Apropos the debate we've been having regarding how more facts would affect political views, Tapped links to some interesting polling which sheds light on the question at least with regards to health care.
My thesis on these issues is that facts can persuade when the underlying dispute is factual, but not when it is based on values. So if American opposition to the Obama health care plan is based off a principled objection to government intervention in health care, then more facts won't change that value assessment. By contrast, if opposition is premised off factual miscues about what is in the bill, then providing the truth would result in some changes.
The poll in question first starts by simply asking respondents whether they support or oppose the Obama plan, and come out with a 40/48 split against. They then poll the individual components of the plan, finding that Americans like pretty much all of it, except the parts where they have to pay for it (Californians empathize). Finally, the poll informs the respondents that the previously named components are the Obama plan, and ask them again to register support or opposition. The results in this second round are a 48/43 split in favor -- not overwhelming, but still a pretty substantial 13 point swing (as TAPPED notes, this might understate things given the reticence some people might have to effectively admitting that their prior opposition was based off a lack of information and their own fickleness).
Consequently, it seems apparent that some, though certainly not all, opposition to the Obama plan is traceable to factual error. More importantly, the poll here indicates that this matters at the margin -- that is, providing the facts makes the difference between plurality opposition and plurality support.