By the time you finish reading this sentence, a 17-year-old in the United States will have turned 18 and become eligible to vote. And by the time you finish reading this sentence, someone will have died. In this nation of nearly 330 million people, a young person turns 18 roughly every 8 seconds, and someone dies every 11 seconds. This dynamic and fast-paced demographic change is transforming the country’s politics – and is also the glaring analytical hole in the growing school of thought coming to be known as “popularism.”

After Democrats under-performed in the 2020 Congressional and down-ballot races, some have taken to arguing that the party paid a political price for supporting policies that were “unpopular,” resulting in disappointing electoral outcomes. The New York Times columnist Ezra Klein wrote a lengthy analysis of the phenomenon, most recently popularized by Democratic data analyst David Shor. Klein summarized the essence of popularism as the belief that Democrats “should do a lot of polling to figure out which of their views are popular and which are not popular, and then they should talk about the popular stuff and shut up about the unpopular stuff.”

In the real world of winning elections in the context of a country roiled by white racial resentment and systemic racism, the results of this week’s gubernatorial election in Virginia exposed the naivete and folly of the popularism approach. Republican Glenn Youngkin stoked the flames of white racial fear by promising to outlaw so-called Critical Race Theory (CRT), code for any educational instruction about the realities of racism and oppression in this country. CRT is a law school construct and is not taught in any pre-college schools in Virginia or anywhere else, but it nevertheless sufficiently alarmed white voters that they turned out in record numbers and propelled Youngkin to victory over Terry McAuliffe, the Democratic nominee. McAuliffe certainly didn’t campaign on CRT and did his best to ignore it. But his cautious and conflict-avoidant ostrich approach failed miserably, in part because Democratic voters were not nearly as motivated by the bland popularism approach as Republicans were by what they saw as a threat to their racial identity.

More broadly, beyond the obvious dangers of a moral slippery slope associated with such thinking as popularism (the Nazis were quite popular in 1930s Germany, for example), the analysis suffers from several fundamental mathematical flaws. Ironically, given the future-oriented image of tech-savvy Democratic data crunchers who champion the framework, the primary flaws of popularism stem from adherence to anachronistic models that are of limited use in today’s rapidly changing society.

First, popularism fails to address the question of “popular with which groups of people?” Back in the 1970s when white voters accounted for nearly 90% of all voters, standard practice was to look generically at questions of popularity without doing a deeper demographic dive. In 2020, however, the electorate is no longer so monochromatic, with the white share of the electorate having fallen to 67%. Given that nearly 90% of African Americans consistently vote Democratic, and the majority of white voters almost always take the exact opposite position, a generic race-neutral analysis misses one of the major determinative data points in American politics.

The second problem with the popularism school is its dependence on old metrics that don’t capture new realities. Historically, analysts have compared year-over-year changes in vote margins to try to identify meaningful trends in voter behavior, but such data is of only limited utility in times of extraordinary developments and can result in wildly inaccurate interpretations. It would be like a financial analyst trying to explain Delta Airlines’ $17bn drop in income in 2020 without factoring in the “black swan” event of the worldwide pandemic.

In US politics, the 2020 black swan was the highest level of voter turnout in over a century, making it an even rarer occurrence than a global pandemic. The electoral tsunami also wiped out the usefulness of the traditional metric of comparing vote margins, and popularists who have tried to analyze 2020 electoral behavior with pre-2020 analytical tools have arrived at dangerously inaccurate conclusions about last year’s results.

Many analysts looking at Democrats’ reduced margins in some heavily Latino counties in Florida and Texas have come to believe that large numbers of Latino voters abandoned Democrats in 2020 in favor of Trump. For example, in Hidalgo County – a 93% Latino county on the Texas border which Obama won by a margin of 42% in his 2012 re-election campaign (72% to 29% for Republican nominee Mitt Romney) – Trump shrunk the margin to less than 1.9%, securing 48.5% of the vote. This type of data point is a cornerstone of the popularism theory of the case, but it obscures the very inconvenient fact that the number of Latinos voting Democratic has actually increased over the past decade, even in Hidalgo County. It is difficult to square the circle of Democrats losing voters while at the same time increasing their vote totals.

Looking up from the myopic microscope of vote margins to examine the bigger and more important picture of historic voter turnout reveals the real 2020 story. Simply put, Trump generated greater enthusiasm and turnout among his previously non-voting supporters than Biden and the Democrats did. While Biden attracted 22,000 more Democratic voters to the polls in Hidalgo County than Obama did nearly a decade ago, Trump galvanized 75,000 previously non-voting Republicans. The same phenomenon also explains the Democratic losses in the House of Representatives. In the 16 seats flipped by Republicans last year, an average of 34,000 more people came out to cast ballots for the Democratic candidate than in 2018. The challenge for Democrats was that Republican votes jumped by 54,000 votes per district.

This is definitely a problem for Democrats, but the remedy is very different than what the popularists are offering. Rather than distancing themselves from issues that are unpopular with Trump supporters, Democrats need to double down on the issues that resonate with and inspire infrequent voters who are progressive.

The third flaw in the popularism school of thought is reliance on stale data sets that fail to account for the electorate’s changing demographic composition. Every year, more than 4 million people, half of them people of color, turn 18, and these new voters are already transforming US politics. Biden won Arizona by fewer than 11,000 votes, and 39,000 18-year-olds cast ballots for the first time in that election, according to the Democratic data firm Catalist. In Georgia, where Biden prevailed by a similarly small margin of 12,000 votes, 59,000 18-year-olds voted, a tribute to the massive voter mobilization machine built by Stacey Abrams and her allies over the past decade.

The diversification of the electorate is only going to accelerate in coming years, particularly in battleground states. In Texas, where Beto O’Rourke narrowly lost his 2018 senate race by 214,921 votes, more than two-thirds of those under 18 are people of color (67%), and nearly 300,000 people of color turn 18 every year; that’s equivalent to 33 new eligible voters of color in the Lone Star state every hour. O’Rourke is now considering a 2022 gubernatorial bid, and by the time ballots are cast in November 2022, 1.2 million more young Texans of color will be eligible to vote than was the case four years ago.

Not only are these younger voters in the process of upending old electoral assumptions, but they are far more progressive and bring a different set of priorities in terms of determining what is popular. They backed Biden by a 34-point margin nationally, and that level of support was replicated in the critical states of the South and Southwest – including in a 31% Democratic advantage among 18-24-year-old voters in Arizona, and 20% advantage in Texas.

On policy issues, younger voters have a very different definition of what is “popular.” A 2020 Gallup poll found that 70% of 18–34-year-olds supported “reducing the budgets of police departments and shifting the money to social programs,” more than twice the 32% level of support of those over 50. On immigration, the percentage of young people who support a pathway is twice as large as those who advocate deportation. Among those 45 and older, the immigration positions are split 50/50 according to a September Economist/YouGov poll.

Popularists are right about one thing and that is that Democrats should champion popular policies. Looking at the correct data set shows that the electorate is becoming more diverse – and more progressive – by the minute. If you’ve made it this far in this column, 25 people have now turned 18 in the US. A future-focused popularism would pay greater attention to engaging these new voters changing the composition of the electorate than to polling data about the behavior of sectors of the electorate that had more sway in the past.

• Steve Phillips is the host of Democracy in Color with Steve Phillips, a color-conscious podcast about politics, and is the author of Brown Is the New White: How the Demographic Revolution Has Created a New American Majority