The study of decision rules is hard-core political science but sometimes essential to understand real-world politics. The Alberta Progressive Conservative leadership race, which will soon choose Ed Stelmach's successor, is one of those cases.
Conservative members vote for their leader using a runoff system, with the second ballot containing the top three finishers from the first ballot. The winner is then chosen by alternative ballot, which allows voters to rank the three finalists in order of preference. Having three finalists was originally meant to placate rural members, who were afraid that Calgary and Edmonton candidates would always dominate; but it also allows a finalist to emerge from each of the party's three main factions: Red Tories, Blue Tories and non-ideological party loyalists.
The third slot allows a candidate with little support to get to the second ballot as long as he can finish third on the first ballot. He then has a chance to win, as Mr. Stelmach did in 2006, if he can get support from candidates who have now been dropped from the ballot. Trailing far behind Red Tory Jim Dinning and Blue Tory Ted Morton, party loyalist Mr. Stelmach placed third on the first ballot. By offering cabinet jobs to those who had been knocked out of the race, he gathered additional support while Mr. Dinning and Mr. Morton went full-bore negative against each other.
Equally important, Mr. Stelmach stayed outside the polarization between the two front-runners, thus becoming the preferred second choice for both camps. Second choices ultimately pushed him over the top, since no one won an absolute majority of first preferences on the second ballot.
Another important feature of the system is the ease of obtaining memberships, which can be purchased for $5 right up to the close of voting at 7 p.m. Contrary to practice elsewhere, one can buy a membership in the two weeks between the first and second ballots. Truly, in Alberta's Conservative world, "it ain't over till it's over." Many voters pay little attention until after the first ballot. They wait to see who the finalists are, then decide whom to support. In 2006, about 50,000 more people voted on the second ballot than on the first.
The race becomes much more public in the final two weeks between the first and second ballots. Realizing that one can't win just by persuading existing members, campaigns resort to TV and radio advertising to arouse public interest and sell new memberships up to the very end, a tactic that would be pointless in leadership races in other jurisdictions.
In effect, an Alberta Conservative leadership race resembles an open primary in U.S. politics. Winning requires attracting support outside the usual party membership, with the ironic outcome that victory is often decided by outsiders. Colleague David Stewart has shown that, in 2006, Mr. Dinning targeted Liberals, Mr. Morton appealed to Wildrose Alliance supporters, and Mr. Stelmach recruited Ukrainians. These tactics are being repeated in 2011, as Gary Mar reaches out to Liberals, Mr. Morton appeals to Wildrose supporters, and Doug Horner tries to keep Mr. Stelmach's coalition of party loyalists, Edmontonians and other northerners intact.
If the vote on the first ballot is very low on Sept. 17, one candidate could win an outright majority and the game will be over. More likely is that there will be a second ballot with representatives of all three factions, and the winner will be the one who can best attract support from the losers and second preferences from his opponents while frantically selling new memberships in the final two weeks.
Don't bet on whoever is allegedly the front-runner now. History shows that's a minor factor compared to the ability to understand and exploit the complex rules of the game.
Tom Flanagan is a professor of political science at the University of Calgary and a former Conservative campaign manager.