In Book II of his Politics, Aristotle critically assesses the account of an ideal political constitution set forth by one Hippodamus, a natural scientist described as having no background in political affairs but evidently aiming to impose on the political world the same kind of mathematical regularity that is found in nonhuman nature (say, the annual changes of seasons). As Aristotle demonstrates, Hippodamus’s proposals — among other things, he had a penchant for wanting to divide things into threes (three classes of citizens, three kinds of land, etc.) without any practical reason for doing so — would self-destruct in practice, since he failed to take into account the nonmathematical realities of human behavior and motivation. Hippodamus also wanted the city to reward innovators like himself. Among his ideas, one stands out for its oddity, since it embodies the sort of abstract political morality to which intellectuals — in our time, philosophy professors like John Rawls and Robert Nozick — are prone. To simplify, Hippodamus objected to the fact that in civil or criminal trials, individual jurors were required to vote either “innocent” or “guilty,” rather than allowed to specify the degree of a defendant’s guilt or debt. Requiring them to choose between only two verdicts, he contended, compelled jurors who thought a defendant only partly guilty to “perjure themselves.”
As Aristotle points out, Hippodamus’s proposed juridical reform would make it impossible for a jury to render a collective verdict at all: How would the various individual verdicts be summed? Nor, Aristotle observes, does requiring a juror to choose between only two verdicts entail making him commit perjury, since he wasn’t asked how guilty he thought a defendant was (or how much he owed) but only whether he had or hadn’t committed a crime or owed a debt to the plaintiff.
All this is relevant to understanding and assessing the argument for Ranked Choice Voting (RCV), a system for counting election ballots that has recently been adopted by the state of Maine, is already on the books in some dozen cities nationwide, and is currently on the ballot in the state where I live, Massachusetts, as well as in Alaska. (In Massachusetts, the system would apply to primary and general elections for the U.S. Senate and House, along with both branches of the state legislature and elections for executive offices from the governor on down to certain county officials.) Under RCV, voters would no longer be constrained to vote for a single candidate for a given office (as jurors are required to issue verdicts of either “guilty” or “not guilty”). Instead, they are offered the opportunity to rank up to five candidates in order of preference. Then, if no candidate wins more than 50% of the vote on the first ballot, the candidate with the fewest first-place votes is eliminated, and ballots of voters who chose that candidate are reallocated to other candidates based on their second-place choices. If that pushes another candidate over the 50% mark, he wins. If not, the process of reallocating the votes of the remaining lowest-ranked candidate continues until one candidate, out of a maximum of five, wins.
RCV advocates make two related but partly inconsistent arguments. On the one hand, they maintain, it would eliminate the so-called spoiler effect, in which votes for a third-party candidate determine the outcome of an election by disproportionately taking away votes from one of the major-party candidates in a way that alters the election outcome. For instance, it is alleged, on the basis of exit polls, that more voters for third-party candidate Ralph Nader in the 2000 Florida presidential election would have preferred Al Gore over George W. Bush, so that if all voters had had the opportunity to record a second choice, the Florida election—and, in consequence, the national one—would have been won by Gore. In a rather different instance, in the 1994 Maine gubernatorial election, independent Angus King (now a U.S. senator) won office with only 35.37% of the vote, beating out both Democratic candidate Joseph Brennan (33.83%) and Republican Susan Collins (23.08%). Arguably, given King’s narrow margin of victory, had RCV been in place, a sufficient number of his voters might have given Brennan their second-choice vote, enabling him to win the election.
But neither of these examples demonstrates any departure from proper democratic procedures. One can safely presume that Nader voters in the 2000 election were well aware that choosing Nader might entail decreasing the likelihood that their favored candidate between the major-party contenders would win; but they decided that Nader was such a superior candidate (even if he was unlikely to be victorious) as to merit that risk — or, more simply, that “making a statement” on behalf of Nader was more important than contributing to a victory by Gore over Bush, or vice versa. Nor is there anything objectionable in the fact that King outdrew the two major-party candidates — just as Lisa Murkowski, having lost the Republican Senate primary in Alaska in 2000, went on to win her state’s election as a write-in candidate, with a relatively small plurality over the official Republican candidate, when the combined votes of that candidate and the Democratic contender considerably exceeded her own. In fact, both the King and Murkowski elections demonstrate that third-party candidates aren’t necessarily “spoilers” but, given sufficient voter appeal, may actually win.
A connected but somewhat inconsistent claim by advocates of RCV is that it eliminates “wasted” votes — votes for a single candidate who loses an election — by still giving voters a chance to have their preferences “matter” (unless the winning candidate wins an absolute majority) in the final outcome. Supposedly, according to RCV advocates, this prospect will increase voter turnout. And, according to a mailing from Massachusetts’s Ranked Choice Voting 2020 Committee, it will encourage candidates “to seek support from all of us,” generating “more positive and issue-focused campaigns,” since candidates will have a disincentive to disparage rival candidates, lest they lose the support of those favoring lower-ranked candidates, which they may need in successive stages of ballot counting.
In fact, as Boston Globe columnist Jeff Jacoby has pointed out, RCV will make our elections less, rather than more, democratic. This is because it potentially disenfranchises the many voters who normally have little interest in ranking minor-party candidates but simply choose between one of the two major-party candidates and leave it at that. In consequence, such voters may discover that the candidate for whom they, along with a plurality of voters, marked their ballot nonetheless loses the election, once the votes of the second-place finisher are enhanced by those reallocated from the third- through fifth-place candidates.
The real agenda of advocates of RCV — which, in Massachusetts, has been heavily funded by out-of-state organizations (helping to generate $5.7 million in total funding, vs. $250 [sic] for RCV opponents, as of September 21, according to Ballotpedia.org) — can be gleaned from the list of its supporting groups. Aside from the Democratic — but not the Republican — Party, this list includes numerous Democratic elected officials (U.S. senators as well as several members of the state’s House delegation), several public-employee unions, the Libertarian Party, the Green-Rainbow Party, the Sierra Club, the League of Women Voters, and Common Cause, along with such obscure lobby groups (or groups specially created to promote RCV) as Amplify Latinx, the Humanity First Collective, Indivisible, Our Revolution, and RepresentUs. The list does include two Republicans, the only noteworthy one being the quirky ex-governor Bill Weld, a liberal member of his party who resigned midway through his second term to become ambassador to Mexico, only to later pursue election, unsuccessfully, as governor of New York and in 2020 threaten to challenge President Trump.
As this list makes clear, RCV supporters fall overwhelmingly into two (mostly overlapping) categories: Democrats and groups whose members vote heavily for Democratic candidates; and groups that (Libertarians aside) have practically no chance of winning elections even under RCV, except at the local level. Given Massachusetts’s status as a heavily Democratic state (the state’s congressional delegation consists solely of Democrats, who also have long held a substantial supermajority in the legislature), Democrats have little to fear from losing elections to Republicans as a result of RCV. Rather, they need to court members of left-liberal fringe groups, as well as public-employee unions, to ensure that they turn out to vote — knowing that those groups’ supporters would almost surely make the Democratic candidate, at worst, their second-choice candidate, further guaranteeing the defeat of any Republican contender. At the same time, RCV must be seen as a vehicle on the road to breaking up the two-party system, by incentivizing all sorts of partisan groups, from the relatively mainstream Libertarians and Common Cause to “Our Revolution,” to run candidates at all levels of American elections. Even if they have little hope of winning state elections, even with RCV, they can heighten pressure on the major parties (mostly the Democrats) to move their positions closer to those of the minor “parties,” probably to a degree disproportionate to their actual level of support but proportionate to the loudness of their leaders and the financing that they can obtain.
Political scientists commonly describe one of the essential functions of America’s two-party system as “interest aggregation.” This means that, in a large and pluralistic nation like ours, each party consists of a coalition of diverse interest and principle-based groups to which it must appeal, while luring voters away from the other party’s coalition. Yet at the same time it must persuade each group to moderate its demands for the sake of preserving and enlarging the coalition. Somehow the Democrats must retain the support of liberal Jews as well as Muslim critics of Israel; of philanthropic corporate CEOs and union members; of blue-collar workers and those who think that they are “deplorable”; etc. Republicans must pursue a similar process in order to win elections. This process corresponds to the scheme outlined in Federalist no. 10, according to which the need for members of Congress to compromise with members representing blocs of voters quite different from their own in order to get legislation that they favor enacted is likely to moderate their demands and offer a plausible excuse to voters in their district for not having fully satisfied their desires.
Transferring the function of interest aggregation from within the parties to the election itself, as RCV would do, with multiple parties usually battling for victory, threatens gradually to replace our two-party system with a multiparty one, in which no party commands a majority in either house of Congress (or in many state legislatures). While the Constitution, with its requirement that a winning presidential candidate gain an absolute majority of votes in the Electoral College, otherwise transferring the final vote to the House voting by state, prevents this from occurring when it comes to choosing our chief executive, the so-called National Popular Vote movement, which cannot constitutionally provide for a run-off election among the top two finishers, actually makes it likely that a candidate favored by only a small minority of voters would gain the office.
To understand why moving to a multiparty system would be detrimental to the national interest, whether one is a Republican or Democrat, liberal or conservative, one need only look to the long-chaotic postwar history of multiparty systems in countries such as Italy (which has averaged a new government every year or two since the restoration of democracy) or Israel (where, with the two leading parties closely balanced, the small Orthodox parties, commanding only a few votes in the 120-member Knesset, are able to blackmail the governing party into making concessions that are opposed by a large majority of voters, lest the Orthodox switch sides and unseat the current government).
Beyond this potential problem, the RCV movement rests on a fundamental misconception of voting, alien to the American Founders and leading statesmen who followed them, as a matter of expressing individual “preferences” (as the flyer distributed by the Massachusetts campaign puts it), as if one were choosing among hairstyles or brands of toothpaste, rather than engaging in rational deliberation about the public good. Altering the voting process to accommodate the feelings of Our Revolution or Indivisible (whoever they are) threatens to enhance the ideological fracturing of our electorate (already, sadly, on the increase), enhance the power of unabashedly self-interested groups like Amplify Latinx and the public-employee unions, and weaken a sense of common citizenship and public duty. As already noted, far from strengthening constitutional, democratic governance, RCV will tend to disenfranchise voters who are not ideologues and who have little interest in taking the time to assess the claims of small competing groups. (Who wants to spend time even trying to find out what Our Revolution’s program is?) That effort of assessing the merit of voter demands and their degree of popularity is what our major parties are for. It is then up to voters, just as with jurors, to decide which side seems more likely to promote the common good, as well as their own — while understanding that their wishes will never be fully satisfied. That’s what comes from being political animals who must get along with their diverse fellow citizens while appreciating the need to obey common rules.
Anyone who cares seriously for the preservation of constitutional, democratic government in America should vote against RCV and persuade everyone he can to do likewise.
David Lewis Schaefer is professor of political science, College of the Holy Cross.