I am a PPE student at Oxford University, where I was previously the student president of the EA Oxford group. I am interested in artificial intelligence (especially governance and strategy), great power conflict, and interdisciplinary research.
Thanks for this comment, and for the suggestions of literature and potential next directions! I'm excited to hear that arms control agreements are your main research focus and would like to hear about what you've found. My claim about the link between these track II discussions and the ultimate outcome of the negotiations is that (if they had any effect!) they contributed in a positive way because they were necessary for the Soviet negotiators to propose extensive limitations on anti-ballistic missiles and for the Soviet side to be enthusiastic about reaching some kind of extensive agreement on ABM. I don't think Pugwash discussions were necessary for the ultimate outcome of the ABM Treaty. That's because I think there's some chance that if the Soviets had not proposed extensive limits on ABM, the specifics of bargaining during the negotiations would have made them want to accept an American proposal for such limitations, even if other considerations left them enthusiastic about developing missile defences . My guess is that those scenarios are quite unlikely, so that Pugwash made a very signfiicant positive contribution to achieving an agreement on ABM. That said, I did not spend very much research time looking at the details of the SALT negotiations. I think Pugwash probably had only a very limited influence, if any, on the agreement on offensive systems (also, I think you're right that the agreement on offensive systems is underemphasised and probably had more important effects on the arms race - as I discuss briefly in 4.6, I think ABM turned out to be a bit of a side issue in slowing the progress of the arms race).
I really like your idea of emphasising that n=1 case studies can reveal mechanisms, and I agree that's an important strength of this kind of study.This is the most important original contribution of this research - as I mentioned in the report, the argument for the influence of Pugwash has largely already been made by Matthew Evangelista, and I'm mostly trying to figure out what it means for the mechanisms of impact for track II.
When it comes to the 'selecting on the dependent variable' concern - I feel like this would be more of a worry if my study was trying to say something about the likely impact of a typical track II dialogue. However, I see this study as instead trying to say something about how high the high upside of track II can be, which matters a lot if the impact of these kinds of programs is fat tailed (as I briefly discuss in 2.2). That said, understanding the impact of more typical cases of track II is also important for figuring out how rare big wins are and whether these programs consistently have modest positive effects or just usually fail to do anything. And Pugwash certainly is one of the most major and sustained track II dialogues between countries not in active armed conflict and focussed on arms control - so it is a natural one to look at if you're focussing on track II for improving great power relations.
Thanks again for the feedback and suggestions!