On Thursday, May 4, London directly elected a mayor for the first time. It also elected its city council by a form of proportional representation which played a major role in shaping the composition of the council.

For those interested in reforming elections to executive positions (such as mayor, governor and president), the London elections were very important because the mayor was elected using a modified form of instant runoff voting. While inferior to instant runoff voting as described on our web pages, the London system -- called "the supplementary vote" -- did clarify the winner's mandate and lessened the problem of "wasted votes." The election also resulted in an independent candidate winning. While this candidate -- Ken Livingstone, a popular Labour Party member of parliament who was denied his party's nomination by party insiders -- would likely have won under plurality rules as well, the system made it much easier for him to decide to run as an independent because of the power given to voters to rank a second-choice candidate.

Caleb Kleppner, director of the Center's majority project, provides the following summary of the mayor's race and the media coverage of the election and the election system.

London's Mayoral Election, Results and Media Coverage
by Caleb Kleppner

Independent candidate Ken Livingstone was elected Mayor of London using a system known as the supplementary vote. The supplementary vote is a modified form of instant runoff voting. In the supplementary vote, voters can only rank first and second choice candidates -- even when, as in this election, there are four or more strong candidates. If no candidate receives a majority of the first choice votes, all but the top two candidates are eliminated simultaneously. Ballots that rank eliminated candidates first are then counted for whichever remaining candidate is listed second on each ballot. Ballots which rank another eliminated candidate second are treated as if no second choice was provided.

Livingstone received 667,887 of the first choices (39% of ballots cast). He then received 108,550 second choices from eliminated candidates, giving him a total of 776,427 votes. This total was 45% of the total ballots cast, but a full 58% of the combined votes for the top two candidates making the "runoff."

Second place finisher Steve Norris, the Conservative Party candidate, received 464,434 first choices (27%) and 99,703 second choices, for a total of 564,137. Hid totals represented 33% of total ballots cast and 42% of the total for the top two finishers.

Of the 581,761 first choice votes for eliminated candidates (34% of all first choice votes), only 208,253 second choices counted for one of the two remaining candidates. The remaining second choice votes, 373,508 (22% of the total votes), were wasted because they either listed another eliminated candidate second or were left blank.

Note that as of now, neither the government nor the British or international media reported the distribution of second choices from ballots for eliminated candidates. They did list the total number of second choice votes each candidate received, but these numbers are irrelevant to the outcome of the election and, actually, misleading. (For example, these totals include many of the second-choice preferences for people who ranked Livingstone and Norris first -- and whose ballot would not have counted for the second choice.)

Table 1. Results of Election of the Mayor of London, May 4, 2000

Candidate	Party	1st Choices	%	2nd Choices	Total
Ken Livingstone	Independent	667,877	39%	108,550	776,427
Steve Norris	Conservative	464,434	27%	99,703	564,137
Frank Dobson	Labour	223,884	13%
Susan Kramer	Liberal Democrat	203,452	12%
Ram Gidoomal	Christian Peoples Alliance	42,060	2%
Darren Johnson	Green	38,121	2%
Michael Newland	British National Party	33,569	2%
Damian Hockney	UK Independence Party	16,234	1%
Geoffrey Ben-Nathan	Pro-Motorist Small Shop	9,956	1%
Ashwin Kumar Tanna	Independent	9,015	1%
Geoffrey Clements	Natural Law Party	5,470	0%
Total		1,714,072		208,253	1,340,564

Final Tally

Candidate	Party	Votes	As % of total	As % of final
Ken Linvingston	Independent	776,427	45%	58%
Steve Norris	Conservative	564,137	33%	42%

Important Numbers

1) As mentioned above, the votes of 373,508 people (22% of all votes cast) did not help to choose between Livingstone and Norris. These voters either gave their second choice to an eliminated candidate or left it blank.

2) Some 293,000 voters (17%) did not list a second choice candidate. We do not know how many of those ballots listed Livingstone or Norris first (and hence were not wasted, as their first choice stayed in the running) and how many ballots listed an eliminated candidate first and hence were wasted.
 

Media Coverage of the Mayor's Race

This analysis is based on a reading of several U.K. and U.S. papers, the BBC coverage, and coverage of international politics on American news sources such as CNN, Washington Post and New York Times. It was a surprise to me that the coverage by news media in the U.K. and U.S. as well as the reporting of results from government sources in the U.K. were generally quite poor. This reminds us never to take media understanding of new systems for granted.

With few exceptions, the media did not accurately describe the "supplementary vote" system. The American media typically didn't even mention the use of the system: a reader of the New York Times or Washington Post would have no idea an alternative voting system was used for either the mayoral or city council elections. The media also did a very poor job of reporting the relevant second choice votes. The media was generally unclear about which second choice votes were counted, often implying or stating that all second choice votes are counted (including those on ballots whose first choice made the runoff, which is inaccurate).

The media typically reported the total number of second choice votes cast for the Labor, Liberal Democrat and the minor candidates. These numbers are completely irrelevant to the outcome of the election, although they do show that large numbers of voters gave 2nd choice votes to candidates who had no chance of being in the top two, which shows that voters did not fully understand the system.

Even the total second choice votes for the top two candidates were misleading because they included second choice votes of people who listed the other top candidate as a first choice, so they overstated the number of second choice votes for Livingstone and Norris that actually counted. Of those who ranked Livingstone first, one in eight listed Norris second. Of Norris' first-choice supporters , nearly one in six listed Livingstone second.

The results of an instant runoff or choice voting/STV election typically include the destination of votes for each eliminated or defeated candidate in each round, and I was unable to find these numbers anywhere. The media and the government, as best I could tell, completely failed to describe the destination of the 2nd choices from each eliminated candidate. For a good example of this, see the BBC's website, which actually did describe the system accurately in the small type.