What’s wrong with Thomas Sowell?

Photo by NeONBRAND on Unsplash

There is something of an obsession among my fellow evangelicals with an economist named Thomas Sowell. I’ve seen his name trotted out all over the place — most recently, Thaddeus Williams’s book Confronting Injustice without Compromising Truth calls him “the other St. Thomas.” Williams repeats a claim I’ve heard many times: he’s never heard anyone engage the arguments of Thomas Sowell (or any number of other Black conservatives). Many of Sowell’s avid fans, of which the Internet contains multitudes, make the conjecture that this is because people on the left are afraid of Sowell’s no-nonsense, fact-based challenges to their arguments.

To a degree I also think it is a shame that there are so few counter-arguments to Sowell in public and in accessible language. To be sure, reviews, rebuttals, and critiques of Sowell’s work have appeared in academic publications, but these are frequently not within reach to people without academic affiliation, and compared to other conservative favorites such as Jordan Peterson or Ben Shapiro, relatively few popular critiques of Sowell are to hand. However, I don’t share the assessment of Sowell’s fans that this is because of his unmatched brilliance. Sowell’s arguments are, in the main, sophomoric in construction and ideologically resistant to intervention from the real world. Despite his bluster about Evidence and Facts that purportedly come to knock down the house of left economic and racial ideas, Sowell is unique even among the conservatives he’s usually cited with for his immunity to real knowledge and his social-scientific sophistry.

Given the gargantuan volume of Sowell’s popular writing, I’m going to need to focus on the central cluster of claims Sowell is known for. This is essentially aggregated from a variety of Sowell’s work, since much of it is repetitive—the primary books I’m addressing are Discrimination and Disparities and Black Rednecks, White Liberals, combined with the vast expanse of Sowell video interviews and clips available on the Internet. I’ve broken down what I see to be this central cluster into its constitutive claims in order to deal with them in sequence:

  1. Disparities do not prove discrimination, particularly pertaining to Black-White economic inequality in the US.
  2. Black Americans, especially the worst-off living in urban centers, have a “redneck” culture that was handed to them by White southerners via Britain. It is this culture which produces bad behavioral patterns, such as crime and single parenthood.
  3. These behavioral patterns are exacerbated by the interventions of the welfare state.
  4. It is these behavioral patterns from the combination of culture and welfare that lead to Black-White economic inequality.
  5. Black people are blameworthy for their inequality-producing behavior.

Each of these claims, as well as the argument as a whole, are riddled with problems of argumentative logic and empirical evidence. In many cases even just one of them would tank the entire enterprise; the sum of all of them is utterly damning.

Sowell is correct that intentional racial discrimination (according to Sowell’s classification, Discrimination 1b and 2) at a given juncture — say, racial discrimination by an employer — cannot fully explain Black-White racial disparities in economic outcomes. It does not follow, however, that therefore the remaining racial disparity not explained by acute racial discrimination is not caused by racism in society. Sowell concludes that, for instance, employers and realtors and bankers will make choices about hiring or real estate or loans based on the relevant qualities the individual brings to the table, such as education, credit scores, criminal or eviction history, and so on (this is what he calls Discrimination 1a). People have differences in the quantity and quality of these they can bring to the table, and thus it is perfectly reasonable to find inequalities in economic outcomes.

What is not answered by this, however, is why these inequalities would be unevenly distributed by race. It’s certainly not a realtor’s fault that there is a Black-White disparity in credit score, but that difference is not a natural fact of the universe. I and most other social scientists believe that there is inherited inequality from the entire history of American social life that at least in part accounts for why the distribution of these sorts of things are unequal by race. The literature on inherited racial inequalities in sociology, economics, and history is simply massive and cannot be hand-waved away. As it stands, Sowell’s argument on this point is hopelessly endogenous.

This argument is also wrapped up with the claim that any government intervention intended to reduce disparities will result in unintended consequences that will most likely make things worse for the people you’re trying to help. This is, in the terms of Albert Hirschman, the “perversity thesis” — the common conservative argument that whatever change you want to implement in society will actually do the opposite of what you want it to do. The reason why this rhetorical move has power is because it seems intuitively true, and sometimes policies do have consequences of this kind. But it simply does not logically follow from the example that some such policies backfired that any such policy inevitably will.

Where Sowell thinks underlying disparities originate from, rather than inherited inequality, is inherited culture. This argument is perhaps one of the worst ones Sowell makes and betrays an unbelievable historical and social-scientific ineptitude. The basics of his claim is that British Americans introduced “redneck” culture into the South prior to the Civil War, this culture was transmitted to Black people, and that it was brought to Northern cities with Black migrants. This, he says, explains why Black people commit so much crime and have such high rates of poverty, single motherhood, and unemployment.

This argument is rife with historical and conceptual problems. For instance, if “redneck” culture accompanying Black migrants to Northern cities was the cause of increases in crime in those cities, why did homicide rates increase after the second wave of the Great Migration, but not the first? Why wasn’t there a similar racial disparity in crime in the South, where Black people were moving to cities from? Why did crime rates only begin to rise in the 1960s (at the same time low-skilled Black unemployment rates began to soar)? Why did the rise in single parenthood that coincided with the rise in crime and unemployment wait until some 20 years after the second wave of the Great Migration to take effect if we’re to believe that this is all caused by “redneck” culture, passed along in the 18th and 19th century and not taking full effect until the second half of the 20th?

Sowell’s theory of culture is also incredibly bizarre. To read his argument, you would imagine culture to be a free-floating thing, unaffected by material circumstance, passed unilaterally from one group to another and retained, unchanged, until you strip it off and put on a new one. Culture does not work that way. At the very least, culture adapts to material circumstance, especially culture regarding how one should behave in order to be successful in life. If your group is presented with a series of poor economic chances in historical succession culminating with the segregation into neighborhoods of concentrated unemployment following de-industrialization, it would be perfectly rational that your group would develop a culture adapted to that economic environment. There is little reason to believe Sowell’s theory of unchanged cultural traits passing unilaterally to a population, while the theory of cultural adaptation to material circumstances has much to recommend it given the actual historical record.

I’ve previously written an entire post about this argument of Sowell’s — that the welfare state helped create the behavioral pathologies that maintain Black-White inequality. There simply is not a lot of evidence for this, and even where it may have merit — a welfare cliff when you get married is bad — it cannot go the whole way to explaining the persistence of these disparities. On the other hand, the historical record of mass Black migration to Northern cities followed by housing discrimination and de-industrialization has immense explanatory power here, and Sowell basically doesn’t acknowledge that possibility.

I want to briefly point out that this is exactly the sort of adaptive theory of culture and behavior that Sowell ignores when it’s not convenient to his underlying ideological assumptions. Why would it be that Black culture would adapt to the material conditions of the welfare state but not the material conditions of, say, housing discrimination and deindustrialization leading to unemployment? The rational conclusion is that Sowell wants to downplay the history of material deprivation and inequality that led to the behaviors he wants to condemn and magnify the role welfare had to play in those behaviors because he has an ideological predisposition to oppose welfare, being a Chicago-school libertarian economist.

One of the big payoffs to this entire line of Sowell’s argument is that culture and behavior are the cause for the continuation of Black-White disparities in economic outcomes. Setting aside the continuing problem Sowell has with endogeneity (how do we determine that differences in culture and behavior are not shaped by present and past discrimination?), Sowell also has a major empirical obstacle with this argument. The best recent scholarship on poverty has adopted a new framework called the “prevalences and penalties” framework that tackles the question of whether the prevalence of behaviors considered to be poverty risks (the four commonly agreed on are single motherhood, young head of household, unemployed head of household, and below-high school education) actually explain the level of poverty in society. Sociologists David Brady, Ryan Finnigan, and Sabine Hübgen found that bringing single motherhood to zero would only result in a dip of US poverty of 1.3 percentage points. Even worse, this is only because the US has one of the highest penalties for single motherhood among rich democracies. In some countries, due to their better-developed welfare states, a single-mother household is no more likely to be poor than other households. A recent talk by David Brady cites from two pre-publication papers which I do not have permission to cite, but which both go into more specifics about the effect reducing single motherhood would have on racial disparities in poverty — and the findings are not promising for Sowell’s thesis.

This brings us to one of the big problems with cultural and behavioral explanations of economic inequality. Someone’s culture and behavior is only one side of the transaction; the distributive institutions of a society are the other. A behavior is only a risk for poverty if the distributive institutions of a society make it so. Imagine a teacher is giving an exam and has decided to grade on a curve. She already knows exactly what proportion of the class will get a given letter grade. That’s the distributive institution. The students study to varying degrees, some studying quite hard and others slacking off. A social scientist could observe the studying behaviors of each student and then look at who received what grade as a result. They would most likely find that the students that worked most diligently got the best grades and the students who slacked off got the worst grades. But it would be completely incorrect to conclude that how much the students studied determined what grade they received. The teacher set up a distributive institution such that even if everyone behaved perfectly (studied for the test), everyone would still not receive an A.

Let’s think of the real-world example of unemployment and health insurance. Unemployment is a big risk for being uninsured in the United States because we primarily distribute health insurance through employers. However, in (for instance) the UK, Canada, or Finland, there is no relationship between unemployment and health insurance because they distribute health insurance to everyone. This is also why there are countries where single motherhood is not a significant risk for poverty: their welfare states provide much more comprehensive family benefits. We have chosen to structure our distributive institutions such that unemployment, low education, and single parenthood are risks for poverty — and such that even if no one in our society had any of these risks, we would still have high poverty and racial inequalities in poverty. Even if Sowell’s arguments had survived to this point, this simple descriptive reality would render them entirely moot.

We finally come to the last piece of Sowell’s argument: the normative claim that Black people are at least in part (to Sowell’s mind, in large part) to blame for their economic position in society by nature of their inequality-producing behavior. The fact that very little poverty can be explained by the typical poverty risk behaviors can lead us to the conclusion that even if we conceded this point, the amount of Black-White economic inequality in poverty that can reasonably be explained by the behavior of Black people is quite a lot smaller than Sowell imagines. But we do not have to concede the point of Black blameworthiness, either. Here I commend Adaner Usmani’s debate with Glenn Loury on the subject of the persistence of racial inequality and summarize the crux of Usmani’s case against Black blameworthiness for Black-White economic inequality. If we take Sowell’s argument to be the case — that Black culture produces behaviors that further their disadvantaged economic position — this is not something created by any individual Black American, but rather inherited as a set of cultural circumstances. Can they be blamed for acting in a way that gains them what anyone is looking for — social acceptance and getting their needs met — within that cultural setting? Borrowing from philosophers Christopher Lewis and Julia Markovits, Usmani argues that a person of average willpower would, in aggregate, behave the same given that cultural environment as the average Black person. To say otherwise is, ultimately, to claim that Black people as a whole have behaved more poorly given their inherited culture than other people would. To make such a move would be to assign an essential racial inferiority to Black people. This is only a brief and thus truncated summation of Usmani’s entire subtle and compelling argument, and I certainly recommend watching the entire debate.

I believe this claim of the blameworthiness of Black people for their economic position is why many of Sowell’s critics call him a racist. It is not always a knee-jerk reaction to disagreement, but is rather based on the implicit logic of saying Black people are blameworthy for their own circumstances. There is a microscopically thin line between saying that Black people are inherently inferior and are thus blameworthy for their circumstances and saying Black people have a bad culture and are thus blameworthy for their circumstances. Sowell tries to get around this by saying that they received but did not create “redneck” culture, but even if they had received that culture from other people, the argument in the paragraph above would completely pertain. I don’t think Sowell intends to espouse racist views, but I don’t think it is mere hysteria that brings people to call Sowell a racist, even if that’s not what I would say.

Thomas Sowell, I think it is fair to say, is first and foremost a pundit. He has made his career less on scholarly arguments accountable to the rigorous critique of his peers and more on quotable quips, book-length tirades, and debate clap lines for the adulation of his libertarian fans and conservative think tank colleagues. Even though they are hollow when you knock on them, Sowell presents his arguments with confidence and frames the story as being one of an incompetent, mean-spirited economic left against a sensible, evidence-based economic right. When that is a story you already believe, Sowell’s arguments appear compelling, and his demeanor is confident and charismatic. But ultimately, Sowell is better at rhetorical flourish than thoughtful empirical analysis or philosophical consistency. I worry that too many Christians assume with Sowell that those on the left are simply ignorant and naive or resentful, even while failing to recognize the devastating problems with every part of Sowell’s program.

PhD student in sociology at Ohio State University studying religion, capitalism, and race in the US. Cofounder, Evangelical Labor Institute.