Page 167 - Contributed Paper Session (CPS) - Volume 3

CPS1982 Dmitri J. et al.
Even in countries with nearly complete vital statistics this problem may exist.
The main source of underestimation of infant mortality in such cases is a
restrictive definition of live birth. Perhaps the most cited example of such
definition is the Soviet definition of live birth that existed with some
modifications till (Anderson and Silver, 1986).

6. Impact of migration at working/reproductive ages on mortality and
fertility estimates
As we discussed above, accurate annual population estimates are still a
problem in the demographic statistics. While reliable census data are available
in almost all developed countries, the annual estimates require much more
efforts to reach sufficient data quality. Usually each new census brings
inconsistency between the census results and earlier post-census population
estimates based on the previous census and vital and officially registered
migration events over the inter-censal period. In the previous section we
discussed situations when statistical offices do not recalculate the annual
inter-censal populations back from the newly available census. In some cases,
however, official inter-censal estimates exist but their quality is much worse
that one may expect. If there no additional or alternative annual data are
available, it may be better to use the inter-censal estimates calculated using
the standard HMD method based on the assumption of a uniform distribution
of migration across time. If the overall level of migration is not too high, such
estimates seem to be an optimal choice.
Noteworthy, mortality data are relatively insensitive to biases in the
population denominator due to migration since mortality at ages of maximum
migration is low. But this is not a case for fertility data. Quality of the
population exposure is much more important for fertility indicators.

7. Mortality estimates at old ages
The old age population in developed countries has increased very rapidly
throughout the second half of the 20th century. Improvements in survival are
pushing it up to new limits: today more than half of all males and two thirds
th
of all females born in Western countries may reach their 80 birthday. The
proportion of centenarians has increased by about ten times over the last
th
thirty years, and more and more people celebrate their 100 birthday (Robine
and Vaupel, 2001). The importance of high quality mortality data increases
with every decade. Despite it, internationally comparable high quality
demographic data on old-age populations remain insufficient. The HMD is the
only demographic database which provides such data. Population estimates
for ages 80+ in the HMD are recalculated using extinct/almost extinct cohort
and survival ratio methods (Wilmoth et al., 2007). But even such extensive

156 | I S I W S C 2 0 1 9