Microsoft Word Socialized Choices 31-12 pod docx

Socialized Choices - Labour Market Behaviour of Dutch Mothers 
118 
Independent variables: demographic and situational factors 
To allow for objective background characteristics, several control variables are 
included. These variables allow weighing the importance of attitudinal factors as 
compared to objective characteristics. Moreover, the path analysis enables us to 
distinguish between the (direct) influence of these background characteristics on 
women’s labour market behaviour and the indirect influence through her work 
preferences.  
Educational attainment. Many studies have shown that the higher the 
educational level attained, the higher the labour market participation of women 
(Merens et al., 2011). Higher educated women more frequently continue to work 
after giving birth than low educated women, because the former earn a higher 
wage, which allows them to pay for child-care facilities (Doorewaard et al., 2004, 
p.11). Furthermore, it is known that women’s and men’s educations are positively 
associated with egalitarianism (Kroska and Elman, 2009, p.373,), which might be 
the result of exposure to ideas about equality or the establishment of career-
oriented networks (Cunningham et al., 2005, p.887). Thus, the study expects 
higher educated mothers to (prefer to) work more hours than lower educated 
mothers.
40
 
Number and age of children. It is expected that the larger the number of 
children, and the younger their age, the lower the mother’s (preferred) number of 
work hours will be. The age of youngest child is measured in three categories: 0 
to 4 years, 5 to 8 years, and 9 to 12 years.  
Age. Age may refer to the life phase as well as to the generation (birth cohort) 
of the respondent. In a cross-sectional analysis it is not possible to distinguish 
between age and cohort effects. Recent research has shown that the number of 
hours women work after giving birth to their first child is higher for younger 
generations than for older ones (Lut, Van Galen and Latten, 2010). Younger 
mothers are expected to work more hours than older mothers. Yet, the older a 
mother is, the older her children tend to be, and so her caring tasks diminish. 
Thus, the number of work hours older mothers prefer is expected to be higher 
than the number younger mothers prefer. After a certain age, as mothers approach 
retirement, their preferred labour market orientation may change again (Román et 
al., 2007). To account for this possible non-linear relationship between age and 
labour participation, the variable age squared (divided by 100) is included in the 
analyses.  
Partner. A partner can affect his spouse’s employment decisions in various 
ways: his income, the number of hours he works, his career perspectives, his 
attitude towards her income and her career perspectives, and his acceptance or 
                                                           
40
 The categories are: 1: primary school, 2: VMBO (intermediate secondary education (equivalent to 
US: junior high school); 3: HAVO/WVO (higher secondary education/preparatory university; 4: 
MBO (intermediate vocational education, equivalent to US junior college); 5: HBO (higher 
vocational education, equivalent to US college); 6: WO (university). 

---

Chapter 4 - The vital and stabilising role of work preferences 
119 
rejection of the male breadwinner model may all influence a mother’s labour 
market preferences and behaviour (Hoffnung and Williams, 2013; Kangas and 
Rostgaard, 2007; Van Wel and Knijn, 2006). I expect that cohabiting mothers 
work fewer hours than single mothers, since single women cannot financially rely 
on a partner and, thus, have to work more hours in order to earn a living. And yet 
the preferred number of work hours for single mothers might be lower because 
they cannot share their caring tasks with a partner.  
In line with microeconomic theory, the higher the income of the partner, the 
fewer hours his wife works, because his income is sufficient to make ends meet.
 
41
 Moreover, a strongly career-oriented husband, with high earnings, may work 
long hours and leave the bulk of household responsibilities to his wife, which 
may hinder her labour participation (Cloïn, 2010). However, a high income might 
not necessarily reduce a mother’s work preference, which might be higher than 
one would expect in view of her partners’ income.  
Religion. Religiosity is also expected to coincide with more traditional general 
and personal gender attitudes, since most religions endorse traditional views on 
the role of women (Bolzendahl and Meyers, 2004; Cunningham, 2001; De Valk, 
2008; Kraaykamp, 2012; Thompson, 1991). I thus expect that religious mothers 
have lower work preferences and more traditional general values and personal 
attitudes. The respondents were asked how often, aside from special occasions 
such as weddings and funerals, they attend religious gatherings nowadays: 1: 
every day; 2: more than once a week; 3: once a week; 4: at least once a week; 5: 
only on special religious days; 6: less often; 7: never. 
4.6 Results 
Hypothesis 1:  
A mother’s labour participation is largely determined by her preferred number 
of work hours. 
Participation decision  
Initially, all relevant demographic variables and the number of preferred work 
hours are included in the analysis, in order to examine which variables affect 
whether a mother has a paid job or not. Remarkably, only the number of preferred 
work hours exerts a direct significant effect on participation in paid work (beta 
.673) (table 6). Work preference explains 45 per cent of the variance of the 
participation. In itself, this is not very surprising, but the impact of this variable is 
so strong that the background characteristics of the mother, such as age, 
educational level, the number of children, the presence and the income of a 
                                                           
41
 
  Since we include the log of the income of the partner in the analysis, we imputed a value of zero 
for the log of zero income, which does not exist. To correct for this, we added a dummy variable 
of one when the partner has no income. 

---