Why Endorsements Matter in Presidential Nominations

In the New York Times last week, Michael Powell looked at the value of endorsements in presidential nominations and came up with little: “A majority of endorsements are political popcorn, signifying nothing.”

This misses the big picture. Marty Cohen, David Karol, Hans Noel, and John Zaller have just finished a new work on the role of political parties in the nominations process, which is forthcoming from Chicago University Press. Hans presented an overview of the book at GW last month.

A very interesting finding: the best predictor of how many delegates a candidate wins is not money raised, the amount of media coverage, or support in the polls. It’s the number of endorsements the candidate has received from party elites. As Powell himself notes, endorsements can serve as cues to the endorser’s followers and to donors. They may also signal the willingness of the endorser’s political “machine” to work on the endorsee’s behalf—precisely the kind of benefit that would not be picked up by polls or FEC reports.

Cohen et al.’s work suggests that party leaders play an important role in presidential nominations, even after reforms (e.g., the McGovern-Fraser Commission) that empowered rank-and-file partisans at the expense of leaders. To be sure, we are a long way from the proverbial smoke-filled backrooms of yore. But party leaders ain’t popcorn.

Find a copy of the book manuscript here.

P.S. Who is leading in endorsements in 2008? According to some data Hans presented, it is Romney and Clinton. Stay tuned.

2 Responses to Why Endorsements Matter in Presidential Nominations

  1. Erik November 27, 2007 at 6:35 pm #

    Could it be that elites are simply “smarter than the average bear” – predicting who is likely to get their parties nod/win primaries and thus making a strategic decision regarding who (and when) to make endorsements?

  2. Arnold Kling November 28, 2007 at 6:53 pm #

    What Erik said. If the authors don’t have a way to control for reverse causality, the study loses a lot of its meaning.