Stephen King: Prospect of low inflation puts Fed in a delicate predicament

If inflation falls after interest rates have come down close to zero, policy makers have a serious problem

"The probability of an unwelcome substantial fall in inflation, though minor, exceeds that of a pickup in inflation from its already low level."

"The probability of an unwelcome substantial fall in inflation, though minor, exceeds that of a pickup in inflation from its already low level."

Well, who would have thought it? The Us Federal Reserve has finally admitted that inflation can be too low. The Fed's policy statement last week suggests that inflation – rather than growth – is top of the policy agenda. But, unlike the 1970s and 1980s, the Federal Reserve will now be aiming to keep inflation up: America's central bank needs to find the monetary policy equivalent of Viagra.

Why should anyone worry about inflation being too low? There are two obvious reasons.

First, low inflation might simply be a symptom of inadequate demand. If inflation is too low, it's likely to be associated with depressed profits and high unemployment. Neither of these is desirable, thereby suggesting that the central bank should choose to boost demand through lower interest rates. This is the standard cyclical concern.

Second, low inflation could be a cause of inadequate demand. Let's say, for example, that you've borrowed £100. You did this on the expectation that inflation in the first year would be, say, 2.5 per cent. But suppose that actual inflation is zero. Under these circumstances, the real, inflation-adjusted, debt level at the end of the first year is 2.5 per cent higher than you thought it was going to be. You might choose to react to this either by borrowing less in the future or repaying the existing debt more promptly. Either way, you choose to spend less. Repeat this at the national level and you suddenly discover that there's a major shortfall in the level of demand which prompts a further inflationary undershoot. This is a less familiar structural concern.

Of course, the central bank can offset this second problem in normal circumstances by lowering the interest rate. If the interest rate is also lower than expected, the debtor is back to square one. Yes, the real debt level is higher but the cost of servicing that debt – the interest payment – is lower. The two factors cancel each other out.

Or at least that would be true if interest rates were always able to fall. At very low interest rates, however, this might not be possible. If inflation is 2.5 per cent lower than expected and interest rates start off at 5 per cent, the central bank merely has to cut interest rates by 2.5 per cent. If, however, inflation is 2.5 per cent lower than expected and interest rates start off at 1.25 per cent – which is where they are in the US today – then it's simply impossible for the central bank to provide the necessary offset: nominal interest rates cannot fall below zero.

Because of this, the Federal Reserve finds itself in a bit of a quandary. America's central bank has spent the last two years trying to bring about a sustained recovery in economic growth. To a degree, it has been successful: consumer spending had, until recently, been quite buoyant and, in 2002, economic growth surprised on the upside. Despite all this, inflation – excluding the volatile food and energy components – has continued to ebb away, suggesting that, despite the pick up in economic activity in the first half of 2002, there is now an increasing risk of a downward debt-driven spiral. If inflation persistently surprises on the downside, real debt levels will persistently surprise on the upside. But if interest rates are already very low – as they are today – the Federal Reserve may no longer be in a position to bail out the economy via more and more interest rate cuts.

Indeed, it may be the case that the attempts to prevent deflation have, in reality, made the situation even worse. To understand this point, it's worth going back to the arguments of the Austrian economists who, in the 1920s, were all the rage but were then quietly forgotten about – apart from the free market revival that took place in the Thatcherite 1980s.

The Austrian view – proposed by Friedrich Hayek – was quite straightforward. The Austrians made three general points.

First, during bubbles, interest rates are too low because the central bank will not allow them to rise to a "market clearing" level (in America, the justification for keeping rates low in the late 1990s was partly based on the Asian crisis, Long Term Capital Management scandal and the Y2K fever).

Second, the low level of interest rates will give companies the impression that consumers prefer to save rather than spend (after all, the greater the supply of savings the lower the interest rate), thus companies should invest because it will appear that future demand will be higher than current demand.

Third, consumers themselves will take one look at the low level of interest rates and conclude that there is absolutely no point in saving from current income because returns are insignificant. So they spend. The net result is that, in the boom, there is too much demand and, at the point when the investment comes on stream, there is too little demand. Seven years of famine then follows seven years of feast.

A non-Austrian central bank will look at this situation and see a standard cyclical shortfall in demand. Concerned about this shortfall, it will decide to lower interest rates still further. Consumers carry on spending and, hopefully, companies carry on investing. But this simply defers, rather than cures, the problem. Indeed, the policy potentially makes the problem bigger: consumption will only remain strong if consumers can be persuaded to carry on borrowing. But this is very much a finite process. Either nominal incomes will be hit, thereby reducing the desire to borrow, or interest rates will fall to zero and, therefore, be unable to decline any further. In both cases, the consumer will eventually hit a brick wall, increasing the risks of a shortfall in demand and an undesirable further decline in inflation.

Could the Federal Reserve now be facing this Austrian problem? Two years ago, it cut interest rates with the intention of stimulating an economic recovery. But if interest rates were too low during the period of the boom, it becomes difficult to see why even lower interest rates can be part of any subsequent solution. Indeed, by creating more debt, the danger is that the ultimate deflationary impetus becomes even more unpleasant.

Ultimately, this is an issue about prevention and cure. Interest rate cuts are seen as a way of preventing deflation. But if inflation is still falling after interest rates have come down to something close to zero, policy makers have got a serious problem. The irony is that the policy used to prevent deflation – getting people to borrow on the back of low interest rates – becomes the very worst policy to use if deflation arrives in any case.

So, should we worry? The good news is that, unlike the Bank of Japan in the first half of the 1990s, the Federal Reserve has been quick to recognise the possibility of deflation and, therefore, should stand a better chance of coming up with either prevention or cure. The potential bad news is that the prevention polices pursued by the Federal Reserve are, according to Austrian arguments, doomed to failure. Low interest rates only make the problem worse: ultimately, time alone will provide a resolution to America's economic difficulties. On this basis, policy makers should aim only to minimise the pain of adjustment, not pretend that the pain can be removed altogether.

Stephen King is managing director of economics at HSBC.

stephen.king@hsbcib.com