Tuesday, September 07, 2004

The Glass Cliff?

British psychologists Alex Haslam and Michelle Ryan are the researchers in a study which found that women who break through the proverbial glass ceiling may find themselves teetering at the edge of a glass cliff. (All this glass is getting on my nerves; I fear that they might shatter. And yes, I know that I have used similar metaphors before, but I did it much better.)

The 'cliff' refers to their finding that firms, political parties and so on tend to nominate women to important positions when times are bad, when the job is seen as impossible to perform. According to Haslam, this is not only true for the first appointments of new, eager managers or politicians but a recurring problem for women and, he speculates, for racial and ethnic minorities as well.

The Haslam-Ryan study can be summarized as follows:

Prof Haslam and a colleague, Dr Michelle Ryan, compared the stock market performance of FTSE 100 companies with their appointment history.
The 19 that had appointed women to the board in the previous year had done worse than those whose appointments had been all male.
"But in all cases women had only been appointed after the company performance had slumped," Prof Haslam told the British Association science festival in Exeter yesterday.
"If everything had gone swimmingly, then they carried on with the old jobs for the boys."
A similar pattern emerged in a study of candidates for the Scottish parliamentary elections. Women were more likely to be nominated as candidates for harder to win seats.
The glass cliff also emerged in experiments in which 300 students were asked to pick a candidate for a fictional senior management post.
Given a man and woman with identical qualifications, students were far more likely to pick a female candidate than a male one if the company was doing badly.
The reason for the discrimination is unclear. Prof Haslam believes it could be explained by overt sexism - that men were handpicked for the good jobs, leaving women to take posts in failing companies.
More subtle forms of discrimination could be in play, he said. The predominantly male managers of companies were likely to recommend desirable jobs to their predominantly male friends, but give "poisoned chalice" jobs to those they did not know.
Another explanation was that women were perceived as being better at crisis management, he added.

See how we can't get rid of the glass? Now it's a poisoned chalice.
But the results are interesting and even suggestive. I have often wondered if Condie Rice wasn't one of those appointments, given that her background made her a pretty astonishing choice for advising about current security concerns. Her selection could have been seen as a win-win solution for the Republicans: if she fails it's because she's a woman and a member of a racial minority, if she succeeds, well, then the Republicans can take the credit for being smart in their choice and gender- and color-blind.

What is serious about all these cliffs and ceilings is their real impact on fairness and on productivity, of course. Note the way the correlation with poor performance of stocks and the appointment of women into positions of power can be distorted:

It was reported in 2003 that UK business had gone downhill in the previous year as the number of female directors on the boards of FTSE 100 companies had risen by 20%. A newspaper commentator said: "The triumphant march of women into the country's boardrooms has wreaked havoc on companies' performance and share prices."
Professor Haslam said: "What we found was that in all of those cases, women had only been appointed after company performance had slumped quite dramatically."

The problem here was that old tiresome assumption that correlation implies causality or that the order of the causality is obvious from the correlation. If the bad performance preceded the appointment of women as Haslam stated, then to argue that the 'triumphant march' (what framing!) of women has 'wreaked havoc' (what framing!) is lying. In fact,

Haslam also said: "The appointment of a woman director was not associated with a subsequent drop in company performance. Indeed companies that appointed a woman actually experienced a marked increase in share price after the appointment."

I think that the real glass cliff is in the tongues of some journalists and public commentators who never learn to appreciate basic statistics as well as the glib sounds that so easily slide out of their mouths.