The History and Future of African Census Analysis Project (ACAP)¶
Tukufu Zuberi; Martin W. Bangha (2006). The History and Future of African Census Analysis Project (ACAP). Paper presented at the session 'Censuses: New trends, new formulas', Chaire Quetelet 2006, Louvain-la-Neuve, Belgium, 29 November-1 December 2006. University of Pennsylvania (Population Studies Center).
| Type | conference paper |
| Scope | Comparative and methods |
| Method | synthesis review |
| Period studied | 1652–2006 |
| Regions | Sub-Saharan Africa, Southern Africa, South Africa, SADC region |
| Themes | methods data sources, demography urbanisation |
| Citation key | zuberi2006 |
Argument¶
The paper argues that African censuses have been gravely under-utilised, badly stored and poorly documented, so that census results have not informed policy or development planning. It surveys the history of census taking in Africa from colonial headcounts and hut-tax estimates to post-independence censuses, and compares Anglophone and Francophone practice. It presents the African Census Analysis Project (ACAP) at the University of Pennsylvania as the remedy: recovering old census micro-data tapes, archiving them in the Pan-African Census Explorer (PACE), building African research capacity, and publishing a multi-volume series. It is an institutional and promotional account, not an analytical study of the economy or of apartheid.
Evidence and method¶
Compilations of census dates from national statistics office websites, the ACAP data navigator, Domschke and Goyer (1986), Groupe de Démographie Africaine (1988) and UN (1992); ACAP's own archive inventory; the authors' account of recovering the 1977 Malawi census tapes; and lists of ACAP-linked dissertations and publications.
Findings¶
-
Census data in Africa have been grossly under-utilised, with the focus on enumeration and publication rather than quality evaluation or further analysis. (p. 2)
the history of census taking in most of Africa is characterized by gross under-utilization of census data, the primary preoccupation having been the execution of the processes of enumeration
-
Early colonial headcounts were designed for fiscal and military or labour purposes, and were crude estimates for the African population. (p. 5)
these enumerations were mainly administrative censuses or simple headcounts that provided estimates or assessment of the residents of the colonies for fiscal purposes and/or eventual defense needs
-
Colonial population estimates relied on hut or tax-roll multipliers, which varied by country. (p. 6)
This factor varied from country to country, from 3.5 in Zimbabwe through 5 in Tanganyika, 6 in Algeria to 7 in Sierra Leone
-
Early census content was limited by its tax and labour-force motive, with only broad age groups collected. (p. 6)
In line with the taxation and labor force motivation only very broad age classifications were used.
-
Many post-independence censuses have been lost, and some countries have lost all of theirs. (p. 9)
having taken a census and losing the data is worse than not taking at all. The two Congos are in this situation of having lost all their censuses.
-
Poor metadata makes surviving census data hard to interpret. (p. 10)
It is still a major problem trying to figure out how most computations and imputation were made in the data even for recent censuses because these are rarely preserved.
-
ACAP's archive holds a large collection of census micro-data, including South African and former homeland censuses. (p. 14)
The expanding PACE collection currently consists of census micro-data of 55 censuses from close to 26 African nations. ACAP has the largest collection of complete census data in the world
-
Anglophone African countries have held censuses more regularly than Francophone ones, which drift towards 15-year intervals. (p. 12)
most countries of Anglophone Africa have conducted at least 4 censuses and these are generally within close range of the recommended interval of 10 years.