I'm afraid to need to be a Mark Penn defender, but I think Ezra's point here, while true, is also a bit unfair. Ezra accuses Penn-the-pollster as miraculously finding the exact same shape of public opinion in every single year -- Democrats must move to the right on economics. And, indeed, that does seem to be Penn's perennial conclusion. That said, finding the same result every time he goes out to do a survey doesn't, as best I can tell, differentiate Penn that much from other big-time political pollsters. Ideology shouldn't matter in the field of public opinion research but, in practice, given pollsters tend to give remarkably consistent advice year after year.

This is actually what makes the fact that Penn is working for Hillary Clinton so significant. Penn isn't the kind of advisor you hire because you wonder what advice he's going to give you. Rather, he's the kind of advisor you hire because you know perfectly well what advice he's going to give and you've decided that's the advice you want to get. Penn is hardly unique in this regard (many people have noted that 2004-vintage John Edwards sounds more like 2008-vintage Barack Obama than he does like 2008-vintage John Edwards and that David Axelrod worked for Edwards in '04 and Obama in '08) but that's just the point -- politicians aren't naive about this stuff, they pick strategists who are going to give congenial strategic advice.