Is social media good or evil? “The answer is 'yes',” says Sinan Aral, the David Austin Professor of Management and Director of the Initiative on the Digital Economy at MIT. His tongue-in-cheek response speaks to his hope that we can transcend the good-or-evil debate and think instead about how we can achieve the promise and avoid the perils of social media. During a virtual CHM Live event on November 11, 2020, Aral shared with veteran tech journalist John Markoff the four levers we have available to us to steer social media technology toward promise and away from peril: money, code, norms, and laws. Within each category, Aral describes the challenges we face and suggests concrete actions that companies, governments, and individual consumers can take to tackle them.
In what Aral calls a decade of “techno-dystopianism,” social media has become a pariah even as it is deeply ingrained in many of our lives. But not all of social media’s damaging effects are new to humanity. The phrase “fake news,” for example, first appeared in a Harper’s Magazine article in the 1930s. The difference today, Aral says, is the speed, breadth, and depth with which social media allows misinformation to spread. A 2018 study showed that falsity spreads farther and faster than the truth in every category of information. We can blame social media, at least in part, for the spread of misinformation that caused the resurgence of measles in the US and the erosion of public confidence in the COVID-19 vaccines under development.
Social media also permits a degree of customization and personalization that was never possible in traditional media like books or television. Through what Aral has termed the “hype machine loop,” social media newsfeed algorithms read our behavior to determine the kinds of content we like, and then show us more of the content we prefer while excluding content we don’t like. This narrows the scope of information we are exposed to until our newsfeeds become echo chambers that isolate us from different views and beliefs, contributing to polarization in such areas as politics. However, large-scale experiments show that when the algorithms are turned off, the human desire for diversity of information returns. “We don't become permanently changed by the algorithms to crave narrowness,” Aral says.
So should we just get rid of the algorithms? The result would be a reverse-chronological newsfeed with no filtering whatsoever, flooding users with far more information than they can absorb. Fortunately, the algorithms can be rebuilt to mitigate negative effects. Software engineers can minimize polarization, for example, by adding multi-variate objective functions that put more weight and importance on diversity and alternate viewpoints. They can also tackle misinformation by systematically nudging users to consider whether a news story is true or false. When users stop to reflect, large-scale studies show that they are less likely to believe other fake news stories several minutes to an hour afterward.
If so much can be done to make social media platforms safer and more equitable, why haven’t social media companies made those changes? The current “hype machine,” engagement-focused business model has been very profitable for them—Facebook, for instance, brought in $70 billion in ad revenue last year. Newsfeed algorithms are designed to show us news we are interested in because it helps platforms get our attention, which they can then sell as advertising inventory to businesses, political campaigns, NGOs, and even countries. If the algorithms end up feeding us polarizing fake news, social media companies don’t see the negative consequences on their bottom lines—it might even make them more money. In the short term, there does not seem to be any economic incentive for the platforms to limit misinformation or polarization. Aral believes this a market failure that requires intervention.
Some see antitrust action, against Facebook in particular, as the regulatory solution. Aral believes the economics of the social media marketplace would make this ineffective, arguing that breaking up Facebook would only allow another Facebook-like company to dominate the market. In addition, because the social media economy runs on network effects, the connections that platforms enable create tremendous value. Research estimates that every year in the US, consumers are willing to pay about $370 billion more than Facebook charges for the services it provides—this amount is called "consumer surplus" and is interpreted as consumers’ net economic benefit from being able to access Facebook. In some countries, Aral says, “Facebook is the internet. If you tear apart those connections by breaking up Facebook, you're going to destroy value.”
Before we can consider anti-trust action, Aral believes we need structural economic reforms that will encourage competition in the social media economy, allowing startups to grow and take some market share from the dominant players. Because social media companies have no economic incentive to implement these changes, this is where legislation can help.
Aral advocates for legislation that requires social network portability and data portability, which would allow users to take their data and connections from one platform to another. Rather than having a number of distinct online identities and networks on, say, Facebook, Instagram, and Twitter, a user could have one identity across those platforms. This would give consumers a greater degree of freedom and choice, forcing platforms to compete for their attention through features like better privacy and security, as well as less fake news and interference. Aral gives the example of cell phone number portability, which not only allowed people to maintain their contacts when switching from one carrier to another but also created approximately $800 million in consumer surplus every quarter for years after it was legislated in Europe.
Likewise, Aral emphasizes the importance of legislating platform interoperability, which would require platforms to provide a standard set of messaging formats to allow users to communicate across different platforms. The ACCESS Act, currently before Congress, would require any social media platform greater than 100 million users to be interoperable by law. Interoperability successfully increased competition in the case of the AOL Time Warner merger, where AOL Instant Messenger (AIM) was forced to become interoperable with Yahoo Messenger and MSN Messenger. Within three years, AIM went from 65% market share to ceding the entire market to then-new startups like Facebook and Google. If we legislate platform interoperability, Aral says, we could then pursue antitrust actions against Facebook without destroying the value contained in connections on the platform.
What can we do as consumers to help social media change for the better? The #DeleteFacebook movement, launched in response to the Cambridge Analytica scandal, motivated users to change privacy settings in their accounts and had a negative (if temporary) effect on Facebook’s stock price. Likewise, the recent Stop Hate for Profit movement convinced companies like Unilever, Verizon, and Coca-Cola to pause spending on Facebook ads to protest hate speech on the platform. Facebook soon announced it would add warning labels to posts that broke its rules.
Whistleblowers within social media companies and employee walkouts also play a role in raising public awareness and motivating companies to improve their platforms. All this backlash, Aral says, indicates that the current model is not sustainable. He believes leaders of the new social age will seek to align long-run shareholder value with societal values rather than short-term profit.* This will be key to creating a better, more sustainable social media ecosystem.
When guided by ethics, possibilities abound for the types of values we might inscribe in our platforms. “Why do we have a 'like' button,” Aral asks, “but not a ‘truth' button or a ‘this taught me something’ button or a ‘this improved my health’ button?” Shifting from popularity to learning, health, or other metrics of success on social media could fundamentally change the kind of content we see and engage with. Rather than humanity serving the social media economy, social media could begin to serve humanity.
*To learn more about how companies can align shareholder value with societal values, watch Rob Chesnut’s Intentional Integrity workshop at CHM.