Paul Krugman and the liquidity trap
I thought Paul Krugman was the only senior economist afflicted by the scourge of the liquidity trap. But apparently the disease is catching. The New Yorker quotes Larry Summers as saying this month that while searching for insights about what to do in a debt crisis, "I was heavily influenced by the basic I.S.L.M. framework augmented to take account of liquidity traps." In what follows I offer an effective vaccine against the liquidity trap. But of course if people decline to be vaccinated … [shrug shoulder here].
The liquidity trap thesis originates with what Keynes called the speculative motive for liquidity preference. In Keynes the argument is put somewhat like this. When interest rates are very low, everybody expects them to go up shortly. If this were to happen any bonds they hold would depreciate and they would sustain a loss. Therefore there is a general preference to opt out of bonds and instead hold cash.
It is troubling that 75 years after The General Theory was published Krugman does not realise that a) there is an unstated assumption behind the liquidity preference thesis and b) the assumption is invalid. But of course if you believe, as Krugman does, that all you need to understand recessions and their cure is to reread Keynes, it gets a bit difficult to notice any errors that the master might have made.
Let's take a concrete example. If I have $100 after spending on consumption from income I have two choices. I can buy a bond with it or retain it as liquid cash. (Of course I can choose to retain $50 as cash and buy a bond with $50 but the argument would not change; it would still apply to each of the $50s, or the two parts of whatever proportion the $100 is divided into). It is an EITHER/OR situation. If I choose one I cannot have the other. Sounds logical.
So where is the assumption? Although it is not obvious at first glance, Keynes is assuming an all-currency bankless system. The EITHER/OR choice applies only to currency. If I place the $100 in a bank demand deposit, I have complete command over the liquidity it offers. At the same time the bank is at complete liberty to buy a bond with it. For the economy as a whole there is no EITHER/OR situation. By splitting the act of retaining liquidity from the act of buying a bond it can have both. In a primitive all-currency system, which might have existed a couple of centuries ago, we would assume that Keynes's arguments applied, but in such a system there would have been no bond trading so actually the liquidity preference thesis is not valid at any time or place.
But let's forget the poor logic and accept a common language definition of the liquidity trap. It is that when interest rates are very low people hoard cash and do not offer them up to the sellers of bonds, or in other words, for lending. During the Great Depression it is true that people kept cash under their mattresses. But this was because banks were going bust and there was no deposit insurance, not because of people's expectations of the way interest rates were going to move. Still, one can give Krugman the benefit of the doubt for that period because of the paucity of data.
However, every recession since then has proved the liquidity trap theory wrong. Take the 2008-09 recession for example. In a blog in the New York Times in March 2010 Krugman wrote that the US was "obviously" in a liquidity trap. So one can test the hoarding thesis for that period.
The figure below shows demand deposits from December 2007 (when the recession officially began) to March 2010 (the point when according to Krugman the US was in a liquidity trap).
Does it look as if people were hoarding cash? If they were, the quantum of demand deposits would plunge instead of soaring; the drop from December 2008 to March 2009 reflects a plunge in bank lending. People were readily putting their cash in banks (obviously without putting any obstruction in the way of its being lent out or being used to buy bonds); it is a different matter that banks were not lending out the cash. So there goes the hoarding thesis.
But Krugman has another version of the liquidity trap. "In my analysis, you're in a liquidity trap when conventional open-market operations — purchases of short-term government debt by the central bank — have lost traction, because short-term rates are close to zero," he wrote in another NY Times blog. But that does not make a liquidity trap. In fact, I would not only agree with Krugman that large-scale purchases of short-term debt by the central bank have no traction but believe that they are the principal obstruction to a recovery.
It is the argument behind the observation that is faulty. Krugman argues that at low interest rates the demand for money by the system expands to absorb any money that the central bank throws at it. But this lands us in a fix. Assume that interest rates are around 2% and that bond holders are already holding plenty of cash. Still, the central bank wants to pump more cash into the economy by buying a lot more bonds. To do that it has to offer to buy bonds at a price higher than the present market price which means that it has to lower interest rates further. Thus we find ourselves in a situation where when the demand for money expands, the price (or really speaking, rent) of money goes down. In other words, we are in a situation where we have to abandon the law of demand and supply, which is the fundamental law of economics. And if we do that it is only a step further (or backwards) to abandoning the idea of marginal utility and returning to the labour theory of value. We therefore have to abandon the idea of the liquidity trap.
Several conclusions follow:
1. Krugman is wrong and there is no liquidity trap.
2. The demand for money is not what Krugman thinks it is. The demand for money does not come from the seller of bonds to the government but from those who demand loans. When this demand goes up interest rates go up as they should if the law of demand and supply is not to be violated.
3. What Krugman thinks is money is not money, because money is created not when the central bank pumps cash into the system, but when banks make loans.
(As for what money is, read The Riddle of Money, Finally Solved.)
It is worth taking a second look at Keynes on the subject of liquidity preference.
In The General Theory he writes:
"It should be obvious that the rate of interest cannot be a return to saving or waiting as such. For if a man hoards his savings in cash, he earns no interest, though he saves just as much as before. On the contrary, the mere definition of the rate of interest tells us in so many words that the rate of interest is the reward for parting with liquidity for a specified period."
Keynes is right in saying that the rate of interest cannot be a return to saving or waiting as such. But it is not for the reason Keynes gives: For if a man hoards his savings in cash, he earns no interest, though he saves just as much as before. The reason is that the act of lending is carried out not by the saver but by banks (and shadow banks) and banks are not constrained to limit the amount of lending to the amount of savings.
Or again he says:
"Thus the rate of interest at any time, being the reward for parting with liquidity, is a measure of the unwillingness of those who possess money to part with their liquid control over it. The rate of interest is not the 'price' which brings into equilibrium the demand for resources to invest with the readiness to abstain from present consumption. It is the 'price' which equilibrates the desire to hold wealth in the form of cash with the available quantity of cash; which implies that if the rate of interest were lower, i.e. if the reward for parting with cash were diminished, the aggregate amount of cash which the public would wish to hold would exceed the available supply, and that if the rate of interest were raised, there would be a surplus of cash which no one would be willing to hold."
But the error in this is obvious. It applies only to an all-currency system, not to a system which contains banks. And empirically too we saw that Keynes' contention was not true. As we saw in the graph, given deposit insurance, people are willing not to hoard cash even when the interest rate is close to zero.
When I started to write this article it was with the limited intention of demonstrating the obvious error behind the idea of liquidity preference. I believed that liquidity preference was at best a peripheral part of Keynes's work. But having come thus far I have changed my mind. Indeed the logic of my argument leads me to believe that liquidity preference is the central error of Keynes, and that proving it wrong destroys the whole of Keynesianism.
First, Keynes was wrong in thinking that people hoarded money at low interest rates because of calculations about the price of bonds. To begin with, the argument only holds on the basis of an unrealistic assumption viz. the non-existence of banks and demand deposits. There is also clear empirical evidence that this is not the case. What actually happens is that banks do not lend money to firms, both because of low interest rates and because the destruction of the previous asset bubble has blown a hole in their balance sheets. Thus the central problem is not hoarding but the collapse of lending by banks because of which firms cut back on the manufacture of goods and provision of services, and thus on factor payments, of which wage payments are the easiest to cut (the wholesale unemployment during the Great Depression or the recent Great Recession is excellent proof of this). And finally, this means that recessions are not a failure of aggregate demand but a failure of aggregate supply.
17 April 2011