Michael Strain

Facebook Fixes Won't Be Easy

Many agree that Facebook needs to make some changes. But those changes are a lot more complicated than the public conversation suggests.

Let’s discuss three of what would seem to be the simplest moves Facebook could make: tighter standards for political speech, user verification, and allowing people to take their social network with them when they stop using Facebook.

Tighter controls on political content seem relatively simple to implement. In fact, a few days before Facebook CEO Mark Zuckerberg came to Washington for congressional testimony, he announced a major policy change to this effect: “Every advertiser who wants to run political or issue ads will need to be verified,” including confirming their identity and location. Facebook “will also label them and advertisers will have to show you who paid for them,” Mr. Zuckerberg wrote on his Facebook page.

This is designed to deter foreign interference in US elections. The company may also hope that disclosures similar to those on broadcast political ads will dissuade officials from regulating Facebook more like a media outlet.

The first difficulty here is identifying what counts as a “political or issue ad.” If a foreign public health organization buys Facebook ads to raise awareness about, say, malaria vaccinations, should Facebook label them as engaging in political activity? If users in the US see the ad while foreign aid is an election issue, does that count as foreign interference? If a US university buys ads to promote the research of its faculty, and that research is on politically salient issues like the minimum wage, should the university be labeled as politically active? What if a religious organization does the same on social issues?

Tighter controls on political content may indeed be a necessary change for Facebook, but it is certainly not straightforward. Neither is user verification.

Facebook has been rightly criticized for the proliferation of fake users -- bots -- on its platform. As a private company, it should be able to allow bots, of course -- but many Facebook users want to interact with human beings, not with complex software pretending to be people.

This problem seems easy enough to solve. After all, Facebook used to require that users be Harvard students, and then allowed accounts for students at select colleges. It eventually allowed anyone with a dot-edu, dot-com, dot-org, dot-gov or dot-mil email address. And in 2006, it opened its doors to any person with a valid email address.

Perhaps the doors have been opened too wide. Facebook could require that potential users submit a driver’s license or government ID as a requirement of obtaining an account, and design a computer program to quickly verify that the license is valid. Bots don’t have ID. Problem solved, right?

This would cut down on bots, but there would be collateral damage. Teenagers and adults without driver’s licenses would be unable to open Facebook accounts. The same types of problems -- individuals who should be able to open an account but can’t because of verification requirements -- would surface for any form of ID, including passports, birth certificates and Social Security cards. (The similarities to long-standing concerns with voter ID laws are obvious.)

And this says nothing about the difficulties of verifying users in developing countries without the widespread use of government-issued identification. In addition, do we really want Facebook holding identification documents for all users? This would only increase the potential damage from future data breaches.

Many are concerned that Facebook is dominant to the point that other social network platforms can’t compete. One proposal to empower consumers and increase competition is allowing Facebook users to take their social network with them -- for instance letting you download the names and email addresses of your Facebook friends, so you could invite your network to join you on a competitor’s platform.

And isn't competition healthy? Well, sure, but in this case the proposed solution looks a lot like the behavior that landed Facebook in hot water in the first place -- after an app scraped data on up to 87 million users.

None of this is to say that Facebook shouldn’t take steps to keep bots and foreign influence off its platform and to more clearly label political ads. Nor does it imply that the federal government shouldn’t keep an eye on competition in tech. But beware. The wrong moves -- by Facebook or governments -- could easily change the status quo for the worse.