Princeton Election Consortium

A first draft of electoral history. Since 2004

Aug 06: Biden 372 EV (D+6.3% from toss-up), Senate 51 D, 49 R (D+3.9%), House control D+3.6%
Moneyball states: Senate MT KS ME, Legislatures KS TX NC

How behavioral science could help get more Americans to vote

June 15th, 2016, 9:03am by Sam Wang

My latest, in the Washington Post.

Tags: Politics

14 Comments so far ↓

  • Tony

    I love the site and visit every day, but I have a question on how Kansas can still be considered safe Republican when a recent poll (and I think the only poll this election cycle) shows Clinton up by 7. If you do polls-only, wouldn’t this make Kansas a tossup? Also, in general how do you handle states with no or few polls?

  • bks

    Congrats, Sam.

    [Unrelated] Those wanting more Clinton-favorable electoral college analysis will find it at Josh Putnam’s site:

  • SFBay

    Interesting article. I believe procrastination is the default thought pattern of human beings. We require a proactive thought to move on any idea or plan. Behavioral science has shown this to be true.

    Maybe the inertia principle that is part of physics translates in some fashion to human behavior patterns.

  • Daniel Barkalow

    I think what’s missing from that analysis is information about the portion of registered voters who voted in 2012 in the different states. It wouldn’t be unreasonable to guess that the people who don’t register in advance also wouldn’t figure out how to get to their polling place in advance, schedule around going there, have the leisure to cope with the lines, etc., and that states that didn’t make it easy to register before won’t make it easy to cast your ballot now.

    I’d agree with the effect of automatic voter registration if, for example, all the necessary vote-by-mail forms (for all voters at the address) would show up with a utility bill due to be paid in the time window where sending in votes is appropriate, but otherwise I still see too much of a tendency to inaction, and I suspect it would be correlated with current non-registration.

  • Rob in CT

    Solid article, which led me to discover that my own state is pursuing this approach (yay!)

    Of course, many people’s response would not be “that won’t work” (of course it will work) but rather “if people can’t be bothered to register why do we want them voting?” That’ll be the hurdle to overcome. I could see this being easier to roll out in solid blue states and maybe even in solid red states (and/or less diverse states?), but “battleground” states would see more opposition, b/c this sort of thing has the potential to shift electoral outcomes leftward.

    • Matt McIrvin

      Solid-blue Massachusetts has historically been somewhat backward when it comes to voting and voter registration, but I have hope: we did manage to get early voting, and it starts this fall.

  • RA

    Off topic, thinking about using Google searches as a new big data form of demographics, I wonder if it is possible to get other data sources that might work just as well or better. Which amazon sales, which netflix movies, which ebay sales best correlate with Hillary or Trump voting and under what terms could researchers access data like that?

    But if all that is available is google search data, then that will be increasingly useful and predictive and I guess it will have to do.

  • Kevin

    Great development. Next step: universal absentee voting, as in Washington state.

  • Jeff Graham

    What if instead of everyone voting, we just had a sample of the voters. If done properly, it seems like it would eliminate the turnout factor in elections, which are currently just a large and potentially biased sample.

    • Lorem

      On a conceptual level: seems reasonable, and I’m largely for it, but it’s also possible that turnout is something we want to care about. Perhaps people who care more about the outcomes should have greater influence. It’s a bit of a philosophical question.

      On a practical level: I cannot think of a single scenario in which this idea has the slightest chance of being enacted in any nation.

    • Ed Wittens Cat

      The GOP will *never* allow automatic registration because demographics is destiny.
      BUT!!…if what we really care about is turnout, lets enable online registration, indeed even online voting. The GOP would concievably support that because they have behavioral tendency enthusiasm advantage– Dems would support it because they have numeric demographic advantage for the forseeable future.
      The vulns are many, sure, but afterall we are about to deploy “smart-car” tech based on 100 million lines of code. Theres only 50 million lines of code in the LHSC.

  • bks

    In re behavioral science: Is there any evidence that political ads on TV work? Is Clinton’s money advantage wasted on TV ads that have no effect?

Leave a Comment