I am a Georgetown Ph.D. candidate studying election reform in the United States. My academic writing has appeared in American Politics Research, National Civic Review, The Monkey Cage, and USAPP. I have three immediate projects: a book on the viability of proportional representation in American politics, article-length work on the recent spread of "ranked-choice voting" in U.S. states, and article-length work on best practices in scaling exotic roll-call data.


Proportional representation (PR) lowers the bar for winning office. If a party or faction gets ten percent of votes, it gets ten percent of seats. Can PR elections work with American legislative institutions? How might PR win in the first place? I look for answers in the last wave of reform, 1915-60, when 24 cities adopted PR and all but one repealed it.

Compared to other charter reform episodes, PR resulted when incumbent party defectors colluded with local minority parties. Defectors sought to end direct primaries, and both sorts of groups sought more power in government. Moderately multi-party systems resulted, often with pre-election coalition deals.

Advocacy also mattered for PR's spread. From 1914-64, the National Municipal League included the single transferable vote in its model city charter. This ensured widespread knowledge of a single PR alternative. It also meant that at-large, plurality elections, which became common in this period, were deviations from that model charter.

Whether PR was stable depended on legislative discipline by junior coalition partners. When a multi-party coalition can be fluid, as it can in the absence of "snap" elections, its largest party can get hurt. This leads the largest parties to attack smaller ones by changing the voting system. PR's stability therefore requires one of two conditions: that only the second-largest party "roll" its coalition, or that coalition-splitting bills be entirely avoided.

This project also gives the first quantitative history of PR in American cities. I draw on newspapers, archives, and newly digitized data from three very different party systems. These data cover 5,127 roll-call votes, 126 unique legislators, 1,011 rounds of vote-counting under the single transferable vote, 1,001 candidates, and their party affiliations over 25 elections. The data come from three representative cities: Cincinnati (1929-57); New York City (1937-47); and Worcester, Mass. (1949-61).

Thanks for reading.