Ten of the 11 recessions in the United States since World War II have been preceded by an increase in oil prices. Does the recent surge in oil prices mean we should be looking for recession number 12?
I’ve noted before that once energy expenditures get above 6% of average consumer spending, we start to see significant changes in spending patterns. We crossed that threshold in March, when 6.27% of every dollar spent went to energy-related goods and services. For lower-income groups, that expenditure share is significantly larger.
And you don’t need me to tell you that energy prices have surged dramatically since March.
|San Diego Historical Gas Price Charts Provided by GasBuddy.com|
A key sector for determining how the U.S. economy reacts is the auto industry. Let’s take a detailed look at what happened in the spring of 2008, which was the last time we saw gasoline prices like these. Because of strong seasonals, I like to summarize auto sales with graphs like the one below, which reports U.S. sales of imported cars. You can see what happened in an individual year by following a particular color horizontally, and can compare a given month with the same month in other years by looking at different colors. If you follow the dark blue bars corresponding to 2008, you’ll see that surging gasoline prices in the first half of 2008 were a boon for fuel-efficient imports, though these collapsed in the second half of 2008 as the recession became much more severe after the bankruptcy of Lehman in September 2008.
On the other hand, as shown in the figure below, if you were a domestic manufacturer of light trucks (a category that includes larger sports-utility vehicles), those rising gasoline prices in the first half of 2008 were a disaster. There was a slight rebound in sales in August of 2008 when gas prices came down, but then the financial crisis delivered a second hammer blow to the industry.
The struggling domestic auto industry was an important factor in the first year of the recession, which the National Bureau of Economic Research (not to mention Econbrowser as well) characterizes as having begun in 2007:Q4. I noted in a recent paper that if domestic U.S. auto production had not fallen, U.S. real GDP would have grown by 1.2% over 2007:Q4-2008:Q3.
The relevant concern at the moment is of course the light blue bars in the graph above, which report the data so far for 2011. It’s interesting to note that sales of domestically manufactured light trucks in April 2011 are not far from where they were in April 2008. But whereas in 2008, that level represented a significant decrease from previous levels, and thus contributed mechanically to a negative GDP growth rate, today they represent a significant increase from previous levels, and thus contribute mechanically to a positive GDP growth rate.
This discussion illustrates a principle that applies to the economy more broadly. Higher oil prices by themselves put a burden on consumers and force them to cut back spending somewhere. But getting unemployed people back to work is a much bigger positive factor in overall spending. The momentum of the ongoing recovery is an important factor that is serving at the moment to offset the contractionary effect of higher oil prices.
I’m also suggesting that, precisely because consumers never went back to the earlier spending patterns on items like bigger cars, $4 gasoline will not have the same disruptive effects now that it did when we experienced these prices for the first time in 2008. In a study I published in 2009, I noted that one could account for much of the overall decline in U.S. real GDP in the first year of the recession (that is, 2007:Q4-2008:Q3, prior to Lehman) on the basis of a relation I first published in 2003 (working paper version here). That simulation was based on a very nonlinear relation. The good news is that, high though current prices are, they actually would not trigger a response according to that relation because we are still below the highs reached less than 3 years ago.
I recently surveyed the academic literature on how strong the evidence is for this kind of nonlinearity. I personally find the evidence to be pretty convincing. However, a range of alternative nonlinear specifications fits the data equally well, many of which would begin to predict effects from the recent oil price run-up. My judgment is that, contrary to my 2003 model, there will be some contractionary consequences of recent developments, but, consistent with that model and many others, they will not be as severe as what we experienced in 2008.