Do Mandates Matter?

by Nolan McCarty on November 6, 2012 · 8 comments

in Blogs

Undoubtedly, the political discussions in coming days ahead will focus on the question of whether the winner of today’s presidential election will have a “mandate” to move the country in his preferred direction.

The idea of a mandate is another one of those ideas that are much more popular among the press corps and pundits than it is among political scientists.  The data simply do not support the idea that presidents who win with large electoral majorities are able to translate that support into legislative victories.

David Mayhew’s recent book Partisan Balance provides some new data with which to look at the question of mandates.  His data include all of the major presidential requests from the first two years of a presidential term from Harry Truman to the second term of Bush 43.  From these data, he can then tally up “wins” (Congress passes what the president wants) and “losses” (Congress does not pass the requested legislation).  Moreover, to take into account the fact that not all requests are equal, he weights each outcome by the importance of the request to president who issued it.

From these data, we can check to see whether presidents who won by large margins had better won-loss records (weighted by importance).  The figure below plots Mayhew’s success rate against the number of electoral votes won by each president.

That correlation is zero to three decimal places.  Now one might quibble about the inclusion of the Watergate-plagued second Nixon term.  Even if that observation were dropped, there would still be no statistically or substantively significant relationship (and one could as easily argue for the elimination of the second Eisenhower term where Ike batted 1.000 but only made one request).

The story looks little different is we examine the total wins rather than the won-loss ratio.


Here LBJ looks like an outlier.  Apart from his success following his landslide victory over Goldwater, there is significant negative relationship between wins and electoral votes.

Even when I control for divided party control, the results are the same.  Winning big is no more valuable than winning ugly.  Although President Obama is not included in Mayhew’s dataset, his experience does not deviate from the pattern.  Despite a relatively modest 365 electoral votes, he had major legislative victories on the stimulus package, health care reform, and financial reform.

So while there are many legitimate worries about whether the winner of today’s election can have a major impact on our nation’s problems, size (of his victory) does not matter.



(Thanks to Ben Farley for catching the mistaken duplication of figures in the original post.)



brianS November 6, 2012 at 6:05 pm

might want to fix the second figure (it is identical to the first)

Nolan McCarty November 6, 2012 at 6:07 pm

Fixed. Thanks.

Jack November 6, 2012 at 6:39 pm

Hmmm. Looks like you might have to have a conversation with your humanist colleague down the hall:

kerokan November 6, 2012 at 7:18 pm

I wonder if Mayhew controls for the “difficulty” of a request in addition to its “importance to the President”. A mandate may encourage a leader to make bolder, more “far-reaching” proposals. The bigger the proposals, the bigger the resistance will be. As a result, even if a bolder leader with a stronger mandate may not have a larger ratio of victories or a larger number of victories than leaders with smaller mandates, the leader with a mandate could achieve more by winning “bigger” victories.

Anonymous Coward November 7, 2012 at 10:00 am

Isn’t a simpler argument just “Mandates are fairy tales, because Arrow.”?

Jeffrey Lax November 7, 2012 at 7:17 pm

Without taking into account the endogeneity of requests (not just “importance to the president” but the nature and size of the requests, I’m struggling to figure out what to take from this analysis. I suppose a raw success score, or weighted one, matters as a …. scorekeeping exercise, but I’m not sure what political science questions this really answers. I also assume there is other analysis including legislative composition, etc. I don’t find this takedown of mandates (even if I agree with conclusions) as persuasive as his takedown of realignment arguments.

John Patty November 7, 2012 at 11:05 pm

I agree with Lax about requests. But another more easily dealt with objection is the use of electoral vote to measure the perceived strength of the victory. I won’t belabor the point except to admit that the electoral vote somewhat captures the bicameral structure, but this could be captured in more nuanced ways, too, in a multivariate fashion. But, yes, I also agree with Lax that I don’t think electoral mandates are as simple as aggregate election returns. Thanks for the post-got me thinking in the fun way.

Currency euro mandate November 9, 2012 at 7:05 am

these articles are very usable and gracefull thanks for contacting us.

Comments on this entry are closed.

Previous post:

Next post: