Imagine if a report came out showing evidence that Wells Fargo violated the Fair Housing Act by hiding certain home listings from African-Americans. Every politician in Washington would condemn the bank for illegal practices. The Justice Department would be inundated with letters demanding prosecution. Congressional committee chairs would schedule hearings to give members an opportunity to yell at executives. Wells Fargo would put out a sober apology expressing deep sorrow and vowing to make everything right. In other words, we have a context for bank misconduct, and everyone dutifully plays their part.
When the same circumstance occurs with Facebook in the role of the villain, however, nobody knows how to react. There are no assigned roles when a tech firm with a glimmering reputation creates a controversy. We implicitly give them a break, regardless of the merits. That’s a bias we should probably correct.
On Friday, ProPublica revealed that Facebook allows advertisers a tool that enables them to exclude “ethnic affinities” like African-Americans or Hispanics from viewing their ads. (Facebook does not ask users about their race, but collects data based on posts they like or comment on.) This goes well beyond targeting different styles of advertising to certain groups, which is common. Instead, it specifically prevents a black or Hispanic Facebook user from seeing a particular ad.
This shifts from market segmenting to violating the law when the product is an ad for housing or employment recruiting, especially because Facebook carries such granular data on its users. The problem is the exclusion and how precise Facebook’s big data allows that to work. Federal law prevents discrimination in such matters by race or gender. ProPublica tested this with a sample ad inviting people on Facebook to a forum at the Brooklyn Public Library about challenging illegal rent increases. African-Americans, Asian-Americans and Hispanics weren’t allowed to see it.
If the ad were selling a house or recruiting for a job fair, it would almost certainly be illegal. “This is about as blatant a violation of the federal Fair Housing Act as one can find,” civil rights lawyer John Relman told ProPublica.
Facebook, of course, disputes the allegation. It says the tools are meant to test marketing campaigns with specific groups for their effectiveness, not to discriminate. In a buzzword-laden post on its website, Facebook head of multicultural sales Christian Martinez writes, “we want Facebook to be a platform that’s respectful and empowering,” and that “we will take aggressive enforcement action” against targeting that violates anti-discrimination laws.
But ProPublica’s housing-based ad got approved for use on Facebook within 15 minutes, suggesting that there isn’t any close scrutiny of exclusion targeting, or understanding of its negative consequences. At the least, Facebook hasn’t thought critically about these issues.
While the ProPublica story has generated some chatter, no politician has demanded that Facebook immediately take down its exclusion targeting tool. Nobody has called for hearings. Nobody has sought clarification on whether the Justice Department would open an investigation. Nobody has written a negative word about Facebook at all.
Banks have a pre-existing reputation that would turn a story like this into a serious negative pile-on. Large tech firms don’t have that baked into their reputation. It’s reasonable to ask why, considering that we have a smattering of examples of Silicon Valley’s indifference to federal law, from the industry-wide wage-fixing scandal to Amazon’s sweatshops.
Because tech is new and shiny, it gets the benefit of the doubt when negative stories get published about industry practices in a way that Wall Street simply doesn’t. And it’s not just Facebook ads. Uber has over the years claimed to solve a market failure of African-Americans having trouble hailing taxicabs. But a massive study from the National Bureau for Economic Research showed that riders with African American-sounding names experienced higher cancellation rates and longer wait times when using the Uber app. Airbnb has struggled with the same biases.
Outside of a couple members of the Congressional Black Caucus contacting Airbnb, none of this has raised much of a peep in Washington. In fact, former Attorney General Eric Holder rode to Airbnb’s rescue to help the company arrange an effective anti-discrimination policy. This kind of assistance isn’t present when Bank of America or some other bank disadvantages customers by race.
Federal anti-discrimination laws are just one area where the tech industry has found itself in trouble. ProPublica has another series on the tech industry’s behind-the-scenes practices, finding that Amazon favors its own products with its “buy box,” which often charges consumers more for the same product. That skirts antitrust laws, which is something of a trend for tech. Google is fighting multiple antitrust charges in Europe, where Apple has separately been accused of illegally avoiding $14.6 billion in unpaid taxes. Facebook also has a tax issue over assets the company moved to Ireland to take advantage of lower rates.
If we thought differently about Silicon Valley, we could build a profile of tech’s biggest businesses as corporate miscreants, constantly seeking advantage by avoiding or outright ignoring the law. But they don’t have that reputation, and so these stories remain disassociated rather than cumulative.
We hear a lot about the bias of journalists who report on stories, but not much about the bias of those who read them. Reputational history creates blind spots in how readers process the news. It’s pretty obvious that Facebook values the profit it can gain from advertisers by giving them a tool to segregate their ads well beyond the potential for illegal abuse. But we don’t read that as part of a consistent set of value judgments by Silicon Valley firms. Maybe we should.