There is a lot of confusion about quantitative easing, especially now that the US has joined the UK in trying it. The usually clear Robert Peston has finally given up trying to understand. This is from his blog entry:
(When economists claim that something only works under an assumption of irrationality, make sure to check that it is not just a way of saying it is all a muddle to them).
The clear head, as so often, is Brad de Long:
The answer to Peston's question about how quantitative easing works without irrationality all-round is that monetary policy has always worked through market power: it is because the central bank can be a bigger player than any other that it can move prices. Nothing irrational in recognising that.
In ordinary times, the central bank shifts its own portfolio towards short term bonds when it is pushing down interest rates and towards longer term bonds when it is pushing them up. This, in ordinary times, can be done through simple portfolio movements and requires no new cash in the system. It can influence liquidity preference by making it more or less attractive to hold cash. As it sells longer dated bonds and buys short dated bonds, short term interest rates fall relative to long term interest rates and people find holding cash less expensive. So they hold more of it, and banks can lend more. That is monetary easing in normal times.
But when short term rates are effectively zero, you can't do that anymore. But you can certainly make cash more attractive by conbtinuing to buy bonds where interest rates are non-zero, like longer dated government bonds. True, to do this on a large scale, you now need an overdraft (with yourself---which is what seems most troubling to Peston). But the mechanism by which monetary policy is working is still exactly the same. It works because the central bank is large enough to push prices around.
Peston worries about Ricardian equivalence --- the notion that public borrowing doesn't help the economy because everyone knows that taxes will have to rise in the future to pay for the borrowing, and so will save a compensating amount today. This is a completely different point, and depends on whether we are indeed in a Keynesian recession. As I have written before, Osborne and the Conservatives have tried to argue the Ricardian equivalence point, but not very persuasively.
When the central bank performs quantitative easing, it ends up holding more long dated assets like government debt in its portfolio. If it comes to sell these assets in the future, it will reduce the money supply correspondingly. And if we accept that we are in a Keynesian recession with a liquidity trap, there is nothing wrong or strange about the transaction. Indeed, the taxpayer could make money out of it.
Brad de Long's important point is that this process relies on the credibility and trustworthiness of the institutions involved in carrying through the policy over its entire term. The credibility point is that fiscal stimulus requires borrowing, and even quantitative easing requires enough people to believe that inflation will not instantaneously erode the value of cash balances by exactly compensating amounts. So the liquidity created today must indeed be mopped up when times get better. The point about Keynesian equilibria is that these assumptions need not be illusory -- they actually hold. So, in Brad's words:
Willem Buiter has been offering the powerful argument that the credibility of the UK and the US are shot in this regard, here on openDemocracy and here on his Maverecon. This is the interesting argument against massive injections of liquidity (or massive fiscal stimulus).
