Creating Technology for Social Change

How do Social media Shape Collective Action? Helen Margetts at the MIT Media Lab

How does the changing use of social media affect politics?

Today at the Media Lab, Helen Margetts of the Oxford Internet Institute joined us to talk about a new book with Peter John, Scott Hale and Taha Yasseri, Political Turbulence: How Social Media Shape Collective Action. Ethan Zuckerman facilitated the conversation.

Margetts and her colleagues wrote the book because they wanted to look at the relationship between collective action and social media, putting meat on the conversation. There’s a widespread feeling that political mobilization is on the rise and that it has something to do with social media. In this book, Margetts and her colleagues delve into the data on that question. As a political scientist, Helen is in a continual state of excitement about the availability of behavioral data. Typically, political scientists have studied what people think they did, think they might do, or what they might prefer. Social media is exciting because it provides researchers with new sources of data about the political world that political scientists have never had before.

Social media allows “micro-donations of time and attention” to political causes, tiny acts of politics. With just one click of a button, you can encourage others to take action too. These tiny acts of participation can scale up to very large mobilizations, says Helen. She shows us an image of Tahrir square. The people there are not carrying out tiny acts. How did they get there? How did they get enough signals of viability suggesting that there would be a million people there? That happened via a wide range of media, says Helen, with each like sending a signal of viability leading to a mass mobilization.

These series of chain reactions of tiny acts of participation are affecting people’s individual lives for whole groups of people. How many refugees changed their course, asks Helen, after seeing an image on social media stating “refugees welcome”? These mobilizations also achieve policy change, although 99% of them fail. What factors predict their success or failure? It’s hard to generalize, says Helen. Successful mobilizations on the UK government e-petitions site develop fast, going up straight away in the first few hours. It’s not what political scientists expected. Political scientists expect an s-shape– that things would grow until they reach a threshhold and then take off. Instead, successful mobilizations take off almost immediately. When things rise quickly, attention also decays very quickly. If your petition hasn’t become popular within ten hours, “it’s digital dust.”

What does this mean for politics? When successful movements move rapidly, the result is unpredictability. How might we get a handle on this unpredictability? Margetts and her colleagues tried to get at these things through experimentation, looking for the forms of influence that they felt were most influential on social media. She talks about (a) social information, real-time information about the participation of others, as well as the (b) ways that our actions are visible to others. In the Ice Bucket Challenge, people’s actions were being made visible, adding more social information to shape others’ actions.

Changes in platforms also offer the potential for conducting natural experiments. When the UK government introduced trending information on their petition platform, it concentrated signatures toward the most popular petitions at the expense of the least popular. Rather than expand the number of signatures, it just concentrated them towards the most popular petitions.

Can these rapidly-unfolding movements lead to sustained, enduring change? Supported by digital communications, says Margetts, large mobilizations can develop without more classic forms of leadership. But these “act-first, collective-identity later” movements can struggle to lead to enduring change. To sum up, Margetts argues that online organizing has led to a system of political turbulence, a dynamic system with unpredictable behaviour. This turbulence, which is shaped by personality but lacks leaders, institutions, and organizations, has a high sensitivity to initial conditions, non-linear relationships, high interconnectivity, and tipping points. The book offers a start, but there’s still a way to go to better understand these issues.

Discussion with Ethan

Ethan: in many ways, you’re offering a response to Malcolm Gladwell’s way of seeing online activism, that activism is all about physical risks rather than online clicks. At some levels, you have a straightforward argument that many little things can add up to something more powerful. But when we lose risk and danger, are we talking about a very different form of mobilization.

Helen: it’s not just Gladwell. Many political cultures, especially British political culture thinks that if it doesn’t hurt or involve some long meeting or protest on the streets, or something that infringes on your life and making it worse– then it’s not politics. We are arguing against that view quite strongly. However, there is another sense in which the post-revolution will not be tweeted — there needs to be some kind of institutional catch-up.

Ethan: In many ways, the cost of participating in activism is being reduced. In a pre-digital age, your first form of activism might have been going out to attend a speech or rally, or purchasing the bumper sticker. In the grand scheme of things, those require effort. Now, any time there’s a political joke, we can retweet it and feel like we’re doing our part. How would you see the fear that the requirements of activism have become so low that the line between jokes and civics has begun to blur. What do you make of that?

Helen: most people never go to the demonstration or participate in the other, lumpy things. It’s blurring the line; the jokes have always been important. I wouldn’t agree that sharing a joke is not a political act. Social media is drawing people into politics and into fights against injustice who are young. Political scientists and political commentators have often worried that young people aren’t involved in politics, and now they’ve just voted Jeremy Corbyn into being the leader of the labor party.

Ethan: how does retweeting get us to electing Jeremy Corbyn? What is it about the ability to inexpensively make and spread media that makes an apparently-outsider, countercultural figure so powerful?

Helen: before, there would have been no way for an outsider to garner support without those institutional mechanisms. Now, it’s possible. Corbyn and his supporters ran a campaign that spread on social media in ways that the other candidates didn’t. It’s having a cumulative effect. It’s giving people the feeling that the extraordinary can happen. When people see that figures from outside the mainstream can become leaders, it’s having a pervasive effect on what people think about politics.

Ethan: that notion that the unlikely can happen seems well supported by the theory that you’re putting forward. If 99% of mobilizations fail, how do we deal with the issue of confirmation bias? Is it surprising that so many Arab spring movements failed?

Helen: Most books about political participation either have no evidence, or they are surveys, or they offer narrative and stories. We can’t get a sense of failures from that. There are many other datasets I wish we had that we don’t.

Ethan: I am struck by your three badgers example, where one petition gets only a few signatures, one gets a few thousand, and another gets over a hundred thousand signatures. Is there any possibility that if you could synchronize your research with people who study the spread of ideas on social media, could you figure out which petitions will be successful? Or will that be difficult to predict at the end of the day?

Helen: It could be part of the story. We did use data on celebrity tweets and institutional support, and we didn’t find evidence for that. Even if Justin Bieber were to retweet to something, it wouldn’t be guaranteed success. Our dataset doesn’t indicate where an institution is supporting the petition or not. But some of the success of a petition must be down to luck– if people don’t happen to notice the petition at the right time, it might not succeed.

Ethan: And you found that extroverts are less sensitive to social pressure than introverts? Helen: extroverts are more likely to do things when no one has done it before– they’re more independent. It’s complicated– we found a much weaker effect among people who saw themselves as the captain of their fate.

Ethan: You’ve argued that people can participate in ways that are much easier, they can take smaller acts. People can see each other’s actions, and when they see each other, they can coordinate and make decisions about what to do. These can turn into powerful mobilizations, except it often doesn’t. What I want to know is: what happens when we build these new social movements? Citing Zeynep Tufekci, Ethan talks about the Gezi Park protests of 2013. The result is the world’s least likely political collaboration: gay and lesbian turks and ultra nationalists are marching together because they’re both interested in mobilizing against Erdogan. But when they decide what to do together, they have a fragile movement of individuals who made their own choices, rather than a movement of groups who are committed to working together. Do you think this new form of mobilization is a good thing?

Helen: In social movement theory, you first get collective identity, and then you get collective action. Here, you get “act first, identify afterwards.” It’s more difficult to do that. But you might have said the same about the Indignados in Spain, which was based on many different dissatisfactions. Yet, we have seen that turn around; Podemos has emerged as a force in Spanish politics. The mainstream politics is still unable to accommodate it, but it still has a major part of the vote.

Ethan: is this unique moment resulting from changes in technology or something else? We have a lot of insight into how our institutions work, including their failures. Is this a moment of high mistrust in institutions, or are these movements uniquely enabled by technology?

Helen: I’m a social scientist– I’m not going to take the deterministic argument. But technology is definitely a factor. And yet our mistrust might be related to social media as well. Transparency has often been seen as a good thing, but it often leads to less trust. The more we know about our institutions and how they work, the less we trust them. There’s no way we’re going to get systemic democratic changes without society and technology alike.