Page 164 - Contributed Paper Session (CPS) - Volume 3

CPS1982 Dmitri J. et al.
Discontinuities in the population trends are due to several factors. First, it
seems that population estimates have not been recalculated backward using
the latest censuses. Thus, the available data series consist of post-censal
estimates rather than the inter-censal population estimates. Second, there was
a significant amount of unregistered emigration (especially during the period
from 1989 to 2010) not accounted for in the official statistics (Glei et al., 2015).

Figure 1. Trends in the official (left panel) and HMD (right panel) population
by sex, Bulgaria, 1961-2011. Source: (Glei et al., 2015)

The standard HMD inter-censal method is not applicable to the period
from 1985 to 1992 because of an irregular pattern of out-migration. During
1985 to 1988, international migration was very restricted in Bulgaria. After the
collapse of Communism in 1989, the Government lifted the ban on free
movement abroad, which led to mass emigration (mostly of the Turkish
minority) over the next several years (IOM, 2003). After analyzing additional
data (including indirect migration estimates), the HMD team decided to split
the 1985-1992 inter-censal period into two sub-periods: 1985-1988 (a period
of stable and negligible migration) and 1989-1992 (a period of substantial out-
migration). The official population estimates were used for the former period,
but new population estimates were calculated for the latter period.
Specifically, the year 1988 was treated as a “pseudo-census point” for the start
of the interval and the 1992 census counts were used to end the interval. The
standard HMD inter-censal method was then used to derive adjusted annual
population estimates (Glei et al., 2015).

3. Numerator-denominator bias
Problems with registration of out-migration together with relatively well
functioning system of vital registration within country and inability to get
information about vital events related to citizens living abroad may lead to a
numerator-denominator bias. In this case a formally de facto population turns

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