Stephen King: Central bankers need to be radical in our post-bubble world

The problem lies in persuading the public that central bankers really can change their spots

How do you convert conservative central bankers into radical policy makers? This is no idle question. Most central bakers earned their spurs establishing reputations as inflation fighters. But how relevant is an inflation fighter in a post-bubble environment? What can an inflation fighter do to help when the biggest risks facing an economy are high debt levels, stretched balance sheets, deflation and potential failure of traditional monetary policies?

Fortunately, some central bankers are already on the case. Ben Bernanke, a member of the US Federal Reserve's interest rate setting committee, has already outlined some of the things that central bankers could do in a world where short-term interest rates hit zero (see http://www.federalreserve.gov/ boarddocs/speeches/2002/20021121/ default.htm). Chief among these is resorting to the monetary printing press. The more money that is swishing around, the more likely it is that people will spend rather than save, thereby lifting income growth and reducing the burden of existing levels of debt.

All well and good. The problem, however, lies in persuading the public that central bankers really can change their spots. Will people really believe that central bankers can turn to the printing press? After all, the current crop of central bankers has been telling the public over the last 30 years that the printing press is the ultimate source of economic ruin. So why should anyone believe that central bankers are capable of performing a complete 180-degree turn?

One way to force a change in perceptions would be for central bankers to replace their inflation targets with targets for the price level. An inflation target is simply an ambition for inflation in the future. It is largely indifferent to what has happened to inflation in the past. Yet past inflation – or, in current circumstances, the lack of it – can lead to significant effects on real debt levels. A price-level target takes these effects into account because it is both forward and backward looking. Policy has to be changed not just on the basis of future expectations but also on the basis of past errors.

A simple numerical example easily illustrates the difference between the two approaches. Assume that a central bank has a choice between an inflation target of, let's say, 2.5 per cent per year or, alternatively, a price-level target. Let's say that the price-level target is set 10 years ahead and is based on the idea of a 2.5 per cent inflation rate per year. If the price level at the beginning of the period is equal to 100, this would give a price level target 10 years down the road of 128.

In the first year, actual inflation comes in at 2.5 per cent. On that basis, the central bank is on track to meet either its inflation target or its price-level objective. In the second year, however, inflation drops to zero. Under the inflation targeting approach, the central bank simply has to restore the inflation target to 2.5 per cent in the following year.

The price-level approach, however, requires the central bank to go further. By undershooting the implied inflation rate in the second year, the central bank will be obliged to overshoot the inflation target in each of the subsequent eight years. In this example, the inflation rate for the remaining eight years would have to rise to 2.8 per cent to hit the price-level target. If prices refused to budge in the third year, the central bank would have to throw caution to the wind even more, having to achieve inflation of 3.2 per cent per year in the remaining seven years.

It might seem to be rather bizarre to be actively encouraging higher inflation. There is, however, a very good reason to be worrying about the impact of lower than expected inflation – or possibly deflation – in a world of high debt levels. When a bubble bursts, people are typically left with excessively high levels of debt that they either want to repay or are forced to repay. This process often leads to a prolonged period of weak demand which, in turn, depresses inflation. If the inflation rate is lower than expected – or, worse, still, turns into deflation – people's real debt burdens will become even bigger, forcing a further round of retrenchment which depresses the price level still further.

In this world, a permanent shift downward in the price level implies a permanent increase in the real debt burden. An inflation target does little to change this state of affairs. A price-level target explicitly forces central banks to take action against this potentially nasty vicious circle. The further the price level falls short of target in the short term, the more the central bank is obliged to print money to rectify the situation.

By pre-announcing this policy, there is a better chance of avoiding a debt-deflation downward spiral in the first place. If you know that the central bank is committed to preventing real debt burdens from rising unexpectedly in real terms, there will be less need to take an axe to debt in the short term, thereby reducing the chances of a shortfall in demand.

This week's charts provide a real life example of failure to embrace this particular approach. The left-hand chart shows the performance of Japanese inflation since the mid-1980s. The right-hand chart shows the growth rate of money supply in Japan, focusing on both narrow and broad monetary measures. Japan started to show signs of inflationary undershoot in the early 1990s and outright deflation arrived in 1995 (the later temporary rise in the inflation rate was simply the result of tax changes).

It was not until 2002, however, that the Bank of Japan started aggressively to pump liquidity into the system – reflected in the surge in narrow money growth – but, by that stage, it may already have been too late: the banking system had become so weak after years of sluggish demand and deflation that it was unable to support a multiple expansion in bank lending. Had a price-level target been adopted at the beginning of the 1990s, the Bank of Japan would have been forced to print money much earlier, potentially giving a much more successful stimulus to economic activity.

Price-level targets are useful not only during periods of deflation or excessive debt. They're also potentially attractive during periods of rapid asset-price inflation. Too often in the past, central banks have been happy to countenance a period of asset-price inflation on the basis that there is no immediate threat to the inflation objective. However, to the extent that an asset-price bubble may burst and, therefore, lead to subsequent inflationary undershoots, there is still a case for action. A longer-term price-level regime would make it that much easier for central banks to quell asset-price inflations at an early stage.

Of course, no policy framework is ever going to be perfect for the simple reason that economic challenges change from decade to decade. An inflation-targeting regime, for example, may be at its best when inflation is too high in the first place: it provides a clear and transparent target for both central banks and for the public at large.

A price-level target would inevitably increase uncertainty about the path for inflation in the short term. But, in a low-inflation world whereby the biggest single threat is an unintended – and undesirable – increase in the real debt burden, a price-level regime could provide the kinds of guarantees that would force conservative policy makers to take a radical plunge. And perhaps the time is ripe for radical thoughts: Paul O'Neill's departure from the US Treasury suggests that the political establishment may become increasingly fed up with policy makers who are unable to deliver the answers in our new, post-bubble world.

Stephen King is managing director of economics at HSBC.

stephen.king@hsbcib.com