Testimony of Robert Richie

Executive Director, The Center for Voting and Democracy
to the House Government Operations Committee, Vermont House of Representatives
   February 1, 2000


Good afternoon. It is a pleasure for me to have the opportunity to speak to you today about such an important issue as House Bill 199, the legislation to enact instant runoff voting for your statewide and federal elections. I am aware that there are other pressing concerns in Vermont -- ranging from the debate over same-sex marriage to the primary elections today in New Hampshire -- and I appreciate your recognition that debate about election laws may at times sound dry on paper, but have important long-term repercussions for the health of a representative democracy.

I am executive director of The Center for Voting and Democracy. Based in Washington, D.C, our Center is the premier national organization to research and disseminate information on how election systems affect voter participation and representation. We focus in particular on how to reform plurality elections to promote majority rule and participation. We are a 501(c)(3) educational organization.

I have been executive director of the Center for Voting and Democracy since 1992. Among my activities at the Center are:

  • addressing the Voting Section of the U. S. Department of Justice, the Texas Commission on Judicial Efficiency, the annual meetings of the National Association of Counties and the National Conference of State Legislatures and several groups of foreign dignitaries through the auspices of the United States Information Agency;
  • testifying in special sessions before charter commissions in localities such as Nassau County (NY), Miami Beach (FL), Cincinnati (OH) and Detroit (MI) and before state legislative committees in Alaska, North Carolina, Vermont and Washington;
  • providing advice to several state legislators and Members of Congress in drafting legislation relating to election systems and reapportionment;
  • speaking as a panelist at national conferences of such other organizations as the Joint Center for Political and Economic Studies, U.S. Term Limits and National Black Caucus of State Legislators;
  • appearing as a guest lecturer in numerous classes at such universities as Georgetown Law School, the American University School of Law, Princeton University, the University of Pennsylvania and Duke University;
  • working closely with our board and advisory board, members of whom include Arend Lijphart, 1995-96 president of the American Political Science Association, and leading electoral system scholars such as Douglas Amy (Mount Holyoke College), Mark Jones (Michigan State University), Rein Taagepera (UC-Irvine) and Joseph Zimmerman (SUNY-Albany).
  • writing numerous articles for publications such as the Washington Post, New York Times, Legal Times and Wall Street Journal -- I also am co-author of an extensive analysis of instant runoff voting that will appear in the next issue of the National Civic Review and of a recent commentary in the Jackson Clarion-Ledger that argued instant runoff voting would be a sensible reform of their method of electing their governor to replace the current provision that the legislature elects the governor if no candidate wins a majority -- a similar situation as in Vermont that in led to the legislature having to elect the governor last month.

As regards House Bill 199, I am not a Vermonter and will not presume to tell you whether instant runoff voting is the ideal system for your state. Nevertheless, I can discuss the following matters with great confidence: 1) how instant runoff voting works; 2) the history of instant runoff voting; 3) how and why instant runoff voting is gathering growing support around the nation and in other countries. I also want to stress that our Center will do whatever we can to help you answer any questions that arise in this hearing. In fact, some of you already may have had an opportunity to talk with Caleb Kleppner, director of our "majority rule project." With his Vermont roots, Caleb has a particular interest in your consideration of instant runoff voting and is an excellent resource for explaining instant runoff voting.

How instant runoff voting works

Instant runoff voting is a well-established voting system for electing a single winner from a field of candidates for a particular office. Note that the system has been called other terms as well. In the United Kingdom, where a version of it will be used to elect the mayor of London this spring and where it recently was recommended by a high-level commission for elections to the British House of Commons, instant runoff voting is called "the alternative vote." In Australia, where it has been used for decades, instant runoff voting is called "preferential voting."

Each of these names suggest how the system works. "Preferential voting" is an appropriate name because voters are given the opportunity to rank candidates in order of preference rather than just vote for one. "The alternative vote" makes sense because voters have the chance to indicate alternative selections if their first choice candidate is too weak to win. "Instant runoff voting," our preferred name for the system, is derived from the fact that the method of counting ballots simulates a series of runoff elections, with the goal being a majority winner.

With instant runoff voting, each voter has one vote, but is able to express preferences for more than one candidate. The voters rank candidates in order of choice in a way that, for them, is "as easy as 1, 2, 3." Ballot-counting simulates a series of runoff elections. In the initial round, all first choices are counted, just as they would be in a traditional plurality election. If no candidate wins a majority of first choices, however, the election is not finished. Instead, the weakest candidate -- the candidate with the fewest first-choices -- is eliminated. A second round of counting then takes place, with ballots counted for the top-ranked candidate on each voter's ballot. (In practical terms, this generally means that election administrators tabulate the second choices of the voters who ranked the eliminated candidate first; the remaining ballots continue to count for the first-choice candidate on them because that candidate is still in the running to win the election. As I understand H.199, this tabulation would occur in Vermont by a court-appointed recount committee rather than by exhausted local polling officials that night.) Last-place candidates are successively eliminated and ballots redistributed to next choices until a candidate has won over 50% of votes or only one candidate remains. Note that this recount process only takes place if a candidate fails to win a popular majority.

The experience of other nations using instant runoff voting shows that the system is easy for voters, although I of course would recommend an educational outreach effort about the new system, as with any change to the electoral process. Voters of course can vote just as they do now, ranking only one candidate -- indeed, it is highly likely that most voters' ballots will count for their first choice throughout the ballot-counting, and I like the provision in the Vermont legislation that requires the ballots to look very much as they do now. At the same time, voters now have additional options that in many ways make voting easier. Instead of having to make uncertain calculations about candidate viability and "spoilers," voters have every incentive to simply rank their favorite candidate first, their next favorite second and so on. The reason is simple: they know that their ballot can still count toward a winner if their first choice loses. They also know that ranking a lower choice will never count again one of their higher choices. A ballot only counts for a lower choice if higher choices have lost and failed to advance to the "runoff."

There are strong arguments in favor of instant runoff voting -- ones that are generally accepted by political scientists who study comparative politics like those mentioned above. For example, instant runoff voting:

  • creates a clearer mandate for the winner through that candidate's accumulation of a majority of the vote and the reduction of the number of votes cast for losing candidates;
  • increases the range of choices available to a voter without resulting in "spoiler" candidacies that can undercut popular will;
  • reduces excessive negative campaigning by creating incentives for candidates to be the second choices of supporters of other candidates;
  • tends to boost voter turnout for these reasons.

In Vermont, of course, you have a special interest in a system that produces a majority winner because of your constitutional provision that the legislature will elect your governor and other important statewide offices if no majority winner is produced in the election. The legislation's sponsor, Rep. Terrill Bouricius, contacted us when exploring options for solving the potential problem of the adoption of public financing in Vermont making it more likely that there will be plurality winners in these elections. We offered the instant runoff voting system as a sensible match for the needs of Vermont. (Note that while our national and international expertise has prompted Rep. Bouricius to seek out our information and assistance, likewise we have also found Rep. Bouricius's political science degree and extensive background in electoral matters, to be a great resource for the Center as well.) It is encouraging to see the broad range of support for the concept in Vermont barely two years after we were contacted by Rep. Bouricius.

Finally, note that while instant runoff voting simulates a two-round runoff election, it has definitive advantages over these traditional runoffs. For example, two-round runoffs: require candidates to fund two rounds of elections; require taxpayers to fund a second election; and can vary greatly in voter participation and enthusiasm between the first and second rounds, particularly when the first-round takes place during a high-profile November election.

The history of instant runoff voting

[Note: I don't want to go in to more detail than the committee wishes, but wanted to ask if you would be interested in a thumbnail sketch of the history of instant runoff voting. Shall I proceed or skip over to information about concerns about the system and about where states and cities are now debating IRV?]

A key to the development of instant runoff voting was the invention of the single transferable vote (STV) in the 1850's by political theorist Thomas Hare in England. The essence of the transferable vote is the concept that a citizen would continue to have one vote in a particular contest, but that this one vote could have a greater chance to be effective by counting for a lower choice if it could not help elect a higher choice.

Hare devised the transferable vote balloting and counting procedure in creating a system of proportional representation for which he received great praise from the renowned political philosopher John Stuart Mill. Instant runoff voting, however, is not a system of proportional representation. Instead, it makes use of the transferable vote innovation in a winner-take-all context. The instant runoff voting form of transferable votes in fact was invented in 1870 by an American: W. R. Ware, a Professor at the Massachusetts Institute of Technology.

The first known use of a variant of instant runoff voting in a governmental election was in 1893 in Queensland, Australia. The "graduated runoff" concept that we understand today as instant runoff voting was first used in a public election in Western Australia in 1908. The system soon after was adopted to elect the Australian house of representatives, which is the main seat of power in the nation. Instant runoff voting is widely popular in Australia. The system is also used in the Republic of Ireland for its presidential elections, while the proportional representation variant is used to elect the parliament. Both uses of instant runoff voting are well-established and very unlikely to be changed; we have correspondence to this effect from the chief election administration official in the 1997 presidential race in Ireland.

Instant runoff voting has been debated in the United Kingdom for decades, and twice this century was nearly adopted for electing the House of Commons. Most recently in the United Kingdom, a high-level commission (headed by Roy Jenkins, one of the nation's most influential politicians) released a report in October 1998 that recommends instant runoff voting for electing most seats the House of Commons. The Jenkins Commission proposal almost certainly will be put before the voters in a national referendum within the next few years, although it is possible that instant runoff voting might be adopted without a referendum; in the meantime, a version of instant runoff voting will be used to elect the first elected mayor of London in this spring.

In the United States, instant runoff voting was used in 1975 in Ann Arbor Michigan. The presence in Ann Arbor of a third party, the Human Rights Party, had created lively three-way elections that had resulted in splintering of the Democratic Party's vote. After the adoption of instant runoff voting, the 1975 defeat of the Republican mayor by a Democrat -- very clearly on the strength of second-choice votes transferred from the Human Rights Party candidate -- prompted backers of the former mayor to push to eliminate the system. A legal challenge failed, but the bitter partisanship behind the results of the election led to a repeal in a low-turnout special election.

The proportional representation variant of instant runoff voting has been used to elect the city council and school committee in Cambridge, Mass. The main lessons to be drawn from that experience for those interested in instant runoff voting are: 1) voters clearly can handle the opportunity of ranking candidates, as there are very few spoiled ballots despite the fact that Cambridge has many low-income and elderly residents; 2)d voter turnout is higher than in jurisdictions near Cambridge that have similar demographics and levels of education; 2) the election administration can be modernized, as Cambridge has fully automated its ballot-count, although I realize this is not an issue for Vermont with H. 199, given that no new voting equipment will be necessary here.

Among private organization elections using instant runoff voting, perhaps the most instructive is the American Political Science Association (APSA), the organization of political science professors. Recognizing that the voting system cannot be manipulated and is a fair, simple way to elect a majority winner from a field of candidates, the APSA uses IRV to elect its national president.

Addressing objections

I want to stress that no voting system is perfect, in part because there is no consensus on exactly what we want representative democracy to look like. I can say with confidence that the instant runoff is a clear improvement over plurality voting in one bedrock definition of democracy: electing a candidate with majority support. Still, I wanted to address briefly three concerns that have been raised:

1. Instant runoff voting is not too confusing: As mentioned before, any change to the electoral process should lead to education to voters about the change. That should be done, and Australia and Ireland have done a fine job in creating models for this outreach -- we certainly would be happy to assist your election officials in that process in any way possible. But note that the empirical evidence is strong that voters like the opportunity to rank candidates and can quickly learn the new rules. Voters in Australia and the Republic of Ireland, where turnout is far higher than in the United States, have used the system for decades and are quite content with the system. In fact, Americans make decisions through ranking choices all the time: in choosing what meal to order at a restaurant, what television program to watch and so on. In elections, most people in elections with a multi-candidate field choose a candidate to support after an internal process of weighing and ordering candidates. They mentally divide candidates into "worth considering" and "not worth considering," then further divide the candidates into serious contenders for their vote and long-shots, and finally settle on one, their top pick. Certainly that has taken place among Republican primary voters in New Hampshire. And as one example, consider asking a small child about her favorite ice cream. Chocolate, she might way. And what if there is no chocolate? Then she'll have strawberry. And if there's no strawberry, she'll settle with vanilla. The child just ranked three candidates: chocolate, strawberry, vanilla. That�s all there is to the instant runoff.

2. Instant runoff voting does not undercut the two-party system: As most elected officials in the United States are members of the two major parties, some have expressed concern that instant runoff voting will undercut the two-party system. While it is true that the system makes it easier for a voter to consider a minor party candidate, it does not make it easier for such a candidate to win -- he or she must earn majority support. In fact, the country that has used instant runoff voting for the longest period of time is Australia. It has a clear two-party system. Most races are still decided on the first count even after nearly eight decades of elections using the system, yet Australian observers believe it quite important to have the system in order to resolve those situations where a minor party rises to the point of fracturing an electoral majority. Note that some major party leaders here are concerned about any change to the status quo, but we have now seen several examples where instant runoff voting holds up well to their scrutiny.

3. Instant runoff voting is consistent with our electoral norms: Some may be concerned that instant runoff voting will lead to dramatic changes in our politics and to people winning who never could win in our current system. I believe this is quite untrue. Rather, the system refines our current system by eliminating the problem of "spoilers" and encouraging more positive politics. It simulates a runoff election, pure and simple, but without the costs and turnout problems associated with two-round runoffs. Runoff elections are common in the United States and indeed in Vermont's electoral history.

The recent rise in interest in instant runoff voting

Just in the past three years there has been a surge of interest in instant runoff voting in the United States. In New Mexico, for example, the state senate in 1999 passed a constitutional amendment very similar to your H.199. (Note that an analysis by the Brennan Center in New York City indicated that the new Mexico constitution needed to be amended to accommodate IRV because of a provision requiring that the plurality vote-getter must win. Its analysis of Vermont's constitution reveals that there is no constitutional obstacle in Vermont.) The New Mexico bill stalled in the house of representatives, largely over concerns relating to election administration -- unlike Vermont, New Mexico lawmakers want to have results tabulated instantly, which certainly is possible, but can require new machines.

In Alaska, another bill (HB 141) similar to H.199 was introduced in 1999 and received hearings in two committees. Its backers have made it a major priority for 2000 and asked me to speak in Juneau and Anchorage later this month. In addition, a petition to enact the bill's provisions is being circulated in Alaska; backers expect to have the signatures later this spring necessary for a November 2002 vote.

There also is rapidly growing interest in Utah, where a bill to enact instant runoff voting for most state offices is being prepared and where Republicans are exploring its adoption for internal elections. The California Democratic party also is studying its use for internal elections, and the Reform Party plans to use it for its national, vote-by-mail presidential primary this summer. In North Carolina, an electoral laws commission was created last year that will include consideration of instant runoff, particularly as a complement to possible new campaign finance legislation.

We also know of legislation that has been introduced or likely will be introduced in Massachusetts (where a one-party district's congressional primary was won in 1998 with only 23%, sparking much talk of reform), Oregon, Pennsylvania and Texas. As far as we know, none of these states had debated instant runoff voting legislation before 1997.

At a city level, a charter commission in Austin, Texas -- a city of 500,000 people -- has recommended that a stand-alone measure to enact instant runoff voting to replace two-round runoffs; the proposal has a good chance to be on the ballot in Austin this year. Two localities -- Vancouver, Washington and Santa Clara County, California -- in recent years have adopted stand-alone charter amendments to allow instant runoff voting. The measures gained the support of leading political players in both of those localities, including the San Jose Mercury News and Vancouver Columbian. We expect to see other localities to consider ballot measures on instant runoff voting this year.

I believe that the major reason for this surge in consideration of instant runoff voting is simple: that information about the system is reaching more decision-makers. To know about instant runoff voting typically leads to appreciation of its merits. Republicans and Democrats alike are showing support when they recognize that fractures in the electorate need to be healed to elect a majority winner. This recognition seems to be particularly appropriate in the light of current conditions in our politics that trouble thoughtful political observers from across the spectrum: shrinking voter turnout; increasing chances of "spoilers" because of participation by minor party and independent candidates; and excessively negative campaigning that drives good people from participation in politics.

Exercises recently carried out in Vermont -- in which junior high school, senior high school and college students were able to use instant runoff voting in mock elections -- are quite encouraging. These young people overwhelmingly preferred instant runoff voting to current methods of voting and had a near-perfect validity rate in voting with the system. As to the matter of increasingly fractured candidate fields, note that as of early 1998, nearly half of all governors had been elected by less than a majority in one of their gubernatorial races.

The dawn of a new century -- which in this case, is a new millennium -- is helpful for drawing attention to first principles. Limiting voters to a single choice is an unnecessary restriction that puts limitations on our democracy itself. Instant runoff voting's simple change -- allowing voters to express the preferences they already have -- is beginning to capture the imagination of a broad range of elected officials and civic leaders. I look forward to Vermont being among the new wave of states seeking to give their voters the freedom to fully express their choices that I suspect someday will be a nearly universal right in American elections.

Once again, thank you for inviting me to speak with you, and I look forward to answering any questions today and having our Center be a resource to you in any way we can in the future.

Attachments:  Pages 1: A listing of scholars and civic leaders directly linked with our Center