There’s little doubt that Donald Trump has changed the rules of campaigning in the 21st century. Or, more accurately, he’s the first to understand how to effectively campaign in a 21st century media environment. Yet, winning the media cycle is different from winning a campaign. And, as my colleagues and I have written about extensively, that requires him to do better with the electorate that existed in 2012, or to change the composition of the electorate in 2016 to his benefit. Thus far, it doesn’t look as if he’s doing either.
First, let’s take a look at some of the most recent national public polling. The May NBC/Wall Street Journal poll found Trump winning white non-college voters by 27 points and tied with Clinton among those with a college degree. If those numbers hold until November, that’s very good news for Clinton. Romney carried non-college white voters in 2012 by 24 points and carried college educated white voters by 12 points. In other words, Trump is doing just three points better among non-college white voters, but 12 points worse among college educated voters.
The Washington Post’s Scott Clement found similar results in their own polling data, with Trump running four points ahead of Romney’s 2012 showing with white non-college voters and 10 points behind Romney with white college voters. The newest Quinnipiac national poll found Trump ahead 17 points among whites, three points lower than Romney’s overall showing in 2012. That Quinnipiac poll also showed Clinton running about two points better than Obama’s 2012 showing among African-Americans and Latinos.
So, what do these numbers mean for the overall Electoral College map? For some in help in that department, I plugged these numbers into the handy-dandy “Swing-O-Matic” tool created by my colleague David Wasserman and Aaron Bycoffe of fivethirtyeight.com. This model allows you to see how shifts in party preference and turnout by different demographic groups would affect the 2016 election. If we assume that Trump wins white, non-college voters by 65 percent and Clinton and Trump tie among college white voters, the electoral map would look very similar to 2012 with the exception of North Carolina which would swing to Clinton. Go ahead, plug in your own estimates and see what you get.
Of course turnout is a critical component in every election. Clinton has to be able to hit her marks among non-white voters, especially African-Americans in states like Ohio, Pennsylvania, Virginia and Florida. But, let’s suppose turnout among all other groups stays similar - Latino turnout remains at 50 percent and African-American at 66 percent - but assume a higher than normal turnout among white, non-college voters. Even if I adjust turn-out to 74 percent - a whopping 16 point increase in turn-out from 2012, and give Trump 65 percent of that vote, while moving the college-white vote to even, the electoral map would look identical to 2012 - 332 for the Democrat to 206 for the Republican. In order for Trump to win with white working class voters alone, he’d need to get at least 70 percent of their votes. And, at this point, there’s no empirical evidence that I’ve seen that he’s doing that.
Qualitative and survey research among working class white voters - defined as adult white voters who have less than a four-year college education - conducted by Jill Normington of Normington Petts & Associates and Pete Brodnitz of Expedition Strategies for a Democratic-based independent expenditure group, found Trump’s support among these voters to be “comparable to Romney’s 2012 performance.” Obama won 36 percent of the vote among white working class voters, notes the independent expenditure source, while their data shows Clinton getting 35 percent.
This polling is consistent with other private data out I have been privy to that shows Trump consolidating the GOP faithful, but not expanding his support into traditional Democratic constituencies of white working class voters.
In other words, Trump isn’t doing better than Romney among white, non-college voters, the one group that everyone concedes is his strongest constituency. Moreover, the more he doubles down on his rhetoric that alienates non-white and college educated women, the harder it will be for him to make up for his vote deficit on white non-college men alone. Or, as the Post’s Clement puts it: “Trump needs to increase his overall share of white voters if he is unable to outperform Romney's support among African American and Hispanic voters. The more divided the white vote is — whether on gender or education lines — the more problematic it is for his chances.”