Does party support help candidates win?

Abstract

The parties’ congressional campaign committees have made it their business to strategically provide contributions to candidate campaigns in order to help their candidates win. However, the effectiveness of these contributions in terms of increasing the competitiveness of party-sponsored candidates remains untested. Using contribution data from the U.S. Federal Election Commission in a series of mixed effects models as well as a matching analysis, the receipt of direct party contributions and coordinated support is shown to significantly improve the competitive position of challengers but not open seat candidates in races for the House. Further, independent expenditures by the parties do not significantly increase candidates’ competitiveness. The implications of these results for future party strategies are explored.

Keywords:

• Political parties
• Political campaigns
• Congress
• Campaign strategy

Notes

1 A personal interview was conducted on January 14, 2010 and two telephone interviews on May 16, 2010 and October 17, 2011 with three party officials, one Republican and two Democrats, all of whom have had experience working for the congressional campaign committees of the respective parties. In keeping with the consent agreement, their names will be kept confidential.

2 The FEC files list the reporting committee/candidate and independent expenditures against that candidate rather than independent expenditures made against that candidate's opponent. Thus, it was necessary to reorganize the data to add independent expenditures against the candidate's opponent to the candidate's tally of party support.

3 This is done to avoid confounding the slope of party support, a problem that would arise if the candidates of different parties were pooled together because the results of winners and losers from the same sets of races would be in the same model effectively canceling out one another. This problem does not arise in the challenger-incumbent models because the dependent variables all pertain to outcomes for the challengers but not the incumbents.

4 Insignificant results are not shown in tables.