Hello Cameron,
Thank you for the thoughtful reply and for laying out the two families so clearly.
I agree with your framing, and I am happy to revise the paper accordingly. I will add an explicit limitation stating that the deviation results hold under the no-reactive-defender assumption shared by Budish (QJE 2024) and the Bonneau papers, citing Moroz et al. (2020) as the counterpoint where a counter-forking victim produces deterrence. Framed that way, the paper should be read as a conservative, worst-case analysis rather than a prediction of the full contest.
On Budish specifically: I read that paper closely and found it interesting enough that I actually drafted a rebuttal at one point, though I never submitted it. That said, for personal reasons I do not expect to write another paper on this model, so modeling the full capacity-bidding game will have to remain future work for someone else to pick up. I hope the explicit scoping above is a fair middle ground.
Thanks again for engaging with the work so carefully.
Best regards,
Junhyuk