T. Jefferson wondered:
. . .
Maybe. But let me play devil's advocate: Someone should when everyone starts agreeing, right?
Suppose for instance that there is some third thing--call it "enemy relative strength"--that contributes (negatively) to both our probability of battle success and the enemy's probability of negotiating with us. Suppose also that enemy relative strength has (at least some) unique causal factors that are not simply due to past battle successes or negotiations (that is, it has at least in part its own independent dynamics). Under these circumstances, independent dynamic changes in enemy relative strength will generate a positive correlation between our battlefield successes and negotiation with the enemy, even if there is no direct causal link between the former and the latter. We or for that matter the historian would observe a relationship between the two, but it wouldn't be because one caused the other. Rather there would be an independent underlying third thing driving both.
Don't get me wrong; I suspect Ken White and Jones_re are right in their interpretation of such evidence. But what "story" would make it right? The obvious theoretical story--I'll bet the one that Ken or Jones_re have in mind, and that I would too--would be that enemy decision makers only know their own relative strength imperfectly, that battlefield failures cause them to revise their own estimates downward and that this then causes them to have a higher probability of negotiating. But that is just a theory: The mere fact of a correlation between enemy losses and willingness to negotiate isn't clear evidence about that theory.
Sorry for being a pain-in-the-butt empirical inference wonk. I can't help it.
Bookmarks