Instant Runoff Voting:  No Substitute for Proportional Representation

Douglas J. Amy

Mount Holyoke College

     Clearly the newest and “hottest” electoral reform idea in the United States right now is instant runoff voting or IRV.  Many electoral reformers have jumped on the IRV bandwagon and there are now several efforts to promote the adoption of this alternative voting system on the city and state level.  But while this voting system does has some advantages, it also has some serious limitations and drawbacks.  And whereas IRV is appropriate for single-office elections like mayor and governor, it is clearly not the best system for legislative elections.  The best system for those elections is proportional representation (PR), and an overenthusiastic effort to promote IRV may only make it more difficult to adopt PR.  This article first considers the case for IRV elections, and then explains why that system is a poor substitute for proportional representation.

  

The Promise of IRV

 

     Instant runoff is designed to address several of the problems of our current system of plurality voting, where the winning candidate is simply the one that gets the most votes.  In IRV, voters mark their preferences on the ballot by putting a 1 next to their first choice, a 2 next to their second choice, and so on.  A candidate who receives over 50% of the first preference votes is declared the winner.  Otherwise, the weakest candidate is eliminated and his or her votes are reallocated to the voters’ second choices.  This reallocation process continues until one candidate receives a majority of the votes.  

 

     This voting system has a number of obvious advantages over plurality voting.  First, IRV ensures that the winning candidate receives the support of the majority of voters.  That is not always the case in plurality voting.  Assume for instance that there are three candidates vying for office and the Republican receives 43% of the vote, the Democrat receives 40% and the Green candidate 17%. Under plurality rules, the Republican wins – even though the majority of voters opposed that candidate and actually voted for candidates on the left.   This example also illustrates another problem of plurality voting: the spoiler.   A spoiler is a minor party candidate that takes away enough votes from one major party candidates to ensure the election of the other major party candidate, who would not have won otherwise.  In the case above, those who voted for the Green candidate inadvertently helped the Republican candidate win.

     Instant runoff voting eliminates both minority winners and spoilers.  In the above example, since neither major party candidate received over 50% of the vote, the weakest candidate – the Green – would be eliminated and his or her votes transferred to their second choices.  Assuming 11% of the Green vote goes to the Democrat, that candidate would win with 51% of the vote.   In this way, IRV ensures a winner supported by the majority of voters, and also that votes for minor party candidates do not inadvertently aid in the election of a candidates those voters want the least.

     It was Ralph Nader’s role as a spoiler in the 2000 presidential elections that sparked much of the growing interest in instant runoff voting.  Nader came in for a lot of criticism for siphoning off enough votes from Al Gore in Florida to enable George W. Bush to take that state and the presidency.  But reformers have suggested that it is unfair to castigate candidates like Nader when it is clearly the plurality voting system that is to blame for making such spoilers possible.   It is the plurality voting system that is the spoiler, they say, not the candidate. In any case, the Nader candidacy did help to highlight the problem of spoilers and to spur interest in ways to avoid this problem – such as instant runoff voting.

Interest in IRV vote has also been growing in states that have been experienced this spoiler problem.  In New Mexico, a particularly strong Green party has taken away votes from Democratic congressional candidates and assured the election of Republicans in traditionally Democratic districts.   In Alaska, conservative minor party candidates have been making it more difficult for Republicans to get elected.  Not surprisingly, a bill was recently introduced in New Mexico to switch to alternative voting for state elections.  In Alaska, a referendum to use the alternative vote for federal elections will be on the 2002 ballot. In addition, bills to adopt the alternative vote have been introduced in a dozen other states. At this point it is difficult to assess political prospects of these bills, but the very fact that this many states are even considering this new voting system attests to its growing appeal.  It is also significant that a major national newspaper, USA Today, recently gave a strong endorsement to instant runoff voting on its editorial page.

 The Problems of IRV

Serious questions must be raised, however, about this rush of promote the adoption of IRV.  This reform is clearly a good idea for single-office elections, such as major, governor, etc.   In those elections, IRV would be an improvement over plurality voting. Unfortunately, however, some politicians and election reformers have also proposed that IRV be used for legislative elections, instead of other electoral reforms like proportional representation (PR). The idea would be to keep single-member districts, but use IRV instead of plurality voting to fill the seats in the legislatures.  But using IRV for city councils, state legislatures, or even Congress is a bad idea – for several reasons.

 First, instant runoff voting is a very poor substitute for proportional representation.  It offers very few political benefits compared to using PR for legislative elections.  Although it has a few advantages over plurality voting, IRV is still a winner-take-all system and so is prone to all the other serious drawbacks of these systems.  And as one veteran electoral scholar, professor Wilma Rule, has observed, there are several important things that IRV does not do – but PR does. “[IRV] is a majority system which leaves out the political minority especially women and ethnic minorities, and third and other small parties.”  Thus IRV does nothing to help solve our voting rights problems, or aid in the election of more women.  Nor does it ensure fair and accurate representation for all parties, including minor parties, as PR would.  IRV slightly reduces but does not eliminate most of the enormous numbers of wasted votes in plurality elections.  It also does not produce multiparty legislatures that truly reflect the variety of political views in the electorate. Finally, unlike PR, IRV eliminates none of the problems associated with redistricting, such as uncompetitive districts and partisan gerrymandering.  In short, in legislative elections, IRV is not much better than plurality elections; and as a winner-take-all system, it remains grossly inferior to PR.  Adopting proportional representation elections would bring a number of badly needed changes to American elections and American politics – adopting IRV would not.  In a sense, using IRV for legislative elections is like putting a new set of seats into an old, run-down car, while doing nothing to address its worn out engine and malfunctioning transmission.  The car will feel a bit better to ride in, but it will still not run well and will continue to break down.

If instant runoff voting is so inferior to PR, why is it being so enthusiastically promoted for legislative elections?  One reason is that it seems to be an easier reform battle to win.   IRV is simpler to explain to policymakers and the public than the relatively complicated processes of proportional representation.  Moreover, it is easier to get support for IRV from Republican and Democratic lawmakers.  While PR is often seen as a direct threat by the major parties, the alternative vote can actually work to the benefit of these parties by eliminating the problem of spoilers.  For instance, in New Mexico, some Democratic lawmakers want IRV so that Green candidates can no longer act as spoilers and allow Republicans to win what have been traditionally Democratic districts.  In Alaska, it is the Republicans that have been hurt by minor party spoilers and are now interested in this reform.

But the support of these politicians simply demonstrates how weak a reform IRV actually is.  Major party politicians can embrace IRV because it is relatively harmless and does not pose a threat to their dominance, or to the two-party system.  It will not allow the routine election of minor party candidates.  Indeed, it will primarily help the major party candidates because they will no longer have to worry about spoilers.

Minor parties would be marginally better off under IRV than under plurality rules.  They would be able to attract more votes for their candidates and this could help to increase their political legitimacy.  But in the end, such parties would remain largely excluded from the governing process.  They still would get neither seats nor any real power in our legislative institutions.

Some IRV proponents acknowledge its inferiority to proportional representation for legislative elections, but maintain that it can serve as a “stepping stone” to PR.  As one activist explained: “Although IRV is a majoritarian system that would not significantly improve the representation of political minorities, it represents a substantial improvement over plurality and runoff elections, and it helps pave the way for future PR efforts.”  It is thought that once people are used to ranking candidates in IRV, it will be easier to convert them to the single transferable vote or other forms of PR.  Perhaps – but this may be wishful thinking.  There is no empirical evidence that adopting the IRV aids in further electoral reforms.  In addition, it is unclear how reformers would be able to easily shift political gears after the adoption of IRV and begin to promote a change to PR.  They would have to turn from enthusiastic supporters of IRV to severe critics of that system – an inconsistency that would surely not go unnoticed by the public, the media, and policymakers.

Even worse, the adoption of instant runoff voting for legislative elections could actually set back the PR movement in the US, or at least distract from it.  Resources in the electoral reform movement are limited, so any effort for IRV must come at the expense of promoting PR.  In addition, using instant runoff voting to eliminate spoilers and ‘fix’ some of the problems of plurality elections could give politicians an excuse for not considering other reforms such as proportional representation.  Indeed, in electoral reform efforts in New Zealand during the 1990s, IRV became the preferred choice of many of those who opposed PR and wanted to maintain a winner-take-all voting system.

For all of these reasons, it would probably be best if the advocacy of IRV were confined to the area of single-office elections.  Working for “better” winner-take-all elections is only appropriate when no other alternatives are possible.  It makes little sense in legislative elections, where much better alternatives – such as proportional representation – are available.   Advocating IRV’s use in legislative elections thus might prove to be a strategic error for the electoral reform movement.  The struggle to adopt PR – and all the major political benefits that it will bring – is difficult enough without making it harder by promoting IRV for these elections.  Those interested in genuine electoral reform for legislative elections would be better off keeping their “eyes on the prize,” and continuing to press for the adoption of proportional representation elections in the United States.