Tuesday, May 7, 2013

Keynes’s Mistakes in the General Theory

Matias Vernengo raises the issue of Keynes’s errors in the General Theory in this post.

The mistakes and oversights that Keynes made in the General Theory are arguably as follows:
(1) the assumption of an exogenous money supply;

(2) the marginal efficiency of capital (MEC) idea. Keynes, in developing the MEC, failed to free himself from the neoclassical marginal productivity of capital (King 2002: 209):
“[sc. Keynes] made a fatal mistake in offering a quasi-long-period definition of the inducement to invest as the ‘marginal efficiency of capital’, that is, the profit that will be realised on the increment to the stock of capital that results from current investment and, still worse, identified the profitability of capital with its social utility. This was an element in the old doctrine from which he failed to escape. He had an alternative concept of the inducement to invest as the expected future return on sums of finance to be devoted to investment. Minsky (1976) points out that he did not seem to recognise the difference between the two formulations. If he had stuck to his short-period brief, he would have used only the second.” (Robinson 1979: 179–180).
The MEC seems to suggest that there exists a rate of interest which is low enough to induce full utilization of capital goods. But this is just smuggling in the Wicksellian natural rate of interest, when Keynes had wanted to abandon the natural rate.

A number of Post Keynesians reject the MEC, because it is based on the neoclassical or marginalist theory of distribution.

(3) Keynes did not sufficiently stress the role of uncertainty and expectations in undermining the coordinating role of interest rates (King 2002: 14). In Chapter 18 of the General Theory, Keynes played down the role of uncertainty (which he had stressed in Chapter 12) and, if he had really maintained the crucial role of uncertainty (as he did later in Keynes 1937), this would have “ruled out any stable functional relationship between investment and the interest rate” (King 2002: 14). The door was thereby left open for neoclassical synthesis Keynesians to reformulate the General Theory as a general equilibrium model where the interest rate has a pivotal role (King 2002: 14).

(4) In Chapter 2 of the General Theory, Keynes used the marginal productivity of labour concept. Later he was criticised by Lorie Tarshis and Dunlop, who invoked empirical evidence on pro-cyclical wages, and in Keynes (1939) he came to reject this marginalist idea, apparently giving some endorsement of Kalecki’s theories.
Other possible problems include:
(1) Did Keynes properly understand the heterogeneous nature of capital goods? Possibly he did (see Hayes 2007), though the Cambridge capital debates were long after he died;

(2) Did Keynes understand the extent and significance of fixprice markets? One charge against Keynes is that the General Theory does not consider fixprice markets properly. By contrast, Michał Kalecki did understand fixprices, in his ideas on cost-determined pricing. Kalecki and later Post Keynesians understood that as long as excess capacity exists in fixprice market firms, then government stimulus produces direct increases in output and employment in the latter markets, not just inflation.
“Keynes’s Marginal Efficiency of Capital: A Mistake?,” January 1, 2012.

“Post Keynesian Policy on Interest Rates,” March 12, 2013.

“Interview with Bob Rowthorn,” March 4, 2012.

Hayes, M. 2007. “Keynes’s Z-Function, Heterogeneous Output and Marginal Productivity,” Cambridge Journal of Economics 31.5: 741–753.

Keynes, J. M. 1937. “The General Theory of Employment,” Quarterly Journal of Economics 51: 209–223.

Keynes, J. M. 1939. “Relative Movements of Real Wages and Output,” Economic Journal 49: 34–51.

King, J. E. 2002. A History of Post Keynesian Economics since 1936. Edward Elgar Publishing, Cheltenham, UK and Northampton, MA.

Robinson, J. 1979. “Garegnani on Effective Demand,” Cambridge Journal of Economics 3: 179–180.


  1. Kaldor has some articles in Vol 9 of his Collected Essays on the shortcoming of the GT

  2. (1) I'm coming less and less to think that money exogeneity was that big a deal for the General Theory. These days, as in Keynes' time, the real argument was over the natural rate of interest. Its far more post-Keynesian to accept exogenous money and reject the natural rate than it is to accept endogenous money and accept the natural rate. So, I don't think the GT suffers too much from relying on a variable velocity rather than a variable quantity.

    (2) Yes, this is a serious problem. But Minsky is correct. Keynes had an alternative formulation. I think he just made this concession in the GT so that neoclassicals would understand what he was talking about. Keynes knew all too well that investment has to do with animal spirits/effective demand and not some rational calculation based on the interest rate.

    (3) Agreed. As in point (2).

    (4) Again, I think that Keynes was well prepared to throw that old marginalist stuff under the bus. And I think that anyone who seriously reads the GT would be prepared to do so too.

    (1b) I really don't think this was within the scope of the GT.

    (2b) I don't think Keynes appreciated this dimension or its importance at all. I think it was really Kaldor that developed this thoroughly -- although it can be found in Kalecki.

    Oh, and a more general point about the accusation of "Keynes worship" that is sometimes thrown at post-Keynesians. This is completely ludicrous. PKs tend to be quite scholarly in that they try to pinpoint who said what and when. That way we can get a clear map of when and where certain ideas developed. I find this enormously helpful in that it forces you to recognise when there is a hole or a discrepancy in the "General Theory" of macroeconomics.

    What is derided as idolatry in economics is recognised as good scholarship in almost every other social science. The accusations of Keynes worship reflect not so much on PKs as on the poor state of the discipline.

  3. Regarding point number 3, Kregel had a paper, Economic methodology in the face of uncertainty. His points are quite telling, I think they refute particularly item 3. As for this paragraph: "The MEC seems to suggest that there exists a rate of interest which is low enough to induce full utilization of capital goods. But this is just smuggling in the Wicksellian natural rate of interest, when Keynes had wanted to abandon the natural rate." I thoroughly disagree. I fail to see how we can call "natural" a highly conventional, monetary interest rate, which does not adjust to bring full employment. I will say more. All the elements of the General Theory are already present in the Treatise on Money (really, even the liquidity preference, which is called "bearishness"). The difference is that in the Treatise there is (and in the General Theory there's not) a natural rate. The word "natural" has implications, when it is used to refer to the "natural" interest rate or the "natural" rate of unemployment. None of those implications are present in the attributes of the MONEY interest rate in the General Theory. We may not like the MEC, some may not like liquidity preference, etc., but there are no grounds to say that Keynes hold a natural rate in the General Theory.

    1. I am not saying Keynes that did not abandon the natural rate in the GT, but that the MEC seems to leave the door open to it in an ambiguous, inconsistent way.

    2. It seems to me that Keynes was talking about some sort of natural rate in the General Theory. Indeed Keynes says this explicitly:

      "I had, however, overlooked the fact that in any given society there is, on this definition, a DIFFERENT NATURAL RATE OF INTEREST for each hypothetical level of employment. And, similarly, for every rate of interest there is a level of employment for which that rate is the "NATURAL" RATE, in the sense that the system will be in equilibrium with that rate of interest and that level of employment." (Keynes, GT, Chapter 17, My Emphasis)

      For Keynes in the GT there is a weird dichotomy. On the one hand, investment is driven by "animal spirits" and wage-growth is determined institutionally, yet on the other hand he claims that there is a different natural rate for each level of employment. What this actually means seems to me entirely vague. Does it mean a rate of interest that would stabilise prices at a given level of employment? So, at, say, 10% unemployment there is an interest rate that ensures that no inflationary demand is realised for this level of employment? Keynes appears to be saying something like this and it is very odd altogether.

      This is definitely one of those "Keynes got it wrong" moments for me. Or, more specifically, he got it wrong in the GT and clarified later.

    3. I suspect that Keynesian of the non post- variety would see this as Keynes getting it right.

      He is saying that you can change the equilibrium level of employment by adjusting the interest rate - which is at the heart of most Keynesian models outside the zero bound.

  4. Can someone explain to me what in God's name the 'natural rate of interest" has to do with the marginal efficiency of capital. I always thought that the productivity of capital, both marginal and otherwise, has to do with the unique nature and output per unit of time of the piece of machinery or capital equipment itself, (like a computer running on 3 petaflops per second, or a nineteenth century loom being 10 times faster than a hand tailor.)

    1. I believe the idea is that the natural rate induces optimal investment. Deviation from that rate can result in failure to invest wisely, productivity can then stall or even decline resulting in disequilibrium.

  5. Whilst this is a belated response, Lord Keynes, there was a 2008 comment posted on Mark Hayes's 2007 article in the Cambridge Journal of Economics that I think you might want to read.


  6. I just came back to this today while doing some reading. I actually now think that Vernengo is wrong on most of these points... Especially the MEC point. That seems terribly, terribly wrong.

    1. You mean the marginal efficiency of capital concept is "terribly, terribly wrong"?

      Or do you mean MEC is actually correct?

    2. MEC is fine. Robinson and Vernengo's interpretations are wrong. Here is a very accurate summary by GLS Shackle that I cannot fault:

      "There is in [Keynes'] respectful treatment of the inducement to invest a most striking contrast with the relatively contemptuous dismissal of saving as a mere automatic residue of income depending mechanically upon the size of the latter. And we must not be misled by that imposing vehicle called the marginal efficiency of capital. This is a mere basin in which the investment-decision maker can finally stir together his ingredients, it tells us nothing about of what those ingredients should be. The marginal efficiency of capital is a mere abstract and general statement of a sum to be done. It entirely lacks any suggestion of any of the 'givens' without which, of course, no real, arithmetical sum can be done. It really says no more than that expected installments of profit from proposed equipment must be discounted and summed, and the result compared with the first cost of equipment. It says nothing of how the profits themselves are to be estimated, and nothing, either, as to how those estimates are to be adjusted for the flimsiness of any possible base upon which they can be erected." ('A Scheme of Economic Theory', GLS Shackle, p99)

      Victoria Chick and Engelbert Stockhammer had a wonderful discussion of this at Kingston University recently where we all read the General Theory and this was the consensus view on what the MEC is all about. It's actually a very useful concept. And Keynes states quite explicitly in the GT that it has nothing to do with marginal productivity.