The author of the LA Times piece I linked to in this earlier post about gasoline thermal expansion, Elizabeth Douglass, was nice enough to send some comments and useful links (for example, NIST discussion of point-of-sale fluid measurement). For a working reporter with deadlines, I consider this pretty classy behavior.
I think her piece was a good roundup of the political players engaged in the issue and the policy events on the table; I can tell from her email and the story that she’s a real shoe-leather reporter. Nevertheless, I’m going to continue to gnaw this bone because it raises a larger journalistic issue Marty Linsky once characterized as the facts versus the truth: she presents facts, but not enough of them, or not the right ones, or not the right way, so the reader comes away without the truth.
To review: the face of the issue is that when gasoline warmer than 60F is sold at retail, as it often is, the customer gets less gasoline by weight (or energy content, or what matters) because the fluid has expanded and the pump measures volume. If it expands X%, the retailer gets to sell ten (60 F standard) gallons and be paid for 10*[(100 + X)/100)].
The policy under debate is to install temperature compensated meters in all gas pumps, so what reads as a gallon sold is always a 60 F gallon, with the same amount of gas in it, even if it’s physically fluffed up by being hot. This would be equivalent to selling gasoline by weight.
What I think Douglass’ piece doesn’t get across, and that I think are essential to a reader’s getting the truth from a piece like this, are the following:
(1) How big could this problem be?
The State of California has begun monitoring delivered temperatures of gasoline for a year. A couple of months’ data are available here, unfortunately not in a tabular form that can be analyzed without a lot of retyping and copying. But the average sale temperature in May was 70 F, just over 5 C above the standard. The volume coefficient of thermal expansion of gasoline is 0.00096/C, so last month, California residents were dinged a half a percent (some more, some less: readings seem to range from 55 to 75 F). Half a percent on a summer’s gas purchases is about five bucks per car. I think if a reader knew this number, a number easily accessible to a reporter, she would not regard this as a big issue from the get-go. (In the winter, in many places, the gas will be delivered colder than 60 F, so customers are getting a dribble more for their money.) Percentages and per-capita effects are essential to really understanding policy numbers.
Still, for someone filling a big rig with a couple of hundred gallons of diesel, the issue may well have legs; 2c a gallon adds up on eighteen wheels. So a second question arises:
(2) Is the problem real?
The answer to this question goes beyond physics, and needs interpretation. My interpretation is, “if we installed temperature compensating pumps, would people be better off?” I think it’s an easy case that we would not, and that properly reported, this story should make that clear. Douglass mentions (in her email) a conversation with a dealer who says gasoline is priced lower in hot weather because of the volume expansion; this is useful anecdotal evidence for my claim, but not entirely persuasive. I would consider this thought experiment: if temperature-compensating pumps were suddenly installed for this summer, what would happen? If posted prices didn’t change, gasoline (and diesel) retailers would experience an instant half-percent drop in gross revenues, just as though one of every two hundred customers didn’t pay for his fillup. I believe it’s beyond argument that they could not eat this for long, since the retail markup on gasoline is only about 20c per gallon: it’s a 10% hit. So on the whole, consumers would go from paying, say, $3.50 for a hot summer uncompensated gallon to $3.52 for a corrected 60 F gallon, that is, exactly the same per mile or per joule as they were before. When the gas on sale is, say, 50 F, competition for customers would drive retailers to reduce prices by the same 2c, leaving customers again in the same place.
I think finding that nothing will change if gas sales are temperature compensated is equivalent to learning that the problem it “solves” isn’t real.
Markets don’t always work right, not all prices are efficient, etc. etc. But interventions into markets call forth price responses that aren’t random or mysterious: this is a fact about the world that’s as elementary as the physics of gasoline temperature-volume relationships, or whether the Governor gave a speech yesterday and what he said in it. Are gas prices pretty much competitive? Well, either they are or they aren’t. If they aren’t, we have to ask why gas prices aren’t $3.52 already, or maybe $3.54, or maybe $13.54, or why they ever go down. I believe it’s part of reporting the truth, and I don’t mean reporting in an analysis column or op-ed, to tell the reader the nearly incontrovertible fact that if gas prices are set in a market, price changes will erase any advantage to be gained by temperature compensation. Should Douglass come out against, or for, compensated pumps? Of course not. But I think she should, and her editors should hold her to, report the facts that would allow the reader to infer the truth, including facts that don’t always come from an interview with a politician or a gas retailer, but sometimes from a textbook or a general education.