Could you separate the difference between polling and voting on specific issues from the difference between polling on specific issues and turning out for party politicians?
I’m not sure it’s possible to do so entirely, if only because the party a politician joins can itself be information about the sort of matters they’ll be able to put forward. You can almost always find better proxies than the first reported by media sources, though. Simple averaging together the various individual components of health care reform is a really stupid and obviously inaccurate tool—people don’t value individual components equally—but it shows a far more interesting picture of the full system.
If you start with an assumption of massive electoral dysfunction, that can explain a pretty large number of things—but it seems to have explanatory power, rather than predictive power.
I’m not sure it’s possible to do so entirely, if only because the party a politician joins can itself be information about the sort of matters they’ll be able to put forward. You can almost always find better proxies than the first reported by media sources, though. Simple averaging together the various individual components of health care reform is a really stupid and obviously inaccurate tool—people don’t value individual components equally—but it shows a far more interesting picture of the full system.
And you probably have other complicated variables to deal with. People don’t generally know the details of any specific law. Again, using health-care reform since has some of the best research, there’s pretty clear evidence that one in four people have never been aware of even the most popular parts of the law.
If you start with an assumption of massive electoral dysfunction, that can explain a pretty large number of things—but it seems to have explanatory power, rather than predictive power.