MBM talks a lot about collecting and sharing knowledge. However, we sometimes don’t stress the need to understand exactly what that knowledge means. Often what appears a clear connection is really quite meaningless. I was reminded of this by an article in today’s Arizona Republic (HT: Coyoteblog). The article complains that consumers are being “gouged” by faulty gas pumps. Apparently, about 9 percent of Arizona’s inspected gas pumps failed tests by the State’s Department of Weights and Measures. The article is full of innuendo and numbers implying there is something rotten in the state of Denmark (or at least Arizona). Despite the tone of the article, the fourth sentence begs some curiosity:
Half of those [pumps] were malfunctioning to the detriment of customers.
Faulty pumps are not good, and people getting ripped off is not good; but if half of the malfunctions cause detriment to customers, then what about the other half? Are some of the good people of Arizona getting a little free gasoline? Of course, I do not doubt the need for governments to standardize and inspect weights and measures (the practice dates to the Magna Carta (clause 35) in the Anglo-American tradition), but the little nugget on the impact of these inaccuracies tells us that this situation is probably happenstance or, at worst, neglect on the part of the pump owners. Interestingly, the article also mentions that it can cost up to $800 to recalibrate a pump, which considering the thinness of margins in retailing gasoline (only a few pennies on the gallon), is probably incentive enough to recalibrate infrequently—as opposed to deliberately fiddling the pumps. Furthermore, as the fine for a first time offense is only $300, it makes perfect sense for the gas station owners to have the state come out and do the inspections for them (paying the fine and fixing the affected pumps should be much cheaper than constantly re-calibrating of all of them). So, while malfunctions of 1 out of 11 inspected pumps causes us to quickly jump to conclusions, a little vetting of information reveals nothing more than perhaps the nature of pumps, happenstance, poor incentives, or the combination of all three—meaning only a few lines tell the whole story, and they don’t match the rest of it.
Comments