As a student attaining their Bachelor's in political science, I have in my thesis tried to bridge some of the gaps between political science and EA that critics often allude to, focusing on how to best adapt the ITN framework to political interventions and highlighting political polarisation as a potential topic of interest. I submitted this last month and thought it would be worthwhile to share with you all. Looking forward to your thoughts and feedback!
Thesis: https://drive.google.com/file/d/179yTM3BhF2zPMnHh3iOjpvooqFG3DJ8S/view?usp=sharing
Abstract:
This paper builds on approaches from political science and cause prioritisation to create a framework that can effectively compare solutions to different political institutions, arguing that these solutions have tended to be undervalued in cause prioritisation. It will show that frameworks from cause prioritisation can effectively be adapted to the political context by changing measurements and adding categories that can be excluded in non-political contexts. Then, this framework will be applied to compare solutions to political polarisation, or the increase of ideological and emotional cleavages around political issues in both the public and the political elite. This paper concludes that the most effective solution to political polarisation depends on context, but that increasing intergroup contact through citizens’ assemblies appears to be the most generally promising solution reviewed.
Great work, thanks for sharing! It's great to see this getting more attention in EA.
Just for those deciding whether to read the full thesis: it analyses four possible interventions to reduce polarisation: (1) switching from FPTP to proportional representation, (2) making voting compulsory, (3) increasing the presence of public service broadcasting, and (4) creating deliberative citizen's assemblies. Olaf's takeaway (as far as I understand it) is that those interventions seem compelling and fairly tractable but the evidence of possible impacts is often not very strong.
Thanks Tobias, that's helpful.