Voters cast their ballots and wait in line at Monelison Middle School on Nov. 8, in Madison Heights, Va. (Jay Westcott /News & Daily Advance via AP)

In the aftermath of the presidential election, the Democratic Party has been in the midst of a heated debate over “identity politics,” ignited partly by an op-ed by the Columbia University intellectual historian Mark Lilla.

Lilla and his supporters argue that Hillary Clinton’s campaign focused too much on reaching out to specific groups, especially minorities, instead of developing a broader message that could have also appealed to the white working class. Others view Lilla’s argument as a call to abandon the party’s support for minorities in a society that remains rife with bias based on race, gender, religion, and so on.

Underlying this debate is a problem of electoral math. What would it take for the Democrats to achieve a governing majority at the national level, winning not just the presidency (where they won the popular vote this year) but also control of Congress (where Republicans won the popular vote)?

Despite winning a declining share of the white vote in recent elections, Democrats have remained competitive by relying on large majorities of minority voters. And in a diversifying America, where the share of the white population is falling, many progressives believe that it is only a matter of time before demographic change leads to a natural Democratic political majority.

Why demography isn’t necessarily destiny

But this hope of an inexorable Democratic tide based on demography could easily be wrong, for at least two reasons.

First, the forecast of a majority-minority society in the near future is more problematic than has been commonly assumed.

Second, the progressive hope relies on precarious assumptions about the future behavior of minority voters. In particular, recent patterns of assimilation, especially among U.S.-born Asians and Latinos, complicate the white/minority division of the population.  These patterns also suggest that assimilating voters may behave politically more like whites than the conventional wisdom allows.

Without winning a larger share of the white vote, optimism about future Democratic political strength may be premature.

Let’s do the math

To understand the Democratic predicament, start with some simple arithmetic: Democrats have lost the white vote by an almost 60-40 margin in the last two presidential elections. In 2016, whites made up 70 percent of the electorate. Suppose for the sake of argument that they constitute two-thirds in 2020. Unless the Democrats’ margin changes, white voters would give them barely more than a quarter of all votes cast.

To get to a safe majority then — one that would likely ensure an electoral college victory and a majority of seats in the House — Democrats would need an overwhelming number of minority votes. To get to just 53 percent nationally — hardly a landslide — they would need to win 80 percent of the nonwhite vote. Since the 2016 exit polls showed that only two-thirds of Asians and Latinos voted for Clinton,  that may be a challenge (though the exit polls may also underestimate Democratic support among minorities).

But things get better for the Democrats the further we go into the future, right? Not necessarily. When one looks behind seemingly objective demographic facts and examines the social processes (and Census categorization) shaping the data, the certainty of a Democratic majority becomes more tenuous.

So what’s wrong with census forecasts of fewer whites?

To begin with, the census data that these forecasts are based on exaggerate the extent of white demographic decline; even the prediction of a majority-minority society is not guaranteed. The reason lies in the census misclassifications of a fast-growing group of young Americans from ethno-racially mixed backgrounds.

Currently, 14 to 15 percent of infants born in the United States are multiethnic or multiracial, a number that was just 11 to 12 percent in 2000. But despite the fact that most of those children have a white parent, inadequacies in the census classifications mean that the great majority of them are identified as nonwhites.

This is important, because most partly white individuals behave like whites in sociological terms. They grow up in neighborhoods with many whites, have white friends as adults, think of themselves mostly as white or partly white, and marry whites. We don’t know yet whether they vote like sociologically similar whites, but it is quite plausible that they will. We will likely find out in the near future, because young people from mixed backgrounds are numerous in the cohorts coming of voting age in the next few decades.

Individuals of mixed background are the leading edge of assimilation, which also includes many socially mobile and intermarrying minorities. Assimilation does not have as large an effect as it did for the white ethnics in the post-World War II period, but it happens often enough to carry political weight. For example, the top tiers of the U.S. workforce are increasingly inhabited by nonwhites, especially Asians and second- and third-generation Latinos.

The Democratic vision of an inexorable demographic tide assumes that their votes in the future will be determined primarily by their minority status — rather than by, say, their economic position or their family relationships to whites.

But there is no reason to presume that what happened to Catholic ethnics decades ago as a result of assimilation — they became the “Reagan Democrats” — won’t repeat itself. This may be especially true for light-skinned Latinos, who, like Catholic ethnics, come from cultural milieus where socially conservative values dominate. If this happens, it does not mean that assimilating Latinos will uniformly vote Republican — the assimilating Catholic ethnics didn’t, either — but they may vote more like the whites to whom they are otherwise similar. That would still be bad news for the Democrats.

The current Democratic strategy, founded on the so-called Obama coalition, recalls in key respects the late 20th century United States. In it, Americans are segmented into clearly defined ethno-racial blocs, with the white one dominant and holding power partly through systematic racism.

In the early 21st century, systematic racism is still powerful, but other patterns — of group-melding and assimilation — are becoming important as well. Precisely because whites remain the dominant group, assimilation usually requires degrees of social integration with them, in effect expanding the white majority.

In their strategic thinking, Democrats need to take account of these consequential new realities.

Richard Alba is Distinguished Professor of Sociology at the Graduate Center, CUNY.

To learn more: