Monday, September 24, 2012

Callahan on Sowell on Say

He (Gene) writes:

"I am re-reading Sowell's wonderful book, Say's Law: An Historical Analysis. If you want to read a fascinating history and understand the theoretical issues at play more clearly, do pick it up. In any case, a will, as usual, comment occasionally as I read, starting here.

The first thing of note is, come on, this is Thomas Sowell. He is a fairly market friendly, "right-wing" economist, and no "born again Keynesian," as I was recently accused of being. So if he says that Say came to agree with Malthus, well, perhaps he is wrong, but it's not because he wants to throw the decision to the opponents of Say's Law; he is just being honest.

And when he writes, in his very first sentence, "The idea that supply creates its own demand -- Says' Law..." he is not trying to back Keynes, he is letting us know that is a pretty darned good way to summarize Say's Law
."

I can't speak exactly for Keynes (I'd have to review a few things before I tried to), but another thing to remember regarding Gene's last paragraph is that most Keynesians don't think Say's Law stated that way is all that crazy in most circumstances. Supply is calibrated to expectations about what the market will bear, as well as cost and profit requirements. As everyone knows, those wage earners and rentiers that get payments from the entrepreneur (costs, from the entrepreneur's perspective) and the profits of the entrepreneur himself are earmarked either for reinvestment, a cash cushion, or consumption. Demand is implicit in supply in "normal times" in exactly the sense that Say was discussing. There really has to be something exogenous to this whole logic to upset Say's Law. It's not that Say's Law is nonsense - it's that it doesn't consider a few other things.

I also think one of the biggest problems with discussions of Say's Law is this confusion between "general gluts" and recessions. They're related, but not the same. The former is a fundamentally Walrasian concern, while the latter is a fundamentally Keynesian concern.

It's the exact same confusion that comes up when people mix up excess supply of labor and unemployment. They're not the same thing.

14 comments:

  1. "I can't speak exactly for Keynes (I'd have to review a few things before I tried to), but another thing to remember regarding Gene's last paragraph is that most Keynesians don't think Say's Law stated that way is all that crazy in most circumstances. Supply is calibrated to expectations about what the market will bear, as well as cost and profit requirements. As everyone knows, those wage earners and rentiers that get payments from the entrepreneur (costs, from the entrepreneur's perspective) and the profits of the entrepreneur himself are earmarked either for reinvestment, a cash cushion, or consumption. Demand is implicit in supply in "normal times" in exactly the sense that Say was discussing. There really has to be something exogenous to this whole logic to upset Say's Law. It's not that Say's Law is nonsense - it's that it doesn't consider a few other things."

    In other words, you're saying that Keynesians would argue that "Say's Law" is a special case. I've given links to Dr. Michael Emmett Brady's reviews of books on Say's Law before, and I believe that he would agree with you on this point.

    However, I think you have neglected to specify Keynes's mathematical formulation demonstrating the special case of Say's Law:

    MPC (Marginal Propensity to Consume) + MPI (Marginal Propensity to Invest) = 1

    Keynes also has an elasticity analysis in Book V (check Chapters 20 and 21 of The General Theory) also spelling this out.

    ReplyDelete
    Replies
    1. Special case, or maybe first approximation is a better way to put it. Certainly not a guide to business cycles.

      Right, that's practically a definition, right? Every marginal dollar is spent on something. Whether you call it a "special case" or a first approximation depends on how you think actual MPI and MPC relate to 1. Do they settle around it or is it truly random?

      And as I alluded to in this post - all of this is still very different from saying everyone will be employed when MPI+MPC=1. They may not be. This is one thing that frustrates me about Keynes - he has a tendency to treat unemployment as being the same as excess labor supply too.

      Delete
    2. And if you want to view the case when savings exceed intended investment as just an unintended investment in inventory, you could see Say's Law as even holding in that case.

      Delete
    3. Daniel Kuehn: I don't believe that Keynes treats unemployment as excess labor supply necessarily. There is room for a "natural rate of unemployment" in Keynes's General Theory. Also, as a side note, Thomas Sowell's book is cited in this article by Rogerio Arthmar and Michael Emmett Brady that was published in 2011 in the International Journal of Applied Economics and Econometrics. It's a good article summarizing the classical position and contrasting it with that of Keynes's position.

      http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1790083

      Delete
    4. Blue Aurora:

      "However, I think you have neglected to specify Keynes's mathematical formulation demonstrating the special case of Say's Law:

      MPC (Marginal Propensity to Consume) + MPI (Marginal Propensity to Invest) = 1"

      Aren't these variables *average* propensities, not marginal propensities?

      "I don't believe that Keynes treats unemployment as excess labor supply necessarily. There is room for a "natural rate of unemployment" in Keynes's General Theory."

      As I understand him Daniel is concerned about the act of presenting labour at the market. If a recession occurs then many potential workers may stop looking and remove themselves from the short-run supply curve for labour. I agree with him, this kind of thing will lengthen recessions, though who know how much of an effect it will have in practice.

      Daniel:

      "Right, that's practically a definition, right? Every marginal dollar is spent on something. Whether you call it a "special case" or a first approximation depends on how you think actual MPI and MPC relate to 1"

      That's not right, at least not within any given unit of time. Firstly, a marginal dollar may be saved. I save all my marginal dollars at present. Secondly, it may be spent on something that isn't a component of GDP.

      Blue Aurora is saying the same thing as you, AFAICT.

      Delete
    5. I'm not sure there is room for a natural rate of unemployment in Keynes. What about this quote?

      "The Conservative belief that there is some law of nature …that it is financially ‘sound’ to maintain a tenth of the population in idleness for an indefinite period, is crazily improbable – the sort of thing which no man could believe who had not had his brain fuddled with nonsense for years and years…"

      Delete
    6. Keynes doesn't define "full employment" as the situation where everyone is employed, if he did his ideas would be much simpler. In fact he defines it in a few ways which aren't necessarily consistent. (William Darity wrote about this calling it part of Keynes' greatness, in much the same way that Marxists think that Marx's labeling of all his own contradictions as "contradictions of capitalism" was part of his greatness).

      For example, he defines it like this in Chapter 2 of GT:
      "If this were true, competition between entrepreneurs would always lead to an expansion of employment up to the point at which the supply of output as a whole ceases to be elastic, i.e. where a further increase in the value of the effective demand will no longer be accompanied by any increase in output. Evidently this amounts to the same thing as full employment. In the previous chapter we have given a definition of full employment in terms of the behaviour of labour. An alternative, though equivalent, criterion is that at which we have now arrived, namely a situation in which aggregate employment is inelastic in response to an increase in the effective demand for its output."

      This kind of thing happened in Ireland during the boom, NGDP and price inflation increased but employment stayed the same. Keynes is saying that that unemployment is frictional or structural.

      Delete
  2. Supply creates its own demand: Pet rocks.

    ReplyDelete
  3. What people should remember here is that to most people who discuss the subject Say's Law isn't necessarily about the very short run. It isn't about the length of time where money holdings and money demand are highly variable, and capital is essentially fixed. And, our definitions are not the same as those of others. But, different people talk about different things here, that's why there are so many formulations of Say's law, as Gene mentions. From what I've read it looks like Say himself did mean it in the sense criticized here.

    For example, to later economists and some Austrian economists "commodities" or "good" constitutes everything for sale, not just output. For example, in his discussion of Say's law J.S.Mill writes: "Besides, money is a commodity; and if all commodities are supposed to be doubled in quantity, we must suppose money to be doubled too, and then prices would no more fall than values would."

    It's a complicated question, this is my opinion.... In general short-run forms of Say's law are wrong not just because of demand for money but also because of demand for goods that aren't components of output. Of-course if you define Say's law to be about all assets then only money is a problem, Gene posted recently a quote showing that this was Yeager's view. Longer run forms are right with the following caveats:
    * Changes in demand for leisure, as Malthus noted.
    * Capital misallocation (Ricardo & J.S.Mill mention this).
    * The issue Daniel mentions about people taking themselves out of the labour force. I'm not sure if this comes under "path dependency".

    * Also, the following may cause long-run changes in the demand for money...
    ** Changes in the law. (Banking regulations can cause changes in the demand for outside money, for example).
    ** Exogeneous uncertainty. (Probably governments).
    ** Redistribution of wealth changing the demand for money in the long run. (Not necessarily tax, could also be caused by real-nominal difference in debt relationships.)

    ReplyDelete
  4. "I also think one of the biggest problems with discussions of Say's Law is this confusion between 'general gluts' and recessions. They're related, but not the same. The former is a fundamentally Walrasian concern, while the latter is a fundamentally Keynesian concern."

    A recession is a generally accepted term for a period where GDP declines, you don't have to be a Keynesian to talk about them. Calling "general gluts" a "Walrasian concern" when they were invented by pre-marginalist economists is pretty weird too.

    ReplyDelete
    Replies
    1. Why is that weird at all?

      The point is pretty simple, Current. It's a Walrasian concern insofar as most rebuttals of Say's law are statements of when a general disequilibrium can occur. When we talk about recessions the concern is a departure from potential output or full employment.

      If you confuse the two, you start defining away possibilities like equilibrium unemployment.

      I don't understand what the concern is about applying relatively clear descriptives to older discussions. Would you feel better if I just said "proto-Walrasian" and "proto-Keynesian"?

      That just doesn't sound as good to me.

      Delete
    2. One of the problems with the Say's law debate is that we're always concerned with what people say about it recently and what historical economists thought. That makes many things difficult, see my longer post above for example.

      I agree with the essential point that you make, but I think it's useful to talk about "proto-Walrasians" since the historic economists we're discussing here, Say & the Mills, for example, weren't actual Walrasians. I haven't read them in enough detail to know if each of their renditions of the law of markets is proto-Walrasian, but I doubt you have either :). Also, there are more than a few people who take Say's law to be a more long-run idea, the idea that *eventually* supply creates it's own demand. That's not a short-run GE idea.

      Delete
    3. OK, but I didn't say they were Walrasians Current! I wouldn't even argue they were proto-Walrasians. They were dealing with Walrasian problems in this particular case - on this particular issue - and there are other Keynesian problems that have nothing (directly) to do with whether or not we are in general equilibrium.

      Delete
    4. OK, you're not saying those folks are Walrasians, that's fair enough. But you still say that Say's law is necessarily a Walrasian idea, that's what I'm not so sure about. I agree that many renditions of the problem look like GE ideas, but not all do.

      Delete

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