Last year saw a conceptual breakthrough on that front. In a paper titled “The Antitrust Case Against Facebook,” the legal scholar Dina Srinivasan argued that Facebook’s takeover of the social networking market has inflicted a very specific harm on consumers: It has forced them to accept ever worse privacy settings. Facebook, Srinivasan pointed out, began its existence in 2004 by differentiating itself on privacy. Unlike then-dominant MySpace, for example, where profiles were visible to anyone by default, Facebook profiles could be seen only by your friends or people at the same school, verified by a .edu email address. “We do not and will not use cookies to collect private information from any user,” vowed an early privacy policy.

As the company grew, Srinivasan argued, Facebook tried to backslide on its privacy commitments, but it faced discipline from a market that it still hadn’t cornered. In 2007, it rolled out Beacon, a product that allowed it to track user activity even when they were off the site. Facing fierce backlash—Beacon publicly reported your purchase habits on friends’ NewsFeeds—the company discontinued Beacon within the year. Zuckerberg called it a “mistake.” After rivals like MySpace exited the stage, however, Facebook had less to fear. Today, its “pixel” tracks users all around the internet, just as Beacon did (but without the ill-considered NewsFeed posts). According to Srinivasan, this is just one of many ways in which Facebook rolled back privacy protections once it sensed users couldn’t take their business elsewhere.

Srinivasan’s theory provided an elegant theoretical solution to the consumer harm puzzle but left open some empirical questions: Did Facebook actually compete for users by offering better privacy protections? And did it really renege on those commitments later on simply because the company’s leaders thought they could get away with it?

The case filed by the state attorneys general provides new evidence suggesting that the answer to both questions is yes. It cites an internal report from 2008 in which the company identifies strong privacy controls as one of four pillars of “Facebook Secret Sauce.” The report observed, “Users will share more information if given more control over who they are sharing with and how they share.”

The most revealing insight comes from the summer of 2011, when the company was gearing up to fend off the threat of Google’s rival platform, Google+. The complaint quotes an email in which Facebook COO Sheryl Sandberg wrote, “For the first time, we have real competition and consumers have real choice … we will have to be better to win.” At the time, Facebook had been planning to remove users’ ability to untag themselves in photos. One unnamed executive suggested pumping the brakes. “If ever there was a time to AVOID controversy, it would be when the world is comparing our offerings to G+,” they wrote. Better, they suggested, to save such changes “until the direct competitive comparisons begin to die down.” This is close to a smoking gun: evidence that, as Srinivasan hypothesized, Facebook preserves user privacy when it fears competition and degrades privacy when it doesn’t.

The states and FTC make a number of other claims about the harm caused by Facebook’s monopolistic practices, but they are relatively vague. Sure, Facebook’s tendency to gobble up potential competitors or cut them off from its developer tools has probably reduced the level of innovation in the field, but who’s to say what social networking would look like in the counterfactual scenario? The privacy theory, by contrast, has the virtue of being concrete: Facebook really did backslide on privacy commitments as it grew more dominant, and that appears not to have been coincidental. That isn’t to say the government will therefore glide through litigation; antitrust law remains stacked in favor of big business, and the federal judiciary is full of judges who were indoctrinated into a narrow consumer welfare model. But the privacy argument will at least get the enforcers’ foot in the door. Facebook may not charge users a fee, but that doesn’t mean users haven’t been paying a price.