It is interesting to compare the policy environment now with the situ- ation in the late 1960s and early 1970s. At that time policy was still geared to the postwar Keynesian consensus view that active monetary and fiscal policy should focus primarily on avoiding high unemploy- ment. However, inflation was gradually accelerating and a huge policy debate opened up between “Keynesians” and “Monetarists.” The initial Keynesian solution was to combat inflation with income policies, with- out changing the rest of the approach. But incomes policies were dis- mal failures in most countries and had to be abandoned. Gradually, the old Keynesian consensus gave way to the current orthodoxy, which sees the level of unemployment as determined by structural factors over the long run and focuses monetary policy entirely on controlling inflation.
Now we face a new environment. While ordinary inflation has faded as a problem, asset price volatility and financial instability have emerged as major worries.2Yet conventional policy is still rooted in the anti-inflation orthodoxy of the last couple of decades, which takes very limited account of asset prices. Is it time for a new approach? In par- ticular, is it time that central banks leaned harder against emerging bubbles?
Asset prices affect inflation and growth in various ways. If asset prices are high and rising, economic growth is likely to be well sup- ported, which will eventually, when the job market becomes tight or capacity utilization rates rise particularly high, lead to higher inflation.
There is therefore a clear case for higher interest rates when asset prices are surging and the economy is strong, even though current inflation may be well behaved. Similarly, if asset prices are falling and the econ- omy is weak, cutting interest rates is the best antidote to both and a recovery in asset prices will help reinforce the economic recovery and vice versa. In these cases, then, there is no conflict between monetary policy aimed at managing inflation and policy aimed at controlling asset prices.
However, suppose inflation is at or below target and the economy is seen as either relatively weak or as growing satisfactorily, while asset prices are rising strongly. Then, conventional policy suggests that the central bank should ignore asset prices because raising interest rates might hurt the economy. This was the situation for the US economy in much of the 1990s.
In the 1990s the Fed almost certainly held interest rates too low over- all. It is easy to be critical after the event, but many commentators argued the same at the time. The combination of a strong economy and soaring stock prices certainly pointed to the need for a more restrictive stance.3Moreover, the apparent increase in the trend rate of growth of the economy, from around 2.5 percent to 3.5 percent per annum, also pointed that way because of the implication that the “neutral” rate of interest should be higher.
A simple measure of the stance of policy is to subtract the growth of nominal GDP (including real growth and inflation) from the Federal Funds rate (see Chart 11.1). When the funds rate is below nominal GDP
growth, we can say that policy is easy; when it is above, policy is tight.
Overall, Fed policy was generally looser in the 1990s than in the 1980s.
We can also see that policy since 2001 has been much looser than in the early 1990s. But I would focus the main criticism on policy in the period 1997–99.
The Fed’s overstimulatory policy at that time was partly a response to two shocks. One was the near bankruptcy of the hedge fund Long Term Capital Management (LTCM) following the Russian default, in the autumn of 1998. This led to large-scale unwinding of bond posi- tions and turbulence in the financial markets that threatened to hurt the economy, so the Fed was quite right to respond swiftly. But it was then slow to tighten monetary policy again in 1999, apparently because of fears of a shock from the Millennium Bug at the end of the year.
The risk of major problems with computers on January 1st, 2000 meant that the Fed had to be particularly careful to ensure ample liq- uidity for banks at the year end. But it has never been satisfactorily Chart 11.1
Fed funds rate less nominal GDP growth
Source: DATASTREAM
explained why the Fed felt it needed to dawdle in raising interest rates during 1999. The two issues are separate when the liquidity needs were only likely for a limited period. But the Federal funds rate was moved up only very cautiously in 1999, from 4.75 percent to 5.5 percent. This lapse encouraged a new inflation of the stocks bubble, with the S&P 500 index rising from 1,000 points after the slump in the autumn of 1998 to over 1,500 points in just 15 months.
The dilemma for the central bankers is that if a bubble grows large and then bursts, it may bring a serious recession, making it very likely that they will undershoot their inflation target in two or three years’
time. But if they raise interest rates to stem the inflation of the bubble, unless the economy is strong they may kill the bubble and end up trig- gering the very problem they were trying to avoid. And of course, if the economy is indeed strong, higher interest rates may have little effect on the bubble at first, until suddenly there is a sharp swing in expecta- tions. It is always hard to deflate a bubble gradually.
Central bankers are well aware of the problem and some are arguing that it may be desirable to “lean” more heavily against bubbles than Mr.
Greenspan attempted to do in the 1990s. Sometimes this argument is couched, neatly within the prevailing orthodoxy, as a need not just to target inflation one or two years ahead, but to take a longer view.4This flexibility sounds like a step in the right direction, though it still leaves the same dilemma of what exactly is the right level of interest rates and how central bankers should explain why they want to raise them when inflation is still low. It also raises some disquiet over the balance of
“rules versus discretion” for monetary policy. Rules that are simple, like
“target 2 percent inflation in 24 months’ time,” are much more attrac- tive than vague and potentially conflicting longer-term approaches.
One way to keep a strictly rules-based approach would be to include asset prices in the targeted measure of inflation.5 In principle, this would remove the conflict by giving central bankers one target to hit with their one available instrument, interest rates. However, it is diffi- cult to know which assets to include and what weight to give them.
Moreover, the behavior of asset prices can be fundamentally different from the behavior of consumer prices, particularly during bubbles. So the idea has not gained much traction.