Pigou, Samuelson, Solow, & Friedman vs Keynes and Krugman

Thoma Bait

I might as well be honest. I am posting this here rather than at rjwaldmann.blogspot.com , because I think it is the sort of thing to which Mark Thoma links and my standing among the bears is based entirely on the fact that Thoma occasionally links to me.

I think that Pigou, Samuelson, Solow and Friedman all assumed that the marginal propensity to consume out of wealth must, on average, be higher for nominal creditors than for nominal debtors. I think this is a gross error which shows how the representative consumer (invented by Samuelson) had done devastating damage already by 1960.

The topic is the Pigou effect versus the liquidity trap. Keynes argued that economies might fall into liquidity traps in which the safe nominal interest rate is zero so money and bonds are perfect substitutes. This matters a lot, because Keynes claimed that the way in which wage flexibility lead to full employment was as follows wages fall so prices fall so the real value of the money stock increases so interest rates fall so investment increases. This only works if a higher real value of the money stock causes lower interest rates and that can’t happen if interest rates are already zero. The possibility of a liquidity trap is the possibility that flexible wages and prices aren’t enough to get to full employment.

On many important topics Milton Friedman disagreed with Samuelson and Solow, but both dismissed the possibility of a liquidity trap. They were very casual about the argument with the tone of people restating something which is well known an, once you think of it, obvious. They appealed to the Pigou effect. Pigou’s argument is that money is not just a means of payment but also a store of value. A huge decrease in the price level makes the money which the average person holds much more valuable. That increased wealth causes the average person to spend more so the economy returns to full employment.

The problem is that they decided to simplify their model of the world by considering only two agents, the state and a representative consumer (who owns the firms). The state is the only debtor. They further treat public spending on goods and services (G) as exogenous.

In the real world some private sector agents are nominal debtors too. They include everyone with a mortgage and almost all firms. A reduction in the price level causes debt deflation increasing the value of debts and not just the liabilities of the state (including money). Irving Fisher had already stressed the imortance of debt deflation long before 1960, so what happened ? In the real world, nominal debtors including firms and liquidity constrained households have much higher marginal propensity to spend out of wealth than creditors. Very much higher. So debt deflation has a much greater depressing effect than the tiny stimulating Pigou effect.

Now there is another problem with appealing to Pigou. This is in particular a very very serious problem for Friedman. In other contexts, Friedman argued that “to spend is to tax” that the effect of the state on private wealth (with other things including GDP equal had to be minus the present value of G no matter how G was financed. In particular he stressed that financing government spending by printing money created an inflation tax which was his vivid name for seignorage. In other contexts, Friedman insisted that transfers from the state to the private sector did not affect consumption at all. This means he argued that, given a nominal interest equal to zero, if the real value of the money stock increased, because the price level fell, then people would spend more, and also argued that if the real value of the money stock increased just because the government printed more money and gave it to people, then people wouldn’t spend more. I have claimed that I can get any result I want out of a model with rational utility maximizing agents. The reason I make this claim is I know that, as a last resort, I can totally cheat as I do below.

The point is that Pigou ignored the public sector budget constraint, or rather assumed that both policy makers and private agents would ignore it. If money is worth more then either G must be cut sooner or later or taxes must be increased or there must be inflation bringing the value of the money back down. Just printing money doesn’t make the economy as a whole richer — the social budget constraint is the same. It can and does cause higher spending by creating the illusion of wealth. In every context except the debate over whether there could be a liquidity trap, Friedman assumed that people have no such illusion of wealth. He claimed to believe both in the Pigou effect and in Ricardian equivalence. It is not possible that a person as intelligent as Milton Friedman could believe both things. I think this is just another (of many) proofs of his intellectual dishonesty.

OK so how to reconcile the claims of a Pigou effect and Ricardian equivalence (with assumptions that it is not possible to believe). Assume people believe that Milton Friedman has an imortal soul not exactly in heaven. Assume that people believe that Milton Friedman’s soul is happy if data confirm his hypotheses. Assume people want Milton Friedman’s soul to be happy. Assume this is all they care about. Then it is rational for them to act in the apparently contradictory manner he predicted consuming no more if there are temporary tax cuts and consuming more if deflation increases the real value of the money stock. The last resort is that your prediction is consistent with rational utility maximization if the sole aim of everyone alive and everyone who will ever live is to confirm your prediction. Notably, a week ago I hadn’t thought of any other set of claims so utterly and obviously false that I had to resort to this example.