Page 165 - Contributed Paper Session (CPS) - Volume 3

CPS1982 Dmitri J. et al.
into a de jure population, since it includes people registered in the country but
living abroad, while vital events produced by these people are being registered
only if these events occurr within the country.
Moldova provides an example of substantial numerator-denominator bias.
This is despite a functioning population register. The registration of deaths
and births in Moldova covers only the events that occur within the territory of
the country (the de facto population), whereas the population estimates
include Moldovan citizens who live abroad (Penina et al., 2015). In addition,
Moldova experiences a very high level of out-migration. About 490,000 people
left the country between the censuses of 1989 and 2004, and 322,000 people
left between the censuses of 2004 and 2014. Thus, Moldova lost close to
800,000 its residents due to out-migration, or more than 20% of the total
population, as enumerated by the 1989 census. The official net migration
numbers are far more modest: 206,000 and 17,000 for the two inter-censal
periods, respectively (Penina et al., 2015). All this (hidden) migrants are still
registered in Moldova but the corresponding vital events are not included in
the official population statistics of the country. This bias produces substantial
differences between the de facto and the de jure population numbers.
(Penina et al., 2015) proposed an alternative population estimates based
on unofficial data of the 2004 census which refer to the de facto population
instead of the official counts of people registered as residents of Moldova.
According to the alternative population estimates, net migration was equally
redistributed over the inter-censal years 1989-2004 and 2004-2009. From
2009 onward, the annual net migration was estimated from the border
crossing migration statistics. As one may expect, the crucial point by this
reconstruction is availability of border crossing data, which is quite unusual for
modern statistics. It allows obtaining more reliable estimates of migration.
It is often assumed that error due incorrect counting of migration has a
rather minor impact on the aggregate mortality indicators such as life
expectancy at birth because it accumulates around the youngest and the most
mobile population groups, which have relatively low mortality. In Moldova,
corrected population estimates are 18 percent lower than the official estimates
(figure 2). The adjusted estimates of life expectancy at birth in 2014 were 64.94
years for males and 73.74 years for females. Compared to the official estimates
these figures are by 2.58 years and 1.65 years lower, respectively.

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