There has been misinformation bouncing around on the Internet about instant runoff voting (IRV) in the guise of "researched analysis" that claims to show it has serious flaws. By and large, such "analysis" tends to be faulty, selective, agenda-driven manipulation. We have attempted to address many of these false claims and myths on this FAQ page.
Table of Contents
VOTER EXPERIENCE UNDER IRV
1.1 Is instant runoff voting simple for voters to use?
1.2 Is instant runoff voting prone to strategic voting?
1.3 Do voters like using IRV?
1.4 Has IRV been repealed in some jurisdictions?
IMPACT OF IRV
2.1 Does IRV address the "spoiler" problem as well as separate runoff elections do?
2.2 Does IRV favor extremists?
2.3 Is IRV fair to racial minorities?
2.4 Does IRV unfairly favor third parties?
2.5 Does IRV entrench the "duopoly" two-party system?
2.6 Does IRV undermine political stability by encouraging fringe candidates?
CONSTITUTIONAL ISSUES
3.1 Does IRV comply with the one-person-one-vote principle?
3.2 Is IRV constitutional?
LOGIC AND FAIRNESS OF IRV
4.1 Aren't plurality winners okay, even if they don't get a majority?
4.2 Is IRV a proven and accepted voting method?
4.3 With sequential elimination of candidates from the bottom, do the supporters of the weakest candidate get extra clout by getting their second choice counted before the second choice of other voters?
4.4 Is every voter treated equally under IRV?
4.5 Is IRV better than using traditional separate runoff elections?
4.6 Does IRV elect true majority winners?
4.7 Are any other alternative voting methods better at solving the "spoiler problem" than IRV?
4.8 Does IRV violate important election fairness criteria?
4.9 Is the so-called "winner-turns-loser" or "monotonicity" issue have any real-world significance?
4.10 Will IRV generally elect the "Condorcet winner?"
4.11 Is ranking candidates in order of preference a more valid procedure than making up "scores" as in Range Voting?
ADMINISTRATION OF IRV ELECTIONS
5.1 Is IRV too complicated for election administrators to handle?
5.2 Does IRV need to cost more or require buying new voting machines?
5.3 Can IRV elections be counted manually without much additional time or effort?
5.4 Opponents of IRV have quoted a fiscal note for an IRV bill before the Maryland legislature in 2006 estimating a high cost for implementation. Is this estimate valid?
5.5 Can current voting machines handle IRV ballots without improvements to hardware of software?
5.6 Is federal certification of voting machines that can handle ranked-voting an insurmountable challenge?
5.7 Does IRV require centralized counting of ballots?
5.8 Does IRV mean moving to DRE touchscreen voting machines?
5.9 Does IRV weaken election integrity?
5.10 Can IRV be manually audited?
5.11 Does IRV delay announcing election results?
5.12 Opponents of IRV have asserted that using IRV in Pierce County cost over four dollars per registered voter. Can that be true?
VOTER EXPERIENCE UNDER IRV
1.1 Is instant runoff voting simple for voters to use?
Yes. The voters' task is very simple. Voters can simply mark their ballots in exactly the same way as they always have in the past. However, the voter has the option of ranking alternate choices, in case there is no majority winner and the voter's favorite candidate doesn't make it into the final runoff count. Since a vote for a minor candidate won't be wasted, as long as the voter ranks other choices, the voter can generally avoid the conundrum of voting for a favorite or a lesser evil. For many voters this makes voting with IRV easier than under the current plurality method.
Several U.S. cities that adopted IRV in recent years have had exit polls to assess voter acceptance of their new system. In every city where such a scientific survey was done, voters without exception have favored IRV over the old method. Also, studies of IRV elections have demonstrated at most very small changes in uncountable ballots (spoiled or skipping the IRV race) with the adoption of IRV. In the Burlington IRV election in 1999, for example, 99.99% of ballots cast in the IRV race for mayor were valid. People had no difficulty voting, and there were almost no "spoiled" ballots.
Two nations with the highest voter participation rates in the world, Australia and Malta, both use instant runoff voting. The only "complicated" aspect of instant runoff voting is the tabulation that occurs if there is no initial majority winner. But the voters don't need to absorb these details. A voter can dial a telephone without understanding the intricacies of the internal electronics or vote for president without understanding the intricacies of the Electoral College, which has a provision for election by the House of Representatives when no candidate has a majority.
1.2 Is instant runoff voting prone to strategic voting?
No. There is another side to voting that makes IRV simpler than most other voting methods, including systems that might at first seem more simple such as vote-for-one plurality voting. While every voting method is potentially subject to manipulation by strategic voting in some situations, IRV is uniquely resistant to such strategies (see James Green-Armytage analysis for more) Under most voting methods, a potentially beneficial voting strategy can be recognized by at least some voters (who may gain an advantage over other voters). Thus, voters may face a dilemma deciding whether to engage in strategic voting.
For example, in a simple plurality election there is the "lesser-of-two-evils" problem where voters often realize that voting for their favorite candidate, who is likely to get very few votes, may deny their second choice candidate enough votes to beat out that voter's least preferred choice. Such a voter might find the plurality election decision to be extremely difficult. Under IRV, this voter's ballot will automatically transfer to that second choice candidate if the first choice is at the bottom, and the second choice candidate can win with that transferred vote. This makes the voter's task much simpler with IRV. The default strategy is always to honestly rank candidates in the order of choice. Of course, voters may still face strategic dilemmas with IRV in some rare situations, but these are far less common than under most other voting methods. For example, under Range Voting and Approval Voting, giving any support to a second choice can cause that voter's first choice to lose (violation of the Later-No-Harm Criterion), thereby leading most voters who understand the system to strategically truncate their true preferences to gain an advantage over less informed voters. IRV complies with the Later-No-Harm Criterion, and is thus immune to such strategic calculation.
1.3 Do voters like using IRV?
Yes. Voters prefer IRV to the previous election method, according to exit polls conducted in U.S. cities that use IRV. In every case of a scientific poll, a substantial majority of voters preferred IRV compared to those who preferred the old system. Excluding those who said the methods were equal or who had no opinion, the gap between those favoring IRV and those favoring the old system ranged from 71% vs. 29% in the lowest city to 89% vs. 11%.
[More information and links to the various exit polls]
The City of Cary, NC used IRV in a pilot program authorized by the state. Both election officials and voters were extremely pleased with IRV.Cary conducted a general public opinion survey of residents, and included some questions about IRV. Residents favored IRV over the former method by huge margins. The survey results can be viewed by clicking here: (note that this survey was of residents, not just voters, and so included many people who had no exposure to voting place voter education.)
1.4. Has IRV been repealed in some jurisdictions?
Yes, but not many. In the 21st century, voters in at least seventeen American jurisdictions have voted for charter amendments or statutory changes to move to IRV. Voters in three jurisdictions have passed charter amendment to replace it with more traditional two-round runoff systems: Pierce County (WA), Aspen (CO) and Burlington (VT). Each had their own contexts, but it's important to underscore that none went to a plurality voting system; all decided to uphold majority rule using a two-round runoff approach.
Some opponents like to list other repeals, but are inaccurate. For instance, Cary (NC) voted by 4-3 to use IRV in a one-time pilot program in 2007. Two post-election surveys found that most voters preferred the new system, but the default was to use a two-round runoff system in subsequent elections.
They even like to keep saying Georgetown University has repealed IRV because they happened to come across a news story saying it did. But a few weeks later, Georgetown students restored IRV and have used it ever since, along with some five dozen additional colleges and universities? Will all those schools keep IRV? Probably not. But we expect the great majority of them will keep it and support it. That's the same with IRV in city elections.
IMPACT OF IRV
2.1 Does IRV address the "spoiler" problem as well as separate runoff elections do?
Yes, but better. IRV solves the spoiler problem better than separate runoffs and at least as well as any alternative voting method. A "spoiler" is a negative term for a minor candidate with not chance of winning the election. Under plurality rules such "spoiler" candidacies can throw an election to a candidate the majority of voters oppose. IRV tackles the "spoiler problem" in a manner similar to a traditional two-election runoff, except that, due to its sequential elimination procedure, it is more effective at eliminating the spoiler effect.
Under a traditional two-election runoff, if there are several candidates in the first round who have similar policy views, it is possible for them to split the majority in such a way that none of these similar candidates reaches the runoff round, while two candidates with substantial core support but very narrow appeal advance. Under IRV, voters who divided among these similar candidates will see their votes automatically re-unite for the strongest candidate in the group, and thus advance to the final runoff. Some advocates for other alternative voting methods have claimed that IRV does not solve the "spoiler problem" 100%. IRV solves the problem as well or better than the alternatives.
[Examples of spoiler scenarios under Plurality, Range, Approval and other voting methods]
2.2 Does IRV favor extremists?
No. Current plurality voting is far more likely to elect extremist candidates than is IRV. In a plurality election with several candidates, a candidate does not need any support beyond his or her ideological core supporters to get the "most" votes - even if that is a relatively small percentage of the voters. With IRV, a candidate must be able to garner both strong core support and broad appeal in order to win. With more than 80 years of use in Australian elections for the House of Representatives, IRV has proven that extremists are not benefited by IRV.
At the other extreme from favoring extremists, some alternative voting reform proposals (such as Condorcet, Approval, and others) do not strike IRV's desirable balance of core and broad support, in the name of favoring "centrist" candidates. These methods do not necessarily favor "centrist" candidates, but may favor inoffensive and merely unknown candidates (in the event voters are sincere -- if, as is more likely, voters increasingly engage in tactical voting, the outcomes will be quite erratic). They also allow a "compromise" candidate to be elected even if not a single voter considers him or her to be the best choice. These other methods allow a candidate who would not get a single vote under the current voting method, to win. For some offices, such as a club treasurer, this "compromise" bias may be desirable. But in high-stakes elections of political leaders, these methods may discourage candidates from revealing their stands on controversial issues, as they seek to avoid alienating any voters at all.
[More on Australian Elections]
2.3 Is IRV fair to racial minorities?
Yes. IRV repeatedly has been used effectively by racial minorities, but it neither hinders nor promotes the election of racial minorities. To generally advance the cause of representation for racial minorities, forms of proportional voting are needed, rather than any winner-take-all method such as IRV and nearly all other election methods in the U.S. However, by focusing the decision on a single election, rather than a two-election runoff process, IRV does tend to enhance voter turnout and success of racial minorities in the decisive election. In a multi-candidate IRV election, candidates have an interest in appealing to the supporters of other candidates for second preferences, and there is some risk that appeals to racism may backfire by alienating potential second preferences. However, in a head-to-head separate runoffs, demonization of opponents and appeals to racism are often an effective strategy.
Exit polls by San Francisco State University and analyses by the New America Foundation and FairVote of San Francisco's IRV elections found that racial minorities used the ranked-ballot effectively, and preferred it to the former election method and that IRV dramatically increased voter turnout by low-income and racial minorities in the decisive election round. In 2005's citywide election, IRV increased turnout in san Francisco’s six most socio-economically diverse neighborhoods by over 300% compared to what would have been the likely turnout in the previous runoff election system. Scholarly studies have also found that IRV had the effect of reducing the number of uncounted ballots (over-votes and under-votes combined), especially in districts with high percentages of racial minorities.
[More information on IRV and race]
[More information on San Francisco voter turnout and IRV]
2.4 Does IRV unfairly favor third parties?
No. IRV simply helps level the playing field so that voters will tend to vote for the candidate they genuinely prefer without worrying about "spoiler" risks. Some major party advocates have mistakenly suggested that IRV unfairly benefits minor parties - implying that some quirk of the vote tally procedure may allow a third party candidate to win merely with second choices, despite a narrow base of support. This is incorrect. As with any runoff election, under IRV a candidate must have both substantial core support and be more popular than the other finalist candidate. Unlike some other alternative voting methods (such as Borda, range, and approval voting), under IRV it is impossible for a candidate to win on the strength of second choices alone, without substantial first preference support.
2.5 Does IRV entrench the "duopoly" two-party system?
No more than any other single-winner election method. Some opponents of the American "two-party system" have suggested that IRV would simply entrench what they call "duopoly." IRV neither over-throws, nor entrenches the current domination of two major parties. IRV does allow minor parties to exist, contend for office, and possibly eventually supplant one of the existing major parties without being labeled as "spoilers." However, since IRV is a majority voting method, third parties that do not appeal to the majority of an electorate would not generally win. Only a proportional voting method (which IRV is not) is likely to result in the election of substantial number of candidates from more than two parties. Like every non-proportional, winner-take-all voting method (including plurality, Approval, Range, Condorcet, Borda, etc.) "Duverger's law" suggests two parties will tend to predominate in such winner-take-all electoral environments.
2.6 Does IRV undermine political stability by encouraging fringe candidates?
No. While fringe candidates can undermine stability through a "spoiler" effect under plurality voting, IRV solves this problem. Australia's experience with IRV indicate that it is likely to promote stability. Rather than wildly swinging results, depending on whether a minor candidate has siphoned enough votes to throw an election to a different major party candidate, IRV allows these fractured majorities to re-combine through the vote tabulation process. Thus while minor parties may cause wild swings under current voting methods, that is not the case under IRV.
[More on the Australian experience]
CONSTITUTIONAL ISSUES
3.1 Does IRV comply with the one-person-one-vote principle?
Yes. Both federal and state courts have upheld IRV as complying with the one-person-one-vote standard. Under IRV each voter is only allowed a single vote for a single candidate to be counted in each round of counting. Since every voter has an equal right to rank the candidates, courts in the U.S. have ruled that IRV complies with the constitutional one-person, one-vote mandate. As a Michigan court ruled in 1975 when upholding IRV as used at that time in Ann Arbor:
3.2 Is IRV constitutional?
Yes. There are no federal constitutional obstacles to IRV for any federal, state or local office. Indeed, IRV has been tested in courts on federal constitutional grounds, and been upheld, including in an unanimous ruling by a three-judge panel of the 9th circuit court of appeals in May 2011. Nothing in the federal constitution specifies a particular voting method to be used for any office. Some state constitutions may have provisions that would need to be amended to apply IRV to certain offices. But even those states with constitutional mandates for plurality winning thresholds may not have constitutional barriers since a candidate with a majority after an IRV tally, in fact also has a plurality (more votes than any other candidate). There is a strong argument that an IRV statute structured such that an IRV process that automatically reduces the field to two finalists, declaring whichever of these two has the plurality at this final stage elected, would comply with such plurality state constitution language, though it has not yet been tested in court.
Nor does the system of counting subsequent choices of voters for eliminated candidates unequally weight votes. Every voter has the same opportunity to rank candidates when she casts her ballot, and in each round every voter's vote carries the same value.[Full Minn. Supreme Court decision]
LOGIC AND FAIRNESS OF IRV
4.1 Aren't plurality winners okay, even if they don't get a majority?
Not if you believe in representative democracy. While electing a plurality winner who has not received a majority is not necessarily a disaster, it can be very undemocratic. The candidate with the most votes (but less than half) in a race with three or more candidates is often not the candidate preferred by the majority of voters, but not always. With plurality rules, a candidate whom the majority of voters may believe is the worst choice can be elected. The fact that some of our presidents or good politicians have won without a majority does not mean the voting method should be maintained. Pointing to a benevolent king or dictator is likewise not a convincing defense of monarchy or totalitarianism over having representative democracy. A basic principle of democracy is majority rule, as recognized by Robert's Rules of Order, which does not favor the plurality threshold for electing officers. Of course, it is also possible that a winner with only a plurality of the votes may,in fact, be the majority choice if there had been a runoff, but the failure of plurality voting to reveal that fact may weaken the elected officials mandate.
4.2 Is IRV a proven and accepted voting method?
Yes. It is used by millions of voters in government elections in the U.S. and around the world, and is a voting procedure detailed in current editions of Robert's Rules of Order, called "preferential voting," as the recommended procedure when repeated voting without eliminating candidates is not possible. IRV has also been adopted by the American Political Science Association (college professors), who know a thing or two about elections, to elect their national president. It was originally invented by a professor at the Massachusetts Institute of Technology around 1870, and has been used by millions of voters in governmental elections around the world for generations, especially in English-speaking countries, including Australia, England, and the Republic of Ireland.
[Text of Robert's Rules of Order on IRV]
4.3 With sequential elimination of candidates from the bottom, do the supporters of the weakest candidate get extra clout by getting their second choice counted before the second choice of other voters?
No. Supporters of the bottom candidate (who is eliminated first) do not have any additional clout in an IRV tally. Sequential elimination of bottom candidates in an "exhaustive" runoff is a standard, and accepted procedure, used by U.S. Congressional caucuses for electing their leaders, and mandated in various state election laws such as for party nominations by committee.
Imagine an IRV tally with five candidates. If eliminating the fifth-place candidate and transferring those voters' ballots to their second choice gives one of the other candidates a majority (more than half the votes), that means eliminating the third or fourth-place candidate first instead could not possibly give any other candidate a majority. There are not enough votes among all of the other candidates (together they have less than half the votes) to allow any other candidate any possibility of winning, regardless of the order that candidates are eliminated (unless the rules perversely eliminated one of the top candidates first). In most cases, IRV procedures eliminate all of the bottom vote-getters, with no mathematical chance of winning, in one bulk elimination. In most cases that would effectively mean eliminating all but the top two candidates. Another version of IRV, as used in Cary (NC) immediately eliminates all but the top two candidates to exactly mimic the logic of a traditional runoff election.
4.4 Is every voter treated equally under IRV?
Yes. Every voter gets an equal vote. In every round of counting, every ballot counts as one vote for the highest-ranked candidate still in the running. If your candidate is still viable, your vote will continue to count for your favorite candidate. If your candidate has been eliminated, rather than getting zero vote, your vote automatically counts for your next favorite candidate. The misunderstanding that some voters get more votes than others was the basis for a legal challenge to IRV in Ann Arbor, Michigan. The court ruled that IRV fully complied with the principle of "one person, one vote," giving equal weight to each voter. The judge wrote in his decision:
Likewise, a federal court challenge against IRV in San Francisco was rejected by a district judge in 2010 and unanimously in May 2011 by a three-judge panel of the 9th circuit court of appeals, while in 2009, a legal challenge to IRV in Minneapolis in 2009 resulted in a unanimous decision by the Minnesota Supreme Court upholding IRV. Affirming the fact that IRV treats all voters equally, the Minnesota court decision states:
"Nor does the system of counting subsequent choices of voters for eliminated candidates unequally weight votes. Every voter has the same opportunity to rank candidates when she casts her ballot, and in each round every voter's vote carries the same value."
4.5 Is IRV better than using traditional separate runoff elections?
By and large, yes. While more democratic than one-round plurality elections, two-round runoffs have distinct disadvantages. A traditional runoff extends the campaign season and can be met with a collective sigh of "Oh no, here we go again." IRV increases the likelihood that the ultimate decision will be made at the election with the greatest level of citizen participation.
Runoffs tend to have a lower voter turnout, though there are, of course, exceptions. Imagine the turnout for a runoff for a more minor office. The winner of a runoff may get fewer votes than an opponent earned in the original election, leading to doubts about the "will of the people," hobbled legitimacy, and lack of a perceived mandate. Traditional runoffs are also costly, both to the taxpayer who must pay for the duplicate election and to the candidates who must resume campaign fund-raising and prolong the stress on their families and business lives. The cost of ballot tabulation in the case of IRV is a fraction of the cost of holding a new election.
For those who want a "second look" at a reduced field of candidates, IRV can be incorporated into such an approach. In the first round, it be used to avoid vote-splitting. It also allows advancing more than two candidates, thereby creating more choice in the final runoff election.
[Comparison of IRV and other single winner election methods]
4.6 Does IRV elect majority winners?
Yes, if you definite "majority winner" as "majority support against the winner's top opponent." In that sense, IRV elects majority winners using the same logic of a regular runoff election. This majority vote in the final round may not be a majority of initial votes cast, but IRV always gives the majority the power to vote for change -- e.g, it provides a clear "majority veto" where the majority of voters always can reject a candidate. If you allow voters to abstain, you can never guarantee that a majority will vote for someone. But unlike plurality voting, where the winner can be the candidate who is the last choice of a majority of voters, IRV guarantees that the majority always can reject a disliked candidate.
The Michigan court decision that in 1975 upheld the constitutionality of IRV in Michigan was clear that IRV in this sense was a system that upholds the principle of majority rule in a single election, helping to explain why so many groups like the American Political Science Association uses IRV to elect their own leaders. The Michigan judge wrote in his decision:
But it's true that if enough voters fail to rank back-up choices, the ultimate winner may not have over 50% of the ballots counted in the initial round of counting. Ballots that have no more rankings are called "exhausted" ballots, and are akin to stay-home voters in a traditional separate runoff election because such voters have expressed no preference between the final candidates. The notion that a traditional runoff election always finds a majority winner uses the same majority winner logic as IRV. Voters who don't express their preference in the final runoff are discounted -- as abstainers. In fact, under traditional separate runoff rules, it frequently happens that, due to a drop-off in voter turnout, the winner of the runoff receives fewer votes than the loser received in the first round. That reduction in votes can never happen with IRV, of course.
4.7 Are any other alternative single-winner voting methods better at solving the "spoiler problem" than IRV?
No. Along with traditional runoff elections, IRV is quite simply the best viable solution to "spoiler" problems. Some advocates of other reforms, such as Range Voting or Approval Voting may claim that IRV does not solve the spoiler problem in all situations -- specifically when the "spoiler" can out-poll one of the major party candidates. They are incorrect as this situation does not meet the usual definition of a "spoiler," and is instead an example of the little-known "center-squeeze" dynamic common to all runoff methods.
Note that other alternative methods suffer from their own version of the "spoiler problem." For example, while likely less prone to spoilers than plurality elections, Approval Voting is not immune because it violates the later-no-harm criterion and many voters will simply "bullet vote" for their favorite, reproducing the same spoiler dynamics typical of plurality voting. In other words, whether a voter "approves" (votes for) a particular candidate depends on what other candidates the voter has to compare the candidate to, creating the same kind of "spoiler" problem one sees with plurality voting. Here is an example of a spoiler scenario under Approval:
Suppose in an Approval Voting election, a voter thinks candidate A is okay, and candidate B is horrible. In a two way race, that voter will likely approve A and not vote for B. If there are 100 voters and 55 prefer A>B and 45 B>A, this two-way race would likely result in a total vote of 55 for A to 45 for B. Thus A is both the de facto majority choice as well as the Approval winner.
Now comes the spoiler dynamic. What if candidate C decides to run as well? It happens that a significant portion (let's say 20 out of the 55) of the former A supporters who care most about issue X view candidate C as fantastically superior candidate to A or B (though they still prefer A over B as well). Some of these voters would feel the need to withdraw their approval of A so they can indicate how superior C is to A and also give C a better chance to defeat A), while others would continue to approve A while adding an approval vote for C. If elven of these 20 former A supporters decide to maximize the chance of electing C by withdrawing their approval from A, then B now wins, with the results being A=44, B=45, C=20. Thus C has "spoiled" the race for A, with the entry of C causing B to go from a loser to a winner without any new votes.
Because the score assigned to candidates under Range Voting is also dependent on what other candidates are running, a similar spoiler dynamic can occur under this method as well. In some scenarios Range Voting may relieve the spoiler dynamic, but not in others. Here is a simple spoiler scenario under Range Voting.
If the voter thinks candidate A is okay, and B is horrible in a two way race, the voter will likely score A as a 10 and B as a 0. If there are 100 voters and 55 prefer A>B and 45 B>A, this two-way race could end with a total score of 550 for A (55 voters giving a 10 and 45 giving a 0) to 450 for B. Thus A is both the de facto majority choice as well as the Range Voting "score winner."
Now comes the spoiler...What if candidate C decides to run as well? It happens that a significant portion (let's say 25 out of the 55) of the former A supporters who care most about issue X view candidate C as a fantastically superior candidate to A or B (though they still prefer A over B as well). It seems likely that many of these voters would feel the need to reduce the score of ten they otherwise would give to A to make room on the scale so they can indicate how superior C is to A. These 25 voters might now score the candidates as follows, A=5, B=0, and C=10. In other words, the score that A now receives from some voters depends on whether C has entered the race. The B supporters who generally don't care much about issue X view C as just another version of A, so give this new candidate a 0 as well. Under this entirely plausible scenario, with C in the race, now the total scores might be: candidate A now only gets 425 (30 x 10 and 25 x 5), while B still gets 450 (45 x 10) and C gets 250 (25 x 10).
Thus C has "spoiled" the race for A. The entry of C caused B to go from a loser to a winner.
Note that both of these spoiler dynamics for Approval Voting and Range Voting (and other methods that transparently violate later-no-harm like the Borda Count and Bucklin Voting) do not require the voter to have any particular information about the election. Indeed, the less they know of the relative standing of the candidates, the more likely it is that they will cast a bullet vote to give maximum benefit to their most preferred candidate. This is in marked contrast to IRV, where doing anything other than casting one's sincere preferences is far more likely to hurt your favorite candidate than help your favorite candidate absent very precise information about all voters' full preferences.
4.8 Does IRV violate important election fairness criteria?
Kenneth Arrow won the Nobel Prize for his mathematical proof that it is impossible to have a voting method that does not violate one of four criteria he offered. In other words, every voting method ever conceived fails some evaluation criteria in certain scenarios. Nevertheless, IRV does better than currently used methods, and does not violate any of the most important criteria even though some IRV opponents have made erroneous claims t o the contrary. A large number of criteria have been proposed to evaluate voting methods, and Arrow's Impossibility Theorem made no pretense of setting forth especially important ones. However, many criteria (as Arrow showed) are mutually exclusive -- that is, any voting method that satisfies criterion A must of necessity, fail criterion B. There is no perfect voting method, since no voting method satisfies all of these criteria, so it is a matter of weighing the relative severity of a potential "pathological" outcome, and the likelihood of such a scenario occurring in real-world public elections.
Many political scientists have in fact concluded that IRV is the best and fairest single-winner election method because it satisfies more of the crucial criteria than other methods, and its potential failings are relatively minor and unlikely to occur. Some of the crucial criteria that IRV satisfies and that other methods, such as Plurality, Approval, Borda, Range fail include the Later-No-Harm Criterion, the Mutual-Majority Criterion, and Resistance to Strategic Voting.
[More on these and other criteria]
4.9 Is the so-called "winner-turns-loser" or "monotonicity" issue have any real-world significance?
No. The "monotonicity criterion" has received an inordinate amount of attention by IRV opponents on the Internet. This is an arcane criterion with almost no real-world significance. As with traditional two-round runoff elections, it is conceivable in certain unique scenarios, that if enough (but not too many) voters were to switch their first preference from their favorite candidate to their least favorite candidate, that this least favorite candidate could be changed from a winner into a loser, and their favorite choice could go from loser to winner. However, simply adding new first-preferences for a candidate can never cause the candidate to lose -- IRV is monotonic as far as additional votes are concerned. It is not the additional vote in favor of a candidate that can cause the candidate to lose, but rather the change in relative support among the other candidates resulting from a vote switch. It is the switch away from another candidate, whether that switch be to the current winner, or some other candidate, that changes which candidates make it into the runoff and can cause a winner to turn loser. IRV advocates argue that it is highly unlikely that the monotonicity "winner turn loser" dynamic will occur in real-world elections. Election method experts such as Austan-Smith and Banks have argued that "monotonicity/nonmonotonicity in electoral systems is a nonissue."
Because voters need accurate information on how other voters are likely to vote in order to utilize this theoretical strategy, combined with its counter-intuitive nature and great risk of back-firing, make its significance questionable. There is no evidence that this non-monotonicity dynamic has ever led to strategic voting in any of the thousands of highly contested IRV elections that have taken place. Exploiting this non-monotonicity possibility is at least somewhat more plausible in a traditional two-election runoff system, however. In runoffs with two rounds of voting, it at times can be a relatively safe strategy because you can move your vote back to your favorite in the final round. It is much more difficult in an instant runoff with a single ballot where insincere votes will stay with the competitor, and thus can back-fire.
Also, it is important to emphasize that any voting method that satisfies the monotonicity criterion in every situation must be necessity fail the later-no-harm criterion (where ranking an alternate choice may cause your favorite candidate to lose), which most political scientists consider to be of much greater importance due to the direct incentives its creates for strategic voting in elections. The later-no-harm criterion has been shown to have a dramatic impact on actual voter behavior, with voters "bullet-voting" for a favorite rather than risk hurting this candidate by indicating a second choice. As with all evaluation criterion, every system presents certain trade-offs.
[More on monotonicity]
4.10 Will IRV generally elect the "Condorcet winner?"
Yes, unless that candidate has relatively little core, first-choice support. Dome advocates of other voting methods point out that IRV does not necessarily elect the Condorcet candidate. Named after a French mathematician, the Condorcet candidate theoretically could defeat every other candidate if paired against them one-on-one. Note that the majorities for the Condorcet candidate in different match-ups may be made up of mostly different and oppositional segments of the electorate, rather than a clear single majority. In addition, in some scenarios there isn't any such winner, because of a cycle where candidate A would beat, B, B would beat C, but C would beat A (think of the Rock Paper Scissors game). Nevertheless, IRV has in fact shown that is far more likely to elect the Condorcet candidate than a plurality voting system. As one example, an examination of ballots in Bay Area IRV elections in California in 2010 demonstrated that in four elections, the IRV winner did not lead in first choice rankings, including in one race with nearly two dozen candidates in which the winner was initially third in first choices. In each case, however, the IRV winner was the Condorcet candidate.
Ironically, IRV may elect Condorcet winners more consistently than supposedly Condorcet voting methods, due to the incentive for such voting strategies as "burying" and "bullet-voting" under Condorcet voting that are absent under IRV.
But it is also true that IRV will not always elect a Condorcet candidate. That is because the Condorcet candidate may not be what we would think of as a "strong" candidate, but rather more of a non-alienating or compromise candidate. In other words, the Condorcet candidate may be one with little strong or "core" support and who would get only a few votes in a typical plurality voting election - becoming the Condorcet winner only by being every voter's second choice. Under IRV, however, a candidate needs to have a substantial amount of first preferences, or core support, relative to other candidate in order to advance in the instant runoff round, and thus such a "weak" Condorcet winner would not win under IRV. Advocates of IRV argue that IRV strikes a desirable balance between first-choice core support and broad support, and its failure to always elect the Condorcet winner is desirable, rather than a fault.
4.11 Is ranking candidates in order of preference a more valid procedure than making up "scores" as in Range Voting?
Yes. Research in the field of psychology has shown that people are far more accurate, consistent and reliable when ranking choices (ordinal numbers) than when assigning scores (cardinal numbers). A person will consistently rank A over B if that is their preference, but may give them wildly different scores depending on circumstances (such as what other candidates are being considered, current mood, etc.) Most election methods experts, including Nobel Prize winner Kenneth Arrow, dismiss cardinal scoring methods, such as range voting, as invalid means of aggregating group preferences. Of course, there are other failings of Range Voting as well. In fact, one of the world's foremost experts on voting methods, Prof. Nicolaus Tideman, wrote in his book Collective Decisions and Voting that Range Voting is one of six unsupportable methods because they "have defects that are so serious as to disqualify them from consideration,"
[More on this topic]
ADMINISTRATION OF IRV ELECTIONS
5.1 Is IRV too complicated for election administrators to handle?
No. Administering an IRV election can save a significant amount of time and money on the part of election administrators compared to running two separate elections in a traditional two-round runoff system. Even in jurisdictions which have switched to IRV, replacing single-round plurality elections (Takoma Park, for example), have found the change to be quite manageable. Whenever new systems are used, there is a transition learning curve, however, every jurisdiction in the U.S. that has switched to IRV has made the transition without significant problems. San Francisco, has faced challenges in some recent elections as a result of problems with their voting machines unrelated to the use of IRV (optical scanners not detecting certain ink colors).
While the voter's task is extremely simple, the election administrator's task is somewhat more involved. The tally procedure takes some explaining but is fundamentally simple in concept, following the logic of a series of runoffs.
[More on San Francisco and election integrity]
5.2 Does IRV need to cost more or require buying new voting machines?
Not necessarily. Some jurisdictions in the U.S. have switched to IRV and purchased new voting machines, and some have figured out how to conduct the IRV election using existing voting machines. Those which have purchased new machines, have still saved money compared to their former system using separate runoff elections. Some jurisdictions, which did not buy new machines were able to upgrade software at a nominal cost, while others have organized a "workaround." The typical workaround involves counting first choices using existing machines and software, and then conducting a hand-count in the event that an IRV tally is needed.
5.3 Can IRV elections be counted manually without much additional time or effort?
Yes. Tallying an IRV election by hand does not typically involve significantly more effort than conducting a traditional plurality election hand re-count. This is because the ballots first get sorted by first choice, and most ballots never need to be handled again, because the vast majority of ballots typically are in stacks of the two leading candidates. It is only the small number of ballots that are in the stacks of eliminated candidates that need to be looked at again, and sorted into the stacks of the continuing candidates. However, if there are numerous IRV races on a ballot, the procedure would need to be repeated for each race, so data-entry of ballot rankings, or optical scanning of ballots is faster.
5.4 Opponents of IRV have quoted a fiscal note for an IRV bill before the Maryland legislature in 2006 estimating a high cost for implementation. Is this estimate valid?
No. The assumptions underlying that fiscal note do not apply to any other jurisdiction, nor even to Maryland any more. Some opponents of IRV taut the fiscal note attached to the 2006 IRV bill in Maryland as proof of a very high cost for implementation. According to Donna Duncan, of the Maryland Election Management Division, who provided the cost estimates to the legislative staff, those estimates were based on an extremely tight three-month implementation timeline based on the bills effective date. All of those cost estimates were removed from the fiscal note for the 2008 IRV bill. Jurisdictions that have actually gone through the IRV implementation process have seen dramatically lower costs than the Maryland estimate. Both Burlington (VT) and Cary (NC) estimate the incremental costs of implementing IRV to be around $10,000, which includes voter education, and substantially less than the cost of holding an entirely separate runoff election. The costs of the Cary (NC) IRV election are discussed in an article by the chair and Secretary of the County Elections board here.
5.5 Can current voting machines handle IRV ballots without improvements to hardware of software?
This depends on the specific jurisdiction and machines in use. The two largest election machine vendors in the United States have produced optical scan voting machines that have been used for ranked-choice elections in the United States. Although these systems are not ideal (some have a practical limit of three rankings per race, for example), they can be used. It is desirable to have these vendors upgrade their systems. Since ranked-ballot elections are rapidly becoming more common in the U.S. voting machine vendors will likely soon be selling fully compatible machines. There are also "work-around" options, such as those used in Cary (NC), in which only first choices were counted by existing voting machines, with a hand tally of alternate rankings only if there was no initial majority winner.
A report discussing work-arounds for implementing statewide IRV elections that avoid the need to purchase updated voting equipment can be found here.
5.6 Is federal certification of voting machines that can handle ranked-voting an insurmountable challenge?
No. Federal certification is voluntary under federal law. There are no federal rules applying to many local elections and even federal voting machine guidelines for state elections are typically voluntary, although many states have piggy-backed onto these guidelines. As of 2011, there has been a bottleneck for getting systems tested in a timely way by the federally certified independent labs, though this can be expected to ease over time -- and there are indications that some of the major voting equipment companies will start selling machines that come ready to run IRV elections.
5.7 Does IRV require centralized counting of ballots?
No. IRV tallies are often done in local counting centers, but the count must be centrally coordinated. Some opponents have suggested that all ballots would need to be brought to a central location, in order to conduct an IRV tally. This is not the case whether a hand count, or computer aided tally is done. For example, in Ireland each counting location simply reports their initial tally to a central election authority, which in turn provides guidance about which candidate is to be eliminated and have his or her ballots redistributed with a hand count. Other jurisdictions keep the optical scan ballots secure in the customary manner, and simply centralize the ballot data to conduct the IRV tally. It is recommended in such situations that a random sample of voting machines be had audited to assure the ballot data conforms with the paper ballots.
5.8 Does IRV mean moving to DRE voting machines?
Absolutely not. No jurisdiction in the U.S. has switched to DRE machines as a result of adopting IRV, or even considered it. Nearly all IRV elections are conducted using optical scan paper ballots. The key national organization promoting IRV, FairVote, advocates against the use of DRE machines, particularly without paper trails, and favors the heightened security and election integrity provided by optical scan ballots using both paper ballots and a redundant computer record.
[FairVote's position on voting machines and election integrity]
5.9 Does IRV weaken election integrity?
No. In fact, it can enhance election integrity. A couple of election integrity activists have made the erroneous assertion that IRV elections would be difficult to monitor for integrity. In fact, because ranked ballot optical scanners capture individual ballot records (rather than just running totals), they can add a higher level of security and fraud detection than paper-only elections. It is the redundancy of ballot records (both paper and computer), made possible by the new generation of optical scanners, that makes fraud so much more difficult to accomplish and easy to detect (the perpetrator needs to utilize two distinctly different strategies using different kinds of resources and overcoming different kinds of security measures to change BOTH records, to get away with it.)
[More about enhanced election integrity made possible by IRV]
5.10 Can IRV be manually audited?
Yes. The procedure for manually auditing a ranked-choice ballot election is a little more involved than a typical plurality election. There are two elements to such an audit: confirming that the machine record of ballot rankings matches the rankings marked on the paper ballots, and confirming that the IRV vote tallying procedure was properly done. To audit the ballot rankings, a random sample of voting machines are selected. San Francisco compares the total number of each ranking reported by the machine to the total number of each ranking manually counted on the same sample of paper ballots. A better method is to print a list of each ranking combination recorded by the machine (suppose, for example that seven ballots in a precinct ranked the candidates in following order: candidate B first, candidate A second, candidate F third) and then look at each ballot in the sample and check off each corresponding ballot type on the list, until every ballot has been looked at and every ballot ranking on the list is checked off.
There are several options for confirming the IRV tallying algorithm. San Francisco runs the software again for the sample of ballots only, while also doing a manual IRV tally with the paper ballots, to confirm the results match. A better procedure is to run an IRV tally of all ballot data using different means (such as independent software, or just a basic spreadsheet program).
[More on IRV auditing procedures]
5.11 Does IRV delay announcing election results?
Not necessarily. As with typical plurality elections, unofficial results can usually be announced quickly. And as with typical plurality elections, there is always some chance that the final official results could be different. With IRV, all precincts must report before it can be determined if a candidate has an immediate majority, and if not, which candidates need to be eliminated in the next round of the count. However, it is certainly possible to immediately provide unofficial results by running the tally software on partial returns (recognizing that, as with any unofficial results, there remains a possibility that the outcome may change when all provisional ballots or absentee ballots are included).
As an example, when Burlington (VT) held its first IRV election for mayor in 2006, the polls closed at 7:00 p.m. and the IRV tally was done and the final results were announced just over an hour and a half later. In Pierce County (WA), using the same voting system used in Bay Area cities like San Francisco and Oakland, the preliminary RCV tallies were reported on election night, as they could be done in those Bay Area cities.
If a jurisdiction does not use voting machine hardware or software that can automate the tally, the final election results may well be delayed, typically for a day or two, while a hand count is done. Nearly all such jurisdictions can at least immediately release first-choice results, which may reveal a majority winner, or indicate which candidates are most likely to win the instant runoff when it is completed.
5.12 Is it true that going to IRV in Pierce County (WA) cost more than four dollars per registered voter?
Not really. There were significant costs for Pierce County to upgrade its voting software when first going to IRV in 2008, to be sure, but this figure derives from questionable accounting. When a jurisdiction purchases new voting equipment or software designed to last for many years, it is not the usual accounting practice to count that investment as an expense for a single election rather than to spread that cost out over the life of the software. The quoted Pierce County tally also includes hundreds of thousands of dollars for staff time for salaried employees that would have been spent anyway and costs associated with voter education that, while admirable, are not necessary to implement IRV effectively. Opponents also fail to mention the savings achieved through IRV that can be achieved by eliminating the need for a separate election, as the case in 2009 when Pierce County taxpapers did not need to pay for an August primary due to using IRV.
As one comparison, Wake County (NC) reported that the City of Cary saved money the very first election with IRV in 2007 due to very low implementation costs and money saved from not needing to hold a runoff election.
Table of Contents
VOTER EXPERIENCE UNDER IRV
1.1 Is instant runoff voting simple for voters to use?
1.2 Is instant runoff voting prone to strategic voting?
1.3 Do voters like using IRV?
1.4 Has IRV been repealed in some jurisdictions?
IMPACT OF IRV
2.1 Does IRV address the "spoiler" problem as well as separate runoff elections do?
2.2 Does IRV favor extremists?
2.3 Is IRV fair to racial minorities?
2.4 Does IRV unfairly favor third parties?
2.5 Does IRV entrench the "duopoly" two-party system?
2.6 Does IRV undermine political stability by encouraging fringe candidates?
CONSTITUTIONAL ISSUES
3.1 Does IRV comply with the one-person-one-vote principle?
3.2 Is IRV constitutional?
LOGIC AND FAIRNESS OF IRV
4.1 Aren't plurality winners okay, even if they don't get a majority?
4.2 Is IRV a proven and accepted voting method?
4.3 With sequential elimination of candidates from the bottom, do the supporters of the weakest candidate get extra clout by getting their second choice counted before the second choice of other voters?
4.4 Is every voter treated equally under IRV?
4.5 Is IRV better than using traditional separate runoff elections?
4.6 Does IRV elect true majority winners?
4.7 Are any other alternative voting methods better at solving the "spoiler problem" than IRV?
4.8 Does IRV violate important election fairness criteria?
4.9 Is the so-called "winner-turns-loser" or "monotonicity" issue have any real-world significance?
4.10 Will IRV generally elect the "Condorcet winner?"
4.11 Is ranking candidates in order of preference a more valid procedure than making up "scores" as in Range Voting?
ADMINISTRATION OF IRV ELECTIONS
5.1 Is IRV too complicated for election administrators to handle?
5.2 Does IRV need to cost more or require buying new voting machines?
5.3 Can IRV elections be counted manually without much additional time or effort?
5.4 Opponents of IRV have quoted a fiscal note for an IRV bill before the Maryland legislature in 2006 estimating a high cost for implementation. Is this estimate valid?
5.5 Can current voting machines handle IRV ballots without improvements to hardware of software?
5.6 Is federal certification of voting machines that can handle ranked-voting an insurmountable challenge?
5.7 Does IRV require centralized counting of ballots?
5.8 Does IRV mean moving to DRE touchscreen voting machines?
5.9 Does IRV weaken election integrity?
5.10 Can IRV be manually audited?
5.11 Does IRV delay announcing election results?
5.12 Opponents of IRV have asserted that using IRV in Pierce County cost over four dollars per registered voter. Can that be true?
VOTER EXPERIENCE UNDER IRV
1.1 Is instant runoff voting simple for voters to use?
Yes. The voters' task is very simple. Voters can simply mark their ballots in exactly the same way as they always have in the past. However, the voter has the option of ranking alternate choices, in case there is no majority winner and the voter's favorite candidate doesn't make it into the final runoff count. Since a vote for a minor candidate won't be wasted, as long as the voter ranks other choices, the voter can generally avoid the conundrum of voting for a favorite or a lesser evil. For many voters this makes voting with IRV easier than under the current plurality method.
Several U.S. cities that adopted IRV in recent years have had exit polls to assess voter acceptance of their new system. In every city where such a scientific survey was done, voters without exception have favored IRV over the old method. Also, studies of IRV elections have demonstrated at most very small changes in uncountable ballots (spoiled or skipping the IRV race) with the adoption of IRV. In the Burlington IRV election in 1999, for example, 99.99% of ballots cast in the IRV race for mayor were valid. People had no difficulty voting, and there were almost no "spoiled" ballots.
Two nations with the highest voter participation rates in the world, Australia and Malta, both use instant runoff voting. The only "complicated" aspect of instant runoff voting is the tabulation that occurs if there is no initial majority winner. But the voters don't need to absorb these details. A voter can dial a telephone without understanding the intricacies of the internal electronics or vote for president without understanding the intricacies of the Electoral College, which has a provision for election by the House of Representatives when no candidate has a majority.
1.2 Is instant runoff voting prone to strategic voting?
No. There is another side to voting that makes IRV simpler than most other voting methods, including systems that might at first seem more simple such as vote-for-one plurality voting. While every voting method is potentially subject to manipulation by strategic voting in some situations, IRV is uniquely resistant to such strategies (see James Green-Armytage analysis for more) Under most voting methods, a potentially beneficial voting strategy can be recognized by at least some voters (who may gain an advantage over other voters). Thus, voters may face a dilemma deciding whether to engage in strategic voting.
For example, in a simple plurality election there is the "lesser-of-two-evils" problem where voters often realize that voting for their favorite candidate, who is likely to get very few votes, may deny their second choice candidate enough votes to beat out that voter's least preferred choice. Such a voter might find the plurality election decision to be extremely difficult. Under IRV, this voter's ballot will automatically transfer to that second choice candidate if the first choice is at the bottom, and the second choice candidate can win with that transferred vote. This makes the voter's task much simpler with IRV. The default strategy is always to honestly rank candidates in the order of choice. Of course, voters may still face strategic dilemmas with IRV in some rare situations, but these are far less common than under most other voting methods. For example, under Range Voting and Approval Voting, giving any support to a second choice can cause that voter's first choice to lose (violation of the Later-No-Harm Criterion), thereby leading most voters who understand the system to strategically truncate their true preferences to gain an advantage over less informed voters. IRV complies with the Later-No-Harm Criterion, and is thus immune to such strategic calculation.
1.3 Do voters like using IRV?
Yes. Voters prefer IRV to the previous election method, according to exit polls conducted in U.S. cities that use IRV. In every case of a scientific poll, a substantial majority of voters preferred IRV compared to those who preferred the old system. Excluding those who said the methods were equal or who had no opinion, the gap between those favoring IRV and those favoring the old system ranged from 71% vs. 29% in the lowest city to 89% vs. 11%.
[More information and links to the various exit polls]
The City of Cary, NC used IRV in a pilot program authorized by the state. Both election officials and voters were extremely pleased with IRV.Cary conducted a general public opinion survey of residents, and included some questions about IRV. Residents favored IRV over the former method by huge margins. The survey results can be viewed by clicking here: (note that this survey was of residents, not just voters, and so included many people who had no exposure to voting place voter education.)
1.4. Has IRV been repealed in some jurisdictions?
Yes, but not many. In the 21st century, voters in at least seventeen American jurisdictions have voted for charter amendments or statutory changes to move to IRV. Voters in three jurisdictions have passed charter amendment to replace it with more traditional two-round runoff systems: Pierce County (WA), Aspen (CO) and Burlington (VT). Each had their own contexts, but it's important to underscore that none went to a plurality voting system; all decided to uphold majority rule using a two-round runoff approach.
Some opponents like to list other repeals, but are inaccurate. For instance, Cary (NC) voted by 4-3 to use IRV in a one-time pilot program in 2007. Two post-election surveys found that most voters preferred the new system, but the default was to use a two-round runoff system in subsequent elections.
They even like to keep saying Georgetown University has repealed IRV because they happened to come across a news story saying it did. But a few weeks later, Georgetown students restored IRV and have used it ever since, along with some five dozen additional colleges and universities? Will all those schools keep IRV? Probably not. But we expect the great majority of them will keep it and support it. That's the same with IRV in city elections.
IMPACT OF IRV
2.1 Does IRV address the "spoiler" problem as well as separate runoff elections do?
Yes, but better. IRV solves the spoiler problem better than separate runoffs and at least as well as any alternative voting method. A "spoiler" is a negative term for a minor candidate with not chance of winning the election. Under plurality rules such "spoiler" candidacies can throw an election to a candidate the majority of voters oppose. IRV tackles the "spoiler problem" in a manner similar to a traditional two-election runoff, except that, due to its sequential elimination procedure, it is more effective at eliminating the spoiler effect.
Under a traditional two-election runoff, if there are several candidates in the first round who have similar policy views, it is possible for them to split the majority in such a way that none of these similar candidates reaches the runoff round, while two candidates with substantial core support but very narrow appeal advance. Under IRV, voters who divided among these similar candidates will see their votes automatically re-unite for the strongest candidate in the group, and thus advance to the final runoff. Some advocates for other alternative voting methods have claimed that IRV does not solve the "spoiler problem" 100%. IRV solves the problem as well or better than the alternatives.
[Examples of spoiler scenarios under Plurality, Range, Approval and other voting methods]
2.2 Does IRV favor extremists?
No. Current plurality voting is far more likely to elect extremist candidates than is IRV. In a plurality election with several candidates, a candidate does not need any support beyond his or her ideological core supporters to get the "most" votes - even if that is a relatively small percentage of the voters. With IRV, a candidate must be able to garner both strong core support and broad appeal in order to win. With more than 80 years of use in Australian elections for the House of Representatives, IRV has proven that extremists are not benefited by IRV.
At the other extreme from favoring extremists, some alternative voting reform proposals (such as Condorcet, Approval, and others) do not strike IRV's desirable balance of core and broad support, in the name of favoring "centrist" candidates. These methods do not necessarily favor "centrist" candidates, but may favor inoffensive and merely unknown candidates (in the event voters are sincere -- if, as is more likely, voters increasingly engage in tactical voting, the outcomes will be quite erratic). They also allow a "compromise" candidate to be elected even if not a single voter considers him or her to be the best choice. These other methods allow a candidate who would not get a single vote under the current voting method, to win. For some offices, such as a club treasurer, this "compromise" bias may be desirable. But in high-stakes elections of political leaders, these methods may discourage candidates from revealing their stands on controversial issues, as they seek to avoid alienating any voters at all.
[More on Australian Elections]
2.3 Is IRV fair to racial minorities?
Yes. IRV repeatedly has been used effectively by racial minorities, but it neither hinders nor promotes the election of racial minorities. To generally advance the cause of representation for racial minorities, forms of proportional voting are needed, rather than any winner-take-all method such as IRV and nearly all other election methods in the U.S. However, by focusing the decision on a single election, rather than a two-election runoff process, IRV does tend to enhance voter turnout and success of racial minorities in the decisive election. In a multi-candidate IRV election, candidates have an interest in appealing to the supporters of other candidates for second preferences, and there is some risk that appeals to racism may backfire by alienating potential second preferences. However, in a head-to-head separate runoffs, demonization of opponents and appeals to racism are often an effective strategy.
Exit polls by San Francisco State University and analyses by the New America Foundation and FairVote of San Francisco's IRV elections found that racial minorities used the ranked-ballot effectively, and preferred it to the former election method and that IRV dramatically increased voter turnout by low-income and racial minorities in the decisive election round. In 2005's citywide election, IRV increased turnout in san Francisco’s six most socio-economically diverse neighborhoods by over 300% compared to what would have been the likely turnout in the previous runoff election system. Scholarly studies have also found that IRV had the effect of reducing the number of uncounted ballots (over-votes and under-votes combined), especially in districts with high percentages of racial minorities.
[More information on IRV and race]
[More information on San Francisco voter turnout and IRV]
2.4 Does IRV unfairly favor third parties?
No. IRV simply helps level the playing field so that voters will tend to vote for the candidate they genuinely prefer without worrying about "spoiler" risks. Some major party advocates have mistakenly suggested that IRV unfairly benefits minor parties - implying that some quirk of the vote tally procedure may allow a third party candidate to win merely with second choices, despite a narrow base of support. This is incorrect. As with any runoff election, under IRV a candidate must have both substantial core support and be more popular than the other finalist candidate. Unlike some other alternative voting methods (such as Borda, range, and approval voting), under IRV it is impossible for a candidate to win on the strength of second choices alone, without substantial first preference support.
2.5 Does IRV entrench the "duopoly" two-party system?
No more than any other single-winner election method. Some opponents of the American "two-party system" have suggested that IRV would simply entrench what they call "duopoly." IRV neither over-throws, nor entrenches the current domination of two major parties. IRV does allow minor parties to exist, contend for office, and possibly eventually supplant one of the existing major parties without being labeled as "spoilers." However, since IRV is a majority voting method, third parties that do not appeal to the majority of an electorate would not generally win. Only a proportional voting method (which IRV is not) is likely to result in the election of substantial number of candidates from more than two parties. Like every non-proportional, winner-take-all voting method (including plurality, Approval, Range, Condorcet, Borda, etc.) "Duverger's law" suggests two parties will tend to predominate in such winner-take-all electoral environments.
2.6 Does IRV undermine political stability by encouraging fringe candidates?
No. While fringe candidates can undermine stability through a "spoiler" effect under plurality voting, IRV solves this problem. Australia's experience with IRV indicate that it is likely to promote stability. Rather than wildly swinging results, depending on whether a minor candidate has siphoned enough votes to throw an election to a different major party candidate, IRV allows these fractured majorities to re-combine through the vote tabulation process. Thus while minor parties may cause wild swings under current voting methods, that is not the case under IRV.
[More on the Australian experience]
CONSTITUTIONAL ISSUES
3.1 Does IRV comply with the one-person-one-vote principle?
Yes. Both federal and state courts have upheld IRV as complying with the one-person-one-vote standard. Under IRV each voter is only allowed a single vote for a single candidate to be counted in each round of counting. Since every voter has an equal right to rank the candidates, courts in the U.S. have ruled that IRV complies with the constitutional one-person, one-vote mandate. As a Michigan court ruled in 1975 when upholding IRV as used at that time in Ann Arbor:
- "Under the [IRV system], however, no one person or voter has more than one effective vote for one office. No voter's vote can be counted more than once for the same candidate. In the final analysis, no voter is given greater weight in his or her vote over the vote of another voter, although to understand this does require a conceptual understanding of how the effect of a "[IRV] System" is like that of a run-off election. The form of majority preferential voting employed in the City of Ann Arbor's election of its Mayor does not violate the one-man, one-vote mandate nor does it deprive anyone of equal protection rights under the Michigan or United States Constitutions." In May 2011, the federal 9th circuit court of appeals unanimously upheld the constitutionality of IRV in San Francisco, specifically dismissing the issue of IRV being a one-person, one-vote system.
3.2 Is IRV constitutional?
Yes. There are no federal constitutional obstacles to IRV for any federal, state or local office. Indeed, IRV has been tested in courts on federal constitutional grounds, and been upheld, including in an unanimous ruling by a three-judge panel of the 9th circuit court of appeals in May 2011. Nothing in the federal constitution specifies a particular voting method to be used for any office. Some state constitutions may have provisions that would need to be amended to apply IRV to certain offices. But even those states with constitutional mandates for plurality winning thresholds may not have constitutional barriers since a candidate with a majority after an IRV tally, in fact also has a plurality (more votes than any other candidate). There is a strong argument that an IRV statute structured such that an IRV process that automatically reduces the field to two finalists, declaring whichever of these two has the plurality at this final stage elected, would comply with such plurality state constitution language, though it has not yet been tested in court.
Nor does the system of counting subsequent choices of voters for eliminated candidates unequally weight votes. Every voter has the same opportunity to rank candidates when she casts her ballot, and in each round every voter's vote carries the same value.[Full Minn. Supreme Court decision]
LOGIC AND FAIRNESS OF IRV
4.1 Aren't plurality winners okay, even if they don't get a majority?
Not if you believe in representative democracy. While electing a plurality winner who has not received a majority is not necessarily a disaster, it can be very undemocratic. The candidate with the most votes (but less than half) in a race with three or more candidates is often not the candidate preferred by the majority of voters, but not always. With plurality rules, a candidate whom the majority of voters may believe is the worst choice can be elected. The fact that some of our presidents or good politicians have won without a majority does not mean the voting method should be maintained. Pointing to a benevolent king or dictator is likewise not a convincing defense of monarchy or totalitarianism over having representative democracy. A basic principle of democracy is majority rule, as recognized by Robert's Rules of Order, which does not favor the plurality threshold for electing officers. Of course, it is also possible that a winner with only a plurality of the votes may,in fact, be the majority choice if there had been a runoff, but the failure of plurality voting to reveal that fact may weaken the elected officials mandate.
4.2 Is IRV a proven and accepted voting method?
Yes. It is used by millions of voters in government elections in the U.S. and around the world, and is a voting procedure detailed in current editions of Robert's Rules of Order, called "preferential voting," as the recommended procedure when repeated voting without eliminating candidates is not possible. IRV has also been adopted by the American Political Science Association (college professors), who know a thing or two about elections, to elect their national president. It was originally invented by a professor at the Massachusetts Institute of Technology around 1870, and has been used by millions of voters in governmental elections around the world for generations, especially in English-speaking countries, including Australia, England, and the Republic of Ireland.
[Text of Robert's Rules of Order on IRV]
4.3 With sequential elimination of candidates from the bottom, do the supporters of the weakest candidate get extra clout by getting their second choice counted before the second choice of other voters?
No. Supporters of the bottom candidate (who is eliminated first) do not have any additional clout in an IRV tally. Sequential elimination of bottom candidates in an "exhaustive" runoff is a standard, and accepted procedure, used by U.S. Congressional caucuses for electing their leaders, and mandated in various state election laws such as for party nominations by committee.
Imagine an IRV tally with five candidates. If eliminating the fifth-place candidate and transferring those voters' ballots to their second choice gives one of the other candidates a majority (more than half the votes), that means eliminating the third or fourth-place candidate first instead could not possibly give any other candidate a majority. There are not enough votes among all of the other candidates (together they have less than half the votes) to allow any other candidate any possibility of winning, regardless of the order that candidates are eliminated (unless the rules perversely eliminated one of the top candidates first). In most cases, IRV procedures eliminate all of the bottom vote-getters, with no mathematical chance of winning, in one bulk elimination. In most cases that would effectively mean eliminating all but the top two candidates. Another version of IRV, as used in Cary (NC) immediately eliminates all but the top two candidates to exactly mimic the logic of a traditional runoff election.
4.4 Is every voter treated equally under IRV?
Yes. Every voter gets an equal vote. In every round of counting, every ballot counts as one vote for the highest-ranked candidate still in the running. If your candidate is still viable, your vote will continue to count for your favorite candidate. If your candidate has been eliminated, rather than getting zero vote, your vote automatically counts for your next favorite candidate. The misunderstanding that some voters get more votes than others was the basis for a legal challenge to IRV in Ann Arbor, Michigan. The court ruled that IRV fully complied with the principle of "one person, one vote," giving equal weight to each voter. The judge wrote in his decision:
- "Under the [IRV system], however, no one person or voter has more than one effective vote for one office. No voter's vote can be counted more than once for the same candidate. In the final analysis, no voter is given greater weight in his or her vote over the vote of another voter, although to understand this does require a conceptual understanding of how the effect of a "M.P.V. System" is like that of a run-off election. The form of majority preferential voting employed in the City of Ann Arbor's election of its Mayor does not violate the one-man, one-vote mandate nor does it deprive anyone of equal protection rights under the Michigan or United States Constitutions."
Likewise, a federal court challenge against IRV in San Francisco was rejected by a district judge in 2010 and unanimously in May 2011 by a three-judge panel of the 9th circuit court of appeals, while in 2009, a legal challenge to IRV in Minneapolis in 2009 resulted in a unanimous decision by the Minnesota Supreme Court upholding IRV. Affirming the fact that IRV treats all voters equally, the Minnesota court decision states:
“[e]very voter has the same opportunity to rank candidates when she casts her ballot, and in each round every voter’s vote carries the same value.”
and "Nor does the system of counting subsequent choices of voters for eliminated candidates unequally weight votes. Every voter has the same opportunity to rank candidates when she casts her ballot, and in each round every voter's vote carries the same value."
4.5 Is IRV better than using traditional separate runoff elections?
By and large, yes. While more democratic than one-round plurality elections, two-round runoffs have distinct disadvantages. A traditional runoff extends the campaign season and can be met with a collective sigh of "Oh no, here we go again." IRV increases the likelihood that the ultimate decision will be made at the election with the greatest level of citizen participation.
Runoffs tend to have a lower voter turnout, though there are, of course, exceptions. Imagine the turnout for a runoff for a more minor office. The winner of a runoff may get fewer votes than an opponent earned in the original election, leading to doubts about the "will of the people," hobbled legitimacy, and lack of a perceived mandate. Traditional runoffs are also costly, both to the taxpayer who must pay for the duplicate election and to the candidates who must resume campaign fund-raising and prolong the stress on their families and business lives. The cost of ballot tabulation in the case of IRV is a fraction of the cost of holding a new election.
For those who want a "second look" at a reduced field of candidates, IRV can be incorporated into such an approach. In the first round, it be used to avoid vote-splitting. It also allows advancing more than two candidates, thereby creating more choice in the final runoff election.
[Comparison of IRV and other single winner election methods]
4.6 Does IRV elect majority winners?
Yes, if you definite "majority winner" as "majority support against the winner's top opponent." In that sense, IRV elects majority winners using the same logic of a regular runoff election. This majority vote in the final round may not be a majority of initial votes cast, but IRV always gives the majority the power to vote for change -- e.g, it provides a clear "majority veto" where the majority of voters always can reject a candidate. If you allow voters to abstain, you can never guarantee that a majority will vote for someone. But unlike plurality voting, where the winner can be the candidate who is the last choice of a majority of voters, IRV guarantees that the majority always can reject a disliked candidate.
The Michigan court decision that in 1975 upheld the constitutionality of IRV in Michigan was clear that IRV in this sense was a system that upholds the principle of majority rule in a single election, helping to explain why so many groups like the American Political Science Association uses IRV to elect their own leaders. The Michigan judge wrote in his decision:
- "Each voter has the same right at the time he casts his or her ballot. Each voter has his or her ballot counted once in any count that determines whether one candidate has a majority of the votes. . . . Far better to have the People's will expressed more adequately in this fashion, than to wonder what would have been the results of a run-off election not provided for."
But it's true that if enough voters fail to rank back-up choices, the ultimate winner may not have over 50% of the ballots counted in the initial round of counting. Ballots that have no more rankings are called "exhausted" ballots, and are akin to stay-home voters in a traditional separate runoff election because such voters have expressed no preference between the final candidates. The notion that a traditional runoff election always finds a majority winner uses the same majority winner logic as IRV. Voters who don't express their preference in the final runoff are discounted -- as abstainers. In fact, under traditional separate runoff rules, it frequently happens that, due to a drop-off in voter turnout, the winner of the runoff receives fewer votes than the loser received in the first round. That reduction in votes can never happen with IRV, of course.
4.7 Are any other alternative single-winner voting methods better at solving the "spoiler problem" than IRV?
No. Along with traditional runoff elections, IRV is quite simply the best viable solution to "spoiler" problems. Some advocates of other reforms, such as Range Voting or Approval Voting may claim that IRV does not solve the spoiler problem in all situations -- specifically when the "spoiler" can out-poll one of the major party candidates. They are incorrect as this situation does not meet the usual definition of a "spoiler," and is instead an example of the little-known "center-squeeze" dynamic common to all runoff methods.
Note that other alternative methods suffer from their own version of the "spoiler problem." For example, while likely less prone to spoilers than plurality elections, Approval Voting is not immune because it violates the later-no-harm criterion and many voters will simply "bullet vote" for their favorite, reproducing the same spoiler dynamics typical of plurality voting. In other words, whether a voter "approves" (votes for) a particular candidate depends on what other candidates the voter has to compare the candidate to, creating the same kind of "spoiler" problem one sees with plurality voting. Here is an example of a spoiler scenario under Approval:
Suppose in an Approval Voting election, a voter thinks candidate A is okay, and candidate B is horrible. In a two way race, that voter will likely approve A and not vote for B. If there are 100 voters and 55 prefer A>B and 45 B>A, this two-way race would likely result in a total vote of 55 for A to 45 for B. Thus A is both the de facto majority choice as well as the Approval winner.
Now comes the spoiler dynamic. What if candidate C decides to run as well? It happens that a significant portion (let's say 20 out of the 55) of the former A supporters who care most about issue X view candidate C as fantastically superior candidate to A or B (though they still prefer A over B as well). Some of these voters would feel the need to withdraw their approval of A so they can indicate how superior C is to A and also give C a better chance to defeat A), while others would continue to approve A while adding an approval vote for C. If elven of these 20 former A supporters decide to maximize the chance of electing C by withdrawing their approval from A, then B now wins, with the results being A=44, B=45, C=20. Thus C has "spoiled" the race for A, with the entry of C causing B to go from a loser to a winner without any new votes.
Because the score assigned to candidates under Range Voting is also dependent on what other candidates are running, a similar spoiler dynamic can occur under this method as well. In some scenarios Range Voting may relieve the spoiler dynamic, but not in others. Here is a simple spoiler scenario under Range Voting.
If the voter thinks candidate A is okay, and B is horrible in a two way race, the voter will likely score A as a 10 and B as a 0. If there are 100 voters and 55 prefer A>B and 45 B>A, this two-way race could end with a total score of 550 for A (55 voters giving a 10 and 45 giving a 0) to 450 for B. Thus A is both the de facto majority choice as well as the Range Voting "score winner."
Now comes the spoiler...What if candidate C decides to run as well? It happens that a significant portion (let's say 25 out of the 55) of the former A supporters who care most about issue X view candidate C as a fantastically superior candidate to A or B (though they still prefer A over B as well). It seems likely that many of these voters would feel the need to reduce the score of ten they otherwise would give to A to make room on the scale so they can indicate how superior C is to A. These 25 voters might now score the candidates as follows, A=5, B=0, and C=10. In other words, the score that A now receives from some voters depends on whether C has entered the race. The B supporters who generally don't care much about issue X view C as just another version of A, so give this new candidate a 0 as well. Under this entirely plausible scenario, with C in the race, now the total scores might be: candidate A now only gets 425 (30 x 10 and 25 x 5), while B still gets 450 (45 x 10) and C gets 250 (25 x 10).
Thus C has "spoiled" the race for A. The entry of C caused B to go from a loser to a winner.
Note that both of these spoiler dynamics for Approval Voting and Range Voting (and other methods that transparently violate later-no-harm like the Borda Count and Bucklin Voting) do not require the voter to have any particular information about the election. Indeed, the less they know of the relative standing of the candidates, the more likely it is that they will cast a bullet vote to give maximum benefit to their most preferred candidate. This is in marked contrast to IRV, where doing anything other than casting one's sincere preferences is far more likely to hurt your favorite candidate than help your favorite candidate absent very precise information about all voters' full preferences.
4.8 Does IRV violate important election fairness criteria?
Kenneth Arrow won the Nobel Prize for his mathematical proof that it is impossible to have a voting method that does not violate one of four criteria he offered. In other words, every voting method ever conceived fails some evaluation criteria in certain scenarios. Nevertheless, IRV does better than currently used methods, and does not violate any of the most important criteria even though some IRV opponents have made erroneous claims t o the contrary. A large number of criteria have been proposed to evaluate voting methods, and Arrow's Impossibility Theorem made no pretense of setting forth especially important ones. However, many criteria (as Arrow showed) are mutually exclusive -- that is, any voting method that satisfies criterion A must of necessity, fail criterion B. There is no perfect voting method, since no voting method satisfies all of these criteria, so it is a matter of weighing the relative severity of a potential "pathological" outcome, and the likelihood of such a scenario occurring in real-world public elections.
Many political scientists have in fact concluded that IRV is the best and fairest single-winner election method because it satisfies more of the crucial criteria than other methods, and its potential failings are relatively minor and unlikely to occur. Some of the crucial criteria that IRV satisfies and that other methods, such as Plurality, Approval, Borda, Range fail include the Later-No-Harm Criterion, the Mutual-Majority Criterion, and Resistance to Strategic Voting.
[More on these and other criteria]
4.9 Is the so-called "winner-turns-loser" or "monotonicity" issue have any real-world significance?
No. The "monotonicity criterion" has received an inordinate amount of attention by IRV opponents on the Internet. This is an arcane criterion with almost no real-world significance. As with traditional two-round runoff elections, it is conceivable in certain unique scenarios, that if enough (but not too many) voters were to switch their first preference from their favorite candidate to their least favorite candidate, that this least favorite candidate could be changed from a winner into a loser, and their favorite choice could go from loser to winner. However, simply adding new first-preferences for a candidate can never cause the candidate to lose -- IRV is monotonic as far as additional votes are concerned. It is not the additional vote in favor of a candidate that can cause the candidate to lose, but rather the change in relative support among the other candidates resulting from a vote switch. It is the switch away from another candidate, whether that switch be to the current winner, or some other candidate, that changes which candidates make it into the runoff and can cause a winner to turn loser. IRV advocates argue that it is highly unlikely that the monotonicity "winner turn loser" dynamic will occur in real-world elections. Election method experts such as Austan-Smith and Banks have argued that "monotonicity/nonmonotonicity in electoral systems is a nonissue."
Because voters need accurate information on how other voters are likely to vote in order to utilize this theoretical strategy, combined with its counter-intuitive nature and great risk of back-firing, make its significance questionable. There is no evidence that this non-monotonicity dynamic has ever led to strategic voting in any of the thousands of highly contested IRV elections that have taken place. Exploiting this non-monotonicity possibility is at least somewhat more plausible in a traditional two-election runoff system, however. In runoffs with two rounds of voting, it at times can be a relatively safe strategy because you can move your vote back to your favorite in the final round. It is much more difficult in an instant runoff with a single ballot where insincere votes will stay with the competitor, and thus can back-fire.
Also, it is important to emphasize that any voting method that satisfies the monotonicity criterion in every situation must be necessity fail the later-no-harm criterion (where ranking an alternate choice may cause your favorite candidate to lose), which most political scientists consider to be of much greater importance due to the direct incentives its creates for strategic voting in elections. The later-no-harm criterion has been shown to have a dramatic impact on actual voter behavior, with voters "bullet-voting" for a favorite rather than risk hurting this candidate by indicating a second choice. As with all evaluation criterion, every system presents certain trade-offs.
[More on monotonicity]
4.10 Will IRV generally elect the "Condorcet winner?"
Yes, unless that candidate has relatively little core, first-choice support. Dome advocates of other voting methods point out that IRV does not necessarily elect the Condorcet candidate. Named after a French mathematician, the Condorcet candidate theoretically could defeat every other candidate if paired against them one-on-one. Note that the majorities for the Condorcet candidate in different match-ups may be made up of mostly different and oppositional segments of the electorate, rather than a clear single majority. In addition, in some scenarios there isn't any such winner, because of a cycle where candidate A would beat, B, B would beat C, but C would beat A (think of the Rock Paper Scissors game). Nevertheless, IRV has in fact shown that is far more likely to elect the Condorcet candidate than a plurality voting system. As one example, an examination of ballots in Bay Area IRV elections in California in 2010 demonstrated that in four elections, the IRV winner did not lead in first choice rankings, including in one race with nearly two dozen candidates in which the winner was initially third in first choices. In each case, however, the IRV winner was the Condorcet candidate.
Ironically, IRV may elect Condorcet winners more consistently than supposedly Condorcet voting methods, due to the incentive for such voting strategies as "burying" and "bullet-voting" under Condorcet voting that are absent under IRV.
But it is also true that IRV will not always elect a Condorcet candidate. That is because the Condorcet candidate may not be what we would think of as a "strong" candidate, but rather more of a non-alienating or compromise candidate. In other words, the Condorcet candidate may be one with little strong or "core" support and who would get only a few votes in a typical plurality voting election - becoming the Condorcet winner only by being every voter's second choice. Under IRV, however, a candidate needs to have a substantial amount of first preferences, or core support, relative to other candidate in order to advance in the instant runoff round, and thus such a "weak" Condorcet winner would not win under IRV. Advocates of IRV argue that IRV strikes a desirable balance between first-choice core support and broad support, and its failure to always elect the Condorcet winner is desirable, rather than a fault.
4.11 Is ranking candidates in order of preference a more valid procedure than making up "scores" as in Range Voting?
Yes. Research in the field of psychology has shown that people are far more accurate, consistent and reliable when ranking choices (ordinal numbers) than when assigning scores (cardinal numbers). A person will consistently rank A over B if that is their preference, but may give them wildly different scores depending on circumstances (such as what other candidates are being considered, current mood, etc.) Most election methods experts, including Nobel Prize winner Kenneth Arrow, dismiss cardinal scoring methods, such as range voting, as invalid means of aggregating group preferences. Of course, there are other failings of Range Voting as well. In fact, one of the world's foremost experts on voting methods, Prof. Nicolaus Tideman, wrote in his book Collective Decisions and Voting that Range Voting is one of six unsupportable methods because they "have defects that are so serious as to disqualify them from consideration,"
ADMINISTRATION OF IRV ELECTIONS
5.1 Is IRV too complicated for election administrators to handle?
No. Administering an IRV election can save a significant amount of time and money on the part of election administrators compared to running two separate elections in a traditional two-round runoff system. Even in jurisdictions which have switched to IRV, replacing single-round plurality elections (Takoma Park, for example), have found the change to be quite manageable. Whenever new systems are used, there is a transition learning curve, however, every jurisdiction in the U.S. that has switched to IRV has made the transition without significant problems. San Francisco, has faced challenges in some recent elections as a result of problems with their voting machines unrelated to the use of IRV (optical scanners not detecting certain ink colors).
While the voter's task is extremely simple, the election administrator's task is somewhat more involved. The tally procedure takes some explaining but is fundamentally simple in concept, following the logic of a series of runoffs.
[More on San Francisco and election integrity]
5.2 Does IRV need to cost more or require buying new voting machines?
Not necessarily. Some jurisdictions in the U.S. have switched to IRV and purchased new voting machines, and some have figured out how to conduct the IRV election using existing voting machines. Those which have purchased new machines, have still saved money compared to their former system using separate runoff elections. Some jurisdictions, which did not buy new machines were able to upgrade software at a nominal cost, while others have organized a "workaround." The typical workaround involves counting first choices using existing machines and software, and then conducting a hand-count in the event that an IRV tally is needed.
5.3 Can IRV elections be counted manually without much additional time or effort?
Yes. Tallying an IRV election by hand does not typically involve significantly more effort than conducting a traditional plurality election hand re-count. This is because the ballots first get sorted by first choice, and most ballots never need to be handled again, because the vast majority of ballots typically are in stacks of the two leading candidates. It is only the small number of ballots that are in the stacks of eliminated candidates that need to be looked at again, and sorted into the stacks of the continuing candidates. However, if there are numerous IRV races on a ballot, the procedure would need to be repeated for each race, so data-entry of ballot rankings, or optical scanning of ballots is faster.
5.4 Opponents of IRV have quoted a fiscal note for an IRV bill before the Maryland legislature in 2006 estimating a high cost for implementation. Is this estimate valid?
No. The assumptions underlying that fiscal note do not apply to any other jurisdiction, nor even to Maryland any more. Some opponents of IRV taut the fiscal note attached to the 2006 IRV bill in Maryland as proof of a very high cost for implementation. According to Donna Duncan, of the Maryland Election Management Division, who provided the cost estimates to the legislative staff, those estimates were based on an extremely tight three-month implementation timeline based on the bills effective date. All of those cost estimates were removed from the fiscal note for the 2008 IRV bill. Jurisdictions that have actually gone through the IRV implementation process have seen dramatically lower costs than the Maryland estimate. Both Burlington (VT) and Cary (NC) estimate the incremental costs of implementing IRV to be around $10,000, which includes voter education, and substantially less than the cost of holding an entirely separate runoff election. The costs of the Cary (NC) IRV election are discussed in an article by the chair and Secretary of the County Elections board here.
5.5 Can current voting machines handle IRV ballots without improvements to hardware of software?
This depends on the specific jurisdiction and machines in use. The two largest election machine vendors in the United States have produced optical scan voting machines that have been used for ranked-choice elections in the United States. Although these systems are not ideal (some have a practical limit of three rankings per race, for example), they can be used. It is desirable to have these vendors upgrade their systems. Since ranked-ballot elections are rapidly becoming more common in the U.S. voting machine vendors will likely soon be selling fully compatible machines. There are also "work-around" options, such as those used in Cary (NC), in which only first choices were counted by existing voting machines, with a hand tally of alternate rankings only if there was no initial majority winner.
A report discussing work-arounds for implementing statewide IRV elections that avoid the need to purchase updated voting equipment can be found here.
5.6 Is federal certification of voting machines that can handle ranked-voting an insurmountable challenge?
No. Federal certification is voluntary under federal law. There are no federal rules applying to many local elections and even federal voting machine guidelines for state elections are typically voluntary, although many states have piggy-backed onto these guidelines. As of 2011, there has been a bottleneck for getting systems tested in a timely way by the federally certified independent labs, though this can be expected to ease over time -- and there are indications that some of the major voting equipment companies will start selling machines that come ready to run IRV elections.
5.7 Does IRV require centralized counting of ballots?
No. IRV tallies are often done in local counting centers, but the count must be centrally coordinated. Some opponents have suggested that all ballots would need to be brought to a central location, in order to conduct an IRV tally. This is not the case whether a hand count, or computer aided tally is done. For example, in Ireland each counting location simply reports their initial tally to a central election authority, which in turn provides guidance about which candidate is to be eliminated and have his or her ballots redistributed with a hand count. Other jurisdictions keep the optical scan ballots secure in the customary manner, and simply centralize the ballot data to conduct the IRV tally. It is recommended in such situations that a random sample of voting machines be had audited to assure the ballot data conforms with the paper ballots.
5.8 Does IRV mean moving to DRE voting machines?
Absolutely not. No jurisdiction in the U.S. has switched to DRE machines as a result of adopting IRV, or even considered it. Nearly all IRV elections are conducted using optical scan paper ballots. The key national organization promoting IRV, FairVote, advocates against the use of DRE machines, particularly without paper trails, and favors the heightened security and election integrity provided by optical scan ballots using both paper ballots and a redundant computer record.
[FairVote's position on voting machines and election integrity]
5.9 Does IRV weaken election integrity?
No. In fact, it can enhance election integrity. A couple of election integrity activists have made the erroneous assertion that IRV elections would be difficult to monitor for integrity. In fact, because ranked ballot optical scanners capture individual ballot records (rather than just running totals), they can add a higher level of security and fraud detection than paper-only elections. It is the redundancy of ballot records (both paper and computer), made possible by the new generation of optical scanners, that makes fraud so much more difficult to accomplish and easy to detect (the perpetrator needs to utilize two distinctly different strategies using different kinds of resources and overcoming different kinds of security measures to change BOTH records, to get away with it.)
[More about enhanced election integrity made possible by IRV]
5.10 Can IRV be manually audited?
Yes. The procedure for manually auditing a ranked-choice ballot election is a little more involved than a typical plurality election. There are two elements to such an audit: confirming that the machine record of ballot rankings matches the rankings marked on the paper ballots, and confirming that the IRV vote tallying procedure was properly done. To audit the ballot rankings, a random sample of voting machines are selected. San Francisco compares the total number of each ranking reported by the machine to the total number of each ranking manually counted on the same sample of paper ballots. A better method is to print a list of each ranking combination recorded by the machine (suppose, for example that seven ballots in a precinct ranked the candidates in following order: candidate B first, candidate A second, candidate F third) and then look at each ballot in the sample and check off each corresponding ballot type on the list, until every ballot has been looked at and every ballot ranking on the list is checked off.
There are several options for confirming the IRV tallying algorithm. San Francisco runs the software again for the sample of ballots only, while also doing a manual IRV tally with the paper ballots, to confirm the results match. A better procedure is to run an IRV tally of all ballot data using different means (such as independent software, or just a basic spreadsheet program).
[More on IRV auditing procedures]
5.11 Does IRV delay announcing election results?
Not necessarily. As with typical plurality elections, unofficial results can usually be announced quickly. And as with typical plurality elections, there is always some chance that the final official results could be different. With IRV, all precincts must report before it can be determined if a candidate has an immediate majority, and if not, which candidates need to be eliminated in the next round of the count. However, it is certainly possible to immediately provide unofficial results by running the tally software on partial returns (recognizing that, as with any unofficial results, there remains a possibility that the outcome may change when all provisional ballots or absentee ballots are included).
As an example, when Burlington (VT) held its first IRV election for mayor in 2006, the polls closed at 7:00 p.m. and the IRV tally was done and the final results were announced just over an hour and a half later. In Pierce County (WA), using the same voting system used in Bay Area cities like San Francisco and Oakland, the preliminary RCV tallies were reported on election night, as they could be done in those Bay Area cities.
If a jurisdiction does not use voting machine hardware or software that can automate the tally, the final election results may well be delayed, typically for a day or two, while a hand count is done. Nearly all such jurisdictions can at least immediately release first-choice results, which may reveal a majority winner, or indicate which candidates are most likely to win the instant runoff when it is completed.
5.12 Is it true that going to IRV in Pierce County (WA) cost more than four dollars per registered voter?
Not really. There were significant costs for Pierce County to upgrade its voting software when first going to IRV in 2008, to be sure, but this figure derives from questionable accounting. When a jurisdiction purchases new voting equipment or software designed to last for many years, it is not the usual accounting practice to count that investment as an expense for a single election rather than to spread that cost out over the life of the software. The quoted Pierce County tally also includes hundreds of thousands of dollars for staff time for salaried employees that would have been spent anyway and costs associated with voter education that, while admirable, are not necessary to implement IRV effectively. Opponents also fail to mention the savings achieved through IRV that can be achieved by eliminating the need for a separate election, as the case in 2009 when Pierce County taxpapers did not need to pay for an August primary due to using IRV.
As one comparison, Wake County (NC) reported that the City of Cary saved money the very first election with IRV in 2007 due to very low implementation costs and money saved from not needing to hold a runoff election.
