In 1995, the Duke University economist Timur Kuran published Private Truths, Public Lies, a study of what he called “preference falsification”—a direct challenge to the neoclassical economic view that human preferences are fixed, purely rational, and designed to maximize personal utility. In fact, Kuran suggests, human beings keep two sets of books for their preferences. When they find themselves presented with social incentives to support a particular position, both individuals and social institutions can and often do adopt public preferences that are out of sync with their private preferences.

“Preference falsification” is economically and socially inefficient, Kuran argues, because it leads to an impoverishment of public discourse. In the long term, such a split between public and private leads to knowledge falsification. This is the concealment of pertinent knowledge that might support opinions that are viewed as socially unacceptable. And it leaves societies vulnerable to dramatic and unexpected changes, because private opinion can shift imperceptibly beneath the crust of public opinion until a single (quite often minor) event triggers a bandwagon that topples the status quo.

Kuran’s observations are immensely helpful for understanding our current social and political moment. Obviously, Donald Trump’s “surprise victory” in the 2016 U.S. presidential election was a shock to many precisely because the news media had heaped such opprobrium on Trump. The perceived stigma of being a Trump supporter created a wide gap between citizens’ public and private opinions, a gap that collapsed only when people were able to express their preferences through the secret ballot. In the short term (but notably not in the long term), suppressing private opinions from entering the public sphere does not make them go away—it merely makes them difficult to gauge.

Since late May, America has been rocked by a series of controversies and crises—in particular, the brutal and unjustified killings of George Floyd and Breonna Taylor. Americans have been watching as months of clashes between federal agents and protesters come to a head in the Pacific Northwest. And in the national conversation surrounding these events, many views held by a substantial segment of the American electorate have become unspeakable in polite society.

How do we know this? As recently as 2017, when social circumstances were far tamer than they are now, a Cato Institute Poll found that 57 percent of moderates felt that the political climate prevented them from saying what they believed. Cato recently published a follow-up study that revealed unequivocally how Americans are migrating toward self-censorship as a form of self-protection. Seventy-seven percent of conservatives, 64 percent of moderates, and 52 percent of Democrats now feel the need to self-censor. Thirty-two percent of all Americans worry that their political views could do damage to their career. The only group without a majority of members who felt the need to self-censor: strong liberals, and even 42 percent of them said they felt they needed to watch themselves. It’s axiomatic that when people say they are frightened to say what they mean, we are in a moment in which we cannot quite believe what they do say. So what are they saying that we might need to take a second look at?


A graphic published in the New York Times on June 10 showed a massive spike in public support for the Black Lives Matter movement after Floyd’s death—67 percent, according to the Pew Research Center as of June 12, 2020. According to the co-authors of the Times article, public support for the movement has steadily increased since its inception in 2013, a trend that was in place even before Floyd’s murder. Just as with same-sex marriage, Nate Cohn and Kevin Quealey of the Times noted, “American public opinion tends to drift toward the side advocating equal treatment.”

These claims, while true on their face, are symptomatic of the misunderstanding of the environment before the 2016 election. They don’t take into account how the social environment is splitting private and public preferences. A close look at other questions from the same June 12 Pew Poll points to cracks in the seemingly broad support for the current movement. For example, Pew found 59 percent of Americans “say some people taking advantage of the situation to engage in criminal behavior has also been a major contributing factor in the protests.” Only 19 percent believe that organizing protests would be a very effective strategy for helping black Americans achieve equality, and only about one third of whites “strongly support” the movement. Another third of whites said they “somewhat support” the movement—but as the wording suggests, the depth of such a conviction is questionable and might instead be a perfect reflection of “preference falsification.” In passing, Cohn and Quealey acknowledge that “defunding the police” (which has since begun in many areas) may not find popular support—a seemingly minor point that makes all the difference, since recent polls have found that 65 percent of Americans are opposed to it.

On June 16, the Associated Press reported that more than four dozen progressive groups had issued an ultimatum to the presidential campaign of Democratic nominee Joseph Biden: Support the wholesale defunding of the police, or lose the black vote. The argument goes that if Biden fails to galvanize the black and progressive electorates, enough people will stay home or vote third-party to cost him the election. This logic fails for at least two reasons. First, contrary to what the activists say, more blacks are opposed to defunding the police than are in favor of doing so. Second, despite demographic projections of the “browning of America,” white voters without a college education accounted for 44 percent of the 2016 electorate while all blacks accounted for only 10 percent. Between 2012 and 2016, the number of non-college-educated whites who identified as Republican increased by 6 percent while those who identified as Democrats declined by 4 percent. Since then, the trend has only continued—no change in Democratic affiliation and an additional 2 percent of new support for Republicans from previously unaffiliated voters. It is true that minority apathy is worth worrying about (Hillary Clinton captured 4 percent less of the black vote in 2016 than Barack Obama did in 2012), but there are good reasons to believe that the 2020 election will still be influenced heavily by working-class-white voting preferences.

Kuran’s dichotomy of “the unthinkable” versus “the unthought” is instructive here. According to Kuran, some beliefs start off as “unthinkable”—that is, they are so taboo that no respectable person can hold them publicly or even entertain them privately without some degree of discomfort. Outside of the political right, one example of an “unthinkable” belief would be that police shootings are not systemically racist (more on that later). By contrast, “unthought” beliefs are those that most people never even entertain. Kuran shows how once “unthinkable” beliefs gradually disappear from popular discourse and become “unthought” by future generations. In the current fraught moment, it seems that positions out of step with the Black Lives Matter movement are hovering somewhere between “unthinkable” and “unthought” on the mainstream left. Older left-leaning Americans have seen enough of history to remember a time before the movement’s ascendency, but younger ones have not. If the taboos surrounding our current conversation on policing and race continue unabated in the long term, we are at risk of an intellectual bottleneck event—a younger generation growing up without heterodox voices that help to refine and improve what the left should stand for.

Meanwhile, a torrent of public pressure has been loosed on companies, apolitical organizations, and individuals to support the movement openly, even vociferously. “Staying silent” on systemic racism in policing is no longer a credible option for any respectable person with a public footprint, and silence is equated with complicity. The protest movement’s foisting of this uncompromising stance onto the rest of the left has forced many relevant facts into the realm of the “unthinkable,” and so they can now be discussed only on the right. Whether Floyd’s killing was a priori an act of racism (rather than the result of general police incompetence and callousness), whether it necessarily indicates an epidemic of police killings of black suspects, or whether the obvious solutions to these problems are lower police budgets and a smaller police presence in high-crime areas, the current answer on the mainstream left to all these questions is “yes,” and it’s not up for debate.

Even if these views turn out to be correct in part, they are unlikely to provide a complete story or the best solutions. For example, a paper by the Harvard economist Roland Fryer found that “viral” instances of deadly force in five American cities have led to changes in police behavior resulting in a total of 900 excess homicides and 34,000 excess felonies in the 24 months following the incidents. What should we make of this fact, which has since been dubbed the “Ferguson Effect”? In a separate paper, Fryer also found that while police are more likely to go “hands-on” in roughing up a black or Latino suspect, they are actually less likely to shoot them than they are white suspects. Heather Mac Donald, who has covered these issues for decades (including in the pages of Commentary), delivered congressional testimony in June citing Fryer’s work and has written many data-driven articles that challenge the left’s received views on policing and race. In July, however, Mac Donald’s citation of an article with findings similar to Fryer’s caused the paper’s co-authors to repudiate her and disown their own scientific conclusions. While one might dispute Mac Donald’s interpretation of the data, it is troubling to see scientists reject that data simply because someone they dislike cited it.

Right now is not a popular time outside of the political right for defending the police, supporting law and order in the face of heightened crime, or seeking to address police brutality in nonracial terms. And yet we must engage with these points if we are to make meaningful progress. From the outset, the vast majority of Americans have believed that George Floyd’s death was horrific and indefensible. But many of these Americans will also dissent in private from the take-it-or-leave-it, package-deal orthodoxy offered by the current protest movement; it is hard to imagine that most will want to see their police department defunded. If the American media had learned anything from the last election, it would make an effort to allow the airing of these views in public without overbearing stigma, the better to debate them openly.


One of the most contentious debates to emerge recently in the social sphere concerns the topic of “cancel culture”—what it is, whether it exists at all, and what it is doing to our public discourse. Put briefly, cancel culture represents a breakdown in the mechanics of conversation in favor of reputational punishment. Any potentially offensive opinion (mistakes or false accusations included) has become grounds for firing, public shaming, or forced resignation. Despite claims from Vox founder Ezra Klein and Democratic Representative Alexandria Ocasio-Cortez that cancel culture is either immaterial or merely a cover story for the powerful to protect themselves from criticism, the facts tell a different story. Heterodox thinkers such as Bari Weiss of the New York Times and Andrew Sullivan of New York Magazine, who once injected intellectual variety into their respective publications, have recently been forced out by hostile work cultures.

A July 7 open letter in Harper’s put it thus: “The free exchange of information and ideas, the lifeblood of a liberal society, is daily becoming more constricted.” The letter called for a more open and searching civic culture—in other words, less preference falsification. Its core argument was vindicated when it was subject to a muscular backlash, including condemnation of the signatories, attempts to fire some, and a number of signature retractions.

It’s easy to quibble about the precise definition of cancel culture or to what extent it currently pervades society, but that misses the point. What we already know about the direction in which our vindictive and harshly punitive civic culture is going gives us enough cause for worry, with 50 percent of “strong liberals” favoring the firing of people who donate to the Trump campaign and 36 percent of strong conservatives with the same opinion about Biden donors. Both sides increasingly feel a strong sense of political desperation and exhaustion, a magnetic pull toward in-group loyalty, and an apocalyptic willingness to do anything to rid themselves of the evil of the other.

Trump’s private support among Republicans may actually be much lower than it appears in public, and his poor handling of COVID-19 has not helped his cause at all. But we should not assume that his missteps will cost him the election in what is now truly a postmodern age of madness. Democrats have an important role to play in police reform, but they risk handing the election to Trump in 2020, as they did in some ways in 2016, by allowing a new orthodoxy to blind them to common-sense realities. They should recognize the significance of “preference falsification” within their own ranks and work to minimize it, even if its results are flattering and comforting to them. And in general, they should strive to be intellectually honest because we all should. If they fail, they might learn the hard way once again that while they may think themselves on the right side of history, the Americans who actually choose the president in November will quietly, and almost secretly, disabuse them as they did in 2016.

We want to hear your thoughts about this article. Click here to send a letter to the editor.

+ A A -
You may also like
Share via
Copy link