« Networks in politics | Main | A Tournament of Party Strategies »

Annals of Improbable Research

Forthcoming from James Fowler in PS: Political Science and Politics

Stephen Colbert, the host of Comedy Central’s The Colbert Report, claims that politicians who appear on his show will become more popular and are more likely to win elections. Although online discussions cite anecdotal evidence in support of his claim, it has never been scrutinized scientifically. In this article I use “facts” (sorry, Stephen) provided by the Federal Election Commission to create a matched control group of candidates who have never appeared on The Colbert Report. I then compare the personal campaign donations they receive to those received by candidates who have appeared on the program’s segment “Better Know a District.” The results show that Democratic candidates who appear on the Report receive a statistically significant “Colbert bump” in campaign donations, raising 44% more money in a 30-day period after appearing on the show. However, there is no evidence of a similar boost for Republicans. These results constitute the first scientific evidence of Stephen Colbert’s influence on political campaigns.

Update: See also Fowler’s LA Times op-ed for the popularized version with extra truthiness …

Comments

This is subject to the usual correlation != causation discussions, because Colbert might seek out candidates about to break through into new levels of fame instead of causing them to do so.

Ideally he'd have examples where someone was supposed to attend, but for some reason could not and compared them with their replacements.

Here's the companion LA Times op-ed if you are interested.

Post a comment