Why do smart people say crazy things?

[I added some graphs to the previous post on income and government spending.  Many readers assumed I was trying to explain why some countries are richer than others.  Not so.  I did that elsewhere.  Also, I recently did a NGDP targeting post for FT Alphaville, in case anyone is interested.]

Some commenters criticized me for putting the term ‘lunatic’ in the title of a recent post.  I suppose they are right.  But first let me explain.  Suppose back during the Zimbabwe hyperinflation someone had complained about the excessively contractionary nature of Zimbabwean monetary policy, and insisted that a more expansionary policy was needed.  Would anyone object to me calling them a “lunatic.” Zimbabwe has had the most inflation of any country during recent decades.

The country with the least inflation is Japan, a country still in the throes of deflation.  So if it would be crazy to call Zimbabwean policy too contractionary during their hyperinflation, why isn’t it crazy to call Japanese policy too expansionary?

But of course there must be more to it that that.  All of the people cited in the article are brilliant; there isn’t a lunatic among them.  So how could they hold seemingly loony opinions?

It might be based on something I learned from reading Milton Friedman back in 1972, when I was in high school.  When you change the growth rate of the money supply, something kind of strange happens.  The equilibrium price level jumps discontinuously.  Consider Graph A, where the money supply growth rate increases.  This will lead to higher inflation, and higher inflation expectations.  The higher inflation expectations will lead to lower real money demand, which causes a discontinuous jump in the equilibrium price level.  So prices rise faster than the money supply, when the money supply growth rate increases.  (Or velocity rises, if you prefer that terminology.)  And this is true even if money is 100% neutral in terms of once-and-for-all changes in the money supply.  There is a ton of data (Cagan, etc) supporting this graph.  Of course in the real world the price level doesn’t jump discontinuously to the new equilibrium, as some prices are sticky.  But the basic idea is true.

Conversely a slowdown in the money supply growth rate causes a discontinuous drop in the price level, followed by a slower rate of inflation (Graph B.)

But central banks don’t like to do policies that would cause a discontinuous drop in the price level.  They are destabilizing.  So more likely if you were to slow the rate of money growth, you’d have a once-and-for all jump in the money supply to accommodate the extra real money demand at the lower inflation rate.  This is shown in Graph C. And this is pretty much what has happened in America, Japan, and a number of other countries.  Notice that during the transition period toward slower money growth, the money supply actually grows much faster than normal.

What does all this mean?  Suppose you live in a world with three kinds of countries.  One group has high and persistent inflation (say India, Vietnam, Argentina, Iran, etc.)  Another group has normal inflation and positive interest rates.  A country like Australia.   And a third group is transitioning to a new steady state with lower money supply growth, lower inflation, lower NGDP growth, and near-zero interest rates.  Let’s say this group includes the US, the Eurozone and Japan.

In that sort of world a graph with inflation on the horizontal axis and money growth on the vertical axis will be U-shaped.  The countries with near zero interest rates will see the demand for base money soar, perhaps from 5% of GDP to 20%. During that transition the average rate of money growth will be quite rapid, despite the low inflation.   The very high inflation countries will have relatively high money growth, for the normal quantity theory reasons.  And the countries in the middle (such as Australia) will often have lower rates of money growth than either extreme.

So the relationship between money growth and inflation is not always monotonic, at least when you are transitioning from one money growth rate to another.  This might be what confuses people.  The countries that seem to have the more expansionary monetary policy (in terms of money growth) will actually include those at both extremes, those that really do have ultra-easy money, and also those with ultra-tight money.

These graphs involve some tricky interaction between changes in growth rates and changes in levels.  But still, I learned this stuff in high school and I’m not that smart.  Famous policymakers should be able to understand this stuff.

Even more confusing, the relationship between interest rates and monetary policy can also demonstrate a U-shaped pattern.  Take a country that currently has low but positive interest rates, like Australia.  You can make a good case that the RBA could increase interest rates sharply with either a much easier policy or a much tighter policy.  The tighter one is what most people are familiar with—the so-called liquidity effect of less money raising short term rates.  But there is also at least one monetary policy, so outrageously hyperinflationary, that it would cause lenders to immediately demand higher rates, even on fairly short term loans.  So the U-shaped pattern also shows up with interest rates.

No wonder everyone’s so confused!  No wonder smart people say things that, when you actually stop to think about it, are completely loony. Like complaining about an easy money policy from the most contractionary fiat money central bank in all of world history.  The Zimbabwe of tight money.  The very special country of Japan.

BTW, although I’ve never really been to Japan (beyond the airport), it’s one of my favorite countries.  If there are any other Japanophiles out there, check out this delightful one hour podcast.  Colin Marshall interviewing Pico Iyer on Japan.  They had me at Colin Marshall interviews Pico Iyer.

HT:  Thanks to Konstantin Mikhailov for the graphs.

 


Tags:

 
 
 

29 Responses to “Why do smart people say crazy things?”

  1. Gravatar of Bill Woolsey Bill Woolsey
    26. January 2013 at 09:47

    The post on alphaville was good.

    Like usual, the comments were scary.

  2. Gravatar of Geoff Geoff
    26. January 2013 at 10:11

    Dr. Sumner:

    I really enjoyed the explanations and the charts. Great visual way of understanding the discontinuous jumps in prices brought about by changes in velocity due to an inflation or deflation.

    Question: For Graph C, is it an accurate assumption that a permanently higher level of money supply, brought about by a one time large increase, after which the same growth rate resumes, will have the effect on prices that you have drawn? I would think that maybe this chart is accurate in the short run, but in the long run, the permanently higher money supply would have a gradually increasing effect on prices, which would make the P line eventually slope up and then stabilize, like an “S” shape.

  3. Gravatar of ssumner ssumner
    26. January 2013 at 10:19

    Geoff, Prices would respond as in Graph A, except that the new slope would be the same as the original slope, since the M growth rates were the same.

  4. Gravatar of ssumner ssumner
    26. January 2013 at 10:19

    Bill, Yes, but that’s par for the course.

  5. Gravatar of Geoff Geoff
    26. January 2013 at 10:22

    Dr. Sumner:

    “Some commenters criticized me for putting the term ‘lunatic’ in the title of a recent post. I suppose they are right. But first let me explain. Suppose back during the Zimbabwe hyperinflation someone had complained about the excessively contractionary nature of Zimbabwean monetary policy, and insisted that a more expansionary policy was needed. Would anyone object to me calling them a “lunatic.” Zimbabwe has had the most inflation of any country during recent decades.”

    Which would be the reason for saying “lunatic”: That Zimbabwe had relatively high inflation compared to other countries, and someone said the central bank should loosen? Or that Zimbabwe had analytically high inflation compared to some supra-country standard, and someone said the central bank should loosen?

    I imagine that if every country except Zimbabwe had inflation in the million range percent, while Zimbabwe had inflation in the billion range percent, then you and most people would likely say that all countries are highly inflationary. So I assume you are arguing inflation should be measured against an analytic standard, not a relative standard.

    Given that, then the question I have is: “What analytic standard for inflation is being used?” For me, I know that this judgment requires information, and the only source for that information is, IMO, a competitive market. Unfortunately we don’t have a competitive market in money matters, so there is no observable information that can allow us to judge whether the production of X = money in country Y is “too high” or “too low.” Again, save for perhaps Zimbabwe.

    But I suppose we can make guesses, and for every country in the world that isn’t Zimbabwe. But still, the inflation rates that exist cannot, IMO, be conclusively judged as “too high” or “too low”, because it is possible that a competitive market in money MAY have generated the outcomes close to the prevailing outcomes. Thus no country, save probably Zimbabwe, can be judged conclusively as having rates of inflation that warrants people in disputes to call each other “lunatics.” A person who says Japan should loosen isn’t a lunatic, and someone who says Japan should tighten isn’t a lunatic either.

    It’s one thing for someone to say inflation in Zimbabwe is too high or too low. It’s quite another to say that inflation in Japan or the US is too high or too low.

    Nobody who argues inflation is too high, or too low, in Japan or the US, is a lunatic. Each country has inflation rates that are likely in the range of possibility that a competitive market in money would have generated.

    It is also possible that both Japan and the US, and most countries of the world, have too high inflation, on the basis that a competitive market in money would have had less inflation. I can’t say for sure, but it’s possible. You have to accept that possibility instead of always insisting that anything less than 5% level growth is “too tight.” How in the world can you know that a competitive market would always and forever generate 5% NGDP level growth? You don’t know enough to make such a conclusive judgment. I’ll give you that you are very bright and intelligent, but you’re not so intelligent so as to know exactly what a market oriented world would actually look like. Nobody is that intelligent. Nobody should claim to know such things. It is pure mysticism appearing as profound insight, like the priests of yore.

    The point is we can’t make conclusive judgments (save for perhaps Zimbabwe) on these issues, because we cannot observe the (market) information that is required in order to make informed judgments.

    Next time you hear someone say “Money is too loose in Japan/US/Australia/etc”, try to imagine how they could be correct, by considering the possibility that a competitive market in money might have generated LESS money and spending in the territory those countries encompass. I know that I don’t think you’re a lunatic for saying money is too tight in Japan in the US, even though I know you are wrong to be so certain, and that I could call you a “lunatic” if I was as fixated on my own ideal of X% growth of Y inflation metric, as you are with NGDP.

  6. Gravatar of Geoff Geoff
    26. January 2013 at 10:51

    Dr. Sumner:

    “Geoff, Prices would respond as in Graph A, except that the new slope would be the same as the original slope, since the M growth rates were the same.”

    OK, but then doesn’t that assume that there was a permanent rise in the demand for money?

    I can see how the same growth rate in M would lead to the same growth rate in P in the long run, but if there was a prior large jump in M (after which there is the same growth rate in M going forward), then wouldn’t that require that initial large jump in M to be accompanied by a corresponding rise in the demand for money, which never gets attenuated in P over the long run, thus generating the same growth slope in P as compared to the other chart?

  7. Gravatar of When you change the growth rate of the money supply, something kind of strange happens. The equilibrium price level jumps discontinuously. « Economics Info When you change the growth rate of the money supply, something kind of strange happens. The equilibrium price level jumps discontinuously. « Economics Info
    26. January 2013 at 13:02

    [...] Source [...]

  8. Gravatar of Steve Steve
    26. January 2013 at 13:10

    “Why do smart people say crazy things?”

    To get promoted!

    From staff economist, to regional president. From junior squid to senior vampire squid. From Bundesbank to UBS Chairman.

    Go back and read the Milgram experiments. Replace “electroshocks” with “hard money”. Then remind the subject delivering the shocks that the recipient was profligate and has structural deficiencies that deserve more intense shocks.

  9. Gravatar of Don Geddis Don Geddis
    26. January 2013 at 13:12

    Geoff: you are assuming that “the right” amount of money, is the amount that a (hypothetical) “competitive market in money” would have come up with. And then you complain that Sumner couldn’t possibly know this value.

    But nobody here accepts your definition of the optimal amount. Sumner has explained this many times. Money growth that is too little, causes the economy to suffer from depressed output and high unemployment (due to sticky wages and nominal debts). Money growth that is too much, causes deadweight losses in taxes and planning for the future. The societally optimal money growth is at the bottom of this U-shaped curve, where the negative effects from too little or two much are balanced and minimized.

    You don’t have to guess what some non-existent market would have produced. Just look at the real economy, and observe the effects.

  10. Gravatar of Bill Woolsey Bill Woolsey
    26. January 2013 at 13:47

    Don:

    I don’t favor minizing a supposed “u shaped”.. what, exactly?

    I think a monetary order that requires everyone to adjust their prices and wages due to a shift in the demand to hold money is worse than one that adjusts the nominal quantity of money to meet the demand to hold it. Making everyone nominal contract a speculation on the demand to hold money, as well as various relative prices, is undesirable.

    This suggests that a stable price level is best. However, a stable price level requires that shifts in the supply of some particular goods be offset by changes in all the other prices in the economy. A monetary order that has this requirement is undesirable. If there is a bad wheat harvest, it is best to just let the price of wheat rise rather than require everyone else in the economy to slightly reduce product prices as well as wages. It is not desirable to make every nominal contract a speculation on shifts in suppply for any particular good. For example, a contract between an entrepreneur in ice cream market and a supply or debt investor does involve speculation about ice cream, but why should it also involve a speculation about the wheat market?

    A monetary regime that keeps spending on output growing at a slow stable rate implies shifts in the quantity of money to accomodate changes in the demand for money while avoiding shifts in the quantity of money sufficient to force offsetting changes in the supplies of particular goods.

    For example, a gold standard doesn’t cause all other prices to fall if the price of wheat rises due to a poor harvest. But, any change in the demand for gold (or gold discoveries) requires everyone in the economy to change their prices.

    A price level rule adjusts the quantity of money to accomodate changes in the demand to hold money but requires changes in the quantity of money to offset shifts in supply of particular goods.

    A nominal GDP rule adjusts the quantity of money to accomodate changes in the demand to hold money but does not requie changes in the quantity of money to offset shifts in the supplies of particular goods.

    The way I see it, it provides a macroeconomic environment that is best for the people using the money.

    If, instead, you start trying to exploit money illusion, saying that people don’t want to have lower wages, but “we” can exploit their money illusion and lower their real wage and keep them working, you are not providing a monetary regime that they find useful. “We” are manipulationg “them.”

    With a stable nominal GDP growth path, then people in people in declining industries get laid off faster (or more of them end up laid off, I guess.) And then, they go to growing industries and get started on their new careers sooner. They may start at even lower nominal wages. Or maybe they start at higher nominal wages. It depends.

    Why is this bad?

    Forcing nominal wages down across the board so prices can fall to bring the real quantity of money up to the demand to hold it, or forcing nominal wages down so that a decrease in supply of some good and increase in its price is made consistent with a stable price level, is unnecessary and bad. If there is extra unemployment for these reasons, it is a crime. In the case of the increase in money demand, there is no need for any shift in labor at all. With the decrease in supply for a particular good, the decrease in wages everyone else is massive overkill for an at most ambiguous requirement for shift in the allocation of labor.

    But when there is a need to reallocate labor, when it is sensible for people to shift from declining industries to growing industries, why is desirable to try to confuse people and keep them in the declining industries longer?

    I also think creating trend inflation creates an expectation of “cost of living” increases, and the minute they are applied to higher prices due to aggregate supply shifts, the result is very bad. And why shouldn’t workers who are told that they are getting a cost of living raise to keep up with inflation not expect to receive a nominal wage increase to compensate them for higher gasoline prices due to threats of war in the Persian gulf?

    The lesson people should have is that inflation occurs when there are adverse supply shocks, and there is really nothing their employer can do about it. Supply is less and so their wages don’t go as far. (And when supply increases, the lower prices of other products allows their wages to go further.)

    What is the supposed benefit of inflation all the time? It is easier for employers to tell their workers that there is no cost of living increase for them than it would be to say that they take a wage cut? The employers will keep workers on with no cost of living increases but they will lay them off before they cut their wages? This is it? So some workers in a declining industry won’t have to get new jobs quite as fast?

  11. Gravatar of Geoff Geoff
    26. January 2013 at 14:21

    Don Geddis:

    I think you misunderstood what I said. I am not “complaining”, or chastising Dr. Sumner for not knowing “the right” quantity of money as would exist in a competitive market, as if I think he has an obligation to go out and learn it in a way I allegedly have.

    No, I am saying that he can’t know it, the same I cannot know it, the same way you cannot know it, the same way everyone else cannot know it, given that we don’t have a competitive market in money.

    I am saying that nobody can know “the right” quantity of money without recourse to the information that only a competitive market can reveal.

    The reason I am bringing this up is because I am seeing Dr. Sumner, and many others, argue things like country X’s monetary policy as conclusively being “too tight” or “too loose”, as if they really do have access to the requisite market based information.

    You say “nobody here accepts your definition of the optimal amount”. You say that as if I am not the one bringing up the very fact that you guys are arguing over definitions, not the reality of people’s desires, behavior, and productivity, constrained to market rules.

    Are you saying “nobody agrees with your definition” as some sort of intimidation tactic, that because I am numerically outnumbered, I have to change my mind? If so, I find that surprising, since Dr. Sumner is outnumbered and yet that doesn’t stop him (good!), and yet I don’t see you making a point to him that many people disagree with his definition, as some sort of refutation of his ideas. Are you an intellectual or not? I don’t appreciate being told I am in some sort of minority as if that is sufficient to showing I am wrong.

    I really don’t see the significance of telling me the statistics of various definitions people are using for money looseness or tightness.

    My “definition” isn’t merely a definition, BTW. It is an explanation of what I argue to be the only source of information that can enable anyone, you, me, Dr. Sumner, etc, to have an INFORMED judgment of the extent of money creation in terms of quantity, which definitions are then made going forward.

    Definitions of real world things refer to reality, do they not? I am arguing not over definitions, but about the underlying reality. I am trying to convince you and others that the reality of which you need to know how much money there should be, doesn’t exist, because we don’t have a competitive market in reality.

    You say that money growth that is too excessive leads to deadweight losses in taxes and future planning, and that money growth that is too little leads to depressed output and unemployment, and so there exists an optimal rate of money growth. Fair enough, as far as that abstract argument goes, I am on board.

    But, then you say:

    “You don’t have to guess what some non-existent market would have produced. Just look at the real economy, and observe the effects.”

    This is a non-sequitur. Knowing that optimal rates of money growth exist, does not mean that you can KNOW what any of them are, by observing past non-competitive market data! The optimal rates of money growth are not a constant of nature, like Planck’s constant or Avogadro’s constant. They are not discoverable by studying the past, especially non-market pasts!

    Optimal rates of money growth are subject to the prevailing knowledge and preferences of people, which can of course change. You don’t know of what unborn people want, let alone what currently living people want, so you can’t possibly know what you claim to know.

    “Looking at the real economy” only tells you about PAST outcomes, and as mentioned, of a non-competitive market in money to boot. It doesn’t tell you what the right growths of the quantity of money are going forward trillions of years into the future as if 5% NGDP is optimal for everyone for all time.

    Dr. Sumner once said to me that NGDP targeting isn’t optimal for all countries for all time. But unfortunately he didn’t elaborate, and so there is still a hole in his version of NGDP theory.

    At best, all you can surmise from past data is that the particular rates of growth of money were either too high or too low at the time, given the prevailing knowledge and preferences of people at the time. Those are unique experiences. They can’t be used to generate a constant rate of growth judgment going forward indefinitely into the future. Not only are humans not robots, not only are you different from me and we are different from past and future generations of people, but your own knowledge and preferences change within your lifetime. What’s good or bad for you can’t be assumed as being good and bad for billions of others, trillions of years into the future. That’s just preposterous.

    Like I said, and I am actually still hoping for a solid rebuttal to this, and not just a “I don’t define things that way” kind of a response: The information that you need, that I need, and that everyone else needs, in order to have INFORMED judgments as to “the right” quantity of X for the present and the future of currently living people, can ONLY come out of an actual real world competitive market in X.

    I am hoping that you aren’t so arrogant that you would claim to know the right growth rate in the quantity of typewriters, or rollerskates, or every other traded good, for everyone else, forever more, based on naively extrapolating from past data. Why do it for money? Is it because it’s “important” and “important” things are revealed to you differently from everything else? Is it because money is a homogeneous concept that is often treated as an abstract number, which leads you to believe that because it’s only a single number (growth rate), that you can know what “the right” growth rate is by studying past relationships between that number and things like output and employment? That would be bordering on cult-like quantophrenia.

  12. Gravatar of Don Geddis Don Geddis
    26. January 2013 at 16:20

    @Bill Woolsey: I’m not sure I understand all of what you said, but most of it doesn’t seem in conflict with what I said. I am curious, though: why do you think essentially every (“inflation-fighting!”) western central bank has adopted something like a 2% inflation target, rather than 0%? Or 1/2%?

    @Geoff: I’m just letting you know that most of your comments are non sequiturs, because you’re assuming the “correct” quantity of money comes only from a “real world competitive market in money”, but the rest of us don’t agree to that assumption. Everything you say after that is ignored, because there is no agreement about basic axioms. Also, Sumner has many times explained why 5% NGDPLT is a good policy for the modern US, but why other policies might be slightly better in other situations. Which is why you’re being silly to talk about fixed growth for “trillions of years in the future”. Nobody is making that claim, so you look foolish when you try to criticize such an obvious strawman. I’m sure Sumner would be perfectly happy to have the Fed reconsider whether 5% is the right number for the NGDP target, perhaps a few times every century.

  13. Gravatar of Bill Woolsey Bill Woolsey
    26. January 2013 at 16:59

    Don:

    Voters stop complaining about inflation when it gets down to 2%.

    If other central banks are doing 2% inflation, then you won’t have your currency depreciate all the time if you have higher inflation.

    What happens if you get get in the position of Japan with deflation and short term money market rates near zero?

    Better to leave inflation at 2% than bring it down to zero.

    I disagree.

  14. Gravatar of Benjamin Cole Benjamin Cole
    26. January 2013 at 18:18

    Scott Sumner is wrong to call the central bankers of Japan and other Western nations “lunatics.”

    Very wrong.

    He should identify the central bankers as “sadistic lunatics.”

  15. Gravatar of dtoh dtoh
    26. January 2013 at 20:00

    Scott,
    I don’t think you have the mechanism right. At low interest and inflation rates, economic participants don’t adjust prices based on changes in these rates. (My view may be anecdotal, but I’ve seen enough different kinds of economic players up close to know that this just does enter into anyone’s thinking.)

    What does drive inflation is short term imbalances in expected supply and demand. Increasing the rate of growth of the money supply will push up real financial asset prices resulting in an exchange into real goods and services producing a marginal increase in the rate of AD growth and an imbalance which impacts the price level.

    Because this increase in spending has a multiplier effect (just like government spending) you get an increase in AD disproportionate to the initial money injection causing the jump in the price level.

    I understand how in theory an increase in real interest rates (including an increase in inflation) will reduce money holdings, but in the real world the absolute amounts are so small and cash balances are so well managed that there is at best a miniscule effect. It’s the impact on marginal demand through the financial asset price mechanism which causes the change in price levels.

  16. Gravatar of Lorenzo from Oz Lorenzo from Oz
    26. January 2013 at 20:51

    Ben Cole wins the internetz for today :)

  17. Gravatar of Benjamin Cole Benjamin Cole
    27. January 2013 at 05:33

    Thanks Lorenzo.

  18. Gravatar of ssumner ssumner
    27. January 2013 at 06:31

    Geoff, No, there is no change in the real demand for money if trend inflation doesn’t change.

    dtoh, You said;

    “I understand how in theory an increase in real interest rates (including an increase in inflation) will reduce money holdings, but in the real world the absolute amounts are so small and cash balances are so well managed that there is at best a miniscule effect.”

    Actually there is a mountain of empirical studies that show just the opposite—changes in nominal interest rates have a big impact on the demand for money. Also note that there is $3 trillion in base money in the US economy–I wouldn’t call that trivial.

    In any case, I think you are describing a disequilibrium situation, whereas in this post I am looking at equilibrium changes.

  19. Gravatar of Bill Woolsey Bill Woolsey
    27. January 2013 at 07:02

    Scott:

    I agree with all of your analysis, but like all of the monetarist work, it is too much based upon a framing of given growth rates of the quantity of money.

    We have one equilibrium growth rate, and then we switch to a new equilibrium growth rate.

    Monetary regimes that target the growth rate of the quantity of money are close to nonexistent.

    I will grant that this analysis does tell us something about inflation targeting regimes or hypothetical nominal GDP targeting regimes.

    In particular, I think that if the regime changes to a higher or slower inflation rate or nominal GDP trend growth rate, the shifts in the demand for money would exist. (Though the discountinous shifts in the price level do not.)

    But I also think care must be taken in other contexts in applying monetarist reasoning. A regime where a growth path of nominal GDP is set, and the quantity of money adjusts as needed, is not the same as one where the quantity of money is set, and nominal GDP adjusts so that the demand to hold money meets the given supply.

  20. Gravatar of Geoff Geoff
    27. January 2013 at 13:34

    Don Geddis:

    “@Geoff: I’m just letting you know that most of your comments are non sequiturs, because you’re assuming the “correct” quantity of money comes only from a “real world competitive market in money”, but the rest of us don’t agree to that assumption.”

    If you disagree that the only source for the requisite information to make informed judgments for “the right” quantity of money FOR the market IS the market, then you are introducing a mystical element into your thinking. It would be like you telling me that you know the right quantity of clothes or computers for me and others, without engaging us and without communicating with us as to our preferences.

    It isn’t an “axiom” that the only source of information to make informed judgments is the market. That is a complex conclusion, based, among other things, on the same economic reasoning that you use for things like computers and flu medicine.

    It is not sufficient for you to simply declare that you “don’t agree”. You have to show why you are using inconsistent reasoning, and how you can know the right quantity of money for millions of others going forward, without even engaging with them in market activity and communication.

    You aren’t making a strong case by basing your entire thought process on nothing but “I just don’t agree with your assumption.” I am the one saying that of course disagreement is all you can go by, because you don’t have any rational grounds to show who is right and who is wrong.

    “Also, Sumner has many times explained why 5% NGDPLT is a good policy for the modern US, but why other policies might be slightly better in other situations. Which is why you’re being silly to talk about fixed growth for “trillions of years in the future”.”

    You misunderstand. It doesn’t matter if the specific rate of growth is constant for trillions of years into the future. It’s that you want there to always be SOME constant growth rate, set by non-competitive market mechanisms, for trillions of years into the future.

    You even made this case yourself, when you said:

    “I’m sure Sumner would be perfectly happy to have the Fed reconsider whether 5% is the right number for the NGDP target, perhaps a few times every century.”

    Oh wow, 5% NGDPLT, with a few resets a few times per century. And you’re really saying that I was off base for saying “Constant NGDP growth trillions of years into the future?” You’re just restating what I am saying.

    I am saying that a judgment of constant NGDP growth, even those subject to numerical adjustments, is not going to be an informed judgment.

    If NGDPLT is going to be a good theory, then it can’t be something that is applied in only one situation at one time, and no others. The theory’s worth derives from the extent of its universality. The theory is being proposed because massively declining NGDP is NEVER a good thing, provided of course that some assumptions such as price stickiness and so on are present.

    Nobody is setting up any straw man. Yes, there are reasons for why 5% NGDPLT would “work” in the US. These reasons are by definition universal, in that should the same assumptions be present, then 5% NGDPLT would ALWAYS follow as optimal.

    Pinning your hopes on future circumstances before you “commit” is precisely what I argue shows that you need the very market information that you say is not even needed in the first place!

    If you don’t need market information, then you should be able to tell me the optimal rates for each and every circumstance right here and right now. If you can’t, then you are really just showing the validity of what I am arguing.

    Finally, maybe you can leave the “foolish” charges at the door. I like to keep things professional.

  21. Gravatar of Geoff Geoff
    27. January 2013 at 13:45

    Dr. Sumner:

    “Geoff, No, there is no change in the real demand for money if trend inflation doesn’t change.”

    What does “real” demand for money mean? Does that mean the demand for money that is intended to be utilized for making real goods purchases, as opposed to being hoarded for its own sake? If so, then I understand all demand for money to be real demand, because we all seek dollars to make short and long term purchases. Well, for paper money that is the case, but for money that is also a consumer good, that wouldn’t apply, and the distinction might be warranted.

    Anyway, thanks for the explanation.

    OK, if the demand for money doesn’t change if trend inflation doesn’t change, then:

    If there was an initial, significant, one time rise in the money supply to offset the rise in demand for money that accompanies a lower trend inflation going forward, then this rise in the demand for money would not be reduced going forward because trend inflation isn’t changing going forward?

    Sorry for not getting this quicker.

  22. Gravatar of dtoh dtoh
    27. January 2013 at 14:44

    Scott,
    You said,
    “Actually there is a mountain of empirical studies that show just the opposite—changes in nominal interest rates have a big impact on the demand for money. Also note that there is $3 trillion in base money in the US economy–I wouldn’t call that trivial.”

    I’m not that familiar with the data. I did see the long term you provided and commented on that a couple of days ago. Let me make just a few comments.

    1) At higher rates of inflation and interest, I would not dispute that a change in rates can cause a change in demand for base money

    2) ZLB is special case, because money becomes dual purpose and people start holding it both as medium of exchange and as a store of value.

    3) The long term data on NGDP/base versus interest rate you provided for last 30 years show a correlation but I’m suspicious of the causality. Changes in ATM availability and fees, growth of the grey economy, etc., might be responsible for the changes in base holding over the last 30 years.

    4) For me though, the bottom line is that it would be irrational for economic players to change their holdings of base money because of small shifts in interest or inflation rates. The absolute amounts of the opportunity costs or savings are just too small (and are subsumed by the higher/lower transaction costs of holding less/more money) to make it worthwhile for an economic player to adjust their base money holdings.

  23. Gravatar of Bill Woolsey Bill Woolsey
    28. January 2013 at 01:08

    Geoff:

    The real demand for money is the amount of purchasing power people want to hold.

    It is represented by the nominal demand for money (how much dollar’s worth of money they want to hold) divided by the price level.

    It is the command over goods and services people want to hold.

    If prices were to stay constant, and people decided to hold 10% more money, then that is a 10% increae in the real demand for money. That 10% more money they want to hold will buy 10% more goods and services.

    If prices rise 5% and people choose to hold 5% more money, then that is no change in the real demand for money. They do want to hold more money, but the total they want to hold buys the same amount of goods and services.

    Real vs. nominal means the same thing.

    Real wages are wages in terms of what they buy. Nominal wages is an amount of money.

    Real GDP is the real volume of goods and services produced. Nominal GDP is the dollar value of everything produced.

    Real demand for money is the amount of money people want to hold in terms of what it will buy. Nominal Demand for money is the amount of dollars they want to hold.

  24. Gravatar of Geoff Geoff
    28. January 2013 at 05:01

    Bill Woolsey:

    “The real demand for money is the amount of purchasing power people want to hold.”

    For how long?

    “It is represented by the nominal demand for money (how much dollar’s worth of money they want to hold) divided by the price level.”

    Who actually buys the goods of which the price level consists?

    “It is the command over goods and services people want to hold.”

    For how long?

    “If prices were to stay constant, and people decided to hold 10% more money, then that is a 10% increae in the real demand for money. That 10% more money they want to hold will buy 10% more goods and services.”

    If they hold 10% more, they aren’t spending 10% more. Spending 10% more must require their income to rise 10%, or their cash to decrease 10% of their income.

    “If prices rise 5% and people choose to hold 5% more money, then that is no change in the real demand for money. They do want to hold more money, but the total they want to hold buys the same amount of goods and services.”

    Not everyone can hold 5% more money, unless there is an increase in the money supply.

  25. Gravatar of Nick Nick
    28. January 2013 at 06:43

    What are the implications of these strange u-shaped relationships on the Fed’s (eventual) exit from QE3? And especially the misunderstanding of many people about the meaning of changes in interest rates?

    What I mean is this: At some point, the Fed will have an appropriately loose monetary policy. At that point, rates will begin to rise. However, everybody thinks that the Fed is enacting a loose monetary policy by *keeping rates low* so that folks can buy a house with a cheap mortgage etc. So what does the Fed do then? Do they expand QE further, to try to keep rates down (which is what everybody thinks QE should do)? If they do that, won’t rates increase further as we go up the “wrong” side of the u-curve?

    Or, assuming the Fed “gets it,” is the beginning of rising rates the signal that they end QE? But if they do that, won’t that tightening keep rates at the really low levels they are at? Despite the fact that the Fed would actually be doing an appropriate tightening, I would imagine strong political pressure for the Fed to not just end QE, but also to raise rates and end “financial repression.” At that point, if they give in to political pressure and raise rates, it would likely plunge the US into a recession (and the yield curve would invert, as you would expect). If they don’t mess with the Fed Funds rate, rates stay at the zero bound, until…when? Are 10 year risk free rates always going to be around 2%, and this is the best case scenario?

    Without assuming a can opener (NGDP futures market), what is the best thing that the Fed can do over the next year or so (when everybody predicts QE will end) to get back to a normal yield curve, and what do you think they actually will do?

  26. Gravatar of ssumner ssumner
    28. January 2013 at 07:55

    Bill, You said;

    “Monetary regimes that target the growth rate of the quantity of money are close to nonexistent.”

    I agree, but my post does not depend on the assumption that M is being targeted. I was simply presenting stylized facts about shifts in Md as trend inflation changes.

    dtoh, There are literally 1000s of studies of money demand (and no I’m not exaggerating.) Money demand does depend on interest rates–I believe the elasticity is around 1/2, but am not sure. At a 10% interest rate the opportunity cost of holding a trillion in cash is $100 billion. That’s serious money. We’ve seen currency demand rise rapidly in the last couple of years—probably due to low interest rates.

    Nick, I hope they target NGDP expectations when exiting. But the danger you point to is real. Most central banks exit the zero boundary prematurely.

  27. Gravatar of o. nate o. nate
    28. January 2013 at 13:24

    Scott writes:
    “I agree, but my post does not depend on the assumption that M is being targeted. I was simply presenting stylized facts about shifts in Md as trend inflation changes.”

    It may not depend on the assumption that M is being targeted, but doesn’t it depend on the Fed providing guidance on the future rate of M growth? Otherwise why would inflation expectations shift today? If the Fed just suddenly expands M but doesn’t announce that the future rate of M growth will be slower, then why should inflation expectations decrease?

  28. Gravatar of Doug M Doug M
    28. January 2013 at 15:19

    The Zimbabwe situation was kicked off by a program of land redistribution and a collapse in agricultural productivity, followed by increasingly heavy-handed tactic by the regeim to maintain control.

    Whetever misteps they may have made on the monitary side of things, their fundamental problems were real.

  29. Gravatar of dtoh dtoh
    28. January 2013 at 23:50

    Scott,
    You said,

    “There are literally 1000s of studies of money demand (and no I’m not exaggerating.) Money demand does depend on interest rates–I believe the elasticity is around 1/2, but am not sure. At a 10% interest rate the opportunity cost of holding a trillion in cash is $100 billion. That’s serious money. We’ve seen currency demand rise rapidly in the last couple of years—probably due to low interest rates.”

    Yes, it’s a $100 billion, but if you assume a big corporation holds a a few thousand in cash and and an individual holds a few hundred: with a 1% change in rates, that’s $30 per annum for a corporation and $3 for an individual. Corporate treasury personnel always work their tail off to minimize cash regardless of rates, and for an individual the ATM fees from more frequent bank visits required by lower cash holdings, wipes outs any savings/costs almost immediately.

    As I said, I’m not disputing that a) there’s an impact at higher rates of for bigger changes, and b) the ZLB is a special case because money becomes dual purpose…not just a medium of exchange but also a store of value (financial asset).

    But when you’re bouncing around with rates below 5%, I think you would be hard pressed to find any meaningful impact.

Leave a Reply