Until not long ago, scanty information on the profile of the Vietnamese population and
the poor quality of the available data have complicated efforts to present reliable
estimates of mortality levels in Vietnam. Only recently, have official data on Vietnamese
mortality become available through the 1979 and the 1989 population censuses. This
paper makes use of such data to estimate Vietnam s mortality levels during the intercensal
period.
The problems posed by census data to estimate mortality are well known. First,
one must rely on the reported age distributions. In Vietnam, little is known about the
accuracy of age reporting, except that, by virtue of Vietnam s proximity to the East-Asian
cultural sphere, age reporting is considered to be fairly accurate. There is little evidence
of heaping from the single-year age structure of the 1989 census,1
and only a small
amount of exaggeration at the oldest ages (80+) is seen in the 1979 census age structure.2
Second, enumerations in the two censuses must be equally complete. Yet, the population
at the time of the 1989 census was very mobile as a result of loosened controls under the
economic reforms of the 1980s,thus complicating the effort to enumerate persons away
from their usual place of residence.