Should We Choose Ranked Choice Voting?

Approval Voting: Works Great, Less Complicated

Response Essays
December 7, 2016

Thomas L. Knapp recommends the use of approval voting, in which candidates simply signal their approval of as many candidates as they find acceptable, without ranking them; the candidate with the most votes then wins. Knapp’s case for approval voting is that it is simple and easy to understand, and that it will produce fewer headaches for voters than Ranked Choice Voting. In the universe of possible systems, none are perfect, but simplicity is a factor that should not be discounted entirely: Simple elections encourage turnout and foster confidence in the legitimacy of the results. Whatever problems our democracy may have, Knapp concludes by arguing that voting method is perhaps not so high on the list.

Nearly a quarter of a millennium into the American experiment in constitutional representative democracy, I am far more surprised the system has survived for so long than that it suffers from numerous problems. I have lived through less than a quarter of the history of U.S. elections, and heck, I’m falling apart already.

A basket of methods intended to serve a population of not much more than 3 million now serves a population of more than 300 million. Yes, it has evolved. In 227 years we have gone from suffrage only for white male landowners to “universal” (okay, not really - ask a convicted felon) adult suffrage. We have gone from hand-written or party-printed ballots to a government-printed “Australian” ballot with candidate access controlled and limited in each state by the ruling parties. U.S. Senators, once chosen by state legislatures, are now chosen by popular vote. And so on.

Replacing plurality voting with Ranked Choice Voting (RCV) - as lead essayist Rob Richie recommends - or with some other new voting system, strikes me as more in the nature of a knee replacement than a heart transplant, but hey, why not? I’m a market anarchist myself, but if we are going to take democracy seriously, I suppose there is something to be said for continuously tweaking technical voting methods to most accurately reflect real voter preferences. Along the way we may still do well to recall H. L. Mencken’s conception of democracy as “the theory that the common people know what they want, and deserve to get it good and hard.” Democracy by my lights is never going to be perfect.

Rob makes two especially strong arguments for RCV:

  • It provides for majority rather than plurality winners. Even if there is nothing magical about majority versus plurality vis a vis “the popular will,” there is a longstanding and visceral appeal to majority rule.
  • It eliminates the need for separate “runoff” elections when a majority is required and no candidate receives one.

So far, so good. But I cannot approve of the complexity of RCV versus my own preferred voting alternative, Approval Voting.

Those who paid attention to the Florida presidential vote recount in 2000 probably recall quite a bit of controversy over the use of “butterfly ballots.” Many voters claimed to find those ballots confusing and difficult to understand. Some even claimed that in retrospect they had probably voted for Pat Buchanan when they meant to vote for Al Gore.

No, I’m not saying that voters are stupid, although some undoubtedly are. I am saying that those of us with an ongoing, even obsessive, interest in politics tend to forget that most American voters don’t spend a lot of time thinking about the technical aspects of casting a ballot. They show up every two or four years to spend an hour standing in line and five minutes in a booth filling out an unfamiliar form. The more complicated that unfamiliar form becomes, the greater the likelihood of error.

Some voters seem to have a difficult time looking at lists of candidates, picking one for each office to be elected, and accurately recording their choices. We can hardly expect those voters to do a very good job of looking at lists of candidates and ranking all those candidates on all those lists in order of preference. If they even bother to try, the ratio of noise to signal in results will presumably rise.

Some voters will blithely choose one candidate, unaware that they should rank their choices. Others will get the numbering system or other measuring mechanism backward when ranking their choices and end up weighting the power of their votes in favor of the candidates they like least instead of most. Human sacrifice, dogs and cats living together… mass hysteria! Well, okay, maybe not that last part. But if the goal is to align outcomes with real voter preferences, there are real potential defects in RCV as the method of doing so.

Approval Voting is simpler. Here is how it works:

The voter votes for all the candidates he or she likes.

And that’s it! Let’s consider an example. Donald Trump, Hillary Clinton, Gary Johnson, and Jill Stein all appear on the ballot for president.

The voter is okay with Clinton or Stein, but not with Trump or Johnson, so the voter votes for Clinton and Stein.

Or the voter is okay with Trump or Clinton, but not Johnson or Stein, so the voter votes for Trump and Clinton.

Or the voter is only okay with Johnson, so the voter votes only for Johnson.

The candidate who receives the most votes wins. Easy peasy. One variant of Approval Voting requires that the winning candidate, in addition to receiving the most votes, must receive votes from a majority of voters.

Are there problems with Approval Voting? Sure. In the majority requirement variant, a runoff might be required. RCV avoids that. Approval Voting can also be gamed. If I prefer Johnson but am okay with Stein, I might vote only for Johnson because I would rather have Johnson and don’t want to put any gas in Stein’s tank. And Johnson’s campaign might encourage his “base” voters to do exactly that.

But of course RCV can be gamed in the same way. A voter doesn’t have to rank all the available choices. In the above example she could rank Johnson first and rank no other candidates, so that only her vote for Johnson counts no matter the circumstances.

Simpler is better. If the purpose of a voting method is to align outcomes with voter preferences, simplicity matters, because complexity produces noise that interferes with our ability to understand those preferences. Is simplicity the only consideration? Of course not. But it’s an important one, and in my view Ranked Choice Voting’s virtues are not sufficient to offset its complexity versus Approval Voting.

A final thought: Maybe voting method isn’t really as earth-shakingly important as those of us who spend time thinking about it want it to be.

In his essay, Rob points out that “[e]xtensive data analysis from more than 125 RCV elections in the Bay Area show that … every single winner has been the ‘Condorcet’ candidate who would defeat all others head-to-head…” But if something was broken in Bay Area election outcomes, RCV didn’t obviously fix it, and it may have been that nothing was broken at all. To be fair, I’m not sure Approval Voting would have fixed anything here either. Perhaps we’re looking in the wrong place for better outcomes.

Hacking America’s Antiquated Elections

Lead Essay
December 5, 2016

Rob Richie of FairVote explains why he thinks ranked choice voting is better than the system that most of us now use. Rather than marginalizing all outsiders and shoehorning everyone into two big, unwieldy parties, a ranked choice system would tend relatively more toward the representation of multiple viewpoints and toward the enactment of all policies that could command a majority, even if they do not command a majority within a given party or division of the electorate. RCV isn’t just better for our politics, says Richie; it’s been tried successfully in a wide array of settings, and it deserves to be seen not as an experiment, but as a well-known alternative. Given the shortcomings of our current system, RCV looks better and better, he argues.

American democracy today is working more poorly than it has in generations. Even as the toxic 2016 presidential campaign featured the two most unpopular major party candidates in modern history and Congress has historic lows in approval, minor party presidential challengers were marginalized, and nearly 98% of congressional incumbents won re-election. New voices are demeaned as spoilers, which suppresses debate about innovative ideas and shoehorns our diverse political views into two fiercely partisan camps. With the overwhelming majority of elections predictably going to a district or state’s partisan majority, most voters lack meaningful choice even among two candidates. In conflict with the spirit of the Constitution, our electoral rules punish representatives who seek to govern outside their party boxes, blocking sensible changes that have majority support.

Absent reform, it is a near certainty that these problems will continue. No single change can unlock voters and spark a democracy where the best ideas rise to the surface and policymakers are able to implement the will of the people with respect for all. But this year we saw a true glimmer of hope for change: with 52% of the vote, Maine voters adopted ranked choice voting (RCV) for all their elections for governor, U.S. Senate, U.S. House, and state legislature in a campaign endorsed by the Libertarian Party, the Green Party, and hundreds of major party elected officials from across the spectrum. Starting in 2018, Mainers will be able to vote for the candidates they like the most without helping elect the candidates they like the least. They will earn what we all deserve: a fair vote and a truce in the battle over whether minor party and independent candidates can have an enduring seat at the electoral table.

Ranked choice voting (sometimes called “instant runoff voting” and “preferential voting”) is a proven voting method designed to accommodate having more than two choices in our elections. When used to elect one candidate, RCV essentially simulates the math of traditional majority runoffs, but in one trip to the polls. Voters have the freedom to rank candidates in order of choice: first, second, third, and so on. Their vote is initially counted for their first choice. If a candidate wins more than half the votes, that candidate wins, just like in any other election. If no candidate has more than half the votes, then the candidate with the fewest votes is eliminated. The votes of those who selected the defeated candidate as a first choice are then added to the totals of their next choice. This process continues until the number of candidates is reduced to two or the winner earns more than half of the active votes.

RCV upholds majority rule while accommodating increased voter choice. It creates incentives for winning candidates to reach out to all voters in order to get a higher ranking and allows a voter to consider more choices with a greatly reduced likelihood of “splitting” their vote in a manner that might otherwise result in an unrepresentative outcome. Based on the context of its use, RCV can mitigate partisan inflexibility, foster greater accountability for incumbents, increase civic engagement, and reduce the impact of campaign spending. When used in multi-winner elections, RCV becomes a candidate-based form of proportional representation that expands the percentage of people who elect preferred candidates, increases competition, and provides a natural means to elect more diverse legislatures that include accurate representation of the left, right, and center, as well as representatives who break free from the two-party box.

Maine’s victory was grounded in grassroots energy, effective organizing, and a well-run campaign. RCV had been debated in the legislature for years and been widely hailed as a success in mayoral elections in the state’s largest city of Portland. In the midst of yet another campaign for governor where the winner received less than half the votes – as has been the case in call but two gubernatorial elections since 1974 – reformers seized a change to launch an initiative campaign. With barely a week to organize, Election Day volunteers collected more than half the signatures required to put it on the 2016 ballot. The Committee for Ranked Choice Voting and its allies, like the League of Women Voters of Maine and FairVote Maine, launched a two-year campaign of education and advocacy that resulted in more than 300 published letters to the editor, more than 175,000 one-on-one conversations about RCV with Mainers, nearly 3,000 donations from Mainers, and community presentations across the state. A surge of funding allowed for television and digital media that helped push the measure over the top despite being a new idea to most voters.

RCV also won in a local campaign in Benton County, Oregon. These wins and more than a dozen other victories for RCV in cities since 2000 demonstrate that RCV is politically viable and impactful in practice. Cities using RCV for mayor and other local offices include Minneapolis (MN), St. Paul (MN), Oakland (CA), San Francisco (CA), San Leandro (CA), Takoma Park (MD), Telluride (CO), and Portland (ME), while Cambridge (MA) has used RCV to elect its city council and school board for decades. Cities awaiting implementation after voter approval include Memphis (TN), Santa Fe (NM), and Sarasota (FL). Internationally, RCV has been used for years to elect Ireland’s president, Australia’s House of Representatives, and the mayors of London (UK) and Wellington (New Zealand). With recommendations by procedural guides like Robert’s Rules of Order, RCV is widely used in nongovernmental organization elections, ranging from major private associations like the American Chemical Society and American Psychiatric Association to nearly every major party in Australia, Canada, Scotland, and the United Kingdom, as well as Republican and Democratic parties in Iowa, Maine, Texas, Utah, and Virginia. Young people have adopted RCV for their student elections at some 60 American colleges and universities and are the most like to support it on the ballot.

RCV’s track record in those elections is impressive. Although still a winner-take-all system that isn’t designed to elect those with minority views, RCV gives everyone a fair shot to run. Australia typically has more than six candidates per house race, and the strongest minor parties run in every district without any fingerpointing or talk of spoilers. Instead, they can make their case, see the best of their ideas adopted by the major parties, and grow their vote such that these parties are now winning fair shares of seats in senate elections held with the multi-winner proportional representation form of RCV.

In city elections in the United States, there has been a string of open seat elections where the best-financed favorites run traditional campaigns focused on their base and lose to enterprising challengers who engage directly with more voters in grassroots campaigns designed to earn not only first choice support, but second and third choice support from backers of other challengers. The pattern seems to be that the best-financed candidates rely on traditional techniques of identifying their stronger supporters, getting them to vote, and going more negative on other candidates – and the best challengers can win by putting more effort into direct voter contact regardless of first choice support.

Extensive data analysis from more than 125 RCV elections in the Bay Area shows that (1) every single winner has been the “Condorcet” candidate, or the one who would defeat all others in simulated head-to-head contests, even though several winners trailed in first choices and one winner initially was in third; (2) voters regularly rank more than one candidate, including close to nine in ten voters in competitive mayoral elections; (3) fewer voters now skip city elections when at the polls for president and governor; (4) voter turnout in decisive elections has on average risen sharply from prior systems with primaries and runoffs; and (5) and more than 99% of voters cast valid ballots, which is often higher than their valid ballot rate in other races with large candidate fields.

RCV’s promise and track record have helped earn notable support. American political leaders backing RCV include President Barack Obama (prime sponsor of RCV legislation as an Illinois state senator), Sen. John McCain (recorded a robo call in support of a ballot measure to implement RCV), former Vermont governor Howard Dean (author of several pro-RCV op-eds, including in the New York Times this fall), former Republican Congressman John Porter (author of a piece in a Brookings Institution report on policy proposals), Sen. Bernie Sanders (who testified on its behalf to the Vermont state legislature in 2007 on a bill that passed the legislature) and this year’s presidential nominees for the Libertarian Party (Gary Johnson) and Green Party (Jill Stein).

 

Ways to Expand Use of Ranked Choice Voting

Ranked choice voting is imperfect, but perfection is literally impossible – and advocates of other, untested systems should be cautious about overstating their potential absent experience. But RCV is viable, legal, and successfully tested as a flexible tool for addressing problems in our elections. Once it becomes easy for all jurisdictions to use, as is likely within the next four years, both legislators and populist reformers will find RCV to be valuable. With each new advance, voters’ conceptions of what it means to vote will change from marking an “X” to ranking choices. The RCV ballot has drawn support in several different contexts, including the following.

Replacing plurality voting: The great majority of American elections are held with plurality voting, where candidates with the most votes win, even if they do so with less than half the votes. As Maine showed, voters are ready to support RCV when they are frustrated by elections that mean either having to vote for the lesser of two evils, or else for unrepresentative winners.

Replacing runoff elections: Holding a separate runoff between the top two finishers is a means to eliminate “spoilers.” But runoffs have downsides. The strongest candidates may not reach the runoff due to split votes. Runoffs exacerbate demands for campaign contributions and often have disparate voter turnout between elections. More than 96% of the nearly 200 regularly scheduled congressional primary runoffs since 1994 experienced declines in turnout, with an average turnout decline of more than 30% – a far steeper decline than the number of voters who don’t rank finalists in RCV races. Finally, runoffs increase election costs and burdens on voters, making them an easy target for budget-cutting policymakers. These problems explain why more than a dozen cities have voted to replace runoffs with RCV.

Replacing problematic means of nominating candidates: Traditionally, parties used conventions to choose nominees, which ensured nominees were accountable only to the parties’ most active members. But the main alternative, the primary system, has unrepresentative turnout, with steadily declining percentages of Americans registering with a major party.

RCV can help solve problems associated with nominating candidates. RCV could be built into the major party presidential candidate nominating processes, starting with party-run caucuses, and RCV could be used more generally to ensure nominees for all offices earn greater support. More dramatically, states could stop paying for primaries entirely and use RCV to accommodate voters having more general election choices among independents and party nominees.

One form of RCV is drawing particular attention: modifying the Top Two primary to advance four candidates, with RCV to be used in November. As used in California and  Washington, Top Two establishes that all candidates seeking an office run in the same primary contest, and the top two finishers face off in November regardless of party. FairVote’s analysis of California’s 2012 congressional elections found that advancing four candidates to an RCV contest in November would nearly triple the number of general election races with third party or independent candidates and more than quintuple the number of general elections with more than one candidate from the same major party.

Opening up legislative elections to better choice and fairer representation: The combination of winner-take-all rules and rising partisanship has led to a sharply rising percentage of districts in which only one party has any real prospect of winning, and more legislatures where one party has a lock likely to last for generations. It has entrenched incumbents, depressed participation, promoted unrepresentative homogeneity within parties, and created barriers for women, racial minorities, and minor parties to win more seats. Redistricting alone has limited impact on these problems, as suggested by distorted partisan outcomes in California and not a single congressional seat changing hands in 2016. Truly unlocking democracy depends on adopting RCV in multi-winner elections. The first step is to have larger districts with more voters and more seats; for example, one might combine five adjoining districts into a larger district with five representatives. These would be chosen by RCV, with the percentage of the vote necessary to win declining in relation to the number of seats in the district – about 17% of like-minded voters being able to elect a candidate in a five-winner district.

Multi-winner RCV is used in at least one governmental election by every voter in Australia, Ireland, Malta, New Zealand, Northern Ireland, Scotland, Minneapolis (MN), and Cambridge (MA). FairVote’s congressional election simulations show that not a single voter in a state with more than two representatives would be represented by only one party. Congress would have a far broader mix of perspectives. New opportunities would arise for independents and third parties to hold the major parties accountable, and more cross-cutting representatives would be likely to forge compromises. Expect to see the Fair Representation Act based on this form of RCV introduced in Congress next year, and for more cities and states to consider it.

Looking forward, American politics is at a tipping point. Our current system simply isn’t working, and all trends suggest it will keep getting worse. Maine shows that voters are ready for change, and reformers are planning city and state campaigns for RCV across the nation. Now is the time to think big – and rank the vote.

 

Coming Up

Essays by Jason Sorens, Thomas L. Knapp, and Jason McDaniel. Conversation through the end of the month.