Politically Viable U.S. Electoral College Reform.

The U.S. Electoral College's winner-take-all method of awarding electoral votes creates a high susceptibility to disputes and errors, but past reform attempts have glossed over their likely disruptions to power balances among states and between the two political parties. This gap is filled by connecting pragmatic models of power shifts and election disputability to a historically informed probabilistic model of future elections. This methodology is then applied to a continuum of reform proposals between the current system and the Lodge-Gossett version of a national popular vote. The results show that a modest smoothing of winner-take-all near the toss-up point delivers a good tradeoff between reducing dispute frequency and distorting power balances, enabling meaningful reform in an era of high polarization. This conclusion holds for extrapolation of the current national landscape into near-future elections, as well as for more arbitrary distributions of partisans among states to represent far-future landscapes. However, as electoral award smoothing diminishes the frequency of disputed elections, it inevitably broadens their scope. [ABSTRACT FROM AUTHOR]