Thinking strategically in a multi-candidate election

All right, you say, this British election stuff is mildly interesting, but what does it have to do with the situation in Tulsa? Plenty. Britain's "first-past-the-post" electoral system has a fundamental flaw that works against enacting the will of the majority, and Tulsa's upcoming City Council special election -- no primary, no runoff, no majority required -- has the same flaw, only to a greater degree. The flaw requires voters to do more than simply vote for their favorite candidate, if they want to ensure that the outcome is at least acceptable to a majority of voters. Strategy is required.

First-past-the-post means no majority is required to win a seat. Whoever gets the most votes wins, no matter how small the percentage of the total vote. If 20% of the electorate loves Candidate Smith, and the other 80% hates him, but are split evenly between five or six candidates, the hated Mr. Smith wins anyway. In the 2001 UK general election, the winning candidate received a majority of the vote (greater than 50%) in less than half the constituencies. 333 seats out of 659 had a winning percentage below 50%, 26 seats had a winning percentage below 40%, and in two seats, Argyll and Bute, and Perth, both in Scotland, the winning percentage was just under 30%. (Thanks to the UK Elections Directory for making the 2001 results available in spreadsheet form.)

Winning without a majority is common in the UK because there are three nationally competitive parties -- Labour, Conservative, and Liberal Democrat -- and in Scotland and Wales, there's also a pro-independence party that has a significant base of support, (Scottish National Party and Plaid Cymru, respectively). If voters who supported the winning candidate in the last election became disaffected, there were at least two other places they could take their votes, and an unpopular MP could stay manage to in office if his opponents split the rest of the votes evenly. This is how Tony Blair's Labour Party managed to retain a majority of the seats in the House of Commons, despite taking only 35.2% of the nationwide vote, down from 40.7% in 2001. If the entire 5.5% swing away from Labour had gone to the Conservatives, there would have been moving vans at Number 10, but most of it went to the Liberal Democrats, the third place party. If a voter's main goal had been to get rid of Tony Blair as prime minister, the smart thing would have been to vote for whichever party stood the best chance of beating the Labour candidate in his constituency, even if that meant a Tory voting for a LibDem or vice versa. It's called tactical voting, and anti-Tory activists tried to convince anti-Tory voters to give it a try in 1992 and nearly succeeded -- the Conservatives barely managed a majority of the seats despite winning nearly 41% of the vote.

In Tulsa on Tuesday we have an election which will determine which of two factions -- the Cockroach Caucus and the Reform Alliance will gain overall control of the City Council. The two factions are fundamentally divided over policy and philosophy (although the Cockroach Caucus would like you to think that it's all a matter of personality and temperament). There are four serious candidates in the race -- two, Bill Martinson and Andy Phillips, have the support of elements of the Cockroach Caucus, and two, Charlotte Harer and Al Nichols, are supported by pro-reform activists. There is a real danger that pro-reform voters could form a majority of those who vote on Tuesday but still lose the election by splitting that pro-reform vote between two candidates.

I can illustrate the danger by telling you about the results in the South Belfast constituency. In Northern Ireland, there are two main political sympathies but four main parties -- two nationalist parties that want all Ireland united in the Republic of Ireland (Social Democratic and Labour Party, Sinn Fein), two unionist parties that want Northern Ireland to remain part of the United Kingdom (Democratic Unionist Party, Ulster Unionist Party). To oversimplify, within each grouping there are two parties that share a central aim but differ on how best to achieve that aim. All four parties competed for the South Belfast seat, which had been held by an Ulster Unionist.

In Thursday's vote, unionist parties received 51.1% of the vote, while nationalist parties received 41.3% -- the rest of the votes went to three minor parties which are neither unionist nor nationalist. Even though a majority of voters supported unionists, the winning candidate was a nationalist. Most of the nationalist votes went to the SDLP candidate, who took 32.3% of the vote, while the DUP and UUP candidates split the unionist vote almost down the middle -- 28.4% and 22.7% respectively. If there were a runoff, the DUP candidate would almost certainly have won, but there isn't going to be a runoff -- just a "winner" who had two-thirds of the voters against him.

In such a situation, the only way to ensure that the winner is someone acceptable to the majority is for each voter to conduct the runoff in his head before he goes to vote. A voter should ask himself, "Of the candidates acceptable to me, which one has the most support already? Which one has the best chance of winning?" That question is easier to answer if polls have been conducted, but you can still make an educated assessment by looking at the segments of the electorate that are likely to back each candidate.

The South Belfast scenario may well play itself out in Tulsa's District 5, but the stakes are higher here. If the reform candidates split a majority of the vote and lose the election, it will be cold comfort to say that we would have won if there had been a runoff. We'll have to suffer for the next 11 months with the Cockroach Caucus back in control of the Council.

The situation here is complicated by the fact that each of the factions has a serious candidate from each political party. That's part of the data that has to be considered in casting a strategic vote. You may not care what party label a Council candidate wears, but some voters will, and that affects which candidates have the best chance of winning. Charlotte Harer's long list of endorsements by elected Republicans like Senator Jim Inhofe and Congressman John Sullivan will carry a lot of weight with those voters who care most of all about electing a good conservative Republican.

Al Nichols, a Democrat and a reformer, is a good man and would be a good councilor. His credentials as a neighborhood activist are impeccable, and he has the best understanding of land use issues of any of the candidates in the race. I don't believe that neighborhood activists and voters concerned about land use constitute a sufficient base of support to win the election. Nichols would have to win a certain number of voters who are simply looking for a Democrat to vote for. Unless they disagree with him on a specific issue, most Democrat voters will vote for Andy Phillips -- they saw his name on the ballot and voted for him just over a year ago.

I believe that Charlotte Harer is in the best position to put together a winning coalition. Her support from prominent Republicans gives her a head start, and the district does lean Republican. Now that she's the only woman in the race, she'll get all the votes of those who simply want more women on the Council. She's solidly pro-reform at City Hall and that should win her some crossover votes from pro-reform Democrats.

I'm going to ask some people I greatly respect and admire to do a hard thing. We share the same goals for city government. I wish it weren't necessary to make this request, and if we had a better voting system, like Instant Runoff Voting, it wouldn't be necessary. Under IRV, you could give your first preference to your favorite without worrying that you might help elect your least favorite candidate. But we don't have IRV, or even a simple runoff. So here goes:

I urge Al Nichols and his supporters to throw their support behind Charlotte Harer for the sake of keeping the Council in the hands of the reformers. You've fought a good fight, and as a former candidate I have some idea of how hard it would be to drop out at the last minute after knocking on hundreds of doors and making hundreds of phone calls. For the sake of the ultimate aim -- reforming city government -- I believe such a sacrifice is necessary. The only thing that would change my mind is a scientific poll or comprehensive survey showing Al in first or second place.