The classical theory of unemployment

There are some bizarre discussions in the literature showing the lengths the (pre-Keynesian) classics would have had to go to to produce a theory of unemployment given fixed money wages, and showing that their attempts would inevitably come to grief. See the diagram on p52 of Hayes (based on the supply curve for labour) and the ‘graphical representation of Pigovian labour demand’ on p66 of Ambrosi – one of the most indigestible diagrams I’ve ever encountered.

On the other hand Klein gave a perfectly sound (but well buried) account under which the classics needed have no difficulty. This is expressed by my simple diagram on p87. I suspect that Klein kept quiet about his line of reasoning because it demolishes half of Keynes’s Chapters 2 and 19. Hayes and Ambrosi can be excused for not having seen it... but really the argument is obvious. I certainly found it so myself. I hummed and ha-ed for a long while over whether I dared put it on Wikepedia without a source; eventually took the plunge; and found Klein’s explanation soon after (giving me the necessary reference).

Ambrosi on hoarding

I interpret Ambrosi as embracing a hoarding explanation of effective demand. He says that ‘savings, in the Keynesian view, are typically not in kind’ (p133). I have a blind spot for these things: if someone says an asset is ‘not in kind’ I hear him as saying that it’s monetary; if he says that it’s not monetary, I hear him as saying that it’s real.

Unfortunately there are reasons to believe that some Keynesians posit a third class of asset, exemplified by loans, whose properties are as follows:

• Items in it are not real assets, because they do not contribute to aggregate demand.
• They are not monetary because they are not constrained by the externally fixed money supply.
• They do not cancel out in aggregate because they contribute to unemployment.

It may be that Ambrosi’s savings which are ‘not in kind’ belong to this third category, and that I was wrong to attribute a hoarding explanation to him. You can only summarise someone’s views if they make sense.

The classical theory of lending, I assume, has different properties:

• Loans of money are monetary assets.
• Loans of goods are real assets.
• They all cancel out (lending with borrowing), so you can forget about them.

At one point I scoured Marshall’s Principles  looking for confirmation of this, but Marshall (understandably) saw nothing which needed to be explained.

Ambrosi uses the empty term ‘financial asset’ to denote a saving which is not mere cash. Of course loans are assets, company shares are assets and so are goods; the word in itself explains nothing.

If a third category exists, it would be nice to know more about it. I have always assumed that company shares were real assets. If not, it would be useful to know whether assets cease to be real when they are owned jointly, or when their owners do not live on the premises, or for some other reason.

If company shares are real assets but debts belong to the third category, then it is difficult to draw a line between them. It is possible to lend money to a company on terms such that the interest is specified as a function of company profits rather than being defined in advance. Everything works exactly the same as for shares, so the question arises of why shares contribute to aggregate demand and share-like loans do not. Or if, on the other hand, share-like loans are in fact real assets, we need to know at what point a change in the method of calculating interest deprives an asset of its ability to contribute to aggregate demand.

The meaning of Ambrosi’s ‘not-in-kind’ savings is one question, but there are others:

• ‘Typically’ is an ugly academic word used in place of ‘mostly’ or ‘usually’, felt to be more prestigious but simply wrong.
• If we are told that saving, in the Keynesian view, is mostly not in kind, then once we’ve resolved its meaning we need to consider the truth of the statement. Being blandly told us that Keynesians see it this way is no help.
• And there is the further question of its significance. The only possible relevance of the claim that saving is mostly not in kind is under the assumption that saving in kind does not have the deleterious effects which Keynes attributes to other forms of saving.
• Although there are passages in which Keynes appears to embrace this view, there are more in which he overtly rejects it.
• If the harmful consequences of Keynes’s saving are limited to not-in-kind saving, then properties of saving in general – such as its identity to investment – have no bearing on his argument; and the support for his conclusions therefore collapses.

Shortly before he casually undermines Keynes’s theory Ambrosi advises us that:

The belief that ‘substantial error’ should be involved in the presentation of an idea so central for his entire theory [sc. Keynes’s concept of effective demand]... is so sensational that it could not be accepted lightheartedly (p120).

Fleeming Jenkin on unemployment

I quote Jenkin’s account of unemployment on p14 of my book, rather giving the impression that he used the ‘first postulate’ with fixed real wages. His argument was framed in terms of supply and demand curves; the demand curve was probably the one people would have obtained later from the ‘first postulate’ with fixed real wages; but he did not derive it in that way himself, and could hardly have done so, writing a year before the ‘marginalist revolution’. Of course the significance of his theory is not that it agreed exactly with what people were saying 60 years later, but that he was saying it 60 years in advance of them.