In the run-up to the 2016 campaign, Republicans were told that they had to adjust to demographic change if they were to have a future. They even said it to themselves.
The Republican National Committee produced an "autopsy" after Mitt Romney's 2012 defeat urging the party, among other things, "to embrace and champion comprehensive immigration reform" or be written off by the country's growing Hispanic population.
Demography was destiny - until it wasn't. House Republicans declined to pass a big immigration bill, and Donald Trump ran a campaign that sometimes seemed designed to alienate rather than court Hispanics. Yet Trump not only won the election, he slightly improved on Romney's share of Hispanic voters who chose a major-party candidate. (Asian Americans were the only racial group that moved toward the Democrats in 2016.)
My sense is that since then, the folk wisdom of Democrats on this subject is that the previous consensus was right but premature. History is still on their side; it is just taking a while to prove it. Among Republicans there is still concern about adverse demographic trends. But the example of 2016 has also increased the number of Republicans who discount or ignore such trends.
Four think tanks have an ongoing project, called States of Change, that brings data to bear on these questions. Its researchers are serious data crunchers (although they let this history major serve on their advisory board). In a new report, Democratic strategist Anna Greenberg and conservative-leaning election commentator Sean Trende offer separate analyses based on the best available data.
The project allows the researchers to run different simulations of future presidential elections based on assumptions about turnout and voting behavior for different groups. For example: If whites with college degrees, whites without college degrees, Hispanics, African-Americans and Asian-Americans vote at the same rates and for the same parties they did in 2016, the change in their share of the population would lead to a Democratic victory in 2020.
All else equal, whites' falling share of the population will help Democrats and hurt Republicans. But Republicans can counteract that effect if, for example, they continually win somewhat higher shares of the many white voters without college degrees.
Several of the scenarios in which Republicans win the presidency resemble 2000 and 2016: They're electoral victories earned with fewer popular votes than the Democrats are projected to get. That might erode the legitimacy of the electoral system. It might also reduce Democratic turnout, as liberal-leaning voters decide their votes do not matter.
The analyses by Greenberg and Trende raise at least three issues worth tracking.
First: What percentage of people of Hispanic descent are going to define themselves, in the future, as Hispanic? Trende raises the possibility that "intermarriage and the shedding of ethnic identity" will affect Hispanics as it previously affected Italian-Americans and other groups, especially as "Hispanic immigration rates level off."
Second: Will Democrats find a way to replicate President Barack Obama's performance among African Americans? Between 2012 and 2016, black turnout dropped a little and their Democratic vote share dropped a lot, from 93.7 to 88.2 percent. The decreased vote share was especially devastating for Democrats, as Greenberg notes: If Hillary Clinton had gotten 93.7 percent too, even with the lower turnout, she would have tied Trump in the Electoral College. Black votes matter. Trende suggests that black voting patterns are returning to their pre-Obama norm, and Democrats will be hard-pressed to regenerate the excitement that greeted the first black president.
Third: How long will the leftward inclinations of young voters last? The generation gap in voting has widened in the elections of the 21st century, although in the last election young voters moved right in some states and left in others. Will age, and the marriages and mortgages that accompany them, pull these voters toward the Republicans?
In his analysis, Trende returns to a theme of his very good book "The Lost Majority": Seemingly inexorable trends can suddenly reverse. In 1925, it would have been natural to assume that the South would stay Democratic, blacks would stay Republican, and the working class was moving toward the GOP. All of those assumptions were upended, some of them within a few years.
There's probably something that even the smartest political minds take for granted that will soon turn out to be untrue.