Making a Fermi estimate of creationisms costs


Fermi in Missouri | 
[Via  NCSE]

Once in a while, a journalist will ask a question that really makes me think. Such a question arose recently, when I was asked whether Missouri’s House Bill 1472—which I earlier said “would eviscerate the teaching of biology in Missouri”—was the worst antievolution bill to come down the pike in a long time. At first, I was inclined to respond by saying that they’re all horrible, which indeed they are. But pondering it further, I realized that there was enough survey data available for me to make a back-of-the-envelope estimate of the expected effect—measured in lost student-hours of effective evolution education—of the two antievolution bills currently before the Missouri General Assembly. Such estimates are sometimes called Fermi estimates, after the Italian physicist who famously estimated the yield of the Trinity A-bomb test accurately by observing the movement of scraps of paper; hence the title of and the illustration for the present post.


Fermi made estimates all the time from incomplete data. When done well, they can provide interesting conclusions.

Determining an estimate through a series of prior estimates with the goal of getting close (say within 10-fold) has become a standard tool of determining whether someone;s conclusions are reasonable and whether all the proper inputs have been included.

Here we have a Fermi estimate of two creationism bills – one that would allow parents to remove their children from any class that discusses evolution and the other that allows teachers to openly discuss creationism without any penalty.

The conclusion: the former has a much greater impact on child education and is therefore a much worse piece of legislation than the latter.

But, as we know, this is like determining the number of angels dancing on the head of a pin. Because in the real world, both bills are horrible.