I like to read newspaper articles and blog posts with statistics because it’s so much fun to question them.
For instance, I just read a summary of deaths due to accidents where the driver had been drinking. This article said that deaths from alcohol-related accidents were down 9% from the previos year, which in turn was 9% lower than the year before.
Great. So the tough DUI crackdown measures are working, right?
Maybe. But before I’d get too excited and start voting for even tougher DU penalties, I would ask for a different statistic. What’s the ratio of auto deaths to miles driven each year? That’s a more relevant statistic. For instance, suppose people are driving 9% less overall. Or suppose people are driving less in specific situations, such as driving to and from work.
This possibility is not unlikely. Here in Seattle, bus ridership has gone up substantially.
Another possibility is that with the recession and lifestyle changes, people are just staying home. Going out to the clubs costs money. If you have kids, you need babysitters.
So I’m certainly not advocating more DUIs. But I am advocating more judicious application of statistics, because we have limited societal and governmental resources. We want to use them wisely.
