122
more likely to build this type of as-family network than individuals born
in the older cohort. Regarding gender, women have a higher chance than
men of having an as-family network focusing on their siblings.
The Beanpole type of network is predicted by country of origin and by
birth-cohort. The significant effect of country is transversal to the three
models, meaning that national context is a strong predictor of this type
of network. Individuals born in Switzerland have a much lower chance of
having a Beanpole type of as-family network than individuals born in
Portugal. The introduction of cohort effects also increases the explanation
power of the model significantly: from R^2 = 0.06 to R^2 = 0.30. Individuals
who were born between 1970 and 1975 are less likely to show this type
of as-family network.
The Nuclear-daughter and Nuclear-son are both predicted by country
of origin, but the statistical effects of this factor in the case of these
types of nuclear networks are notably weaker. Individuals born in
Switzerland are less likely to show the Nuclear-daughter configuration
than individuals born in Portugal; whereas those who were born in
Lithuania are more likely to build the Nuclear-son when compared to
the Portuguese. Meaningful explanations for this type of effect, even if
the latter is weak, are however difficult to find. We cannot presume that
individuals in mid- life in Switzerland would elect their sons rather than
their daughters as close family persons; on the contrary, given stronger
female responsibility for care and household tasks, we might expect the
opposite effect. In this case then, it would be important to gather more
data, for example on the demographic reservoir of these individuals or
other variables, such as the migration of young men and women, before
attempting to interpret these effects. The Nuclear-son type of network is
also predicted by birth- cohort as individuals born between 1970 and
1975 have a lower chance of having this type of as-family network,
probably, as mentioned above, as they have not yet made the transition
to parenthood. Finally, regarding the Restricted-nuclear configuration,
we can see that country has no significant effect. Instead, birth-cohort
and gender are relevant predictors of this configuration. Belonging to
the youngest cohort and being a woman decreases the chances of hav-
ing this type of as-family network.
K. Wall et al.