Should the UK abandon its inflation target? The case for doing so is that inflation targeting works badly when an economy is hit by supply shocks, as it is now, with food and oil prices soaring. Such shocks depress economic activity anyway, so attempts to control inflation by raising interest rates exacerbate the downturn.
Inflation targeting works best when inflation is the result of demand fluctuations. When this happens, the Bank can both control inflation and stabilize demand by cutting rates in a downturn and raising them in a boom.
So, what's the case for keeping the inflation target now? This chart shows it. It shows that when we've had a credible monetary policy - inflation targeting recently or Bretton Woods in the 60s - inflation has been lowly or negatively serially correlated. By contrast, when we had no such credible policy in the 70s and 80s, inflation was positively correlated.
Back then, people reacted to high inflation by thinking: "Our costs are going up. We need to raise our prices or wages to cover them." So inflation fed on itself. By contrast, a credible anti-inflation policy makes folks think: "The Bank is going to raise interest rates. So if we raise prices or wages, demand for our services will fall." Rises in inflation are then only temporary.
Better still, because inflation stays low because people keep prices and wages down voluntarily, the Bank doesn't actually need to raise interest rates to control inflation. We therefore don't get policy-induced recessions as we saw in 1980 and 1990.
It's no coincidence that a low serial correlation of inflation - recently and in the 60s - was accompanied by stable growth, whilst positive serial correlation was associated with output volatility. For this reason, some believe more credible monetary policy has been a major cause of recent macroeconomic stability.
Herein lies the case for retaining the inflation target. Abandoning it will change wage and price-setters' expectations, causing them to expect high inflation to persist. The result will not just be higher inflation, but greater economic volatility as - eventually - bigger rises in interest rates will be needed to bring inflation down.
Luckily, though, we don't need to formally abandon the inflation target anyway. The Bank of England's remit (pdf) acknowledges that inflation can depart from its target as a result of "shocks and disturbances", and that rigid targeting in such circumstances "may cause undesireable volatility in output." This is why the penalties the Bank must pay for such deviations - a mere letter of explanation - are so small.
I would expect Mervyn King to draw attention to this clause.
His problem is two-fold. First, it might be that the negative serial correlation of inflation might have been due less to credibility than to dumb luck (pdf) - shocks to prices happen to have been short-lived since the mid-90s - which is running out. Second, policy credibility might be diminishing anyway; the eight-year breakeven inflation rate is 3.45%, implying that the gilt market expects the inflation target to be consistently breached.
Mr King's dilemma, therefore, is that if he relieves the economic pain now by cutting interest rates, he faces a much bigger job in future, because inflation will be harder to bring under control.
There's no simple solution to this.