Does Facebook increase political participation? Evidence from a field experiment

ABSTRACT

During the last decade, much of political behaviour research has come to be concerned with the impact of the Internet, and more recently social networking sites such as Facebook, on political and civic participation. Although existing research generally finds a modestly positive relationship between social media use and offline and online participation, the majority of contributions rely on cross-sectional data, so the causal impact of social media use remains unclear. The present study examines how Facebook use influences reported political participation using an experiment. We recruited young Greek participants without a Facebook account and randomly assigned a subset to create and maintain a Facebook account for a year. In this paper we examine the effect of having a Facebook account on diverse modes of online and offline participation after six months. We find that maintaining a Facebook account had clearly negative consequences on reports of offline and online forms of political and civic participation.

KEYWORDS:

• Political participation
• experiments
• political behaviour
• Facebook
• social media

Acknowledgements

The authors would like to thank the participants of the ‘Social Media and Political Participation’ panel of the 2014 EPSA General Conference, as well as the participants of the ‘Experimental Research’ panel of the 110th APSA Annual Meeting for commenting on earlier versions of this paper. Our special thanks go to Jonathan Nagler, Don Green and Steven Smith for their encouraging and constructive comments and suggestions. Yannis Theocharis would especially like to thank Jan W. van Deth for the extensive conversations about, and ideas for, the study’s design.

Disclosure Statement

No potential conflict of interest was reported by the authors.

Notes on Contributors

Yannis Theocharis is Senior Research Fellow at the Mannheim Centre for European Social Research (MZES), University of Mannheim. His research interests are in political behaviour, new media, and social capital. [email: [email protected]]

Will Lowe is Senior Research Specialist in the Department of Politics at Princeton University. He is a political methodologist focusing on quantitative political text analysis. [email: [email protected]]

Notes

1. According to a recent study by the PEW Internet and American Life Project, only 4% of social media users block, unfriend, or hide someone because they disagreed with something they posted about politics (http://pewinternet.org/~/media/Files/Reports/2012/PIP_SNS_and_politics.pdf).

2. Many analysts argued that the result of that election would define whether the country went bankrupt, and consequently be expelled from the Eurozone, or remain within the Euro economic area.

3. Greece is among the least saturated with Facebook countries in Europe with around 4 million users and less than 37% penetration in the population (http://www.checkfacebook.com). As the age group of the highest adoption of Facebook in the country is that of 25–34 years, followed by those 18–24 years, we chose to focus on these age categories aiming to capture people likely to adopt Facebook for the duration of our survey.

4. The unequal split between treatment and control groups was decided based on the expectation that panel attrition would be much higher in the treatment group due to the type of treatment.

5. Note that similar results can be obtained by the application of Factor Analysis, provided that we work with an appropriate type of item correlation matrix, for example, polychoric correlations for our Likert items. We prefer to work with the data directly.

6. For each reported participation type, we construct what is effectively a weighted aggregate of possibly different numbers and types of items, so it makes sense to normalise the factor scores to have fixed variance. This allows us to interpret coefficients in terms of standard deviations of response, abstracting from the details of our survey items. By aggregating in any way, from additive index construction to the IRT models we fit here, we should, provided the items do all tap the same participation type, improve the accuracy and efficiency of measurement. The price paid is that it is not easy to map back to particular items. It is also not possible to directly compare levels of different types of reported participation. To offset this disadvantage, we also construct a general participation measure.

7. The education variable also generates statistically significant effects. There are too few cases to confidently distinguish subgroup effects, although higher levels of education correlate weakly with larger drops in reported participation.

Additional information

Funding

This work was supported by the Alexander von Humboldt Foundation and the Mannheim Centre for European Social Research (MZES).