of $300b actually spent, then the true stimulus is $350b – $170b (last year’s amount), or roughly $170b. So we have a situation where fiscal stimulus — whatever its multiplier — is paltry, and monetary stimulus is negligible. This statement, believe me, would be a surprise to the vast majority of economists and market analysts.

  19. Gravatar of Richard A. Richard A.
    17. July 2009 at 21:21

    On July 26, Federal Reserve Chairman Ben Bernanke will join Jim Lehrer in Kansas City, Mo., for an on-the-record forum about how the central bank has responded to the financial crisis and its changing role on the road to economic recovery. The forum will air on the NewsHour during the week of July 27.

    What questions do you have for the Fed chief? Submit inquiries here
    for Bernanke about the Fed and the current state of the economy.

    Scott, maybe you could summit a question asking Bernanke why he doesn’t target nominal GDP.

  20. Gravatar of Current Current
    18. July 2009 at 03:10

    Bill: “Monetary policy is about increasing the quantity of money (the amount of currency, deposits, and the like) beyond the amount people want to hold.”

    Why would it be useful to increase _beyond_ what people want to hold?

  21. Gravatar of Bill Woolsey Bill Woolsey
    18. July 2009 at 04:47


    When the Fed purchases T-bills from a member of the nonbanking public, this increases the balance in that persons checkable deposit and increases that person’s bank’s balance in its reserve account at the Fed. Other things being equal, there is an increase in the quantity of money (the checkable deposit of the seller of the bond at his bank,) and an increase in reserves (the balance at the Fed of the bank of the seller of the bonds.) If the national debt isn’t impacted by this, there is been no increase in lending to the Treasury. The Fed is lending more and the seller of the T-bill is lending less.

    The quantity of money has increased. If it rises above the demand to hold money, spending rises. It could be that person selling the bond to the treasury will use the money to purchase some other security, and so lending increases in that way. But maybe they instead use the funds for a giant party. The “problem” is that there is a shortage of money, not too little lending.


    If nominal income has been depressed by a shortage of money, creating a surplus of money caues it to recover.

    For example, given todays level of output and prices, an excess supply of money should be generated to get nominal income back up to where it was (and I believe on on a 3% growth path.)

    Another way to say this is that the quantity of money now should be at a level consistent with the amount of money people would want to hold if nominal income were on target.

    But, with nominal income too low, that would be a surplus of money given the current level of nominal income.

    My view is that interest rates and output both can adjust to bring the real demand for money to equal the real supply. But those changes in interst rates and output are disequilibria. We are moving interest way from levels that provide for intertemporal coordination (the natural interest rate,) and output away from levels consistent with preferences regarding consumption, investment, and leisure (potential output.) The goods prices and wages were all perfectly flexible, then they could just adjust, but they aren’t. And so, having the nominal quantity of money adjust so it equals the demand to hold money with nominal income “on target,” is the best approach. But if nominal income gets off target, then if the quantity of money is at the right level, then that generates an excess supply or demand for money which gets it back on target. If the quantity of money is too low and nominal income is too low, then an excess supply of money given the current level of nominal income should be created. (Which is another way of saying that the quantity of money should be at a level where there will not be an excess supply or demand for money if nominal income is “on target.”

    The Hayekian Austrians (Selgin, etc,) Sumner, and I all have slightly different views as to what “on target” should be. (I admit that I need to better understand the productivity norm. For years, I thought Selgin advocated constant nominal income.)

  22. Gravatar of David Pearson David Pearson
    18. July 2009 at 05:52


    So it seems we agree. You state that to be stimulative the Fed must exceed the demand for holding money; I state that the Fed must create money that is not parked in excess reserves. Same thing. So you must also think the Fed is not stimulating the economy, which is exactly my point.

    Where we disagree is on the impact of “excess money” on the aggregate banking system balance sheet. If those excess reserves are spent in aggregate, that will show up as a change in composition –and an increase– in banking system assets. Those assets will fall into two types: 1) required reserves; and 2) credit. Can you think of a way that required reserves can rise, in aggregate, without credit also rising? I can’t. Sure there’s the case where all the money is held as vault cash, but really that’s the same thing as ER.

    But again, these points are academic. The question is, “is the Fed stimulating.” A related question: “is there significant deflation risk?”. If the answers are “no” and “yes”, then this Fed is making a monumental mistake, a mistake no different in character than that of the 1930-1931 Fed.

  23. Gravatar of ssumner ssumner
    18. July 2009 at 05:57

    Aaron, Thanks for the tip on the annoying ads, I’ll try to fix it.

    William, Two good points. I agree that Bryan missed the timing of the health bill. Your argument on “Keynesian economics” is also defensible, but I still think my point is valid. There is a difference between “Keynesian economics” and the various assertions made in the GT. Mainstream Keynesian economics does not accept the upward slopping AD curve hypothesized by Krugman. So I think Bryan’s interpretation of mainstream economics is defensible. I might add that you can find something in Keynes to support any point of view. For instance Keynes once said fiat money is the worst possible system, even worse than a rigid gold standard. So someone could use that to defend Austrian arguments as being “Keynesian.” I also think it’s clear that Krugman wasn’t just saying Caplan’s arguments were non-Keynesian, he was saying they were wrong. And apart from the timing issue, Caplan is right, the health bill may well reduce aggregate supply in the economy, and lead to less employment.

    Nevertheless good point, I missed the “Keynesian” angle the first time through.

    Jim, I have always been aware of the liquidity problem. For the past ten years I have been teaching my students that TIPS spreads may underestimate inflation expectations during liquidity crunches. I have also been tirelessly promoting the need for a subsidized NGDP market. Despite many commenters who feel otherwise, I don’t see any reason why the NGDP forecast in a futures markets would be strongly distorted by liquidity considerations.

    In addition, the relevant expected inflation rate is more like 1 or 2 years, and that is clearly much lower than the 5 year forecast. I probably shouldn’t have talked so much about the 5 year TIPS spread, because it suggests that if that rises to 2%, all will be well. I use it because it is easy to find on Bloomberg.com. BTW, there is a CPI futures market, which on theoretical grounds should overstate inflation expectations, as people use it to hedge. And that currently shows an inflation forecast that is higher that the TIPS spread, but still less that 2% for 5 years and less than 1% for 2 years.

    Furthermore, although we don’t have RGDP futures, they would almost certainly show low growth expectations for the next 12 months. I am very confident that if we could measure NGDP growth expectations, they would be far under 5% for the next year. (And my model says we now need 6% growth, because we have fallen one percent behind since February.) But you make a good point, I should not focus so much on the 5-year TIPS spread, as it gives the wrong impression.

    Lord, If IP rises 57% and then falls 20%, you can be sure RGDP is moving in the same direction. I doubt you could find a single economist in the US who would dispute that assertion. Services and farming are mush less cyclical than manufacturing.

    Patrick, I read that quickly and didn’t see anything on monetary policy. Just to be clear, I think monetary policy is very relevant to the issues they are considering, and it speaks volumes if monetary policy is in fact not being considered as a trigger for the late 2008 crash. Did I miss something when I looked at the list of topics?

    David, Monetary policy is very complex and in earlier posts I do deal with this in more detail. My simple take is:

    1. The most effective stimulus would be a price level or NGDP target.

    2. The second most effective would be a penalty rate on ERs.

    3. In the absence of the first two, OMOs may be ineffective. Nevertheless, a substantial amount of QE could have some effect (although perhaps not a lot) in two ways. One is the that at some point it may satiate bank demand for ERs. Also it may indirectly increase inflation expectations by signaling to the markets that the Fed is willing to do something unconventional to boost the economy. I have always been consistent on this issue. My first post right after the March QE announcement was entitled “Don’t get your hopes up” and suggested it might help a bit, but wouldn’t be decisive by itself. That is still my view. And that is also the stock markets view, which rose 2% on the announcement. But I repeat myself, because my view IS the stock market view.

    David, I am more sympathetic to your second comment, but I think you ignored one of Bill’s points. The QE should be accompanied by a negative rate on excess reserves, to insure the money goes into circulation (cash or bank deposits.)
    But I agree that monetary policy is woefully inadequate. I am skeptical of fiscal policy, partly for the policy lags issue implicit in your comment. It takes a long time for the government to actually spend the money. On monetary policy we are actually very close. It may be one of my posts worded something poorly which falsely suggested that I thought the Fed was doing well. Again, I think the Fed caused the crash of late 2008 by errors of omission, or whatever one wants to call them. And policy remains too contractionary.

  24. Gravatar of David Pearson David Pearson
    18. July 2009 at 06:32


    Thanks for your reply. Still, you are implying that QE has produced some stimulus, and I wish you would be specific about it as this is an important point.

    Why is it important?

    The reason has to do with the market’s (and economists’) thinking on the issue of monetary stimulus. If I had to encapsulate it, it goes like this:

    “The Fed is stimulating the economy — a lot — with ZIRP and QE.”

    “This will help produce a recovery in 2H09.”

    “It is not inflationary, however, as all the stimulus is parked in ER.”

    “We should worry about inflation when the economy recovers and the banks start lending out those ER’s, and that is a problem we want to have.”

    Okay, there’s a logical contradiction in this thinking. If we don’t have to worry about inflation because all the money is parked in ER’s, then we ALSO should recognize there is no stimulus because all the money is parked in ER’s. See the problem? The Fed is making everyone believe — inadvertently or not — that it can have its cake (stimulus) and eat it too (no inflation due to all those ER’s).

  25. Gravatar of ssumner ssumner
    18. July 2009 at 06:42

    Richard, I would ask two questions:

    1. Why not put a small penalty rate on ERs to discourage banks from hoarding reserves?

    2. Why not adopt an explicit price level target path so that if inflation fell short one year investors would expect more rapid inflation the next?

    Current, I think Bill meant “monetary stimulus” not “monetary policy.” Bill can correct me if I am wrong.

  26. Gravatar of Patrick R. Sullivan Patrick R. Sullivan
    18. July 2009 at 07:06

    Scott, your mission, should you choose to accept it, would fall under #4 of the commission’s 22 specific areas:

    ‘monetary policy and the availability and terms of credit’

    Hennessey is a very fair minded guy (who has a well written blog with lengthy posts on policy issues). Holtz-Eakins strikes me as another reasonable economist.

    Peter Wallison is, I believe, a lawyer. He has been a burr under the Freddie/Fannie saddle for years. He did an informative hour with Brian Lamb last year:


    Can’t tell you anything about the rest, but there are three guys who I think will give you a chance to make your case.

  27. Gravatar of StatsGuy StatsGuy
    18. July 2009 at 08:37

    Nicely boiled down by BW:

    “Monetary policy is about increasing the quantity of money (the amount of currency, deposits, and the like) beyond the amount people want to hold.”

    But we’re too concerned about assauging bond market and Chinese fears, and pacifying our hard-money ideologues… Every time the Fed hints at doing something halfway intelligent, the media clatter is unbelievable.

    Maybe the solution is to simultaneously expand money and give them extra tools for mopping it up, while explicitly stating the targeting intent?

    It’s possible this will happen, after (in the words of Churchill) we’re tried everything else.

  28. Gravatar of David Pearson David Pearson
    18. July 2009 at 09:02


    As has been argued by some, a penalty rate would not necessarily cause excess reserves to be converted to required. It depends on the banks’ profit expectations from that lending.

    Assume that banks are lending as much as they want to given their desired level of underwriting risk. For them to lend more, the lost income on the penalty rate must be more than the expected credit losses from lowering underwriting standards.

    So assume ER’s are 8% of banking system assets. A penalty rate of 4% would imply a cost of 32bp on these assets. What would be the impact of lowering underwriting standards on losses? Let’s say that 8% is lent out and incurs losses of 5%. Based on that loss rate, then even at a penalty rate of 4%, banks would not convert excess reserves into required reserves. (I’m ignoring lending spread profits, but the concept of expected credit losses offsetting lost penalty profits remains the same).

    So fine, why not make the penalty rate 8%? Because its possible that there is a wide distribution of expected loss rates. If banks think the median is 5% but there is a reasonable chance the number is 20%, then even an 8% penalty rate — which assured — does not offset the chance of 20% losses.

    Of course there is some penalty rate that will force the banks to act. But is there? Can’t the banks merely refuse additional deposits by lowering deposit rates below zero? What would be the impact on the money supply and deflation expectations if this were to occur? One can assume it would not be positive.

    I think you assume that the Fed’s adoption of a penalty rate will also have the effect of lowering credit loss expectations. Possibly, but not necessarily. It could also lead to higher loss expectations if the resulting fall in deposit rates (and t-bill rates) to below zero is seen as a deflationary signal.

  29. Gravatar of travis travis
    18. July 2009 at 09:24

    Scott, you state that the SRAS is flat and thus it would take a lot of growth to produce 2% inflation. Why do you think prices jumped a lot when FDR devalued the dollar against gold in 1933? In other words how is that consistent with this post of yours? (See the ‘update’ part)


  30. Gravatar of Rob Rob
    18. July 2009 at 14:25

    I’m probably going to challenge the EMH more than is warranted here, but only because I can’t help it. Since it seems relevant to your NGDP futures proposal, discussion of it must have some relevance. I know there has been tons of academic work done on it, but as far as I know no one has tried what I’m about to suggest:

    I propose that the best test of the EMH would be to create an electronic market about nothing. One unit of imaginareum is traded at whatever price level it trades at. There are no dividends or interest payments. If the EMH holds either:
    1. No one would participate in this market or
    2. Individual participants would not win more than expected by random luck

    I would be willing to bet big money (a derivatives market on the imaginareum market) that such a market–with a modicum of publicity and public confidence in its integrity–would be highly liquid and that some individuals would dominate such a market more than would be expected by randomness.

    First, why would someone rationally participate in such a market? Because they think they will outsmart the other participants: Just like in poker. Unlike securities markets, no one would need fear that anyone else had inside information. Perfect knowledge would exist, because the only knowledge would be the price history and trading volume.

    Then one could merely analyze whether in fact some participants dominated such a market more than would be expected by randomness.

    In a way, the market would be much like the Texas Hold ‘Em market. The difference would be that instead of matching wits against a few opponents at a time, you would be matching your wits against the group think of the market.(Keynes’s beauty pageant.)

    Just like a poker player might enter a pot with a nothing hand because they want to entice action, some traders would likely take positions to entice action–or because they are outright gamblers. (In practice, you would start the market by letting people risk pennies. From there, the market would grow.) Once some price history was established, traders would look for predictive windows of idiosyncratic price behavior. These windows would open and close as they are exploited into non-existence, creating new windows of idiosyncratic non-random price behavior. Or would the law of large numbers wipe out all such idiosyncratic predictive windows as the market grew? I believe that is the question.

    I suppose the general response to such a market might be: Who cares? But it might shed some light on the influence of “purist speculators” in a market, particularly ones in which there are no hedgers, such as an NGDP futures market would be. For instance: Would volatility in such a market be static? Would there be market crashes more often than expected by chance?

    If it were legal I would be willing to put money into creating such an electronic market. With small commission charges, I believe such a market would be commercially successful. If nothing else, it could be attractive for portfolio diversification, as it would be completely uncorrelated with any other markets. (Or would it?)


  31. Gravatar of Rob Rob
    18. July 2009 at 15:07

    Also: I believe the best way to get your NGDP futures targeting plan implemented is to first get someone to launch the prediction market, then wait for the Fed to start paying attention to it. My belief is that the market would not need to be subsidized. Build it and they will come.

  32. Gravatar of rob rob
    18. July 2009 at 19:22

    as i sit here drunk at the bar, another imaginareum prediction occurs: the market would inflate over time because thw tendency to create paper wealth is hard wired. ponzi schemes are probably in our genetics. vegitable spirits will push the market up instead of down, if only because down is bounded by zero and thus ponzi’s scheme doesnt work in that direction.

  33. Gravatar of ssumner ssumner
    19. July 2009 at 17:44

    Thanks Patrick, I missed that. I’ll take a closer look tomorrow and see what I can do.

    Statsguy, Well put. The thing I find so amazing is that half the people wonder whether the fiscal stimulus will “work,” and the other half worry about high inflation, and yet THEY DON’T SEEM TO REALIZE THEY ARE DISCUSSING THE SAME CONCEPT. It’s like we don’t even talk the same language. Higher inflation would result from monetary policy pushing up AD. And faster real growth would result from fiscal policy pushing up AD. But it’s all about AD. This country really needs to think about whether we want more AD or not. Half the people talk about inflation, and half talk about RGDP. There is no communication. I have a suggestion; let’s all talk about NGDP, and figure out how fast we want it to grow.

    David, I frequently say that I expect banks would put the ERs into T-bills and T-notes. Where is the risk in that? The point is not to encourage “lending” it is to discourage money hoarding. You may have missed that, but I try to mention it fairly often.

    travis, Good question. In those 4 key months industrial production rose 57% and the WPI rose 14%. Both overstate the broader aggregates, and I would estimate the overall price level rose about 7% and real GDP rose about 15% to 20%. But using either numbers it supports my “flat SRAS” hypothesis. I am claiming that real growth would exceed inflation. So a 2% inflation might be associated with 4% real growth, which I think is about ideal (6% NGDP from here to next year, then 5% thereafter.)

    Rob, I think some people do like to gamble, but the marginal investor is probably risk averse, otherwise why would long term rates of return in equities by much higher than T-bills?

    Rob#2, I agree and have proposed that it be tried out experimentally before the Fed risks tying monetary policy to the market. I want to do this in baby steps, with the Fed able to pull back quickly if something is clearly going wrong. Surprisingly, the proposal is very flexible in that way. We could even let the Fed gamble a bit if they thought investors were off track.

    rob#3, Stop drinking right now.

  34. Gravatar of David Pearson David Pearson
    19. July 2009 at 19:54


    How can the banks in aggregate lend out ER’s to the government if the government doesn’t borrow more money? At rates of close to zero, t-bills and cash are close substitutes, so its not clear that sellers of t-bills would want to spend the money. So again, the money shows up as a deposit to the banking system, and again, banks must lower their lending standards in order to lend the money out.

    Of course, the government may want to borrow more money. Doesn’t this require an increase in the deficit? And if it does, then isn’t fiscal stimulus necessary to make monetary stimulus work? You have stated repeatedly that it is not.

    Perhaps you feel that t-bills and cash are not perfect substitutes. After all, the rate on a one month bill is, what, 12bp? That 12bp would disappear and become negative if banks try to buy up bills, in which case cash would be a superior investment, never mind a perfect substitute.

  35. Gravatar of ssumner ssumner
    20. July 2009 at 17:12

    David, What would be so bad about negative rates on T-bills? Don’t most economists think that lower rates would be a “good thing?”

    Banks can buy existing debt, there is trillions in circulation, so the government doesn’t need to issue any more.

    Once again, however, the key is not negative interest rates, it is an inflation target. We don’t have one, the markets don’t trust the Fed to raise the price level 10% in 5 years, in my view we’ll be lucky to get 5%.

  36. Gravatar of travis travis
    22. July 2009 at 13:39

    Scott, Not to belabor the point, but do you agree with the ‘bottleneck’ view of inflation? It seems that in the update mentioned in my first comment, you don’t. “Others might argue that Keynesians have a “bottleneck” view of inflation, which occurs when output pushes against capacity. Unfortunately the most perfect monetary experiment in American history, the 1933 dollar depreciation program, decisively refuted that mechanism.”

    Your statement that the SRAS is flat seems to imply that you expect low inflation because of excess capacity — an absence of a bottleneck. The only way that I make sense of these two positions is that you think there will be *some* inflation, while a pure keynesian wouldn’t expect *any* inflation. Is that a correct intrepretation?

  37. Gravatar of ssumner ssumner
    23. July 2009 at 05:49

    Travis, Yes. And the data strongly support that intermediate position. Some inflation, but less than at full employment.

Leave a Reply