The Heritage Americans Have Already Been Replaced

What everyone is missing in the debate about American identity.

Courtesy Anna Shipman/Unsplash.

What, exactly, is a heritage American? It’s the question that nobody can seem to answer, despite a smoldering dispute on X and among right-wing activists. Vivek Ramaswamy—currently seeking the Republican nomination for Governor of Ohio—reignited the debate in a provocative New York Times op-ed, which argued that the American right is now divided by its very understanding of what an American is.

“One vision of American identity is based on lineage, blood and soil: Inherited attributes matter most. The purest form of an American is a so-called heritage American — one whose ancestry traces back to the Founding of the United States or earlier,” Ramaswamy wrote. “The alternative (and, in my view, correct) vision of American identity is based on ideals. … You are an American if you believe in the rule of law, in freedom of conscience and freedom of expression, in colorblind meritocracy, in the U.S. Constitution, in the American dream, and if you are a citizen who swears exclusive allegiance to our nation.”

Ramaswamy was replying to a growing tendency on the right to identify Americanness with descent from the nation’s Founding population. If your ancestors weren’t here at the Founding (or in 1860, or 1942, or 1965—nobody can quite agree which it is), then you aren’t a real American. At the very least, you aren’t as American as a member of the DAR.

Ramaswamy’s volley, of course, prompted its own replies, with critics insisting that America is not an idea, but a common heritage. “America does have a creed, but we also have a common culture and a common stock, as do all other nations,” Colin Redemer wrote in a representative response in First Things.

Nothing about this debate is new; the fight over whether America is defined by creed or descent has split the right since at least the 1990s, when intellectuals tussled over the ideas of Samuel Huntington and Pat Buchanan. But while this debate is in some senses about ideas, it is also fundamentally about demographics. And there’s a basic problem with one side’s position: most Americans simply aren’t descended from the Founding generation, or even those who were here after the Civil War.

To understand why, start with who was present at the Founding. As Schoolhouse Rock helpfully reminds us, “America was founded by the English, But also by the Germans, Dutch, and French.” To that list we can add, per one analysis, the Welsh; the Scotch-Irish; the Scottish and Irish separately; and a small number of Swedes. To be comprehensive, we should also count the ancestors of today’s black Americans, between 75 and 90 percent of whom are plausibly descended from slaves.¹ And we should include descendants of the Native Americans.

What fraction of today’s Americans are descended from this “Founding stock”? One way to get at this question is simply to ask people their ancestry—as, indeed, the American Community Survey (an annual survey product from the U.S. Census Bureau) now does. The ACS gives respondents the ability to list two different ancestries; the figure below uses data from those who gave responses to at least the first one. Notably, this is an overestimate—if, for example, you’re an English person who arrived last week, you’d still count toward the Founding-stock-attributable share here.²

The image above offers two visualizations of people’s responses to this question on the 2024 ACS. Each square in the waffle plot represents 1 percent of the relevant population. (There may be some slight error due to rounding.)

Figure 1 depicts the breakdown of respondent’s first self-reported ancestry (among all those who gave any answer). It implies that about 44 percent of respondents identified their ancestry as being in our Founding stock groups. The remaining majority—56 percent—don’t claim such descent.

Figure 2 looks within respondents who identified with a Founding-stock ancestry and asks what their secondary ancestry was, if any. About 56 percent gave no second answer, while 29 percent named some other Founding-stock ancestry. The remaining 15 percent said they had some non-Founding-stock secondary ancestry.

What does all this mean? At its greatest extent, the population solely attributable to Founding stock—i.e. descended from Founding-origin groups, without any out-marriage—is about 37 percent of Americans.³ In other words, a little bit more than one in three Americans is of the appropriate ancestry to potentially be a “heritage American.” There are (at most) about 126 million heritage Americans—substantially outnumbered by the (at least) 216 million Americans who are at least partly attributable to post-Founding arrivals.

Where did all of those new ancestors come from? People who post about “heritage Americans” often fixate on the wave of arrivals since the passage of the Immigration and Nationality Act of 1965. That law significantly increased immigration, especially from Latin America and Asia, creating a wave that has yet to subside. Per Pew, about a quarter of the population was either a first- or second-generation immigrant as of 2012—equivalent to about 85 million people today.

But what about the remainder—the people who are neither recent arrivals nor plausibly Founding stock? You may remember that we are currently in the second great wave of immigration in American history. (If not, consult the Schoolhouse Rock video above.) The first—beginning after the Civil War and lasting until entry was significantly curtailed by the Immigration Act of 1924—saw the arrival of some 30 million people. That’s a significant increase compared to the 38 million people who were counted in the 1870 census. As I wrote in a 2023 Manhattan Institute report:

As before 1860, the vast majority of these new immigrants—90% from 1860 to 1919—were European; a further 5% were Canadian. But they came from many more countries, including Italy (12% of arrivals); Austria-Hungary (12%); Germany (12%); Russia (9%); the United Kingdom (9%); Ireland (7%); and the joint kingdom of Norway-Sweden (5%). A large number of Jews also arrived from across Eastern Europe.

These arrivals significantly altered the demographic composition of the United States. They were, of course, mostly European (a point that people enamored of “heritage Americans” like to bring up when I observe this). But that didn’t stop the native population from looking askance at them. Discrimination against Italians and Irish was not uncommon in early 20th century America. And indeed, the level of disruption that natives perceived was part of why immigration was effectively paused in the 1920s, to give the new arrivals time to assimilate.

In that report—as part of a separate inquiry—I asked a relatively simple question: how much of today’s population is attributable to post-Civil War arrivals? Imagine we had closed the borders in 1870 and kept them sealed; how much smaller would our population be? For purposes of the paper—which was about reparations for slavery—I looked in particular at the non-black population. But that group is relevant to our topic insofar as the black population is (as previously mentioned) mostly not attributable to immigration, and especially not to the first wave of immigration.

The results of my efforts to answer that question are depicted in the chart above.⁴ In short: I estimate that the “natural increase” of the vintage 1860 population—the rate at which that population would have grown in the absence of any immigration or emigration—implies that today there would be about 88 million non-black Americans. In reality, there are 286 million non-black Americans, of whom about 197 million are non-Hispanic whites.

What this implies—although it’s just an estimate—is that 69 percent of today’s non-black population, and about 55 percent of the non-Hispanic white population, is attributable to post-1860 immigration. Put this together with a presumed 90 percent of the black population—the share assumed to be descended from slaves—and you get an implied pre-1860-attributable share of 39 percent.⁵

You can quibble with this method as with the ACS estimate, but the basic point remains the same: however you slice it, the super-majority of Americans are not “heritage.” They are instead descended—either party or exclusively—from one of the tens of millions of people who came here after the Founding period, and even after the Civil War. Nor are these people all recent Asian or Hispanic arrivals—a sizable share of them are attributable to the mass migration of the late 19th and early 20th century.

If you believe that America is for heritage Americans, in other words, I regret to inform you that the heritage Americans are no longer the majority. And that change isn’t recent; it’s been baked in for over a century.

Moreover, and in part because of this demographic mishmash, your view isn’t shared by most of your fellow Americans. Take, for example, the above 2025 polling from the left-leaning Public Religion Research Institute. PRRI asked respondents what characteristics are important to truly American. Across partisanship, large majorities emphasized things like believing in individual liberties, the Constitution, and our basic institutions. Respondents were narrowly divided on qualities like being born in America and believing in God. But only a quarter—including just a third of Republicans—say that you need to be of Western European heritage in order to be an American.

For a similar result, consult a module from the 2024 General Social Survey, a subset of respondents to which were asked about how important they felt different things were to American identity. While again large majorities identified respecting our institutions and having American citizenship, only about a third said that being American requires having American ancestry. Interestingly, white respondents—the racial group most frequently identified as deserving some unique heritage status—were the racial group least likely to say that having American ancestry was important to being an American.

That’s not all that surprising: in 2021 Pew polling, white respondents were the group least likely to self-identify with their self-defined origin. Less than a quarter said that their origin was “central to their identity,” relative to over half of black and Hispanic respondents. None of the above data allow breakout by self-identified ancestry. But it seems quite likely that those Americans who can trace their identity back to the Founding (or similar) are in large part not actually strongly identified with that lineage.

In short: when someone like Redemer claims that America has a “common stock,” he is factually incorrect. The large majority of today’s Americans are not descended from the people who founded our nation. That’s not a recent phenomenon. In fact, it’s been a part of our culture so long that most Americans do not consider common descent as an essential part of American identity—including those who would most obviously benefit from that being the case.

Does that mean that Ramaswamy is right, though, that “American identity is based on ideals?” I don’t think so—or, at least, not exactly.

The ideals that Ramaswamy says define an American—belief in “the rule of law, in freedom of conscience and freedom of expression, in colorblind meritocracy, in the U.S. Constitution, in the American dream”—are what respondents to the above polls pointed to. But these ideas are, importantly, not untethered from a particular heritage. The concepts of liberty encoded in the American tradition at the Founding come from somewhere: the English philosophical and legal tradition in which the Founding fathers were embedded. Those ideas have an ineluctable cultural substratum. The Constitution of North Korea also provides for freedom of expression, but the words plainly have a different meaning in the DPRK.

To be an American, moreover, it is not enough to mouth the word “liberty.” One has to actually have the experience of practicing it, as de Tocqueville saw it practiced in the early American republic. And the form of liberty on display there is, like it or not, an inheritance of the particularly English political context from which our institutions sprang.

That culture, moreover, was and can be again challenged by immigration. The Irish and Italians who arrived in the late 19th and early 20th century came from distinct political cultures and struggled to integrate with America’s norms. Most other nations are not as free as is our own—some days it seems as though Europeans like our concept of free speech almost as little as the North Koreans do. And, as George Mason economist Garrett Jones has argued, American culture is a “goose that laid the golden egg”—even from the disinterested perspective of global total utility, interfering with our delicate institutional balance through uncontrolled mass immigration is an obvious error.

None of this means that we should seal the borders. Rather, it means that we need to pay careful attention to cultural as well as economic fit in selecting immigrants. We should want smart people who love our country and know how to practice American values. Conversely, there’s nothing wrong with rejecting people who hate America, no matter how skilled they might be.

It’s quite possible that Ramaswamy agrees with all this. But we should not undersell the challenge posed by the end of heritage America. We no longer live in a nation populated primarily by the English who founded the country. But we are still inheritors of their tradition, a tradition that is both essential to our success and hard to replicate.

In other words, most of us aren’t heritage Americans. But even those who aren’t have a responsibility to that heritage nonetheless.

1. This figure is estimated in the report I cite later in this piece.

2. If you care to replicate this analysis in IPUMS, I used the following codes for the ANCESTR1 and ANCESTR2 variables: 11 (“British”), 12 (“British Isles”), 21 (“Dutch”), 22 (“English”), 26 (“French”), 32 (“German”), 50 (“Irish”), 87 (“Scotch Irish”), 88 (“Scottish”), 89 (“Swedish”), 97 (“Welsh”), 900 (“Afro-American”), 902 (“African-American”), 920 (“American Indian”). For fig 1, I drop respondents who give answers > 994.

3. 0.44 * 0.85 = 0.37.

4. If you want to know how I got there, read the report.

5. (1 - 0.69) * (1 - 0.14) + 0.9 * 0.14 = 0.39.