This paper was submitted as my dissertation to the MSc in Social Data Science, University of Oxford, where I was awarded the Thesis Prize. Final grade: 78
Abstract
The British party system is known for its discipline and cohesion, but it remains wedged on one issue, and that is on European integration. This was observed both in the days of the EEC in the 1970s and the EU-Maastricht treaty in the 1990s; my thesis aims to investigate whether this holds true in the Brexit era.
And while most empirical work on this topic focus on party cohesion scores and other aggregate metrics, I argue that understanding the voting dynamics of individual Members of Parliament (MPs), whose actions ultimately determine the outcome of the legislation, may provide deeper insights into Brexit-related dissent. Thus, for this study I utilise social network analysis to unpack the patterns of dissent and rebellion among pairs of MPs.
Using data from Hansard, the official repository of the British Parliament, I compute similarity scores between pairs of MPs from June 2017 until April 2019 and visualise them in a force-directed network. Then, comparing Brexit-and non-Brexit divisions, I analyse whether patterns of voting similarity and polarity differ among pairs of MPs.
My results show that Brexit causes a wedge in party politics, consistent to what is observed in history. Further, a visual inspection of the network identified sixteen rebels, 5 from the Right-wing and 11 from the Left-wing. It appears from the analysis that the rebels’ position in the network is justified by their ideology. Finally, I further validate the positions of the rebels in the network by producing a quantitative measure, a rebellion score for each MP.
This study is timely and relevant particularly because the factionalisation in Parliament has left the country faced with indecision and uncertainty. Understanding the social interactions in the House of Commons may help us appreciate the individual dynamics of the voting process, and could facilitate a tangible proposal on how to best move forward.
Link to ArXiv: https://arxiv.org/abs/1908.08859v1
Publication forthcoming in Applied Network Science Journal.
