Can Trump win in November?

Based on the primary elections to date, politicos and pundits are already anointing Donald Trump as the presumptive Republican nominee for president. This begs the question: What will he have to do to win in November?

Given the sharp racial divide between Trump’s Republican primary support among whites and Hillary Clinton’s strong support among blacks and other racial minorities, I was facetiously asked by a colleague: How much white support would Trump need if he got no support from minorities? Assuming that whites made up 72 percent of voters, I calculated that if Trump got zero percent of minority votes he would need 70 percent of whites to win the national popular vote. This would be well above Mitt Romney’s 59 percent in 2012, or Ronald Reagan’s watershed 66 percent in 1984. Of course, this is an unrealistic exercise since Trump will get votes from racial minorities, and because winning the Electoral College reflects the clustering and dispersal of minority and white voters across states.

Fortunately more sophisticated models exist based on both demography and geography. One of these is a new study, America’s Future Electorate that I authored with Ruy Teixeira and Robert Griffin as part of a collaboration between Brookings, the Center for American Progress and the American Enterprise Institute. In it, we simulated the 2016 election under several different scenarios based on the projected race and age composition of the electorates in the 50 states and D.C. The simulations assume different sets of turnout and voting patterns which get applied to the updated demography.

As might be expected, scenarios which apply the turnout and Democratic/Republican voting patterns from the 2008 and 2012 Obama victory elections, yield similar Democratic Electoral College wins in 2016. But when we instead apply the turnout and voting patterns from the 2004 election, when George W. Bush defeated John Kerry, the simulation produces a 2016 Democratic popular vote win but a Republican Electoral College win, though smaller than in 2004. In other words, the changing demography (read: increased racial diversity) of the electorate affected the outcome in just 12 years.

The fact that a 2004 voting scenario could yield a slight Republican win in November does not necessarily bode well for Trump. That 2004 scenario built in markedly lower voter turnout and Democratic voting preferences for blacks, Asians and Hispanics than occurred in the last two presidential elections. While it might be argued that minority support could be more tepid for a potential Hillary Clinton candidacy than it was for Barack Obama, the opposite argument could be made if her opponent is Trump. In fact, negative minority pushback against both a Trump candidacy and the fact that a Republican president would nominate the next Supreme Court justice might increase minority enthusiasm for Democrats especially among Hispanics who still have a high ceiling for voter turnout. The latter could steer swing states like Florida, Virginia, and Colorado- those that voted Republican in 2004- to the Democrats.

The most favorable Republican scenario for 2016 that the study produced assumes the same minority voting patterns as 2012, but increases white Republican support by 5 percentage points in all states—almost the levels in Reagan’s 1984 re-election. This simulated increase in white support does produce a substantial Republican win not only in Sun Belt states but also in Northern states like Ohio, Pennsylvania, Iowa, Wisconsin, and Minnesota. This would seem an unrealistic scenario. A huge rise in white Republican support, occurring uniformly across the country, will probably not occur given the wide ideological and age spectrum of white voters.

However it does paint a picture of a possible game plan for Trump. That is, the white support for Republicans in those older, whiter Northern states has traditionally been lower than in much of the South and Great Plains—creating a higher ceiling especially in an atmosphere of white blue collar anger. If there is a possible path for his victory, it might be to elevate white support in those states, particularly Ohio, where recent elections have been close.

On the other hand, the demography of diversity could well work in favor of the Democratic candidate in states like Florida, Virginia, North Carolina, Nevada, and Colorado. If that occurs, each party’s candidates will be trying to make gains in the other’s traditional strongholds.

Win or lose, Trump’s possible candidacy could make the 2016 election a pivotal one with respect to the divide in white and minority politics. According to the study’s long term projections, the nation’s demography will become increasingly favorable to the Democrats if recent voting patterns by race and age persist. This should lead both parties to reassess their long term strategies toward cultivating new bases of support.