Lucas Da Silva
PhD Candidate, Political Science
| Email: | ldasilva@tcd.ie |
| LinkedIn: | https://www.linkedin.com/in/lucas-paulo-da-silva/ |
Doctoral Thesis Summary
My dissertation accomplishes two related tasks. First, it examines whether there are gaps in the ideological supply of media on the economic and cultural ideological dimensions. To do so, I create (1) an original dataset of one million randomly-selected online media articles from almost 9000 websites in over 26 countries and (2) a novel way to measure ideology using large language models. On the two-dimensional (economic/cultural) map of ideology, I find gaps in two ideological places: (1) the economically-leftist and culturally-conservative space and (2) the economically-rightist and culturally-progressive space.
Second, I test the effects of these media supply gaps on the people who inhabit these ideological spaces – “cross-pressured voters” (CPVs). I use a survey with eight experiments, three panel datasets, and a quasi-experiment to test how CPVs select media – since they lack like-minded outlets – and how this media influences their voting behaviour. Early results indicate that media influences CPV voting considerably, especially by altering their perceptions of party positions. Since CPVs are a large and growing population in many democracies, this process could be highly consequential for elections.
Biography
I am completing my PhD in Political Science at Trinity College Dublin. I have been at Trinity for a while – I also received my MSc in International Politics and took all of the taught modules for the MSc Applied Social Data Science at Trinity! My bachelor's degree was a double major in Digital Media and History, with a minor in Business. Before commencing the MSc, I worked for five years in the digital media industry as a director of Southern European media operations, based in Austria and Cyprus. My research interests include political media, political behaviour, and public opinion. I am particularly interested in ideology, as expressed by media outlets, parties, voters, and even businesses. My methods include computational text analysis, experiments, quasi-experiments, surveys, and panel data approaches.