Causes and Consequences of the Oil Shock of 2004

James D. Hamilton

University of California, San Diego

August 11, 2004


All but one of the recessions since World War II have been preceded by a dramatic increase in crude petroleum prices.  Recent turbulence in energy markets has some analysts speculating that, in the immortal words of Yogi Berra, it could be déjà vu all over again.  However, it is my opinion that, if prices increase no further than they already have, the oil price spike of 2004 will slow GDP growth by about one percent, but is not enough to derail the ongoing robust economic recovery.

The current behavior of oil prices is unlike the spike that preceded earlier recessions in two key respects.  First, oil prices have gone up to a large degree not because of a shortfall of supply but rather because of an increase in demand.  The world is producing 3 million more barrels of oil each day relative to last year, nearly a 4% increase.  But demand is up even more dramatically.  China alone is consuming over 1 million more barrels of oil each day than it did in 2000, which is double the increased demand coming from the United States over that same period.

This is quite a different situation from other historical oil shocks that were caused by military conflicts that physically disrupted the production or delivery of petroleum, forcing consumers and firms to make less use of this vital input.  The current situation, to a large extent, is simply that we have to share the increased supply with other consuming nations.  There should be no quarrel with the proposition that a booming world economy overall is good economic news, not bad.

To be sure, concerns about supply have also made a contribution to the recent run-up in prices.  Iraqi production is not as strong as was expected at this point, and even maintaining current levels could prove problematic.  The market is sufficiently tight that any new supply concerns, such as the ongoing negotiations over Russian production, quickly show up in higher prices.  Production worldwide has dropped slightly over the last few months, and this would certainly be a source of  much concern if that trend continues.

The second way that the current oil price spike differs from those that preceded earlier U.S. recessions is that a good part of the recent increase is merely a correction to an earlier dramatic drop in oil prices.  The current oil price of $45 a barrel is 47% higher (logarithmically) than the $28 price we saw last September.  However, it is important to remember that before those September lows, oil had been selling for $36 back in February of 2003, so that the current price is only 22% (logarithmically) above what we saw just a little over a year ago.  There were similar corrections (an oil price spike following an earlier downturn) in 1987 and 1994 with no apparently adverse economic effects.  By contrast, those episodes that were followed by recessions were invariably associated with dramatic new highs, not simply a correction to an earlier decrease.  Using a formula described in my 2003 paper in the Journal of Econometrics, this 22% net increase would lead us to predict 1.1% slower growth of GDP for the second quarter of 2005; [see equation (3.8)].

There are some other dimensions of the current situation that aggravate the problem and warrant policy changes in the United States.  The U.S. has only 150 refineries, down from 250 a quarter century ago.  This is compounded by the differing gasoline standards across U.S. communities.  For example, the additive MTBE has been selectively mandated in certain communities and then banned outright in others.  The reduction in the number of refineries and the proliferation of legally required gasoline formulations raise costs and greatly reduce the competitiveness of individual markets, making the gasoline price mark-up over crude costs unacceptably vulnerable to small supply disruptions.  Easing the rules for new refinery construction, as proposed in the energy plan that President Bush submitted to Congress in 2001 but on which the U.S. Senate has yet to act, and agreeing on a single nation-wide standard for gasoline formulation, are clearly important steps that need to be taken.

A second on-going concern is the possibility that military or civil conflict could disrupt petroleum production or shipments from Iraq or Saudi Arabia.  Alternatively, terrorists may target oil refineries anywhere in the world.  Given the current oil supply-demand balance, that would easily turn the current oil price spike into a major economic concern.

However, the oil futures markets are not betting very heavily on such pessimistic scenarios—the current price of oil futures contracts shows a steady decline for the next year and beyond.  Assuming that the market is right on this one, we will see a slowdown of U.S. GDP growth rates, but no recession in 2005.