There is a lot of information in the cross tabs in this poll from McClatchy-Marist.   The sample size is small, so the margin of error in the cross tabs is high. That said, it is still a fascinating look into voter motivations.

I was especially intrigued by the question

Did you choose Donald Trump because you are for him or because you are against Hillary Clinton?

Page 10 contains the breakdown for Trump voters. In every demographic breakdown and sometimes by a factor of two to one, those polled who said they were voting for Donald Trump also said their motivation was to vote against Hillary Clinton.

Page 9 is the same question asked of Hillary voters.  In most categories, those who were polled said theirs was a positive vote for Clinton rather than a vote against Trump.   The four exceptions were:

This suggests there is a huge divide within Whites who are not college graduates, as 59% of those in that class who are voting for Trump report it as a vote against Clinton.  The same can be said of Independents. In that category, 68% of the vote for Trump is anti Clinton. Ditto for Soft Republican affiliation, which is 64% anti Clinton.  It appears there may not have been enough respondents in the 18-29 demo-graphic who were voting for Trump for a statistically meaningful breakdown on motivation. Among those voting for Trump, MClatchy-Marist only provides two age classes- under 45 and 45 and older.

The antipathy of younger voters for Trump is confirmed on Page 11, the four candidate race broken down by demographics.  Note that among voters under 30, Trump ran in dead last place, with only 9% support.

Note, too, that only 2% of Black voters in the sample supported Trump.  That 2% number is consistent with other national polls. Last week Trump supporters were talking up a poll that shows Trump running strongly among Black voters in North Carolina.  May I suggest that there are polls out there that are nothing more than click bait for those who sponsored it.   An organization sends out a press release reporting a startling poll result. This gets picked up by the media and traffic to that organization's website goes up. Note that I am familiar with the organization that released the above poll. They have burned me twice with polls that showed the Republican challenger neck and neck with the Democrat. Then on Election Day the Republican was crushed.  In each case, my money could have gone to a race where it might have made a difference had I not relied on their poll.