Welcome to 'PhD Perspectives' – a monthly showcase of the dynamic research in the Department of Political Science at Trinity College Dublin. Each month, we interview a PhD student to explore their research journey, from inspiration to real-world impact.
Lucas Da Silva: 4th Year PhD Student
Can you give us a brief overview of your research and what inspired you to choose this topic?
In general, my research uses machine learning, experimental, quasi-experimental, and panel data methods to examine topics related to mass ideologies, voter behaviour, polarization, and political media.
More specifically, my PhD dissertation focuses on how the media influences the voting calculations and voting behaviour of “cross-pressured voters.” These voters hold economic and cultural ideological positions that are often considered contradictory. They are either economically left-wing and culturally conservative (left-conservatives) or economically right-wing and culturally progressive (right-progressives). These voters are a large and growing population (between a quarter and half of most European electorates) and are likely to be particularly susceptible to media effects.
I have long been interested in the changing cleavages and demographics of contemporary democracies, particularly the working-class shift away from center-left parties. I noticed that media outlets often do not represent these (mostly) cross-pressured voters, which inspired me to examine how this might influence their voting behaviour.
What has been the most surprising or interesting finding in your research so far?
So far, I have found two primary fascinating results (more to come hopefully!). First, I examine whether cross-pressured voters lack like-minded media outlets on both the economic and cultural dimensions. I use machine learning and large language models to classify the topics of almost half a million online media articles and then classify economic and cultural ideological positions. Based on 324 outlets from 26 countries, I use regression and clustering models and find that there are significant gaps in the left-conservative and right-progressive ideological spaces.
Second, using a quasi-experiment, I find that the longstanding boycott of the right-wing tabloid The Sun in Liverpool has increased Liverpool support for the Labour Party. The primary reason appears to be that people in Liverpool were no longer exposed to exaggerated depictions of Labour’s ideological positions. This suggests that the media influences voting behaviour largely through our perceptions of party positions.
How do you think your research could impact the field of political science or society as a whole?
This topic is relevant for both academia and the general public. First, the distinct concept of cross-pressured voters has been relatively ignored by much of the literature, partly because there were few cross-pressured voters until relatively recently. Although scholars are increasingly paying attention to these voters now, to my knowledge, there is no research on how media influences them. This is a significant lacuna, because they are now such a large population and are likely to be susceptible to media influence.
Second, society will also benefit from this research. If cross-pressured voters are influenced powerfully by media outlets, then elections may favour the interests of media elites to the detriment of the general public. Moreover, my research examines the particular pathways through which the media influences cross-pressured voters. Thus, it can provide insights for addressing this influence -- for example, by reducing how outlets distort perceptions of party positions.
What advice would you give to other students who are considering pursuing a PhD in political science?
Don't do it! Just kidding. First, think about what research topics will still be interesting and important in ten years. Unlike most pre-doctoral research (which usually lasts no more than several months), your PhD dissertation is a multi-year journey. That's a long time to focus on one topic. If you select a topic that is trending now but will be forgotten next year, you probably won't enjoy it and you'll struggle to have an impact.
Second, be sure to differentiate your work from research that exists and also research that does not yet exist. Do a thorough literature review to avoid reinventing the wheel. It is also frustrating to discover that someone recently published something almost identical to your topic years after you already picked it. So think carefully about what others might research and how you can make your project original by including several unique elements to it.
March 2025