Social media is associated with the prevalence of mood disorders, depression, and anxiety. With no regulations to address the dangers of addictive digital content, antitrust may be the best tool we have since competition may create choices that would improve consumer welfare in the future.


In 2005, five percent of American adults were connected to social media. Today, over 70 percent have at least one social media account. While platforms such as Facebook, Instagram, and YouTube embedded in ubiquitous smartphones allow consumers to reach friends and access content at incredible ease, evidence from ongoing mental health studies suggests that social media content may be both addictive and, especially for young people, harmful. In light of the medical evidence and insights from behavioral economics, we argue that the way we regulate these platforms and the way we think about consumer benefit—including the way we enforce the antitrust laws—must adapt to account for the addictive features of social-media technology that can lead to overuse and subsequent harm.

Medical Evidence For Addictive Features And Harms

Neuroscientists have established that neural pathways associated with substance addictions (e.g. smoking) are similar to those associated with behavioral addictions (e.g., gambling). Social media may not rely on a physical product (e.g., a cigarette) or even the promise of something tangible (e.g., money from winning a bet), but there are affective stimuli that platforms optimize for higher levels of engagement. For example, variable reward schedules that have long been shown to lead to addictive behavior in mice and other animals, appear to drive identical behavioral patterns in humans who check social media with ever-increasing regularity for “likes” and other forms of engagement titrated from platforms. 

In order to enhance this habituated, constant engagement, platforms use dynamic algorithms designed to personalize content to keep users online, learning from previous user choices to better hone the content presented and the timing of that presentation. Though users experience decreasing marginal utility past some threshold of consumption, and may reach a level of social media usage that is harmful for them, this reward shape is different for the profit-maximizing platform: businesses with advertising models continue to earn more money from ad sales as more users are online for more minutes. Social media platforms are like cigarette companies that earn more profit from more sales of cigarettes, even as an individual addicted smoker is harmed from the smoke.

While the use of social media is relatively new, its explosive growth provides useful settings for scientific study, though it is still early to expect the full effect of platform engagement to be seen. Large cohort studies already show that social media is associated with the prevalence of mood disorders, depression, and anxiety. The strongest effects to date are among teenage girls. The effects are more intense for groups who spend more time on platforms than the average user. Firms have incentives to target these groups, such as teenagers, by making it easier to access content and personalizing feeds.

“Social media platforms are like cigarette companies that earn more profit from more sales of cigarettes, even as an individual addicted smoker is harmed from the smoke.”

Behavioral Economics Supports the Need For Regulation

In the neoclassical model, a rational consumer always chooses in a way that improves her welfare. In reality, consumers exhibit “behavioral biases” such as accepting defaults or placing more weight on short-term benefits than long-term consequences. These biases have been carefully modeled by economists, shown in the lab in myriad ways, and demonstrated in the field. The behavioral economics literature is decades old and has generated three Nobel Prize winners, so it is ready for consideration by antitrust enforcers and regulators.

Digital platforms have the ability to frame what the user sees, control stimuli, respond to user choices, and repeatedly test what content will cause users to remain attentive to the platform. Further, platforms have the financial incentive to addict users and extend their time on a platform in order to sell more advertising. Consumers, who try to exercise self-control over their time and choices, are likely not aware that they face an environment controlled by a supercomputer behind the scenes. They also may be unaware of the mental health and addiction risks of using social media. The setting is ripe for exploitation.

In other markets with asymmetric information, regulators have stepped in to choose defaults and frame choices for consumers who face behavioral hurdles to acting in their long-run self-interest. For example, the Credit Card Accountability and Responsibility Disclosure (CARD) Act of 2009 restricts who can open credit card accounts and how issuers can communicate with consumers. The CARD Act requires transparency of payments and fees and regulates defaults, all of which help a consumer resist the temptation of consuming something that may harm her in the future. This type of regulation improves her welfare. However, there are currently no policies in the US that directly regulate how social media platforms interact with consumers.

Regulation

The dangers of addictive digital content raise the very real question of when society will choose to regulate this area. Cigarettes were known to be a health hazard for decades before tobacco companies lost the political power that kept them from being regulated. Many addictive drugs are regulated (e.g., OxyContin), and financial products like credit cards and gambling are also regulated. Known types of regulations include age restrictions, protection by intermediaries (e.g. a physician), time and place restrictions, disclosure of risks, limits on advertising, and prohibited business models. Likely society will need new ideas as well as existing methods to create regulations that make social media safe for society.

Antitrust Analysis

In the absence of regulation, antitrust may be the best tool we have for combatting social media’s addiction and mental health harms. The antitrust laws safeguard competition. A competitive market can help improve product safety as firms competing for the business of consumers may make that aspect of the product salient. Think of Volvo in cars or Disney in family programming. Safety improvements may reduce ad revenue, and without competition, digital platforms may not have incentives to innovate along these lines. Acquisitions of nascent competitors may suppress competition from different business models (e.g., WhatsApp supported itself by subscription when it was independent). In the current digital environment, enforcers should be particularly attentive in preventing harms to innovation and quality.

The antitrust treatment of efficiencies in digital platform cases must also adapt to the addictive nature of these goods. In particular, output is no longer a reliable proxy for consumer welfare: more is not always better when goods are addictive. The claim that consumer welfare is raised by giving more Oxycontin to Oxycontin addicts is clearly problematic. Yet social media platforms with advertising-supported business models may claim that more “user engagement” is always great, regardless of the marginal utility the content provides that consumer. This is because even if the consumer is harmed, the platform earns profit from the additional ads. In a setting with addictive goods, the traditional antitrust shortcut, that output and welfare move together, is no longer true. Rather, antitrust must return to the root concept of interest—consumer welfare—in order to evaluate whether particular conduct benefits consumers.

Other antitrust issues that will loom large when platforms deliver harmful content include acquisitions of nascent firms that may eliminate competition on features such as safety and privacy. Such competition may create choices that would improve consumer welfare in the future. Society also gains from competition from alternative business models, such as the subscription model, that do not rely on advertising revenue to create better incentives for quality. Finally, defendants should have the burden of showing how specific conduct affects consumer welfare, not just that “output” has increased.