Page 15 - IJPS-2-1

Anastasia Kostaki, Javier M. Moguerza, Alberto Olivares and Stelios Psarakis

age-specific fertility rates for the populations of Sweden, Norway, and Denmark (1996 and 2000);
Belgium (1993 and 1995); Greece and Italy (1995 and 2000); UK (1992 and 2000); Ireland (1995
and 2000); the white and black populations of the USA (2003); and for Spain (1942 and 1963). The
empirical data sets were obtained from Eurostat New Cronos database (http://www.eui.eu/Rese-
arch/Library/ResearchGuides/Economics/Statistics/DataPortal/NewCronos.aspx). Additionally, single
year age-specific fertility rates for the US were derived for the 2003 Natality Data Set, obtained after
a request from the US National Center of Health Statistics (http://www.cdc.gov/nchs/). Cohort data
are also used for Spain for the generations born in 1943 and 1962, obtained from the Eurostat New
Cronos database. It should be noted that even for cohorts not yet completed, Eurostat provides esti-
mates of the fertility rates for older women by using the rates observed for previous generations,
without waiting for the cohort to reach the end of the reproductive period. Parity-specific birth rates
were computed as occurrence exposure rates based on parity in marriage.
Then, we applied SVMs and kernels and also provided the fits of the alternative parametric models
to these data sets, the latter initially calculated by Kostaki and Peristera (2007). In populations with no
apparent early-age hump, except of kernels and SVMs, the fits of Hadwiger, Gamma, and Beta models
(Chandola, Coleman, and Hiorns, 1999; Hoem, Madsen, Nielsen et al., 1981), P-K model (Kostaki

Figure 4. Empirical and graduated q x -values, French males, 1991. Figure 4A. Empirical and graduated q x -values, French males, 1991.

Figure 5. Empirical and graduated q x -values, Japanese males, 1990. Figure 6. Empirical and graduated q x -values, Swedish males, 1981–1985.
International Journal of Population Studies | 2016, Volume 2, Issue 1 9