Kemp: Cheaper Prices Spur US Gasoline Demand

Strong growth in demand for refined fuels such as gasoline and diesel is helping to rebalance the global oil market, according to most observers.

"Recent oil market strength ... stems partly from unexpectedly strong oil demand growth," the International Energy Agency explained in its June monthly oil market report.

In the first quarter of 2015, global oil consumption was 1.7 million barrels per day higher than in the prior-year period.

"In particular, gasoline prices have found support from robust U.S. demand, leading to a surge in crack spreads," according to the IEA.

But if there is a consensus that demand is rising more quickly than predicted at the start of the year, there is fierce debate about whether growth is being driven by economic recovery, cheaper fuel prices or a combination of the two.

And there is a mystery about why the consumption of gasoline is growing more strongly than diesel.

    


I'm Dead Wrong
In one camp are those, myself included, who attribute a big part of the increase in fuel demand to the sharp drop in fuel prices since the middle of 2014.

"With such a major price incentive, there is no surprise motorists are using their cars more for discretionary, non-work-related trips" ("Driving is making a comeback in the U.S.", John Kemp, May 26).

In the other camp are those who insist consumption is largely insensitive to prices and that the increase is being driven by other factors including economic recovery and a rebound in U.S. construction.

"(Kemp) asserts that driving is up because gasoline prices are down. (He) is DEAD WRONG about this, as good macroeconomists will understand," veteran oil analyst Phil Verleger wrote earlier this month ("Notes at the Margin", June 8).

"Had he bothered to look beyond the oil sector, he would have realised that American consumers are sitting on their wallets ... shoppers have been saving rather than spending the benefits from lower oil prices."

Instead "the surge in building explains the growth in gasoline and diesel fuel sales. Construction activity requires much more petroleum than leisure use", Verleger said. "Most construction workers live long distances from their jobs."

Homebuilding is booming, and with it the number of people employed in construction as well as the amount of gasoline and diesel fuel consumed.

"Consumers do not buy gasoline for enjoyment in the way they buy chocolate. They buy gasoline because they must if they want to move ... Gasoline use is up not because consumers are taking more leisure trips but rather because economic activity is burgeoning," Verleger said ("Notes at the Margin", June 15).

             

California Drivers
This is the classic economists' view that fuel demand is insensitive to price changes in the short term and is mostly driven by changes in incomes, employment and economic activity.

In practice, of course, fuel demand is driven by a combination of price and income factors. Economy-related factors are generally more important in the short and medium term. But the sensitivity of fuel consumption to price shifts is not zero.

A very large change in fuel prices over a short period, such as happened between June 2014 and March 2015, can generate a significant response from consumers, employees and businesses.

To understand what has been happening, it is worth starting with some basic data on fuel consumption in California, the largest consumer of gasoline and second-largest consumer of diesel in the United States.

California's gasoline consumption began rising in mid-2013 after falling for more than seven years, according to state tax records. But there was a marked acceleration in the volume sold after August 2014, which corresponds to an upturn in traffic volume growth and coincides with the fall in fuel prices (http://link.reuters.com/kyc94w).

Diesel consumption has followed a rather different pattern (http://link.reuters.com/qyc94w). Diesel use peaked two and a bit years later than gasoline in October 2007 and then plunged during the recession of 2008-2010 but has been growing fairly consistently since early 2011 (http://link.reuters.com/nyc94w).

In contrast to gasoline, there has been no noticeable acceleration in diesel consumption since the middle of 2014 even though prices have fallen sharply.

That still leaves unresolved the question of what factors exactly have been driving gasoline and diesel demand and why they have been behaving differently.



Washington Study
State governments take a close interest in consumption because fuel taxes account for a significant share of their revenues.

In 2010, the Washington State Department of Transportation (WSDOT) conducted a detailed review of its econometric models after they failed to predict fuel consumption correctly.

WSDOT's review included a survey of how 17 other states predicted gasoline and diesel sales ("Statewide Fuel Consumption Forecast Models", November 2010).

It found that most states include a mix of fuel prices, fuel efficiency, state population, and economic variables such as personal income and employment in their gasoline demand models.

Separate diesel models are rarer but more often weighted towards measures of economic activity such as state GDP.

"Most states had fuel efficiency, fuel prices or both in their gasoline forecast model," WSDOT wrote.

"The independent variables in diesel consumption forecast models were usually broader economic indicators like real national gross domestic product, gross state product, real personal income and population."

In principle, the factors that determine the amount of fuel consumed per capita are straightforward: the number of vehicles on the roads, the distance driven per gallon by vehicles on the roads, the number of trips, and the distance travelled on each trip. But "identifying key factors that can be used in an econometric model can be complex," WSDOT admitted.

"Some of these factors can be hard to predict (e.g. cultural shifts in driving habits, gasoline prices, fuel efficiency) while some factors are random (e.g. weather conditions)."


             

Full Demand Factors
In updating the state's demand models, WSDOT's technical working group of economists evaluated a range of variables.

Factors they considered included changes in real and nominal fuel prices; motor vehicle registrations; non-farm employment; personal income; wages and salaries; total and driver-aged population; consumer confidence; industrial production; U.S. imports and exports; national demand for petroleum products; and business income and profits.

Most of these variables were rejected because they did not improve the predictive accuracy of the models or made them worse.

WSDOT found the best gasoline consumption model used changes in state non-farm employment, state population and a composite of gasoline prices and fuel efficiency.

The best diesel consumption model used state employment in the trade, transportation and utilities sectors, plus state real personal income.

In the gasoline model, state population and changes in non-farm employment have by far the biggest impact on consumption.

"The composite variable of gasoline prices times fuel efficiency is not as important in the forecast model as the other two drivers ... but it is still negatively correlated with gasoline consumption and statistically significant."

By contrast, real and nominal fuel prices play no role in the diesel forecasting model.

Now we are getting closer to a model of fuel demand that explains what has been happening in the United States, specifically the acceleration of gasoline demand since mid-2014 and the far weaker performance of diesel.

While the correlation between gasoline prices and consumption is weak, a very large shift in gasoline prices, such as experienced since the middle of 2014, is enough to trigger a significant acceleration in demand.

Economic expansion and employment gains have been driving a broad-based upturn in gasoline and diesel consumption for the last two years in the case of gasoline and four years in the case of diesel.

The sharp drop in fuel prices since the middle of 2014 has given an extra lift to gasoline consumption associated as it is with consumers, but not diesel fuel, which is more closely tied to general business activity.

(By John Kemp)