Gasoline Prices are NOT Proportional to Oil Prices

Several weeks ago, a friend asked me why, when oil prices had fallen to a third of what they were a year ago, gasoline prices had fallen by less than half of what they were then. And the other day, I heard some talking-head mediot* ask,

"So tell me this: if oil has fallen from $140/barrel to $50, why have gasoline prices only fallen from $1.29/litre to $.79/litre? Huh?"

Here is a graph of a few of the observed data points to which they were referring:

[Update: several people were confused by the original graph, so here is a better-labeled one. It is a scatter plot of the price of gasoline versus the price of oil at three different points in time:]

Applying the famous "Palmer Eyeball Algorithm" to these data and completely ignoring the lack of degrees of freedom, it appears that gasoline in Ontario is roughly priced at about $.55  x [1+  (price of oil)/100]. In other words, about 55 cents of the price of a litre of gasoline is due to other costs that do not vary with the price of oil. These costs include, of course, taxes, but they also include the bulk of labour and capital costs as well. These costs cannot be altered much as the price of oil changes, and hence their influence on the price gasoline cannot change much when oil prices change.

It is nearly a linear relationship, but because of the comparatively fixed costs, the relationship is not proportional.

Finally, did you ever notice that the people who complain when gasoline prices don't fall in proportion to a fall in oil prices do not express gratitude when prices rise less than proportionally with oil prices?

*mediot: a contraction of the words "media" and "idiot".

Comments

[Mike Moffatt]

There's also one other major factor the media always leaves out - exchange rates. The oil prices in the news are, almost always, presented in U.S. dollars. But gasoline prices are reported in Canadian dollars.

When oil was $140 US/barrel, the Canadian dollar was at or above parity, so a barrel of oil was around $135-140 Cdn/barrel.

Today Nymex crude futures are at $50.24 US/barrel. At today's exchange rates that equals around $66.13 Cdn/barrel as the Canadian dollar has fallen (largely because of falling oil prices).

A Canadian saying oil is at $50 without stating currencies is hugely misleading, as the price of oil is closer to $80 than it is to $50 when priced in Canadian dollars.

[M. Cooper]

Retail price of each litre would be less likely to inflame consumers if the profits reported by the producers were relatively stable or even moderately growing. However, we are told that record producer income has occurred at the same time.

[EclectEcon]

Actually, M. Cooper, oil profits have plummeted over the past year, as have the prices of their shares. But if you want to get in on those profits for the next oil price increase, you can consider buying oil stock (or a mutual fund that deals in oil stocks) now and take a chance.

But if you don't want to take that chance, then you have to admit that a good chunk of the oil profits are a return to taking risks that others are not so willing to take.

Mike: What exchange rate are you using to say that $50US is close to $80 Cdn these days. I get something in the range of $60 - $63 Cdn.

[Mike Moffatt]

"Mike: What exchange rate are you using to say that $50US is close to $80 Cdn these days. I get something in the range of $60 - $63 Cdn."

I get $66.. which is closer to $80 than it is to $50. :)

[Tom Hanna]

Some of the frustration is that gas prices do jump immediately on oil prices rising, but are much stickier when oil prices fall. There may be good reasons for it, but it's still perfectly legitimate for customers to ask the companies why they're charging what they are and to try to modify their own shopping accordingly and the appearance of impropriety is always news. It's also a PR "problem" that could probably be dealt with by using better inventory controls and, more importantly, a potential marketing bonanza for the company that figures out how to solve it.

[EclectEcon]

An interesting empirical question, Tom. Do prices move more quickly in response to oil price changes on the upside or the downside? My own casual empiricism suggests there is no significant difference, but my experience may differ from yours.