Principles and Recommendations for Population and Housing Censuses, Revision 3216
5.8. Household deaths in the past 12 months
133
(core topic)
4.250. Information on household deaths in the past 12 months classified by sex of deceased
and age at death is used to estimate the level and pattern of mortality in countries that lack
satisfactory continuous death statistics from civil registration. In order for estimation derived
from this item to be reliable, it is important that all deaths to household members occurring
during the 12 months preceding enumeration be reported as completely and as accurately as
possible. Typically, reports of deaths in censuses underestimate the overall number of deaths
if only because some deaths result in the disintegration of households so that household sur
-
vivors, if any, may not report their occurrence (in particular, deaths of persons living alone at
the time of death are unlikely to be reported). Nevertheless, provided that there are no serious
errors in the reporting of age at death, estimates of completeness of death reporting can be
derived via indirect estimation and adequate mortality estimates obtained.
134
4.251. Ideally, information on mortality should be collected for each household in terms of
the total number of deaths in the 12-month period prior to the census date. For each deceased
person reported, name, age, sex and date (day, month and year) of death should also be col
-
lected. Care should be taken to clearly specify the reference period to the respondent so as
to avoid errors due to its misinterpretation. For example, a precise reference period could be
defined in terms of a festive or historic date for each country.
4.252. When information is collected on household deaths in the previous 12 months (or
some other reference period), countries may wish to ask a pair of follow-up questions concern
-
ing cause of death. After ascertaining the name, age and sex of the deceased person and date
of death, two additional questions could be asked:
(a) Was the death due to an accident, violence, homicide or suicide?
(b) If the deceased was a woman aged 15
135
to 49, did the death occur while she was
pregnant or during childbirth or during the six weeks after the end of pregnancy?
4.253. Data derived from such questions can help to assess trends in levels, and some causes,
of adult mortality. At the data-processing stage, reported deaths can be tabulated accord
-
ing to broad categories of cause of death: external, pregnancy-related, other and unknown.
Ignoring the “unknown” responses, both external and pregnancy-related deaths can provide
valuable information in countries where no other sources of information to systematically
obtain causes of death are available. Of course, such information is approximate and must be
interpreted with caution after careful evaluation and often adjustment. Nevertheless, using
these simple questions should make it possible to derive some useful information about major
trends in mortality that are otherwise difficult to obtain.
4.254. ere is no universal agreement about the feasibility of collecting reliable cause of
death information as part of a population and housing census. More research is needed on
both the feasibility and methods of collecting cause of death information as part of a national
census.
5.9. Maternal or paternal orphanhood
136
4.255. Some countries may also wish to collect information on maternal or paternal orphan-
hood in another attempt to ascertain the level and patterns of mortality in the population.
Census data from these two topics are intended for indirect estimation of mortality by sex.
Estimates are based on the proportion of persons classified by age whose natural mothers or
fathers are still alive at the time of the census.
4.256. For the collection of information on orphanhood, two direct questions should be
asked, regardless of whether or not the mother and father are enumerated in the same house
-
hold, namely:
133
See Handbook on the Collection of
Fertility and Mortality Data, Stud-
ies in Methods, Series F, No. 92
(United Nations publication, Sales
No. E.03.XVII.11), available from
http://unstats.un.org/unsd
/publication/SeriesF/SeriesF_92E.
pdf; Methods for Estimating Adult
Mortality (United Nations publica-
tion ESA/P/WP.175), available from
www.un.org/esa/population
/techcoop/DemEst/methods
_adultmort/methods_adultmort
.html; and Rob E. Dorrington,
“The Brass Growth Balance Meth-
od and the Preston-Coale Method
for One Census, and the General-
ized Growth Balance Method and
Synthetic Extinct Generations
Methods Upon the Availability
of Deaths from Two Censuses”, in
Tools for Demographic Estimation,
T.A. Moultrie, R.E. Dorrington,
A.G.Hill, K. Hill, I.M. Timæus and
B.Zaba, eds. (International Union
for the Scientific Study of Popula-
tion, 2013), available from http://
demographicestimation.iussp.org.
134
See chapter 4 on methods for
data evaluation and adjust-
ment in WHO, WHO Guidance for
Measuring Maternal Mortality from
a Census (Geneva, World Health
Organization, 2013), available
from http://apps.who.int/iris/bit
stream/10665/87982/1/97892415
06113_eng.pdf; and T.A. Moultrie,
R.E. Dorrington, A.G. Hill, K.Hill,
I.M. Timæus and B. Zaba, eds.,
Tools for Demographic Estima-
tion (International Union for the
Scientific Study of Population,
2013), available from http://demo
graphicestimation.iussp.org.
135
It may be appropriate in some
countries to reduce the lower age
limit by several years.
136
For methodological details
on the uses of the data,
see I.M. Timæus, “Indirect
estimation of adult mortality
from orphanhood” in Tools for
Demographic Estimation,
T.A. Moultrie, R.E. Dorrington,
A.G. Hill, K. Hill, I.M. Timæus
and B. Zaba, eds. (International
Union for the Scientific Study of
Population, 2013), available from
http://demographicestimation
.iussp.org/content/indirect
-adult-mortality-orphanhood);