My undergraduate degree was in Natural Sciences. Everybody studying with me—botanists to astrophysicists—had to do statistics, whether they liked it, or not. In that environment, letting your research interests be guided by an aversion to a particular set of methods would have seemed as weird as becoming an art critic who refuses to look at the colour red. I worked as a journalist when I graduated, and started hanging out with people who’d at some point gained a degree in politics and often didn’t much like numbers. A disturbing proportion of them sought to be swayed purely by the eloquence of an argument when a little bit of data analysis would have gone a long way. Frustration with this kind of silliness led me to comparative government studies at Oxford. My DPhil here uses quantitative methods to look at the impacts of corruption.