Krugman and Sumner on the Zero-Interest Lower Bound: Some History of Thought

I indicated in my first posting on Tuesday that I was going to comment on some recent comparisons between the current anemic recovery and earlier more robust recoveries since World War II.  The comparison that I want to perform involves some simple econometrics, and it is taking longer than anticipated to iron out the little kinks that I keep finding.  So I will have to put off that discussion a while longer.  As a diversion, I will follow up on a point that ScottSumner made in discussing Paul Krugman’s reasoning for having favored fiscal policy over monetary policy to lead us out of the recession.

Scott’s focus is on the factual question whether it is really true, as Krugman and Michael Woodford have claimed, that a monetary authority, like, say, the Bank of Japan, may simply be unable to create the inflation expectations necessary to achieve equilibrium, given the zero-interest-rate lower bound, when the equilibrium real interest rate is less than zero.  Scott counters that a more plausible explanation for the inability of the Bank of Japan to escape from a liquidity trap is that its aversion to inflation is so well-known that it becomes rational for the public to expect that the Bank of Japan would not permit the inflation necessary for equilibrium.

It seems that a lot of people have trouble understanding the idea that there can be conditions in which inflation — or, to be more precise, expected inflation — is necessary for a recovery from a depression.  We have become so used to thinking of inflation as a costly and disruptive aspect of economic life, that the notion that inflation may be an integral element of an economic equilibrium goes very deeply against the grain of our intuition.

The theoretical background of this point actually goes back to A. C. Pigou (another famous Cambridge economist, Alfred Marshall’s successor) who, in his 1936 review of Keynes’s General Theory, referred to what he called Mr. Keynes’s vision of the day of judgment, namely, a situation in which, because of depressed entrepreneurial profit expectations or a high propensity to save, macroequilibrium (the equality of savings and investment) would correspond to a level of income and output below the level consistent with full employment.  The “classical” or “orthodox” remedy to such a situation was to reduce the rate of interest, or, as the British say “Bank Rate” (as in “Magna Carta” with no definite article) at which the Bank of England lends to its customers (mainly banks).  But if entrepreneurs are that pessimistic, or households that determined to save rather than consume, an equilibrium corresponding to a level of income and output consistent with full employment could, in Keynes’s ghastly vision, only come about with a negative interest rate.  Now a zero interest rate in economics is a little bit like the speed of light in physics; all kinds of crazy things start to happen if you posit a negative interest rate and it seems inconsistent with the assumptions of rational behavior to assume that people would lend for a negative interest when they could simply hold the money that is already in their pocket.  That’s why Pigou’s metaphor was so powerful.   There are layers upon layers of interesting personal and historical dynamics lying beneath the surface of Pigou’s review of Keynes, but I won’t pursue that tangent here, tempting though it would be to go in that direction.

The conclusion that Keynes drew from his model is the one that we all were taught in our first course in macro and that Paul Krugman holds close to his heart, the government can come to the rescue by increasing its spending on whatever, thereby increasing aggregate demand, raising income and output up to the level consistent with full employment.  But Pigou, whose own policy recommendations were not much different from those of Keynes, felt that Keynes had left out an important element of the model in his discussion.  As a matter of logic, which to Pigou was as, or more important than, policy, an economy confronting Keynes’s day of judgment would not forever be stuck in “underemployment equilibrium” just because the rate of interest could not fall to the (negative) level required for full employment.  Rather, Pigou insisted, at least in theory, though not necessarily in practice, deflation, resulting from unemployed workers bidding down wages to gain employment, would raise the real value of the money supply (fixed in nominal terms in Keynes’s model) thereby generating a windfall to holders of money, inducing them to increase consumption, raising aggregate demand and eventually restoring full employment.  Discussion of the theoretical validity and policy relevance of what came to be known as the Pigou effect (or, occasionally, as the Pigou-Haberler Effect, or even the Pigou-Haberler-Scitovsky effect) became a really big deal in macroeconomics in the 1940s and 1950s and was still being taught in the 1960s and 1970s.

What seems remarkable to me now about that whole episode is that the analysis simply left out the possibility that the zero-interest-rate lower bound becomes irrelevant if the expected rate of inflation exceeds the putative negative equilibrium real interest rate that would hypothetically generate a macroequilibrium at a level of income and output consistent with full employment.  If only Pigou had corrected the logic of Keynes’s model by positing an expected rate of inflation greater than the negative real interest rate rather than positing a process of deflation to increase the real value of the money stock, how different would the course of history and the development of macreconomics and monetary theory have been.

One economist who did think about the expected rate of inflation as an equilibrating variable in a macroeconomic model was one of my teachers, the late, great Earl Thompson, who introduced the idea of an equilibrium rate of inflation in his remarkable unpublished paper, “A Reformulation of Macreconomic Theory.”  If inflation is an equilibrating variable, then it cannot make sense for monetary authorities to commit themselves to a single unvarying target for the rate of inflation.   Under certain circumstances, macroeconomic equilibrium may be incompatible with a rate of inflation below some minimum level.  Has it occurred to the inflation hawks on the FOMC and their supporters that the minimum rate of inflation consistent with equilibrium is above the 2 percent rate that Fed has now set as its policy goal?

One final point, which I am still trying to work out more coherently, is that it really may not be appropriate to think of the real rate of interest and the expected rate of inflation as being determined independently of each other.  They clearly interact.  As I point out in my paper “The Fisher Effect Under Deflationary Expectations,” increasing the expected rate of inflation when the real rate of interest is very low or negative tends to increase not just the nominal rate, but the real rate as well, by generating the positive feedback effects on income and employment that result when a depressed economy starts to expand.

20 Responses to “Krugman and Sumner on the Zero-Interest Lower Bound: Some History of Thought”


  1. 1 Cantillon Blog July 8, 2011 at 1:27 pm

    David,

    Very interesting post, and I wish I had read this a few years back – you make an important point very clearly.

    Why do you think it is the case that history for students of the business cycle begins with the first data point on a Bloomberg terminal (or if one is taking an extra-long term perspective with the period following World War 2)? I can certainly acknowledge that the structure of the economy changed, permanently in recent years but is it perhaps also the case that we are limiting our study to a period in which only a certain limited range of economic phenomena manifested and by doing so crippling our ability to understand the range of phenomena that might occur in the future? (For example the adoption of the bromide that US home prices have never fallen at a national level would surely not have been quite as smooth had people extended the period under consideration).

    With regards to the relationship between the real interest rate, and the rate of inflation there is one other question that is somewhat distinct theoretically from the issues you discuss, but rather relevant to the situation today. That is how should the central bank react to a burst of productivity growth or to a substantial improvement in the terms of trade, both of which may tend to depress observed inflation, possibly below the levels that might be consistent with an inflation target. The appropriate response might actually be to raise the nominal rate (given the higher prospective return on capital, and the benefit to incomes and wealth – the latter operating via asset prices), whereas an inflation targeting approach would have you cut rates.

    I do believe this is how we got into the mess we are in today (the widespread adoption of inflation targeting at just the wrong time). Compared to the indiscipline of the 70s, any kind of nominal anchor seemed an improvement. But in the end focusing on the price level rather than monetary stability can lead to tremendous problems.

  2. 2 Mikko July 8, 2011 at 1:52 pm

    I have been reading Scott Sumner for a few years now and I’ve been trying to understand the reason why the inflation targeting approach that worked so well for a few decades doesn’t seem to work now.

    Might it be that over the past decade or so, our economy has changed because the developing countries have become big players in the global economy. Before this time period, the price of commodities was basically determined by the existing supply and the changes in demand in developed countries. As all developed countries had approximately the same pace of growth, the price changes in commodities did not swing the boat.

    Now, we have developing countries which grow much faster and hence the the demand for commodities grows much faster than the developed countries grow. This puts pressure on the economy in the developed countries.

    Consider the inflation to come from price changes for external reasons (supply shocks and changes in the demand for commodities in foreign countries) and internal reasons (wage growth, etc.). Now, when exogenous reasons push prices up and central bank targets inflation, it must push down the rate of growth in the country. Otherwise the level of inflation rises from the existing level.

  3. 3 João Marcus Marinho Nunes July 8, 2011 at 2:02 pm

    David
    Very clear an useful post. As Cantillon Blog mentions, I also think inflation targeting was useful to bring inflation down from lofty levels but it has now become a constraint on good MP.
    http://thefaintofheart.wordpress.com/2011/05/01/insisting-on-targeting-the-fed-funds-rate-can-be-bad-for-the-economy%C2%B4s-health/

  4. 4 João Marcus Marinho Nunes July 8, 2011 at 2:10 pm

    Mikko
    Inflation targeting has become “rotten”. That´s due to its “internal workings”, not because of the economic rise of emerging markets.
    http://thefaintofheart.wordpress.com/2011/05/20/from-bullard%C2%B4s-presentation-it-comes-out-that-%E2%80%9Cinflation-targeting-is-rotten%E2%80%9D/

  5. 5 Cantillonblog July 8, 2011 at 4:43 pm

    JMMN – the speech you link to seems to argue rather that targeting core inflation has become rotten. Targeting core never made any sense because core inflation does not predict future inflation better than broader measures. In fact food inflation tends to lead inflation more generally. (Incidentally note that everybody believes inflation in the 1970s was kicked off by OPEC. That’s not true. It all started with soybeans).

    Inflation targeting would have been a great idea for the 1960s and 70s (when they were messing around with trying to push unemployment down towards some imagined low natural rate). But it was adopted broadly just at the worst possible time.

    I do in fact posit that it is a golden rule of policy making that any approach will only be broadly adopted as some kind of panacea when we are reaching the climactic phase of the various cycles coming together that made such an approach appear for a while to be an answer for all time rather than for that particular regime.

    Anyone can be an inflation targeter when the terms of trade are improving, when productivity growth is improving (due to questions of the global transmission of knowledge – embedded and explicit) and when global inflationary mass psychology is turning in a cooling direction. In the 90s and 2000s inflation fell everywhere, whether or not the central bank was particularly skilled.

    But human nature being what it is, instead of pondering the underlying reasons why this might be the case, an orgy of self-congratulation broke out – “what a good job we have done, my dear fellow”.

    The question is what happens from here. Emerging markets first impact was disinflationary as the supply shock was more pronounced. I suspect that from here the demand shock will be of more importance. The future in the west is likely to see steadily deteriorating terms of trade for quite some time. I don’t think this will be terribly much fun. I don’t believe central bankers have the _vir_ (manly courage) to impose sustained disinflationary pressures on the domestic sector in order to offset over time imported inflation. Therefore I posit that it is a matter of time until central banks globally abandon inflation targeting. This will initially be done informally (as we see with the Bank of England) but over time will be recognized and made official.

  6. 6 bill woolsey July 8, 2011 at 9:24 pm

    The inflation as equilibrium makes zero-interest currency too central to macroeconomics.

    Most money takes the form of deposits and can have negative nominal interest rates.

    If nominal interest rates get too low, then issuing zero-interest hand to hand currency becomes unprofitable.

    To me, the obvious answer is to quit issuing it.

  7. 7 anon July 8, 2011 at 9:48 pm

    “Now a zero interest rate in economics is a little bit like the speed of light in physics; all kinds of crazy things start to happen if you posit a negative interest rate … ”

    Viewing the ZLB on nominal interest rates as a fundamental limit akin to the speed of light seems needlessly confusing. Even if we ignore quantitative easing as a means of keeping policy on target, in principle (AFAICT) nothing stops central banks from setting negative interest rates. The only real problem is that the interest rate on currency is fixed at zero, but this could be addressed by taxing currency creation–effectively, a kind of demurrage. One could think of it as the government + central bank standing ready to borrow real resources at a zero nominal rate: obviously this can lead to problems, but the problems do not seem very dissimilar from any other price support policy.

  8. 8 Scott Sumner July 8, 2011 at 10:33 pm

    Mikko, Don’t forget that we experienced deflation in 2009, so the policy stopped working when we stopped adhering to the policy. Having said that, I agree with David that inflation targeting is less than optimal, even if done sincerely.

    David, I’m not just anti-inflation targeting, I’m opposed to the whole concept of inflation. In my view NGDP growth is the “right” metric for deflating nominal wages, nominal interest rates, etc.

    Alan Meltzer once asked why didn’t Keynes simply propose a positive trend rate on inflation, (making most of the General Theory completely superfluous.) It seems to me that the gold standard mindset was so deeply entrenched back then that no one could imagine a positive trend rate of inflation as a policy.

  9. 9 Lars Christensen July 9, 2011 at 3:56 am

    Scott, I generally agree with your view on NGDP, but is there really a difference when we are in a situation where the economy is in more or less monetary equilibrium as during the Great Moderation. Furthermore, there is pretty strong empirical evidence that some central banks, which officially have been inflation targeting have in fact been price level targeting. A former student of mine has shown that for example the Swedish Riksbank and the New Zealand Reserve Bank (I remember correct) have de facto been targeting a path for the price level rather than the inflation rate. To me there is no major difference between price level targeting and NGDP targeting – so maybe Riksbanken has de facto been following your advise and targeted NGDP?

  10. 10 bill woolsey July 9, 2011 at 12:01 pm

    Lars:

    The difference between targeting the growth path of NGDP (level targetting) and targeting a growth path for the price level, is the impact of aggregate supply shocks.

    With a price level target, the central bank must contract to offset the effects of an adverse aggregate supply shock. If gasoline is 10% of the economy, and the supply of gas falls such that gas prices rise 50%, then this raises the price level 5%. The central bank must lower other prices 5% (more or less.) And so, gas ends up about 47.5% higher than before, and the prices of everything else is about 5% lower. The price level is back on target. If the price level is supposed to rise 2% over the year, then other prices must only fall roughly 3%. Of course, that is still 5% below expecations.

    With inflation targeting, on the other hand, adverse aggregate supply shock occurs. There is 5% inflation, more or less. The central bank then just targets the price level to continue growing as usual from there. No need to reverse the inflation. Other prices just rise 2% as usual, and gas goes back to rising 2% from its 50% increased level.

    (Of course, this assumes that we wake up one morning, with the supply shock already having increased the price of oil 50% What if a central bank, observing a series of gas price hikes decides to dampen those hikes so that the inflation rate is only 2%?)

    Price level targeting is a nonstarter unless you assume that using monetary policy to deflate all the other prices 5% from their trend causes no problem–in other words, perfect price flexibility.

    With NGDP targeting on a 3% growth path, the decrease in the supply of gas results in gas prices rising 50%. The price level rises 5% (more or less). And real output falls approximately 2%. If the demand for gasoline is inelastic, this should create an output gap. (real expenditure falls more than potential output.) Depending on the income elasticity of gasoline, and the responsiveness of other prices to this output gap, then the effect on the relative price of gas and the price level will be dampened, and also the impact on real output.

    If prices (including wages) are perfectly flexible, then there will be no output gap. Real output will remain at potential, and the price level will by higher in inverse proportion to the decrease in potential output due to the decreased availability of gasoline.

    Assuming the supply shock is persistent, then with sticky prices, this should also be long run equilibrium. Real output rises back to potential, and the price level falls back towards its long run level, remaining elevated only to the degree potetential output remains below trend.

    During the period in which real demand is depressed for the rest of the economy, basically because people are spending more on gasoline in real terms and less on everything else, resources will be freed from the rest of the economy to shift to the energy sector. To the degree that the supply of gasline is inelastic, then a return to equilibrium will require wages and other money incomes in the rest of the economy to grow below the long term trend, lowering unit costs and prices. Again, the end result is output back to potential, depressed only by the degree made inevitable by the reduced availability of gasoline.

    One major criticism of NGDP targeting is that the preformance above doesn’t necessarily optimize the social welfare function. The counter criticism of that is–who is running the central bank? God?

  11. 11 Scott Sumner July 9, 2011 at 3:40 pm

    Lars, I agree that the difference between NGDP targeting and price level targeting is smaller than the difference between NGDP and inflation targeting. But I still don’t trust price indices. The US core CPI greatly understated disinflation during 2008-09, partly because it showed housing prices rising, even in real terms, during one of the great housing crashes in history. And housing is 39% of the core index.

    Glad to hear Sweden is moving in that direction–I’ve been a fan of their policy since 2009, when I did some posts on Sweden. They’ve done pretty well.

    NZ’s central bank issued an official statement condemning my NGDP targeting idea–which thrilled me. Not because they opposed it, but because they responded to a written proposal of mine. At least central banks are paying attention. I don’t really think NGDP targeting would be right for small countries anyway.

  12. 12 Lars Christensen July 9, 2011 at 3:52 pm

    Scott, just to stress neither Riksbanken nor RBNZ are officially targeting the price level – and as far as I know they are not moving in that direction. However, the research done by my former student seems to indicate that both central banks are de facto targeting the price level rather than inflation. I will be happy to get back to you when his final results are in. He is writing his master thesis on price level targeting at the moment.

  13. 13 Lars Christensen July 9, 2011 at 4:00 pm

    Scott, I by the way disagree. NGDP targeting would work very well for small open economies as long as it is combined with a proper instrument. My idea is to implement a NGDP target, but implementing it through announcing a depreciation/appreciation path for the country’s exchange rate against a basket of currencies. This would be a combination of Sumner-McCallum and Singapore’s actual monetary and exchange rate policy. Obviously one could discussion whether to target domestic final sales instead of NGDP…

    Countries that in my view could benefit from such a set-up could be Iceland and the Baltic countries (Estonia of course has joined the euro…).

  14. 14 David Glasner July 10, 2011 at 1:18 am

    Cantillon, Thanks. I am glad that you found what I wrote helpful. I am more convinced now than I was back when I wrote my book that aiming to stabilize wages is the best target. Doing so allows inflation to cushion supply shocks and allows deflation to spread the benefits of productivity increases throughout the economy. In the near term, however, we need inflation to help lift us out of the whole that we have fallen into. And I do agree with you that a longer historical perspective would be very conducive to a better understanding of what is happening today.

    Mikko, I am not convinced that rising commodity prices are such a terrible problem. Commodity prices today are still 25 percent below where they were three years ago before the crash. There was a lot of scare mongering about commodity price inflation as a way of stirring up opposition to the very modest monetary easing that the Fed undertook last fall. But the rapid increase in prices from their lows after the crash still leaves them well below their pre-crash peaks, so I don’t understand the basis for hand-wringing about commodity prices.

    JMMN, Thanks for the links. I am hoping to do a future post on inflation targeting, and will use Bullard as my foil. But the point will be that the best inflation or level target is wages, and wages are now rising at a rate that is historically very low. In an ideal world we might want an even lower rate of increase in wages, but that can’t be our short-term policy goal now.

    Cantillon (responding to JMMN). I agree that core inflation is a slippery concept. I think that it is designed basically to give the Fed or whoever some extra wiggle room in justifying a policy that would be harder to justify if it didn’t have the alternative inflation measure. I think the real problem with inflation targeting is that changes in inflation can result from either supply or demand side disturbances and depending on whether the origin is supply or demand side the appropriate policy response is not the same for changes. Thus an increase in the price of soybeans reflecting an easing of monetary policy may call for a tightening of policy, but an increase in oil prices owing to a supply disruption or a deliberate withholding of supply may call for monetary expansion to cushion the supply shock. So looking just at prices without being able to identify whether the underlying disturbance is on the supply side or demand side can lead the monetary authority astray as it did disastrously in 2008 and in 1973-74. On the other hand looking at wages doesn’t give you an incorrect signal. A negative supply side shock reduces wages and signal monetary expansion which we want and a positive demand side shock increases wages and signal tightening which we also want. So your prediction need not turn out to be true if we shift to some form of wage-inflation targeting. Of course, Scott Sumner will say that NGDP targeting would give us essentially the same outcome as wage-inflation targeting and I am not ready to say that he is wrong.

    Bill, Zero-interest currency is central because all other forms of money are convertible into it. There may be some way of creating a monetary system without convertibility into some “outside” asset, but no such system exists as far as I know. Taking the existing monetary systems as given, the price level is determined by the stock of non-interest-bearing currency and the demand to hold it and the demand for and supply of all financial claims that are convertible into currency determine their yields relative to the yield on currency. Of course the demand for currency depends on the yields on the financial claims that are convertible into currency, so there are a lot of equations that have to be solved simultaneously. I don’t exactly follow your conclusion about whether it is profitable to issue zero-interest-bearing currency and why it ought to be stopped.

    Anon, I take your point about my analogy between zero interest and the speed of light. But it was only an analogy. We could impose a tax on currency, as Irving Fisher once proposed, as did others, I believe Silvio Gessel, whom Keynes praised in the General Theory, was one. But that proposal doesn’t seem to have caught on.

    Scott, We have discussed this before on your blog, and I am happy to have the chance to welcome you to this blog to continue the discussion. I don’t think that there is any single right or wrong way to talk about this, and you make a persuasive case that NGDP can do everything that the price level or the rate of inflation can. How would you define the value of money i.e., the purchasing power of money? Remember money is a stock and NGDP is a flow, so how do you translate one into the other? Where did Meltzer ask this question about Keynes?

    Bill (responding to Lars) Wouldn’t wage-level targeting produce the optimal adjustment path?

  15. 15 Lars Christensen July 10, 2011 at 7:37 am

    Dear Bill, you are preaching to the converted,-)

    I totally agree that NGDP level targeting is better than price level targeting – whether it is political realistic is another story, but there is no reason that that should stop us from making the case for NGDP targeting.

    However, my point is just that price level targeting and even inflation targeting have it clear advantages compared to what we used to have before in both the US and Western Europe – that is vulgar keynesisan discretionary monetary policies.

    Being non-American, however, I am slightly less concerned about the general GLOBAL deflationary risks and as such there is little practically difference between NGDP, price level and inflation targeting in the near term for countries where there is not a major monetary disequilibrium. In fact in many Emerging Markets we are in the opposite situation of the US. My favourite example is Turkey where money supply growth clearly is outpacing underlining money demand and inflationary pressures as a consequence is clearly on the rise. An NGDP approach would obviously tell you the same story.

    Maybe is what I am looking for is a more global perspective on the discussion of NGDP level targeting. As you (David) have noted it seems like Sweden now have closed the “NGDP gap” and as such monetary tightening clearly warranted in Sweden. Furthermore, I think some focus should be given to what instruments are used to implement NGDP targeting. As I have suggested for (some) small open economies an exchange rate based might be the way ahead.

  16. 16 bill woolsey July 10, 2011 at 7:43 am

    David:

    Define the dollar in terms of a reserve balance at the Fed. Replace Federal Reserve notes with privately-issued banknotes redeemable in those balances on the same terms as bank-issued checkable deposits.

    Bank issued checkable deposits are redeemable for reserve balances at the Fed. In other words, those receiving payment by check or banknote can shift the funds to whatever bank they prefer.

    Money is now defined in terms of an interest bearing balance. Hand-to-hand currency is issued privately when banks find it advantageous to do so. If nominal interest rates become too low for the issue of hand-to-hand currency to be profitable, then it won’t be issued.

    If that happened, the likely result would be either a termporary disruption in sectors of the economy that depend on hand-to-hand currency, or an expansion of noncurrency methods of payment, like more debit cards, or the development of poorly capitalized firms that issue hand to hand currency to fund risky investments.

    Hand to hand currency isn’t the basis of the financial system in the scenario, and it is not a good store of wealth.

    After considering that scenario, then leave off the privatization of hand-to-hand currency. Money is still defined as before, but reserve balances aren’t redeemable in hand-to-hand currency. If the central bank wants, it just quits issuing currency. Checkable deposits still clear using reserve balances. Interest rates on those balances can be positive or negative.

    It isn’t that hard to see. Just imagine that technological change makes hand-to-hand currency disused. And when everyone is buy using their electronic payment media, the currency is removed.

    As for a wage index, presumably on a growth path equal to the trend rate of growth of labor productivity, the difference with nominal income targeting has to do with changes in the share of income between labor and capital. Nominal income would shift to a new growth path when that changes. I think this is an advantage of wage index targeting.

    If wages are really sticky, then an excess demand for money might result in reduced output and employment, while wages are expected to continue on trend. Those who went short on the contracts don’t profit, even though they helped reduce the excess demand for money. Those who went long on the contracts don’t lose, even though they exacerbated the problem. (I actually understand the operations of the system well, but just now–short, long, which is it?)

    With only a bit of flexibility in wages, you get the profits and losses, but they are too low because of the stickyness.

    With money income targeting, an excess demand for money should result in lower prices for flexibly priced goods and reduced output for all sorts of goods. Those who went short on the futures profit and those who went long lose. As they should.

  17. 17 David Glasner July 11, 2011 at 2:18 pm

    Bill, You’ve packed a lot into this comment, so it’s taking me a lot of time and effort to work my through it. You say that private money is defined in terms of an interest-bearing balance at the Fed. It may or may not be interest-bearing, and I think that we need to work our way through both cases. But to do that now would give me a headache, so I will skip that point. About whether private banks issue their own banknotes, it seems obvious to me if there is a demand for banknotes that can be physically exchanged, if only by underground economic transactors, some bank will offer to issue banknotes to its depositors as a way of increasing its market share even if they were not generating any explicit return to the bank. The underlying problem with any fiat monetary system is what accounts for the demand to hold the monetary asset. If there is no demand hold it, it will not serve as money. There is a bootstrap problem here which is rarely addressed. But if you posit that there is a technological change that totally eliminates the demand for hand-to-hand currency so that the only outside asset is the reserve balance at the Fed, I am still not sure if there is any determinate solution to the implied system of equations because you are missing a nominally fixed variable in terms of which to solve the rest of the variables. The model seems undetermined to me.

    I’m not sure I follow your reasoning for the shift in income shares between labor and capital. Could you spell it out for me? If there is an excess demand for money that causes output to fall, presumably there would be downward pressure on nominal wages which would trigger short-sales of the futures to the monetary authority expanding the money supply, eliminating the excess demand and easing downward pressure on wages. Thanks, by the way, for your posting about Uneasymoney.com on your blog. I will try to respond tonight.

  18. 18 bill woolsey July 12, 2011 at 10:30 am

    Thanks for the reply.

    I favor targeting a growth path for GDP (nominal). That is the nominal variable that determines the system. I favor pegging the interest rate paid on central bank reserves to a market rate, and then changing the quantity of reserves using ordinary open market operations in bonds parellel to index futures contracts. But, leaving aside index futures convertibility, just ordinary open market operations based on hitting the target growth path for GDP.

    Private banknotes are nothing like fiat money. They are like checks. There may be a demand for them, but banks won’t supply them unless the nominal interest rate earned on asset portfolios is high enough to make it profitable. At negative nominal interest rates, this is unlikely. No doubt, this will disrupt black market activities. I don’t know what drug dealers would do. I don’t care that much.

    The smooth response to a shift in income shares is a benefit of wage targeting relative to GDP targeting. With GDP targeting, a shift away from labor to capital requires a decrease in the growth path of nominal wages. That is because real wages drop. With wage targeting, nominal wages stay on target, but nominal capital income grows (that is how it increaes in share.) This implies growing nominal income. If real output didn’t change, remaining on its growth path, the higher nominal income and constant real income implies a higher price level. Real wages fall to a lower growth path. Real capital incomes rise less than nominal capital incomes because of the higher price level.

    The disadvantage of wage targeting relative to GDP targeting involves monetary disequilibrium. You are correct about an excess demand for money putting downward pressure on wages. But with wages being sticky, lots of pressure (unemployment) is needed to get much decrease. With GDP targeting, you get the impact on flexible prices and real output too. Of course, there are sticky prices of goods as well, and they won’t respond much.

    If all prices and wages were perfectly flexible, and we assume no impact on output, then wage and GDP targeting would be the same. All that would happen with the excess demand for money is lower prices and wages. Wages and GDP (nominal) would fall in proportion. With stickness, output and employment fall and some prices fall more than others. If wages are relatively flexible, then wages respond more to the excess demand for money than other prices. If, as seems likely, wages are less flexible than most prices, wages respond less than other prices. If all prices are highly inflexible, then wages don’t respond much, and output and employment do. GDP responds to the excess demand for money much more than wages or prices.

    If index futures convertibiltiy always works perfectly, the GDP or wages stay on target all the time. There is no monetary disequilibrium. In reality, it won’t work perfectly, and so wages and GDP will deviate from target. Profits and losses will be generated. Which is better? The one that deviates more when there is monetary disequilibrium, or the one that deviates less?

  19. 19 Current July 13, 2011 at 8:36 pm

    Nice blog, I look forward to reading it more.

    I agree with Bill that demand for money involves demand for money-in-the-broader-sense, that is current account balances and notes. I don’t think analysing it based on zero interest rate hand-to-hand currency makes sense.

  20. 20 David Glasner July 14, 2011 at 11:48 am

    Bill, I am not sure that merely targeting the path of nominal GDP growth is enough to determine nominal values for the system. What ensures that the target is hit? Convertibility of some sort, e.g., index-futures convertibility, seems to be required. I even must admit that I have some doubts about whether index future convertibility is enough, but, having already gone into print saying that it is, I don’t want to go down that particular road, at least not right now. Sorry, but I had trouble following your argument about monetary disequilibrium with wage-index targeting and stick wages. Sometimes I have to read an argument several times before I get it. I also think that the adjective “sticky” in this context is highly misleading and confuses at least two distinct concepts that should be kept very much distinct. Sorry to just throw that out without clarifying, but see my reply to comments by Nick Rowe and Scott Sumner on this post.

    Current, Thanks for your encouragement. I hope that you will continue to like what you read here.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s




About Me

David Glasner
Washington, DC

I am an economist in the Washington DC area. My research and writing has been mostly on monetary economics and policy and the history of economics. In my book Free Banking and Monetary Reform, I argued for a non-Monetarist non-Keynesian approach to monetary policy, based on a theory of a competitive supply of money. Over the years, I have become increasingly impressed by the similarities between my approach and that of R. G. Hawtrey and hope to bring Hawtrey's unduly neglected contributions to the attention of a wider audience.

Enter your email address to follow this blog and receive notifications of new posts by email.

Join 441 other followers

Follow Uneasy Money on WordPress.com

%d bloggers like this: