War and Responsibility
War and Responsibility (DOI: https://doi.org/10.1017/S0003055425000206) was written by M. Patrick Hulme in 2025. It was published in American Political Science Review.
"Since Truman’s precedential “police action” in the Korean War ... smaller uses of force are virtually always undertaken absent Congress’s formal blessing. Why is it the case that in the years since 1950 smaller uses of force have almost always been undertaken unilaterally, and yet full-scale wars have only been initiated after legally binding authorization was acquired?"
The author constructs a model in the style of blame avoiding.
On one hand, congresspeople have disparate electoral incentives and some will oppose a proposed or ongoing military operation. They can oppose it through...
- rejecting authorization
- cutting the defense budget, making the operation less feasible
- various other oversight mechanisms
There is still a cost to position taking on such an issue.
On the other hand, presidents are inextricably tied to their military operations. They face both the risk of failure (loss responsibility costs) and the risk that Congress will make failure more likely. There are always benefits to seeking authorization.
Authorization is a sort of insurance for the president, guaranteeing support from a clear majority. Congresspeople can and will evade blame. By not holding a vote on authorization, they freeride on a president's decisions, benefiting from the operation's potential successes without being on the record in the event of its failure. The author therefore concludes that unauthorized military operations are evidence of Congressional power, not presidential power.
This is explored in a game theoretic model in terms of costs (esp. casualties and blame) and some benefit.
The author then also explores how Congress informally conveys support for or opposition to a military operation, so as to constrain the president's decisions, even absent an authorization.
Reading Notes
I think this is an interesting argument, but not a convincing one.
There are two negative factors in the democratic politics of war: dissatisfaction with the war's achievements and dissatisfaction with the war itself. The former can be placed in a risk framework, the latter cannot. I think there is decent reasoning about this subject matter in the risk framework nonetheless. However, the author jumps to game theory for the models because there needs to be a cost-benefit analysis, with an unspecified public good, for anything to make sense. (Otherwise there is a dominant strategy in simply not going to war.) Also because the author clearly wants this model to fit into a larger, rationalist model.
- The cost is American casualties. The primary determinants of how many casualties are created by a military operation are the size and duration of the deployment; neither choice is exogenous of the risk evaluation. But risk is also a function of those decisions. I believe the author tried to shoehorn in the game between the president and the adversary precisely to avoid this inconvenient fact, giving a cleaner interpretation of the cost term. But it remains true that soldiers can only be killed if they are put in harms way: the president has a mechanism to put a hard ceiling on costs. It's not at all clear to me that the set of observations we have to work with is unbiased. And while there is a plausible explanation of behavior, there is no causal theory to cut through the challenges to analysis.
- Personally, I expect that the interplay of presidents and Congress here is more conventional than described. Congress is a good proxy for public sentiment; when the president lacks broad public support for military action, they plan quieter and quicker operations. If the public is clearly opposed, Congress would block anything more substantial. When the president has public support, there is also authorization.