Chandan Khanna/AFP via Getty Images
A Black Lives Matter supporter confronts members of the Proud Boys in Miami on May 25, 2021Chandan Khanna/AFP via Getty Images
Navigate to News section

The Assault on Empiricism

From crime to climate change, the hostility of ‘movements’ to data is making it impossible to address real-world problems

by
Wilfred Reilly
August 12, 2021
Chandan Khanna/AFP via Getty Images
A Black Lives Matter supporter confronts members of the Proud Boys in Miami on May 25, 2021Chandan Khanna/AFP via Getty Images

While we’re talking about things that matter—Black lives, blue lives, all lives, vaccines, guns—how about facts?

A remarkable aspect of today’s culture war debates, across a whole range of topics, is the fact that many massively popular positions bear no resemblance to measurable truth. Many core claims of Black Lives Matter (BLM), the “systemic racism” school of sociological thought, the Stop AAPI (Asian American Pacific Islander) Hate movement, and, for that matter, QAnon and election fraud devotees on the political right, don’t gel at all with empirical facts. Often this is no accident. Many activists and a surprising number of academics and media figures seem to have returned to the ancient idea that truth is relative, and hard data of the Bureau of Justice Statistics variety are less valuable than individual “lived experience.” Others, so far as I can tell, simply lie to facilitate personal or political goals. The trend is a dangerous one: People who fear nonexistent demons are also likely to propose costly and unnecessary witch hunts. In the face of the new Hounds of God, empiricism must again be defended.

It is simply not debatable that the claims of many popular modern movements are far removed from reality. BLM’s contentions about a near-genocide of African Americans directed by police may be the best example of this. To give only two famous examples out of dozens, Black Trans Lives Matter activist Cherno Biko stated on prime-time television in 2015 that an innocent Black person is “murdered” by American police “every 28 hours,” while star attorney Benjamin Crump hinted at an even higher total in a 2019 book he titled Open Season: The Legalized Genocide of Colored People.

These sorts of claims have become conventional wisdom on the political left. A well-run and large-N study from the Skeptic Research Center in February 2021 found that 54% of Americans who “identify as very liberal” believe that the average number of unarmed Black men killed annually by U.S. police is somewhere between “about 1,000” and “more than 10,000.” A major empirical survey conducted by the political scientist Eric Kaufman in April 2021 found that 80% of African Americans and 60% of educated white liberals believe that more young Black men die annually at the hands of police than in car wrecks.

OK. The actual number of unarmed Black men killed by police last year was 17. Given the grave importance of this issue, it’s worth repeating that number—17—across the tens of millions of annual police-citizen interactions. As it turns out, about 1,000 people of all races, sexes, and ages are shot to death by on-duty police officers in a typical year (1,021 in 2020), and about 250 of those are identified as Black. Even this gap—between the share of Americans who are Black (13%) and the share of police shooting victims (24%-25%)—largely vanishes when a simple adjustment is made for the gap in reported crime statistics (and thus in police encounter rates) between Blacks and whites, which was 2.4 to 1 when the Bureau of Justice Statistics’ National Crime Victimization Study collected data on violent crime victims in 2019. Even leaving such academic points aside, it is clear that estimates of police violence on at least the political left are several orders of magnitude higher than reality.

The same is true for popular perceptions of racial conflict and interracial crime. American media is chock-full of shocking stories about cruel whites harassing or attacking people of color as we do the most basic things in our daily lives—barbecue in parks, swim in public pools, watch birds, ride the Chicago L or New York 6 train with a hijab on, or just walk home from Subway with a tuna sandwich. Basketball centimillionaire LeBron James declared last year that Black people as a group are “terrified” even to leave home because of the risk of police misconduct and white violence. The hard edge of the political right engages in a fair amount of this as well, with websites and books like White Girl Bleed a Lot and Stuff Black People Don’t Like joining the decades-old VDare and American Renaissance in chronicling almost every attack on a white pensioner.

This is all, again, empirically nonsensical. According to the 2018-19 Bureau of Justice Statistics Criminal Victimization report (table 14), the vast majority of crime is intraracial, with 70% of attackers targeting Blacks and almost 65% of those targeting whites being individuals of the same race as their victims. Violent interracial incidents involving Blacks and whites, on the other hand, make up roughly 3% of total crime: During 2018-19, there were 20,828,040 “Index” violent crimes or serious property crimes in America, 607,726 of which were violent offenses involving a white offender and a Black victim or a Black offender and a white/Caucasian victim.

What’s more, of those 607,726 violent offenses between Blacks and whites, 547,948 (90%) were Black-on-white crimes, versus 59,778 white-on-Black. That 90% figure, reported during a year that witnessed significant racial tensions, is more than a little atypical. But interracial crime involving only Blacks and whites has been more than 70% Black-on-white in every recent year on record. This is, in some sense, not even particularly surprising: There are five times as many whites as there are Blacks in America.

These examples do not stand alone. Almost every time the narrative of systemic racism is subjected to empirical analysis, it collapses. The core claim of the systemic racism (or “systemic bias”) school in academia is that pervasive racism must still exist within modern society because measurable and fairly large gaps between races persist in variables such as income. At least according to such thinkers as Boston University’s Ibram X. Kendi, the “pervasive racism” explanation must be accurate because the only other potential explanation for these gaps in outcomes would be the unacceptable specter of genetic inferiority.

While some more-than-residual racism obviously remains in American society (a 2015 Gallup poll found that 8% of Americans would not vote for a well-qualified Black presidential candidate from their own party), the second part of Kendi’s famous argument is simply wrong. Roughly the same level of bias exists against many groups—Jews, Hispanics, Blacks, women—which nevertheless have dramatically different outcomes. Adjusting for other important variables that differ among racial and ethnic groups like age and region of residence, the gaps glibly attributed to prejudice all but close.

The discontinuity between reality and perception is by no means confined to the political left, as the popularity of the QAnon movement and bizarre election fraud theories demonstrate. In fact, the anti-empirical trend seems to be picking up steam on all sides. The past few decades have featured a slew of books with titles like The Culture of Fear, Apocalypse Never, False Alarm: The Truth about the Epidemic of Fear, and my own Hate Crime Hoax, all aimed at poking holes in nationwide panics about phenomena like plane crashes, stranger kidnappings, racial conflict, and the immediate effects of climate change.

We live in a very postmodern era when it comes to fear and nonsense more generally, with the “signifier” (constant media-driven panic about literally everything) bearing no resemblance to the “signified” reality of continually improving life in our technological era. In reality, virtually all metrics measuring Western and global life expectancy, IQ, governmental corruption, poverty, racism, women’s rights, and so forth have been improving for decades—but no one seems to know it. This rather technical and academic point matters: The more our baseline assumptions are based on (forgive me) bullshit, the less capable we are of addressing real-world problems in real time. The COVID pandemic provided a very real example of this conundrum. According to the consulting firm KEKST-CNC, as of July 2020, Americans on average believed that 9% of our population had died during the pandemic, and at least liberal Americans believed that 50% of all those infected with the virus wound up in hospital or the ER.

In reality, and without minimizing the tragic consequences of COVID, it appears that, as of July 2020, the infection fatality rate (IFR) for the virus was actually between 0.26% and 0.6%, and and that the hospitalization rate for those infected was under 10% even among the minority of officially recorded cases. What’s more, about 1 in 1,000 Americans died from the virus during the first year of the pandemic. It is also worth noting that the average age for a COVID fatality in the West has been well above 70, which is consistent with the first data that came from Italy in early 2020. Confusion about these empirical facts explains at least some of the enduring American and European support for COVID lockdowns in the face of major studies showing that well-done nonpharmaceutical interventions work about as well, and of direct evidence from counterexamples like lockdown-averse states such as Florida and Texas.

The risk of misdirected action based on faulty inputs is even larger when it comes to highly politicized issues, in which one or both sides have an incentive to lie or misdirect. The debate about how best to “stop AAPI hate” provides the best recent example. Following a genuine surge in shocking street attacks against people of East Asian heritage, many prominent activists and media figures attempted to tie the violence to “white supremacy” and held major Black Lives Matter-style marches to oppose this boogeyman. The actual data provides, quite literally, no support for the argument that white racism has motivated most anti-Asian attacks.

Notably, most of these attackers haven’t been white. According to the most recently available national crime data that includes Asian American victims, 27.5% of violent attackers were Black, 24.1% were white, 21.4% were Latino and “other” combined, and 24.1% were Asian. In a separate data set focused only on the 98 most prominent recent attacks on Asian Americans, a research associate and I found that 29% of the attackers who were identified in racial terms (24 out of 84) were white, while 71% (60 out of 84) were people of color.

Virtually all metrics measuring life expectancy, IQ, governmental corruption, poverty, racism, and women’s rights have been improving for decades—but no one seems to know it.

In the climate of year 1 AF (After Floyd), the flow of politicized information, particularly from the left, seems likely to intensify for some time. Things will get worse before they get better. Today’s hot-button issue—the teaching of various forms of critical race theory (CRT) in colleges and secondary schools—is directly relevant to this point. While it is often defended by supporters as “just teaching about racism,” the reality is that CRT (broadly defined) is very specific: It is among the most extreme left-wing paradigms of legal and historical analysis.

Per Richard Delgado, co-author of Critical Race Theory: An Introduction, the core claim of the paradigm is that racism is “everyday” and everywhere, and that most race-neutral policies and systems are in fact designed to oppress, evidence of which is the disparity in group performance. As we have seen, much of this is empirically, demonstrably, false. But in practical terms, today’s political fight between the woke and anti-woke centers on whether to teach it anyway. If we do, the result will almost certainly be a generation of high school and college graduates able to recognize the Weathermen but not the Wright brothers, burdened with incomplete ideas about American society and history that rest largely on false assumptions.

Citizens can resist this outcome in particular, and the tide of nonsense in general, by keeping a few baseline principles in mind. First, logic and data are not discipline specific: If arguments on the left or right, or within a specific field such as education, seem nonsensical to you, they probably are. Second, academic credentials are useful as signals of probable IQ, but are very often—outside a few specialized fields like medicine—no more than that. Experts are no more immune than anyone else to extreme political bias or irrationality, and it is worth remembering that there are many experts, and they disagree about everything. Finally, most factual, primary-source information is rather easy to find. If you are curious about whether the media is overhyping the national crime rate, simply search the phrase “BJS NCVS national crime data,” or pull the same report out of a local archive, and see it for yourself.

The truth will set you free, and most of it can be found in a library.

Wilfred Reilly, a political science professor at Kentucky State University, is the author of Taboo: 10 Facts You Can’t Talk About.