آیا تنها یک موضوع برای شخصیت وجود دارد؟ در مورد نقش بهزیستن ذهنی در رفتار باروری

Abstract This paper analyses the role played by individual subjective well-being (SWB) in childbearing behavior. We use the German Socio Economic Panel (GSOEP) survey, which contains repeated information about SWB, childbearing events and, importantly, also measures of respondents’ personality, to estimate the way SWB matters for having a(nother) child, controlling for personality traits (PTs). We find that SWB positively predicts childbearing for women and men, with the effect significant (and sizeable) for both genders only for the second child. Furthermore, we assure that – although PTs are a strong component of SWB variability – the effect of SWB on fertility is not determined by PTs.

. Introduction A burgeoning recent literature is focusing on the way SWB associates with childbearing (e.g., Aassve et al., 2012, Billari and Kohler, 2009, Kohler et al., 2005, Margolis and Myrskylä, 2011 and Myrskylä and Margolis, 2014). At the same time, on one hand, several studies have suggested that personality matters for childbearing behavior (e.g., Dijkstra and Barelds, 2009, Jokela et al., 2009 and Jokela et al., 2011), on the other hand, a key finding from psychology is that subjective well-being (SWB henceforth) is strongly mediated by the respondents’ personality (e.g., Costa and McCrae, 1980). So far, however, there has not been any systematic joint analysis of the relationship of these dimensions. Aiming at bridging the gap between these research strands, this paper tackles the issue of the effect of SWB on fertility controlling for the role of personality traits (PTs hereafter). There are many important reasons for considering the relationship among SWB, personality and fertility behavior. From the demographic side, the key interest lies in the fact that most developed countries are now facing fertility levels well below the replacement rate, but with stark differences across countries. The recent trends have sparked a very lively debate, not only in terms of policy perspectives, but also for the theoretical understanding of fertility behavior: existing theories of fertility dynamics no longer adequately respond to why people still have children in contemporary advanced societies, and also why, in some societies, fertility is even rebounding, as appears to be the case in Nordic and Anglo-Saxon countries as well as in France (Goldstein et al., 2009). Nevertheless, the interest in how fertility behavior links with SWB stems from the fact that childbearing in modern societies is very much viewed as part of a series of choices aimed at the self-realization of the individual. As Van de Kaa (1987) pointed out already more than three decades ago, one side of new demographic behavior is that individuals put stronger importance to their own realization and their psychological well-being. Consequently strong emphasis has been given to the way SWB plays a role in how individuals make decisions about childbearing. The cornerstone in this literature, although not always expressed explicitly, is that individuals’ decision-making process derives from the quest for happiness, of which offspring presumably make up an important component. A corollary of this argument is that fertility is higher in those societies where couples derive a higher level of SWB from childbearing (Aassve et al., 2015). An important weakness of this literature, however, is the implicit assumption that SWB is a function of childbearing. The analysis is consequently based on regressions where the dependent variable is the standard overall measure of SWB, typically measured by self-reported happiness or satisfaction. But this approach appears to have come about more as an empirical bandwagon rather than being derived from strong theoretical arguments. Quite on the contrary, there are compelling arguments for taking the opposite approach: childbearing is a decision that couples make in which their mental well-being might very well play an important role. Intuitively, it would make sense if optimistic and satisfied people feel they are better prepared to start the monumental task of forming a family, in which case the level of satisfaction should be positively related to the likelihood of childbearing. However, as the recent literature exploiting panel data has pointed out, any relationship and causal effect between childbearing and SWB would critically depend on the moment in which SWB is compared. One frequently observed pattern is that SWB increases prior to childbearing, whereas following the childbearing event there is a great deal of adaptation, and then, in many instances, any increase in SWB is neutralized after some time. In other words, there is ample evidence suggesting that there is an anticipation effect, in the sense that SWB may increase as a result of the anticipation of the childbearing event (Balbo and Arpino, 2014, Clark et al., 2008 and Myrskylä and Margolis, 2014). For instance, if one measures SWB at the moment of childbearing and one or two years after, it is most likely a negative trend. In any case, the fact that one often observes an anticipation effect rises the interesting issue to what extent increased SWB associates with a higher likelihood of childbearing. Here, we tackle this issue head on, by considering the impact of SWB on the likelihood of experiencing childbearing in the consequent time period. The past literature, indicating the presence of an anticipation effect, also suggests that SWB is potentially endogenous with respect to childbearing event. The vast majority of studies considers however, the impact of childbearing on SWB, and does not consider directly how any change in SWB may lead to a higher likelihood of childbearing. In the present paper, performed with the German Socio-Economic Panel Survey (GSOEP), the dependent variable is derived from observed childbearing events. More precisely, we run separate models by child parity, since low overall fertility is driven both by high rates of voluntary childlessness and low progression from the first to the second birth (Frejka, 2008). Our main explanatory variable is the SWB before the pregnancy. We also embed PTs into our analysis by using the standard “Big-five” construct (Costa and McCrae, 1980), being consequently able to assess to what extent SWB matters for childbearing, but explicitly elaborating on the role played by personality. Differently from Myrskylä and Margolis (2014) and Clark et al. (2008), we are interested on the effect of the SWB on the decision of having a child, and not on the consequences of this event on the parental SWB. Other than being a longitudinal survey of very high quality, the German case serves as a particularly interesting example when considering fertility behavior. Despite being the economic locomotive of Europe, the country also suffers from a long lasting very low fertility, and, with the current Total Fertility Rate (TFR) of less than 1.4 children per woman (Eurostat data for 2012), it is joining the club of lowest-low fertility countries, such as those of the Southern and Eastern Europe (Billari and Kohler, 2004).

. Conclusions In this paper we have analyzed the effect of SWB on fertility, with the control of PTs, using suitable data from the German Socio Economic Panel Survey. Our analysis suggests that an increase in SWB may indeed result in an increase of the likelihood of having a child. In addition, we have shown that even if fertility is predicted by both SWB and PTs, the latter do not unequivocally determine the behavior in which we are interested. The key finding from our analysis is, therefore, that SWB certainly affects the likelihood of childbearing, and, even if the PTs are significant in terms of childbearing behavior, they do not alter the hypothesis that SWB increases the likelihood of having a(nother) child. This fact that reported variation in SWB associated with childbearing behavior is not only a matter of differences in PTs is undeniably an important finding, not least because it suggests that there is room for policy measures aiming at sustaining fertility (with the obvious caveat, in establishing this causal relationship, that our study is based only on a German sample). Furthermore, even if PTs are measured at two different points in time, allowing us to use an average of the two, we are able to show that these traits remain remarkably stable during the childbearing age. Another important contribution of this paper is that we consider the effects by parity, i.e., how the potential effect of SWB on childbearing differs by whether we consider the first, second or the third birth. It is a relevant result that the effect of SWB on fertility is only significant for the progression to the second birth. Thus, SWB does not appear to matter for the transition to become a parent for the first time, nor parity progressions beyond the second birth, and these results appear rather robust. The intuition behind this result is relatively straight forward. Becoming a parent is for most individuals an important life goal and may be a strong marker of personal success, thereby leading to social recognition and feelings of pride, esteem and satisfaction (e.g., Michalos, 1985). The desire for parenthood may also have a strong biological component, gratifying basic human needs (Veenhoven, 1996). Consequently, the variation in SWB may not have a large impact on this decision. Having a second child in contrast, is a question of quantity where the decision is very likely is a function of having become a parent first time around – and importantly – its aftermath. The experience of becoming a parent is one of the most important in shaping individual's life course and, reasonably, fertility intentions and expectations are modified by the experience of parenting itself, in particular after the “revolutionary” experience of having the first child. The onset of parenthood increases family commitment, reducing time for working and leisure, and possibly, at least in the short term, also decreases individuals’ overall satisfaction with life. In short, for childless couples it is rather difficult to anticipate how it will be to be a parent, whereas once parenthood has taken place, couples will have gained superior information, which they use for deciding on having a second child. If the experience of the first childbirth is unexpectedly harder (or, ideally, even better) than what they had foreseen, subjective wellbeing may decrease, and also make them revise their fertility plans. A fruitful avenue for further research would therefore be to look into the way subjective wellbeing gets affected by the first childbirth. In terms of fertility trends, and especially with an eye to the very low fertility setting of Germany, this finding is of great importance. In Germany total fertility rates have been declined constantly throughout cohorts. Starting with a steady decline from cohorts of women born during the 30s of the last century, for whom final fertility at the end of reproductive life was above replacement level, has reached 1.5 children on average. The level of definitive childlessness, although also increasing for these cohorts, is nevertheless moderate (Miettinen et al., 2015) and it is now around 15 percent. Therefore, the German fertility decline is mainly due to the reduction of women with two or more children. The increase in one-child families is prominent in lowest-low fertility countries and it is becoming the distinctive trait compared to those countries where the fertility rate is still near the replacement rate such as the Nordic ones (and also France), where the progression ratios to first and especially second births are much higher ( Frejka, 2008). In other words, the fundamental driver for low fertility lies in the lack of progression from the first to the second child. The fact that SWB following the first child matters for the progression to the second birth means that this parity should constitute the main target for family and fertility policies. We also show that any enhance in SWB increases the likelihood of childbearing, and this effect is not merely an anticipation of the childbearing event itself (e.g., using the one period lag if SWB has a very similar impact on childbearing compared to the two-period lagged version). Starting from the average SWB level for both men and women (which stands around seven points in a 0–10 scale), a one-point increase in SWB would raise the probability of having a second child by 27 percent for women and 22 percent for men. Establishing exactly which policy mix would achieve such an increase of at least one point on the average SWB is beyond the scope of this paper. The existing socio-demographic literature regarding very low fertility settings highlights the crucial importance, both for SWB and fertility behavior, of several aspects concerning women's role in the labor market. Other than generosity of social security, and more generally labor market conditions, a fruitful direction for further research lies in disentangling how the reconciliation between work, family and childbearing relates to individual SWB and to fertility behavior (as already indicated in Kohler, 2012 and Kotowska et al., 2010), with the specific focus on the drivers of SWB for those parents having had the first child, in order to understand how it affects progression to having the second.