African Households: Censuses and Surveys (A General Demography of Africa) (review)

African Households, edited by the late Etienne van de Walle, is the second volume in the series A General Demography of Africa. This collection is intended to showcase studies carried out under the African Census Analysis Project, a collaborative initiative between researchers at the University of Pennsylvania and African institutions specializing in demographic data collection, research and training. ACAP focuses specifically on the use of African census data for research and policy purposes, and has developed a bank of census data collected in sub-Saharan Africa over the past 30 years.

This volume presents a collection of papers initially discussed in an internet-based conference in November 2001. The conference focused on two areas: 1.) comparative research on household structure, and 2.) analyses of relationships between household members. Accordingly, the book is divided into two sections, the first Approaches to the Recording of Household Structure, and the second Intra-household Relationships. While these analyses focus mainly on national census data, some reports from "demographic surveillance systems" are also included. A DSS comprises longitudinal data collection on a set population or group of households, with intensive follow up of member households over time.

A central debate in household survey research is the definition and nature of households themselves. As noted in van de Walle's Introduction to the collection, researchers differ as to the most appropriate definition of a household, and more fundamentally, over whether the concept of a household accurately reflects the social reality of Africa. This volume usefully summarizes the viewpoints of different researchers on this topic, and then extends the debate by offering case studies drawn from the experience of field-based data collection.

To some extent, the debate over the definition reflects the viewpoints of different disciplines. For demographers and others whose aim is to survey populations, the household serves as a basic statistical and administrative unit for measurement. In contrast, anthropologists view the household as a social unit, and one that is inherently dynamic. To some, these differences are reflected in the tendency of anthropologists and social scientists from similar disciplines to see society as socially constructed, and dependent on lineage and family ties, versus a model that derives from the economic theoretical tradition. Thus, a household is the basic social unit which encapsulates kinship, residence patterns and economic organization. Who is a household member is not merely a statistical question, but rather an important social question. The household is not merely a unit, but rather the basic element in a broader social organization.

Papers by Hosegood and Timaeus, and Townsend and colleagues address this debate through the lens of household-level research in two rural sites in South Africa. South Africa offers a somewhat unique example of household composition and organization because the entrenched system of labor migration means that many households include non-present members on a nearly permanent basis. If a census is limited to counting people at one point in time, then how are these people to be accounted for? Both papers present innovative approaches to addressing this question.

In spite of the centrality of the "household" debate, the volume raises other important points regarding household organization and social structures. Comparisons are made between East and West Africa. As Garenne and colleagues note in a chapter on Senegal and the Gambia, French colonial influence wrought profound changes in administrative and social organization throughout West Africa, which facilitate the conduct of censuses and study of household organization. This perspective and attention to regional variations within the African sub-continent are important features of this volume, which provides a sense of a collective African census enterprise and highlights and distinguishes some of the important regional differences. From a demographer's standpoint, these differences become important in discussions of cross-regional or cross-national comparative analyses.

The second half of the volume gives attention to intra-household relations, presenting micro-level data drawn from a variety of sources. While not as cohesive and...