A TEST OF DURKHEIM THEORY OF SUICIDE - WITHOUT COMMITTING THE ECOLOGICAL FALLACY

The data adduced by Durkheim in support of the association between rel igion and suicide have seldom been subjected to scrutiny; when they ha ve been so examined, the scrutiny has been based, of necessity, on dat a subject to the ''ecological fallacy''. Data for the Netherlands, rou ghly contemporaneous with Durkheim's, that have recently come to light allow us To test the statistical support for Durkheim's theory about religion and suicide without risk of committing this ''fallacy''. We f ind the Catholic-Protestant differential in suicide rates to be explic able entirely in terms of the practice of categorizing as ''sudden dea ths'' or ''deaths from ill-defined or unspecified cause'' a large prop ortion of deaths among Catholics which would have been categorized as suicides had they occurred among Protestants. This finding raises doub ts not only about Durkheim's theory bur also about other causal theori es concerning suicide that rely on a sociological rather than a psycho logical (or even idiosyncratic) explanation.