Expectations about Fertility and Field of Study among Adolescents: A Case of Self-selection? Expectations about Fertility and Field of Study among Adolescents: A Case of Self-selection? Micha G. Keijer, Aart C. Liefbroer, Ineke Nagel Abstract: In recent studies on the association between education and fertility, in- creased attention has been paid to the fi eld of study. Women who studied in tradi- tionally more “feminine” fi elds, like care, teaching, and health, were found to have their children earlier and to have more children than other women. A point of debate in this literature is on the causal direction of this relationship. Does the fi eld of study change the attitudes towards family formation, or do young adults with stronger family-life attitudes self-select into educational fi elds that emphasize care, teach- ing, and health? Or do both fi eld of study preferences and family-life attitudes arise before actual choices in these domains are made? We contribute to this debate by examining the relationship between fertility ex- pectations and expected fi elds of study and occupation among 14-17 year-old ado- lescents. We use data collected in 2005 from 1500 Dutch adolescents and Structural Equation Modelling (SEM) to examine the associations between expected fi eld of study and occupation and fertility expectations. Our results show that expectations concerning fertility and fi eld of study are already interrelated during secondary edu- cation. Both female and male adolescents who expect to pursue studies in fi elds that focus on care and social interaction (like health care, teaching etc.) are less likely to expect to remain childless. This holds equally for girls and boys. In addi- tion, girls who more strongly aspire to an occupation in which communication skills are important also expect to have more children. We did not fi nd any relationship between expectations of pursuing a communicative fi eld of study and occupation and expectations of earlier parenthood. In addition, among boys, we fi nd that the greater their expectation of opting for an economics, a technical, or a communicative fi eld of study, the less likely they were to expect to remain childless. Boys who expected to study in the economic fi eld also expect to have their fi rst child earlier, but boys expecting to pursue a tech- nical course of studies expect to enter parenthood later. We also found that those who expect to pursue cultural studies are more likely to have a preference for no children, or if they do want children, to have them later in life. Overall, our fi ndings suggest that the processes of elective affi nity between the communicative fi elds of study and work on the one hand and fertility on the other Comparative Population Studies Vol. 44 (2019): 85-106 (Date of release: 21.08.2019) Federal Institute for Population Research 2019 URL: www.comparativepopulationstudies.de DOI: 10.12765/CPoS-2019-11en URN: urn:nbn:de:bib-cpos-2019-11en2 • Micha G. Keijer, Aart C. Liefbroer, Ineke Nagel86 hand are more or less comparable for boys and girls. With respect to the other do- mains, we fi nd, apart from the gender differences in the relation between fi elds of study and childlessness, hardly or no gender differences in the expected timing of parenthood and the number of children. The genders do differ in their level of pref- erence for communicative and economics-related fi elds of study and occupation, but if they do have the same preference, the association with fertility expectations is more or less similar. Keywords: Education · Fertility · Field of study · Adolescent expectations 1 Introduction Educational attainment is generally viewed as one of the key factors explaining late and low fertility (Blossfeld/Huinink 1991; Kohler et al. 2006; Wood et al. 2014). In the past decade, the attention of research on the impact of education on fertility has shifted in some cases from the level of education to the educational fi eld. The fi eld of study chosen by young adults is thought to infl uence their future labour market options and thus their opportunities to combine employment and family life. Women employed in sectors where it is relatively easy to combine employment and family life are likely to have more children and have them earlier than women employed in sectors where it is diffi cult to combine these roles. Both country-spe- cifi c (Lappegard/Rønsen 2005; Hoem et al. 2006 a/b; Begall/Mills 2012; Oppermann 2014) and comparative studies (Van Bavel 2010) have provided evidence for the existence of a relationship between educational and occupational fi elds and fertility behaviour. Generally, these studies show earlier and higher fertility among women who have specialised in educational fi elds that prepare for traditionally “feminine” occupational domains, such as caring, teaching, and health than among women who have studied in another educational fi eld (e.g. Bagavos 2010; Lappegard 2002; Begall/Mills 2012; Oppermann 2014; Tesching 2012). Two mechanisms have been proposed to explain the relationship between fi eld of study and fertility: fi eld-specifi c socialisation and self-selection. Field-specifi c socialisation implies that enrolment in specifi c fi elds of study infl uences students’ family-related attitudes. With respect to fertility, educational fi elds that emphasize “female qualities” lead women to develop attitudes in favour of becoming a mother early in life and having relatively large numbers of children (Van Bavel 2010). Self- selection assumes that girls who have strong preferences for family life would “self- select” into fi elds of (tertiary) education that emphasize traditionally female qualities like care or teaching (Hakim 2003). The key difference between those two explana- tions is that the fi eld-specifi c socialisation argument suggests a causal infl uence of the fi eld of education on subsequent fertility decisions, whereas according to the self-selection argument, the causal relationship is reversed, with fertility preferenc- es shaping adolescents’ choices concerning the fi elds of education in which they want to study and the sectors in which they seek employment. A more nuanced ver- Expectations about Fertility and Field of Study among Adolescents • 87 sion of the self-selection argument is that preferences and choices in these two life domains develop in tandem and that there is no clear causal relationship between the two. Therefore, both processes have been modelled simultaneously (Hoem et al. 2006a/b; Martín-García/Baizán 2006; Tesching 2012; Lutz 2014). Nevertheless, little is known about the potential role of self-selection as this would require data on fertility attitudes prior to enrolment in a specifi c fi eld of study (Begall/Mills 2012; Oppermann 2014). Data on expectations about prospective fi elds of study, occupa- tion and fertility among adolescents would provide more information on whether expectations about educational and occupational fi elds on the one hand and fertil- ity on the other hand are already aligned in adolescence, which can be regarded as indicative of self-selection processes. For this reason, we use data from adolescents (14-17 years old) to examine whether the association between expectations about future fi elds of study and occupation on the one hand and future fertility on the oth- er is already present before choices have been made in these domains. Therefore, the fi rst research question is How are adolescents’ expectations concerning fertility related to their expected fi elds of education and occupation? Almost all studies on the association between fi eld of study and occupation and fertility, with the exception of Martín-García (2009) and Oppermann (2014), focus on women. Martín-García (2009) found that, contrary to women, men who pur- sued health and care-related studies did not have a child at an earlier age than men who studied in other fi elds. Opperman (2014) found no association between fi eld of study and fertility behaviour among men in Germany. Yet men also develop expec- tations about fertility and employment in adolescence and this relationship might differ from that of women. Boys are also subject to gender socialisation and might develop different expectations of their role in family formation than girls (Guetto/ Panichella 2013). Therefore, our second research question is To what extent do gen- der differences exist in the relationship between adolescents’ expectations con- cerning fertility and their expected fi elds of education and occupation? To answer our questions, we use data from more than 1,500 Dutch adolescents aged 14 to 17 who were surveyed in 2005 on their plans for the future, life-course expectations and ambitions. The combination of information on future plans for fertility, educa- tion and occupation is unique. Furthermore, these respondents, who were surveyed in 2005, are currently (2019) in the prime ages in which they actually make decisions on whether and when to have children, and if so, how many. 2 Theory The emergence of females’ preferences concerning professional career and family life One reason to expect that expectations concerning fertility on the one hand and educational and occupational fi elds on the other hand develop early in life is that both sets of expectations are shaped by interactions during childhood and adoles- cence with family and friends (Youniss/Smollar 1985). • Micha G. Keijer, Aart C. Liefbroer, Ineke Nagel88 The family in which one is raised is viewed as the most important socialising institution which shapes an adolescent. Many empirical studies have shown that parental education and occupation affect children’s fi eld of study via a process of intergenerational transmission. These studies fi nd that children’s fi eld of study strongly resembles the occupational fi eld of the parents (Goyette 2008; Van der Werfhorst/Luijkx 2010). Another set of studies, drawing on cultural reproduction theory (Bourdieu 1986), emphasize the role of more general preferences. Bourdieu argued that the parental (cultural) environment is important in choosing a fi eld of study, since children’s choices refl ect standards and practices within the social “fi eld” in which one is raised. The parental home is also known to infl uence people’s fertility behaviour (Thorn- ton 1980). The similarity between parents’ and children’s fertility patterns is often explained by the transmission of values and preferences from parents to children (Barber 2000). Axinn et al. (1994) fi nd that parents’ fertility values and preferences to a certain extent infl uence their daughters’ fertility timing and quantum. Other studies found evidence for observational learning. For instance, Murphy and Wang (2001) show that coming from a large family increases the likelihood of children hav- ing a large family themselves. Based on these processes, expectations concerning both family and career can be assumed to crystallise during adolescence. However, at least two different argu- ments can be put forward which suggest that this crystallisation process leads to associations between specifi c types of educational and occupational expectations and specifi c types of fertility expectations. The fi rst argument for their intercon- nectedness is based on gender role theory (Stockard 2006; Risman/Davis 2013), the second one on ideas about the development of personality and social orientation among adolescents (Woods et al. 2016; Tavares 2016). We will discuss both argu- ments in turn. Stockard (2006) provides an overview of the different theoretical approaches in which socialisation processes lead to the development of gendered identities that infl uence how girls (and boys) refl ect upon their roles in key life domains relating to work and family. Charles and Bradley (2009) argue that the persistence of gender differences in the choice of educational and occupational fi elds is due to gender as a critical part of self-expression and identity, even in modern societies. However, the actual expression of the gender identity of girls may vary substantially, as is suggested in Hakim’s “preference theory” (Hakim 2000, 2003), that offers an expla- nation for the way women combine family life and employment in contemporary so- cieties. Hakim (2003) differentiates between three types of identity that are largely shaped during adolescence. First, some women consider family life and children to be their main priority and reduce their employment time to the minimum. A second category of women prioritises work over family life, and these women devote much of their life to work. The majority of women, however, try to balance work and fam- ily. Hakim (2000) labels these women as adaptive because they have no prevailing preference orientation. If girls develop a preference for such an “adaptive” lifestyle during adolescence, it is likely that they will expect to combine having children and employment. This could lead to the development of an expectation to opt for fi elds Expectations about Fertility and Field of Study among Adolescents • 89 of education and fi elds of occupation that offer good opportunities for combining both roles. This will favour the development of expectations for studies in fi elds like care, health and teaching that can relatively easily be performed on a part-time basis.1 Another explanation for the joint development of expectations concerning oc- cupational and fertility goals in adolescence can be found in the literature on per- sonality development. Personality characteristics have been found to be related to fertility outcomes (Jokela 2012; Skirbekk/Blekesaune 2014; Tavares 2016), choice of educational fi eld (Korpershoek et al. 2010), and occupational choices (Wells et al. 2016; Woods et al. 2016). A high score on agreeableness, for instance, is related to a higher propensity to have children (Jokela 2012; Tavares 2016, but not Skirbekk/ Blekesaune 2012), a low propensity to opt for educational tracks in science and economics (Korpershoek et al. 2010) and a lower propensity to opt for managerial- type occupations (Wells et al. 2016) and a higher propensity to opt for health care specialties that emphasize social interaction (Woods et al. 2016). Thus, adolescent girls who score high on agreeableness might be expected to be strongly oriented towards care and social interaction, and thus to be attracted both to caring for chil- dren and to fi elds of study and occupation in which care and social interaction are central. Based on these theoretical ideas, we expect that, even in adolescence, expecta- tions about the fi eld of study and occupation are related to fertility expectations. This implies that the association between expected fi eld of study and expected fer- tility is, at least partly, already present before entering tertiary education. The devel- opment of expectations in the fi elds of education and occupation on the one hand and in the fi eld of fertility on the other hand could be considered as intertwined, as it is not clear if adolescents opt for a fi eld of study because it accommodates their fertility preference or vice versa. Therefore, we view their relationship during adolescence as essentially interdependent rather than causal, and formulate the fol- lowing three hypotheses: H1a: The greater the preference among female adolescents for fi elds of study and occupation in care, teaching, and health, the less likely they are to ex- pect to remain childless. H1b: The greater the preference among female adolescents for fi elds of study and occupation in care, teaching, and health, the earlier they expect to have children. H1c: The greater the preference among female adolescents for fi elds of study and occupation in care, teaching, and health, the more children they expect to have. To our knowledge, only one study has examined this relationship as yet (Kanji/ Hupka-Brunner 2015). They analysed data of Swiss female adolescents at age 16/17 1 In 2017 of all jobs in the Netherlands, those in health care and teaching have a relatively low number of working hours [https://opendata.cbs.nl/statline/#/CBS/nl/dataset/81433NED/ table?dl=12C3E, consulted October 18, 2018] • Micha G. Keijer, Aart C. Liefbroer, Ineke Nagel90 and at age 23. Their results showed that a strong preference for children (which they termed “Kinderwunsch” (p. 125)) at age 16 made it more likely for young women to enter highly gender-segregated occupations at age 23 than their female peers with a weak preference for children. Women who had a strong preference for children and who were not in employment by the age of 23 were also more likely to be en- rolled in fi elds of education which prepare for occupations with higher proportions of women. Thus, their results clearly support the view that the association between fi eld of study and fertility develops early in life. However, Kanji and Hupka-Brunner (2015) used rather crude measures of fertility preference and expected occupation. They asked adolescents if they expected having children to be very important in their future life, and asked for no further information on timing and quantum prefer- ences. Furthermore, information on fi elds of study and occupation was only ascer- tained at age 23. Male preferences While most research on the association between fertility and fi eld of study has fo- cused on women (exceptions: Martín-Garcia 2009; Opperman 2014, 2017), we ex- amine the relationship between expectations concerning fertility and fi eld of study among both female and male adolescents. Are associations between expected fer- tility and expected fi eld of study and occupation different for male and female ado- lescents? Two diametrically opposed points of view can be put forward. On the one hand, one could argue that, if personality and social orientation are important, boys and girls will show the same level of association. Boys with an interest in caring for others and socialisation of the younger generation are more likely to choose to study in the fi eld of care, health and teaching. On the other hand, one could also argue that preferences for traditionally male occupations among male adolescents, such as in technical and analytical fi elds (Davies/Guppy 1997), are an expression of their male identity (Charles/Bradley 2009). These fi elds usually lead to jobs with a (higher) stable income and good job security which increase the prospects of men in the partner market (Kalmijn 1998). In addition, male earning potential and fertility are often correlated within the traditional breadwinner model (Pascall 2006). According to the patriarchal norm, the breadwinner is a male work- ing outside the home to provide the family income. The breadwinner model sug- gests that male adolescents who prefer this model will combine a preference for traditionally male-dominated fi elds of study, like economics and technical studies, with a preference for earlier and higher fertility transitions. Although we are not completely sure which of these lines of reasoning will apply, for the sake of argu- ment we assume that the latter reasoning will be the stronger one, and therefore formulate the following hypotheses: H2a: The greater the preference among male adolescents for educational fi elds and occupations related to economics and technology, the lower their likeli- hood of expecting to remain childless. Expectations about Fertility and Field of Study among Adolescents • 91 H2b: The greater the preference among male adolescents for educational fi elds and occupations related to economics and technology, the earlier they ex- pect to have children. H2c: The greater the preference among male adolescents for educational fi elds and occupations related to economics and technology, the more children they expect to have. 3 The Dutch case Our study is situated in The Netherlands. The Dutch system of secondary education is markedly stratifi ed. Children are sorted at ages 12 and 13 into a number of verti- cally organised secondary education tracks, and there is relatively little movement between the tracks. First choices for specifi c fi elds of study have to be taken after two or three years in secondary education, leading to horizontal tracking of adoles- cents as well. The Netherlands is an interesting setting for a study on the relationship between fertility and fi eld of study expectations in adolescents, as the reconciliation of em- ployment and family life remains an important issue (Lewis et al. 2008). On the one hand, the Netherlands is a country with liberal gender attitudes and levels of female labour force participation and fertility that are close to the European average. On the other hand, most women work part-time, the timing of the fi rst birth is very late and the costs of formal child care provision are quite substantial. There is also a strong minority who feels that mothers should not work full-time if they have children un- der age 12 (Liefbroer et al. 2015). Thus, in the Netherlands, if men and women want children, choosing fi elds of education and occupation that offer good opportunities for part-time work broadens their options for realising their fertility preferences. The opportunities for part-time work are particularly numerous in the public sector (in- cluding the health and care sector) and in the service sector (Statistics Netherlands 2019a). These factors make the Dutch case interesting for examining the relation- ship between adolescents’ expectations regarding fi eld of study and occupation and their fertility expectations. 4 Method 4.1 Sample The data for this study were collected in 2005 (Ganzeboom et al. 2005-2006) as part of the research project Youth and Culture (Ganzeboom/Nagel 1998-2002) which in- cludes several studies on students in secondary education in the Netherlands. The data were collected in 14 municipalities located throughout the Netherlands. The selected municipalities include two major cities, eight medium-sized municipalities, and four small municipalities. Within these municipalities, 69 schools were contact- ed, of which 60 were willing to participate in the research. All of the various levels • Micha G. Keijer, Aart C. Liefbroer, Ineke Nagel92 of education of Dutch secondary education are represented in the sample. These levels are VMBO-B (lower vocational education), MAVO/VMBO-T (lower general secondary education), HAVO (higher general secondary education) and VWO (pre- university education). Within each school, a sample was drawn from three classes of different level-grade combinations. This procedure resulted in 190 classes of which 148 actually participated in the project. The response rate was 87 percent at the school, and 78 percent at the class level. In each class, the students were asked to complete a survey during a regular lesson (45-50 min). As the data collection was part of a larger project, the classes were randomly divided into two: one half of the class was given a survey on future plans and cultural participation, the other half was asked to complete a survey on computer use. In the end, 1,544 secondary school students (49 percent males and 51 percent females) aged 14-17 (34 percent VMBO, 34 percent HAVO, 33 percent VWO) fi lled out the questionnaire about future life plans. Selective non-response at the student level can be assumed to be lim- ited, as questionnaires were fi lled out during class time and hardly any individuals refused to cooperate. 4.2 Measurement instruments Fertility measurements Fertility expectations of adolescents were measured by asking whether they thought they were going to have children in the future, with possible answers yes or no. In our study, adolescents responding “no” are counted as expecting to remain child- less. Adolescents who responded “yes” were asked at what age they expected to have a fi rst child and how many children they expected to have. Answers to these two questions were used as indicators for expected fertility timing and quantum. Expected fi eld of study Adolescents’ expectations with regard to their future fi eld of study were measured by asking: “If you are thinking about going into higher education, in which fi eld of study do you think that will be?” A list of thirteen fi elds of study were offered: (1) humanities, languages, history or theology, (2) social and behavioural studies, (3) art, (4) economics, commerce, business administration, (5) law and legal services, (6) teacher training or education, (7) medical, health services, nursing, (8) personal care services, (9) agriculture, (10) mathematics or science, (11) technical & engineer- ing, (12) transport, and (13) public order and safety. The adolescents were asked to indicate the likelihood of each fi eld of study on a 4-point scale (1 = defi nitely not, 2 = probably not, 3 = probably, and 4 = defi nitely). Based on these scores, we com- puted composite scores of preference for four general fi elds of study (cultural, eco- nomic, communicative and technical), using a classifi cation developed by Van der Werfhorst and Kraaykamp (2001). Various classifi cations are used in the literature on the relation between fi eld of study and fertility and on sex segregation within fi eld of study, but all differentiate between health and/or care, teaching and relational skills Expectations about Fertility and Field of Study among Adolescents • 93 and other fi elds of study (Hoem et al. 2006a/b; Charles/Bradley 2009; Barone 2011; Morgan et al. 2013; Michelmore/Musick 2014). The classifi cation of Van der Werf- horst and Kraaykamp (2001) focuses on the main type of occupation-related skills developed in a fi eld of study and is also used by Begall and Mills (2012), Martin- Garcia and Baizán (2006) and Martin-Garcia (2009). Like these authors, we adjusted the original coding scheme by deeming a preference for becoming a teacher to be communicative rather than cultural. Cultural fi elds of study comprise humani- ties, languages, history, theology, social and behavioural studies and art. Economic fi elds of study comprise economics, commerce, business administration, law and legal services. Communicative fi elds of study comprise teacher training, education, medical and health services, nursing and personal care. Technical fi elds of study comprise agriculture, mathematics or science, technical and engineering, transport and public order and safety. This compressed educational fi eld categorisation gives a better view of which type of skills the adolescent wants/expects to develop in the future. A preference score for each of the four main fi elds of study was calculated by summing the scores of all the constituting specifi c fi elds of studies and dividing by the number of specifi c fi elds. Descriptive information on the means and standard deviations for all four expected fi elds of study are presented in Table 1. Aspired fi eld of occupation To ascertain what kind of occupations adolescents aspired to, the following ques- tion was posed: “How much would you like to practise the following occupations?” Respondents were presented with a list of 30 different occupations. Answers were scored on a four-point scale: absolutely not, rather not, maybe and gladly. Examples of occupations that had to be scored were physician, lawyer, teacher, truck driver, and hairdresser. To compress these 30 different occupations, they were classifi ed into fi ve broad categories: cultural, economic, communicative, technical and lower (Van der Werfhorst/Kraaykamp 2001; Begall/Mills 2012). This classifi cation largely mirrored the one used to classify fi elds of study. The category “lower” was added, as some of the occupations (for instance cleaner, assembly line worker, shop atten- dant, etc.) asked for few specifi c skills and could not easily be classifi ed into one of the four main fi elds of occupation. Using the same procedure as for fi eld of study, a composite score was calculated for the aspiration of each student with regard to each of the fi ve fi elds of occupation (see Table 1 for means and standard deviations). Educational level Adolescents’ current level in secondary school is used as a control variable in all analyses. The educational level was measured on a scale from 1 (lower vocational) to 4 (pre university). Subsequently, this level of education was recoded using the newly developed International Standard Level of Education (ISLED) scoring system (Schröder/Ganzeboom 2014). ISLED is an empirically obtained, internationally com- parable interval scale for educational level, the value of which can vary between 0 and 100. • Micha G. Keijer, Aart C. Liefbroer, Ineke Nagel94 N o te s: S co re s fo r fi e ld s o f st u d ie s ar e c al cu la te d b y t ak in g t h e m e an s co re s o f al l s tu d ie s w it h in a s p e ci fi c fi e ld . A ll st u d ie s ar e s co re d o n a 4 -p o in t sc al e f ro m 1 ( = d e fi n it e ly n o t e x p e ct t o e n ro l i n t h is s tu d y ) to 4 ( = d e fi n it e ly e x p e ct t o e n ro l i n t h is s tu d y ). S co re s fo r as p ir e d t y p e s o f o cc u p at io n s ar e c al cu la te d b y t ak in g t h e m e an s co re s fo r al l o cc u p at io n s w it h in a s p e ci fi c fi e ld . A ll o cc u p at io n s ar e s co re d o n a 4 -p o in t sc al e f ro m 1 ( = a b so lu te ly w o u ld n o t lik e t o p ra ct is e t h is o cc u p at io n ) to 4 ( = g la d ly li ke t o p ra ct is e t h is o cc u p at io n ). E x p e ct at io n o f h av in g c h ild re n is t h e r e sp o n se t o t h e q u e st io n “ D o y o u t h in k y o u w ill h av e c h ild re n ?” ( 0 = n o , 1 = y e s) . E x p e ct e d a g e a t fi rs t ch ild b ir th is t h e r e sp o n se t o t h e q u e st io n “ A t w h at a g e d o y o u e x p e ct t o h av e a fi r st c h ild ?” E x p e ct e d n u m b e r o f ch ild re n is t h e r e sp o n se t o t h e q u e st io n “ H o w m an y c h ild re n d o y o u t h in k y o u w ill h av e ?” IS LE D is a n e m p ir ic al ly o b ta in e d , in te rn at io n al ly c o m p ar ab le in te rv al s ca le f o r e d u ca ti o n al le v e l, th e v al u e o f w h ic h c an v ar y b e tw e e n 0 a n d 1 0 0 . ** *T h e m e an s o f g ir ls a n d b o y s ar e s ig n ifi ca n tl y d if fe re n t at p < 0 .0 1 . V a ri a b le s M e a n S ta n d a rd d e v ia ti o n N (E x p e c te d fi e ld o f st u d y /A sp ir e d o cc u p a ti o n s o n a s c a le o f 1 t o 4 ) G ir ls B o y s T o ta l G ir ls B o y s T o ta l G ir ls B o y s T o ta l E x p e c te d F ie ld o f S tu d y C a te g o ri se d a s C u lt u ra l 1. 8 0 * * * 1. 4 6 1. 6 3 0 .6 3 0 .5 5 0 .6 2 7 3 7 7 0 2 14 3 9 E x p e c te d F ie ld o f S tu d y C a te g o ri se d a s E co n o m ic 1. 9 9 2 .0 2 2 .0 0 0 .8 2 0 .8 1 0 .8 2 7 3 0 7 0 8 14 3 8 E x p e c te d F ie ld o f S tu d y C a te g o ri se d a s C o m m u n ic a ti v e 2 .0 8 * * * 1. 6 0 1. 8 5 0 .6 0 0 .5 5 0 .6 3 7 3 8 7 0 5 14 4 3 E x p e c te d F ie ld o f S tu d y C a te g o ri se d a s T e ch n ic a l 1. 3 4 * * * 1. 8 1 1. 5 7 0 .4 2 0 .5 7 0 .5 5 7 3 4 7 12 14 4 6 A sp ir e d O cc u p a ti o n s C a te g o ri se d a s C u lt u ra l 1. 7 9 1. 7 2 1. 7 6 0 .6 6 0 .7 0 0 .6 8 7 12 6 8 6 13 9 8 A sp ir e d O cc u p a ti o n s C a te g o ri se d a s E co n o m ic 2 .0 0 * * * 2 .1 5 2 .0 7 0 .7 4 0 .7 7 0 .7 6 7 12 6 8 6 13 9 8 A sp ir e d O cc u p a ti o n s C a te g o ri se d a s C o m m u n ic a ti v e 2 .0 6 * * * 1. 6 9 1. 8 8 0 .6 8 0 .6 4 0 .6 9 7 12 6 8 6 13 9 8 A sp ir e d O cc u p a ti o n s C a te g o ri se d a s T e ch n ic a l 1. 74 * * * 2 .0 4 1. 8 8 0 .5 5 0 .7 1 0 .6 5 7 12 6 8 6 13 9 8 A sp ir e d O cc u p a ti o n s C a te g o ri se d a s L o w e r 1. 4 9 * * * 1. 5 6 1. 5 2 0 .3 2 0 .4 6 0 .4 0 7 12 6 8 6 13 9 8 E x p e c ta ti o n o f h a v in g c h ild re n 0 .9 2 * * * 0 .8 7 0 .8 9 0 .2 7 0 .3 4 0 .3 1 6 9 9 6 6 6 13 6 5 E x p e c te d a g e a t fi rs t ch ild b ir th 2 6 .7 4 * * * 2 7. 6 3 2 7. 13 3 .2 3 3 .9 2 3 .5 7 5 7 0 4 5 0 1 0 2 0 E x p e c te d n u m b e r o f ch ild re n 2 .3 2 2 .2 1 2 .2 7 1. 0 2 1. 0 2 1. 0 2 6 6 2 5 7 3 12 3 5 A d o le sc e n ts ’ c u rr e n t le v e l o f e d u c a ti o n ( IS L E D ) 5 8 .4 4 * * * 5 4 .9 5 5 6 .7 1 12 .1 4 13 .9 1 13 .1 6 7 5 7 74 4 15 0 1 T a b . 1: D e sc ri p ti v e in fo rm a ti o n o n d e p e n d e n t an d in d e p e n d e n t v a ri a b le s Expectations about Fertility and Field of Study among Adolescents • 95 Analysis Strategy Structural equation modelling (SEM) was used to estimate the relationships be- tween adolescents’ expected fertility on the one hand, and expected fi eld of study and aspired occupational fi eld on the other. To account for item non-response in the data, the models were estimated in Mplus using the “full information maximum likelihood” (FIML) method (Muthén/Muthén 2004), in which a cluster correction was applied at the level of school classes. FIML allows us to make maximum use of all the available data in our dataset. It has comparable statistical properties to multiple imputation, but is more effi cient (Allison 2012). All models are evaluated using three widely-used overall goodness of fi t criteria (Root Mean Square Error of Approxima- tion [RMSEA], the Comparative Fit Index [CFI], and the Tucker-Lewis Index [TLI]). The fi t of models is deemed good if RMSEA < 0.06, and CFI and TLI > 0.95 (Hu/ Bentler 1999). First, a logit model with the expectation of remaining childless as the dependent variable was estimated. Next, models with the expected timing of fertil- ity and the expected number of children as the dependent variables were estimat- ed simultaneously among those adolescents who expected to have children. In all analyses, the expectations regarding the educational fi eld and aspirations regarding the occupational fi eld were used as independent variables. In addition, we con- trolled for the adolescents’ current level of education. To test whether the outcomes varied by gender, interaction effects were added between expectations on fertility and fi eld of education and occupation and gender (Hardy 1993). Unstandardised parameter estimates are presented for all models in Table 2. In additional analyses (results not shown), we examined whether the results would change if we did not include the scores on “lower” aspired occupations in the models. Results for other parameters did not differ, suggesting that the results were robust with regard to the in- or exclusion of lower aspired occupations. 5 Results Descriptive fi ndings Separately for boys and girls, Table 1 shows means, standard deviations and number of responses for the variables used in the analysis. Clear gender differences could be observed in the expectations for broad fi eld of study. Girls had a stronger expec- tation to opt for communicative and cultural fi elds, whereas boys had a stronger expectation to opt for technical fi elds. No gender differences were observed for an expectation to opt for economic fi elds. Boys most strongly expected to opt for economic fi elds of study, followed by technical, communicative and cultural fi elds, whereas girls expected to opt for communicative, economic, cultural and technical fi elds, in that order. With regard to aspired occupational fi elds, no gender differenc- es were observed for cultural occupations, but girls aspired to communicative jobs more than boys, and boys aspired to economic and technical jobs more than girls. Among boys, jobs in the economic fi eld were the strongest preference, whereas • Micha G. Keijer, Aart C. Liefbroer, Ineke Nagel96 among girls, jobs in the communicative fi eld were the strongest preference. Con- cerning fertility, most adolescents expected to have children, and fewer girls (8 per- cent) than boys (13 percent) expected to remain childless (t=-3.22, p<0.01). Girls expected that they would have their fi rst child around age 26.7, while boys expected this on average about a year later. The average number of children which adoles- cents expected to have was a little over two children, and did not differ between boys and girls. Clearly, there is a substantial gap between fertility ideals and actual fertility behaviour, because in the Netherlands in 2005 the average age of women having their fi rst child was 29.4 years and the average number of children was 1.71 (Statistics Netherlands 2019b). This gap between “optimistic” expectations and “ob- stinate” reality is consistent with earlier fi ndings within the so-called fertility gap literature (Liefbroer/Billari 2010). Expectations for educational and occupational fi eld and for childlessness Next, multivariate analyses were performed to study the association between fi eld of study and occupation and fertility expectations. The results are presented in Ta- ble 2. All models show an excellent fi t based on standard goodness of fi t criteria (RMSEA, CFI, TLI). Model 1 presents results for boys and girls together, whereas in Model 2, interactions between gender and educational and occupational fi eld are included to test whether the same patterns apply to boys and girls. A logit regres- sion was performed to examine the link between expected fi eld of study and the expectation of remaining childless. Model 1 suggests strong associations between expected fi eld of study and childlessness. The greater the expectation among ado- lescents to study in the fi eld of communication, the more likely they were to ex- pect to have children (b=0.352). The opposite was true for expectations of pursuing cultural studies; this was associated with a lower expectation of having children (b=-0.271). Only weak evidence was found for an association between aspired oc- cupational fi eld and expectations regarding childlessness. The greater the expecta- tion among adolescents to work in the fi eld of communication, the more likely they were to expect to have children (b=0.215), whereas the greater their expectation to pursue a lower occupation, the less likely they were to want children (b=-0.325). Clearly, preferences regarding childlessness were more strongly related to adoles- cents’ expected fi eld of study than to their aspired fi eld of occupation. Interactions between gender and the other variables in the model were included in Model 2. Several statistically signifi cant interactions were observed. The upper half of Model 2 presents the effects for boys. The bottom half shows the estimates for the extent to which the effects for girls deviate from those for boys. First, the negative association between the expectation of entering a cultural fi eld of study and wanting children among boys was -0.444 (p<.01). To calculate the estimate for girls, the estimate for the interaction variable for a cultural fi eld of study and sex (0.380) was added to the parameter for boys (-.0444), leading to an effect for girls of -0.064. Additional analysis revealed that this parameter is not statistically signifi cantly different to 0. Thus, the negative association between the expectation of pursuing a cultural fi eld of study and expecting to have children is only observed Expectations about Fertility and Field of Study among Adolescents • 97 D o y o u t h in k y o u w ill h a v e c h ild re n ? A t w h a t a g e d o y o u t h in k In d e p e n d e n t V a ri a b le s y o u w ill h a v e c h ild re n ? M o d e l 1 M o d e l 2 M o d e l 1 M o d e l 2 L o g S t. S ig . L o g S t. S ig . S t. S t. S ig . S t. S t. S ig . o d d s E rr o r o d d s E rr o r C o e ff E rr o r C o e ff E rr o r In te rc e p t/ T h re sh o ld -0 .6 2 2 0 .3 8 4 -0 .1 6 2 0 .4 9 4 2 2 .8 7 1 0 .8 0 7 * * * 2 3 .4 11 1. 2 3 0 * * * E x p e c te d F ie ld o f S tu d y C a te g o ri se d a s C u lt u ra l -0 .2 7 1 0 .0 9 5 * * * -0 .4 4 4 0 .1 3 5 * * * 0 .5 3 0 0 .2 0 9 * * 0 .8 4 2 0 .4 3 9 * E x p e c te d F ie ld o f S tu d y C a te g o ri se d a s E co n o m ic 0 .0 8 6 0 .0 7 6 0 .2 2 2 0 .0 9 8 * * -0 .4 0 6 0 .1 9 5 * * -0 .5 3 7 0 .3 5 7 E x p e c te d F ie ld o f S tu d y C a te g o ri se d a s C o m m u n ic a ti v e 0 .3 5 2 0 .1 1 8 * * * 0 .2 0 3 0 .1 7 2 0 .0 3 9 0 .2 3 3 0 .0 4 0 0 .4 9 2 E x p e c te d F ie ld o f S tu d y C a te g o ri se d a s T e ch n ic a l 0 .1 7 6 0 .1 2 6 0 .3 7 3 0 .1 5 8 * * 0 .7 9 2 0 .2 7 6 * * * 0 .7 74 0 .4 4 0 * A sp ir e d O cc u p a ti o n s C a te g o ri se d a s C u lt u ra l -0 .1 19 0 .0 7 9 -0 .1 3 5 0 .1 2 9 0 .2 4 8 0 .2 15 0 .0 6 9 0 .3 6 5 A sp ir e d O cc u p a ti o n s C a te g o ri se d a s E co n o m ic 0 .0 9 6 0 .1 0 1 0 .1 8 8 0 .1 4 5 0 .3 4 9 0 .2 3 9 0 .4 1 0 0 .4 0 3 A sp ir e d O cc u p a ti o n s C a te g o ri se d a s C o m m u n ic a ti v e 0 .2 15 0 .1 2 9 * 0 .2 17 0 .1 8 9 -0 .2 7 7 0 .2 11 -0 .2 6 3 0 .4 5 2 A sp ir e d O cc u p a ti o n s C a te g o ri se d a s T e ch n ic a l -0 .1 2 1 0 .0 9 7 -0 .1 4 3 0 .1 2 6 -0 .0 0 2 0 .2 17 0 .0 9 1 0 .3 3 4 A sp ir e d O cc u p a ti o n s C a te g o ri se d a s L o w e r -0 .3 2 5 0 .1 7 1 * -0 .3 9 5 0 .2 41 -1 .1 8 1 0 .4 9 6 * * -1 .3 4 3 0 .7 8 7 * S e x a d o le sc e n t (f e m a le = 1) 0 .1 74 0 .1 2 5 1. 2 7 1 0 .8 8 3 -0 .8 8 9 0 .2 4 4 * * * -2 .2 0 8 1. 5 8 9 A d o le sc e n ts ’ c u rr e n t le v e l o f e d u c a ti o n ( IS L E D ) 0 .0 0 5 0 .0 0 4 0 .0 12 0 .0 0 5 * * 0 .0 8 0 0 .0 11 * * * 0 .0 7 2 0 .0 15 * * * In te ra c ti o n S e x * C a te g o ry E x p e c te d F ie ld o f S tu d y C a te g o ri se d a s C u lt u ra l 0 .3 8 0 0 .1 9 2 * * -0 .4 9 1 0 .5 15 E x p e c te d F ie ld o f S tu d y C a te g o ri se d a s E co n o m ic -0 .1 9 3 0 .1 7 6 0 .2 3 5 0 .4 2 9 E x p e c te d F ie ld o f S tu d y C a te g o ri se d a s C o m m u n ic a ti v e 0 .2 4 7 0 .2 5 3 0 .0 0 5 0 .5 4 5 E x p e c te d F ie ld o f S tu d y C a te g o ri se d a s T e ch n ic a l -0 .4 7 5 0 .2 4 0 * * -0 .0 7 6 0 .5 8 0 A sp ir e d O cc u p a ti o n s C a te g o ri se d a s C u lt u ra l 0 .1 0 8 0 .1 7 5 0 .3 0 5 0 .4 2 8 A sp ir e d O cc u p a ti o n s C a te g o ri se d a s E co n o m ic -0 .1 8 9 0 .2 4 0 -0 .0 8 9 0 .4 9 4 A sp ir e d O cc u p a ti o n s C a te g o ri se d a s C o m m u n ic a ti v e 0 .0 0 9 0 .2 4 3 -0 .0 0 5 0 .5 1 6 A sp ir e d O cc u p a ti o n s C a te g o ri se d a s T e ch n ic a l -0 .0 4 5 0 .2 0 9 -0 .1 7 2 0 .4 6 9 A sp ir e d O cc u p a ti o n s C a te g o ri se d a s L o w e r 0 .1 7 8 0 .4 0 3 0 .4 4 9 0 .9 6 3 A d o le sc e n ts ’ c u rr e n t le v e l o f e d u c a ti o n ( IS L E D ) - 0 .0 19 0 .0 1 0 * 0 .0 17 0 .0 17 n = 12 6 4 n = 1 0 8 4 T a b . 2 : U n st an d a rd is e d e ff e c ts o f e x p e c te d fi e ld s o f st u d y a n d a sp ir e d fi e ld s o f o cc u p a ti o n o n t h e e x p e c ta ti o n o f h av in g ch ild re n , e x p e c te d a g e o f h av in g c h ild re n a n d e x p e c te d n u m b e r o f ch ild re n . In m o d e l 2 , in te ra c ti o n e ff e c ts w e re a d d e d a s fi e ld o f st u d y a n d o cc u p a ti o n a l c a te g o ri e s m u lt ip lie d b y s e x ( m a le = 0 a n d f e m a le = 1) . • Micha G. Keijer, Aart C. Liefbroer, Ineke Nagel98 T a b . 2 : C o n ti n u a ti o n H o w m a n y c h ild re n d o y o u In d e p e n d e n t V a ri a b le s th in k y o u w ill h a v e? M o d e l 1 M o d e l 2 S t. S t. S t. S t. C o e ff E rr o r S ig . C o e ff E rr o r S ig . In te rc e p t / T h re sh o ld 2 .0 7 5 0 .1 7 9 * * * 1. 8 9 4 0 .2 4 3 * * * E x p e c te d F ie ld o f S tu d y C a te g o ri se d a s C u lt u ra l 0 .0 3 8 0 .0 4 3 0 .0 6 8 0 .0 9 7 E x p e c te d F ie ld o f S tu d y C a te g o ri se d a s E co n o m ic -0 .0 1 0 0 .0 51 -0 .0 0 3 0 .0 8 1 E x p e c te d F ie ld o f S tu d y C a te g o ri se d a s C o m m u n ic a ti v e -0 .0 4 3 0 .0 6 3 -0 .1 15 0 .1 1 8 E x p e c te d F ie ld o f S tu d y C a te g o ri se d a s T e ch n ic a l -0 .0 8 6 0 .0 8 3 -0 .0 7 1 0 .1 19 A sp ir e d O cc u p a ti o n s C a te g o ri se d a s C u lt u ra l -0 .0 6 4 0 .0 6 5 -0 .1 2 5 0 .1 0 0 A sp ir e d O cc u p a ti o n s C a te g o ri se d a s E co n o m ic 0 .0 3 0 0 .0 5 9 -0 .0 5 0 0 .0 9 6 A sp ir e d O cc u p a ti o n s C a te g o ri se d a s C o m m u n ic a ti v e 0 .1 2 1 0 .0 5 2 * * 0 .0 6 6 0 .0 8 9 A sp ir e d O cc u p a ti o n s C a te g o ri se d a s T e ch n ic a l -0 .0 5 8 0 .0 6 3 -0 .0 7 5 0 .0 8 6 A sp ir e d O cc u p a ti o n s C a te g o ri se d a s L o w e r 0 .1 5 6 0 .1 2 2 0 .4 2 4 0 .1 8 9 * * S e x a d o le sc e n t (f e m a le = 1) 0 .0 5 7 0 .0 74 0 .4 2 8 0 .3 9 3 A d o le sc e n ts ’ c u rr e n t le v e l o f e d u c a ti o n ( IS L E D ) 0 .0 0 3 0 .0 0 2 0 .0 0 7 0 .0 0 3 * * In te ra c ti o n S e x * C a te g o ry E x p e c te d F ie ld o f S tu d y C a te g o ri se d a s C u lt u ra l -0 .0 5 5 0 .1 2 0 E x p e c te d F ie ld o f S tu d y C a te g o ri se d a s E co n o m ic -0 .0 0 4 0 .0 9 6 E x p e c te d F ie ld o f S tu d y C a te g o ri se d a s C o m m u n ic a ti v e 0 .1 3 4 0 .1 3 4 E x p e c te d F ie ld o f S tu d y C a te g o ri se d a s T e ch n ic a l -0 .0 8 7 0 .1 5 0 A sp ir e d O cc u p a ti o n s C a te g o ri se d a s C u lt u ra l 0 .1 3 5 0 .1 3 1 A sp ir e d O cc u p a ti o n s C a te g o ri se d a s E co n o m ic 0 .1 51 0 .1 14 A sp ir e d O cc u p a ti o n s C a te g o ri se d a s C o m m u n ic a ti v e 0 .1 0 7 0 .1 12 A sp ir e d O cc u p a ti o n s C a te g o ri se d a s T e ch n ic a l 0 .0 5 4 0 .1 3 6 A sp ir e d O cc u p a ti o n s C a te g o ri se d a s L o w e r -0 .5 74 0 .2 6 3 * * A d o le sc e n ts ’ c u rr e n t le v e l o f e d u c a ti o n ( IS L E D ) -0 .0 0 7 0 .0 0 5 n = 1 0 8 4 N o te s G o o d n e ss o f fi t o f al l m o d e ls w as e v al u at e d w it h t h e f o llo w in g i n d ic e s: R o o t M e an S q u ar e E rr o r o f A p p ro x im at io n ( R M S E A ), C o m p ar at iv e F it I n d e x ( C F I) , an d Tu ck e r- Le w is I n d e x ( T LI ). A ll m o d e ls h av e v al u e s o f R M S E A = 0 .0 0 , C F I= 1 .0 0 a n d T LI = 1 .0 0 , su g g e st in g g o o d m o d e l fi t ( H u /B e n tl e r 1 9 9 9 ). ** *p -v a lu e < 0 .0 1 . ** p -v a lu e < 0 .0 5 . *p -v a lu e < 0 .1 Expectations about Fertility and Field of Study among Adolescents • 99 in boys, not in girls. Second, among boys, a stronger preference for having chil- dren is associated with a stronger expectation for pursuing an economic fi eld of study (b=0.222) and a stronger expectation for pursuing a technical fi eld of study (b=0.373). No such associations are found among girls (b=0.222-0.193=0.029, n.s.; b=0.373-0.475=-0.102, n.s.). Among girls, we only fi nd a positive relationship be- tween expectations of studying in the fi eld of communication and the expectation of having children. The relation between expected studies in the communicative fi eld is equally related to expected childlessness for boys and girls. Among boys, all four fi elds of study are related to the desire to have children. Boys with stronger ex- pectations for studying in the cultural fi eld are less likely to expect to have children, whereas boys with stronger expectations for studying in economic, communica- tions and technical fi elds of study are more likely to want children. For girls, only an expectation of studying in the communicative fi eld is relevant in this respect. The association between job aspiration and childlessness does not vary signifi cantly between boys and girls. Expected educational and occupational fi elds and expected age of entry into parenthood The results in Table 2 on the expected age of entering into parenthood suggest that there are no gender differences in the underlying process (see middle panel of Table 2, Model 2). Therefore, only results from Model 1 are discussed. For both fe- male and male adolescents, the expected age of entry into parenthood was strongly related to their current level of education and their expected fi eld of education. The higher the adolescents’ current level of education, the later their expected age of having children (b=0.080). Stronger expectations of studying in cultural and tech- nical fi elds were associated with a higher expected age of entry into parenthood (b=0.530 and b=0.792, respectively). Stronger expectations for studying in the fi eld of economics was associated with a lower expected age of entry into parenthood (b=-0.406). No association was found between the expectation of studying in the communications fi eld and expected age of entry into parenthood. No associations were found between aspired occupational fi elds and expected age of entry into par- enthood either, with one exception: boys and girls who “aspired” to a lower type of occupation expected to have their fi rst child earlier (p<0.05). Expected educational and occupational fi eld and expected number of children The fi nal panel of Table 2 shows the results for the association between educational and occupational fi eld and the number of children that adolescents expect. Only few statistically signifi cant associations were observed. Model 1 shows that the more strongly adolescents aspired to an occupation in communication, the more children they expected to have (b=0.121). Model 2 shows that boys and girls who “aspire” to a lower type of occupation differ in their views on the number of children they expect. Among boys, a stronger aspiration for a lower type of occupation is associ- • Micha G. Keijer, Aart C. Liefbroer, Ineke Nagel100 ated with a higher expected number of children (b=0.424), whereas no association is found among girls (b=0.424-0.574=-0.150, ns). 6 Conclusion The growing literature on the association between fertility, fi eld of study and occu- pation shows that fertility is higher among women who were educated or are em- ployed in communicative fi elds that emphasize teaching and care. An unresolved issue in the literature is to what extent fi eld-specifi c/occupational socialisation pro- cesses infl uence fertility decisions of young adults, or to what extent young adults are selected into specifi c fi elds of study and occupation because of pre-existing preferences concerning family life and employment. The goal of this study was to examine if such a selection effect is likely. We did so by studying the associations between adolescents’ expectations and preferences concerning fertility, fi eld of study and occupation during the life phase before entering tertiary education. In addition to studying this relationship among adolescent girls, we also examined it among boys. Our fi rst set of hypotheses stated that girls that expect to enter communicative educational and occupational fi elds were also less likely to remain childless and had a greater likelihood of having more children and having them earlier. Several fi ndings are in line with this hypothesis. The greater the expectation among girls to opt for communicative fi elds of study, the more likely they are to have children. In addition, the more likely girls are to prefer jobs in the communicative fi eld, the more children they expect to have. Thus, there seems to be an elective affi nity between preferences for studies and jobs that emphasize caring and social interaction and fertility expectations among adolescent girls. The expected timing of having chil- dren is not associated with expectations of pursuing a communicative fi eld of study or aspired occupation. What is striking, though, is that this elective affi nity also ap- plies to boys. Here, too, a preference for studies and jobs that emphasize care and social interaction goes hand in hand with stronger fertility preferences. Rather than showing gender-specifi c patterns, as has been suggested by earlier studies (e.g. Van Bavel 2010; Hakim 2003), both girls and boys who are attracted to studies and jobs in the social domain are also attracted to parenthood. Clearly, girls are more likely to feel attracted to studies and jobs in the social domain, but boys who feel drawn to these studies show the same pattern of associations as girls. The second set of hypotheses stated that boys that expect to opt for studies and jobs in the economic and technical domain are more likely to expect to have children, to have more children and to have them earlier. The results only partially confi rmed our hypotheses about this link. The greater the expectation among male adolescents to opt for an economic fi eld of study, the less likely they were to expect to remain childless and the more likely they were to have their fi rst child somewhat earlier. Boys who preferred technical fi elds of study and occupation were also less likely to expect to remain childless, but they expected to have children at a later age, which is not in line with the hypothesis. On top of this, similar associations be- Expectations about Fertility and Field of Study among Adolescents • 101 tween expected fi eld of study and expected timing of entry into parenthood and the expected number of children were observed among both boys and girls, again sug- gesting that the processes of elective affi nity between domain of study and work on the one hand and fertility on the other hand are more or less comparable for boys and girls. The genders differ in their likelihood of preferring communicative and economic fi elds of study and occupation, but if they do have the same preference, the association with fertility expectations is more or less the same. A result we did not anticipate is that the greater the expectation among male adolescents to opt for cultural studies, the more likely it was that they expected to remain childless and, if they did expect to have children, to have them later in life. This latter result is congruent with Van Bavel’s (2010) fi nding in the Netherlands that people who are trained in arts and the humanities are the most progressive con- cerning family attitudes. It shows that adolescents opting for these fi elds of study and occupations have signifi cantly different fertility values, suggesting that men with more “traditional/conservative” values are different from men with “artistic/ progressive” values. Our results clearly show that expectations regarding fi eld of study and occupa- tion and fertility are already related to each other during the teenage years and this strongly suggests that the association between fi eld of education and actual fertil- ity found among adults is indeed at least partially the result of processes of self- selection occurring at an early stage in the life course. Our study could not examine which mechanisms “cause” the association between fi eld of study and fertility pref- erences among adolescence: for instance whether it refl ects gendered identities, as suggested by Charles and Bradley (2009), or to what extent these are in turn the result of common causes like early socialisation or personality. Future research could, for instance, examine the role of parental (gender) socialisation, personality traits (Jokela 2012), and other background variables, such as socioeconomic status or religiosity of the parents (Hubert 2014), or the role of schools and peers (Driscoll/ Abma 2015). The cross-sectional nature of our study did not allow us to study how realistic adolescents’ expectations are. Using panel data, it would be interesting to exam- ine to what extent adolescents’ expectations are realised, and whether the same processes infl uence both their expectations and their actual behaviour. In addition, panel data would allow us to study the consistency and development of prefer- ences regarding fertility, fi eld of study and occupation. Such a design would be par- ticularly interesting given that it allows the direct disentanglement of causality and self-selection in the relationship between fertility and educational and occupational fi eld, and identifi cation of common causes, e.g. early socialisation and personality traits, that could explain both types of preferences. Finally, in line with research on country variation in the association between actual choices in fertility and fi elds of study (Van Bavel 2010; Oppermann 2017), it would be interesting to study country variations in the association between preferences in the professional and the fam- ily domain. The degree to which educational systems and labour markets allow for the expression of gendered identities, which was found to be related to gender segregation in fi elds of study (Charles/Bradley 2009), differs between countries. It • Micha G. Keijer, Aart C. Liefbroer, Ineke Nagel102 could therefore be expected that a stronger correlation between preferences in the professional and the family domain will be found in countries that allow stronger expression of such gendered identities. In addition, institutional arrangements like maternity leave, possibilities to work part-time, social acceptance etc. also differ per society. It would be interesting to examine whether countries that facilitate “working-mothers” and “caring fathers” are more likely to exhibit a weaker relation between preferences regarding fertility, fi eld of study and occupation. References Allison, Paul D. 2012: Handling missing data by maximum likelihood. Paper presented at the SAS Global Forum (paper 312-2012) [http://www.statisticalhorizons.com/wp- content/uploads/MissingDataByML.pdf, 22.07.2019]. Axinn, William G.; Clarkberg, Marin, E.; Thornton, Arland 1994: Family infl uences on family size preferences. In: Demography 31,1: 65-79 ]doi: 10.2307/2061908]. Bagavos, Christos 2010: Education and childlessness: the relationship between educa- tional fi eld, educational level, employment and childlessness among Greek women born in 1955-1959. In: Vienna Yearbook of Population Research 8: 51-75 [doi: 10.1553/ populationyearbook2010s51]. Barber, Jennifer S. 2000: Intergenerational infl uences on the entry into parenthood: mothers’ preferences for family and non-family behavior. In: Social Forces 79,1: 319- 348 [doi: 10.1093/sf/79.1.319]. Barone, Carlo 2011: Some things never change: Gender segregation in higher education across eight nations and three decades. In: Sociology of Education 84,2: 157-176 [doi: 10.1177/0038040711402099]. Van Bavel, Jan 2010: Choice of study discipline and the postponement of motherhood in Europe: The impact of expected earnings, gender composition, and family attitudes. In: Demography 47,2: 439-458 [doi: 10.1353/dem.0.0108]. Begall, Katia; Mills, Melinda C. 2012: The infl uence of educational fi eld, occupation, and occupational sex segregation on fertility in the Netherlands. In: European Sociological Review 29,4: 720-742 [doi: 10.1093/esr/jcs051]. Blossfeld, Hans Peter; Huinink, Johannes 1991: Human capital investments or norms of role transition? How women’s schooling and career affect the process of family forma- tion. In: American Journal of Sociology 97,1: 143-168. Bourdieu, Pierre 1986: The forms of capital. In: Biggart, Nicole Woolsey (Ed.): Readings in Economic Sociology. Oxford: Blackwell Publishers: 280-291 [doi: 10.1002/9780470755679.ch15]. Bozick, Robert et al. 2010: Framing the future: Revisiting the place of educational ex- pectations in status attainment. In: Social Forces 88,5: 2027-2052 [doi: 10.1353/ sof.2010.0033]. Charles, Maria; Bradley, Karen 2009: Indulging our gendered selves? Sex segregation by fi eld of study in 44 countries. In: American Journal of Sociology 114,4: 924-976 [doi: 10.1086/595942]. Davies, Scott; Guppy, Neil 1997: Fields of study, college selectivity, and student inequal- ities in higher education. In: Social Forces 75,4: 1417-1438 [doi: 10.1093/sf/75.4.1417]. Expectations about Fertility and Field of Study among Adolescents • 103 Driscoll, Anne K.; Abma, Joyce C. 2015: Changing sociodemographic factors and teen fertility: 1991-2009. In: Maternal and Child Health Journal 19,10: 2159-2167 [doi: 10.1007/s10995-015-1728-8]. Ganzeboom, Harry B.G.; Nagel, Ineke 1998-2002: Youth and culture. A multi-actor panel study (machine readable datafi les). Utrecht: Interuniversity Center for Social Science Theory and Methodology. Available from the authors. Ganzeboom, Harry B.G.; Nagel, Ineke; Liefbroer, Aart C. 2005-2006: Youth and culture. A multi-actor panel study. Cultural participation and future plans (machine readable datafi les). Amsterdam: Vrije Universiteit. Available from the authors. Goyette, Kimberly A. 2008: College for some to college for all: Social background, oc- cupational expectations, and educational expectations over time. In: Social Science Research 37,2: 461-484 [doi: 10.1016/j.ssresearch.2008.02.002]. Guetto, Raffaele; Panichella, Nazareno 2013: Geographical mobility and reproductive choices of Italian men. In: European Sociological Review 29,2: 302-315 [doi: 10.1093/ esr/jcr068]. Hakim, Catherine 2000: Work-Lifestyle Choices in the 21st Century: Preference Theory. Oxford: Oxford University Press. Hakim, Catherine 2003: A new approach to explaining fertility patterns: Preference the- ory. In: Population and Development Review 29,3: 349-374. Hardy, Melissa A. 1993: Regression with Dummy Variables. Newbury Park, Cal.: Sage [doi: 10.4135/9781412985628]. Hoem, Jan M.; Neyer, Gerda; Andersson, Gunnar 2006a: Educational attainment and ultimate fertility among Swedish women born in 1955-59. In: Demographic Research 14,16: 381-404 [doi: 10.4054/DemRes.2006.14.16]. Hoem, Jan M.; Neyer, Gerda; Andersson, Gunnar 2006b: Education and childlessness: The relationship between educational fi eld, educational level, and childlessness among Swedish women born in 1955-59. In: Demographic Research 14,15: 331-380 [doi: 10.4054/DemRes.2006.14.15]. Hu, Li-tze; Bentler, Peter M. 1999: Cutoff criteria for fi t indexes in covariance structure analysis: Conventional criteria versus new alternatives. In: Structural Equation Mod- eling: A Multidisciplinary Journal 6,1: 1-55 [doi: 10.1080/10705519909540118]. Hubert, Sandra 2014: The Impact of Religiosity on Fertility: A Comparative Analysis of France, Hungary, Norway, and Germany. Wiesbaden: Springer [doi: 10.1007/978-3- 658-07008-3]. Jokela, Markus 2012: Birth-cohort effects in the association between personality and fertility. In: Psychological Science 23,8: 835-841 [doi: 10.1177/0956797612439067]. Kalmijn, Matthijs 1998: Intermarriage and homogamy: Causes, patterns, trends. In: An- nual Review of Sociology 24: 395-421 [doi: 10.1146/annurev.soc.24.1.395]. Kanji, Shireen; Hupka-Brunner, Sandra 2015: Young women’s strong preference for chil- dren and subsequent occupational gender segregation: What is the link? In: Equality, Diversity and Inclusion: An International Journal 34,2: 124-140 [doi: 10.1108/EDI-05- 2014-0041]. Kohler, Hans-Peter; Billari, Francesco C.; Ortega, José Antonio 2006: Low fertility in Eu- rope: Causes, implications and policy options. In: Harris, Fred R. (Ed.): The Baby Bust: Who Will Do the Work? Who Will Pay the Taxes? Lanham, ML: Rowman & Littlefi eld Publishers: 48-109. • Micha G. Keijer, Aart C. Liefbroer, Ineke Nagel104 Korpershoek, Hanke et al. 2010: Who ‘fi ts’ the science and technology profi le? Personal- ity differences in secondary education. In: Journal of Research in Personality 44,5: 649-654 [doi: 10.1016/j.jrp.2010.08.010]. Lappegård, Trude 2002: Education Attainment and Fertility Pattern among Norwegian Women. Oslo: Statistisk Sentralbyrå. Lappegård, Trude; Rønsen, Marit 2005: The multifaceted impact of education on entry into motherhood. In: European Journal of Population 21,1: 31-49 [doi: 10.1007/s10680- 004-6756-9]. Lareau, Annette 2011: Unequal Childhoods: Class, Race, and Family Life. Berkeley: Uni- versity of California Press. Lewis, Jane et al. 2008: Patterns of development in work/family reconciliation policies for parents in France, Germany, the Netherlands, and the UK in the 2000s. In: Social Politics: International Studies in Gender, State & Society 15,3: 261-286 [doi: 10.1093/ sp/jxn016]. Liefbroer, Aart C.; Billari, Francesco C. 2010: Bringing norms back in: A theoretical and empirical discussion of their importance for understanding demographic behaviour. In: Population, Space and Place 16,4: 287-305 [doi: 10.1002/psp.552]. Liefbroer, Aart C.; Merz, Eva-Maria; Testa, Maria Rita 2015: Fertility-related norms across Europe: A multi-level analysis. In: Philipov, Dimiter; Liefbroer, Aart C.; Klobas, Jane E. (Eds.): Reproductive Decision-Making in a Macro-Micro Environment. New York: Springer: 141-163 [doi: 10.1007/978-94-017-9401-5_6]. Lutz, Katharina 2014: Compensating dissatisfaction in the job by turning to the family? The impact of current occupation on timing of fi rst births in Germany. In: Advances in Life Course Research 21: 43-58 [doi: 10.1016/j.alcr.2014.03.003]. Martín-García, Teresa 2009: ‘Bring men back in’: a re-examination of the impact of type of education and educational enrolment on fi rst births in Spain. In: European Socio- logical Review 25,2: 199-213 [doi: 10.1093/esr/jcn041]. Martín-García, Teresa; Baizán, Pau 2006: The impact of the type of education and of educational enrolment on fi rst births. In: European Sociological Review 22,3: 259-275 [doi: 10.1093/esr/jci056]. Michelmore, Katherine; Musick, Kelly 2014: Fertility patterns of college graduates by fi eld of study, U.S. women born 1960-79. In: Population Studies 68,3: 359-374 [doi: 10.1080/00324728.2013.847971]. Morgan, Stephen L.; Gelbgiser, Dafna; Weeden, Kim A. 2013: Feeding the pipeline: Gen- der, occupational plans, and college major selection. In: Social Science Research 42,4: 989-1005 [doi: 10.1016/j.ssresearch.2013.03.008]. Murphy, Michael; Wang, Duolao 2001: Family-level continuities in childbearing in low- fertility societies. In: European Journal of Population 17,1: 75-96 [doi: 10.1023/A: 1010744314362]. Muthén, Linda K.; Muthén, Bengt O. 2004: Mplus: The comprehensive modeling pro- gram for applied researchers. Los Angeles, CA: Muthén & Muthén. Oppermann, Anja 2014: Exploring the relationship between educational fi eld and transi- tion to parenthood – An analysis of women and men in Western Germany. In: Euro- pean Sociological Review 30,6: 728-749 [doi: 10.1093/esr/jcu070]. Oppermann, Anja 2017: Educational fi eld and fertility in western Germany: An analysis of women born between 1955 and 1959. In: Vienna Yearbook of Population Research 15: 239-267 [doi: 10.1553/populationyearbook2017s239]. Expectations about Fertility and Field of Study among Adolescents • 105 Pascall, Gillian 2010: Male breadwinner model. In: Fitzpatrick, Tony et al. (Eds.): Interna- tional Encyclopedia of Social Policy. 1st ed. London/New York: Routledge. Risman, Barbara J.; Davis, Georgiann 2013: From sex roles to gender structure. In: Cur- rent Sociology 61,5-6: 733-755 [doi: 10.1177/0011392113479315]. Rønsen, Marit; Skrede, Kari 2010: Can public policies sustain fertility in the Nordic coun- tries? Lessons from the past and questions for the future. In: Demographic Research 22,13: 321-346 [doi: 10.4054/DemRes.2010.22.13]. Schröder, Heike; Ganzeboom, Harry B.G. 2014: Measuring and Modelling Level of Edu- cation in European Societies. In: European Sociological Review 30,1: 119-136 [doi: 10.1093/esr/jct026]. Skirbekk, Vegard; Blekesaune, Morten 2014: Personality traits increasingly important for male fertility: Evidence from Norway. In: European Journal of Personality 28,6: 521- 529 [doi: 10.1002/per.1936]. Statistics Netherlands 2019: Information on Employment by Sector [https://opendata. cbs.nl/statline/#/CBS/nl/dataset/81432ned/table?ts=1559038555556, 28.05.2019]. Statistics Netherlands 2019b: Information on Births [https://opendata.cbs.nl/statline/#/ CBS/nl/dataset/37422ned/table?ts=1559039903598, 28.05.2019]. Steenhof, Liesbeth; Liefbroer, Aart C. 2008: Intergenerational transmission of age at fi rst birth in the Netherlands for birth cohorts born between 1935 and 1984: Evidence from municipal registers. In: Population Studies 62,1: 69-84 [doi: 10.1080/00324720701788616]. Stockard, Jean 2006: Gender socialization. In: Saltzman Chafetz, Janet (Ed.): Handbook of the Sociology of Gender. New York: Springer: 215-227 [doi: 10.1007/0-387-36218- 5_11]. Tavares, Lara Patrício 2016: Who delays childbearing? The associations between time to fi rst birth, personality traits and education. In: European Journal of Population 32,4: 575-597 [doi: 10.1007/s10680-016-9393-1]. Tesching, Karin 2012: Education and Fertility. Dynamic Interrelations Between Women’s Educational Level, Educational Field and Fertility in Sweden. Acta Universitatis Stock- holmiensis: Stockholm University Demography Unit – Dissertation Series 6 [URN: urn:nbn:se:su:diva-63761]. Thornton, Arland 1980: The infl uence of fi rst generation fertility and economic status on second generation fertility. In: Population and Environment 3,1: 51-72 [doi: 10.1007/ BF01253070]. Trent, Katherine; Crowder, Kyle 1997: Adolescent birth intentions, social disadvantage, and behavioral outcomes. In: Journal of Marriage and Family 59,3: 523-535 [doi: 10.2307/353943]. Wells, Robert; Ham, Roger; Junankar, P.N. 2016: An examination of personality in oc- cupational outcomes: antagonistic managers, careless workers and extraverted sales- people. In: Applied Economics 48,7: 636-651 [doi: 10.1080/00036846.2015.1085636]. Van de Werfhorst, Herman G.; Kraaykamp, Gerbert 2001: Four fi eld-related educational resources and their impact on labor, consumption and sociopolitical orientation. In: Sociology of Education 74,4: 296-317 [doi: 10.2307/2673137]. Van de Werfhorst, Herman G.; Luijkx, Ruud 2010: Educational fi eld of study and social mobility: Disaggregating social origin and education. In: Sociology 44,4: 695-715 [doi: 10.1177/0038038510369362]. • Micha G. Keijer, Aart C. Liefbroer, Ineke Nagel106 Wood, Jonas; Neels, Karel; Kil, Tine 2014: The educational gradient of childlessness and cohort parity progression in 14 low fertility countries. In: Demographic Research 31,46: 1365-1416 [doi: 10.4054/DemRes.2014.31.46]. Woods, Stephen A. et al. 2016: Personality and occupational specialty. An examination of medical specialties sing Holland’s RIASEC model. In: Career Development Interna- tional 21,3: 262-278 [doi: 10.1108/CDI-10-2015-0130]. Youniss, James; Smollar, Jacqueline 1985: Adolescent Relations with Mothers, Fathers, and Friends. Chicago: University of Chicago Press. Date of submission: 19.06.2017 Date of acceptance: 08.06.2019 Dr. Micha G. Keijer (). Amsterdam University of Applied Sciences, Amsterdam School of International Business (AMSIB). Amsterdam, The Netherlands. E-mail: m.g.keijer@hva.nl URL: https://www.amsterdamuas.com/amsib/profi le/k/e/m.g.keijer/m.g.keijer.html?orig in=Z%2B1TAmBTT1GFpljJFzuhXQ Prof. Dr. Aart C. Liefbroer. Netherlands Interdisciplinary Demographic Institute (NIDI). The Hague, The Netherlands. E-mail: liefbroer@nidi.nl URL: https://www.nidi.nl/en/staff/overview/liefbroer Dr. Ineke Nagel. Vrije University Amsterdam, Faculty of Social Sciences, Sociology, Social Inequality and the Life Course (SILC). Amsterdam, The Netherlands. E-mail: f.a.nagel@vu.nl URL: https://research.vu.nl/en/persons/ineke-nagel Published by Prof. Dr. Norbert F. Schneider Federal Institute for Population Research D-65180 Wiesbaden / Germany 2019 Managing Editor Prof. Philip Rees Dr. Katrin Schiefer Copy Editor Julia Luther Editorial Assistant Beatriz Feiler-Fuchs Wiebke Hamann Layout Beatriz Feiler-Fuchs E-mail: cpos@bib.bund.de Scientifi c Advisory Board Karsten Hank (Cologne) Michaela Kreyenfeld (Berlin) Marc Luy (Vienna) Natalie Nitsche (Vienna) Peter Preisendörfer (Mainz) Zsolt Spéder (Budapest) Rainer Wehrhahn (Kiel) Comparative Population Studies www.comparativepopulationstudies.de ISSN: 1869-8980 (Print) – 1869-8999 (Internet) Board of Reviewers Martin Abraham (Erlangen) Laura Bernardi (Lausanne) Hansjörg Bucher (Bonn) Claudia Diehl (Konstanz) Andreas Diekmann (Zurich) Gabriele Doblhammer-Reiter (Rostock) Jürgen Dorbritz (Wiesbaden) Anette Eva Fasang (Berlin) E.-Jürgen Flöthmann (Bielefeld) Alexia Fürnkranz-Prskawetz (Vienna) Beat Fux (Salzburg) Joshua Goldstein (Berkeley) Sonja Haug (Regensburg) Hill Kulu (Liverpool) Aart C. Liefbroer (The Hague) Kurt Lüscher (Konstanz) Emma Lundholm (Umeå) Nadja Milewski (Rostock) Dimiter Philipov (Vienna) Roland Rau (Rostock) Tomáš Sobotka (Vienna) Jeroen Spijker (Barcelona) Olivier Thévenon (Paris) Helga de Valk (Brussels) Heike Trappe (Rostock) Michael Wagner (Cologne)