Assessing the effectiveness of e-petitioning through Twitter conversations
MOLLY ASHER Data Science Intern, University of Leeds, UK
CRISTINA LESTON BANDEIRA Professor of Politics, University of Leeds, UK
VIKTORIA SPAISER University Academic Fellow, University of Leeds, UK
Abstract: Recent trends of decline in political support have led political institutions to develop new democratic innovations to promote linkages with the public, beyond the representative democracy model. It is in this context that the UK Parliament introduced a new e-petitions system in 2015, aiming to significantly enhance its relationship with the public, namely by opening up the institution to a wider public and to develop deeper engagement. This paper explores whether this aim is being met, through specifically using Twitter data from conversations on e-petitions. Through the use of natural language processing (computational text mining), machine learning and social network analysis of Twitter data we explore (a) what Twitter data can teach us about the extent of people’s engagement with e-petitions beyond signing them, (b) the nature of Twitter e-petition conversations and the extent to which they reveal how people perceive the e-petition procedure and (c) who is taking part in these conversations and how they interact. Focusing mainly on a case study, we find the public reacts differently to an oral evidence session and a parliamentary debate: whilst the former stays factual and discursive, the latter becomes more emotive and critical. We also show clear patterns of polarization. Our results show that more care needs to be given to how petition debates unfold and the extent to which they’re inclusive of the original petition’s aims.
Keywords: e-petition, UK Parliament, Twitter, public engagement, Natural Language Processing, Social Network Analysis