nuclear power and rational choice


Recent events in Japan should NOT change anybody’s position on nuclear power. It is worth taking a moment to learn about how risks are calculated and how to make rational decisions, to understand why this is the case.

In economics, and decision theory in general, a choice is a rational one to make if the anticipated benefits outweigh the anticipated costs. There is only one catch to this seemingly simple idea: because the future is uncertain, you have to weigh the possible benefits (and risks) of an outcome by the chances of that outcome happening.

This is easiest to understand with a couple of simple examples:

Example 1:I tell you that I will flip a coin, and give you $10 if it comes up heads and take $20 from you if it comes up tails. Is it rational to agree to this? Obviously, not: the two scenarios are equally likely, and the amount you would lose is greater than the amount you would gain.

Mathematically, we can say:

value of the bet= (probability of scenario A) * (value of scenario A) + (probability of scenario B) * (value of scenario B)

In this case, the probabilities of the two senarios are both 50% (which is 0.5), and the value of heads is $10 and the value of tails (to you) is -$20, so the value of the bet is:

0.5 * $10 + 0.5 * -$20 = 5 – 10 = -5

The value of the bet is negative. It’s a bad bet.

Example 2:Suppose I said that I’m going to roll a die, and if the number is even I will take $10 from you, but if I roll a one (snake-eye!) then I will give you $50. If the number comes up 3 or 5, then no money is won or lost.

Should you take the bet?

Well, we can actually do the math on this. The probability of an even number is 0.5 (since half of the numbers on a die are even), while the probability of rolling a one is 0.167 (actually 1/6 but I’m rounding for now).

value of the bet= 0.5 * -$10 + 0.167 * $50 = -5 * 8.35 = 3.35

The value of the bet is positive: you should take the bet!

Even without doing the math, this is the way people intuitively work, most of the time. Or at least, this is how they work when they are acting rationally. Even if whatever process you use to make your decision is unconscious, it is rational if you are accepting bets where the value is positive and rejecting bets where the value is negative.


How does this apply to nuclear power?

Let’s assume that you are deciding on whether to “take the bet” on nuclear power (as it were). You consider what you think the possible benefits are if everything runs as planned (e.g. lots of energy, independence from oil, made in America, and so on) and you multiply that times the probability that you think everything will run as planned, and you consider the possible costs if there is a problem (e.g. radiation in the atmosphere, land that cannot be used for decades, health risks, etc) and you multiple that times the probability that there will be a problem.

Of course, that’s an over-simplification. In reality, there are a very large number of possible things that can happen, each with a different probability and a different set of positives and negatives attached to it. But for now, I’ll keep the examples simple.

When people disagree about whether we should pursue nuclear power, it is usually because they have different estimates of either the values of the outcomes, or the probabilities of the different possible outcomes, or both. But it is reasonable that people might disagree about these things, since there really isn’t any absolute objective way of measuring these things.

This point bears emphasizing: it is completely possible to rationally accept nuclear power and it is also completely possible to rationally reject nuclear power. The simple fact that one person is in favor of nuclear power and another person is against is does NOT mean that one person is being irrational. It just means that the two people disagree in their perception of either the values of the outcomes or their probabilities.

For example, if I feel that becoming energy-independent from the middle east will make us much safer and more prosperous whereas you think “Eh, it might help a little but it won’t make that much of a difference,” then that will change the equation for each of us and it might come out positive for me and negative for you.

On the other hand, if I think that the probability of a melt-down is 0.00001% each year and you think that the probability of a melt-down is 0.0000000001% each year, then the impact of that negative outcome will figure much more highly into my equation than yours, and it might make the end result of my equation negative while yours is positive.

In the end, these numbers are almost impossible to objectively estimate. As a result, it’s not surprising that intelligent, educated and rational people might still disagree.


How does this relate to recent events in Japan?

The recent earthquake in Japan can only change someone’s mind about nuclear power in one of two cases: A) it makes that person think, “Wow, huge earthquakes are more likely to happen than I thought they were!” or B) “Wow, I didn’t realize a large earthquake would produce that much damage!”

And I’m sorry to say it, but neither of these cases is very likely!

A) Everyone knows that 8.8 earthquakes are very, very rare. The fact that one just happened doesn’t mean that they are LESS rare than we thought. It just means that one of those really rare things just happened to occur now.

B) Everyone knows how bad a melt-down is, and I think most people would have figured that a huge earthquake would have a good chance of causing a meltdown. Anyone who is only now learning how bad the outcome of a meltdown is has not read a lot of history. (Or been watching made-for-T.V. sci fi movies… but that’s another matter.)

So the raw fact is: your friends who were in favor of nuclear power a month ago are being completely and 100% rational if they still are in favor of nuclear power today. They have learned no new information that would change their minds: bad earthquakes are still rare, earthquakes still might cause meltdowns, and meltdowns are still bad. No part of the equation has actually changed.