Why it’s hard for women to prove systematic discrimination in court

When Ellen Pao launched her high-profile lawsuit against the venture capital firm of Kleiner Perkins Caufield & Byers, she became a hero to feminists who believe Silicon Valley has a pervasive problem with workplace sexism. But while Pao provided some vivid examples of sexist attitudes by some of her male Kleiner Perkins coworkers, a jury found that she hadn’t proved that these attitudes were the reason she wasn’t promoted as quickly as some of her male colleagues.

While Pao’s case might be over, the broader debate about gender discrimination in Silicon Valley will continue. Women at Facebook and Twitter have alleged that they missed out on opportunities because of gender discrimination.

Pao’s loss in court showed that women can face an uphill battle in proving their claims in court. Many legal experts agree that proving to a jury that an atmosphere of sexism affected particular hiring and firing decisions is difficult. Meanwhile, there are fears that these sorts of cases will lead companies, fearing lawsuits, to hire fewer women.

But even when companies “win” these lawsuits, they still lose in the court of public opinion. A jury might have ruled that Pao didn’t prove her case, but the lawsuit still persuaded a lot of people that Kleiner Perkins has a dysfunctional company culture, which could harm its reputation for years to come.

So smart managers should be thinking about ways to change their company culture so women never become frustrated enough to file these lawsuits in the first place. One way to do this is with data. By collecting detailed data on their hiring and promotion practices and analyzing it carefully, these companies can better understand where they’re falling short on gender equality.

Sometimes proving discrimination is easy. Pao’s case is often compared to the 1982 case of Price Waterhouse v. Hopkins. Ann Hopkins, a Price Waterhouse employee, was denied partnership despite stellar performance. When she asked for feedback, her employer gave her memorable advice: “Walk more femininely, talk more femininely, dress more femininely, wear makeup and jewelry, have my hair styled,” as she later described it.

The firm admitted to giving her this advice, and with that evidence on her side, Hopkins later won the case.

That kind of smoking gun, as it were, is missing from many gender discrimination cases. One of the most common takeaways people have drawn from the Pao case is that subtle sexism is too hard to prove in courtrooms. That’s true, but several of the incidents in the Pao case were anything but subtle.

One of Pao’s coworkers, Trae Vassallo, testified that a male partner came to her hotel room on a business trip in a bathrobe and with a glass of wine one evening. Pao said that her mentor, John Doerr, said she had a “woman chip on her shoulder.” Women at the firm weren’t invited on a ski trip that all of the men went on, nor were they invited to a company dinner because, as one man commented, “women kill the buzz.“

But Pao wasn’t able to prove that those instances directly and substantially connected to her lack of promotion.

“Just the fact that defendants engage in some bias doesn’t establish that it was the cause of plaintiff’s lack of advancement,” says Deborah Rhode, director of the Center on the Legal Profession at Stanford’s law school. “It’s not at the level of intent. No one at Kleiner Perkins said, ‘We don’t want women at the partnership level.’”

That said, there were also incidents in Pao’s case where reasonable people could disagree about whether sexism was involved, as New York Magazine’s Annie Lowrey has written. Women were asked to take notes at meetings, and Pao was criticized for having “sharp elbows,” which may have been an inadvertent commentary on the distastefulness of a woman having an aggressive personality.

It’s not that the law won’t allow juries to award damages because of that subtler sexism, explains one expert.

“If a plaintiff can show that she lost out on a promotion because of implicit bias, she is entitled to prevail under the formal law. The problem is that it’s hard to prove, and judges and jurors are typically looking for more overt proof,” says Samuel Bagenstos, a professor at the University of Michigan law school.

Not only is sexism hard to prove in court, it’s also hard to recognize and stop in the workplace. Especially in a male-dominated workplace, it can be easy to perpetuate subtler forms of sexism without even trying. Asking women to take notes may seem innocuous to some men, but it perpetuates ideas of women as filling peripheral, secretarial roles. Interrupting women is another example of this sort of inadvertent sexism, as Sheryl Sandberg has written.

None of this is to say that Pao should have won — Kleiner Perkins made a strong case that Pao was denied a promotion because of performance reviews, and jurors have cited those reviews as playing a key role in their decisions.

But there is a case to be made that low-grade, chronic sexism can worm its way into performance reviews, and that it may have in the Pao case.

“There was a fair amount of evidence to suggest that her personality was not a good fit for the firm. But I think you have to ask yourself what about the firm’s culture made it a poor fit? And was she judged by a different standard from her male colleagues? And that’s where I think her evidence was strong,” says Rhode. “John Doerr famously said that she had a female chip on her shoulder, but I think it was the firm’s responsibility to ask how it got there. And what about the culture was unwelcoming to women?”

As the Pao case shows, proving that sort of thing to judges and juries can be difficult. But more important, if a firm — or an entire industry — is truly committed to eliminating sexism, it doesn’t want to go before a jury or pay out multimillion-dollar settlements. Undertaking the kind of introspection that Rhode describes will likely be the key. And there’s one easy way to spot a lot of it: statistics.

“One thing you could do, which I am fairly in favor of, is look at, be really aggressive in looking at numbers,” says Bagenstos. “You see if there are bottlenecks in the process where the demographics get skewed in one way or another, and those are the places where you say, ‘Aha. It looks like something funky is going on here.’”

Bagenstos adds that judges and legislators tend to have an instinctive “quota-phobia,” refusing to put affirmative-action policies into practice. Even absent hard quotas, however, companies should be looking hard at a wide range of metrics to see where they can do better (and if not, why not).

Some tech companies are already partway there, reporting their gender breakdowns publicly. Google acknowledges that it’s a heavily male, mostly white company: “We’re not where we want to be when it comes to diversity.”

But it’s not just about hires — it’s about pay grades and promotions. Companies should be looking at their numbers to determine whether the women they hire on are rising through the ranks as quickly as the men. And if women are falling behind on a companywide basis, it may be time to ask why that’s happening. If enough women’s ideas aren’t being put into practice, it could be that those women are being silenced in meetings. Much sexism may be subjective, but numbers are an easy and concrete way to start rooting it out.

Google in fact is a model case of this — when the company found that women weren’t promoted to better jobs as often as men, they discovered that the women weren’t good at self-promotion. When the company started encouraging women to nominate themselves for promotion, the women started moving up the ladder.

To be clear, focusing on the data at Kleiner Perkins may not have gotten Ellen Pao a promotion. And data is only a first step at trying to eliminate sexism. But what data can do is prevent bad patterns from developing. And if enough women are around (particularly in positions of power), it becomes impossible to go on a ski trip without bringing them along.