How good are American elections?

Where would one start in answering this question?

Whenever this question is posed, it is common to answer it from the position of deeply held beliefs, but rarely from the position of a systematic analysis of facts. These beliefs might arise from partisanship: a good election is one that my favored candidate wins. These beliefs might be chauvinistic: a good election is one run according to the rules of my community.

Rarely are these beliefs rooted in hard facts.

When facts intervene, they rarely are presented in a systematic fashion. Opinions about levels of voter fraud might be attributable to a viral YouTube video. Concerns about the effects of a new voter identification law might be informed by a reporter’s interview with an activist who is eager to share stories about how voters she has talked with will be disenfranchised on Election Day. Satisfaction with a new electronic voting machine may be illustrated by a picture of a smiling citizen coming out of the precinct with an “I Voted” sticker stuck to her lapel. Disdain about the ability of local governments to run elections might follow from a newspaper article detailing yet another season of long lines when waiting to vote in Florida (or South Carolina or Maryland or …). At its worst, this approach is evaluation by anecdote.