Uses of Population Censuses and
Household Sample Surveys for Vital
Statistics in South Africa
United Nations Expert Group Meeting on
International Standards for Civil Registration and
Vital Statistics Systems
New York, 27-30 June 2011
1
Main Sources of Data
• Civil registration systems not functioning well
in majority of African countries
– Population censuses and household sample
surveys are primary sources of demographic data
• In South Africa vital statistics are mainly
derived from population censuses and
household sample surveys
2
Vital Statistics Information Collected
in Censuses: 1996-2011
Births
1996




Total number of children ever
born
Number of children surviving
2001


Total number of children ever
born
Number of children
surviving
2011


Total number of children ever
born
Number of children surviving
Date of birth of first child
Births in the past 12 months

Number of children living
elsewhere
Number of children dead
Age at first birth

Date of birth of last child born



Date of birth of last-born child
3
Vital Statistics Information Collected in
Censuses: 1996-2011
Deaths
1996


Survival status of mother
Survival status of father
2001







2011
Survival status of mother
Survival status of father

Survival status of last-born child
Deaths in the past 12 months

Deceased died from accident or
violence
Month and year of death

Deceased died while pregnant or
within 6 weeks after delivery
(females of reproductive age)

Deceased died (i) while pregnant; (ii)
during birth; (iii) within 6 weeks of
delivery (females of reproductive age)

Date of death of last-born child



Survival status of mother
Survival status of father
Survival status of last-born child
Deaths in the past 12 months
Deaths due to natural or non-natural
cause
Month and year of death
4
Vital Statistics Information Collected
in censuses: 1996-2011
Marriages and Divorces
1996

Present marital status
2001

Present marital status
2011

Present marital status
5
Observations on Questions asked
in the Censuses
• Total number of children ever born and surviving as well as the survival
status of the parents were asked in all three censuses
• The number of questions pertaining to vital statistics fluctuated from one
census to the other
• Improvements in asking information on certain questions (e.g. asking 3
questions on maternal mortality in 2011 instead 1 question in 2001)
• Similar questions were asked differently, e.g., date of birth of first child in
1996 and age at first birth in 2011
• Responses for present marital status in 1996 combine divorced and
separated as one category while in 2001 and 2011 they are two separate
categories
• Inclusion of the cause of death question in 2001 and 2011
6
Vital Statistics Information Collected in
Household Sample Surveys: 1994-2010
Vital
Event
Type of Survey
October Household
Survey (1994-1999)
Births



General Household
Survey (2002-2010)
Children ever born
Survival status of
child
Community Survey
(2007)


Birth registration
status



Pregnant past 12 months
Status of pregnancy

Total number of children
ever born
Number of children
surviving
Number of children living
elsewhere
Number of children dead
7
Vital Statistics Information Collected in
Household Sample Surveys: 1994-2010
Vital Event
Type of Survey
October Household
Survey (1994-1999)
Deaths





Age of child at death
Deaths in the past 12
months
Death registration
status
Cause of death
(natural, accident,
violence)
Survival status of
parents
General Household Survey
(2002-2010)
Community Survey (2007)







Survival status of last born
child
Survival status of mother
Survival status of father
Deaths past 12 months
Month and year of death
Cause of death
(unnatural, natural)
Deceased died while
pregnant or within 6
weeks after delivery
8
Use of Population Censuses and Household Sample
Surveys for Vital Statistics
• Census data as denominators in calculation of vital rates and ratios for
vital statistics investigated in the census, e.g., for crude birth and death
rates
• Population projections from census data are denominators for vital rates
and ratios from vital statistics from other sources, e.g., vital registration
system to calculate annual crude birth and death rates
• Number of births and deaths, including due to AIDS are estimated from
census data and/or combined with other source
• Census and survey data are used to evaluate completeness of birth and
death data from civil registration system
• Using indirect demographic techniques, Community Survey data were
used to estimate levels of fertility
9
How the Censuses and Surveys Meet the Core Vital
Statistics Requirements
• Vital statistics produced are limited to births and deaths
• Information on marriages and divorces is based on present marital status
with no indication of when event took place and when it got registered
Therefore cannot determine annual marriages and divorces
• Information is not continuous especially from the census. Therefore
seasonal, monthly, quarterly or annual information is not available
• There is no consistency in information collected over time both for
censuses and surveys
• Detailed cause of death information is not available from either censuses
or surveys
10
How Censuses and Household Sample Surveys
Complement Register-based Statistics
• Background:
– Register-based statistics include data from the national population
register and health services records
• Uses:
– Census and household sample surveys used for the calculation of birth
and marriage rates from the National Population Register
– Data also used to estimate the completeness of birth registration
– Data from the National Population Register on deaths used to verify
data from the Civil Registration System
– Health services records (number of births and immunization statistics)
used to evaluate the number of births from the National Population
Register
11
Conclusion
• Population Censuses and Community Survey have been the
main sources of data from vital statistics in South Africa
• They have also been used to complement statistics from the
civil registration system and the population register mainly for
estimating completeness of death and birth registration
• Population Censuses and Community Survey do not
completely meet the core vital statistics requirements
• There is need to start utilizing information from the civil
registration for vital statistics in South Africa
12