Wednesday, May 4, 2011

More Nick Rowe on Monetary Disequilibrium

Still getting my head around it, but there are a couple points he does a good job detailing things that I've asked him about in the comment section here.

Nick says he's heavily influenced by Patinkin - I was leafing through my copy of Money, Interest, and Prices this morning and I decided I'm going to have to read this soon. This is one of several classic texts I was able to pick up for free when they shut down our library at the Urban Institute and let us scavange the shelves.



  1. You picked up classic texts FOR FREE?


  2. It was heaven... but sad for the librarians, who I knew pretty well. One got to stay to manage electronic databases, etc. but two others were let go. Urban tries to avoid layoffs, but some are inevitable.

  3. Four "Handbooks of Economics" - Labor, two Public Economics, and Econometrics. Usually those cost a lot. Hardback copy of Friedman and Schwartz. Several NBER volumes. Hicks's Theory of Wages and Revision of Demand Theory (I already had Value and Capital), lots of Joan Robinson, Buchanan's "Cost and Choice", a couple Phelps books... a good day. And one of my favorites is actually an old history of the Reconstruction Finance Corporation by one of its directors in the thirties.

  4. Patinkin's book is great. I haven't read it cover to cover, and I don't agree with much that I have read, but it's obvious that Patinkin knows what he's talking about. It seems as if Money, Interest, and Prices is the alternative to Mises's The Theory of Money and Credit (which Patinkin criticizes here and there in his book).

  5. An excess demand for oranges could cause a general glut, but not all by its lonesome -- an excess demand for money is also necessary.

    How can an excess demand for oranges could "spill over" into an excess demand for money and, more importantly, can it be stopped?

    Normally we don't get a recession because of an excess demand for oranges, because price adjustments are quick and smooth. But even with sticky prices (or even rigid price controls), an excess demand for oranges does not "spill over" into money. People who are unable to purchase oranges might just buy apples instead -- demand "spills over" into apples. The consequence for the allocation of resources is very much as though people had preferred apples in the first place, i.e. as though no excess demand oranges existed.

    An excess demand for oranges does not usually cause a recession because money is not a close substitute for oranges. If people cannot get oranges, chances are they'll get the next best thing that satisfies the same end, i.e. something like apples.

    The situation with bonds (specifically government bonds) is almost identical, except that sometimes money can be a very close substitute for bonds. Sometimes the next best thing to satisfy the same end is money. When that occurs, the demand for money rises as demand "spill over" from elsewhere. Unless the supply of money increases to offset rising demand, an excess demand for money will emerge and aggregate nominal income will fall.

    Hypothetically, an excess demand for oranges could cause a recession if it spilled over into money, but in practice no such thing occurs because there are countless other goods and services that people readily purchase if their efforts to get oranges are frustrated. The situation with bonds, especially government bonds with near zero interest, is different only because there are fewer goods and services between money and bonds to absorb any spill over. An excess demand for bonds, by itself, does not cause recession unless it results in an excess demand for money.

  6. Money Interest and Prices is a great book. But also deeply flawed. It tries to do monetary economics in a Walrasian economy. So money is "inessential". If all the stock of money was destroyed, the rest of the economy could carry on as if nothing had happened.

    Let M represent the stocky of bling. We measure prices in bling. We get utility from wearing bling, but it's the real value of that bling, M/P that counts, because we want to show off our wealth. The demand for bling is a positive function of income, Y, and a negative function of the opportunity cost, r, of holding part of our wealth in the form of bling. So there's a demand for bling function: Md/P = L(Y,r).

    The Quantity Theory and Neutrality of bling. If you double the nominal stock of bling, M, the value of each unit of bling will halve, so the real stock of bling M/P and all other real variables will stay the same.

    Everything Patinkin says works fine for bling.

  7. I agree with Lee Kelly in that comment above.

  8. I generally agree with the points that Lee has made, but as usual would qualify his criticism on the bonds point and add that other things could be going on as well.

    First I think we're getting too caught up on explaining the cause of a genuine general glut vs. the cause of a recession. Excess demand for bonds may not always spill over to money, but it may still cause a general enough excess supply elsewhere to hurt a lot of workers because demand does not sufficiently spill over anywhere. That's not a "general glut" but it may sure feel like a recession to workers and look like one in the data. Are we a science of general gluts only or a science of economic activity?

    Second, as I said on the earlier post, if interest rates are determined by both a liquidity preference rule and a loanable funds rule there's no guarantee that both will be satisfied and you could get deadweight loss in the loanable funds market. Output and prices could adjust to bring that into equilibrium. I still couldn't figure out whether that was the same point you were making or not, Nick. Initially I thought it wasn't but then I thought it might be essentially the same point.

    Third, I still think we need to be careful about how much emphasis we place on "excess demand" and "excess supply". Let's we're not in a Walrasian world, but we are in a world where there's no excess demand for money and all markets are clearing. If we equate "labor surplus" with "unemployment", then there would be no unemployment here. But that's not the data that we collect and name "unemployment". "Unemployment" isn't "labor surplus" - "unemployment" is "every point to the north east of the labor market equilibrium" - at least that's how we define it when we collect the data and think about it in every day life. I explain that idea more here:

    In other words, let's assume all you've presented and assume no excess demand for money. You can still have unemployment with all those markets nice and tidy and clearing.

  9. Daniel Said: "Excess demand for bonds may not always spill over to money, but it may still cause a general enough excess supply elsewhere to hurt a lot of workers because demand does not sufficiently spill over anywhere."

    No. Nuh uh. Not possible. Demand always spills over somewhere. People buy bonds with money. If they cannot buy bonds, then they'll either buy something else or hold money. Those are the only two choices. If they buy something else, then we just get a reallocation of spending and resources. Sure there might be some structural unemployment as profits and losses get shifted around the economy, but not a general glut or recession. On the other hand, if they hold money, then we'd better hope prices are really really flexible.

  10. Forgive me using the third person to refer to you Lee – I originally planned this as a post and then decided to make it a comment, and I don’t feel like changing it all!


    Let’s say going into 2008 people had plans expectations to buy and sell 100 units of oranges and 100 units of nice, safe, government bonds. Then, the housing market crashed, which brought down credit markets and sent everyone running to buy safe government bonds. Now we have a demand for 150 units of government bonds and (because oranges seem less important now) 50 units of organges. Of course, this means that orange producers can’t sell 50 of the units they had made and the bond market is short of 50 units of bonds that are being demanded (and a Republican Congress doesn’t want to produce more).

    Now, if the bond demanders are so fickle that they immediately respond by saying “you know – if you don’t feel like printing those extra bonds, we’ll just buy the 50 units of oranges we were originally planning on buying before the crash after all” then clearly there’s no harm done. But clearly they’re not that fickle and there is real concern behind their increased demand for bonds, and so that demand for bonds will persist.

    Lee wants to call this a money demand problem, and for good reason. He knows they’re not going to jump right back in the market for oranges, after all. He knows they can’t get bonds. So they hold (i.e. “demand”) money. If prices don’t adjust immediately, this causes a general glut just like we’ve been talking about across the blogs for the last several days. That increased demand for money (because people aren’t uber-fickle) is going to make money more dear in all markets – because money is the one “good” that is traded in all markets. Lee is right that excess demand for bonds can cause a general glut through its impact on the excess demand for money in this way.

    I’m not denying that.

    What I’m saying is that while money is becoming more dear in all markets and all other goods are becoming less dear as a result, you still have 50 units of undemanded oranges sitting around from way back up in my first paragraph!

    Yes, Lee is right – if you can’t satisfy demand at the very least you’re going to hold money and that causes a general glut as a second order effect. My point is that the initial relative imbalance (in this case between oranges and bonds) that sets that monetary process in motion can still be substantial and hit a lot of workers, and it’s not due to the monetary dynamics that Nick and Lee have been talking about. I’m not saying this to disagree with Nick and Lee on their monetary point - I’m saying this to resurrect the idea that these relative gluts can still have very significant effects. General gluts are crucially important and too many people actually do accept the straw man version of Say’s Law and inappropriately reject the prospect of a general glut. But explaining a general glut is not co-terminous with explaining a recession (although it does go a long way).

    Add on top of this my point about dead-weight loss in the loanable funds market because of the odd nature of the interest rate and my point that labor markets can clear without there being full employment and I think you start to get a much better picture of the nature of recessions: and this picture is incomplete if you exclusively cite a classical monetary disequilibrium general glut.

  11. And I should note - this is why I was saying the other day that I don't think DeLong's point that the excess supply of bonds was all that bad.

    First, the relative imbalance itself is not trivial, particularly after a massive shock where there's lots of demand for bonds or safe asssets.

    And second, because I don't think the monetary disequilibrium point is lost on him. Why are Krugman and DeLong so disconcerted about this downturn? Precisely because at these interest rates bonds and money are such good substitutes: precisely becuase of the point that Lee makes that we should start being concerned if the excess demand is satisfied by spill-over into money.

  12. Daniel: on your 3 points:

    1. I agree with Lee.

    2. Here's the standard ISLM story: Liquidity preference says r adjusts to equilibrate the supply and demand for money LM. Loanable funds says that r adjusts to equilibrate savings and investment IS. We can reconcile the two conflicting theories (for a fixed P) when we recognise that both S and Md depend on Y, so Y adjusts until both are true. (or, for fixed Y and flexible P, Md depends on P, so we reconcile them that way.) I disagree with that ISLM reconciliation. I don't think it makes sense.

    3. Take e.g, a minimum wage, efficiency wage, monopoly union theory of the natural rate of unemployment. Because real wages are stuck too high, there is an excess supply of labour even in equilibrium. Here's what I would say: in the labour market there's an excess supply of labour matched by an excess demand for money *in that market*. All other markets are clearing.

  13. On 3 - that's not what I'm saying. I'm saying in competitive, non-monopoly, reprehensibly idealized market that clears there can still be unemployment.


    Because "the lack of unemployment" is not the same as "competitive market equilibrium".

    When we count unemployment and when we commiserate with friends over their unemployment we do not query them on whether their reservation wage exceeds the equilibrium wage. We ask them two things: (1.) do you want work?, and (2.) are you in the market for work?

    Their wage relative to some equilibrium wage is completely irrelevant to his social concept.

    If we want to use models to talk about how labor surpluses emerge, fine - we can do that.

    But if we want to understand the social phenomenon of "unemployment" we cannot equate it with labor surplus.

    This is precisely why the obsession over "microfoundations" is so dangerous - it forces us into worrying about textbook micro problems rather than trying to explain the real world.

    Take a look at that graphic in the link I provide on point #3 and I think you'll get what I'm trying to say. When I sympathize with an unemployed friend I don't first confirm that his reservation wage exceeds the competitive equilibrium wage.

  14. Daniel: OK. I misunderstood your point 3. Yes, there's sometimes a big step between what we measure or talk about in ordinary language and our theoretical constructs.

  15. One leads toward price rigidity arguments, the other leads toward demand arguments.

    Quite a difference in the implications for the scientific conclusions, and those conclusions in turn have quite a difference in the implications for policy!

  16. Daniel,

    Supposing the excess demand for bonds spills over into money, but the money supply increases to preserve monetary equilibrium, it is not clear that any of your orange producers must become unemployed.

    The increased cash balances of the public are turned into an increase in loanable funds by the Fed and banking system. Perhaps the orange farm will procure a loan to help them get through the temporary fall in sales, or maybe redundant employees will get loans to start new businesses, perhaps the farm will borrow to invest in training for its employees, etc.

    You're right, of course. Preserving monetary equilibrium does not preserve the prior arrangement and allocation of resources, though I think of that more a feature than a bug. You are also right that unemployment is likely to rise, some "relative gluts" are likely, and output will probably slow or fall as resources are reallocated. However, might point above is that I actually expect these adjustment processes to be less tumultuous for an excess supply of bonds than it would be for a consumer good.

  17. To clarify. Supposing monetary equilibrium, an excess demand for bonds might just change the way the orange producers finance their activities. To put it another way, a shift from oranges into televisions is probably more disruptive to the prevailing allocation of resources than from oranges into bonds.

  18. Another thing that I find interesting.

    Yeager points out that just people a good is a close substitute from an individual perspective, it does not follow that each has similar systemic consequences.

    Near zero interest earning government bonds may be a close substitute for money from the perspective of an individual, but an excess demand for bonds does not have the same systemic consequences as an excess demand for money. Money is special in a way that bonds are not.


All anonymous comments will be deleted. Consistent pseudonyms are fine.