Rebutted but not Refuted
I stand rebutted.
Note to self: When you are taking on Brad DeLong, do not write "You will not find it in Keynes" when you mean, "You will not find it in the relevant chapters of The General Theory."
Having made that concession, let me now empty it of force and meaning.
In 1925, Keynes had not completed the "struggle of escape" of which he speaks in the introduction of the General Theory, published in 1936. Indeed, there he writes of his own "lack of emancipation from preconceived ideas" only five years before.
Part of that emancipation, which I suspect happened around 1933, lay in discarding the concept of a labor market. This Keynes achieved by rejecting the supply curve of labor. No supply curve, no labor market. If there is no labor market, then the volume of employment has to be determined by something else.
That is what I meant, when writing earlier that the General Theory is not centered on markets. There are markets in Keynes, of course. There are markets for specific products, with which he wasn't much concerned, and there are financial markets, which he understood very well. But Keynes reorganized two critical concepts away from the markets to which classical economists had assigned them.
The first was employment, which he removed from the labor market. The General Theory is, after all, a theory "of Employment."
The second was the interest rate, which he removed from the classical capital market, or the market which supposedly balanced the demand for investment against the supply of savings.
Keynes instead linked the interest rate to the market for money, something quite different. The General Theory is thus also a theory of "...Interest and Money."
[Keynes called the labor supply curve the "second postulate" of the classical doctrine, and he likened it to the axiom of parallels in Euclid's geometry. As I have written long ago, there is a deep affinity between Keynes and Einstein. "Monetary production economics," the topic on which Keynes lectured in 1933, is the economic analog of "space-time."]
The first problem of the labor supply curve is that workers do not negotiate a real wage. They negotiate a money wage. But if the price level can be made to rise, then real wages will fall. So even if money wages are sticky, real wages need not be.
The second problem is that even if money wages are flexible, prices set by markups would mean that falling money wages produced falling prices. Real wages would not change much, and the supposed adjustment mechanism would not work.
Either way, Keynes argued, the classical economists had gotten themselves into a muddle. There are, of course, many ways around this, to which the MIT school has devoted much effort over the years. The point is, that isn't where Keynes went.
With no labor market, the adjustment of wages is irrelevant to the problem of unemployment. Employment becomes a function of the aggregate effective demand for output as a whole. And we can then go on to a discussion of the determinants of that.
The main determinants of total demand are total consumption and total investment (along with government spending and net exports). That is what I meant by writing, in desperate shorthand, that Keynes' theory dealt mainly with the elements of the national income accounts.
There is no "market" for total consumption or total investment. Rather, they are governed by conventional behaviors, and by the state of expectations.
The phrase "sticky-wage Keynesianism" is a good description of Brad's position. It captures the general alignment of policy positions, while preserving the fact that the underlying theory is not Keynes's own.
James Galbraith
Comments (13)
Not exclusively, though.
In a letter to the poet, T.S. Eliot, dated April 5, 1945, Keynes identified shorter hours of work as one of three "ingredients of a cure" for unemployment (1980: 383-84). The other two ingredients were investment and expanded consumption. Keynes regarded investment as "first aid," while he described working less as the "ultimate solution." A more thorough and formal presentation of his view appeared in a note Keynes prepared in May 1943 on "The Long-Term Problem of Full Employment" (320-325). In that note, Keynes projected three phases of post-war economic performance. During the third and final phase, estimated to commence some ten to fifteen years after the end of the war, "It becomes necessary to encourage wise consumption and discourage saving, –and to absorb some part of the unwanted surplus by increased leisure, more holidays (which are a wonderfully good way of getting rid of money) and shorter hours" (323).
Keynes, J. M. (1980) The Collected Writings of John Maynard Keynes. Vol. 27, Activities 1940-1946: Employment and Commodities: Shaping the Post-War World, D. Moggridge (ed.). London: MacMillan; New York: Cambridge University Press.
June 2, 2007 6:42 AM | Reply | Permalink
Being a microeconomist, I can predict that many microeconomists would argue that there are labor markets.
Now, when you delve into actual empirical "labor markets", they tend to be tricky things, and the New "Keynesians" make much of possible implications of these aspects of labor markets.
However, far more importantly, they do not aggregate to a single overall labor market. That does not mean that they are not connected ... as Keynes observed, they are indeed connected, by the fact that the people capable of performing highly skilled, specialized work are also capable of performing low skilled, generalized work. Or, in other words, establish the convention of paying in the neighborhood of $150-$200/week per class, for 32 weeks of the year, to a PhD "instructor", with little more security than experienced by a temporary light industrial worker, and they can meet or beat that working in a warehouse.
Rather, they do not aggregate as a functional equivalent of a standard market because, in the aggregate, the supply side enters into the demand side and the demand side enters into the supply side.
And if we can no longer pretend that this is a situation that can be handled in a general equilibrium framework ... because outside of very special and unlikely conditions, general equilibrium models are liable to numerous equilibria and to having equilibria as ill-behaved as conceivable ... then we require a different equilibrium concept than the market equilibrium concept in order to understand why output and employment in successive months and quarters exhibit pronounced serial auto-correlation.
The neoclassical school went back to Walras and added more sophisticated mathematics to try to make an end-run around the fact that there is no coherent "labor market". But that involved borrowing models from physics of low dimension, and translating them into models of very, very high dimension, with not only every product in production but every product that could conceivably be produced as a distinct dimension of the model, and as was shown in the 1970's, the end result turns out to be a non-viable model.
And with that end-run a dud, we are left with either Keynes' equilibrium of effective demand, or else with the naive and insupportable process of simply pretending that there is a national "labor market".
June 2, 2007 8:45 AM | Reply | Permalink
One of the great frustrations one often experiences when trying to discuss the importance of a particular generalization in economics is that a detractor can be expected to seize upon any exception he can find and hold it up as an example of either your profound ignorance or of your lack of professional caution. I find this kind of criticism especially annoying.
There is not an Econ 101 textbook anywhere in the world that does not celebrate the Law of Demand (market demand and prices are inversely related) and yet this "Law" isn't much of a law, is it? Usually before the end of the chapter, it is acknowledged that there are exceptions to the 'law', but none of the authors of these textbooks find it necessary to assail the use of the world 'law' as a accurate description of the generalization. Surely the reason why economists continue to refer to the Law of Demand because it serves to emphasize the importance of the relationship between prices and demand that we normally see.
If only Brad DeLong could have understood your generalization about Keynes within this context, Dr. Galbraith.
I think critics like DeLong need to be frequently reminded that they have not successfully discredited the arguments of those who have challenged them by simply pointing out not-so-meaningful exceptions to any sweeping statements that have been made. The only thing that should matter to a serious theorist is whether the exceptions are significant. Dr. Galbraith has successfully demonstrated that the exceptions DeLong identified are not.
I sympathize because I'm one of those who likes to use 'sweeping statements' as a way to emphasize the great significance I see in some of the generalizations I am highlighting. For example, I like to say that "Even a steeply progressive income tax---right up to 99% on the highest incomes---would impose no loss of purchasing power on wealthy income earners" even though in an open economy it is possible to point to some rather minor exceptions to the rule. I am, however, just as comfortable in making that statement as I am in mentioning the Law of Demand to students.
I wonder, Dr. Galbraith, if there isn't some way we could perhaps preempt these annoying side discussions about exceptions by making a simple 'reference' that would tell any serious theorist to only bother to mention the exceptions if she is prepared to make clear why she believes they are significant?
June 2, 2007 1:54 PM | Reply | Permalink
"The crucial step was this: labor and land were made into commodities, that is, they were treated as if produced for sale. Of course, they were not actually commodities, since they were either not produced at all (as land) or, if so, not for sale (as labor).
Yet no more thoroughly effective fiction was ever devised. By buying and selling labor and land freely, the mechanism of the market was made to apply to them. There was now supply of labor, and demand for it. Accordingly, there was a market price for the use of labor power, called wages, and a market price for the use of land, called rent. Labor and land were provided with markets of their own, similar to the commodities proper that were produced with their help."
Karl Polanyi
The commodity fiction has a lot to do with economics' blindness to the environment and the family.
I am aware that orthodox economists have made attempts to bring the environment and the family into economics. At best they seem to have brought economics up to the place where everyone else already was without economics' help. As the powerful science it claims to be, economics should have offered superior insights into the economics the family and the environmental economics, but economists are still really playing catchup, and trying to bring economics minimally in contact with reality.
At worst, for example in Becker's writing on the family, economics is promulgating extraordinary looniness. For which looniness Becker received a Nobel Prize.
June 3, 2007 6:19 AM | Reply | Permalink
"Usually before the end of the chapter, it is acknowledged that there are exceptions to the 'law', but none of the authors of these textbooks find it necessary to assail the use of the world 'law' as a accurate description of the generalization."
In principle, it is easy for the law of demand to be violated, all that is needed is an inferior good in which the income effect is greater than the substitution effect. But in actual economies, exceptions to the law of demand are so exceedingly rare that it is difficult to come up with actual real world examples. Therefore, from an empirical standpoint it is, for all practial purposes, a law.
June 3, 2007 9:11 AM | Reply | Permalink
"This Keynes achieved by rejecting the supply curve of labor."
You do not reference the article in which Keynes rejected the supply curve of labor. But if there is no supply of labor, involuntary unemployment becomes a meaningless concept. In order to have involuntary unemployment, the amount of labor workers want to supply has to be greater than the amount of employment employers are willing to offer.
It would seem that in the General Theory there is supply of labor, only it is based on nominal, rather than real wages.
June 3, 2007 9:17 AM | Reply | Permalink
"With no labor market, the adjustment of wages is irrelevant to the problem of unemployment."
That is not corret. Whether it is, or is not relevant to the problem of unemployment, depends on what happens in the other markets.
For example, if the central bank holds the supply of nominal balances constant, a fall in the price level will increase the supply of real balances, which will decrease the interest rate and therefore increase aggregate demand.
That raises all the traditional debates between Keynesians and classicals about whether or not there is a liquidity trap (Keynes did not thing this had happened in the Great Depression), whether even a 0% interest rate would not increase investment enough, etc., etc., etc.
June 3, 2007 9:27 AM | Reply | Permalink
"The phrase "sticky-wage Keynesianism" is a good description of Brad's position. It captures the general alignment of policy positions, while preserving the fact that the underlying theory is not Keynes's own."
Sticky wage Keynesianism does capture the essence of Keyene's model of involuntary unemployment. Aggregate demand is insufficient to make the quantity of labor demanded by firms as great as the quantity of labor supplied by workers, so that there is involuntary unemployment, but nominal wages will not fall. Of course, MIT Keynesianism does not slavishly adhere to the crude models Keynes used in the General Theory, in arriving at this result and righfully so. The General Theory was the beginning of modern macroeconomics, not the final conclusion.
Of course, Keynes argued in Chapter 19 that if money wages were flexible downwards, things would get worse, not better, but the issues involved in this are also effectively dealt with in MIT Keynesian analysis, such as Tobin's article "Keynesian Models of Recession and Depression."
June 3, 2007 9:40 AM | Reply | Permalink
"There are markets for specific products, with which he wasn't much concerned"
Aggregate demand implies a market for output of final goods and services.
June 3, 2007 9:43 AM | Reply | Permalink
"The crucial step was this: labor and land were made into commodities, that is, they were treated as if produced for sale. Of course, they were not actually commodities, since they were either not produced at all (as land) or, if so, not for sale (as labor)."
This is the kind of fuzzy thinking that tends to pervade the left. The fact that something was not produced for sale does not mean that the people who own these things cannot sell them. Workers can and do sell their labor services regardless of the fact that they were not produced for sale. And people who own land will rent or sell it regardless of the fact that the land was not produced. When they offer them for sale, a market for them comes into existence.
June 3, 2007 9:55 AM | Reply | Permalink
This shows that Keynes was not infallible and should not be turned into a saint. While Keynes' analysis introduced in the General Theory was very important in understanding how the economy works in the short-run, when he discussed the long run he was just plain wrong. This is because he incorrectly believed that as income increased people would save an increasingly large percentage of their income, leading to a increasing secular tendency to deficient aggregate demand. Events since that time have disproved this belief.
Neo Keynesians are correct in taking the parts of the General Theory that are productive and discarding the rest. If idelogues therefore accuse them of not really being Keynesian, so be it.
June 3, 2007 11:48 AM | Reply | Permalink
There aren't many generalizations you can make in economics that do not ultimately yield to some exceptions. The important question is how significant are they? What I find especially annoying is the failure of 'Democrat economists' to heap the proper amount of ridicule on the theories of 'Republican economists' who identify certain minor variables and then exaggerate their importance beyond all reason.
One example that comes to mind is the role of 'expectations' in discussions of inflation. Yes, inflationary expectations can play an increasingly significant role as actual inflation begins to approach hyperinflation levels (when hoarding becomes widespread, supply is further restricted, driving prices up even higher), but at even 'robust' levels of moderate inflation, most suppliers are intent on moving product to take advantage of the higher prices they are commanding. Thanks to the repeated exaggerations of Republican Economists, most Democrat Economists accept the fear of inflationary expectations as sufficient reason to crush any and all inflation pressures with high unemployment.
I won't go into it here, now, but another example I can think of is the "incentive" argument that is used to press for lower tax rates for wealthy people. There are few Democrat economists who will argue vociferously---as they should---that a rich person's displeasure at having to pay higher taxes is hardly likely to discourage them from generating as much revenue as they possibly can.
I could go on and on...
June 3, 2007 12:31 PM | Reply | Permalink
This is the 4,967th debate over who has read Keynes correctly. Observers of this debate might conclude, "well it seems a number of different readings of Keynes are consistent with Keynes' work."
Reasonable people might conclude this.
Unreasonable people would say, "well, only one interpretation of Keynes is correct and all who disagree with this interpretation must be banished from top journals, etc. Some of these (wrong) interpretations are so dangerous to our society that we must keep these ideas from our graduate students because if they read Keynes differently than we do, society might collapse."
June 3, 2007 1:36 PM | Reply | Permalink