Tuesday, June 15, 2010

Response to some recent attacks on IRV

Some opponents of election reforms such as instant runoff voting (IRV) have been circulating certain supposed failures of IRV. These opponents' claims need correction.

1. These opponents have claimed that IRV is costly.

In fact, instant runoff voting can save money immediately – it depends on the context.

Whether IRV will cost money or not depends on the situation. San Francisco’s use of IRV resulted in the City avoiding a number of runoff elections that would have cost millions of dollars more to administer than the cost of implementing IRV. It cost the city of San Francisco $2.4 million dollars to implement IRV, including $1.6 million for a one-time upgrade of voting machines and $800,000 to educate people on the new system. According to Gerard Gleason of the San Francisco Election Commission, now that instant runoff voting (locally called "ranked choice voting") is well established, voter education and poll worker training costs are insignificant. It has had some ongoing costs, but they are minor and far less than the costs of administering runoff elections. As for savings, in the first year of its use, IRV saved the city of San Francisco $1.2 million dollars by avoiding runoffs in district races for the Board of Supervisors, and has avoided the need for at least one citywide runoff that would have cost approximately $3 million as well as several additional district runoffs in other years.

Cary (NC) and Hendersonville (NC) are two cities that have participated in a state pilot program similar to the program envisioned in the New York legislation – a law first passed in 2006 and extended and expanded for three more years in 2008 after two IRV elections in 2007. The Wake County Board of Elections director Cherie Poucher estimates that IRV saved Cary $28,000 in its election in 2007, and would have saved as much as four times that amount if the mayor’s race had gone to a runoff. Hendersonville has implemented IRV with little cost, and while no runoffs have been avoided, savings would have been immediate if there had been runoffs. After voting to use IRV for a second time in 2009, the Hendersonville city council in 2010 voted unanimously to explore how it could make IRV a permanent part of its elections.

2. These opponents have suggested that rather than enhancing voter participation, IRV could reduce turnout, citing Minneapolis as an example.

In fact, IRV generally increases participation for picking decisive winners, but there is no guarantee.

IRV tends to improve voter participation, particularly when replacing two rounds of voting where either the first round or second round can have much lower turnout. But as a general matter, turnout is mostly driven by how exciting elections are, rather than the voting method. In the case of Minneapolis in 2009, the popular incumbent mayor had no serious opponent, and of course won in a landslide. When the outcome of an election is a foregone conclusion, turnout is generally low, regardless of method. Local scholars have dismissed IRV as a reason for the turnout decline in Minneapolis, pointing out that neighboring St. Paul had exactly the same dynamic with its mayor, and its turnout drop in 2009 was larger than that of Minneapolis. At the same time, Minneapolis avoided the need for its September primary, where turnout historically had been very low, yet eliminated most candidates.

When IRV combines two round runoff systems into a single election the increased voter participation is most dramatic. An analysis of voter participation (http://www.fairvote.org/assets/turnout.pdf) in San Francisco by Dr. Christopher Jerdonek found that the use of instant runoff voting in San Francisco’s November 2005 election increased voter participation in the decisive round of the Assessor-Recorder race by an estimated 2.7 times, or 120,000 voters of what would have happened in a December runoff that year. In six out of twenty-five neighborhoods, it is estimated that voter participation in the decisive round tripled due to RCV. (http://www.fairvote.org/assets/turnout.pdf)

3. These opponents have suggested IRV may harm racial minorities.

In fact, IRV has proven easy for voters of all races and has elected diverse representation in places like San Francisco.

IRV is a majoritarian, winner-take-all voting method rather than a proportional voting method designed to represent those in the minority. For that reason, it is neither more nor less likely to elect racial minority candidates than existing single-winner methods in a general matter, although traditional racial minorities can have trouble competing as effectively in expensive, one-on-one runoff elections as better-financed white candidates.

Nevertheless, there is convincing evidence that racial minorities easily adapt to using ranked ballots and, indeed, utilize IRV very effectively. Several studies of the San Francisco elections have shown that minority voters were just as likely to effectively utilize their rankings as other voters, with racially diverse districts actually decreasing the rate of residual votes (under-votes and over-votes) compared to non-IRV elections. Also, because the separate runoff election was eliminated, voter participation in the most racially diverse districts of San Francisco increased more than in white districts -- by an astonishing 307% compared to separate runoffs. http://www.fairvote.org/assets/turnout.pdf

The City with the longest use of instant runoff voting is San Francisco. Its 11 members of its Board of Supervisors include 3 Asian Americans, 2 Latinos, 1 African American and 1 Persian American. A full analysis of the impact of IRV on racial minorities in several cities is available here

4. These opponents claim that IRV usually produces a plurality winner and fails to elect majority winners.

In fact, IRV elects candidates with majority support over their top opponent.

As in traditional runoff elections, it is not surprising that the most common winner of the runoff tally is the leader of the first round. But, when a “spoiler” scenario has split the majority among similar candidates, IRV – just as in a separate runoff system – allows that majority to re-coalesce around the strongest candidate, resulting in a "come-from-behind" victory. Australia has used IRV to elect their federal House of Representatives for generations. An analysis of these elections from 1949 - 2007 shows that on average, in 16% of those contests that went to an instant runoff tally, there was a reversal with the first round leader being defeated. In other words, one out of six “plurality-only” leaders was in fact not the majority choice when the field narrowed to two. http://blogs.abc.net.au/antonygreen/2010/05/preferential-voting-in-australia.html

In addition, as in separate runoff elections, some voters abstain from the runoff (either stay home if a separate runoff, or don't rank either of the two finalists in IRV). This can mean that the "majority" winner of the runoff gets fewer than half the votes cast originally in the first round. However, because IRV combines the two elections into one, the drop-off due to these "exhausted" ballots is generally much less severe than the drop-off occurring in separate runoff elections. Under separate runoff election rules, a "majority" winner often receives fewer votes in this runoff round than the "loser" received in the first round of voting – something that can never happen with IRV. For example, turnout in a 2009 citywide race with IRV in Burlington (VT) declined by 6% between the first round and decisive instant runoff. Turnout in an actual runoff in a city council race (which did not use IRV) that year in Burlington declined by 45%. Of all 116 federal primary runoffs in 1994 to 2008 in states around the country, fully 113 had declines in turnout, with the average drop in turnout of 35.1%.

5. These opponents claim that IRV leads to "2 party domination."

In fact, IRV accommodates voter choice, but does not represent those who can’t win majorities.

Some opponents of the American "two-party system" have suggested that IRV would simply entrench "2-party domination." IRV neither overthrows nor entrenches the current predominance of two major parties. IRV does allow minor parties to exist, contend for office, and possibly eventually win office, 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 defeat candidates who can muster that majority support.

6. These opponents claim there is never enough voter education for IRV.

In fact, voter education is always good, but often not necessary.

Voter education has been more than adequate in every implementation of IRV in the U.S., as indicated by the fact that there have been very small numbers of ballots that did not indicate a valid first choice. In 2009, for example, Minneapolis for the first time implemented IRV to elect the mayor, city council and several other offices. A survey found that 95% of voters found the new election process easy to use and that the entire election produced just one defective ballot that could not be counted for its first choice among 45,968 cast.

The key is a good ballot design, as the voters' task is simple. Voters can 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. This in many ways makes voting with IRV easier than having to calculate who is a credible candidate under the current plurality method.

Scholars have conducted formal voter surveys in several U.S. cities that recently implemented IRV (see http://archive.fairvote.org/?page=2170) to assess voter acceptance of the new system. Without exception, in every city, voters have overwhelmingly favored IRV over the old method. Also, studies of the San Francisco and Burlington IRV elections have proven that there was no increase in uncountable ballots (spoiled or skipping the IRV race) with the adoption of IRV. In the Burlington IRV election in 2006, for example, 99.9% of ballots cast in the IRV race for mayor were valid – and it rose to 99.98% in 2009 (a single invalid ballot). People had no difficulty voting; news reports indicated that poll workers on hand to explain IRV had an uneventful day.

The two nations with the highest voter participation rates in the world, Australia (which also has mandatory voting) and Malta, both use ranked choice ballots

7. These opponents claim IRV "leaves some voters behind."

In fact, voters use IRV effectively even without knowing the details of how the count works.

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 complexities of the internal electronics or vote for president, without understanding the constitutional complexities with the Electoral College. IRV is easy for voters to use, and a well-informed voter does not get an advantage over a less-informed voter who indicates their favorite choice first, second-favorite second and third-favorite third – just as suggested by the instructions.

A traditional vote-for-one method can create great strategic complexity for voters when there are more than just two candidates, due to the "spoiler" dynamic. However, IRV dramatically reduces the need for such calculations, as voters have less concern about "wasting" their votes, and second choices can never hurt that voter's first choice. While no voting method can completely eliminate every possibility of strategic "gaming," IRV allows for less strategic gamesmanship than either plurality elections or two-round runoffs.

8. These opponents claim that IRV is too difficult and complex to count.

In fact, jurisdictions have developed fair and secure IRV counting methods.

Because voters can say more with a ranked ballot, and it effectively combines two elections into one, tallying IRV ballots is more involved than a “vote for one with an “X” form of ballot. Nevertheless, a number of jurisdictions have met this challenge, either using special state-certified voting equipment (as in San Francisco, and, later this year, Oakland in California), new procedures (including a hand-count after tallying first choices at the polls, as done in 2009 in Minneapolis) or procedures using existing equipment (as in North Carolina, where all local elections must use federally certified equipment).

IRV elections can be readily audited, as is routinely done. Many jurisdictions audit a random selection of voting machines to compare the machine record with the paper ballots. There is no need to transport ballots to a central location, though this is an option for a recount. In Ireland’s national elections, the IRV election for the national president is hand-counted at local voting centers in less than a day. Australian jurisdictions get an unofficial full election night tally by hand, then a final official central count.

Jurisdictions in the U.S. typically collect voting machine data to run the IRV tally election night, with a manual audit later. Because IRV supports the capture of complete ballot images from paper ballots, rather than mere candidate totals, it can markedly enhance election integrity. It is the redundancy of having both a paper and electronic record of every ballot that makes fraud especially difficult. http://www.fairvote.org/ranked-voting-and-election-integrity-2/

Featured Quote "I have had the great fortune to be a small part of what could easily be considered the most significant civic exercise in the history of Minnesota government: the implementation of the first Ranked Choice Voting election in Minneapolis and in Minnesota. We proved that it could be well administered, quickly and accurately counted,
and that voters had little problem with the concept."
- Former city elections director Patrick O'Connor, who oversaw implementation of IRV in Minneapolis in 2009

9. These opponents claim jurisdictions are abandoning IRV.

In fact, most jurisdictions with IRV are keeping it -- only two have repealed it.

Some opponents of IRV imply that instant runoff voting has been repealed most everywhere it has been adopted. This is not true. Let's look at the facts, as of the date of this post.

At the start of the 21st century, no city in the United States used instant runoff voting. Since then, 14 cities and counties have passed ballot measures adopting IRV, five of which have already used it. Two cities have voted to use it on a one-time basis as part of a pilot program. In addition, more than 50 major colleges and universities have adopted IRV for student elections, along with even more associations, including several with more than 100,000 members. Internationally, IRV has been adopted and used for electing mayors in cities like London (United Kingdom) and Wellington (New Zealand) for national elections in Papua New Guinea. The new British government has committed to holding a national referendum on IRV for electing the House of Commons.

Returning to the United States, there have been two repeals of IRV among those 14 cities that passed it at the ballot: Burlington (VT) and Pierce County (WA). Repeals took place over the opposition of the local League of Women Voters and were tied to special partisan calculations that rarely will be repeated. At the same time, several cities are moving toward IRV, including action this year by the California city councils of Oakland, Berkeley and San Leandro to use it this November. Despite what some might claim, no other city has repealed IRV; as one example of opposition claims, Sunnyvale (CA) this year changed a never-used feature to have the 7-member city council use IRV when electing one of its members to be mayor – backers of this change stressed that their vote had nothing to do with their views on using IRV in actual elections.