Statistics and liars

(Title: "Where does your gasoline dollar go?")
Anyone who picked up the newspaper on Tuesday would have seen this ad from everyone's whipping post, i.e. the Oil Industry. Unfortunately, they did not help their cause with this sloppy and deceptive piece of work.
The key problem here is: what is the base? The title of the full-page ad seems to suggest that it is the retail price of oil. It is perhaps telling the story that 55% of the price went to raw materials, 26% to various "friction" in the system and 19% went to the government.
Discerning readers would immediately wonder if any of that money ever went to any human being. Further, one might wonder if these companies ever give dividends to shareholders. Or if they re-invest any of the profits. Just to name a few omissions. In assigning each dollar to a use, some dollars became MIA.
Moreover, the most crucial piece of data required to interpret this chart is unavailable. 55% of one million is very different from 55% of one hundred billion, for example.
This chart is one of those that gives statistics an undeserved bad rep.
Reference: API website



Your criticisms are fair but not devastating. Without quoting the likely alternative analysis the criticisms sound like an ill focussed moan about high gasoline prices.
I would be tempted to show the analysis for any European country where >60% of the retail price will go straight to government (and with little complaint as the resulting incentives to lower pollution and dependence on middle eastern oil are worth the pain). See http://www.ukpia.com/Portals/0/Repository/Documents/UKPIA%20understanding%20pump%20prices%202005%20year.pdf
Also from the point of view of retail gasoline, the breakdowns of profit/reinvestment/etc are irrelevant. The input (crude oil) is sold in a competitive market where the only cartel with market power (sometimes) is OPEC and not the oil companies. Sure they make profits, but the price is not high because they make profits, profits are high because the price is high.
Posted by:stephen black | May 04, 2006 at 09:37 AM
I'm deliberately staying away from the politics and focusing only on the statistics.
That said, profits are totally relevant unless they change their base figure. If the base is the retail price, and if the 3 categories together add up to 100%, clearly profits need to be in the equation, otherwise, as I said, some dollars went missing.
Posted by:Kaiser | May 04, 2006 at 12:22 PM
In the big PDF on the API web site, there's a bar under the entire bill that says "Earnings (8.5%)" But there's no legit reason to leave that out of the main graph.
Worse, the bill is distorted! In both your image and the original PDF, the crude oil occupies 59% of the horizontal space and refining only 21%. And the thick bars muddle things further: the visible part of the bill in the refining area is actually narrower than the visible part in the taxes area!
Posted by:Matt L. | May 04, 2006 at 04:17 PM
Good catch, Matt! Now someone needs to explain where that extra 8.5% came from since the other 3 bits already reached 100%
Posted by:Kaiser | May 05, 2006 at 12:27 AM
What do you think of the Oil Poster? Curious if you had any thoughts.
http://www.oilposter.org/posterlarge-x.html
Posted by: | May 09, 2006 at 01:35 AM
I'd love to comment but even the large poster is not large enough to see what is going on. And they are selling posters so I doubt whether I can get a hold of a copy that is legible.
Posted by:Kaiser | May 09, 2006 at 02:42 PM
But of course any neoclassical economist will tell you that normal profit is the return on capital invested, so in this case the profits are an inevitable cost to the consumer. Indeed, some argue that a lack of past profit, and consequent underinvestment in exploration and refining, is why we're facing such a sharp spike; any super profits the oil majors are *currently* making are needed to remedy this, so these too should be seen as an inevitable cost.
I think separating out profits in this case would overly simplify what is a quite complex argument.
Posted by:derrida derider | May 17, 2006 at 02:19 AM
"Statistics and Liars" is one way to say it, but I prefer "Figures don't lie, but Liars can figure."
Posted by:Don Culver | May 17, 2006 at 12:46 PM
JUST AS I HAD ALWAYS SUSPECTED! THERE IS no PROFIT IN THE OIL BUSINESS! EXXON ET AL ARE PHILANTHROPISTS!
Posted by:COLE BANKS | May 17, 2006 at 07:41 PM