I strongly recommend John Sides’s post, and for that matter I also strongly recommend On Deaf Ears and the Strategic Presidency, the most essential readings about the presidency this side of Nuestadt and Skowronek.
One thing I should add is that — while public opinion is something that quantitative methods are especially well-positioned to evaluate — I would never say that a behavioral social science study is the God’s truth. I definitely think that careful case studies that show what seems to be the general trends of the data not applying are important. The impossibility of definitively testing historical counterfactuals and isolating causal relationships makes even the best studies potentially limited. But one reason I believe Edwards is correct is that he not only has the data to support his hypotheses, he has a much more convincing theory than the Cult of the Bully Pulpit people. Most importantly, people who believe in the power of messaging to achieve short term results have never dealt convincingly with the fact that 1)most people don’t pay attention to presidential rhetoric, and 2)the people that do are generally high-information voters with strongly entrenched partisan and ideological commitments. Arguments in favor of the power of the bully pulpit, conversely, in my experience tend to involve pundit’s fallacies, unconvincing ad hoc explanations for obviously disconfirming cases, and other signs of a bad argument.
I haven’t read the Canes-Wrone book, but for similar reasons I find the idea that the appropriation process is an exception to the general rule very plausible. Whether it’s actually rhetoric or messaging per se doing the work I don’t know — perhaps she has convincing data on this point — but certainly a president can be expected to have more leverage in a context where legislation has to be passed. With something like the ACA (or Social Security privatization or whatever), the president’s leverage is inherently constrained by the fact that Congress can just walk away from the table.
Which, speaking of theory and counterfactuals concerning the ACA, is the heart of the issue. The next person who can explain what leverage — via messaging or anything else — that Obama had over Evan Bayh, a greasy conservative not running in a state where Obama isn’t especially popular anyway and beholden to corporate interests for his future career, will be the first. The Green Lantern position seems to be that it’s massive failure of available powers that Bayh, Nelson, Lincoln, Lieberman et al couldn’t be made to support a much more progressive bill. My position (which is ironically sometime portrayed as Democratic apologism) is that it’s increasingly amazing in retrospect that Reid and Obama got these people to vote for anything.