Short version: if something doesn't happen for thirty years, people start thinking it's exceedingly unlikely. But (a) there are a lot of very problematic 'somethings' out there, and (b) thirty years isn't a particularly big multiple of a two-term Presidency. So if you're an incoming Democratic President that hopes to be re-elected, you'd better watch out for the misnamed 'black swans' - they'll kill you.
Today's text is from the book of McClatchy, chapter Obama, verse SAO:
"It's really important to understand you have decades of nothing going wrong," said one senior administration official, who spoke only on the condition of anonymity as a matter of White House policy.
"The last time you saw a spill of this magnitude in the Gulf, it was off the coast of Mexico in 1979," a second senior administration official said. "If something doesn't happen since 1979, you begin to take your eye off of that thing."
Oops.
I'm not questioning the SAO's basic assumption: that if something doesn't happen in a good long while, that's probably an indicator that it isn't particularly likely to happen next year. The problem is the jump to "you begin to take your eye off of that thing."
A probability lesson: suppose the fact that something hasn't happened in N years means that its likelihood of happening next year is 1/N. So an oil spill that hasn't happened in 30 years had 1/30 chance of happening this year.
The problem, or one of them anyway, is that a two-term President is in office for eight years, and an eight-year timeframe is a considerably larger target for an unlikely event than a one-year period. And we reflect that in the numerator of our formula by replacing the 1 with an 8. (Technically, that's not exactly true, but close enough.) So the oil spill that hasn't happened in 30 years has an 8/30 chance of happening on Obama's watch, which is a nearly 27% chance, which is far from trivial right there.
The other problem is that there are all sorts of bad things that haven't happened in a good while. We haven't had a hostage crisis in 30 years either. Or energy shortages and gas lines. And we haven't had a Great Depression in 80 years. So these things have probabilities of 27%, 27%, and 10% (=8/80).
And we're just getting started, but even with that set of 'black swans,' their individual probabilities of afflicting a particular Presidency are hardly trivial, and the probability that at least one will happen in a given 8-year period is around 65%.
So this business of taking your eye off the ball because it hasn't happened in 30 years or so is dangerously misguided. Hell, even if it hasn't happened in 80 years, you should still keep half an eye out for it, if its occurrence would be consequential enough to ruin your Presidency. Really the only reasons to take your eye off these things is if there are causal, structural reasons why these things won't happen anymore, or if their potential ramifications have dwindled to triviality.
In the case of the oil spill, better regulation kept the worst blowouts from happening (maybe), but better regulation was conspicuously undermined during the Bush years. And the ramifications were same as they ever were.
I think the Obama crew had better keep an eye on the possibility of peak oil, because it could really ruin the Presidency of a Democrat that gets caught not only unprepared for it, but having failed to really try to warn the public about the possibility. If I were in his place, I'd be doing it low-key, but I'd still be mentioning in my speeches the real possibility that there may soon be less oil than there used to be, and that when it does, the price will go up radically. And I'd be suggesting that we make major investments in subways, buses, and SUPERTRAINS so that $10/gallon gasoline doesn't catch everyone without any alternatives to driving long distances daily.
