Welcome to a discussion about the upcoming 5th Revolution in the US, which I’ve titled the “Revenge Revolution.” For more about the Revenge Revolution and the author, Entry #1.  Periodically I write a “sense check” to assess whether a revolution in the US is possible or whether the entire exercise is based on a statistical aberration — i.e., a roughly 50-year cycle between major upheavals in the US. Entry #430 was the most recent “sense check.”

Entry #443 raised the question whether the US should link gas-pump prices to spot oil prices. This entry offers some additional information.

First, let me clarify that the presumed link between spot oil prices and pump prices was assumed to be at the macro level, i.e., average national price, and not intended to predict prices at an individual gas station. Obviously, there are numerous factors affecting prices at any individual gas station – rent, competition, geographic location, transportation cost, state taxes, and so forth.

The variables affecting pump prices at a specific station probably account for no more than 20% of the pump price, and likely far less.  If one had data over time of pump prices at a specific station, the correlation to spot prices might parallel the correlation for national prices.

In determining whether national gas-pump prices track spot-oil prices, I was reminded of some guidance conveyed while a graduate student at MIT. One of the professors at the Sloan School was fond of saying, “when in doubt, run a regression.”  The advice has proved valuable over the years, at least to get a handle on possible causes.

Obviously, correlation does not equal causation. At the same time, when the variables appear inextricably linked, a high correlation increases the likelihood of causation.

For the spot-price/pump-price correlation I used two data bases from the Energy Information Administration: (i) spot oil prices for WTI (West Texas Intermediate); (ii) average US gas prices. The data gathered by EIA are reported by week.  Spot oil price was the independent variable. 

To get some idea of the correlation, five (5) different periods were selected, ranging from three (3) months to 15+ months, including prices published during the week of 04/11/2022.  The periods were selected somewhat randomly and might not be the most representative.   Nonetheless, the five periods included different levels of stability and a range of prices. The widest variation in prices occurred during CY2008, when maximum oil price was $142.52, 4.32x higher than the minimum price of $32.98.  The least variation during 1994, when maximum oil prices only varied $3.30/bbl.

The pattern of the correlation was remarkably consistent. The higher the spot price and the greater the volatility in spot prices, the higher the correlation. Only in 1994 when spot prices stayed in a narrow band, and gas prices changed even less, did the correlation drop significantly. Since pump prices remained so stable in 1994, even a small movement in spot prices would reduce the R². 

The correlation did include a two-period lag between spot prices and pump prices as published by EIA.  The data were lagged to allow for some delay gathering and reporting pump prices and for some delay between changes in spot price and changes at the gas pump. The lagging increased R² a few points compared to using EIA data published during the same calendar week.

No adjustment was made to gas prices for changes in state taxes or changes in refining mix. For example, in the upper Midwest, the gasoline “blend” changes in the summer. For many years, and it still might be the case, the capacity to refine that summer blend was limited, so sharp jumps in pump prices were common, even if spot oil prices remained stable or declined slightly. I recall one business trip to northern Indiana in early summer when the pump prices at the same station increased nearly $0.40/gallon between morning and evening.

The correlation, in my view, reinforces the need to address the question, “If the US is so concerned about how fuel-pump prices affect CPI and other costs, then why does the US allow pump prices to be linked to spot-oil prices?” The link seems especially unnecessary since the US has reduced imports and in CY2021 became a net exporter of fossil-fuel products. 

The cost of domestically produced oil does not change whether the spot price for WTI is $100/bbl, $10/bbl or $1/bbl. The cost to refine the oil (WTI) does not change because refining is a processing cost, not a raw material cost.

As noted in the first memo, if the US wants to begin to make fuel prices more predictable for consumers and businesses, a good first step would be for the Biden administration to complete a comprehensive analysis of anticipated crude-oil supply based on different fixed prices.  And, yes, the analysis should review why such efforts have either failed or succeeded in other countries.

With such a large domestic supply and production base, such a study would help bring to light the enormous impact of spot prices on pump prices and how influential a few oil traders are in affecting the disposable incomes of many US citizens.  The decision might be to continue the current approach.  However, having a meaningful debate whether to implement a fixed national price or a hybrid version, having less volatile price for gasoline and diesel (before tax) would help manage inflation, would help sustain economic growth and might help focus members of Congress on attempting to address and solve real problems.

Note: The NY State Attorney General apparently believes the oil companies might be price gouging.  The AG opened an investigation.  A number of media reports 04/14/2022. END ENTRY #445

Other Topics. Interested in more info about climate change, what’s required to electrify a fleet of cars/trucks, what it was like to work day-to-day with Lee Iacocca and an array of other topics? Visit another page of this website, https://usrevolution5.com/jrd-thought-comments/