On Tuesday, the Republican Party chose Cleveland as the site of its 2016 national convention. My initial response, in an effort to determine the electoral effect of the decision, was to look to see whether presidential candidates exceed expectations in the state where their convention is held. But parties (especially the GOP) haven’t seen much of a bump for their candidate in such states.
My analysis did not control for other factors that could affect the vote in a state, such as the home state of the presidential and vice presidential nominees. But someone else did, and even after controlling for these other factors, there was no evidence that candidates benefit when their party holds its convention in a particular state. In 2004, Richard Powell, a professor at the University of Maine, ran an analysis that examined the effects of the home state and region of the presidential and vice presidential candidates; the party of the governor; and where the national convention was held.
From 1932 to 2000, a candidate actually lost a statistically insignificant 0.1 percentage points on his vote share in the state where his party’s convention was held. In the more recent 1972-2000 period, he gained a statistically insignificant 0.4 percentage points on his share. That is consistent with the 0.2 points I found for the 1964-2012 period.
Moreover, Powell determined that the Republican Party did worse in the 1972-2000 period in states where it held conventions. Republicans lost a statistically insignificant 0.3 percentage points on their vote share, which, again, is similar to the 0.35 points I showed for the 1964-2012 period.
Powell concluded that “generally, parties do not derive significant electoral benefits in states selected to host the national convention.”
When we combine this detailed study with my simpler, updated look, it’s fairly clear that Republicans shouldn’t expect to do better in Ohio just because they are holding their convention in Cleveland.