Redirect that data

Race is a political myth, argues Tukufu Zuberi, and it is time we used racial statistics to prove it

The idea of a cause-and-effect relationship between race and social status emerged from colonialism and the enslavement of Africans and Native Americans. But while Africans were enslaved, Europeans were arguing for democracy.

The period of Enlightenment was distinguished by the establishment of European colonies in Africa, Asia and America. When Africans were emancipated, colonisation and segregation ruled the day. These apparent contradictions needed justification, so the study of racial statistics was born and gave a scientific credibility to racial inequality.

This past week in South Africa, the world gathered at the World Conference against Racism, Racial Discrimination, Xenophobia and Related Intolerance. Racial data played an important role. There was much talk of data on health, poverty, wealth, teenage pregnancy, crime and imprisonment - all differentiated by race and presented as the result of politically neutral scientific methods.

Racial data are never politically neutral. The history of racial statistics is inextricably linked with that of social statistics. Justifying racial stratification required more developed statistical techniques than those employed in other areas of science. The early success of Francis Galton, Karl Pearson and Ronald A. Fisher in establishing statistics as a science - so giving birth to modern social statistics - was the first effort to legitimate statistical racism. All three were eugenicists, and eugenics provided a theoretical context for biological and social statisticians to employ racial data for understanding society.

The second world war caused them to reframe their arguments, but since 1950 there has been a renewed interest in the relationship between evolution and social science. At first, this tended to be presented without the overt racial implications of the politically incorrect eugenics, especially its Nazi manifestation. Ultimately, however, although most of human genetics has been freed from the eugenics movement and might be seen as an "independent" scientific vocation, racial statistics continues to reflect its classical eugenics legacy through, for example, the work of Arthur Jensen or Charles Murray.

Academics still use race as a causal factor of social differences. Widely read eugenics books, such as The Bell Curve by Murray and Richard Herrnstein, present arguments on the benefits of using a biological model to interpret racial data and attempt to explain social differences in crime, wealth, test scores and social dislocation by reference to presumed intellectual aptitudes.

Such research makes the case for social policy that supports racial stratification. It has also helped cultivate a public policy environment that has reduced governmental support for child rearing among the poor in the United States.

The conventional European eugenic explanation - not as blatant as the US version - is that racial groups differ statistically in average cognitive ability and that these differences largely account for racial differences in education and economic success.

However, by employing racial statistics and interpreting race as a cause, well-intentioned scholars often legitimate the use of methodologies that perpetuate the problems they seek to solve. The leading journals are full of examples in which academics resort to racial statistics in an effort to refute racist arguments or vindicate past misdeeds.

Race cannot be a cause because race does not exist. The problem with eugenic-based research is that it presents race as being capable of causing a social difference and hence being a determining factor in society. Non-eugenic research would not focus on race as a cause. William G. Bowen and Derek Bok's work on the positive effects of affirmative action, for example, shows how social policy affects racial stratification.

Race is an idea created to perpetuate racial injustice, not a physical reality. It does not depend on science, but political decisions and personal prejudices. It is not biological; the human genome project proved that. Likewise, our prejudices about society guide how we interpret racial data. However, these problems with racial classification should not be confused with the need for racial statistics to solve the legacy of white racism. Whether we ought to authorise the collection and analysis of racial data is a political question.

Scientifically, it would be better to phase out racial data and promote a policy of deracialising society rather than supporting a strong multiracial illusion. But in our racially divided world, this would only allow us to ignore the legacy and reality of racial discrimination and prejudice - it would be like putting the cart before the horse.

A first step could be to change how we interpret racial statistics. Although they have been used to justify racial stratification, recasting them for racial justice has benefited both science and racialised populations.

Tukufu Zuberi is director of the African-American studies programme at the University of Pennsylvania and author of Thicker Than Blood: How Racial Statistics Lie , published in October by the University of Minnesota Press, £20.50.