Ever since he attained the age of majority some four decades ago, Terence Hamilton, now 57, has voted in Mississippi’s Second Congressional District, covering 30 counties across the western part of the state. First drawn in 1986 to comply with the Voting Rights Act, the district’s scope has shifted somewhat over the years. But the bottom line has been a majority-black district in a state where African Americans comprise 38% of the population.

Like many of his black neighbors, Hamilton has reliably backed the district’s incumbent US representative, Democrat Bennie Thompson, who’s held his seat since 1993. But the US Supreme Court has now disrupted the pattern, holding that the Voting Rights Act doesn’t require states to draw majority-minority districts like Hamilton’s. In response, Mississippi Republicans have rushed to redraw the state’s electoral map in a manner that is certain to dilute black votes.

They’re doing this, of course, in the name of colorblindness.

Meanwhile, nearly 1,000 miles away, Ethan Chen, aged 18, is preparing to attend Georgetown University, not far from his family’s home in Bethesda, Md. It is a bittersweet transition for Chen, who had always considered Georgetown a “safety school.” His dream was to attend either Stanford or one of the Ivy League schools. But during the winter and spring, he received one rejection letter after another, despite a perfect grade-point average and equally impressive test scores.

How did this come to be, after another high-court ruling that barred colleges from discriminating against white and Asian applicants in the name of “affirmative action”? Chen, of course, will never know what exactly went into admissions officers’ decisions to reject him. What he does know is that many schools have continued to use essays and other “soft” factors to practice affirmative action and racecraft under another name.

The college officials, too, claim to uphold colorblind justice.

More from this author

How AI will change society

By Glenn Loury

These two stories — based on composite characters — illustrate the paradox of race and colorblind justice 250 years since the Declaration of Independence proclaimed that “all men are created equal.” The two recent Supreme Court decisions — on affirmative action and on voting rights — have clarified something profound about American public life. They reveal that two-and-a-half centuries since the Founding, we are trying simultaneously to transcend race and to manage its consequences and even “win” on racial terms. These ambitions pull against one another, suggesting that race will continue to indelibly stamp the American experience.

Consider the parallel nature of the two decisions. In Students for Fair Admissions v. Harvard, decided in 2023, the Supremes prohibited universities from explicitly using race in admissions. In Louisiana v. Callais, they sharply constrained the use of race in congressional redistricting under the VRA. On the surface, both decisions advance the same constitutional ideal of “colorblindness”: government should treat citizens without regard to race. Yet in practice, both rulings immediately generated what one might call a politics of proxies.

Universities didn’t abandon the pursuit of racial diversity after the high court proscribed affirmative action. Rather, university leaders shifted emphasis toward other considerations, some of which are strongly correlated with race: economic hardship, neighborhood disadvantage, first-generation status, personal adversity, and essays narrating the experience of marginalization. Formally speaking, these are race-neutral criteria. But everyone understands that they may also function as indirect means of sustaining black and Hispanic representation at selective institutions, as they appear to be doing.

Meanwhile, in the voting-rights context, a similar proxy logic holds. Red states attacking minority-majority districts increasingly invoke race-neutral partisanship, rather than race, as the guiding principle of line-drawing. Since black voters in much of the South overwhelmingly support Democrats, partisan gerrymanders often produce substantial racial consequences without having to explicitly classify voters by race.

The symmetry is striking. In both cases, non-racial instruments are being employed that have predictable racial effects. But political partisans see the cases differently. Conservatives who object to universities using race-correlated proxies to preserve minority enrollment often defend partisan districting as constitutionally neutral. Meanwhile, liberals defend “holistic” admissions criteria as part of a legitimate effort to broaden opportunity. This, even as the Left often condemns partisan gerrymanders precisely because their racial effects are foreseeable. Both sides, in other words, alternate between formal and substantive understandings of equality depending on context, thus illustrating what I call the colorblindness paradox.

This suggests that the stakes are deeper than ordinary partisan inconsistency. The real problem lies in the incoherence of colorblindness itself when race remains deeply embedded in the social structure. For decades, the Supreme Court has increasingly embraced a view of the Constitution that is suspicious of racial classification. The central danger, in this account, is the government sorting citizens by race and allocating benefits and burdens on that basis. This principle — suspicion of government racecraft, even when it’s used for “positive” or non-invidious reasons — dominates the high court’s equal-protection jurisprudence. Rightly so, in my view.

But implementing that principle is another matter altogether. The difficulty here is that race in America remains highly correlated with countless other variables that public institutions may legitimately consider when allocating scarce benefits or opportunities. Educational attainment, neighborhood quality, wealth, family structure, social networks, geography, and partisan affiliation — all continue to bear the imprint of racial history. And they all are relevant when exercising the government’s discretionary powers.

As a result, banning explicit racial classification needn’t preclude racially differentiated outcomes. It merely changes the mechanisms through which those outcomes are pursued — or avoided. The colorblindness principle provides little guidance here. This is why our public arguments about colorblindness increasingly revolve around debates over the use of proxies.

Take college admissions. If a university gives special weight to applicants who overcame severe educational disadvantage, is it pursuing a morally defensible concern for human development? Or is it covertly pursuing racial “balancing”? The answer isn’t obvious, since the non-racial criteria that serve as proxies for race are themselves neither arbitrary nor irrelevant. Educational disadvantages are real. Unequal access to stable schools, family resources, mentoring networks, and social capital is real. Moreover, non-racial criteria that correlate with race may genuinely identify talented students (of all races) whose capacities were developed under adverse conditions. Thus, it is legitimate for an admissions process to consider them.

But such consideration also predictably shapes the racial composition of entering classes, and it is plausible to suspect that this is precisely why some institutions are using them more heavily now than before the 2023 anti-affirmative-action ruling. If so, those institutions are violating the spirit if not the letter of the court’s holding, as Chen in our narrative learned to his chagrin.

Or consider redistricting. A legislature produces a map to maximize Republican advantage, knowing that such a map will reduce black electoral concentration. Is this merely ordinary political hardball with unfortunate racial side effects? Or is it something uglier — partisanship masquerading as a cover story for anti-black animus? Again, the answer isn’t straightforward. Political affiliation is itself a legitimate organizing principle in democratic politics. Yet in a society in which race and party remain so tightly aligned in some regions, the distinction between partisan and racial intent is hard to discern.

The difficulty here is that, while the court can forbid the explicit consideration of race, it can’t forbid the use of statistical signals that correlate with racial identity. Everything turns on the motivations of officialdom, which aren’t easy to determine.

This points toward a deeper conceptual confusion in our racial discourse, which arises from our conflating two very different ideas: racial blindness and racial indifference. A society may prohibit explicit racial classification while remaining intensely concerned with racial outcomes, embracing blindness while rejecting indifference. Indeed, that is precisely our current situation. Universities remain deeply invested in racial diversity even after affirmative action. Voting-rights advocates remain deeply invested in preserving black political representation even after the court’s retrenchment. The disappearance of explicit racial language hasn’t produced indifference to racial disparities. It has simply displaced racial management, and racial politicking, into more indirect forms. This distinction matters, because it exposes the limits of purely procedural solutions to racial inequality.

I have spent much of my professional life arguing that durable inequality isn’t sustained merely by discriminatory transactions at the point of selection. It is sustained through differences in social development — differences in the networks, norms, institutions, and cultural practices through which human capacities are formed. Opportunity travels along the synapses of social networks. Human-capital formation depends upon family stability, educational culture, neighborhood order, mentoring relationships, and access to trusted institutional pathways. These developmental processes unfold long before a university admissions office reviews an application or a legislature draws district boundaries. And racial identity figures prominently in these processes of social development in the United States.

This is why legal and administrative efforts to engineer proportional representation at the endpoint often feel unsatisfying. They attempt to correct downstream disparities via anti-white “reverse discrimination,” which the Supreme Court now decries, while leaving untouched the upstream developmental inequalities from which those disparities partly emerge.

And there is another drawback to the method. Formal neutrality can itself become morally evasive if it ignores the historical conditions that continue to shape social outcomes. A society can’t simply declare itself colorblind and thereby escape the consequences of its past. The legacy of racial exclusion remains embedded in institutions and inherited patterns of advantage and disadvantage. The Terence Hamilton in our opening narrative might be the great grandson of a slave and the son of a father who endured de jure racial apartheid. Those historical realities and their consequences don’t disappear the second a court in Washington holds them irrelevant.

Thus our predicament: the paradox of colorblindness. On the one hand, permanent racial management threatens to entrench precisely the kind of group consciousness that a liberal democracy ought to transcend. A political culture organized around endless racial accounting risks weakening common citizenship. The more every institutional disparity is interpreted through the lens of racial representation, the harder it becomes to encounter one another simply as fellow citizens. On the other hand, abstract invocations of neutrality may fail to reckon adequately with the developmental realities that shape unequal outcomes in the first place.

Suggested reading

We are all Eurasians now

By Anatol Lieven

America oscillates between these two poles because we seek both transcendence and remediation. We wish to become a society in which race no longer governs public life. Yet we also seek to offset the stubborn consequences of racial history. These goals aren’t fully compatible.

Indeed, the deeper paradox may be this: the more persistently racial disparities remain visible, the harder it becomes to sustain genuine colorblindness. But the more aggressively institutions organize themselves around racial outcomes, the harder it becomes to achieve a society in which race recedes in civic importance. No constitutional doctrine can fully resolve this tension, because it doesn’t arise merely from law. It arises from the unfinished social condition of the nation itself.

I supported affirmative action many years ago because I regarded it as a temporary bridge across inequalities inherited from Jim Crow. I remain sympathetic to the moral impulse behind such policies. But over time, I have grown skeptical that endless procedural adjustment at the point of selection can substitute for the harder work of social development. Nor do I believe that a healthy republic can indefinitely sustain a politics in which every important institution is expected to produce racially calibrated outcomes.

The Supreme Court’s recent interventions are therefore both understandable and welcome, but they are also incomplete. They reflect an effort to limit explicit racial administration in public life. Yet they can’t eliminate the underlying realities that make proxy politics inevitable, and inscrutable, in the first place.

For that reason, the real challenge before us isn’t simply legal. It is civic and developmental. If we want a politics in which race matters less, then we must build a society in which the capacities required for flourishing are more broadly and deeply cultivated across all communities. No amount of doctrinal refinement can spare us that obligation.

The paradoxes will remain. But 250 years after a 33-year-old Thomas Jefferson penned his majestic words, perhaps we can at least begin speaking about them honestly. Indifference to the roots of racial inequality is actually inconsistent with achieving colorblindness in the conduct of our public affairs.

The article stated that the results of last year’s census marked “the first time the absolute number of people who identify as White alone has shrunk since a census started being taken in 1790”. Similar refrains rang out across numerous other news outlets: “America’s white population set to shrink”; “America’s White Population Shrank for the First Time”; “Vast Stretches of America Are Shrinking. Almost All of Them Voted for Trump”.

There was jubilation among the Left. Michael Moore called the announcement “the best day ever in US history”. For cultural radicals such as he, the census marked the end of “white supremacy” and the start of a millennium of progressive dominance and racial equity. Similarly, Neocon-turned-Never-Trumper Jennifer Rubin gushed that “this is fabulous news — now we need to prevent minority White rule”.

Meanwhile, on the Right, Tucker Carlson, while condemning the Left’s gloating, went on to claim, without evidence, that the opioid crisis in white communities explained the decline. Likewise, the news will only confirm views held by the far-Right, whereby white nationalists have warned that white decline is being orchestrated by globalist elites and minorities, and which will lead to dire consequences for white people and their children.

However, despite the charged rhetoric doing the rounds, the number of people American society considers to be white has almost certainly increased since 2010. White America is not shrinking, and whites are not being racially supplanted. It’s a false narrative with serious political repercussions: study after study has found that reminding white Americans about their impending minority status shifts their policy attitudes in a conservative direction and increases support for Right-wing populism. So while this announcement may make the cultural Left feel good, it largely benefits the Trumpist Right.

But what do the actual findings indicate? Isn’t the census reporting the facts as they stand, that the number of white people in the country is declining? Well, it’s certainly true 5.1 million fewer non-Hispanic Americans ticked their race as “white alone” in 2020 compared to 2010. But to take this number out of context is to paint a misleading picture of social reality.

Any quantitative social science student knows that when you change the wording of a survey question, you can alter — often substantially — the pattern of responses. The 2010 census questionnaire offered single tick boxes for “white” and “black”. The 2020 version was changed so that people were not only asked to tick a box but also prompted to write their ethnic group in a text box below. Research from Pew shows that around 15% of black and white respondents did not write anything in the text box, or they wrote something that doesn’t make sense. Hispanic Americans were also told, for the first time, that Hispanic is not a race[1.  For this survey, Hispanic origins are not races”, was the wording alongside the What is your race? question.] when asked to define theirs. It is likely this prompted people to question their choice of which box to mark, opting for a multiple race or “other race” category instead. This would depress the number ticking the “white alone” box.

In addition to the change in census wording, the politicisation of white identity in the Trump era may also have led liberals to distance themselves from the category. The 2010s bore witness to a sharp rise in cultural progressivism. During this decade, the share of liberal Democratic white citizens, saying they felt “cold” towards white people increased by roughly 10%.

More from this author

What liberals get wrong about race

By Eric Kaufmann

There’s also the fact that many WASP Americans, such as Elizabeth Warren, have some Native American ancestry and some, such as Brad Pitt, believe they do even when such claims cannot be substantiated. Similarly, since 8 in 10 black Americans have white or other racial backgrounds in their family history, and 1 in 10 white southerners has African ancestry, the scope for expansion in the mixed or other race categories is potentially enormous.

So did these changes affect what people wrote on the 2020 census? Much of this depends on three things: whether their answers reflect a) their own identity, b) how they think others see them, or c) some judgement about the preponderance of pan-ethnic origins in their background.

Figure 1 (below) shows that the number of Americans ticking “white alone” as their race dropped by 19.3 million between 2010 and 2020, while the number ticking “white” alongside some other category increased by 23.6 million. Much of this shifting took place within the Hispanic population, but even among non-Hispanics, “white alone” declined by 5.1 million.

At the same time, the census recorded a rise of 24.8 million in the “two or more races” group and a jump of 28.1 million in “some other race”. The increase of 3.6 million in the population of those with mixed “American Indian and Alaska Native” and other racial background is also noteworthy given that this group numbered just 2.3 million in the 2010 census. Hardly any of this shift can be explained by aboriginal birth or death rates.

More from this author

Israel's population time bomb

By Eric Kaufmann

Indeed, it's striking that most biracial white and Native Indian Americans are effectively considered white and view themselves as such. Meanwhile, the assimilation process means that with each generation, fewer descendants of Hispanic and Asian immigrants, especially those of mixed background, identify with their origins. It is, therefore, difficult to conceive that the number of Americans considered white is in decline.

Yet none of this means there is nothing happening with the nation’s ethnic demography. The balance between white births and deaths, taking into account the small flow of some 200,000 white immigrants a year, is nearly equal. The non-Hispanic white population, using the old question, was essentially flatlining between 2010 and 2019. White Americans have maintained a North European fertility pattern, with a total fertility rate between 1.5 and 2, since the Seventies.

Though relatively high for the developed world, this is still below the magic 2.1 replacement level. The “unmixed” white American population, along with white ethnic majorities in the rest of the developed world, will start to gradually decline in the decades ahead. But the big demographic story of the 2010s wasn’t about White Americans, but the big drop in Hispanic fertility, which is rapidly converging with the more stable white rate. So even while the share of “unmixed” whites will probably dip below 50% by 2050, the share who are considered socially white is likely to be a majority into the foreseeable future.

Suggested reading

Why the rich are revolting

By Ed West

Either way, to be more informative, the US census needs an update. Social scientists and pundits are interested in people’s subjective racial identity, but we also want to know about the balance of “objective” racial categories. The latter shapes people’s decisions about where to live, who to marry, who to vote for and can affect their life chances. One option might be to ask people how others perceive their race. In Northern Ireland, people’s growing tendency to state their religion as “none” or “not stated” led census takers in 2011 to add a question asking about the religion people were brought up in.

Ultimately, though, until the results of such a question show a real decline in those considered to be "white", the zeal to foreground the Great Replacement tells us more about ideological hopes and fears than any emerging social reality. One of the few seeking to dispel such illusions is leading American sociologist of ethnicity, Richard Alba. However, his important book The Great Demographic Illusion failed to garner reviews because it doesn’t fit the prevailing narrative. Instead, the media are attracted to demographic stories which heighten white threat perceptions, gratifying progressives while increasing support for Trumpism.

So while the definition of white may be blurring at the edges, the white American majority isn’t going anywhere.

But are its findings really evidence of widespread racism in the US? Hardly. In fact, you need only look at the behaviour of Americans — at the number of interracial marriages or police shootings of minorities, for example — to see that racism has almost never been so absent.

Take interracial marriages. In 1958, 94% of whites opposed it — yet just 10% do today. Similarly, the long-running General Social Survey found that in the 1970s nearly 60% of white Americans agreed with the statement that blacks shouldn’t “push themselves where they’re not wanted”. In 2002, that figure fell to 20% — and the question was discontinued.

It’s much the same story with police shootings of African-Americans. Despite the charged rhetoric of the past year, which culminated in calls for entire police forces to be disbanded, such shootings are 60-80% lower than they were in the 1960s.

So why is there such a misunderstanding between the American public and the reality of US race relations? The answer, I suspect, reflects a media-driven moral panic — one that is rooted in the racialisation of identity-obsessed progressive politics.

Suggested article

Antiracism won't save you

By Tomiwa Owolade

Indeed, the “rise” of racism in America since 2014 is a social construct that reflects perception rather than daily life. This is largely a result of what Amos Tversky and Daniel Kahneman described in 1973 as the “availability heuristic” — the idea that vivid images rather than statistical reality tends to shape people’s perceptions. For instance, people routinely overestimate vivid phenomena, from crime to the share of Muslims in their country, because these stories make the news. Crime in America may have fallen every year before 2019 — but when asked about it by Gallup, most people in all but two years said crime had risen over the past year.

So too with race. In some ways, this is to be expected. Whenever events such as the 1992 Los Angeles riots or the killing of George Floyd make the news, people — shocked by what they are seeing — inevitably perceive race relations to be worse than they are. What is distinctive today, however, is the sustained determination of major progressive news outlets like the New York Times or CNN to centre racism as a dominant theme in American society.

As UnHerd has previously noted, citing ground-breaking research by Zach Goldberg of Georgia State University, the use of the term “racism” and similar terms in mainstream news outlets took off around 2015, (which can be seen here).

What caused this spike? Goldberg cites social media and the new partisan clickbait business model in online journalism as the catalysts. Then, during the campaign for the 2016 US election, Donald Trump energised the enterprise. As the New York Times‘ Ben Smith recounted last year: “Heated Twitter criticism helped to retire euphemisms like “racially charged. The big outlets have gradually, awkwardly, given ground, using “racist” and “lie” more freely, especially when describing Mr. Trump’s behaviour.” Certainly, many of these outlets swiftly retired all objectivity and standards in favour of this new partisan model.

One reason for this was the media’s desire to tap into a mythic script of a white officer killing a black suspect. Indeed, Goldberg’s study of the 2015-20 Washington Post police killings database reveals that black police shooting victims received nine times the media coverage of white victims. As Coleman Hughes has observed, when footage emerged in 2016 showing Tony Timpa, a white suspect, being suffocated as a Dallas officer placed his knee on his neck, the media looked the other way while the officers walked free.

Suggested article

The battle between the two Americas

By Joel Kotkin

This media fixation on identity politics, alongside pre-existing misperceptions, ultimately skews the public’s sense of reality. The number of unarmed black men killed by police in the Washington Post’s own database in 2019 was between 13 and, using a very broad definition of “unarmed”, 27. Yet nearly half of “very liberal” Americans think the number is between 1,000 and 10,000. There were over twice as many unarmed whites killed by police as blacks but, as John McWhorter, author of the new book Woke Racism notes, this never makes the news because it doesn’t fit the narrative of white racial violence against African-Americans.

The narrative of police racism is exacerbated by party politics. A young black man is around ten times as likely to die in a car accident than from a police bullet, but only 10% of African-American Democrats and 20% white liberals get this question right — while more than 70% white Trump voters do. While there are some who are undecided, Figure 2 (below) shows that 95% of blacks and 70% of whites who believe “all white Republicans are racist” agree with the statement.

The impact of such warped perceptions cannot be overstated. But rather than trying to correct untruths, the American Left has leaned into these corrosive myths. Witness Elizabeth Warren tweeting that Michael Brown was murdered by police in Ferguson, despite the fact that the officer had been cleared by a federal grand jury of wrongdoing. Or, in the wake of the Chauvin verdict, Joe Biden reinforcing the irresponsible fiction that black people in America live in fear of their lives from police bullets:

We must do more on police violence, so they don’t have to wake up knowing that they can lose their very life in the course of just living their life. They don’t have to worry about whether their sons or daughters will come home after a grocery store run or just walking down the street or driving their car or playing in the park or just sleeping at home.

No doubt Biden was well-meaning, but he was, in effect, propagating a conspiracy; it was similar to telling an anti-vaxxer that their fears are well-placed and it is up to pharmaceutical companies to eliminate every last side effect before we can expect them to trust a jab.

Suggested article

The fightback against Critical Race Theory

By Douglas Murray

Of course, Black Americans have had a historically uneasy relationship with the police, but that is no excuse for the blend of romanticism and virtue-signalling that seems to have gripped today's progressives.

While Democratic politicians and the media rightly held Trump to account for violating liberal democratic norms, such as claiming the election was stolen or encouraging the Capitol rioters, they were strangely silent when Biden said he was "praying for the right verdict" in the Chauvin trial and Democratic Congresswoman Maxine Waters told a crowd in front of the courthouse to "get more confrontational" and let the judge "know that we mean business", Meanwhile, media interest in the record-setting $2 billion in damage caused by the riots following George Floyd's death — which struck a disproportionate number of minority-owned businesses — was minimal.

It was a disturbing reflection of the distorted, racialised reality they now inhabit. Just don't expect them to take responsibility for it.