TCS Daily

Scientific Prejudice

By Iain Murray - April 1, 2002 12:00 AM

"Unequal Treatment" is the latest report from the Institute of Medicine (IOM), the medical arm of the National Academies and it made headline news across the country because it authoritatively attributed some of the known disparities in healthcare between the races to active racial bias.

"The report found that blacks ... have been deprived of medical care that most whites take for granted," said USA Today ("Study: Racial disparities persist in medicine," Mar. 20). It is almost certainly true that this has happened to some extent, but the emphasis placed on racism in the coverage of the report may be overstating the case.

Much of this disparity in treatment was attributable to insurance variations, income, age, medical history and other such factors. However the report concluded that "stereotyping, prejudice and clinical uncertainty on the part of healthcare providers may contribute to racial and ethnic disparities in healthcare." Notice the word "may." The word is there because this report was not a comprehensive quantitative analysis. Instead it reviewed all the recent studies on racial disparities and placed them in a context of a broad pattern of racial disparities in other areas. The finding that racial bias is a factor in healthcare is a qualitative conclusion, not one backed up by any single undeniable piece of statistical evidence.

The study is at its weakest when it ventures beyond medicine into other areas of racial disparity. It asserts that there are racial disparities attributable to bias in mortgage lending, for instance, when those disparities are almost certainly driven by socio-economic factors such as savings habits and marital status. In fact, few, if any, lenders actually know the race of their applicants, having just names or even social security numbers to go by in addition to the financial information they are supplied with.

The report also alleges that it is proven that minority youth receive harsher treatment in the criminal justice system. This may, however, be an artifact of how crimes are categorized. More white youths are arrested for the minor offense of simple assault and more black youths are arrested for the serious crime of murder, but both crimes are categorized as "personal." It should not be surprising, therefore, if white youths receive less harsh handling than black youths do overall under that broad heading.

The context of institutionalized racism is therefore much less clear than is indicated by the report. Another prime example comes from figures released last week by New York State. The findings show that there is a serious disparity in school test results between white and Asian students on the one hand and black and Hispanic students on the other. However, at no point was this attributed to racial bias, with researchers instead preferring to point to a "complicated interplay of school, family and social influences that can be tricky to untangle." This is hardly surprising. The idea of ascribing institutionalized racism to New York public schools and teachers seems rather far-fetched.

It must therefore be asked whether the IOM researchers have done enough to untangle the complicated interplay of healthcare provider, economic and social influences that make up the web that surrounds the racial disparities in healthcare. To an extent, they admit they have not. They admit that the evidence for bias, stereotyping and prejudice is circumstantial.

This is a wise admission, for previous claims to have found definite statistical evidence of bias have not been able to stand up to scrutiny. For instance, a 1999 study on heart disease treatment published in the New England Journal of Medicine (NEJM) concluded that the less sophisticated tests ordered for minorities were most likely due to unconscious bias about gender and race. But a closer look at the study data revealed that the physicians who evaluated the black and female patients who were less aggressively treated were themselves more likely to be black and female than those who evaluated the white and male patients. The study also found that all physicians were more likely to order stress test data for black patients than they were for white patients. Finally, on a measure of physicians' assessment of the personal characteristics of their patients, white male patients were consistently marked the most negatively. After these points were made, NEJM's editors admitted that "although racism and sexism are prevalent in American life, the evidence of racism and sexism in this study was overstated."

That conclusion is likely to ring true for the larger IOM study as well. The authors identified other possible factors that might be driving the disparity, such as language and cultural barriers, a lack of medical resources such as drugstores in minority communities, and a scarcity of physicians from minority groups. Racial bias and stereotyping may well contribute to the disparity as well, but a lot more work needs to be done before we can identify how much of an influence they are.

The President of the National Medical Association told the New York Times that the study proved "that racism is a major culprit in the mix of health disparities." The study did nothing to prove it a major culprit. However, the weight of the evidence certainly seems to lean towards it being a clearly identifiable factor. The IOM would do well to sponsor rigorous research aimed at quantifying just how important it is.

TCS Daily Archives