Published in the Financial Times on August 23, 2004
As summer nears its end, oil prices are closing in on $50 a barrel; double what they were in nominal terms just over two years ago. For people with long memories, this raises the spectre of the 1970s. Oil prices rose from $4 a barrel in 1973 to $40 in 1980. The result was a rise in inflation to over 18 per cent in the US, and a drop in growth.
Fortunately, policymakers (and economists) learnt an important lesson from that experience, so we can say with confidence it will not happen again. We know that when faced with a persistent increase in oil prices, the most important thing is to raise interest rates - by more than the increase in inflation precipitated by the oil price rise itself.
The best way to think about oil price increases is as a tax on producers and consumers alike. Take the US, which imports roughly half of the 20m barrels of petroleum per day it consumes. At $25 per barrel, Americans are paying foreigners some $90bn for oil imports. Demand is insensitive to price so a doubling of prices does not much change consumption. At $50 a barrel, producing countries receive an additional $90bn a year for the oil they export to the US. That is a tax increase of nearly 1 per cent of gross domestic product that reverses the last cut in personal income tax. An oil price increase creates the possibility of stagflation and is a challenge for monetary policymakers charged with stabilising the economy. They can only change interest rates, and a rise in rates reduces inflation and output. Monetary policymakers have a tool that moves output and inflation in the same direction. But oil price changes move output and inflation in opposite directions. Faced with an increase in petroleum prices, should policymakers lower interest rates to support growth or raise them to fight inflation?
When oil prices rose so dramatically in the 1970s, the policymakers did the wrong thing: they lowered interest rates. There is a consensus that the high US inflation of a quarter-century ago was a direct consequence of the failure of the Federal Reserve to tighten policy in the wake of price increases by the Organisation of Petroleum Exporting Countries in 1974 and 1979. Today we know better. With China, India and south-east Asia booming and Japan recovering, oil demand is likely to continue to grow rapidly. While prices will probably not stay so high, we should plan on their remaining at or above $35 per barrel for some time. Statistical models tell us that when oil prices rise from $25 to $35, US inflation goes up by 0.5 percentage points and GDP falls by 0.2 percentage points (relative to where they would have been).
If nominal interest rates are left unchanged, the rise in inflation lowers real interest rates, resulting in an implicit policy easing. While this will neutralise the output reduction, it raises inflation further and risks a repeat of the 1970s. So interest rates have to go up by at least the forecast rise in inflation - a bare minimum of 50 basis points for each $10 a barrel increase in the price of oil. In the current environment, with the Federal funds rate well below the equilibrium level of 4 per cent or so, the urgency of raising rates is increased. At its current pace the Federal Reserve will not reach a neutral interest rate until late 2005. But unless oil prices recede quickly it should be aiming at least half a percentage point higher. That means picking up the pace swiftly.
Federal Reserve policymakers have made their lives difficult by refusing to announce a long-term inflation objective. Inflation tends to be persistent - when it goes up it stays high and when it goes down it stays low. One of the reasons is that higher inflation today leads to larger wage increases tomorrow, which leads to higher prices - and so on. The result is higher inflation expectations, which are self-fulfilling. A long-term inflation objective can short-circuit this mechanism, anchoring long-term inflation expectations and reducing the danger that oil price increases will get built into the inflation process. When they can count on inflation expectations staying low, central bankers have more room to respond to the damage oil shocks do to growth because everyone believes they are serious about fighting inflation. Inflation-targeting central banks can be more measured in their response to oil price rises.
The textbook response of raising interest rates assumes there is nothing anyone can do about oil prices. This may be true, but it need not be. The alternative is to use the US Strategic Petroleum Reserve, which currently holds 665m barrels of oil, to stabilise prices. The rationale for the SPR is to prevent oil supply disruptions from damaging the US economy, presumably through dramatically increased prices. Why not create an independent SPR administrative board, similar to the Federal Reserve Board, whose role is actively to stabilise the price of petroleum? The SPR Board would announce a low price, below which they would purchase petroleum, and a high price, above which they would sell it. The range could be wide - say $10 per barrel - but significant imbalances would force adjustments. The requirement for physical delivery would prevent speculative attacks of the type that plague fixed exchange rate regimes.
Until a mechanism is found to stabilise oil prices, monetary policymakers remain on the front line. Today that means raising interest rates in a weakening economy - an unpopular move. Let us hope the Fed has the guts to do it at its next meeting, six weeks before a presidential election.
Stephen Cecchetti is Professor of Economics at Brandeis University, Massachusetts, and a Research Associate at the National Bureau of Economic Research.
Archive of his articles published in the Financial Times