Reply to Kling on Tobin

Arnold Kling raises some very good questions regarding my recent post on James Tobin.  I had said:

I wish Tobin had lived long enough to have something to say about the role of money in the current financial crisis. As you know I am one of the few people who think tight money, defined as a policy of suddenly driving NGDP growth 8% below trend, was the most important factor in the financial crisis,

And Kling replied:

I strongly doubt that Tobin would say that Ben Bernanke carried out a policy of “suddenly driving nominal GDP growth 8 percent below trend.” I don’t think Sumner really wants to say that, either. At the risk of putting words in his mouth, I think that what Sumner wants to say is that there was a shock to velocity in 2008 that was going to drive expected nominal GDP growth 8 percent below trend, and the Fed, instead of doing something to offset that shock, mindlessly decided that with nominal interest rates already low there was nothing it could do.

Kling is being very polite in trying to give me another chance to recant my absurd argument, but I remain as stubborn as when Jimmy Carter gave Ford another chance, and Ford kept insisting that Poland wasn’t under Soviet domination.  I do really want to say the Fed drove NGDP 8% below trend with tight money.  Of course they didn’t actually want NGDP to fall, but they certainly could have taken any number of actions to prevent it, and decided that the risks of inflation were greater than the potential benefits of more aggressive rate cuts in September, October, and November 2008.  They also decided to pay interest on reserves in order to prevent a breakout in inflation (according to James Hamilton.)  Even the Fed admitted that the program’s intent was contractionary.  So one can hardly argue that the Fed didn’t do more in late 2008 because they thought there was no more that they could do.  Whether additional steps would have worked is of course an open question.

Kling then states:

I doubt that Tobin would have embraced recalculation, but I doubt that he would have said that the Fed could have done something about nominal GDP in the very short run (two or three quarters).

I have no idea what Tobin would have said.  However, it is well understood that if monetary policy affects AD in the long run, then, ipso facto, it also has an effect on near-term AD.  This is because near-term shifts in AD are (according to Woodford) strongly influenced by changes in future expected AD.

Kling then quotes from my discussion of sticky wages:

It’s like musical chairs; when the music stopped there weren’t enough dollars of NGDP out there to pay the debts and salaries that had been contracted under much different expectations.

And he doesn’t find my metaphor very convincing:

As to salaries, I just don’t think that the rise in real wages (less than 2 percent) is enough to explain 10 percent unemployment. When you say that debts were contracted under much different expectations, you make it sound like we had farmers taking out loans to plan crops when prices are 100, and then by the time the crops are harvested the prices are at 80, due to general deflation. That just does not seem like the story of the last year.

There are a couple problems here.  I wasn’t talking about real wages.  This is because I don’t trust price indices, and in any case I think the relevant variable is the relationship between nominal wages and NGDP, not nominal wages and prices.  If NGDP falls 8% below trend, then nominal wages must also fall 8% below trend to maintain full employment, regardless of what is happening to real wages.  But I also disagree with his characterization of the current situation.  At my company we did not receive any pay cut in the period from mid-2008 to mid-2009.  But if our wages had been flexible, rather than negotiated once a year, I am pretty sure we would have had a pay cut.  Why do I say that?  Because the first opportunity for wage cuts after the crash of late 2008 occurred in April 2009, when our salaries were set for the period from July 2009 through July 2010.  And we got no pay increase, which was 4% below trend.  That was a pretty good indication that even before July 2009 our wages were above equilibrium, they simply had not yet had a chance to readjust them yet.  And even this 4% pay cut relative to trend wasn’t enough to offset the sharp fall in NGDP.  There is definitely some money illusion involved when it comes to nominal wage cuts, and this imparts even more rigidity to nominal wages.  Another zero bound!

Of course Arnold Kling is famous for his recalculation theory, and here he adds one additional factor that might explain the severity of the recession:

I’m still wrestling with the issue of what caused such a deep recession. Another thought is that the domestic auto industry took another ratchet down in what has been a long-term trend. Manufacturing employment in general continued to ratchet down.

There is just one problem with this argument, there is no long term downward trend in domestic auto sales, indeed according to a 2006 article in Business Week, just the opposite:

It’s easy to get the impression that domestic car manufacturing is headed toward the same inevitable extinction as American textile, apparel, and consumer-electronics production.

But the reality is more nuanced. Look past the trouble in Detroit, and the auto industry is anywhere but in decline. In a growing number of Southern hamlets such as Canton, Toyota (TM ), Honda (HMC ), Mercedes (DCX ), and other foreign car manufacturers are providing nonunionized jobs — 33,000 since 2000 — that pay almost as much as United Auto Workers earn farther north. Consumers are enjoying more choice than ever, while the market as a whole is humming. Car sales in the U.S. inched up last year, to 17 million vehicles, the third-highest ever.

And get this: Even as Ford and GM cut production last year, North American plants built 15.8 million cars and trucks, the same as in 2004. That happened thanks to foreign carmakers producing 4.9 million vehicles, an increase of 500,000 from 2004. Overall production is expected to rise to 16.8 million by 2009, when an estimated 5.8 million vehicles will roll off foreign-owned assembly lines. Looking back, car and truck production in the U.S. has nearly doubled since Detroit’s heyday in the early 1960s. “The domestic auto industry is as healthy as it has ever been,” says Eric Noble, president of Car Lab, an industry consulting firm in Santa Ana, Calif. “The names on the plants are just changing.”   (Italics added.)

Of course the auto industry never reached its production goals of 16.8 million units in 2009.  It’s tough to produce that many cars when NGDP is falling at the fastest rate since 1938.  Kling may have been referring to the long run decline in auto employment, but if the current recession was due to long run trends in productivity, we should see more autos being produced with fewer workers, as that is the long run trend.  Instead we are seeing a sharp drop in auto production.

Some will argue that correlation doesn’t prove causation.  Sure NGDP and RGDP are correlated, but there could be some other reason.  The problem with that view is that when we have evidence that a monetary shock is exogenous, it seems to produce a real effect.  There are good reasons why the best and the brightest macroeconomists from David Hume to John Maynard Keynes to Milton Friedman to Robert Lucas to Michael Woodford have been convinced that falling NGDP will depress real GDP in the short run.  If anyone has a better explanation for the correlation, I’d love to hear it.



26 Responses to “Reply to Kling on Tobin”

  1. Gravatar of TGGP TGGP
    17. January 2010 at 19:14

    Usually the Hume story involved a changing supply of money having no effect, as in Arnold’s story. On the other hand, I didn’t read Hume in the original.

  2. Gravatar of OGT OGT
    17. January 2010 at 20:19

    A minor point on the interest on reserves. I haven’t been able think of a rational argument for Bernanke doing this other than to do a backdoor recapitalization of banks. I recently noticed a paper refering to some academic work of Bernanke’s characterizing the financial system as an amplification mechanism of economic shocks, which is in line with that behavior.

    Of course, banks would be better off with higher NGDP, but if one views it has being severely more difficult to engineer that rise without stabilizing the banks then one is in a quandary.

  3. Gravatar of rob rob
    17. January 2010 at 20:26

    I still dont understand what caused the sudden increase in demand for money in mid 2008. Was it the overleveraged position of the banking system combined w the housing crash? If so, then it seems fair to suggest the fed failed to act accordingly to avert the disaster, but it doesnt seem entirely correct to say that they caused it.

  4. Gravatar of Tom Hickey Tom Hickey
    17. January 2010 at 20:47

    Scott, revisiting your proposal for the Fed to adopt QE to increase NGDP, it did and the monetary base is now over 2T, much to the consternation of a lot of inflationistas, some of whom are even warning of impending hyperinflation. That doesn’t seem to be materializing, and the deflationistas seem to be winning this round, with real interest rate still higher than nominal.

    I’d be interested in hearing your assessment of the effect, so far and whether you anticipate the increase in NGDP you expected. Thanks.

  5. Gravatar of Cameron Cameron
    17. January 2010 at 22:58

    It seems to me the primary difference between you/Woolsey and Kling/Meltzer approaches are that you tend to favor rational expectations while Kling favors adaptive expectations.

    From an adaptive expectations perspective NGDP may be too low, but a more expansionary monetary policy won’t really have any immediate affect. An adaptive expectation believer might ask “Do you really think businesses set prices and fire employees based on what the fed is saying and doing?” Businesses and individuals will “believe it when they see it” and therefore it’s difficult to tell if monetary policy is too tight or too easy at the moment. Any additional easing will be too late and will simply show up as inflation later.

    It’s not really surprising that Kling is wrestling with the severity of the downturn under this belief… although one could argue many businesses saw the credit crisis up close and personal and expectations fell more dramatically than they would have if the Fed simply made a future NGDP lowering move.

    Someone closer to rational expectations might point to your Great Depression dollar devaluation example or movements in the stock market/TIPS spreads after fed moves. If RatEx is true(or even partly true) then of course the fed can influence long run NGDP expectations instantly and therefore it can influence short run NGDP fluctuations.

    So thats my simple summary of the debate. I don’t mean to put words in anyone’s mouth, and I’m probably wrong or missing something… but that’s how I see it.

  6. Gravatar of DanC DanC
    18. January 2010 at 08:16

    I must admit some confusion.

    While I think the recalculation story has some merits, I don’t understand the blind spots Kling has toward changing political markets. I think Cochrane and Taylor have talked about government interventions distorting markets and getting people to game the system. Or the Barro argument that forward looking investors see the huge growing government deficits and adjust today for the tax bill tomorrow.

    With higher marginal tax rates coming are some Americans adjusting toward more leisure in the near future (as in the European model)? When work is taxed do some work less, and need fewer workers to help with their tasks?

    So to Professor Kling I would say that executives are not saying “We are in for bad times, I need to cut costs in order to survive.” Instead they are saying we are in bad times and I have a huge tax bill coming due in the near future, I need to run my enterprise in a way that will maximize my after tax income, and I’m not sure how to do that.”

    Again to Professor Kling we are having an adjustment problem, adjusting to a new political, regulatory and tax reality.

    I read Professor Sumners work and I find much of it compelling. I’m not sure how you get his ideas into action. I haven’t read that many of his posts, and I would encourage him to write an accessible book on the topic, but I’m unsure how you get inflation expectations to change.

    Moreover, even if the Sumner model would work to end a recession in most situations could it overcome the planned changes in government spending and taxes. Or will the rush toward a European style economy significantly change incentives and growth going forward.

    After all, an implication of the Obama health plan is that technology has advanced to the point that future advancement in medicine is no longer cost effective. That the marginal benefits of most health care research fails a cost benefit analysis. I heard one of his leading advisors claim that surgical techniques have advanced as far as we can expect.

    Given such a view, who wants to invest in medical technology. Or look how some in government brag that the Bush pharmaceutical plan did not cost as much as projected, because we are seeing fewer breakthrough drugs under the new compensation system. And this is a good thing?

  7. Gravatar of Philo Philo
    18. January 2010 at 08:20

    Rob writes: “it seems fair to suggest the Fed failed to act accordingly to avert the disaster, but it doesn’t seem entirely correct to say that they caused it.” I think this quibble about the notion of causation is also part of what Kling has in mind when he objects to saying that the Fed suddenly *drove* NGDP 8% below trend. There is an intuition that the forces that drove NGDP down had nothing to do with the Fed; the Fed simply *allowed* those forces to prevail, instead of *opposing*, *countering*, or *neutralizing* them.

    I am suspicious of such intuitions about causation. They seem illegitimately imported from physics, where the notion of *force* has a fundamental status that it does not have in the social-political sphere. There may also be a moral illusion: the idea that one is less culpable for *failing to counter* destructive forces than one would be for *oneself providing the fundamental destructive impetus*. (At best, this is a rough rule of thumb, not a fundamental principle. And much of the Fed’s responsibility–its *raison d’être*–is to counter such forces.)

    In any case, if we think the Fed’s actions were a factor–a decisive factor–in the fall of NGDP, we need not worry whether they should be called “the cause.” The real issue between Sumner and Kling concerns whether they were, indeed, a decisive factor: Kling thinks NGDP would have fallen about the same amount, whatever the Fed had done in 2008.

  8. Gravatar of David Pearson David Pearson
    18. January 2010 at 09:33

    It would be helpful, in establishing causality, if you delved into the chronology of 2008. When, exactly, did the Fed receive the signal that it was “crashing NGDP?” What, exactly should it have done differently at that time??

    The stock market, we can all agree, I think, has embedded in it an expectation of NGDP growth as a proxy for corporate revenue growth. The S&P peaked in October of 2007 at around 1540. In July of 2008, it reached 1440. That’s a 6% decline over the space of nine months. Was this a signal that NGDP was crashing?

    Oil prices reached an all-time high of $147 in July of 2008, and in that month the broader CRB index of commodity prices also reached a peak that was 33% above its prior peak. The 10-yr TIPS inflation spread reached 3%.

    So, based on the above evidence, we could conclude that the Fed is innocent of all charges of crashing NGDP expectations as of July of 2008. What happened in August? The failure of Fannie and Freddie, which occurred despite all protestations from the Fed, Government and Wall Street that these institutions were not at significant risk, that the “sub-prime crisis was contained.” In response to the erosion of confidence in Government and Wall Street caused by the potential failure of these two giants, you argue the Fed should have adopted an inflation target of 3%, and that this would have supported expectations. I would say in an ideal world where NGDP targeting had credibility as a policy, then perhaps. In the real world — the world of asymmetric policy and the Greenspan Put — the failure of confidence in large institutions would steamroll over any Fed policy promise.

    So what could the Fed have done differently between August and October of 2008? Would it have worked? How?

  9. Gravatar of rob rob
    18. January 2010 at 10:35

    Philo, you make a good point, but I am left wondering what exactly the fed should have done and exactly when. The market made its massive drop in only a few days. Since expectations are key, this would mean that investors didn’t expect the Fed to stabilize NGDP. If the Fed had reacted to this market drop as a reason to suddenly boost NGDP expectations, without a NGDP futures market how would they know exactly what to do and whether or not they had accomplished their goal? After all, the markets rebounded on Oct 10 — might the Fed have thought NGDP expectations had stabilized at this point?

    If the Fed failed to prevent the collapse in NGDP expectations in mid 2008 — can it really be said that that failure occurred IN mid 2008 — or did it occur much earlier, in the sense that the market didn’t seem to EXPECT the Fed to stabilize NGDP.

    I understand that Scott argues that the short run effects tend to be underestimated — but how short run are we talking? Isn’t the key to expectations that the Fed make a “credible commitment”? Can they really make a “credible commitment” during one week in October?

  10. Gravatar of David Pearson David Pearson
    18. January 2010 at 10:46


    BTW, I would also focus on what the government was trying to tell people in August of 2008: that the GSE’s would require less than $100b in funds, most of which would be repaid, that no other institutions were at risk, that the fundamentals of the economy were strong, etc. All of this falls under the category of “expectations setting”, and by and large it worked. By September of 2008, the stock market had fallen barely 8% more from its July peak. So, my question is, would massive Fed QE in August and September of 2008 have “fit” with the narrative that the Fed and Treasury was trying to spin: that “everything was fine”. I know plenty of people in the markets that were looking at Fed actions as a potential sign of “panic”; a signal that things might be, in fact, a lot worse than government was letting on. In fact, this “panic” QE move would likely have been seen by market actors as a sign that long-term inflation expectations should skyrocket. If the Fed “panics” when “everything is fine”, then what should we expect future inflation to be?

    Of course, in a world where NGDP targeting has credibility, I agree, things could have turned out differently. The Fed did not have that credibility then, and it does not have it now.

  11. Gravatar of Tom Hickey Tom Hickey
    18. January 2010 at 10:46

    Causality is a notoriously difficult issue to pin down since David Hume observed that correlation does not imply causality. The current understanding, in a nutshell, is that causality is expressed in terms of “scientific laws,” that is, conditional statements (hypotheses) that are embedded in the structure of a scientific theory that supports them and which has shown itself capable of making predictions confirmed by facts that are consistent with the body of the theory. Economics has not reached this point.

    Causality can be difficult to pin down in the physical and life sciences, and much more difficult in the social sciences, owing to a variety of factors, not the least of which is dynamic complexity that often renders static models too simplistic and data too difficult to gather and process into information for use in empirical confirmation. An overarching paradigm in which normal science is pursued is also lacking. The social sciences still look a great deal like philosophical debates based on competing assumptions.

    And there is always the temptation to presume that an apparently sufficient condition is a cause.

  12. Gravatar of rob rob
    18. January 2010 at 11:54

    Scott, you said:

    “the Fed could have taken any number of actions to prevent it”.

    Could it? Because they key is the signal those actions would have sent the market, not the actions themselves per se. As David mentions above, the wrong action might have sent a signal of desperation. For example, when the bank of England tried to shore up the pound in 92, its action merely showed a hand of desperation and had the opposite result as intended, or so goes one narrative.

  13. Gravatar of ssumner ssumner
    18. January 2010 at 12:25

    TGGP, No, Hume says a fall in velocity has the same effect as a fall in M. That was the subject of my second post in February.

    James Hamilton (who unlike me doesn’t say a lot of wild and crazy things) says the Fed did it to prevent high inflation. I think he is right, that was the reason.

    rob, It was a combination of a low nominal rate on T-bills (low op. cost of holding cash) plus interest on reserves, plus banks wanted more liquidity because of the crisis conditions.

    Tom, First of all, they did not do what I asked for. As Hamilton argued, if you pay interest on reserves you aren’t doing QE. When I made the proposal the monetary base had already doubled, so I knew that that alone wouldn’t be enough.

    I expect low inflation this year and next, and I anticipate that will increase support for my position, against the Austrians and monetarists who predicted high inflation. They will have no place else to go, unless they want to become left-wingers.

    Cameron, Those are good observations, and I’d just add one point. In mid-2008 we didn’t need to sharply raise NGDP growth expectations (which is hard to do) rather we needed to keep them from falling (which is much easier to do.) So that does bear on one of the issues you raise—how easy would have been for the Fed to boost NGDP growth expectations.

    DanC, Good points, but remember that the recession itself is one reason we are getting these statist policies. That’s one reason I so strongly opposed letting NGDP fall. I have studied the Depression much of my life, and saw how it also led to left-wing policies.

    You said;

    “I heard one of his leading advisors claim that surgical techniques have advanced as far as we can expect.”

    I realize you are just quoting others here, but can’t resist the mentioning statement a few decades back that no home computer would ever need more than 64 kilobites.

    Philo, I agree, and clearly so does Tobin with his rudder/compass distinction. Look, inflation targeting is a policy. If the Fed allows deflation (as they did in late 2008 and early 2009), that is a policy change, especially given it occurred when their other objective was also below target.

    David Pearson, I did some very long posts with timelines last summer. You might want to check them out. You said;

    “The stock market, we can all agree, I think, has embedded in it an expectation of NGDP growth as a proxy for corporate revenue growth. The S&P peaked in October of 2007 at around 1540. In July of 2008, it reached 1440. That’s a 6% decline over the space of nine months. Was this a signal that NGDP was crashing?

    Oil prices reached an all-time high of $147 in July of 2008, and in that month the broader CRB index of commodity prices also reached a peak that was 33% above its prior peak. The 10-yr TIPS inflation spread reached 3%.”

    I’ve always argued that their policy in July 2008 was reasonable. The first big mistake was at the September 16th meeting. 5 year inflation forecasts had fallen to 1.23% and real growth forecasts were also falling fast. They didn’t cut rates. Then they tightened policy with the interest on reserves program in early October, as the stock market was crashing; another huge mistake. They let monetary policy drift and focused on bailing out banks, not realizing that fast falling NGDP expectations were hurting banks much more than bailouts were helping. They shoveled water out of the boat, but it was coming in through a hole even faster. To plug the leak they had to commit to higher NGDP in 2009, instead they made it clear NGDP was falling, and there wasn’t anything they planned to do about it.

    If in the “real world” a Fed commitment to NGDP targeting, level targeting, would have had no credibility, then we need a whole new policy regime from Congress. But I think it would have had credibility. If the markets had not been hopeful the that the Fed was going to do something, the crash would have occurred even earlier than October—it would have occurred after Lehman failed and the banking system froze up.

    David#2, The markets don’t get scared when the Fed “panics,” stocks often soar when the rate cuts are bigger than expected, or when FDR devalued the dollar in 1933. The markets crash when the Fed doesn’t panic, when they smugly sit back and act like all is well as NGDP expectations are collapsing.

    And they did not need QE in September, they needed a sharp rate cut and a promise to hit an inflation of NGDP target, level targeting.

    Tom, I agree.

  14. Gravatar of David Pearson David Pearson
    18. January 2010 at 13:05


    Thanks for the detail. I think we are discussing why NGDP expectations fell. You have argued they fell because the Fed let them from mid-September on because the Fed paid rather than charged interest on reserves. It is just as possible that he banks wanted to hold reserves as precautionary cash balances, and even suffer a penalty to do so, and if that penalty was too high, then they could have passed it on to their lending clients in the form of higher rates — something hardly conducive to an increase in NGDP expectations.

    Look closely at how bank managers make decisions. A charge of 3% of interest on ER’s — roughly 10% of banking assets — would have cost them 30bp. Assuming losses of 5% on each marginal loan, that would cost them more — 50bp. But wait: if they merely re-priced their loan book to pass on the 30bp interest charge, their profitability wouldn’t suffer at all, and they could just maintain the ER balances. A quick repricing of loans of even 1% is not unheard of: the marginal credit card rate, for example, has gone from about 9% to as high as 29% over the past year. We also have abundant evidence that small and medium sized businesses are paying higher interest rates than in 2007. So why do you think banks would not have passed on the cost of interest charges on reserves?

    So if the Fed charged 10% on ER’s in September 16, 2008, the markets may have concluded that 1) bank profits would suffer, hindering their recapitalization; or 2) bank loan rates would rise, negating the impact of “easier” money and hampering NGDP;or 3) the banks were anxious to lend out the money despite the rising loss experience. Which of these expectations was most likely in late 2008?

    I’m not saying the Fed can’t affect expectations. I’m saying it would have taken time and credibility, and the Fed had neither during the period of October, 2008 – January 2009. I’m also saying there was no way for the Fed to fine-tune expectations during that time, and if it tried to force NGDP forecasts up at all cost, it might have seriously damaged confidence in the dollar — confidence that was already eroding in the summer of 2008 with oil prices at $147.

  15. Gravatar of TVHE » Need more behavioural relationships please TVHE » Need more behavioural relationships please
    18. January 2010 at 16:44

    […] Sumner seems to state that the Fed needs to print money with a nominal GDP target in mind, Woolsey suggests that the Fed should change inflationary pressure given a set real GDP target, and […]

  16. Gravatar of Tom Hickey Tom Hickey
    18. January 2010 at 21:02

    David, I think you make a good point here when you say, “if it tried to force NGDP forecasts up at all cost, it might have seriously damaged confidence in the dollar.” Some governments aren’t happy with the US jawboning about a strong dollar while the dollar falls, putting them at a trade disadvantage. Some feel it is not appropriate for the issuer of the world’s reserve currency to be undertaking a beggar-thy-neighbor policy to strengthen its own economy when it should be working cooperatively with its international partners. Certainly, Chairman Bernanke is well aware of this and knows that there are international limitations in a flexible rate regime.

  17. Gravatar of Bill Woolsey Bill Woolsey
    19. January 2010 at 03:51


    It is much more likely that banks would “pass on” the cost having to pay interest on excess reserves by charging depositors for storing money.

    Any one bank that raises its interest rates on loans to pass on this “cost,” would lose loan business to the other banks and end up with more reserves rather than less.

    Any one bank that lowers the interest rate it pays (to the point of charging for storage of money) would lose deposits to other banks. It would also lose reserves and so the interest cost.

    If you think about the interest on reserves rising and falling, it is clear that your approach is faulty. The Fed pays more and more on reserves, and you would suggest the the banks use this source of funds to subsidize commercial loans. More likely would be a result where they pay more on deposits and shift their asset portfolio away from commercial loans to reserves. This should result in higher rates on commercial loans because of the decrease in supply.

    The reverse, lower interest on reserves should have the opposite effect. Lower rates on deposits and a shift from reserves to commercial lending. This increase in the supply of loans results in lower rates on commercial loans.

    If this process continues until reserves have negative rates, then the process continues–even lower rates on deposits and lower rates on commercial loans.

    The process stops when the interest rates on deposits get more negative than the cost of storing paper money, and the result is a currency drain.

    An analogy would be a the impact of a tax on gasoline on the price of diesel fuel. The result of the tax on gasoline is a lower price for crude oil, a higher price of gasoline, and a lower price of diesel fuel (assumed to be untaxed.) If the tax on gas gets so high that the firms lose money on each gallon, they don’t start raising deisel prices to cover the cost of selling gas. They quit selling gasoline and maximize the production and sale of diesel.

    This is just basic supply and demand economics.

  18. Gravatar of Bill Woolsey Bill Woolsey
    19. January 2010 at 04:00


    If the Fed wants to control the exchange rate of the dollar, then there is no doubt that it cannot target inflation or nominal expenditure.

    I favor targeting a growth path of nominal expenditure. So the value of the dollar must be ignored.

    One of the processes by which an expansion of the quantity of money expands nominal expenditure is a lower exchange rate. It results in higher prices of exports and so an increase in demand for domestic import competing goods. It results in lower foreign currency prices of exports, and so increased demand for output in that manner.

    The impact on the trade balance (deficit and net capital inflow) is ambiguous. However, this isn’t a devaluation aimed at gaining foreign exchange reserves or else stopping a loss of foreign exchange reserves. It is just one o the effects of nominal expenditure targeting.

    A “run” on the dollar is stopped because if the dollar falls too much, then nominal expenditure will rise above target. Money creation must be slowed or reversed. That slows and reverses the drop in the value of the dollar.

    Beggar your neighbor? The point isn’t to obtain foreign exchange from them. It isn’t a gold standard. If they want, they can expand their monetary policies as well. Under current conditions, they should.

  19. Gravatar of David Pearson David Pearson
    19. January 2010 at 11:06


    The banks’ decision to hold precautionary balances is independent of their deposit and loan pricing. Loan pricing has to do with oligopolistic competition for lending and the price elasticity of borrowers. Deposit pricing has to do with oligopolistic competition and the price elasticity of depositors. Where is the rate on excess reserves in that set of calculations? By saying lenders will not pass on the of ER’s, you are making an assumption that either perfect competition exists and/or borrowers are price elastic. I don’t believe either is the case, otherwise how would banks have exhibited such enourmous pricing power over credit card, auto, and small business loans.

    Woosely and Summer, are you really claiming that the banking system is holding 10% of its assets as “cash” at the Fed in order to earn 25bps when lending spreads are the HIGHEST they have ever been? When spreads (net of fixed costs) are 350bp or more, it is highly likely that banks would be indifferent to the existence of a 25bp rate earned on foregone loans.

  20. Gravatar of Tom Hickey Tom Hickey
    19. January 2010 at 13:42

    Woosley: “Beggar your neighbor? The point isn’t to obtain foreign exchange from them. It isn’t a gold standard. If they want, they can expand their monetary policies as well. Under current conditions, they should.”

    Agreed. The US is obviously not interested in foreign reserves under the present monetary regime. But the US is interested in devaluing for two principal reasons: 1) inflating, especially asset prices (equities rose in lock step with the fall in the dollar), 2) boosting exports. Boosting exports is clearly secondary to the monetary objective, but it is a consequence that does have international implications. Moreover, letting the dollar fall also disadvantages foreigners (especially Asians) holding US debt, and there’s a bit of squealing from that side, too. Of course, the US is also limited in dropping the dollar by the increasing cost of petroleum imports, which is politically unpopular at the pump.

    Regarding exports, some (Germany in particular) are complaining that the weak dollar is hurting their exports, and they see it as a backdoor tariff that violates the spirit of cooperation. I completely agree that ECB policy, for example, is too tight, especially in light of what is happening on the EMU periphery. The center states that dominate the EMU are not concerned about failure to cooperate with the periphery when it is is in their interest. The policy the ECB is enforcing on the peripheral states is unconscionable socially, and it has even led to unrest in some countries (Latvia). It is straining the EMU and the EU, too, since leaving the EMU is tantamount to quitting the EU.

    My point is that the US doesn’t not have an entirely free hand monetarily because of international relations and its leadership position. The US at least has to give the impression that this working with its international partners and not punking them. There are already shots being fired across the bow, e.g., calls for an end of the dollar as the reserve currency. These are likely more political statements rather than a serious call for global monetary reform.

    But in the end the US will do what it sees best domestically, just as Nixon stiffed the world when he unilaterally shut the gold window. Not only is it politically necessary for the ruling party, but also it is their obligation to put national interests first.

  21. Gravatar of Tom Hickey Tom Hickey
    19. January 2010 at 14:05

    As I understand it, banks loan against bank capital, not bank assets (reserves). Loans create deposit, and deposits, reserves. The Fed always stands by to make reserves available as needed at the going rate.

    Banks are now capital constrained, and lending standards have tightened. All the liquidity in the world is not going to force lending when there are few qualified buyers and conditions are such that not many who are qualified want to borrow much, even at historically low rates. Banks are now focusing on rebuilding capital instead of returning to the loose lending practices of the recent past. Moreover, the public is now delevering and saving, and business is not expanding investment in the face of weak demand for goods.

    Interest paid on reserves, especially at the low rate prevailing, and the correspondingly low FFR, aren’t a material factor affecting lending. Interest rate cost is typically figured in to the cost of a loan along with other factors, and spreads are pretty wide, as David observes. I have good credit and don’t carry balances, and yet, Citi recently informed me that it was increasing my card rate by 2%. I don’t think that’s because of the interest on reserves. Banks are tightening up after a credit binge.

    The key factors are whether banks have the necessary capital to expand lending and whether there is demand for loans at the prevailing rate. Since capital requirements have been suspended to prevent zombification, it would seem that the reason for low levels of lending and correspondingly high excess reserves from Fed ops is a scarcity of (qualified) demand for loans. The big banks can make more on leveraged prop trading, and that is what they are doing.

  22. Gravatar of azmyth azmyth
    19. January 2010 at 14:22

    Pearson: If the Fed charged a penalty rate on reserves, a firm that got rid of their excess reserves would have an advantage over a firm that did not. The low excess reserve firm would be able to offer better terms to their loan customers and still remain profitable. Even if banking were not a competitive industry, firms are still profit maximizing. I don’t think that shareholders would accept lower profit just to maintain an enormous precautionary balance. Finally, firms can also buy government/highly rated corporate bonds, they don’t need to hold reserves.

    “…are you really claiming that the banking system is holding 10% of its assets as “cash” at the Fed in order to earn 25bps when lending spreads are the HIGHEST they have ever been?”

    As implausible as it seems, excess reserves are still skyrocketing:

  23. Gravatar of scott sumner scott sumner
    19. January 2010 at 18:56

    David Pearson, You said;

    “Thanks for the detail. I think we are discussing why NGDP expectations fell. You have argued they fell because the Fed let them from mid-September on because the Fed paid rather than charged interest on reserves. It is just as possible that he banks wanted to hold reserves as precautionary cash balances, and even suffer a penalty to do so, and if that penalty was too high, then they could have passed it on to their lending clients in the form of higher rates “” something hardly conducive to an increase in NGDP expectations.”

    This seems to confuse money and credit. Less demand for credit is deflationary. Less demand for money is inflationary.

    You also forget that banks have a zero rate alternative to ERs, they could hold T-bills, so I am confident that a 3% interest penalty would have sharply reduced the demand for ERs, indeed I think virtually all monetary economists would agree with me on that narrow point.

    The Fed didn’t need to raise NGDP expectations, just prevent them from falling, which is far easier.

    Tom Hickey, Bernanke is a student of the Great Depression. I can’t believe he would have allowed a major recession in order to preserve a “strong dollar.”

    Bill, I agree with both of your comments.

    David#2, Spreads are misleading when interest rates are expected to rise. They can earn a higher rate on ERs than T-bills, that is what matters.

    Tom, The odd thing about your argument is that it is exactly the opposite of the argument made against China. I don’t even agree with those who say the yuan is undervalued, but at least China has a huge trade surplus. How can a weak dollar be hurting German exports when they have a massive surplus and we have a massive deficit?

    Tom, Again, the low level of lending has nothing to do with the high level of ERs. Banks hold ERs because they are more profitable than T-bills. If you made them less profitable, banks would hold T-bills not ERs. Last summer I had lots of posts on the negative interest on reserves idea.

  24. Gravatar of scott sumner scott sumner
    19. January 2010 at 19:00

    rob, The British could have stayed in the EMS, they simply didn’t want to. And the markets knew that (or at least Soros did) The exact same thing happened in 1931, almost on the same day.

  25. Gravatar of StatsGuy StatsGuy
    20. January 2010 at 07:44


    “business is not expanding investment in the face of weak demand for goods…”

    I’ve made this point in the context of the savings/investment gap being the real problem, not excess savings. Ssumner’s reply is that there is a level of expected monetary stimulus such that this gap must close (at some point, banks would rather buy _something_ than hold onto money that is expected to lose value). He is of course right.

    So there are really two questions left:

    Are we really not consuming enough? (really?)

    If we are not consuming enough, then presumably the economy should be investing even at the current level of consumption. But if so, why is this not happening, and WHERE should we be investing? And I do not mean saving or paying down debt. I mean where should the economy be _investing_ (aside from outside the country)? The primary argument I’ve made in this context is that some of the best investment opportunities are currently public investments. Infrastructure (not just physical infrastructure) and system upgrades.

    Instead, most of the federal outlays have not been investment related, but rather consumption preserving. Investment expenditures have been thin gruel compared to the sheer scale of the consumption/transfer programs.

  26. Gravatar of StatsGuy StatsGuy
    20. January 2010 at 07:44

    If we _are_ consuming enough, then…

Leave a Reply