Can Hashtags Change Democracies?

Can Hashtags Change Democracies? (Sunstein, Cass. #Republic: Divided Democracy in the Age of Social Media. New Jersey: Princeton University Press, 2017)

Cass Sunstein has had a long and distinguished academic career in the field of behavioral economics, teaching at prestigious universities such as the University of Chicago and Harvard (the latter to this day). Alongside his career as a professor and researcher, it is also worth noting that between 2009 and 2012 Sunstein worked as administrator of the White House Office of Information and Regulatory Affairs during the Obama administration.

Obama's first presidential campaign is recognized worldwide as a turning point in the use of social media to target specific sections of the electorate, particularly young people aged 18 to 25. During the 2016 presidential campaign, the massive use of social media was effectively exploited by his successor, Donald Trump, demonstrating the power of using the internet for political ends.

A glance at Sunstein's past publications makes clear that he has always been concerned with the role of heterogeneous audiences in building better democratic institutions. #Republic continues to pursue this question, as it explores the possible implications that social media may have for societies and democracy itself.

The book is entitled #Republic: Divided Society in the Age of Social Media, it was recently published (in 2017) and contains 11 chapters. Sunstein's main argument is basically summarized up in the three first chapters, while the others are used to illustrate his arguments by exploring specific technological trends (such as cybercascades), legal frameworks (such as regulation) and contemporary concerns (calling special attention to the use of social networks by terrorists to coordinate attacks and to recruit new members).

Before exploring some of these more interesting topics in greater detail, it is important to outline Sunstein's main argument. This is that societies tend to develop better if their citizens are exposed to different opinions and views, and build common ground and make decisions following debate between opposing parts. This means that censorship is one of the greatest risks to freedom of speech (specifically) and democratic system (in general). However, it is not the only one.

According to Sunstein's book, real life - in the real world - is made through contact between different groups and their ability to establish common ground despite their differences. Every day we face heterogeneous arguments about everything and the democratic system needs to be able to deal with diversity, at least as an ideal aim. However, what we have witnessed with the growth of the world wide web and - more powerfully - with the widespread use of social media, is that people are distancing themselves from others with different views and opinions.

In contrast to the real world, the virtual world does not stimulate the coexistence of heterogeneous audiences, but instead promotes polarization between groups as separated homogeneous audiences. In addition to this argument, Sunstein highlights the fact that younger groups, mostly those born after 1980 (the so-called millennials) do not primarily get their daily news and information about politics, economy, culture, environment (basically everything) by television or newspaper, but instead via social media.

And what are Facebook, Twitter and Instagram doing? These social networks have been developing technological tools to create niche audiences (considering the fact that users are consumers, and their attention a valuable commodity), where similar kinds information orbit around one other. In this sense, if you like a specific kind of music, only similar artists, similar songs and similar fans will orbit around your news feed, creating a well-functioning niche market. The same goes for politics. With the power of highly complex algorithms and the processes of machine learning, people are becoming increasingly enclosed in niches that receive only some parts of the wider reality.

Of course, humankind has never been aware of everything, everywhere, all of the time. However, the effects of polarization seem to have been aggravated by the spread of social media. When algorithms build a personalized world of information for individuals, this new world statistically constrains people's probabilities of encountering other realities, and may lead them to misrepresent reality itself.

This scenario has a powerful impact on politics, altering the possible ways of achieving common ends between groups of different political persuasions. In his main argument, Sunstein emphasizes that the lack of diversity puts people's freedom in jeopardy. At the end of the book, he offers some insights into how to deal with this challenge. However, his main achievement is to shed light on this online system of biased flows of information.

Chapter 01 is called 'Daily Me'. This is a quotation from MIT specialist, Nicholas Negroponte, from back in the 1990s, when he envisioned the creation of a package of information that would be designed only for you, according to your preferences, tastes, etc. This trend - that later became a reality - is supported by the tendency humans have toward homophily (to bond with the similar). The problem with this 'architecture of control' is that it slowly undermines the possibility of serendipity (new discoveries made by chance) and narrows information flows to the individual through greater and greater specialization.

If we apply this trend to politics, we see a greater tendency towards group polarization that may result in partyism (an automatic dislike of people from an opposing political party). To stimulate the argument, the author refers to the work of Jane Jacobs, 'The Death and Life of Great American Cities' and her thesis about the importance for the public of sidewalk contact: shared spaces offer an array of information that you may like, dislike or not even expect, but which reinforces the awareness of the different.

The second chapter (An Analogy and an Ideal) pursues this line of reasoning, focusing on how newspapers and television may be seen as big 'sidewalks' for heterogeneous topics. The author recognizes that mass media also carry political bias (and do not problematize enough in that sense) but emphasizes that a variety of themes are covered (from local to global, from the economy to culture), making mass media a sort of general-interest intermediary. That is not the case for special-interest intermediaries - social networks - and their tendency towards verticalization. Furthermore, algorithms are being designed to increase homogeneity, without people being fully aware of the consequences of this.

That's the argument Sunstein develops in the third chapter (Polarization). Once again, the idea of filtering information is nothing new. Actually, it represents the backbone of Republicanism: as direct democracy is unrealistic, you need to create structures of political representation to discuss what should be considered common goods. This filtering incorporates the 'clash of opinions' as a prerogative. Social media filtering, on the contrary, can create processes of balkanization, establishing the debate merely within echo chambers, where people just say and repeat the same things.

The impacts on politics are huge. The most fundamental of these is the decline of plurality, as like-minded people not only tend to unify around particular arguments, but also to radicalize them. Presenting several studies made by other specialists and researchers, Sunstein shows how similar groups end up endorsing extremist views, as they are not being faced by opposing perspectives. For example, using some empirical behavioral studies, he identifies the tendency of centrist-liberals, when put into contact with extreme-liberals, to move away from the center and towards more extreme positions. This alone has a negative impact on politics by impoverishing the range of possible perspectives, but another downside is how balkanization is affecting politicians' behavior. Instead of working towards the center (or at least with a pro-debate posture), politicians aware of their electorates may compromise on advancing the common good, due to balkanization.

Another interesting feature, notably during elections, is the power of cybercascades - the topic of the Chapter 04. The speed of information flows can really change elections by capturing hearts and minds, whether or not the propagated facts are true or false. They often bypass processes of reasoning due to the power of the 'crowd effect': a large group of people can end up believing in something just because several others have retweeted it. The use of hashtags during elections - including abuse - is notorious. The use of fake accounts to mobilize social media during election time, spreading support or disdain for a candidate is increasing everywhere.

Developing this idea, in Chapter 05 'Social Glue and Spreading Information', Sunstein addresses the dilemma of balkanization versus shared beliefs. The lack of democratic arrangements within our new informational system, as new algorithms (and the suppression of diversity) homogenize people's news feeds, may be progressively undermining equilibrating mechanisms and the important idea of shared beliefs. These may eventually harm one of the main characteristics of democratic societies: the recognition of diversity.

Because of that, Chapter 06 (Citizens) warns us of the consequences that the internet may have on our role as citizens. Social Networks are tools developed for the market and, despite being used for many other purposes (including political ones), their hundreds millions of users are, ultimately, consumers. As mentioned before, internet companies are fully aware that users' attention has value, and intentionally develop their networks to promote services, trade and niche markets.

The problem here is the lack of awareness users may have of the technical aspects of these tools and of their consequences. Considering that users are also citizens and are becoming more and more dependent on social media as their main source of information, the algorithm and its narrowing features may impact on people's sense of shared beliefs and common goods. The main warning made by the author is that the lack of diversity displayed in their news feeds may become a menace to liberty itself, as "deprivation of opportunities is a deprivation of freedom" (SUNSTEIN, 2017, p. 159).

In Chapter 07 (What's a regulation?), Sunstein presents some considerations for how to regulate this new informational world. Media regulation is a complicated topic as it may be confused with censorship. As freedom of speech remains the cornerstone of democracy, all debate surrounding regulation will be controversial. With regards to the virtual world, judges are even more cautious on how to deal with it. The author tries to point out the seriousness of this topic by presenting some examples regarding cybersecurity, but his main warning is outlined in Chapter 10 (Terrorism.com).

Terrorism is one of contemporary society's main challenges and terrorists know very well how to use the architecture of the internet for achieving their goals. Social networks are their main tool for propaganda, organization and attracting new followers. Bringing back Sunstein's argument, recruiters of terrorists seek to exploit the balkanization produced by social media by using polarization techniques. They try to isolate young men and women from their friends and families and intensify their alienation by capturing their attention and showing support.

In 2009, the Obama administration initially took a leading role on developing cybersecurity, creating a new interagency taskforce including the Department of Homeland Security and some Silicon Valley companies. The main idea was to pool different skills and expertise to monitor terrorist cells and other forms of cybercrime. However, internet monitoring can be a very controversial matter, as we can see from frequent news stories highlighting abuses, as intelligence services seek to control politicians, social movements, activists and minorities on grounds of 'national security'.

In response to these dilemmas, Sunstein defends a model of 'structural regulation' (Chapter 08, Freedom of Speech), arguing that the system as a whole should not work to control speech but, on the contrary, to guarantee freedom of speech. In that sense, the author makes some suggestions that could significantly mitigate the growing risk of balkanization of people's news feeds. One of these is the possibility of creating a 'serendipity' or an 'opposite view' button. This suggestion guarantees the personal choice of clicking or not, but still offers users the opportunity of seeing something different from their own perspective. He also suggests the enlargement of 'deliberate forums', built specially to permit the debate among different groups. A third suggestion is to stimulate 'disclosure policies'. When companies open their data to inform the public, this democratic act not only promotes transparency, but can help in the design of better approaches for regulation, based on empirical data.

Cass Sunstein closes his book with a final chapter that sums up his main insights: notably, the argument of diversity as the backbone of democratic systems. The author repeats the warning about the lack of diversity in social media, bringing attention the potential negative consequences, such as radicalism and polarization. At the end of 2017, the net neutrality principle adopted during the Obama Administration was revoked, meaning that telecom companies can now interfere more extensively with content and choose what they want to promote and sell to the public. In this context, #Republic is welcome reading, given the major challenges the virtual world is about to face.

Revised by Matthew Richmond

Sidebar

This publication is registered under a CC-BY Licence.

Footnote

(·) http://dx.doi.org/10.1590/1981-3821201800020007

References

References

Sunstein, Cass (2017), #Republic: Divided Democracy in the Age of Social Media. New Jersey: Princeton University Press. 328 pp.