Evaluative Proportional Representation

3/13/2025 (Bozy):

Evaluative Proportional Representation (EPR) is a proportional extension of Majority Judgment, devised by Stephen Bosworth.

EPR is only related to the Expanding Approvals Rule in that it invites voters to “approve” any number of the candidates. However, EPR also allows the voter to more expressively specify each approved candidate’s suitability for office as either Excellent, Very Good, Good, or Acceptable. Disapproved candidates can be indicated by Poor or Reject. All ungraded candidates are counted as a Reject.

Each voter’s ballot is guaranteed equally to add to the weighted vote of one of the elected candidates in the legislative body. For example, each of the elected members of a seven-member city council will have a different weighted vote in the council exactly equal to one of the seven-largest number of ballots counted for them according to the rules of the count explained below.

Compared to known voting methods, EPR is most like proportional ranked-choice voting (PRCV, also called single-transferable vote (STV)). However, EPR instead invites voters to rank candidates by grading their suitability for office as either Excellent, Very Good, Good, Acceptable, Poor, or Reject. Voters can award the same grade to more than one candidate. Each EPR ballot contains only one vote.

All these grades are counted to guarantee that each EPR voter will be represented equally within the weighted vote of the elected candidate who had received either the voter’s highest grade, their highest remaining grade, or their proxy vote. For example, an elected candidate will receive a voter's highest grade if that grade helped to provide that candidate with one of the seven-largest totals when electing a seven-member city council. If not, that voter's remaining highest grade may help to elect a different candidate if that candidate finally receives one of the seven-largest totals. If not, this voter's ballot becomes a proxy vote. This proxy vote will be added to the weighted vote of the elected candidate publicly judged to be most suitable by the unelected candidate who had received this voter's highest grade. Consequently, EPR maximizes the quality of each citizen’s vote, as well as enabling each citizen’s vote to quantitatively count equally within the weighted vote in the legislative body of the candidate their vote had helped to elect.

Additionally, EPR’s qualitative grades enable voters to express more clearly their different degrees of support or opposition to different candidates. For example, in contrast to an EPR voter, a plurality, Condorcet, approval, ranked-choice, or score voter cannot secretly communicate whether their 1^st choice candidate is their least bad or best candidate. Each citizen’s EPR ballot enables them secretly to reveal the values that led them to grade the candidates as they did. Their values are probably more aligned with the values displayed by a candidate they graded Excellent than a candidate they judged only to be Acceptable or Poor.

Consequently, analyses of each EPR-post-election report could publicly and probably provide everyone with much more information about the priorities of all their fellow voters than could be extracted from studies of all the post-election reports of any other known voting system – to inform everyone about the exact numbers and intensities with which all voters are aligned with the values expressed by each candidate. In this way, the qualitative advantage of using grades by EPR voters would seem to enable everyone to learn more about the realities of their society than is offered by any other election method. This benefit is also made more likely by the fact that EPR removes incentives to vote tactically or strategically almost completely.

As a result, 100% of the voters are equally represented quantitatively in the legislative body through the weighted voting of the member they helped to elect. Qualitatively, the voter’s grade for that winning candidate optimally assures the voter and a researcher of the extent to which the values of this winner are aligned with the voter's. These benefits seem not to be delivered by any other type of PR system currently used in the world. EPR seems optimally to respect the democratic principle that each citizen’s vote should count equally.

Exactly how an EPR election would be counted is provided by the EPR Count: Detailed Description link listed under “Supplementary Materials in GitHub” located on the last page of our published article: "Legislatures Elected by Evaluative Proportional Representation (EPR): an Algorithm"https://www.jpolrisk.com/legislatures-elected-by-evaluative-proportional-representation-epr-an-algorithm-v3/)]

P.S. When reading this 2020 article, please replace "proportionality" with "proportional." "Legislatures Elected by Evaluative Proportional Representation (EPR): an Algorithm"<nowiki>https://www.jpolrisk.com/legislatures-elected-by-evaluative-proportional-representation-epr-an-algorithm-v3/)