Saturday, October 4, 2014

Thus does trust decline

Recent events related to Ebola in the US call to mind Mark Twain's famous adage that there are three kinds of lies: "Lies, damn lies, and statistics." From the president on down, there has been a concerted effort to minimize risks and assure the populace that there is nothing to fear. But as someone has said, the best way to reassure the populace is by demonstrating competence. That has been regrettably absent so far. So the clerisy repeat their lies in the hope that citizens won't notice and the citizens notice and become increasingly concerned.

Richard Fernandez has a useful article on the critical statistical distinction that is being manipulated, allowing politicians to correctly argue that the statistical chances of Ebola reaching America are small while the real likelihood is certain.

From My Hull is So Big and the Leak is So Small by Richard Fernandez.
After all the CDC asserts that the “chances of importing a cluster [are] very small”. But that may be true without contradicting Dr. Mobley’s argument because they are talking about different things. The CDC is talking about probabilities (or “chances”) as expressed in percentages while Mobley is talking about expected values which is the number of Ebola patients you can predict will arrive over time.

What is expected value? “In probability theory, the expected value of a random variable is intuitively the long-run average value of repetitions of the experiment it represents.” Think of it as the long run value of a bet. If you have a ten percent chance of winning $500 the expected value of the bet is 0.1 x $500 or $50. If you’re playing 10 percent for a million the bet is more valuable. It is .1 x $1M or $100,000.

[snip]

When the CDC says the probability of a diseased person reaching the US is “very small” they are talking about the betting odds, which in our hypothetical example was 12/10,000 or .0012. But remember that expected value has a second term: the stake at the table. If there is one Ebola patient with a .0012 chance of reaching America that is one thing, but if there 10,000 people each with a .0012 chance of getting through that is another. The CDC is saying that they have a revolver with thousands of chambers and only of them has a cartridge in it. Mobley is saying that’s true but if you keep playing Russian Roulette long enough you’ll eventually hit upon a loaded chamber.

Spin the wheel enough times and eventually Ebola gets lucky. The more Ebola patients there are worldwide the more times you spin the wheel. Dr. Mobley is saying that if the disease keeps exploding in West Africa the wheel will get spun too often. Someone will get through, because even though the chance of arriving is individually small, there are just too many individuals who might potentially try to get through.
Fernandez points a related item that fuels the belief that the established powers and clerisy are dealing a straight hand. The Precautionary Principle was widely used when it was beneficial to the establishment (for example when dealing with Global Warming).
Parenthetically, what is really interesting is where the “Precautionary Principle” has gone to. The administration used to be really worried about the long shot odds that Michael Mann’s climate model might be right. But it doesn’t worry about Ebola odds. Maybe we should apply the Precautionary Principle all of the time or none of the time. But it’s altogether too confusing to apply it in this uneven manner.
So when we are talking about a theoretical climate position that has not yet been settled, is incremental, and is distant in time but that involves the capture of billions of dollars from citizens, we invoke the precautionary principle. When we are talking about real threat in the present where there are only citizen lives at stake and not the redistribution of money, the precautionary principle is missing in action.

And thus does trust decline.

No comments:

Post a Comment