Page 58 - IJPS-2-2

Latent class models for cross-national comparisons: the association between individual and national-level fertility and partnership characteristics

GDP. General social expenditure is high, although the support for childcare as a proportion of GDP
is somewhat lower than in class 2 and consistent with Eastern European levels. Female labour force
participation is in excess of 70%, albeit slightly lower than in class 2. The school entry is the lowest
seen in all classes.
In general this class demonstrates a high degree of support for cohabitation within the legal
frameworks of countries. Cohabitation is mentioned in nearly one third of relevant laws, and legally
equivalent in nearly 35% of these. The existence of a legally recognised cohabiting relationship is
highly prevalent for countries in this class, with a near universal existence of a formal non-marital
union within the law.

4.2 Individual Characteristics

Growth curves for three demographic processes were extracted, with the growth curve varying by
class. The growth curves for the probability of marriage are presented in Figure 3. Broadly speaking,
the overall pattern of marriage is consistent across all classes, increasing from relatively low levels at
early ages, and peaking around the early 30s. Thereafter, there is some evidence of the decline of the
probability of being married, due in the most part to the dissolution of marital unions. There are
some differences in the overall levels of marriage between classes, as well as marriage timing. Class
1 (Eastern European pattern) shows the overall highest propensity for women to have a marital union,
with the highest probability of marriage at all ages. Class 2 (Western Europe – limited support)
presents a more mixed progression. The probability of marriage is low at younger ages; 10% points
lower than in class 1 at age 16. However, the rate of increase across ages is rapid and increases to the
extent that the probability of marriage is nearly at the level seen in class 1 by age 36.
Class 3 (Western European – extensive support) shows the overall lowest propensity for marriage,
which is consistent across the life course. The overall probability is low at early ages, consistent with
the levels seen in class 2. However, in class 3 the increase in the probability of marriage is not as
dramatic, and the peak in the probability of marriage is slightly below 70%, in comparison to classes
1 and 2 which are both in excess of 85%. Overall then, class 3 is associated with the lowest preva-
lence of marital behaviours, while classes 1 and 2 show greater recourse to the institution of mar-
riage. That said, class 2 is associated with some postponement of entry into marriage, while class 1
is characterised by high levels of marriage at early ages.
The estimated growth curves for cohabitation are presented in Figure 4. The probabilities of coha-
bitation are somewhat lower than they were for marriage (Figure 3) reflecting the preponderance of

Figure 3. Predicted growth curves for probability of marriage for Figure 4. Predicted growth curves for probability of cohabitation for
Western extensive and limited and Eastern classes. Western extensive and limited classes.

52 International Journal of Population Studies | 2016, Volume 2, Issue 2