Anti-establishment feelings are one of the most important concepts used to understand the behavior of parties and politians, and their interactions with multiple audiences. Despite a rich theoretical literature exploring anti-system sentiments such as populism and technocracy, there remains a lack of consistent measures across parties, countries, and over time. In this study, I introduce a novel approach to quantify anti-system sentiments using word-embedding techniques applied to legislative speeches. I present an original dataset encompassing 7 million speeches over a decade from two key democratic states: the UK and Germany. The methodology includes rigorous validation tests and an application of this multilingual approach to examine the impact of rhetoric on party choice. The findings suggest that parties with a distinct rhetorical profile gain increased voter support. I conclude by discussing how this measurement can address both classical and contemporary questions regarding political-rhetorical interactions.