Instant-runoff voting in the United States

Main article: Instant-runoff voting

First used in the United States in 1912, instant-runoff voting (IRV) has been adopted since 2002 in a number of U.S. cities, with some of these adoptions pending implementation. It also has been repealed in three jurisdictions. As of December 2015, IRV elections have been held in a statewide election in North Carolina and for local elections in San Francisco, California; Oakland, California; Berkeley, California; San Leandro, California; Burlington, Vermont; Takoma Park, Maryland; Aspen, Colorado; Minneapolis, Minnesota; Pierce County, Washington; Telluride, Colorado; St. Paul, Minnesota; Portland, Maine and Hendersonville, North Carolina. This article lists the cities in the order of year adopted, the status of implementation, and the results of elections held.

Historical use

1912 Florida, Indiana, Maryland, and Minnesota: for party primaries

In the United States, IRV election laws were first adopted in 1912. Five states (Florida, Indiana, Maryland, Wisconsin and Minnesota) used versions of IRV for party primaries. Of the states with IRV, only the Minnesota and Maryland laws used the standard IRV sequential elimination of bottom candidates, while the others used batch elimination of all but the top two candidates.

After a series of primary elections in which alternate preference votes happened to play no role in determining the winner, this voting procedure was eclipsed in all five states.

By the 1930s, these instant runoff voting systems had been replaced by other primary election reforms, including the use of a second, or runoff primary in the event of a non-majority outcome.

1974, Ann Arbor, Michigan: city mayor

IRV (called preferential voting or PV) was adopted for mayoral races in Ann Arbor, Michigan in 1974 after a successful ballot initiative sponsored by the local Human Rights Party. IRV was used in the 1975 mayoral election. Democratic Party nominee Albert Wheeler, the city's first African-American mayor, won after trailing the Republican incumbent 49% to 40% in the first round of counting, with remaining votes cast for the Human Rights Party nominee.

In April 1976, 62% of voters voted to repeal IRV in a special election.

Since 2002

San Francisco, California

In March 2002, an initiative backed by a broad coalition of civic organizations won 55% of the vote in making instant runoff voting the means of electing candidates for the Board of Supervisors and most citywide offices in San Francisco. It was first used in that city in October 2004 for YouthVOTE, an election held throughout San Francisco’s public schools which elected the San Francisco school board's student delegate; after that, it was used in the November 2004 supervisoral races and every November since that time for at least one election in the city. IRV played a decisive role in at least one city election in 2004, 2005, 2006, 2008, 2010, 2011, 2012 and 2014. Exit polls by San Francisco State University have shown support for the new system from all groupings of voters. A unanimous panel of the US Court of Appeals for the Ninth Circuit has upheld San Francisco's IRV law as constitutional.

The San Francisco Department of Elections prefers the term Ranked Choice Voting (RCV) because "the word 'instant' might create an expectation that final results will be available immediately after the polls close on election night." The department releases first choice totals immediately, but used to wait until more absentee ballots have arrived before running instant runoff ballot counts. That practice has since changed.

Results

Ranked Choice Voting has been used since 2004 to elect its Board of Supervisors and major citywide offices. This implementation allows the voter to rank three candidates and uses sequential candidate elimination until one candidate earns a majority of votes cast for remaining candidates.

As of November 2015, there have been 56 elections using the RCV ballot.

2004 San Francisco results

There were four elections that used the instant runoff process after no majority winner in the first round: Districts 1,5,7, and 11.

• District 1: There were 7 candidates, reduced to 2 candidates in 4 rounds. The winner won 54% of the final round count, which amounted to 48.67% of the total first-round votes, with 9.89% of the ballots exhausted by the final round.
• District 5: There were 22 candidates, reduced to 3 in 19 rounds, when the winner had a majority of active ballots. The winner finished with 50.6% of the final round vote against two runners up, which amounted to 37.63% of the first-round vote with 25.63% of the ballots exhausted.
• District 7: There were 13 candidates, reduced to 2 in 11 rounds. The winner finished with 57% of the votes cast for the two active candidates in the final round. Of first round votes, this amounted to 43.72% of the first-round vote, with 23.12% exhausted ballots.
• District 11: There were 8 candidates, reduced to 2 in 6 rounds. The winner finished with 58% of the final round vote, which amounted to 46.08% of the total first-round vote, with 21.01% of the ballots exhausted.

The District 5 results are included below as the largest election from 2004 and most round of counting. The elimination table shows the candidates reordered by their elimination. The elimination process was stable for the highest 5 candidates, holding their same plurality ranking each round despite the 19 rounds of elimination and transfer votes.

The IRV elimination process was halted when candidate Mirkarimi reached more than 50% of the active ballots, but only 37.6% of the total first-round ballots. This stopping point is pragmatic for picking a winner, but fails to show how many votes the winner had compared to only the strongest runner up candidate.

2004 San Francisco District 5 - Ranked-Choice Voting

Candidate	Pass 1	Pass 2	Pass 3	Pass 4	Pass 5	Pass 6	Pass 7	Pass 8	Pass 9	Pass 10	Pass 11	Pass 12	Pass 13	Pass 14	Pass 15	Pass 16	Pass 17	Pass 18	Pass 19
ROSS MIRKARIMI	9947	9950	9952	9969	9996	10034	10094	10158	10261	10387	10472	10635	10766	10946	11262	11659	11921	12287	13211
ROBERT HAALAND	5124	5126	5130	5146	5180	5192	5226	5254	5318	5384	5461	5538	5628	5740	5956	6319	6409	6636	7272
LISA FELDSTEIN	3257	3265	3274	3289	3309	3323	3381	3430	3484	3566	3671	3765	3851	4070	4313	4636	4759	5064	5628
NICK WAUGH	3025	3025	3027	3035	3053	3070	3090	3118	3187	3243	3296	3391	3441	3540	3732	3900	4063	5041
ANDREW SULLIVAN	2477	2478	2479	2494	2501	2550	2570	2580	2639	2663	2716	2771	2831	2870	2982	3068	3601
JIM SIEGEL	1540	1542	1543	1551	1565	1608	1639	1657	1743	1763	1820	1866	2053	2111	2184	2242
BILL BARNES	1664	1670	1671	1680	1690	1709	1719	1731	1751	1804	1871	1945	1977	2018	2142
DAN KALB	1398	1400	1400	1412	1430	1449	1466	1493	1540	1582	1610	1698	1739	1867
SUSAN C. KING	977	980	984	1007	1034	1051	1072	1116	1147	1206	1237	1293	1371
MICHAEL E. O'CONNOR	868	870	873	882	906	930	944	973	1012	1036	1079	1127
BRETT WHEELER	832	833	835	845	871	881	896	929	951	995	1026
JOSEPH BLUE	802	805	807	814	819	842	851	860	876	908
PHOENIX STREETS	657	658	660	673	699	714	731	752	771
TYS SNIFFEN	686	687	688	692	707	719	730	746
JULIAN DAVIS	418	422	429	443	462	467	481
EMMETT GILMAN	393	394	398	405	407	423
FRANCIS SOMSEL	368	368	370	379	381
ROB ANDERSON	336	341	342	346
VIVIAN WILDER	130	134	135
PATRICK M. CIOCCA	91	91	91
PHILLIP HOUSE	62	62
H. BROWN	57
Eligible Ballots	35109	35101	35088	35062	35010	34962	34890	34797	34680	34537	34259	34029	33657	33162	32571	31824	30753	29028	26111
Exhausted Ballots (-4146 no marks)		8	21	47	99	147	219	312	429	572	850	1080	1452	1947	2538	3285	4356	6081	8998
Total Ballots	35109	35109	35109	35109	35109	35109	35109	35109	35109	35109	35109	35109	35109	35109	35109	35109	35109	35109	35109
Plurality candidate %	28.33%	28.34%	28.35%	28.39%	28.47%	28.58%	28.75%	28.93%	29.23%	29.59%	29.83%	30.29%	30.66%	31.18%	32.08%	33.21%	33.95%	35.00%	37.63%

Notes:

• The source summary listed Eligible ballots each round as nonexhausted ballots. This is shown here.
• The source summary listed Exhausted ballots as ALL exhausted ballots, including ballots with no marks at all on the election. This table subtracts exhausted the 4146 ballots that supported no candidates to be consistent with other tables in this article.
• Similarly the source listed Total ballots to include eligible and exhausted ballots. Here the no-mark ballots were also removed from the total.
• The final column "Plurality candidate %" is based on votes for the top candidate divided by this adjusted total ballots.

2005 San Francisco results

There was one election requiring the instant runoff process to be performed, with 4 candidates and finding a 55% majority winner in two rounds.

November 8, 2005: RCV Assessor-Recorder

Candidate	Pass 1		Pass 2
PHIL TING	94062	47.21%	110053	55.24%
GERARDO SANDOVAL	71850	36.06%	79261	39.78%
RONALD CHUN	33294	16.71%
ANTHONY FABER	18	0.01%
Eligible Ballots	199224	100%	189314	95.03%
Exhausted Ballots (-26146 no marks)	0	0.00%	9910	4.97%
Total Ballots	199224		199224

2006 San Francisco results

There were two elections that required the instant runoff process, districts 4 and 6:

• District 4: There were 6 candidates which were reduced to 2 in 4 rounds. The winner ended with 52.5% of the final round vote for active candidates, which amounted to 42.33% of the first-round vote, with 19.38% exhausted ballots.
• District 6: There were 8 candidates and was stopped on the second round with 4 candidates remaining. The winner had 49.99% of the total first-round votes, with 48.37% divided among the 3 runners up, and 1.64% exhausted ballots.

The detailed runoff results for district 4 are:

San Francisco 2006 MEMBER, BOARD OF SUPERVISORS DIST. 4

Race and Candidate	Pass 1		Pass 2		Pass 3		Pass 4
ED JEW	5184	26.16%	5441	27.46%	6455	32.58%	8388	42.33%
RON DUDUM	5134	25.91%	5521	27.86%	6305	31.82%	7587	38.29%
JAYNRY MAK	4569	23.06%	5012	25.30%	5851	29.53%
DOUG CHAN	3236	16.33%	3414	17.23%
DAVID FERGUSON	1455	7.34%
HOUSTON ZHENG	234	1.18%
WRITE-IN	2	0.01%
Eligible Ballots (-2171 no marks)	19814	100.00%	19388	97.85%	18611	93.93%	15975	80.62%
Exhausted Ballots	0	0.00%	426	2.15%	1203	6.07%	3839	19.38%
Total Ballots	19814	100%	19814	100%	19814	100%	19814	100%

San Francisco results, 2007-2015

San Francisco held a number of IRV elections each November from 2007 to 2015. Several of the elections went to multiple rounds of counting. In 2010, for example, two candidates won who were not the leaders in first choice rankings. In 2011, all three citywide elections up for election - mayor, sheriff and district attorney—were decided in IRV tallies.

In the wake of the November 2012 elections, sixteen of eighteen offices elected by IRV were held by people of color.

Basalt, Colorado

The city of Basalt, Colorado adopted instant runoff voting in 2002 for mayoral elections in which there are at least three candidates. The city is ready to run instant runoff elections, but the elections in 2004, 2008, 2012 and 2016 did not have more than two candidates file for the mayor's office.

Ferndale, Michigan

The city of Ferndale, Michigan passed (68%) instant runoff voting in 2004 pending implementation.

Berkeley, California

The city of Berkeley, California passed (72%) instant runoff voting in 2004. The city used IRV for the first time in November 2010 for elections for four city council seats and the city auditor. Berkeley used IRV for electing its mayor in November 2012. It used RCV again in several city council elections in November 2014.

Burlington, Vermont

The city of Burlington, Vermont approved IRV for use in mayoral elections with a 64% vote in 2005. IRV was repealed 52% to 48% in March 2010.

2006 Burlington results

In 2006 it held its first mayoral election using IRV. Progressive Bob Kiss won in two rounds with 54.4% of the final round vote over Hinda Miller.

2009 Burlington results

• See also: Burlington, Vermont mayoral election, 2009

In 2009 Burlington held its second IRV election for mayor. Again, five candidates were on the ballot. In the initial tally of first choices, there was no initial majority winner, with Republican city council president Kurt Wright in the lead with 33% (2951), followed by the incumbent mayor, Progressive Bob Kiss with 29% (2585), Democratic city councilor Andy Montroll with 23% (2063), Independent Dan Smith with 15% (1306), Green Party candidate James Simpson (35) and write-ins (36). In the final round of the IRV tally, Kiss received 51.5% to Wright's 48.5%.

Candidates		1st Round		2nd Round		3rd Round
Candidate	Party	Votes	±	Votes	±	Votes	±
Bob Kiss	Progressive	2585	+2585	2981	+396	4313	+1332
Kurt Wright	Republican	2951	+2951	3294	+343	4061	+767
Andy Montroll	Democrat	2063	+2063	2554	+491	0	-2554
Dan Smith	Independent	1306	+1306	0	-1306
James Simpson	Green	35	+35	0	-35
Write-in		36	+36	0	-36
EXHAUSTED PILE		4	+4	151	+147	606	+455
TOTALS		8980	+8980

2010 Burlington repeal

Unlike Burlington's first IRV mayoral election in 2006, the IRV winner in 2009 (Kiss) was neither the same as the plurality winner (Wright) nor Condorcet winner (Montroll). The results caused a post-election controversy regarding the IRV method, mostly from supporters of Wright, the leader in first-choice rankings. IRV was repealed a year later by a vote of 52% to 48%.

The repeal reverted the system back to a 40% rule that requires a top-two runoff if no candidate exceeds 40% of the vote. Had the 2009 election occurred under these rules, Kiss and Wright would have advanced to the runoff. If the same voters had participated in the runoff as in the first election and not changed their preferences, Kiss would have won the runoff. In 2011, an initiative effort to increase the winning threshold from the 40% plurality to a 50% majority failed.

Minneapolis, Minnesota

The city of Minneapolis, Minnesota, passed (65%) instant runoff voting in November 2006. Although a citizen group filed a lawsuit in 2007 challenging the constitutionality of the system and to block its implementation, the lawsuit was dismissed in a ruling on January 13, 2009. The Minnesota Supreme Court unanimously upheld this ruling in an opinion on June 11, 2009.

On November 3, 2009, the City used instant runoff voting, now commonly known in Minneapolis as ranked choice voting, to elect the mayor, 13 city council seats and seven other local offices and used a multi-seat variation of IRV, the single transferable vote, for park board elections. In November 2013, it again used IRV for those same elections, including in the open seat Minneapolis mayoral election, 2013

Pierce County, Washington

Pierce County, Washington, passed (53%) instant runoff voting in November 2006 for most of its county offices. Voters upheld the 2008 implementation timing with a vote of 67% in 2007 and made minor adjustments to the charter language involving ballot access and numbers of rankings. Seven instant runoff voting elections took place on November 4, 2008 and one on November 3, 2009. The introduction of IRV was marked by controversies about costs and a Supreme Court ruling that restored Washington State's "top two" election system passed statewide in 2004, but struck down by courts in 2005. On November 3, 2009, voters repealed IRV.

Takoma Park, Maryland

The city of Takoma Park, Maryland adopted instant runoff voting for city council and mayoral elections in 2006 after voters approved it with 84% support in November 2005.

In January 2007 the first IRV election was held to fill a city council vacancy in a three-way race with a majority winner in the first round. Voters selected Reuben Snipper with 107 votes (52.7%), defeating Eric Hensal with 72 votes (35.5%) and Alexandra Quéré Barrionuevo with 23 votes (11.3%) and one write-in. Snipper said the possibility of using the IRV system changed the race's dynamics. "I had every reason to believe this was going to be a close race," he said. "It meant that when I knocked on a door, if a person indicated they were going to vote for another candidate, I didn't just leave right away. I tried to persuade them I would be a good second choice."

In November 2007 the mayor ran unopposed, and, out of six ward seats on the ballot, one was contested. Runoff provisions were not exercised. In November 2009, the mayor and one city councilor each faced one opponent. In November 2011, one city council race drew three candidates; it was won by a candidate securing a majority of first choices.

In July 2012, the Ward Five race again was vacant. In another three-way race, first-time candidate Jarrett Smith was elected. After securing 44% of first choices, Smith won a majority in the instant runoff against Eric Hensal.

In regularly scheduled city elections in 2013 and 2015, no elections went to a second round, but a special election in April 2014 did require an instant runoff.

Oakland, California

The city of Oakland, California passed (69%) a measure in November 2006 to adopt IRV for its city offices. In November 2010, Oakland used IRV to elect its mayor, three city council races and four other local offices, with elections for mayor and council district four requiring multiple rounds of counting. It used IRV in the city's remaining elected offices in 2012. IRV was again used in 2014, including in the mayoral election in which incumbent Jean Quan was defeated by Libby Schaaf.

Oakland mayoral election, 2010

Oakland's 2010 mayoral election was an open seat election in which no candidate earned more than 34% of votes in the first round. In the tally, candidates were eliminated sequentially, but three candidates were far ahead in first choices. After the count of first choices, Don Perata was in first place, Jean Quan in second place and Rebecca Kaplan in third. They remained in that order of votes after all other candidates were eliminated and their votes re-allocated. When Kaplan was then eliminated, Quan picked up 18,864 votes from Kaplan backers while Perata was the next choice of only 6,407 Kaplan backers. As a result, Quan won a final round majority when matched against Perata, which means she was ranked ahead of Perata on a majority of ballots in which one of them received a ranking.(Quan did not get a majority of all ballots cast in the initial round because some people did not rank either of them.)

Candidate	Round 1			Round 2			Round 3
	Votes	%(*)	Transfer	Votes	%(*)	Transfer	Votes	%(*)
Don Perata	40342	32.99%	+5123	45465	37.18%	+6407	51872	43.37%
Jean Quan	29266	23.94%	+5767	35033	28.65%	+18864	53897	45.06%
Rebecca Kaplan	25813	21.11%	+6906	32719	26.76%	-32719
JOE TUMAN	14347	11.73%	-14347
MARCIE HODGE	2994	2.45%	-2994
TERENCE CANDELL	2315	1.89%	-2315
DON MACLEAY	1630	1.33%	-1630
GREG HARLAND	966	0.79%	-966
LARRY LIONEL LL YOUNG JR.	933	0.76%	-933
ARNOLD FIELDS	733	0.60%	-733
Write-In	268	0.22%	-268
Exhausted by Over Votes	355	0.29%	+106	461	0.38%	+65	526	0.44%
Under Votes	2306	1.89%	"
Exhausted Ballots	0	0.00%	+6284	6284	5.14%	+7383	13667	11.18%
Continuing Ballots	119607	97.82%		113217	92.60%		105769	88.43%

(*) Percentages in rounds one and two are computed from the total ballots (122,268), including voters who did not participate in the election.

Oakland Elections, 2012

Oakland used IRV for several elections in 2012, including a citywide election for city attorney and for several seats on the city council and school board. Several races were decided in an instant runoff, including the District 3 city council race where the winner trailed in first choices. Of the 18 Oakland offices elected by IRV in 2010 and 2012, sixteen of the IRV winners received more votes than the previous winner had won before adoption in the last non-IRV election

North Carolina

In November 2010, North Carolina had three IRV elections for local-level superior court judges, each with three candidates, and a statewide IRV election for a North Carolina Court of Appeals seat (with 13 candidates). The Court of Appeals race is believed to be the first time IRV has been used in any statewide general election in the United States.

A 2006 law had established that IRV would be used when judicial vacancies were created between a primary election and sixty days before a general election. That law was repealed by the General Assembly in 2013 as part of a sweeping "voter ID" bill, meaning that special judicial elections with more than two candidates would once again be decided by a simple plurality.

The law also established a pilot program for instant runoff voting in the form of batch-elimination IRV for up to 10 cities in 2007 and up to 10 counties for 2008; to be monitored and reported to the 2007-2008 General Assembly.

Several municipalities considered participating in the IRV pilot in 2007. Cary, Hendersonville and Kinston voted to participate; Kinston dropped out because there were not enough candidates running to use IRV. Other cities declined to participate in the pilot. No NC counties volunteered to pilot IRV in 2008 elections held in conjunction with state and federal races. In August 2008 the governor signed legislation extending the pilot program for local elections to be held in 2009-2011.

There was much debate whether IRV was successful when it was used. This debate continued in the North Carolina legislature when it debated legislation to extend the pilot program. Some verified voting advocates contended that the IRV tabulation procedures used were not legal. Both advocates and opponents of the provision supported amendments to the pilot program to: ensure that the local governing body of any jurisdiction participating in the pilot must approve their participation; the jurisdiction must develop and implement voter education plans; and the UNC School of Government by January 2009 must approve procedures for conducting IRV elections. After these amendments were adopted, the state House of Representatives, by a majority of 65-47, rejected an amendment designed to remove the pilot program from the legislation, and the legislation ultimately won approval by both houses.

In 2009 Hendersonville again used IRV, while the Cary Town Council voted to use a traditional runoff method. Three candidates ran for mayor in Hendersonville in November 2009; five candidates ran for two seats on the city council using a multi-seat version of IRV. All seats were filled based on first choices without the need for instant runoffs.

In 2011, Hendersonville's city council unanaimously voted to use IRV a third time, although utltimately not enough candidates filed for office to trigger the need for the system.

Results (Cary, North Carolina)

In October 2007 the city of Cary, North Carolina used batch-elimination IRV for municipal election for three council seats and for mayor. The mayor's race (with two candidates) and two of the council seats (with four and three candidates on the ballot) were won with a majority in the first round. The remaining council seat, with three candidates, went to a second round of counting under the instant runoff system; the plurality winner in the first round went on to win with 50.9% of the final round vote, amounting to 46.4% of first-round ballots cast, with 8.9% of the ballots offering no preference between the top two candidates.

COUNCIL MEMBER C-B 1
CARY MUNICIPAL DISTRICT B:

Candidate	Round 1		Round 2
Don Frantz	1151	(38.1%)	1401	(46.4%)
Vickie Maxwell	1075	(35.6%)	1353	(44.8%)
Nels Roseland	793	(26.2%)	--
Other	3	(0.0%)	--
Exhausted ballots	--		268	(8.9%)
Total	3022	(100%)	3022	(100%)

Cary used hand or machine-marked paper ballots that are read on optical scanners manufactured by ES&S. First column choices were tallied at the precinct. The second and third column choices were counted at a central location.

Aspen, Colorado

Aspen, Colorado passed IRV in November 2007 for the mayoral race and for at-large council races with two winners. In March 2009, the Aspen council adopted a unique variation of IRV for the council races. A block voting tally based on the first and second rank choices was used to determine first round support. Any candidate with initial majority support was elected. If there were not two first-round winners, there was a batch elimination of low-placing candidates to reduce the number of continuing candidates before the instant runoff. In the latter case, separate IRV runoffs would be conducted for each council seat, with the winner of the first seat eliminated from the race for the second seat.

Aspen's first elections with IRV and the new city council system were on May 5, 2009. The number of voters was the highest in the history of Aspen elections. Mick Ireland was re-elected as mayor in the fourth round of a four-candidate race. Both city council incumbents were defeated in the two-seat IRV election in which nine candidates participated. The winners were selected after IRV tallies. 168 spoiled ballots were recast by voters alerted to errors by their optical scanning machine. The city reported 0% invalid ballots in the mayor's race and 0.9% invalid ballots in the two-seat city council elections.

The elections were close, and some Aspen observers argued that a traditional runoff system would have given more time to consider their top choices. There also was debate over how to implement audit procedures. In 2009 voters rejected an advisory measure to maintain IRV and in 2010 approved a binding amendment to return to a traditional runoff system.

Sarasota, Florida

The city of Sarasota, Florida passed IRV (78%) in November 2007. Implementation is contingent on conditions that were not met as of 2015.

Santa Fe, New Mexico

On March 4, 2008, the city of Santa Fe, New Mexico, passed a referendum for IRV (called Ranked Choice voting) by a vote of 5659 to 3044 (65% for).

• Charter Amendment 5 – RANKED CHOICE VOTING: Proposing to amend the Santa Fe Municipal Charter, Article IV, to create a new Section 4.06 to require that candidates for municipal office be elected by a majority of voters through a ranked choice voting system that combines the initial vote with an "instant" runoff in a single election. The instant runoff is achieved by allowing voters to rank, in order of their preference, each candidate for mayor, city councilor and municipal judge. If, after counting all voters’ first choices listed on their ballots for an office, no candidate has received a majority of votes cast, the candidate with the fewest votes is eliminated. Each ballot listing the eliminated candidate as first choice is then recounted using that ballot’s second choice. If still no candidate for the office has received a majority, the process is repeated until a candidate receives a majority of votes for that office.
• Ranked choice voting would go into effect commencing with the regular municipal election in March 2010 or as soon thereafter when equipment and software for tabulating the votes and allowing correction of incorrectly marked, in-person ballots are available at a reasonable price.

Memphis, Tennessee

On November 4, 2008, voters in the city of Memphis, Tennessee, passed a charter amendment with 71% of the vote to enact IRV for city elections.

Telluride, Colorado

On November 4, 2008, voters in the town of Telluride, Colorado, passed an ordinance with 67% of the vote to adopt IRV for the next three mayoral elections, starting in November 2011 if three candidates file for the office. The system was used for the city's 2011 mayoral election. The incumbent mayor Stu Fraser was re-elected by securing a majority of first choices. in the 2015 mayoral election, Sean Murphy handily won an open seat election for mayor after trailing in first choices.

Saint Paul, Minnesota

On November 4, 2009, voters in the city of Saint Paul, Minnesota, passed a charter amendment with 52% of the vote to adopt IRV for future elections for mayor and city council. In February 2011, the city council adopted rules governing the November 2011 elections. IRV elections took place for city council races, with two council races requiring multiple rounds of counting.

IRV was used in St. Paul's 2013 election for mayor and in an open seat election for city council. IRV also was used for city council elections in 2015, including one election decided in an instant runoff.

San Leandro, California

In November 2000, the voters of San Leandro, California approved a charter amendment by 63% to 37% requiring use of a two-round runoff or IRV if no candidate won a majority of first round votes. In January 2010, the city council voted 5-2 to use IRV for its elections for mayor and three city council seats in November 2010. The mayor's race required multiple rounds of counting.

In November 2012, San Leandro held IRV elections for three city council seats. One election was decided in first choices, and two with an instant runoff. In November 2014, San Leandro used RCV for four city council seats.

Portland, Maine

Main article: Portland, Maine mayoral election, 2011

In November 2010, voters in Portland, Maine, adopted a charter amendment with 52% to establish a directly elected mayor, using instant runoff voting. The first election was in November 2011. Fifteen candidates ran. The winner Michael Brennan led with 27% of first choices and won decisively in the final instant runoff voting.

In November 2015, Brennan ran for re-election against two opponents and was defeated by Ethan Strimling.

Political party usage

Democratic Party of Charlottesville, Virginia

In May 2009, the Democratic Party of Charlottesville, Virginia, held its first "open caucus" to select its nominees for city council and sheriff, using instant runoff voting. Voter turnout was close 1,600 voters. There were two incumbents and one challenger running for two nominations in the city council race. One incumbent was renominated and another defeated without the need for an instant runoff. Three candidates ran in the sheriff's race. No candidate won an initial majority. In the instant runoff, James E. Brown III defeated Mike Baird.

In August 2011, the Party again used to nominate candidates. Voter turnout rose to 2,582 in the city council race for three nominations. Two candidates were nominated with a majority of the first round vote. The final nomination was determined by IRV.

Republican Party of Utah

After voting to authorize its use, the Republican Party of Utah used instant runoff voting in 2002, 2003 and 2004 at its statewide convention, including in a contested race to nominate a governor in 2004. In 2005, Republicans used repeated balloting for its statewide convention and has done so in subsequent years. Some county Republican parties like Cache County continue to use instant runoff voting at their conventions, and IRV was used by Republicans to fill several state legislative vacancies in 2009-2011.

Democratic Party of Virginia

In 2014, the Democratic Party of Arlington used IRV in two "firehouse primaries" for countywide office that each drew several thousand voters, and it joined with the Democratic Party of Fairfax county that year to use IRV in a seven-candidate primary election for a special election for the House of Delegates. IRV was also used by leaders of the Henrico County Democrats in a three-candidate special election nomination contest for the House of Delegates in December 2014

Independence Party of Minnesota (2004 Presidential poll)

In part to increase awareness of the voting method and to demonstrate it in a real-world situation, the Independence Party of Minnesota tested IRV by using it in a straw poll during the 2004 Minnesota caucuses.

The poll allowed a none of the above option which could not be eliminated. Their rules eliminated one weakest candidate at a time, or all candidates in a tie at the bottom. They continued the elimination until only one candidate remained to confirm that this candidate had more support than NOTA.

This summary table shows the first round, and final five rounds, excluding five rounds during which 18 weak candidates were eliminated.

Candidate/Round	1	…	7	8	9	10	11
John Edwards (Democrat)	94 (20.75%)		106 (23.40%)	131 (28.92%)	166 (36.64%)	210 (46.36%)	335 (73.95%)
John F. Kerry (Democrat)	100 (22.08%)		110 (24.28%)	117 (25.83%)	132 (29.14%)	149 (32.89%)	0 (0.00%)
George W. Bush (Republican)	77 (17.00%)		84 (18.54%)	85 (18.76%)	94 (20.75%)	0 (0.00%)
Ralph Nader (none)	70 (15.45%)		73 (16.11%)	78 (17.22%)	0 (0.00%)
Dennis Kucinich (Democrat)	32 (7.06%)		40 (8.83%)	0 (0.00%)
18 others (<10 votes each)	48 (10.60%)		0 (0.00%)
None of the above	32 (7.06%)		39 (8.61%)	40 (8.83%)	50 (11.04%)	61 (13.47%)	66 (14.57%)
Exhausted ballots	0 (0.00%)		1 (0.22%)	2 (0.44%)	11 (2.43%)	33 (7.28%)	52 (11.48%)
Total	453		453	453	453	453	453

Absentee use

Several states jurisdictions that hold runoff elections allow certain categories of absentee voters to submit IRV ballots, because the interval between votes is too short for a second round of absentee voting. IRV ballots enable long-distance absentee votes to count in the second (general) election round if their first choice does not make the runoff. Alabama, Arkansas, Mississippi, Louisiana and South Carolina used IRV ballots for overseas voters in 2014 and will do so in 2016. A city using this practice is Springfield, Illinois after voters approved it with 91%. Louisiana uses it also for out-of-state members of the United States military. or who reside overseas.

Implementations rejected

According to FairVote, an organization advocating IRV, dozens of states have entertained instant runoff voting legislation since 2000. For example, in 2008, Vermont governor Jim Douglas vetoed legislation to establishing instant runoff voting for that state's congressional elections starting that year despite testimony in support by Vermont's U.S. Senator Bernie Sanders and its U.S. House Member Peter Welch. In 2003, an amendment to the California State Constitution was proposed with wide-ranging goals of election reform, including instant runoff voting for statewide offices. In the state of Washington, an initiative seeking to adopt instant runoff voting in 2005 failed to garner enough signatures. The city of Vancouver, Washington voted in 1999 to adopt instant runoff voting and the state legislature enacted enabling legislation in 2004, but the city in 2006 chose not to exercise its option. Instant runoff voting for all state and federal elections was on Alaska's statewide ballot in August 2002, when it was defeated. It also was defeated by voters in Glendale, Arizona, in 2008 in Fort Collins, Colorado in 2011 and in Duluth, Minnesota in 2015.

In the U.S. Congress, the "Voter Choice Act of 2005" sought to require the use of instant runoff voting for general elections for federal office.

External links

• Explainer video: How does ranked-choice voting work?, Minnesota Public Radio MPR.org