Category Archives: Models and modeling

Actuaries aren’t economists

I love to tell the story of Milton Friedman’s brush with actuarial science, as he does above. (Shorter Milton: The exams were too hard.)

Today we get the cautionary tale of two economists who bollixed the AVERAGE formula in Excel. That and a couple other dubious adjustments led them to publish a paper that, while incorrect, was highly influential. (See also here. Authors’ response given here.)

I’ve reproduced the error below, taken from a Salon article:

economistsThe average should have been cells L30 to L49.

By coincidence, I recently made a similar error in a pricing model. My formula was supposed to sum the numbers in a column but failed to capture all of them.

I was corrected by a sharp peer reviewer. Why weren’t these guys?