Pay no Attention to the Woman Behind the Curtain

TOWM's premature conclusion of the day comes from the Catalyst Group, which recently released a report claiming that corps with women on the Board do better.

The argument goes something like this. From the years 2001-2004, Catalyst looked at three economic indicators (perhaps tellingly, these indicators did not include stock prices or that insignificant number known as the "bottom line") for Fortune 500 companies - by which they actually mean 520 companies, since companies come and go from the list.

In return on equity, on average, companies with the highest percentages of women board members outperformed those with the least by 53%, the study found. In return on sales, the companies with more women board directors outperformed by 42% on average and in return on invested capital, by 66%, Catalyst said

Conclusion?

The results suggest that "diversity, well managed, produces better results" for companies, said Catalyst president Ilene Lang. "Bringing women on their boards to represent the stakeholders really gives them a better company and better performance."

Uh-huh. Leaving aside the nagging little fact that none of the indicators considered directly benefit the shareholders (remember that stock prices and dividends, the main things shareholders are actually interested in, were not considered), let's see if we can think what else might account for this correlation?

Here are some things that spring to mind:

1. Cherry-picking. We don't get a real rationale for why the years 2001-2004 were chosen, but they seem like an odd choice by any standard. Catalyst claims that it's because they covered an economic downturn, but it's hard to see what that has to do with measuring the effectiveness of women on the board? I mean, if I'm a shareholder, I'm mostly just interested in whether or not my company outperforms the market. That's not a concern I have only when the market isn't so hot! But even if we accept this convoluted "reasoning," I get why we didn't go all the way back to the '81 recession (I'm guessing the glass ceiling was for real back then) - but why not the '91 recession? Nope, sounds to me like we've chosen our years so that they fit the conclusions we want to reach. (Honestly, who's ever heard of a three year survey period for something like this?) On top of that, we're allowed to play fast and loose with what counts as a Fortune 500 company. This could've gone either way, of course. We could've surveyed only the 480 (or however many it was) that were on the list for all three years. So I really have to ask why we erred on the inclusive side rather than the exclusive side? After all, if you're making your point based on the "most powerful companies" meme, the more exclusive standard would seem to be the one you want. Probably what's going on here is that they needed data from companies that fell off the chart to round out their negative claims. Finally, the "return on sales" indicator seems suspicious to me. Is that or is that not codespeak for "profit margin?" Because if it is, it's irrelevant to their claims. A high profit margin is a good thing to find, but that's more a characteristic of the product than of the success of the company peddling it. Volume often beats margin (see Mart, Wal).


2. Inducing the wrong cause from a correlation. I notice there isn't anything provided that reassures me that it's actually the presence of the women on the board that accounts for our (selectively measured) superior performance. In fact, the article later tells us that "Women are still a small minority in the boardroom, holding 14.6% of Fortune 500 board seats in 2006, down slightly from 14.7% in 2005, but up from 9.6% in 2005, according to prior research from Catalyst." Meaning one of two things, as far as I can tell. Since women are a small minority in corporate boardrooms, it seems likely that this group is (a) a truly capable lot or (b) affirmative action window dressing. Probably it is a mixture of both. The point is, at 14% they're likely to be more in demand as a group (for whichever reason - probably depending on the individual case). Companies that can best afford to hire them do. Causation might be working the other way here. A company that is already successful might be able to afford the luxury of hiring more women, rather than it being the women who account for the success. In order to decide this, we would have to see evidence that hiring a woman had improved the company's performance, not simply that they are correlated with it; we need evidence that they are actually the cause and not just a byproduct. This is a well-known problem in statistical reasoning.


3. Hasty conclusion. In any case, I don't see any evidence that "diversity" helps a company's performance. Indeed, in order to reach such an odd conclusion, we would perversely have to abstract away from the individual skills of the women involved and conclude that it was something about their femininity, rather than their individual capabilities, that leads to success. (Not to mention - it's offensive to equate "diversity" with only gender balance, but I digress.) And even if we could come to such a conclusion, the survey hasn't established anything like a monotonic increase in economic performance based on the number of women hired. It might be that three women is ideal, but a majority would be a disaster, or whatever else. In short, the data cited don't support the conclusion that hiring women in and of itself is a good thing. What they more likely suggest (if indeed anything - see items (1) and (2) above) is just that sexism is a penalty. That is, it seems plausible to me that some of the corporations that fell off the Fortune 500 during this survey might have done so because their boardrooms got too focused on their good-ol-boy network and weren't, as a consequence, open to merit-based hiring. On the theory that women and men are equally capable of running a business effectively, then it stands to reason that gender-blind, merit-based hiring will lead to a number of women on the board. It takes anti-female discrimination to keep a board a boys club - which just translates into a willingness to sacrifice profit for purity in some important sense if confronted with a qualified female applicant. In short, even if we buy the evidence Catalyst is citing, the situation might be (indeed, seems more likely to me to be) that the market is punishing sexism rather than rewarding "diversity." I see no evidence for the conclusion that "diversity" should be a goal. It rather just falls out of the Leistungsprinzip - the Meritocracy.

So, pay no attention to the woman behind the curtain. If hiring women is a net benefit in and of itself in reality, there's certainly no iron-clad case for it here. Best strategy: hire the person who seems most likely to benefit your company, regardless of her gender.