Block That Inference

George Stephanopolous reads exit polls:

We did ask a question I know in the exit polls about Reverend Wright, Barack Obama’s former pastor and whether that was influencing voters. What did we find? Right down the middle. About half said it’s important, about half said it was unimportant. Of those who said it was important, look at this in Indiana, 70% went for Senator Clinton. Of those who thought it was unimportant, again right down the middle, 65% for Barack Obama. So what you thought about the importance of Reverend Wright basically determined your vote.

As Lee Siegelman points out the causal inference here is all wrong. Much more likely is that voters already committed to Clinton -- or strongly predisposed to commit to Clinton -- adopted the view that Reverend Wright was an important issue because they knew it was an issue that reflected poorly on Obama. Note, for example, that both pre- and post-Wright, both Clinton and Obama took a fairly constant share of different demographic categories.

The thing is, you shouldn't need to be especially sophisticated about statistics to figure this out. Clearly, Wright may have swayed some voters, but equally clearly most people had opinions about the election before Wright ever came on the scene. But Stephanopolous is hardly alone here, almost every time I see exit polls discussed on TV it's done with almost no understanding of how to read them properly.