Testing Statistical Identity
In a recent decision about the new law on immigration control, integration and asylum in France, the Conseil Constitutionnel (an institution that verifies that laws voted in Parliament are compatible with the French Constitution) rejected the parts in which the law authorized the production and use of official demographic statistics on ethnicity (some related blogging in French here and here). SOS Racisme (a French association whose main goal is to fight racial discrimination), expressed an enthusiastic approval of the decision. This came after an important campaign against “ethnic statistics” in France in general, and more precisely against a survey on “trajectories and origins” that was being prepared jointly by the INSEE (the official French institute for statistical survey) and the INED (the French institute for demographic studies). Many issues were raised indeed about how surveys (especially official ones) on ethnic conditions, national origins or religious beliefs could eventually feed policies marked by the raise of national identity in Sarkozy’s agenda, or could also have performative effects in fostering racial categorization or communitarian identification (this being at odds with some French supposedly republican ideals).
The controversy is complex. A numerous group of researchers observes that this decision hampers scientific inquiry. They accuse SOS Racisme of mistaking grounds for a struggle against racial discrimination (knowing discrimination would be a first step in fighting against it) but also, above all, of mixing science with politics: see their counter-campaign here and here (see also an article in Le Monde here opposing this other one here, both by researchers from the INED). Their arguments in defense of research are most reasonable. But it’s perhaps a pity that they say that “it’s always dangerous when political actors want to have their say in defining what science ought to be”. We all want to know (“we” refers here to researchers, citizens and politicians altogether) and we may all have a word to say about the how. Incidentally (and despite probably gross epistemological and methodological stopgaps, of course), SOS Racisme proposes a generalization of testing as a research tool. Which is interesting indeed — interesting for the sake of scientific experimental knowledge. (For the polity of this sort of experimental science, see for instance Peter A. Riach and Judith Rich “Deceptive Field Experiments of Discrimination: Are They Ethical?”, Kyklos, 57, 2004, pp. 457-470.)