Using experiments to predict voter turnout
Nice summary of the issue by Dr. Krueger in today's NYT. Excerpt:
The filmmaker Michael Moore is stirring controversy by offering 'slackers' a three-pack of Fruit of the Loom underwear if they promise to vote. Beyond its comical value, Mr. Moore's campaign raises a serious question for candidates in both parties: What is the best way to increase voter turnout?So which method produces the greatest increase in voter turnout/$? Door-to-door canvassing. What I find interesting is the use of experimental methods to uncover this result. As Krueger concludes "Perhaps we are witnessing the rise of what can be called Fisherian Democracy, after the statistician R. A. Fisher, who invented the experimental method."
Two Yale political scientists, Donald P. Green and Alan S. Gerber, have studied turnout for years. Their findings, based on dozens of controlled experiments done as part of actual campaigns, are summarized in a slim and readable new book called 'Get Out the Vote!' (Brookings Institution Press), which is bound to become a bible for politicians and activists of all stripes.
posted by david burnstein at 9:06 AM
1 Comments:
The latter won't pass b/c its too expensive. Every employer would be stuck with a big tab each election day. And there is no guarantee people would vote. Instead, they'll just get hammered the night before and be too hung over to vote on election day.
The former is a good idea, but to promote (more) equal incentives it should increase substantially the value of the awards. From a cost standpoint, it should offer a single, big sum jackpot, say a billion. That would impose less bias than your proposal, which would disproportionately encourage poor folk (but not necessarily rich folk). Either way its not going to be an incentive for those that understand probability theory.
Post a Comment
<< Home