I view the work I've done related to statistics and economics as, roughly speaking, how to do something without having to do everything.