Tue, 04/06/2021
This post is a summary of Episode 13 of The Nebraska Governance & Technology Center’s (NGTC) Podcast Series, Tech Refactored. Host Gus Hurwitz, Director of the NGTC, was joined by Joshua Tucker, Professor of Politics and co-Director of the Social Media and Political Participation (SMaPP) lab at New York University.
It has become axiomatic among the punditocracy that, when it comes to democracy and the health of democratic institutions, social media is the functional equivalent of the four horsemen of the apocalypse, all rolled into one. This week’s guest, Joshua Tucker, has recently released an anthology of essays entitled Social Media and Democracy: The State of the Field and Prospects for Reform that takes a nuanced, data-driven look at the reality behind the relationship between social media and democracy, as well as arguing for more transparency in terms of the data that is made available to social scientists in order to better study the effect of social media platforms on the health of democracies and separate fact from fiction.
The book grew out of Tucker’s work with the Social Media and Political Participation Lab (SMaPP) at New York University. What sets Tucker’s work with the SMaPP lab apart from other organizations that are doing political science research is that they have adopted the kind of “laboratory model” that is more commonly seen in the natural sciences. At the core of the project, then, is the collection and processing of vast amounts of data. As a consequence, the lab has multiple PIs (principle researchers); two political scientists and one biologist/computer scientist, as well as two full-time research engineers, six postdoctoral fellows, and “half a dozen to a dozen PhD students” who are working with the lab at any given time. The lab has dedicated positions that focus on data gathering, fundraising, storing data, and developing analytics, while also using ad hoc teams that are set up to collaboratively work on a host of different projects as they are conceived.
The lab began, in essence, as a solution to an internal bifurcation in political science, and the social sciences more generally. On the one hand, you had a group of researchers, mostly data scientists, who were concerned with questions around how best to reduce text to numbers so as to be able to leverage algorithms to analyze and draw conclusions about a huge body of speech that would be prohibitively large for even a group of individuals to process and categorize. Then, on the other hand, you had a group of more traditional political scientists who were intensely interested in what sort of political speech was being posted on social media, but had no way to meaningfully process the vast amount of content that was being posted. The SMaPP brought both of those groups together to use the tools developed by the data scientists and mathematicians in ways that could draw conclusions that were meaningful within the existing frameworks developed by political scientists.
Tucker walks the audience through the history of academic thinking regarding the interplay between social media and democracy. In the early days of globalized social media, social media platforms proved to be valuable tools for those fighting autocracies- finally they could organize in ways that were free from the censorship of authoritarian governments. This led one author, Larry Diamond, to write an article referring to social media as “liberation technology” in the Journal of Democracy. Sixteen years later, Nathaniel Persily wrote a piece called “Can Democracy Survive the Internet?” in that same magazine.
Tucker argues that the truth lies somewhere in the middle, based on two basic assumptions. First, social media gives voice to people who are excluded from access to mainstream media. Second, despite the fact that social media democratizes access to information, “it can still be used as a tool for censorship.” The same tools that can give access to democratic voices within an authoritarian state also give access to anti-democratic voices within a democratic state. And when you add in factors like Watson-bots, which artificially amplify certain voices, effectively drowning out sincere forms of conversation, it is clear that social media is neither pro- nor anti-democracy, but instead “another tool that can be used by people in their contest for political power- a very fast moving, rapidly changing tool.”
Building on that idea, Tucker reflects that in some ways this is both the best of times, and the worst of times, to be a political scientist. It is the best of times in that, where access to social media data is available, it can provide “granular data about a couple billion people on earth, who are spending some of their time talking about politics. That is just completely different than what we used to do, which was (to rely on) aggregate measures of events, like ‘how many people showed up at a protest, or how many votes did somebody get in an election. (...) But now, all of a sudden we have all this digital trace data, and in a way we’re still just scraping the surface of it.”
On the other hand (“the worst of times” side of the analogy), in the past, when a social scientist gained access to data, that data was theirs, and they had control of it. “Now we’re in this very weird world (...) where all this data that we need in order to face these growing, pressing problems and questions facing society- they’re in the hands of (…) a few small, really powerful private companies. And it leads to all sorts of challenges and conundrums in that regard.
That leaves social scientists with three problematic paths in terms of collecting data on online platforms. First, they can try to work around the platforms: “collecting the data that you can without collaborating with the platforms has huge advantages in terms of independence, but has huge disadvantages” in that a simple change in platform structure/policy can make that data unavailable in an instant. A second option is to “collaborate with a platform, and that has all sorts of other problems, because you have to think about independence and transparency.” The third option is to “try to work with the government to regulate and change the legal infrastructure, to make it so that (the data scientists require) is available.” Each one of those approaches has costs and benefits associated with it, and Tucker advises that researchers, if they are able, pursue all of them concurrently, in the event that one or more of them doesn’t work out.
What, then, is an example of the types of questions that social scientists are trying to answer with all this data? Tucker’s book is an anthology that contains chapters written by various authors on the subject of social media and democracy, and while he recommends them all, one he highlights in particular is an article by Pablo Barbera, analyzing the relationship between social media and political polarization. While conventional wisdom holds that there is a direct causal relationship between the rise of social media and an increase in political polarization, Barbera argues that the academic evidence is in fact much more mixed; indeed there is some evidence indicating that these two events may actually simply be co-occuring with no underlying causal relationship.
While there is conflicting evidence surrounding many elements of the relationship between social media and democracy, there is one takeaway on which Tucker is absolutely certain, and that is the need for online platforms to make usage data available to the public and outside, independent researchers so they can perform rigorous, unbiased analysis to better understand what is happening on social media.
“There are lots of different opinions about what should be done in terms of regulation of these platforms, but what should be underlying all of it is that we want to have a better understanding of what’s happening. And the only way that’s going to happen is if we allow people who don’t work for the platforms, and are not constrained by the platforms, to conduct rigorous scientific research (and to) have access to this data so it can be used for advancing our scientific understanding of the impacts of these platforms on society. And then it can be used to inform public policy.”
For those who are interested in these topics, and would like to explore them from other perspectives, Tucker recommended the work of Joan Donavan who has studied the use of internet technology by white nationalists from a near-term historical perspective.
Tucker also referenced the Knight Foundation, which provides grants to support work in journalism, communities, and the arts, and which highlights some of that work on its website. (the NGTC also has generously received funding from the Knight’s foundation).
Finally, Tucker’s book Social Media and Democracy: The State of the Field and Prospects for Reform is available in a free, open access format here.
