Justin Sullivan/Getty Images
People mill outside Google’s headquarters in Mountain View, California, on Sept. 2, 2015.
Thanks to a ruling on Friday, Google will finally have to cooperate with a U.S. Department of Labor investigation into whether its compensation practices are sexist—but its compliance won’t be nearly as onerous as it could have been. The company pulled that off by making an unusual, and potentially dangerous, argument: that if Google gave the government all the data it wanted for its probe, the firm would be exposing those employees to the risk of the feds being hacked.
If labor regulators limit their investigative activities because of privacy worries, they literally can’t do their job.
In Friday’s ruling, an administrative judge ordered the internet giant to fork over compensation data of up to 8,000 of its employees so the feds can look into whether Google is discriminating against women who work at the company. The agency originally requested data for more than 21,000 Google employees, including their contact information and salary history, for its investigation. But Google refused to comply with that order, arguing that the data request was overbroad and would violate its employees’ privacy.
The company also complained that compliance would cost the company $100,000, which it claimed would be too expensive—a weak argument considering Google’s parent company, Alphabet, generated $90 billion in income last year.
The judge agreed with Google on the privacy front, reducing the Department of Labor’s request for data to about a third of the original request, setting what could be slippery precedent for other companies under investigation for unfair labor practices to resist handing over data under the pretense that they don’t want to risk threatening employee privacy.
The decision points out that even the U.S. government is susceptible to hacking, with the judge noting that his own office has been the victim of a data breach. The logic is that if the government is holding less data, then a potential hack would affect fewer people.
It’s true that in this case the Department of Labor is asking for more information, like email addresses, than it typically would in this kind of investigation, says Michael Selmi, a law professor at George Washington University who specializes in discrimination cases. Even so, “you can’t really question the government’s ability to keep information confidential,” he said. If labor regulators limit their investigative activities because of that worry, they literally can’t do their job.
Reducing the number of records Google has to produce on the basis of privacy, especially if a lot of the data can be shared without providing contact information, could be letting the company off easy. The Department of Labor, after all, is way more likely to find discriminatory trends in 21,000 employee records than 8,000.
While Google says it’s pleased with the decision to narrow the scope of the Department of Labor’s request, the newly provided pay data and an outside audit could still unearth previously unreported inequities at the company, despite the fact that Google boasts to have already fixed its pay gap. In 2016, the company says it conducted its own analysis of 52 different jobs at Google and “found no gender pay gap.”
If Google has analyzed its salary data, as the company claims to have done, then turning over the data to the Department of Labor shouldn’t be that difficult. Since Google has contracts with the U.S. government, the company is required to adhere to equal opportunity employment laws and has to let the Department of Labor inspect its records. (The lawsuit to compel Google to share its employment data was originally sparked during a regular compliance audit of its federal contracts last year, according to the Department of Labor.)
“I’m always suspect when companies collect their own data and analyze it themselves,” said Kellie McElhaney, the director of the Center for Gender Equity and Leadership at the University of California, Berkeley. Even Salesforce, which has spent millions to compensate female employees shorted by sexist pay inequities, hasn’t fixed the problem yet, McElhaney explained.
One reason fixing gender pay gaps can be so difficult for a company, according to McElhaney, is that focusing on compensation alone doesn’t resolve the underlying reason women make less than men in the first place, which is bias against female employees.
Google has voluntarily released its diversity data when it comes to employment trends, having first shared its numbers in 2014. Those numbers don’t paint a rosy picture for Google, which still today only counts 31 percent female employees at the company—just a point better than the number of female employees it counted in 2014.
Sharing data alone doesn’t seem to have spurred more equitable hiring practices. If the company is found to have systemically paid women less than men, Google could lose government contracts. Such a high-profile case may inspire other tech companies to start to fix their pay inequities, too.
Fixing discriminatory salary gaps would certainly be a huge step in the right direction within Google. But as long as it and other firms continue to staff leadership positions with an outsize representation of white men, the insidious, underlying reasons behind unfair compensation and hiring practices at these companies—that those in power exercise prejudice against people who don’t look like them—may not change anytime soon.
While it’s good to protect the privacy of employees, leveraging that concern to blunt a federal investigation might not be to Google employees’ benefit, either. It’ll be even more worrisome when a company with a less evident commitment to equal pay uses the same argument.