I teach Health Economics starting today, and every year I start with data assuring students that learning data will not change their health policy opinions.
He cites a number of studies purportedly showing little relationship between experts' positive and normative views, and concludes:
While most economists think that the facts they spend years studying influence their policy positions, most policy opinions are apparently determined almost entirely by values. Since it is obvious that facts are very relevant, this all makes me ashamed to be ... human.
Robin and I discussed this over lunch. It turns out that his summary of the literature is at best misleading. By Robin's own admission, many studies DO find the expected correlation between expert's (and non-experts!) beliefs about how the world works and their beliefs about what policies we should have. Robin is actually making the subtler point that after controlling for core values, positive beliefs do not have any additional ability to predict policy views.
The upshot: It's quite possible that Robin will change his student's policy views after all. But if he does, he will probably change their values in the process. If this seems unlikely, consider: Can studying health economics undermine the value judgment that "You shouldn't put a price on human life?" In my experience, it can and does.