There has been a lot of reporting of how Trump has not lost much support of those who voted for him.
But do we know how reliable the self-reporting of prior voting behavior is. I wonder if anyone has been checking if the self-reporting of Trump voting is consistent with the actual votes.
We know hat Trump voters are low education, we know that they are often hard to get to answer polls, and we suspect that they may be less prone to tell the truth (at least to pollsters.)
So there may be lots of hidden dissatisfied Trump voters out there. I remember the moment during Watergate when Nixon’s reputation became so poisonous that more people remembered that they had voted against him than for him.
If this theory is right, it holds particular danger for Republlicans facing reelection in 2018, because it will make it harder to track the Trump voters whose reason for turning against a Republican incumbent is his or her support for Trump.
Worth keeping an eye on.