Individual drivers of toxicity in radical right-wing populist legislative campaigns 

with William T Daniel, Max-Robert Valentin, and Laurence Rowley-Abel in West European Politics 

Voters don't like Politicians Using AI LSE British Politics and Policy Blog

with Timea Balogh, Louise Luxton, and Zoe Greene

From Chatbots to the Ballotbox - How Voters used AI to Learn about the Scottish Parliament Election Scottish Parliament Election Analysis 2026. The Campaign, Voters and Policies: Early Reflections from Leading Academics, p. 15-17

with Louise Luxton, Timea Balogh, and Zoe Greene

Local Personalities and Representation in France's Small Parties
Manuscript prepared for Oxford University Press: Oxford Handbook of Personality Politics.

with William T Daniel

Issue Communication Strategies Across Electoral Systems 

Diverging from the Party Line: Candidate Messaging on Social Media Across Electoral Systems

Are campaign strategies structural, partisan, or personal? Individual-Level Evidence from French Legislative Elections

with William T Daniel, Laurence Rowley-Abel, and Max-Robert Valentin

 Machine Minds and Political Divides: How AI-Generated Summaries Shape Our Understanding of Parliamentary Discourse

with Zoe Greene, James Cross, and Derek Greene

Parties' Cross-Channel Communication Strategies during Election Campaigns

Social media has fundamentally altered the landscape of election campaigning, affording candidates unprecedented autonomy to communicate directly with voters outside of party-controlled channels with important implications for democratic accountability. This thesis examines parties’ cross-channel communication strategies during election campaigns, comparing the 2019 UK general election, the 2021 German Bundestag

election, and the 2022 French legislative election. The chapters offer a systematic account of cross-channel communication across party manifestos and candidate tweet during election campaigns, moving from what is communicated on the issue level, to how unified that communication is on the party level, to who diverges from it and under

what conditions on the candidate level. 

Using novel applications of computational text analysis methods, including topic modelling and word embeddings, the findings reveal partial issue overlap across platforms, with more technical issues dominating manifestos and more general language prevailing on Twitter/X. Party unity is consistently high, though British candidates show lower manifesto alignment than their German counterparts when measured through textual similarity. Most candidates do not diverge substantively from their party manifesto positions. These findings offer important insights into how party system characteristics shape cross-channel communication strategies under increasingly fragmented media conditions.