Useless vs Useful Math

by on March 21st, 2004

Readers of my regular blog, A Voyage To Arcturus, know that I’m a big fan of doing the math behind news stories; see, for example, this projection, based on the available numbers, that the final body count from the Iraqi mass graves is almost certainly far higher than the 300,000 figure of current projections.

So I was somewhat pained to see the ordinarily astute William F Buckley Jr quote a bizarre “analysis” as follows:

Looking back on Bush vs. Gore, Professor Joseph Olson of the Hamline University School of Law in St. Paul, Minnesota, gives us a shrewd perspective. Adding up the counties in the U.S. won by the two candidates, it was Gore 677, Bush 2,434. Taking the population of those counties, it was 143 million for Bush, 127 million for Gore. In square miles of land won, Gore 580,000, Bush 2,427,000. The murder rate in Gore counties, 13.2 per 100,000 residents, contrasted with 2.1 in the Bush counties.

This tells us how wrong it is to make facile generalities about the workings of democracy.

It sure does: the actual popular vote was (approximately) 50.2 million for Gore, 49.8 million for Bush, suggesting that the “analysis” as quoted may be safely filed away under “amiable Republican fantasies.” The all-important rate of voting among Republicans was only 88% that of Democrats. (And as this page demonstrates, counting counties is spectacularly useless.)

A truly “shrewd perspective” would incorporate the phenomenal Democratic GOTV effort in 2000, which retook the Senate and came within an eyelash of retaining the White House.

There is nonetheless an underlying trend which may make Democratic vote-getting an uphill climb over the next generation. Turning to page 61 of Demographic Trends in the 20th Century (warning: 3.4 MB *.pdf), and consulting this map of results of the 2000 election (189 kB *.pdf), we find that 8 of the 10 states with the highest percentage of their population at age 15 and under went Republican. Meanwhile, the 10 states with the highest percentage of their population at age 65 and over were nominally equally split, but there is no question that Florida would have been in the Democratic column had not tens of thousands of elderly residents miscast their ballots due to problems with the “butterfly” design.

The picture provided by pages 8 and 9 of Total Population (2.7 MB *.pdf), when compared with the GOP areas on this fundraising map, is even more striking (note especially that the Republican parts of California are where the kids are).

It’s a bit of a simplification, but if Republican-leaning areas are full of younger people and households with children, and Democrat-leaning areas are full of older people and childless households, what will things be like in another 20 years, when all those kids are of voting age and most of today’s older people are gone?

I hereby offer to process-consult to the Democratic National Committee on this problem for the next three years at a million bucks a year. Results not guaranteed.

Jay Manifold