The Monkey Cage

On the article: yes, but remember also that the N is 16, which is small. Importantly, just because she is unable to reject the null hypothesis (i.e. no effect of divisiveness) does not mean that the alternative is wrong.

I would also look at the other direction, that a divisive campaign might actually be beneficial for the party with the division. The fighting candidates and their party get most of the media attention while the real competitor (say McCain) gets little. This is under the (strong) assumption that all news are good news.

Anand Sohkey has written an interesting paper on divisive primaries where he examines the social networks of voters in a single county where there was a competitive primary. Interestingly, he finds that the losers tend to withdraw from networks that make them predisposed to supporting the candidate from their party.

Its not exactly evidence that divisive primaries hurt candidates, but it does suggest that the level of acrimony can seap into micro social environments and make the task of the winning candidate more difficult.

My comments on divisive primaries ended up growing into a post of their own on my blog. Please feel free to come by Frontloading HQ for my take.

The main points of contention:
When does the change from competitiveness to divisiveness take place?

How has the change in timing of delegate selection events affected the development of divisiveness?

Atkeson's 1998 research addresses the impact of divisive state primaries, not national party division. (The same is true of virtually all other studies of divisive primaries.)
Atkeson and I, together with Damon Cann, Nathan Burroughs and Audrey Haynes, have measured the effects of national party division (controlling for 14 other factors). We find that a divided party will lose up to 5% nationally in the general election, as well as losing up to 2% in individal states that had divisive state primaries.

Paul-Henri, do you have a paper with these data? If so, could you post a link?

John,
There's a link to Paul's article (from APSA 2002) in my post linked above (3rd comment down).

Josh, thanks very much. Apologies for the oversight. I've appended that link to the original post.

Oh sure. No problem. Thanks for putting that up.

I emailed Paul to ask him if he had an updated version of the paper that includes Damon Cann. He's been added to the paper since that conference presentation at APSA. So Paul may try to post a link to that later.

Thanks for a great post. This has been a good discussion.

Also relevant to the discussion is Jeff Lazarus's 2005 LSQ article about primary divisiveness in House elections. He argues that it is anticipation of GEs that cause divisive primaries, not that divisive primaries cause close GEs. To at least some extent that seems what is going on in the current Dem primary. If they were competing for the "right" to go on to certain defeat at the hands of McCain, I bet there would be far more enthusiasm for dropping out.

A colleague today reminded me of a similarly themed SNL sketch about a democratic primary debate in 1992 (when Gulf War halo still covered Bush 41 and most thought he would easily win the GE) and the candidates all seem like they're trying to get out of running. Cuomo says something like "but I have mob ties!"

In the recent cases where
the nomination was contested
until the convention, the
party with the contest lost,
and the loss was in part due to the contest.

Humphrey-McCarthy
Ford-Reagan
Carter-Kennedy

John,
Here, finally, is the link to the updated version of Paul Gurian's (et al.) divisive primaries paper.
Josh