Wednesday, July 21, 2010

John Stuart Mill, Keynes, and Employment

In this post on the breakdown of Okun's Law, commenter Sebastian raises a very important point about what is known as John Stuart Mill's Fourth Principle, that "demand for commodities is not demand for labor". He offers this link to the Organizations and Markets blog. I'll quote Mill from his Principles of Political Economy (1848):

"Demand for commodities is not demand for labor. The demand for commodities determines in what particular branch of production the labor and capital shall be employed; it determines the direction of the labor; but not the more or less of the labor itself, or of the maintenance or payment of the labor. These depend on the amount of capital, or other funds directly devoted to the sustenance of labor."
This is something that is often cited as being misunderstood or abandoned by Keynes. I've always been somewhat confused by the point. Keynes has an employment function that is a function of output, which I suppose is the source of the concern. But Keynes:

1. Talks about employment functions for specific industries and firms and contrasts this with the aggregate employment function.

2. Talks about the importance of capital employed in and industry in the determination of the employment function.

3. Talks about how the total employment function can change depending on whether labor is directed towards industries and firms with high or low elasticities of labor demand.

So to me, Keynes seems to hit all of Mill's bases. I honestly don't know what the issue is. I have argued that Keynes could have pushed this analysis farther, but I don't think he completely dropped the ball on the employment function, the way some people suggest. I think where Keynes could most revise his employment function is in bringing his liquidity preference into it, not in abandoning the project for some other formulation.


  1. I may be missing a link (hey, three hours of sleep, etc.), but where does Keynes talk about these three factors exactly? Is concentrated in a few paragraphs, or is scattered throughout his oeuvre?

  2. I got all that from Chapter 20, which is dedicated to the employment function. He definitely mentions the relationship between employment demand and other factors of production earlier, because I referenced back to that this morning - I think it was Chapter 4 - don't remember exactly. I know he also talks about the roundaboutness of production in a chapter titled "Observations on the Nature of Capital", but at the moment I forget whether that discussion gets into employment or not.

    Sorry - my copy of the GT is at home. All of that should be in 20, but it's also elsewhere.

  3. Garrison raises this point, but the odd thing is I don't see how Garrison's treatment of labor demand is any different from Keynes's. As far as I can tell, there is nothing that Garrison mentions that Keynes doesn't.

  4. Ok. Just curious because you hadn't made a citation and you're normally rather good about that sort of thing.

  5. No problem - sorry about that.

    Chapter 20 is here for anyone interested:

    I'm very intrigued by Keynes's discussion of the capital structure here and his discussion of roundaboutness of production in his chapter on the Nature of Capital. Does anyone know of anything that has been written on Keynes's views on capital structure? I think often it's just assumed it was completely ignored. That seems obviously wrong. But I think if you think he completely ignored it, you generally don't get much commentary on what he did say about it.

  6. Garrison raises this point, but the odd thing is I don't see how Garrison's treatment of labor demand is any different from Keynes's. As far as I can tell, there is nothing that Garrison mentions that Keynes doesn't.

    No Austrian has ever subscribed to the rough caricature that aggregate employment is related to aggregate output like Keynes does.

    For one reason, demand for commodities really isn't demand for labor - labor is demanded in proportion to how quickly is can produce commodities, but there is no causal or inevitable link between aggregate output (demand for commodities) and aggregate employment (demand for labor)

    For the second reason, Austrians deny even the notion of aggregation. We deny that there is a such thing as "aggregate output" because there is not one banner you can put all goods under. To my knowledge, a lot of Austrians even shy away from using conventional terms like 'consumer goods' and 'capital goods' given that commodities often switch between these.

  7. Aside from the "no causal link" part (that's absurd, Mattheus - of course there is a causal link), what in your first two paragraphs do you think Keynes doesn't agree with you on?

    Your third paragraph is acknowledged, but I think also silly. Garrison certainly doesn't have a problem with aggregates as long as you acknowledge the heterogeneity within the aggregate (which, by the way, Keynes does).

    The difference seems to me to be not how both sides think about these things (except for ontological silliness like your third paragraph - that is a real difference in how we think about it). The real difference, I think, comes in what factors we think are important. Austrians think the sub-aggregate capital structure is the most important, Keynesians think aggregate behavior of things like liquidity preference is more important.

    That's an emphasis difference - the vast majority of Keynesians recognize heterogeneity as surely as the vast majority of Austrians recognize economy-wide processes.

  8. I don't believe there's a causal connection between employment and output per se. Keynes and, to my knowledge, most Keynesians do. There are functions that relate the "quantity of employment" with output across the economy.

    The aggregation Garrison makes is basic. He aggregates "savings" and "consumption." Two topics we can have broad aggregation on. Like I said, not all Austrians make those aggregates but Garrison does, and it's justified. I can't think of savings or consumption in more than one way, so it's justified to aggregate them. Investment, employment, etc. all have multiple interpretations and provisions. That is what I object to.

    I don't think Keynes agrees with my second paragraph at all. It's an entire repudiation of his theories on employment.

    There's more than an emphasis difference. If it were merely an emphasis difference, we wouldn't be espousing diametrically opposite policy prescriptions. One of us is wrong.


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