This entry looks at Toronto’s city council elections since 2000 – the first year the current 44 ward system was used. Previously, Toronto had used a 28 ward system with 2 members per ward.
Part 1 of this article established that there is a bias against electing female candidates in Australia and Canada’s national elections. The extent that this bias shows up depends on voter’s opportunities to defect to a viable male candidate from another party.
This voter bias doesn’t seem to express itself strongly in left wing parties, where there is no consistent evidence of gender bias among the voters. However, both the Canadian Conservative and Liberal parties appear to be losing seats when they run female candidates. Voters seem willing to defect to the other party rather than vote for a female candidate.
This makes female candidates 10% – 30% less electable than their male counterparts when running for centre or right wing parties.
However, the phenomenon barely shows up in Australia’s 2-party elections. Presumably the inability to defect to a viable male candidate plays a significant role.
So how does this impact Toronto’s city council elections where there are no formal political parties?
One thing that we would expect to see is a less clear distinction between left and right on the political spectrum. However this doesn’t always hold true. Some candidates are strongly identified with one side or the other, or have well known roots in a political party at some other level of government. Some candidates have even openly advertised themselves as part of a party caucus.
What it do however is make the same kind of party electability analysis I did for Australia and Canada very difficult. Since I don’t know the political leanings of the vast majority of candidates, I’m not even going to try.
Instead, I wanted to look at the overall electability of female candidates. It’s actually quite good compared with our federal government.
In a typical election, about a third of the councillors elected are women. This is almost as good as the Australian senate elections. What’s surprising however is that the share of women coming coming in second or third is much lower – only about one in five.
The one exception to this is the 2006 election when a number of women lost close races to men.
So would AV have made a difference in electing more women? It’s difficult to say for sure. My analysis looked at how large a share of the remaining vote the second place finisher would need to overtake the leader. In many races it is less than 60%, indicating there is an excellent chance the results could change.
However, in all the elections except the 2006, more women would lose seats than gain them if the results did change under a different voting system.
The same results hold for assuming the second place finisher could grab 75% of the vote such as could result from an “anyone but” campaign. The point is the ability to change the results doesn’t seem to depend on the exact assumptions used in the model.
Anti-incumbent feelings also seem to hurt women. In the 2003 election, two female incumbents were unseated while all the male councillors who ran for reelection won. In the 2006 election, only a single male incumbent lost his seat. In the 2010 election, it was three men and two women who lost.
While this adds up to four councillors of each gender, there are more than twice as many male councillors as female. This suggests that being a female councillor makes you more likely to be on the receiving end of “boot the bums out” hysteria.
Strangely however, in the elections, female candidates were usually elected when an incumbent was unseated regardless of the gender of the incumbent.
In 2003, two female incumbents were replaced with one male and one female. In 2006, the sole male incumbent removed was replaced with another male. However in 2010 all three male incumbents unseated were replaced by women, while the two female incumbents were replaced by one man and one women.
Overall, the score was five women coming in compared with four leaving while only three men replaced the four leaving.
So how do women get elected to council? Surprisingly, they seem to get slightly higher percentages of support than their male counterparts. The average male councillor won their first election with around 44%. The average female councillor won election with 47% their first time out.
So is it possible that gender bias is absent from municipal elections? Probably not! What may be happening is that female councillors are getting a boost from left-wing voters who are concerned about gender balance on council. This happens even if the councillors aren’t particularly progressive.
An example of this may be the election and eventual defeat of Sherene Shaw. She was hugely popular in the 2000 election, winning over 75% of the vote. In 2003 however, she faced off against a right wing challenger, losing by a mere 400 votes in a two-person race. One established, the new incumbent has retained his office winning two out of every three votes each election. He hasn’t had to face a challenge from a female candidate since the 2003 race.
In the same election, Anne Johnston lost badly to Karen Stintz despite winning convincingly over a male candidate the previous time out. Again, this suggests that there is some effort on the part of progressive voters to gender balance.
At some level this breaks down however. When a progressive male candidate with solid credentials runs, enough left leaning voters may support him to allow him to defeat a female rival. This will also happen when the female candidate doesn’t have sufficient standing in the ward.
We also see examples of excessively right-wing women going down to defeat at the hands of more progressive rivals, such as Mary Fagedakis’s win over former mayoralty candidate Jane Pitfield in 2010. Similarly, Paula Fletcher held off an anti-incumbent surge against Liz West.
So what about voters defecting? Apparently the lack of parties makes this less of an issue. Voter’s either just aren’t as sure about the political leanings of a candidate or there is no leader to attract or repel voters.
In a nutshell then, switching to AV is very unlikely to change the gender balance on council. If anything, in most elections it will result in fewer women being elected.
In part 3 of this series, I will look at how visible minorities fare in Toronto’s elections.
Gary Dale's Blog 