One of the big unresolved arguments from 2014 involved the theory that pro-Democratic groups didn’t just represent a declining share of the electorate in the midterms, but were also showing signs of trending Republican. I wrote about some new evidence from Pew on this subject for Washington Monthly:
 exit polls showed Republicans doing a bit better among millennials, Latinos and even African-Americans, and a lot better among Asians, than in 2012. Similar findings in 2010 led some analysts to conclude that the Obama Coalition only existed when Obama was at the top of the ballot–not a particularly good omen for Democrats looking ahead to 2016. But other analysts argued the logic of figuring that even as midterm elections produced conservative-skewing turnout patterns across demographic lines, they probably do so within demographic groups. Thus, more conservative millennials, Latinos, African-Americans and Asians show up, at least at the margins, and Republicans get a bonus beyond the overall reshaping of the electorate.
There’s some strong if circumstantial evidence of the latter hypothesis in a big new survey from Pew that shows major demographic groups falling more or less into the same partisan preferences (particularly once the hordes of self-identified but bogus indies are pushed to “lean” one way or another) now as in 2012. This includes Millennials, who tilt Democratic by a 51/35 margin; Latinos, Democratic by a 56/26 margin; and Asians (65/23 Democratic after narrowly–at least according to the exit polls–going Republican in 2014).
These numbers should give triumphalist Republicans some pause at a minimum, while reinforcing the belief of many–myself included–that we are in an era of oscillating elections based on different midterm and presidential turnout patterns rather than some sort of steady trend towards the GOP.
These numbers are also a data point against the presumption that the “Obama Coalition” won’t be there for the next Democratic nominee. Indeed, about the only thing placing a thumb on the scales for the GOP in 2016 is the presumption of “voter fatigue” with the two-term Democratic control of the White House. That’s a presumption that relies on a very small sample with mixed results.