Political Misinformation Thrives on Media Competition

In new research, Arseniy Samsonov builds a model showing how having available to the public a multitude of media outlets and social media platforms would not help reduce misinformation from politicians. Rather, monopolistic power could enable these outlets to retain control over the narratives around the information that these politicians provide to journalists and platforms in exchange for publicity and coverage, thus reducing misinformation.

Politicians have exclusive access to important information that the media would like to acquire and sell to their audiences. Politicians may pass along this information to the media, but in exchange they may demand favorable coverage. Media competition worsens this dynamic by giving journalists a stronger incentive to enter this bargain, so that they can provide a better product and gain a larger audience. The same logic applies to social media: politicians may patronize smaller, alternative platforms to avoid fact-checking on established ones.

In 2014, Russian journalist Pavel Kanygin of the independent Russian outlet Novaya Gazeta arrived in the Ukrainian city of Artyomovsk (now called Bakhmut) to cover the independence referendum organized by pro-Russian rebels. Members of the rebel group kidnapped and beat the journalist, later releasing him for ransom. This episode, exemplary of the dangers contrarian reporters can face in regions with low protections or even disdain for journalists, shows why pro-government media have found it easier to cover the Russian involvement in the east of Ukraine and, later, the war between the two countries. As my former colleague, Ashley Blum, shows in her dissertation, Russians favor pro-government over independent sources because the former have better access to information. 

It is not just in Russia where journalists benefit from taking the government’s side. Shortly before the Iraq war, American mainstream media focused on interviewing administration officials and purported Iraqi defectors, which helped attract viewership but also created popular support for the invasion. To take a peacetime example, former President Donald Trump gave outsized attention to the sympathetic right-wing network Fox News compared to more oppositional left-wing mainstream media. Presidential interviews provide insight into an administration’s intentions and are a valuable resource that can boost a channel’s reputation and viewership.

To reiterate, in dictatorships and democracies, the government controls access to exclusive information which citizens care about. Journalists seek to obtain this information to create reports and broadcasts that are of interest to its readers and viewers. If the government only shares its information with loyal journalists, it creates media bias and swings public opinion in its favor. In other words, the government bribes the media with information. My recent paper presents a game-theoretic model that describes this possibility.

Dictators have more straightforward tools to control the media. But in countries where killing, threatening, and corrupting journalists is more challenging, bribing the media with information may be an essential source of media bias. In democracies, it is common to have many media outlets. My model shows that this may be a curse rather than a blessing. 

Suppose that a country’s population consists of loyalists and dissidents with respect to the current government. Loyalists choose the media that side with the government. They are okay with the bias because of the exclusive information such journalists get. Dissidents prefer media that publish contrarian voices. Imagine that you are the president who wants to swing public opinion in your favor. How many loyal journalists do you need? Probably not many. Instead, why not let them compete in how friendly they are and choose the best? Such behavior seems realistic: Trump prioritized Fox News’ Sean Hannity over other journalists. 

In this scenario, media outlets less biased toward the president split the dissident audience. The more numerous these other outlets are, the smaller the portion of the dissident audience each receives. Hence, each media outlet is incentivized to tolerate more bias if it gains them access to president-held information and, consequently, a larger audience. If there are more outlets and thus more competition, the president, in this scenario, is likely to convince more outlets to compete to provide more favorable coverage in return for government information. If there are only a few media outlets, though, there is less competition, and the media outlets command more buyer power to control the narrative in exchange for government information.

This logic may explain why Americans report so much media bias. In the U.S., numerous television stations and news websites compete for readers’ attention. Citizens are skeptical of their quality. In 2017, 45% of Americans saw “a great deal of bias” in the news media, and 51% could not name an objective source. By contrast, only 25% saw “a great deal of bias” in 1989, when the choice of media outlets was more limited. 

Do the rules of the game change as people increasingly use social media to learn the news? Politicians can interact with citizens directly and care less about journalists as intermediaries. However, social media platforms have incentives and behave strategically. Like traditional media, they value publishing engaging content that makes consumers spend more time on their platforms. A politician or political movement that uses Facebook or Twitter creates such content. 

But politics can also be a concern for social media firms. In 2021, Google, Facebook, and Twitter leaders had to testify on disinformation before the U.S. Congress. According to materials that whistleblower Frances Haugen shared, Facebook’s employees also showed concern over disinformation that they believed the platform was spreading. 

These circumstances likely pressed Facebook to create its fact-checking program. Facebook and Twitter also implemented more straightforward tools: removing content that politicians created or banning some of them altogether. An example of the former is when several social media companies removed videos where former Brazilian President Jair Bolsonaro linked Covid-19 vaccines with AIDS. The Trump ban is a famous example of the latter. 

Politicians have responded creatively. Bolsonaro urged supporters to follow him on the messenger app Telegram. Trump first migrated to the social media platform Parler and then created the social media platform Truth Social. In 2021, the Nigerian government had a standoff with Twitter and created an account on Koo. Koo is a microblogging service that the Indian government started actively using when it had its standoff with Twitter earlier that year.

Is relocating a smart move? Bolsonaro has 15 million followers on Facebook and about 2.5 million on Telegram. He lost the recent presidential election in fall 2022. Time will tell if other politicians forced to leave major social media will succeed. My recent paper provides a game-theoretic model highlighting a critical factor: how attractive are major social media platforms relative to competitors? The logic is simple. If a social media platform is ex-ante popular, and the politician leaves it for a competitor, few people will move to the competing platform just because of the politician. Hence, the politician will not reach a large audience. Anticipating this scenario, the politician will comply with the standards of the major platform (in most countries, this is Facebook). Conversely, a politician may risk switching if Facebook’s advantage is limited. Counterintuitively, if this logic is correct, the fact that Facebook is close to a monopoly in many countries may help the company reduce disinformation. 

What policies could affect disinformation? A higher number of media outlets does not reduce bias. While suggesting a different mechanism, Julia Cage’s empirical study reaches the same conclusion as my model. The lack of a positive relationship between the number of media outlets and their quality suggests that media subsidies, a popular European policy, are ineffective. My model of social media implies the same effect: the higher monopoly power of big social media platforms leads to less disinformation. Conversely, policies (such as the U.S. Congress’ Access Act) that encourage platform variety can allow politicians to provide more misleading information. Of course, there are many other considerations that determine how policymakers should regulate social media. For instance, tech firms can abuse their power to control what users see. One of the dangers is that they will do so to promote the political agenda of their employees or senior management. They may also cooperate with repressive governments and silence dissenting voices to retain valuable markets. 

To conclude, traditional and social media may tolerate bias because politicians provide engaging and exclusive content. If this theory is correct, media competition exacerbates the problem by giving politicians more bargaining power. I base this conclusion on anecdotes and game-theoretic models that rationalize such behavior. Empirical studies based on observational data and experiments should follow before making a policy recommendation. Policymakers should also remember that such regulations can affect society in many ways. 

Articles represent the opinions of their writers, not necessarily those of the University of Chicago, the Booth School of Business, or its faculty.