Part of the ugly reality of politics is that there are multiple ways that people identify themselves — parental status, marital status, gender, sexual orientation, religion, age, race, education. In a more perfect world, some of these forms of self-identification would have no political meaning. However, in today’s world, they still do. One reason good pollsters reweight their samples by things like race and gender is because different races and genders tend to vote differently.
In light of this ugly fact, politicians trying to win election have three basic options for dealing with political demographics as they exist: 1) try to change what a demographic group thinks (persuasion); 2) try to increase turnout among groups that favor you (get out the vote); and 3) try to decrease turnout among groups that do not favor you (vote suppression). The latter issue has been in the news again recently with several cases dealing with state voter ID laws, and that will be a topic for another post sometime within the next week. This post is more about where the demographics stand now and what they might mean for the next three months of the campaign.
Earlier this year, the folks at 538 posted a “swing-o-meter” for this election based on demographics. The basic concept was looking at five key demographics — whites with college degrees, whites without college degrees, African-Americans, Hispanics, and Asian-Americans — and seeing how changes in voting preferences and turnout would potentially alter the electoral college vote. For the 2012 numbers, whites with degrees went with Romney 56%-44% and had a 77% turnout; whites without degrees went for Romney 62%-38% and had a 57% turnout; African-Americans went for Obama 93%-7% and had a 66% turnout; Hispanics went for Obama 71%-29% and had a 49% turnout; Asian-Americans went for Obama 67%-33% and had a 49% turnout.
There have been three big trends over the Obama years. First, Hispanics and Asians have been trending Democratic. Second, African-Americans have had record high turnouts in presidential years. Third, having a college degree has been a significant demographic issue for white voters with those with college degrees becoming more Democratic and those without becoming more Republican.
What we have seen from this campaign so far indicates that the first trend is continuing. Just twelve years removed from George Bush being competitive for Hispanic votes, there is every reason to expect that Trump will do even worse than Romney — potentially not breaking 20%. The problem is that Hispanics tend to live in states that are not in play. That result would barely flip North Carolina and no other state. However, there are reports of increased registration among Hispanics and Hispanics eagerly trying to become citizens. If the Democrats could bump Hispanic turnout up to 60% (in addition to taking 80% of the Hispanic vote), that might also put Arizona in play. The same issue applies to the Asian vote. While Donald Trump is doing everything possible to drive Asians to the Democratic party and boost Asian turnout, there really are not that many battleground states where Asians matter.
The big risk for Democrats is with African-American voters. It is unlikely that Trump will do much better than Romney with African-Americans in terms of percentage. Obama in 2012 only did a little bit better than Kerry did in 2004. The big difference was in turnout. Low black turnout might put Ohio, Virginia, and Florida in play.
The big issue will be white voters who are still 70% of the total vote, split almost evenly between those with degrees and those without. Early polling suggests that Clinton might be the first Democrat in half a century to win whites with a college degree, mainly due to women with a college degree. The Republican message has been heard loud and clear by women with college degrees and they do not like it. A slim majority among whites with college degrees would give the Democrats Georgia and North Carolina. However, if instead of gain among whites with college degrees, the Democrats lost 6% among those without degrees and turnout increased to 62%, Republicans would pick up Colorado, Florida, Ohio, Wisconsin, Iowa, and New Hampshire and win the election.
In short, this election will come down to what it always does — getting votes registered and getting them to vote. If you know a college student, make sure they are registered and no what to do to get an absentee ballot. Explain to them why they can’t sit this election out. Otherwise, participate in local voter registration drives and start work on the get out the vote effort now. As a general matter, demographics are trending our way, but — as the mid-year elections demonstrate — favorable demographics mean nothing if your voters don’t show up at the polls. Democrats should be able to hold the states we won last time and pick up Arizona and North Carolina (along with sending John McCain and several Republican Senators that won blue states in 2010 off to retirement), but only if we work to get the vote on November 8.
