By 2022, an estimated 4.7 billion people (out of a total population of 7.9 billion) are regular users of such networks, generally accessed through smart wireless devices. Although Facebook and other networks included in Zuckerberg's metaverse (WhatsApp, Instagram, Snapchat, Messenger) are by far the majority, their dominance is waning. In fact, TikTok is the fastest growing. And a network not long ago considered marginal like Telegram now has 550 million users. Originating in Russia, to escape censorship it is domiciled in the United Arab Emirates and the Virgin Islands. The growing prevalence of social networks, especially among the younger population groups, has raised alarm among the world's ruling elites, who consider them to be a threat to democracy. And it is that the control of information and communication has always been the foundation of power. Control that was in the hands of the State and large media companies. Which have now been disintermediated by the development of the internet and the technical and legal difficulty of controlling it.
The loss of communicative control is all the more serious the more critics and social movements opposed to the established order emerge. In some cases we applaud it, as with the Iranian women who mobilize and rely on social networks to resist the moral police of the ayatollahs. In others, however, we fear it, such as when QAnon induces democratically elected politicians to attack with hammers, or when networks popularize political tendencies that question democracy, such as Brazilian Bolsonarism or far-right acolytes in Spain or Italy. Hence the increasingly widespread opinion that the crisis of liberal democracy is the consequence of the free expression of any uncontrolled opinion in the communication space. Actually the issue is more complicated. And when there is a complication, you have to resort to social science research, which is what they are for. To save you consultation work, I recommend that you look for the exhaustive synthesis of the empirical evidence on the subject carried out by Professor Pablo Barberá of the University of Southern California in 2020.1 As well as the study published by Leikes in 2016.2 For Spain, where there is few studies, a good source is the work of the researcher of the Higher Council for Scientific Research (CSIC) Luis Miller.3
And what does the data tell us? In the first place, we are faced with a crisis of generalized political legitimacy. The majority of the world's citizens, in all countries, do not trust political institutions, nor parties, nor politicians. Something that I documented and analyzed in my book Ruptura (2017).4 Since then, mistrust has grown considerably, according to the Edelman barometer, the Pew Institute and the Eurobarometer. Spain presents a curious fact (which is not from the Center for Sociological Research, CIS). There was a pronounced drop in trust in 2019. Since then, mistrust has decreased, although the majority of citizens still do not trust any party or leader. Simultaneously, social networks have accentuated their role in information, misinformation and contradictory expression of political attitudes. But the fact that the two trends coincide does not mean that social networks are the cause of mistrust.
Among the tangle of studies in this regard, a fundamental result appears. People with established political convictions have become increasingly entrenched in their positions, both on the extreme right and on the extreme left. And they think that those who do not think like them are their enemies, a trait exemplified in Catalonia between independentistas and non-independence supporters while the majority of citizens have become disinterested in politics, according to recent work by Lucía Medina, from the UAB, reviewed in La Vanguardia. The stronger people's political beliefs are, the more they take refuge in networking sites that confirm their views and preferentially interact with like-minded people. In fact, this observation coincides with data traditionally verified in political communication analyses.5 Most citizens choose the media they read, watch or listen to, depending on whether they coincide with their own opinions. Since most of the media have a clearly predictable line of communication, especially the right, those who read El Mundo do not read El País. In the United States it was shown that the Fox News audience almost never listens to National Public Radio and vice versa. They do not read or see to find out but to confirm themselves. What's more, our brain is equipped to reject information that creates discomfort. Only when you have a critical spirit and search for a possible truth (which happens in a minority of people) does one dare to leave the comfort zone.
Something similar happens on social media. Given the ideological homogeneity of the group circles (the "friends" on Facebook or the messages shared by forwarding tweets on each other's list), the interaction in what are called echo chambers is the most intense and the most credible. for those who participate in these communication groups. From the communicative homophily that is generated, the messages go viral in electronic addresses identified as related. The mass communication society becomes a selective self-communication system.
Now, the vast majority of people do not have such drastic convictions. And therefore, they explore social networks without excessive previous prejudices. Moreover, social interaction in networks with political or ideological motivation is largely a minority in the number of people who participate. Hence an important empirical observation: for ordinary mortals, the more they use social networks, the more they are exposed to different points of view. And the more exposure they have to a diversity of opinions, the less polarization they experience. Polarization is exacerbated by networks in already polarized groups, not network users in general. What's more: the youngest groups are the least politically polarized, contrary to what is thought, partly because they believe little in politics. And it is precisely young people who use social networks the most.
In summary: we live in societies polarized in their extreme groups, not in the majority of citizens. But these extreme groups are the most visible and the ones that in certain circumstances act fanatically or violently, inducing hatred in society.
Now, if it doesn't seem that social networks themselves produce generalized polarization, what is the role of the algorithms that manage the networks in increasing the spread of messages that emerge from polarized people? The question is essential because several experiments have shown that polarized discussions increase dopamine in our brain, which leads to increased digital traffic. Precisely what constitutes the business model of social media companies. The more excitement, the more messages, the more messages, the more traffic, and the more traffic, the more free extraction of data for what is known as data capitalism, the basis of the internet economy.
The hypothesis that we can formulate based on the evolution of social networks and the use of algorithms to modulate the content of the interaction is that the algorithms used have become increasingly interventionist, stimulating the self-confirmation of one's own ideas and favoring confrontation with those who hold contrary ideas. Always looking for an increase in data traffic, regardless of its origin.
It should be noted that, after scandals such as the transfer of data by Facebook to Cambridge Analytica during the Brexit campaign, the networks have introduced content moderation mechanisms. For example, TikTok announced in 2022 that they had removed 110 million videos from the network, suppressed 20 million accounts on suspicion of having been created by children under 13, as well as another 33 million considered false. However, the removed videos represent 1% of the videos published between April and June 2022. Furthermore, the platforms do not report what the removed content is. And, above all, the algorithms are not public, so the verification of content moderation can only be done by the company itself. Musk, after the purchase of Twitter, announced the creation of a "Content Council", but he did not specify the criteria to be followed, while he has declared himself a supporter of uncensored freedom of expression. We'll see if he restores Trump's account, as he has implied on occasion.
There is a significant factor that can contribute to polarization in society at large. Namely the rapid and massive production and dissemination of fake news using the most important feature of the internet: the virality of messages and the massive use of bots to forward messages by thousands or millions, something that was decisive for Bolsonaro's victory in his first choice but it was not enough in the second. In other words, the origin of fake news and its political orientation arise from a limited number of Internet users, although its viral distribution exponentially increases the spread of the message. With which a fallacious communicative space could be configured much larger than that of politically radicalized groups. However, receiving a message does not necessarily impact a person's opinion if they are not predisposed to give it credibility based on their motivations and ideology. Communication consists of the adequacy or not adequacy between the emission of the message and the reception of the message. Therefore, it can be thought that receptivity to fake news is a symptom of the political and ideological skepticism that spreads in society and that is deepened by the dissemination of messages contrary to political, moral or scientific authorities. Such could be the case of the massive reception of anti-vaccine hoaxes, responsible for thousands of deaths.
In short: social networks amplify, rather than cause, the existing political polarization in all societies. This polarization is articulated from minority but ideologically homogeneous groups that are reinforced by their grouping in social networks. Being much more active and radical than the majority of the population, they build a controversial and controversial image that is reflected in the media of communication. The secret algorithms of the platforms that control the networks make effective content moderation difficult. In short, the crisis of political legitimacy, which has deep causes derived from the failure of political elites and their reciprocal animosity, motivates these elites to find a scapegoat for their responsibility in social networks. Networks that the elites themselves use as an essential weapon in their mutual delegitimization. In this way, a spiral of lies and hatred is created that undermines deliberative democracy in a post-truth society.
Manuel Castells is Emeritus Professor of Urban and Regional Planning at the University of California-Berkeley.
