Democrats learned painfully in 2016 the meaning of that question, when Donald Trump upended the electoral map by capitalizing on the demographics of today as Democrats pinned their hopes on demographics of tomorrow.
What does the future hold? In the era of President Trump, no predictions are reliable. There are only scenarios and simulations, and a new set of them shows an uncertain future for both parties --- in 2020 and beyond. There's more for Democrats to hang on to, given some of the changes underway in the country, but also enough for Republicans to seize upon, as Trump did, to see paths to victory in future presidential campaigns.
Over time, Republicans probably will have to adapt more than Democrats to avoid bigger presidential defeats. For now, the country's demographics and shifting voting patterns present trade-offs for both parties, as they try to reach out for support from constituencies that have eluded them while holding to their core groups.
These conclusions come from a new report, the fourth in an annual series called States of Change, by Robert Griffin, Ruy Teixeira and William H. Frey. Sponsored by the Bipartisan Policy Center, the Brookings Institution, the Center for American Progress and the Public Religion Research Institute, the report is being released publicly this weekend. An advance copy was made available to The Washington Post.
The authors ran a series of simulations for elections between 2020 and 2036, using different assumptions about the shape of the electorate, while also trying to estimate how tweaks or shifts in levels of support for Republican or Democratic candidates would affect the popular vote in the states and, therefore, the electoral college and the national totals.
One conclusion is that the country should be braced for repeats of what has happened twice in the past five presidential campaigns --- a popular vote outcome different from the electoral college result. "This report finds that quite a few future scenarios could mimic the result of the 2016 election --- a Democratic win in the popular vote with a Republican win in the Electoral College," the authors write. (The same thing happened in 2000.)
The authors suggest that three demographic changes will have the most significant impact on future presidential elections. The first is the country's continuing diversification, which means a decline in the percentage of white voters in the electorate and a rise in the percentage of nonwhite voters. Whites constituted 69 percent of eligible voters in 2016, and projections indicate that will decline to 67 percent in 2020 and 59 percent by 2036.
That obviously means problems for a Republican Party that has been heavily dependent on white voters in recent elections. But what ultimately counts is the makeup of the electorate states, and there Republicans will continue to have opportunities in some traditional battlegrounds. "By 2036, Iowa's EVs [eligible voters] will still be 84 percent white; Wisconsin's will be 80 percent white; and Ohio's will be 77 percent white," the study notes.
The second change is the aging of the population. Americans 65 and older made up 21 percent of eligible voters in 2016. That will tick up to 22 percent in 2020 and 27 percent in 2036. Virtually all of that growth will be as a result of more people of color in the older population.
The third change factored into the simulations is education levels, which became a much starker dividing line in the electorate in 2016. White voters without college degrees strongly backed Trump, but whites with college degrees were far less supportive of him. Whites without college degrees constituted 46 percent of eligible voters in 2016, according to the study. Their share will decline to 44 percent in 2020 and an estimated 37 percent in 2036.
Those are the main demographic variables shaping future elections. Beyond that are the voting patterns: how many voters from one group or another show up on Election Day, and at what levels of support for the candidates on the ballot.
The electorate changed between 2012 and 2016, not just because of changes in the country's demographics but also because of levels of enthusiasm among particular voters. For example, in some places, Hillary Clinton suffered from costly downturns in votes from African-Americans compared with Barack Obama in 2012.
Taking all the variables, the authors simulated future presidential elections under different assumptions of turnout and support. If the 2020 election were to see turnout and support levels equivalent to those in 2012, the Democratic nominee would win the popular vote by about six percentage points and would garner 332 electoral votes - identical to Obama in 2012.
Two 2020 simulations were run based on 2016 statistics: one with support and turnout levels the same as they were then, the other by lowering the third-party vote closer to historical norms and reallocating it. Under the first scenario, Democrats win the popular vote and the electoral college. Under the second, Democrats win the popular vote, but there is a tie in the electoral college.
Looking forward, both parties' nominees must decide which voters to target most. Can Democrats win a greater share of white working-class voters, for example, or would Trump, assuming he runs again in 2020, do even better with that group? Can Republicans substantially increase their share of the Latino vote? And can either party successfully increase its share of one group without suffering erosion with another?
Given future population projections, Democrats have a good chance of winning the popular vote in 2020 under most of the many scenarios considered. "In only two cases do the authors actually see a Republican popular vote victory in 2020," they write. One would involve a swing of 10 points toward the GOP among white non-college-educated voters. The other would be a swing of 10 points toward the GOP among white college-educated voters and a lowering of third party votes with it.
But there are several possibilities examined by the authors in which the Republicans could lose the popular vote and still win an electoral college majority. One, not out of the question, would be a small (five-point) swing toward the GOP among white non-college-educated voters and a swing of equal size toward the Democrats among white college-educated voters.
"The GOP has many roads to the presidency in 2020 even though demographic shifts appear to make a Democratic popular vote victory easier than ever to obtain," the authors conclude. "Even more interesting, some of these fruitful scenarios continue to produce Republican electoral vote triumphs in 2024 and beyond, despite mounting popular vote losses."
The authors note that they are not in the prediction game. After what took place in 2016, that's a prudent course. As their study makes clear, there are enough variables at play to make predictions foolhardy. Which is why they write that "neither party can be assured of long-term dominance simply from shifting demographics."