In British Columbia, the premier is a woman. In Alberta, the premier is a woman. Both women inherited successful political dynasties. Both are likely to lead their parties to defeat.

Call it the "Kim Campbell Phenomenon." Successful political parties that make women leaders have an unfortunate tendency to lose.

And it's not only a political thing. In 2003, The Times of London examined British corporate boards and discovered that, in the previous year, the number of women on corporate boards had risen considerably - and shares of corporations that had added women to their boards had fared worse than those which had remained all-male bastions.

"So much for smashing the glass ceiling and using their unique skills to enhance the performance of Britain's biggest companies," wrote one pundit in response to the Times' exposé. "The triumphant march of women into the country's boardrooms has instead wreaked havoc on companies' performance."

Clearly, women should not lead. Seems obvious, doesn't it? Sadly - for the Archie Bunkers among us - that conclusion is wrong.

Psychologists Michelle Ryan and Alex Haslam had a close look at the Times' exposé and discovered something curious: Corporations that appointed women to their boards "experienced consistently poorer performance in the five months preceding the appointment." So the corporations didn't struggle because they put women in leadership roles. They put women in leadership roles because they were struggling.

Ryan and Haslam called this the "glass cliff": Only when an organization's situation is precarious are women given leadership, which ensures that women in charge often land with a thud.

Researchers have documented the "glass cliff" in a wide variety of settings. In an examination of Britain's 2005 general election, Ryan and Haslam ranked ridings by how secure they were for the Conservatives. Then they looked at the gender of the Conservative candidates who contested the election. As the likelihood of winning the riding declined, they found, the likelihood that the candidate was a woman increased.

One reason for the "glass cliff" is plain old status quo bias. When an organization is doing well, there's no reason to consider change. And, since putting women in leadership positions often represents change, they're often left out. Until things start to go wrong.

Experimental studies have also shown there are underlying stereotypes at work.

Researchers Susanne Bruckmüller and Nyla Branscombe asked 122 university students to read an article about a supermarket chain. In one version, the supermarket was doing well. In another, it was failing. The students were told a new CEO was being hired and they were asked to read profiles of two candidates, a man and a woman, and to rate them in 10 categories. Some were stereotypically male (competitiveness, decisiveness). Others were stereotypically female (communication, ability to encourage others). Then they were asked to choose the CEO. Roughly two-thirds chose the man to head the successful supermarket. Two-thirds chose the woman to be CEO of the failing operation.

It may be that people see stereotypically female qualities as those that are necessary to turn an organization around, which isn't entirely a bad thing, but other research suggests the explanation may be more dismal. "Women may be favoured in times of poor performance," Ryan and Haslam wrote, "not because they are expected to improve the situation, but because they are seen to be good people managers and can take the blame for organizational failure."

The "glass cliff" theory fits Canadian political experience with almost eerie precision.

There was Kim Campbell, of course. And Alexa McDonough.

Christy Clark became premier of B.C. only after Gordon Campbell's handling of the HST drove the Liberal party's popularity into the ground, forcing him to resign.

Alison Redford became premier of Alberta and inherited the Progressive Conservative dynasty only after Ed Stelmach failed to stem the rise of the Wildrose challenge and internal dissent.

Redford is likely to be beaten by a woman, Danielle Smith, leader of the Wildrose party, but Smith wasn't given the leadership of a strong organization. She made it a strong organization.

Pauline Marois, the Parti Québécois leader whose fortunes are once again on the rise, is another apparent exception that isn't. Remember that Marois lost two leadership bids. Only when the PQ suffered a devastating defeat, fell into third place and looked hopeless was she given command.

Similarly, Lyn McLeod became leader of the Ontario Liberals after the Peterson wipeout.

Andrea Horwath, leader of the Ontario NDP, is arguably an exception to the trend. But the only indisputable exception I can think of is Audrey McLaughlin, who took charge of the federal NDP when it was riding high.

Of course leadership means responsibility and a pilot who flies a plane into the ground may be responsible. But before laying blame, we have to ask: Did the plane nosedive because of the woman at the controls, or was the woman given the controls because the plane was nosediving? If the answer is the latter, who is really to blame? Her or those whose irrational thinking bars women from leadership except when they are likely to fail?