Satisfaction with life as an antecedent of fertility: Partner + Happiness = Children?

This paper examines the relationships between satisfaction with life in general, particular domains of life, the partner, and parental relationships with existing children, and subsequent fertility. The data are from 2,948 women and 2,622 men aged 15 to 44 years, from a longitudinal survey of the household population in Australia. For both sexes, a strong positive relationship between prior satisfaction with life and fertility two years later is found. Men’s satisfaction with their partner and with their partner’s relationship with existing children are positively related to fertility. Fertility is also related to age, parity, marital status, education, employment and birthplace. 1 Macquarie University. E-mail: [email protected] Parr: Satisfaction with life as an antecedent of fertility 636 http://www.demographic-research.org 1. Overview Are people who are more satisfied (“happier”) with life more likely to have children? And are people who are happier with their partners more likely to have children? The literature on the interrelationship between fertility and an individual’s ordinal level of satisfaction with life (often termed subjective wellbeing) has tended to focus on the effect of having additional children on satisfaction with life (McLanahan and Adams 1987, 1989; Evenson and Simon 2005; Kohler et al. 2005). Except for Perelli-Harris (2006), the possibility of higher levels of satisfaction with life among parents having preceded the arrival of the children appears to have been overlooked. There is a range of theoretical reasons to expect those with higher levels of satisfaction with life to have higher fertility. First, the presence of a partner contributes to a person’s satisfaction with life, and also, obviously, would affect fertility (Evans and Kelley 2004; Zimmermann and Easterlin 2006; Carmichael and Whittaker 2007a, b). A person’s satisfaction with their partner would affect their satisfaction with life, and would affect their fertility through enhancing the stability and the likelihood of continuation of the union. Third, parents generally aspire to have happy children. If prospective parents are happy with life this may increase the likelihood that their children will be so, due to parent-to-child transmission of biological or social endowments. A perceived prospect of happier children in turn may heighten the fertility desire (Easterlin 2006). Furthermore, since a child will to some extent be in the image of his or her parents, adults who have a more positive view of themselves and of their partner may be more favourably disposed to their prospective offspring. Among those who have already had children, the satisfaction derived from those children will also have contributed to the overall level of satisfaction with life (Kohler et al. 2005). Parental experiences with existing children and the satisfaction they derive from these experiences differ widely and may affect parents’ desire for additional children (Newman 2008). Finally, depression and stress have been found to contribute to (as well as result from) reduced fecundity, and hence would contribute to reduced subsequent fertility (Abbey et al. 1992; Phillips and Slaughter 2000; Arck et al. 2001; Wisborg et al. 2008; Zemishlany and Weizman 2008). Figure 1 presents a diagrammatic representation of these relationships. This study uses longitudinal data to assess the relationship between prior satisfaction with life and subsequent fertility. Since satisfaction with life may arise due to satisfaction with one or more a range of different aspects with life, including personal relationships, health, finances, employment, housing, the local community, and leisure time, the paper also assesses whether satisfaction with particular domains of life is related to subsequent fertility, after controlling for a range of demographic, socioeconomic and cultural confounding variables. Demographic Research: Volume 22, Article 21 Figure 1: Theoretical links between satisfaction with life and subsequent fertility Experiences and Relationships with Existing Children Satisfaction/ Dissatisfaction With Life Expected Child Utility e.g. From Own and Partner’s Traits Apparent in Child Gives Birth to/Father’s (Additional) Child Infecundity, Spontaneous Abortion, Still Birth, Libido, Sexual Dysfunction Contraceptive Use Marital Status and Satisfaction with Partner Biological and Social Endowments Financial Situation, Employment, Housing, Health etc. Selection Into, Continuation, and Stability of Relationships 2. Literature review The literature provides a range of reasons to expect satisfaction with life to be an antecedent of fertility both for men and for women. Principal among these is that there are both empirical and theoretical grounds for expecting satisfaction levels to be both a consequence and a determinant of the formation, continuation, stability of, and satisfaction with unions, and thus to be related to fertility. Even though the fertility of the unmarried has risen markedly in Australia since the 1970s, the married still have considerably higher rates of fertility (Carmichael and McDonald 2003; ABS 2008). Entering and remaining in a registered marriage has also been found to be associated with an increased level of satisfaction with life over the long run, as well as a temporary “honeymoon period” increase (Evans and Kelley 2004; Zimmermann and Easterlin 2006). As well as satisfaction with life being a consequence of partnership, it is also possible that some of this association is due to selection of individuals into unions on the basis of the presence of desirable psychological traits (or absence of undesirable http://www.demographic-research.org 637 Parr: Satisfaction with life as an antecedent of fertility 638 http://www.demographic-research.org traits). Fowler and Christakis (2008) have shown the happiness of one spouse tends to affect the happiness of the other. This may make the happier more desirable as partners. Fertility within unions and the continuation of unions may also be related to satisfaction with the partner and with life more generally. For example, Carmichael and Whittaker (2007a, b) document examples of men and women who were childless because they had never viewed their relationships as being with “the right person” or had become unhappy with relationships and who believed that having children would make it more difficult to search for a more satisfying relationship. The discontinuation of some unions may be due to the impairment of the relationship by one of the partners’ depression, anxiety or other personality traits (Kiernan 1986; Zimmermann and Easterlin 2006; Zemishlany and Weizman 2008). A second reason that satisfaction with life may be an antecedent of fertility is that adults with differing levels of satisfaction with life may perceive the prospect of (additional) children differently; there may be “image of self” and “image of partner” effects. According to Billari and Kohler (2009), the additional happiness which parents anticipate would result from their having (additional) children is a key driver of childbearing decisions. I hypothesize on theoretical grounds that such “predicted happiness from childbearing” would be related to the current levels of satisfaction with life and satisfaction with the partner. The literature suggests that a person’s satisfaction with life reflects biological, personality and social endowments from their parents (Kohler et al. 2005; Easterlin 2006). Thus happier, better-adjusted adults might reasonably anticipate having happier, better-adjusted (and hence more desirable) children. Moreover, since a person’s level of satisfaction with life may be partly attributable to their satisfaction with their own personality and physical characteristics and with those of their partner, and since a child is likely to resemble (to some degree) his or her parents, the highly satisfied (or highly satisfied with partner) face the prospect of children with characteristics they might expect to find highly satisfying, whilst the less satisfied face the prospect of children with characteristics they might expect to find less satisfying. The theoretical model of the relationships between satisfaction with life and subsequent fertility differs between first and subsequent births. This is because for those who have already had children, the expected happiness from having additional children (and hence fertility) may be affected by their own and their partner’s experiences and relationships with their existing children. For example, Newman (2008) describes examples of women and men whose positive parenting experiences had contributed to their desires to have another child, as well as others who had been deterred from doing so by negative experiences. The mother’s relationship with the children may be a more important determinant of subsequent fertility than the father’s, because in Australia the Demographic Research: Volume 22, Article 21 http://www.demographic-research.org 639 additional time spent on the domestic work and parental childcare which results from children is borne mostly by the mother (Craig and Bittman 2004; Craig 2005). As well as providing reasons to expect the more satisfied to have higher fertility, the literature also provides other reasons which might lead one to expect lower fertility among the more satisfied. According to Ramu (1984) and Carmichael and Whittaker (2007a) aversion to lifestyle change is one of the main types of justification the voluntarily childless cite for their remaining childless. If it may be presumed that those who do not wish to change their lifestyle tend to express high existing levels of satisfaction with life, then such high levels of subjective wellbeing might be expected to be associated with lower fertility (Ramu 1984). A second factor tending to produce a negative relationship between satisfaction and subsequent fertility is the experience of infant or child deaths, which would reduce satisfaction with life and may also lead to “replacement” fertility (Olsen 1980). The well-established relationships between lower levels of satisfaction with life and infecundity or reduced fecundity provide a third set of reasons for expecting lower satisfaction with life to be an antecedent of lower fertility. Depression and antidepressant medications have been found to be causes of sexual dysfunction, reduced libido, and impaired interpersonal relationships (Phillips and Slaughter 2000; Zemishlany and Weizman 2008). Stress has also been identified as a trigger for spontaneous abortions and still births (Arck et al. 2001; Wisborg et al. 2008). This said, the interrelationship between infecundity and satisfaction with life is complex. In addition to being a determinant of fertility, infecundity has also been found to negatively affect life satisfaction, with this relationship being stronger among women who are not employed than among women who are employed (Bongaarts and Potter 1983; Abbey et al. 1992; McQuillan et al. 2007). However, where the birth of a child followed assisted reproduction, reduced parenting stress, more positive parent-child relationships, and a higher satisfaction with life have been found (Hahn 2001; McQuillan et al. 2007). Contraceptive use is another proximate determinant which may also be related to psychological wellbeing. Increased rates of discontinuation of use of contraceptive injections and implants have been found among women with pre-initiation of use depressive symptoms, although this may be the due to their generally lower levels of satisfaction with their relationships (Westhoff et al. 1998 a, b). Moreover, fears that use of hormonal methods may affect mood may deter women from adopting these methods, and the depressive individual may be particularly sensitive to such fears. In addition to the previously discussed satisfaction with the partner and satisfaction with relationships with any existing children, satisfaction may also arise from a range of other domains of life, including finances, employment, housing, health, the local community, and leisure time (Easterlin 2006; Fowler and Christakis 2008). Whilst the Parr: Satisfaction with life as an antecedent of fertility 640 http://www.demographic-research.org literature provides numerous theoretical reasons to expect relationships between satisfaction with particular domains of life and fertility, there appears to be a dearth of studies which investigate such relationships empirically. Satisfaction with one’s financial situation may be important, since the reasons given by Australians for not having more children are often financial (Weston et al. 2004). Employment opportunities may be a consideration, since pregnancy and children may diminish opportunities to take advantage of them, particularly for women (Parr 2005). A person’s health has been found to be related both to their satisfaction with life and to fertility. Ryff et al. (2006) have shown that a person’s weight and waist-hip ratio have significant negative associations with their psychological well-being, and that a range of other biomarkers also indicates that higher psychological well-being is associated with a lower biological risk of ill health. Obese and overweight men have also been found to have reduced fecundity (Sallmén et al. 2006), and it is plausible the biomarkers Ryff et al. (2006) describe may also be related to the selection of individuals into unions, the continuation of unions, and fertility desires among those within unions. The home in which a person lives may also have ramifications for fertility: those in unsatisfactory living conditions may delay having additional children until they can move to more satisfactory housing or may forgo having them altogether (Felson and Solaun 1975; Bernadi 2005). In Australia house size would be an issue for those contemplating additional children (Newman 2008). The move towards higher density housing in Australia’s larger cities and increasingly delayed leaving of the parental home due to housing unaffordability and unavailability may thus affect fertility (Flateau et al. 2007). The prospect of moving away from a neighbourhood or local community with which one is dissatisfied may similarly lead to a postponement of some childbearing. Satisfaction with the availability of free time may also be a consideration in view of the reductions to such time use which result from additional children (Craig 2005, 2006). For example, Carmichael and Whittaker (2007a) present examples of people who rationalise their childlessness in terms of the extra free time they could enjoy without children. Analyses of the interrelationship between satisfaction with life as a whole and with particular domains in life and subsequent fertility also need to control for a range of variables which may affect both satisfaction with life and fertility. Since marital status and other partnership variables affect both fertility and satisfaction with life, controls for these variables at the time of observation of satisfaction with life are necessary (Carmichael and McDonald 2003; Evans and Kelley 2004; Kohler et al. 2005; Zimmermann and Easterlin 2006). Socioeconomic factors also need to be controlled for, since income, being employed, and the level of education have also been shown to affect both satisfaction with life and fertility (Carmichael and McDonald 2003; Parr 2005; Carroll 2007; Headey et al. 2008). Ethnicity has also been found to be related Demographic Research: Volume 22, Article 21 http://www.demographic-research.org 641 both to expressed satisfaction with life and to fertility (Carmichael and McDonald 2003; Parr 2005, 2006). Moreover, the effects of age need to be controlled for, in view of the documented variation in life satisfaction with age and the variation in fertility with age (Easterlin 2006; ABS 2008). This paper analyses the fertility of Australian women and men, paying particular attention to whether women and men who are more highly satisfied with life are more likely to subsequently have children. The relationship between satisfaction with life and fertility measured at the same point of time will be affected both by the processes through which past levels of satisfaction may have affected fertility and by the effects of the children on satisfaction with life (Kohler et al. 2005). However, the longitudinal structure of the data used allows the levels of satisfaction to be related to fertility at later points in time, thereby ensuring the observations of explanatory satisfaction-related variables cannot have been affected by the fertility response variable (i.e. reverse causality). In this respect the analysis differs from the overwhelming majority of the existing literature on the interrelationships between life satisfaction and fertility. The relationships with subsequent fertility of satisfaction with a range of particular domains in life, including the relationships with the partner and any existing children, the financial situation, employment, housing, health and free time, and with a range of demographic, socioeconomic and cultural variables are also assessed.