Political Economy Workshop (Gov 3007)

First, Tom O'Grady will present his paper "The role of political elites in lowering support for the welfare state: evidence from British parliamentary speeches," with Chris Lucas as discussant. Second, Matt Lowe will present his paper "The Selection of Top Politicians: Evidence from the UK," with John Marshall as discussant. Abstract for "The role of political elites in lowering support for the welfare state: evidence from British parliamentary speeches" Support for the welfare state has declined dramatically in the UK, despite rising inequality. I document, and aim to explain, several key changes: people are less supportive of redistribution and unemployment benefits, and have become increasingly negative about the effects of welfare and about welfare recipients themselves. These changes are especially prevalent amongst younger voters, but occur across all economic classes. I build a theory that suggests these shifts have occurred due to the influence of political elites’ rhetoric. I then describe a large dataset of speeches about welfare in the British Parliament that I have assembled from 1987 to 2014. Using a structural topic model, I document key changes in elite rhetoric toward welfare that have plausibly affected the opinions of ordinary voters. I end with some suggestions for the next phases of this project, including testing the lnk between rhetoric and opinion - on whch I would welcome feedback. Abstract for "The Selection of Top Politicians: Evidence from the UK" I exploit a natural randomized experiment in UK Parliament to shed light on what drives the selection of the most senior government decision-makers. Each year, hundreds of MPs enter a lottery to get the opportunity to legislate. Using archival data from 1950-1990 I find that high-ranked winners are 34% (8 p.p.) more likely to ever become ministers and hold 28% (0.4) more political offices over their career. Several pieces of evidence suggest that the key mechanism is exposure (as opposed to learning-by-doing or political survival): the effect of winning is larger for women (an under-represented group), smaller if there are more winners from the same party in the same year and smaller when the MP has won before (diminishing returns to signals). Taking the exposure mechanism seriously, I lay out a parsimonious structural model in which party leaders have asymmetric priors (for women vs. men) and learn about the underlying quality of each MP over time. Initial calibration of the model suggests that a lack of learning can significantly affect the efficiency of political selection.