John Sides in Good Authority discusses a new paper that uses statistical analysis to show that “nonpartisan election administration may be the norm.” John add his own note of caution to the paper’s disclaimer against over-reading the implications of their statistical analysis: “The authors are appropriately careful about what their results can and cannot show. Although there is no average effect of partisan election administration on election outcomes, they write that ‘we cannot rule out small differences between Democratic and Republican officials that could determine very close elections. We also cannot rule out rare but very large effects.’”

I look forward to reading the paper. In the meantime, I’ll add my own cautionary note: the reason why I chose the historical methodology I used for the Ballot Battles book on disputed elections in the U.S. is that, after considerable preliminary research, I conclude that focusing on those specific episodes where official decisions over election administration could determine the outcome in a major race was necessary in order to understand how the system worked in the context when every vote truly mattered. The performance of election administration in run-of-the-mill cases, where the margin of victory is so great that the role of official decisions doesn’t make a difference, can’t tell us very much (if anything) about what happens when the system is put to a stress test. Moreover, even when the system is stressed for down-ballot races, where the stakes are not so high, how the system responds in those situations is not a reliable indicator for how it will respond when similarly stressed in the context of a statewide or other major election.