In this guest post, economist Paul Segal argues that the now popular view that oil price spikes lead to recessions is largely unfounded.

Do high oil prices cause recessions? The US economist James Hamilton is famous for his 1983 finding that oil price spikes had preceded all but one post-war US recessions[1]. Hamilton recently claimed that the current recession can be fully accounted for by the high oil prices of 2007-08. But while oil prices are certainly an important macroeconomic variable, it is just not plausible that they have anything like the impact that Hamilton suggests.

Oil prices have a direct impact on output only to the extent that they lead to lower consumption of oil. In a standard competitive model of the economy, the decline in output is equal to the share of oil in GDP times the decline in consumption of oil. From a peak of 7.6bn barrels in 2005, US oil consumption declined by 6 per cent to 7.1bn in 2008. With oil consumption comprising 5 per cent of GDP in 2008, this can account for a decline of only 0.3 percent of GDP – which, alone, is not nearly enough to cause a recession. The largest decline in oil usage that has ever occurred in the US was over 1979-80, for which the same calculation implies a decline in GDP of 0.6 percentage points – again, not nearly enough, on its own, to explain the US recession of 1980.

A cottage industry has developed around the effort to find microeconomic mechanisms such as market frictions that could explain a larger role for oil than this standard model implies. Oil has a romantic place in the US imagination and some US economists cannot quite believe that it is just another commodity. But my own view[2] is that these mechanisms can explain only a small additional impact for the oil price.

Morgan Downey has found a nice correlation between Google searches on the term ‘hurricane’ and actual hurricanes. He’s made this chart:

Whether it’s predictive or not is yet to be seen: as the National Oceanic and Atmospheric Administration points out, the last few years have skewed towards seeing storms earlier in the season – June and July – whereas typically August to October are the worst months.  Last year there were four storms by the end of July, and in 2005 – the year of Katrina – there were seven by that date. This year, Mexico is certainly not out of the woods.

Related links:

On the predictability of search trends (Official Google Research Blog, 17/08/09)
Speculators driving up prices? Use the SPR! (FT Energy Source, 28/08/09)

Opec’s next meeting takes place next Wednesday in Vienna (and a decision will come late, because of Ramadan).

Trying to predict an Opec decision based on comments from member countries in the lead-up is notoriously difficult. But for those who like guessing, Reuters has a handy list of what’s been said to their reporters in the last couple of weeks:

AUG 31
Bulging global oil inventories remain a major concern for OPEC member states, even as signs of a possible global economic recovery grow, and will probably be be a topic at the meeting, Iran’s OPEC governor said.
“Everyone is worried about the current stock levels… for both OECD countries and floating storage,” Mohammad Ali Khatibi told Reuters in a telephone interview. “It may be considered by ministers a policy within one year, maybe a lower or higher period, to bring stock levels back down to normal.”