Understanding Online Collective Action using Big Data: Analysing the Growth Rates of Internet-based Petitions

Now that so much of collective action takes place online, it leaves digital imprints which can generate large-scale datasets, offering political science researchers potential for understanding mobilization using transactional data in a so-called ‘big data’ approach. This paper aims to use big data to understand the mechanics of one form of online mobilization – electronic petitions. It tracks growth and dissemination of over 25,000 petitions to the UK Government by scraping government and social media websites from 2009-2012. The data shows that 95% of petitions fail even to get 500 signatures, while a few obtain millions and widespread attention. We investigate the relationship between petition growth and factors such as petition topic; start date; geographic location; diffusion patterns on social media; and information provided on the petition platform. We find that growth is leptokurtic, rather than normal, suggesting online mobilization is characterized by punctuated equilibria (as identified in agenda-setting research); and most successful petitions grow rapidly ('first day' signatures is the best predictor of total signatures), explaining some volatility in online collective action. Findings demonstrate the potential of ‘big data’ for political science research, with implications for mobilization strategies and design of websites aimed at fostering citizen engagement with policy issues.