Last month, I wrote about a poll that found Sen. Joe Manchin to be a shoe-in for reelection if he votes to confirm Brett Kavanaugh, but in for an extremely close race if he doesn’t. According to the survey, taken by the Trafalgar Group, West Virginia voters favor Manchin over Republican Patrick Morrisey by 29 points if Manchin votes for confirmation, but by only two points (within the margin of error) if he votes the other way.
I expressed skepticism that Manchin’s vote on Judge Kavanaugh will have as much impact as the poll suggested. However, it was easy to accept the view that Manchin’s reelection bid will be better off if he votes to confirm the nominee.
But now, a poll of the Indiana Senate race finds that incumbent Democrat Joe Donnelly improves his chances of re-election if he votes against confirming Kavanaugh. The survey was conducted by Trafalgar, the same outfit that polled the corresponding question in West Virginia.
In Indiana, according to the poll, Donnelly leads by seven points in a scenario that assumes he votes against confirming Kavanaugh. In a scenario assuming he votes to confirm, the lead in only one point. ( FiveThirtyEight points out that, oddly, Donnelly’s lead is 12 points before the Kavanaugh nomination is raised with respondents).
How to explain the differing results Trafalgar obtained in these two states? For one thing, Trump is more popular in West Virginia than he is in Indiana, though he’s popular in both states. It may also be that advocacy advertising regarding Kavanaugh has followed differing patterns in the two states.
My sense is that the question of how voters will respond to a “yes” and a “no” vote on Kavanaugh is too hypothetical at this point to yield reliable information. I think Senate Democrats who plan to base their vote on the nomination on such a calculation should trust their political instincts and not rely on polls.