africahh

The

AFRICAN HOUSEHOLD

Virtual Conference

An ACAP Production

November 21-24, 2001

Last updated 09/26/2002

PRESENTATION OF THE CONFERENCE

The availability of individual-level census data for a series of African countries presents a double challenge, methodological and substantive.
How reliable and usable is the information? What kind of question on relationship should be included in the census to maximize the information that can be used, while not burdening the questionnaire with unusable data? How can the information on individuals be best integrated into an examination of the next unit of aggregation, the household?

The household provides a level of analysis between that of the individual and that of the administrative unit. It has properties and characteristics that transcend those of the individuals, and that are recorded in censuses: type of housing materials, of means of livelihood, of access to public services such as piped water or electricity. Another set of household characteristics -size, structure, characteristics of the head- can be derived from the information on individuals. Finally, individual data provide us with information on other individuals in the family that may of may not be in the same households: spouses (the marital status question), children (the children ever born question), and parents (the orphanhood question).

At the substantive level, African reality presents us with an endless variety of patterns of residence and family structures. We know very little about the typology of households structures encountered in various countries. Simple descriptive studies are often provided in isolation, for instance for a village. Censuses provide us with an opportunity to describe and compare household structures over the continent, and to answer simple, but important questions such as:

Where are the children? The women? The aged?

Beyond the purely descriptive, the chapter in the planned volume will explore patterns of association between household structure and various demographic or social characteristics.

Common approaches.

1) Use the household as a unit of study, and the individual as a constituting element of the household. This contrasts with an approach were the focus would be on the characteristics of the head of household, or even one where the relationships would be analyzed only at the aggregate level.

2) This is likely to involve a scrutiny of the structure of individual households, and in an exploratory first stage, this is best achieved graphically, using common ideographs such as those proposed by Laslett and his colleagues. (See the background of this page for an example.)

3) It will be necessary to link the information on the household with the information collected at the individual level. Moreover, information collected on individuals (such as the information on marital status, on children not living with their mothers, or on orphanhood) can reveal important structural characteristics.

4) The introduction of the volume will include a series of recommendations on what the best aggregate indicators of household structure are.
It is obvious, for example, that the sex and marital status of the head of household is a important item of information, and at the aggregate level, we want to know what proportion of households are headed by a non-married woman. There are other tabulations that may provide simple indices of important social characteristics of the population. For example, the number of households including children not living with their parents (an index of fosterage) or the number of married men of women in the household (either an index either of polygyny, or of household complexity.)

5) We want to recommend a typology of households for the analysis of census data. Can we find simple indicators that will highlight the differences between household types in Tanzania and the Gambia, or between the Mandings and the Fulani in Senegal? Arriving at such a typology will go a long way in answering the fundamental question of the book: what relationship questions should be asked in future censuses to maximize the usefulness of the information?

6) One should never lose sight of the transitory nature of households, and of their life cycle. The age of the head of household, and of the other members, is an essential explanatory factor.

Common questions

1) Questions of quality. What was attempted by the census design, and what was achieved?

2) How does the censuses deal with the issue of the next intermediate category aggregating several households, the compound or concession?
Does it clearly distinguish it from apartments in a building?

3) Where are the children? The women? The old people?

4) How has polygyny been handled? When a husband is co-resident with his several wives, the problem is mostly one of finding the way to represent it ideographically, or analyze the polygamous household in an integrated fashion. When each wife has its residence that the polygynous husbands visits in turn, however, the problems are different, and it is not clear that an appropriate solution has been offered in existing censuses.

Material details about the conference

Common e-mail address: africahh@pop.upenn.edu
[Participants are encouraghed to send ideas, suggestions, references, or even gossip, to be shared among the participants. If you send a message to the above address, it will be received by all the participants.]

Date: 21-24 November 2001.

Methodology: The Virtual Conference will take place in November of 2001. At that time, every participant should be close to a computer outlet, and be willing to participate in the discussion of his own and other people's chapter during a week of concentrated attention.

Participants: We have invited a mix of established scholars and young researchers to contribute papers. Basic requirements to take part in the conference were: interest in household issues, access to a census tape, and competence to analyze it. (Other types of census-like documents are also featured in the papers.) Once the papers have been prepared, and are posted on the web, we can open the conference to a wider audience.