A nuclear lottery



In today's New York Times Magazine, two smart writers, Stephen J. Dubner and Steven D. Levitt, make a really stupid mistake when they talk about nuclear power. The piece is called "the Jane Fonda Effect," and it argues that the reason the United States doesn't have more "clean and cheap nuclear energy" is that the 1979 movie "The China Syndrome" , combined with the accident at Three Mile Island, , irrationally scared the public away from this otherwise wonderful source of energy that doesn't contribute to global warming.

"The big news is that nuclear power may be making a comeback in the United States," the authors, who write the popular column "Freakonomics," note. "Has fear of a meltdown subsided, or has it merely been replaced by the fear of global warming?"

To find that answer, they cite the work of Frank Knight, a legendary U.S. economist who first defined the different in the behavior of people faced with risk (which is quantifiable) and uncertainty, which is, well, uncertain. Here's the drill: You have two boxes filled with red balls and white balls. Box one has exactly half of each; box two has an unknown mix. You want to draw a red ball; which box do you pick?

Most people, of course, pick the first one -- they know the exact risk. That, the authors say, is nuclear power. Then there's the uncertain risk -- global warming. So maybe it makes sense to choose the nukes, knowing that there's a small, but somewhat quantifiable, risk.

But in citing one of the nation's pre-eminent economic thinkers, the two economists miss the point that ought to be part of every economics lecture that discusses choice theory. Choice isn’t just about uncertainty and risk – it’s about possible outcomes. If I told you today that for $1, you could buy a ticket that might get you a free ham sandwich tomorrow – and the odds were about 17 million to one against you winning -- no sane person would take the bet. But what if I told you that the odds were the same, but the reward was a check for $100 million? A lot of people would cough up the buck; if they didn’t, the entire lottery industry wouldn’t exist.

So flip that around: If I told you that the odds of a nuclear accident were very low (they are) and that if you take the gamble on nuclear, you get the reward of emission-free power, you might jump at it. But if you knew – as you should – that the outcome of the accident that probably won’t happen could be the deaths of several million people (as it could be if the Indian Point plant just upriver of New York city melted down or were blown up by terrorists) you might say: Never mind.

Sure, we take risks every day -- getting in a car and driving on a highway is a pretty big one -- and as a society, we're willing to allow individuals to make that choice. But the worst possible outcome of you drivng badly is the death of you and a few others; if a single car crash could wipe out an entire city and render tens of thousands of acres of land uninhabitable for centuries, cars would have been banned long ago.

If I were crazy enough to try to write this out as an equation this late at night, it might look something like this:

V(c)=R*O, or

The value of a risk equals the chance of success (or failure) times the projected outcome.

A tiny chance at winning $100 million in the lottery is still a decent-sized number. A small chance at a disaster of epic proportions is a decent-sized number, too.

The lottery is a sucker’s game, but it keeps right on going, because so many of us are willing to risk a buck on a very, very slim chance at life-transforming riches. But you don’t put pension funds into lottery tickets, and you don’t bet your kid’s future on them. Nuclear power only makes sense from the standpoint of an economist if we’re willing to accept what is an admittedly slim chance at a disaster of cataclysmic proportions. Sorry, Messrs. Dubner and Levitt, but I’m not buying.