The Risk of Assuming Discrimination Causes All Gaps

Thomas Sowell's 2018 book "Discrimination and Disparities" challenged the idea that economic disparities between racial groups are caused primarily by discrimination. That debate remains unresolved, and its implications for regulation and corporate behavior are significant for investors.

Robert Nozick coined the term "normative sociology" to describe the tendency to decide what the causes of problems ought to be and then find a story to support that conclusion. "If X is bad, and Y which is also bad can be tied to X by a plausible story, it is very hard to resist the conclusion that one causes the other," Nozick wrote. The essay cites Nozick's observation that people want one bad thing to be caused by another, and they leap to that conclusion.

The same logic applies in reverse. The cause of poor economic outcomes is never allowed to be something the observer dislikes, such as individual choices or structural factors unrelated to discrimination. Socialists, the essay notes, insist that price controls should not cause supply problems. The cause must be greedy profiteers.

Sowell's work tested the discrimination hypothesis by measuring other variables – educational achievement, family structure, employment rates – across racial groups. He found that the statistical evidence of disparities does not, by itself, establish causality. "The mere omission of one crucial fact can turn accurate statistics into traps that lead to conclusions that would be demonstrably false if the full facts were known," Sowell wrote.

Progressive critics accused Sowell of "correlation hunting" and victim blaming. Jennifer Dolec, in a review of "Discrimination and Disparities," argued that the correlation between having a criminal record and difficulty finding a job is a mystery. She suggested that the government should ban employers from discriminating against people with criminal records until the cause is understood. "If employers' discrimination against people with records is due to simple animus, then we'll need to increase the cost of such discrimination," Dolec wrote.

That example illustrates the risk for investors. A policy premised on the assumption that discrimination causes employment gaps – without considering the individual's credentials or criminal record – could impose new compliance costs on companies. Hiring quotas, expanded background-check restrictions, and mandated diversity spending are all regulatory possibilities that would affect labor costs and operational flexibility.

The essay highlights a double standard. The same economists who dismiss Sowell's correlations as irrelevant to causation never apply that standard to their own focus on measuring gaps. The World Economic Forum claims to have "empirical methods that help show causality" and "actual causal impact of a new policy," but the essay notes that the WEF simply measures different variables that correlate with the outcome being studied – the same type of work Sowell did.

Frank Shostak, an Austrian economist, provided the framework: "Correlations among the various pieces of historical data cannot establish causality. Correlations can only describe, but not explain. We hold that causality can be ascertained by means of establishing the definition of the subject of investigation."

For investors, the risk is that regulation based on flawed causal assumptions will increase costs and reduce market efficiency. The debate over discrimination and economic outcomes is not academic. It directly shapes the policy environment for hiring, pay, and corporate governance. Companies that face new mandates tied to a "normative sociology" view of causation will bear the compliance burden, regardless of whether the underlying assumption is correct.

Sowell's work provides a framework for evaluating those risks. The question is not whether discrimination exists. It is whether the assumption that discrimination is the primary cause of disparities leads to policies that do more harm than good. The evidence suggests that ignoring other variables – education, family structure, behavioral factors – produces regulations that target symptoms rather than root causes.

"Correlations can only describe, but not explain," Shostak wrote. That principle applies to all sides of the debate.