Generally speaking, libertarians tend to have conservative views on issues such as taxes, government spending, regulation, guns, and affirmative action. Also, they tend to have liberal views on issues such as gay rights, abortion, marijuana legalization, and immigration.

When we examine the demographics of support for these various particular positions, there are different big patterns for different kinds of issues. On tax-and-spend issues, the conservatives tend to be folks with higher incomes, especially among whites. On guns and affirmative action, the conservatives tend to be white. On lifestyle issues (e.g., gay rights, abortion, and marijuana legalization), the liberals are less religious and better educated. On immigration, the liberals fall into two camps: Immigrants and high-education natives.

Combining those patterns, one would expect to find libertarians primarily among less-religious whites with more education and higher incomes. In recent decades in the U.S., all of these demographic features—religiosity, race, education, and income—have been on the move. Two trends increase libertarian demographics: people generally have been getting more education and going to church less. One does not: immigration and fertility patterns have lowered the percentage of the population that is non-Hispanic white. And one is complex: poverty rates haven’t declined much overall, but educated folks are getting richer.

On the whole, these trends have led to an increase in libertarian demographics, particularly among whites (where patterns of rising educations, high-end incomes, and religious decline are strongest). The chart below shows the trends in population percentages for those who combine four features: having a 4-year college degree, having an annual family income above around $72,000 in inflation-adjusted dollars, being non-Hispanic white, and not going to church weekly.

In the general population, the percentage with three or four of these features has increased from around 28% in the 1975-1984 period to around 33% in the past decade. If we focus in on the white population, they all have at least one libertarian marker (i.e., they’re white), but the percentage with three or four has increased substantially from around 32% to around 45%. (These aren’t, of course, magic numbers. There are all kinds of ways one could change these percentages by categorizing libertarian demographics in different ways—using, say, different education and income cut-offs or other measures of religiosity. The point is that any plausible set of measures along these lines will show the same thing: This combination of features is experiencing a small rise nationally and a substantial rise among whites.)

(Technical notes: Total sample size is 54,998. Results are weighted.)

Let’s put aside the issue of who calls themselves “libertarian” and focus instead on de facto libertarians, that is, people who maintain, roughly, the particular mix of conservative and liberal positions associated with the libertarian label. We have hints in demographic patterns that their numbers might be increasing.

But here’s the thing. The number of de facto libertarians has actually been decreasing. This will take some explaining.

Demographics have polarized on a libertarian/communitarian axis

Looking at the chart in terms of the total population, the libertarian demographic has expanded, but the communitarian demographic (those with no or only one of the libertarian features) has not declined. Really, there’s been a hollowing out of the middle (i.e., the grey portion of the bars). The libertarianization of white demographics has occurred simultaneously with new sources of communitarian demographics from low-education immigrants.

This has led to increasingly polarized demographics on a libertarian/communitarian axis. Blacks and Latinos have, on average, less education, less income and wealth, and are more religious. Whites are substantially more likely to be college educated, and those who are have had rising incomes over the past few decades. And while church attendance is slowing declining for everyone, attendance has long been generally lower among whites.

But politics have polarized on a liberal/conservative axis

So one might expect this libertarian/communitarian demographic polarization to have led to an increase in folks with de facto libertarian and de facto communitarian packages of political positions. But it hasn’t.

In fact, the past decades have seen a big rise in people who have either generally liberal or generally conservative packages of issue opinions. And the key demographic in this liberal/conservative political polarization has been college-educated whites, which is just where one would have expected to see the rise of de facto libertarians.

Put another way, there was a time, say, 30 or 40 or 50 years ago, when Americans didn’t tend to clump into liberals and conservatives. There were plenty of people who simultaneously held liberal views on some issues and conservative views on other issues. But it was also a time when education levels were lower, people were more religious, and, thus, the libertarian demographic was pretty small. But now that the libertarian demographic is on the rise, high-education whites have actually been increasingly abandoning de facto libertarianism in favor of more consistently liberal or conservative opinion packages.

Demographic changes have been creating space for a libertarian moment, but the simultaneous rise in left-right political polarization—particularly among the libertarian demographic—has kept it from happening.

The spectacular transition of college-educated, wealthier, non-Christian whites from de facto libertarian to outright liberal

The main story here involves college-educated, wealthier, non-Christian whites. (These non-Christians are mostly non-religious folks, but also include some Jews, Buddhists, etc.) They are among the fastest growing American demographic groups, despite having the lowest fertility of any.

This group has long had liberal views on issues relating to religion—abortion, school prayer, gay rights, and so on. But 30 or 40 years ago, they were basically de facto libertarians. Their economic views were generally as conservative as those of college-educated, wealthier, white Christians. But that has changed dramatically. (No, seriously, stare in slack-jawed awe at the black line in the chart from yesterday’s post.)

These days, college-educated non-Christians strongly tend to be liberal across a variety of issues, even on Pew’s issue-based ideology scale that contains a few economic items but only one religious item. And, for the love of God, check out college-educated atheists, where the median response is to land in liberal territory on 9 of the 10 Pew items.

If you were making political projections back in the early Reagan years, you probably would have bet that, as educations rose and religiosity fell, these folks would have been the cornerstone of a new libertarian movement. Instead, they moved to the left on economics and became the flesh and blood of left-right polarization.

Will libertarianism have its moment?

The answer to that question is that nobody knows. Yet I can see a non-crazy speculation combined with a non-crazy hypothetical that could allow a form of libertarianism to take the driver’s seat of one of the two major political parties.

The non-crazy speculation is that the deep source of high-education non-Christians’ newfound economic liberalism is mostly a reaction to the Republican alignment of business and wealthy interests with organized evangelicals that began in the Reagan era. Evangelicals want discriminatory policies and restrictive lifestyle policies that are antithetical to high-education non-Christians, and by so explicitly choosing the side of evangelicals, business/wealthy interests turned wealthier non-Christians from friends to enemies. This may be what gives high-education non-Christians’ economic liberalism a harder edge, where it isn’t just that the poor deserve sympathy, but that millionaires and billionaires and wealthy corporations must be reined in and regulated and made to pay.

The non-crazy hypothetical is that the Trumpian revolt might lead downscale non-evangelical whites to increasingly join white evangelicals in the Republican coalition, while pro-business/cosmopolitan folks increasingly join racial minorities and high-education non-Christians in the Democratic coalition. This would jumble both parties’ positions on redistributive issues and create increased contrasts on discrimination issues and internationalism.

If the speculation is true and the hypothetical comes to pass, then that could reduce high-education, wealthier, non-Christian whites’ motivation to adopt economically liberal opinions—because the pro-business agenda is no longer routinely allied with the evangelical agenda. Which could, over time, lead to a return of de facto libertarian preferences among wealthier, non-Christian whites. Which in turn could, over time, transform the Democratic coalition into, essentially, libertarians plus racial minorities. This coalition would be more focused on some minority priorities (e.g., anti-discrimination efforts, prison and police reform, and immigration reform) than others (e.g., increased government spending)—but the main draw would be simply pointing to the new Republican coalition’s white nationalist agenda and saying “we’re going to do whatever we can to stop those bastards from doing any of that.”

To be clear, I’m not predicting that this will happen. I’m also not saying that I want it to happen. I’m just saying that party coalitions are complex and dynamic, and the Trumpian revolt makes some screwball scenarios less far-fetched than they were before.