High oil prices are creating a fissure in U.S. fuel consumption. A recent Barclays research report, utilizing its proprietary credit card transaction data, reveals early signs of contraction in American consumer fuel demand—both the frequency of refueling and the volume per refill are declining. The suppressive effect of elevated prices on end-consumption is gradually becoming apparent.
According to an analysis based on Barclaycard data, the total gallons of fuel purchased by U.S. consumers on a trailing 30-day basis fell 8% year-over-year. Although total fuel expenditure increased by 13%, driven by a 23% year-over-year rise in the average fuel price, the contraction in consumption volume can no longer be masked by price factors. Based on the AAA's national average for regular unleaded gasoline, the current year-over-year price increase has reached 27%, higher than the 23% seen over the past 30 days, suggesting further downward pressure on consumption volumes ahead.
Barclays analysts noted that the aforementioned trend emerged almost immediately in the first week following the outbreak of the Iran war, when oil prices began to surge. By the fourth week of the conflict, total fuel consumption had officially turned negative year-over-year. This signal of demand contraction, revealed by high-frequency credit card data, holds significant reference value for assessing commodity trends and consumer resilience.
It is noteworthy that official weekly data from the U.S. Energy Information Administration (EIA) for the period ending March 27th, on a four-week rolling basis, still indicated a 1% year-over-year increase in gasoline demand, unchanged from the initial period following the war's outbreak in late February, and not yet reflecting a clear slowdown. The divergence between the two datasets highlights the leading advantage of high-frequency credit card data in capturing demand inflection points.
The decline in fuel consumption is driven by two concurrent factors: a decrease in refueling frequency and a reduction in the amount purchased per refill.
Regarding refueling frequency, the number of refuels per user over the past 30 days decreased by approximately 1% year-over-year. While the decline is modest, a clear downward trend has been evident since the first week of the war. Historically, American consumers refuel about 3.5 times per month, roughly equivalent to once every week to week and a half.
The change in the amount purchased per refill is more striking. The implied volume per transaction has decreased from a historically stable level of around 11 gallons to approximately 10 gallons, a drop of 7% year-over-year. Barclays notes this is a highly unusual shift in historical data, representing the most direct behavioral response by consumers to high prices—choosing to purchase less fuel per visit to control individual transaction costs, even while travel demand remains relatively inelastic.
An anomalous, temporary increase in refueling frequency occurred early in the conflict. Analysts suggest this may have been due to some consumers anticipating further price hikes and choosing to "fill up" preemptively to avoid higher costs. However, as the conflict persisted, this effect gradually faded, and refueling frequency began a steady decline, turning negative year-over-year. Combined, Barclays believes these two metrics constitute solid early evidence that American consumers are curbing their fuel consumption.
Barclays' commodities research team conducted a quantitative analysis of the price elasticity of U.S. gasoline demand, breaking it down into two components: a "vehicle miles traveled effect" and a "fuel economy effect."
Multi-factor model analysis indicates that for every 10% increase in fuel prices, vehicle miles traveled decrease by approximately 0.25% (an elasticity coefficient of about -0.025). After controlling for the miles traveled variable, every 10% increase in fuel prices additionally suppresses gasoline consumption by about 0.45% (an elasticity coefficient of about -0.045). Combined, the comprehensive price elasticity of U.S. gasoline demand is approximately -0.7% (meaning a 10% price increase leads to a 0.7% demand decrease).
Based on this, the approximate 40% cumulative increase in oil prices since the Iran war began theoretically corresponds to a combined decrease in consumer and industrial gasoline demand of about 3%. However, EIA data through March 27th, on a four-week rolling basis, still shows a 1% year-over-year demand increase, not yet confirming this predicted decline.
Analysts also caution that if high oil prices trigger a broader cooling of economic activity, the actual impact of price elasticity could be amplified, creating a risk that the demand decline may exceed model predictions.
The analysis utilizes Barclays' proprietary Barclaycard credit card data, covering millions of active users and containing billions of transaction records, with historical data spanning over a decade. The analysis focuses on transactions with Merchant Category Codes (MCC) for "Automated Fuel Dispensers" and "Service Stations," encompassing both pay-at-the-pump and pay-inside scenarios.
As credit card transaction data records transaction amounts but not actual gallonage, Barclays derived consumption volumes by dividing total spending by the AAA's average retail fuel price. To filter out seasonal factors—particularly fluctuations during the transition from winter to the summer driving season—the analysis employed a year-over-year comparison.
As a sanity check, Barclays' calculated historical averages—approximately 11 gallons per refill and about 3.5 refuels per month—combined with an average fuel economy of roughly 25 miles per gallon, imply monthly travel of about 975 miles, annualizing to approximately 12,000 miles. This figure is highly consistent with the approximately 13,500 average annual miles reported by the U.S. Federal Highway Administration for 2022, validating the dataset's representativeness and reliability.
Comments