Thursday, October 12, 2006

statistic of the day

From Donald MacKenzie's An Engine, Not a Camera: How Financial Models Shape Markets:
"In 1951, just over 2 percent of the pages ... the American Economic Review, contained an equation."
BTW, for those who think I only complain about sociology, the MacKenzie book is marvelous. It's a case study of the effect of academic economics on the economy (more specifically, on finance, or more specifically still, on the creation and operation of options markets).


Brayden said...

I'm going to read this one as soon as the librarians finish searching the stacks (which according to my account they've been doing now for 3 days) and delivers it to my office. Hooray for faculty-laziness privileges!

Tom Bozzo said...

JFW readers thinking that the current figure is 100% less the fraction of bibliography pages should be advised that the AER doesn't exactly beat its readers over the head with equations -- it isn't Econometrica.

There are certainly subdisciplines of economics, and I'd venture even sociology, where 2% of pages with equations is suboptimally low.

jeremy said...

Weirdly, even though I retyped it accurately into Blogger, I still thought it referred to articles rather than pages. 2% of pages isn't really surprising the way reading it as 2% of articles astonished me. The % by 1978 was 44% of pages.

Tom Bozzo said...

It may depend on what you count as an "equation" -- I didn't count tables of results from regression analyses in a canvass of the 9/06 AER on my desk -- but the current fashion seems to favor more English in the main text, with equation-dense appendixes deployed where needed -- I'd think (but admittedly wouldn't bet) that 44% would be on the high side for a typical issue now. Also, I'm probably blind to small bits of math inlined with the text, which I didn't count either.