Abstract
Recent analyses of Third World data, both at the level of national or other large aggregates and at that of individuals studied in sample surveys, have revealed the surprising fact that social characteristics, such as the level of schooling or fertility control, or cultural characteristics, such as ethnic group, are usually more influential in determining mortality levels than is access to medical services, income, or nutritional levels. Evidence from the United States at the beginning of the century suggests that this was not the case earlier in the West. This article examines the evidence, shows why developing countries are currently in an unusual situation, and presents anthropoligical evidence on how cultural, social, and behavioral factors achieve their impact. An attempt is made to begin the construction of a more general theory of mortality transition.