As widely reported, the Bank of England is intending, over the next three months, to purchase some £75bn of assets - mainly gilts - in an operation of quantitative easing (i.e. printing money). The right-wing press and blogosphere seems highly disconcerted by this, even in some cases outright opposed. I think it is right to be nervous about this policy, and I shall propose a refinement, but I support it nonetheless.
Talk of printing money (which is broadly the right way to describe what is happening) is liable to make us imagine that the Bank of England is increasing the money supply. Consequently, of course, we suspect that that will lead to inflation. Inflation may be a kind of solution to our woes - inflating away our debts - but (one might think) might the medicine not be worse than the disease?
There is not nothing in this thought - indeed, I do expect that there will eventually be inflation - but it misses a key piece of opening context. It isn't that we are in a situation in which the money supply is ticking along normally and then the Bank of England decides to print a load of extra money. No. The startpoint - the current situation - is (effectively) that the money supply is shrinking rapidly because of the collapse of the financial system. This shrinkage in the money supply would (if it did not reverse itself or were not countered by policy) lead to considerable deflation - to prices falling.
Thus, quantitative easing is not about expanding the money supply. Rather, it is about stopping the money supply from shrinking. It is not, in the first instance about creating inflation. It is about countering deflation.
But if it works, it will still probably lead to inflation. Why? Well, with the banks bust, the money that the Bank of England injects may not, in the short term, increase the total money in the economy by much more than the £75bn it is printing itself. But if the banks start to function again properly, each £1 the Bank has printed may turn into £15, £20, perhaps even £30 of extra lending by the banks. This makes the exercise of working out how much money to print incredibly difficult.
If you print too little, the deflationary impact of the collapse in the financial system will not be countered, prices may start falling uncontrollably, people's salaries will start falling as well, at lower salaries they will give up the ghost on trying to service their debts by defaulting, and greater defaults will create further losses for the banks generating more deflation. But if you print too much, suddenly it will work, and you will end up combatting excess monetary demand of 150%, 200%, perhaps 300% of GDP - implying inflation rates well into the hyperinflation zone.
Even if you knew very precisely how much to print (in fact we have no very good idea at all - probably we don't know how much is needed even to the nearest 10% of GDP) - even if you knew very precisely, that would not be the end of your problems. For you would then face the problem of what to do at the margin. If you print 1% of GDP too much, that might lead to 30% inflation, but if people expect you to react to 30% inflation by taking 1% of money back (quantitative tightening) the banks might save the extra 1% and you would not break out of the deflationary/money-hoarding regime into normal functioning of the banking system. So the precise amount needed to restore functioning to the financial system depends crucially upon expectations - expectations that could shift suddenly (with large inflationary or deflationary consequences) when the time comes.
It is all incredibly difficult to manage successfully - and no country has ever done so. Quantitative easing has always either been inadequate (leaving deflation in place) or has created uncontrollable inflation when it worked (on the "exit path") - the "ketchup-in-the-bottle effect". To help avoid these problems this time, I believe a price-level target might help - that would tell us how much inflation would be allowed to stay in the system and how much would be reversed later. This would allow the management of expectations at around the margin of the "right" amount of money to print. That limits the problem to that of finding that "right" amount - still very hard but with a bit of trial-and-error we might have a chance.
With an inflation target, rather than a price-level target - well, I say that, though of course the UK's inflation target has collapsed, but let's run with the idea for now - with an inflation target, I don't believe that it will be possible, at the margin, to have sufficient control of expectations to both print enough to be effective and also to avoid a ketchup-in-the-bottle effect when the policy works. I had expected us to be more cautious in quantitative easing than the US and consequently to have less inflation on the exit path. That's not looking so obvious now - permission to print £150bn and some £75bn printed in just the next three months is a large start. We'll probably struggle to keep inflation below 10% on the exit path, and it may go well into the teens or even higher.
That's not Zimbabwe, but it could represent the undoing of all the hard-won achievements of the Major and Thatcher governments on getting down inflation - to add to the undoing of all their efforts to get down public expenditure as a percentage of GDP. I accept this risk (indeed, this likelihood), but, with great regret, I consider it a price worth paying. We just don't want to know what happens from here, as the most privately-indebted country ever, if we allow deflation to get out of control...
