Author:Blanco, Laura C.

    The economic literature consistently provides evidence that children have a negative, or at least a non-positive effect, on life satisfaction (Di Telia, MacCulloch & Oswald 2003; Clark & Oswald, 2002), happiness (Glenn & McLanahan, 1981; Alesina, Di Tella & MacCulloch, 2004; Margolis & Myrskyla, 2011), marital satisfaction (Twenge, Campbell & Foster, 2003), spousal love (Grossbard & Mukhopadhyay, 2012) and financial and leisure satisfaction (van Praag, Frijters & Ferrer-i-Carbonell, 2003). The literature also suggests that the effect of children on well-being varies across groups: children contribute negatively to well-being when people face harsh circumstances (Dolan, Peasgood, & White, 2008) and to happiness when people are poor (Alesina et al., 2004) or have lesser financial resources or live in liberal -rather than welfare- states (Margolis & Myrskyla, 2011), while women, high socioeconomic groups and younger birth cohorts report higher marital dissatisfaction after the birth of a child (Twenge et al., 2003). Women also report a higher loss of spousal love than men (Grossbard & Mukhopadhyay, 2012). Similarly, there is some evidence of decreasing returns on having children: the gains in happiness turn negative or non-significant after the second child (Kohler, Behrman & Skytthe, 2005; Myrskyla & Margolis, 2014) and well-being decreases the most for women with three or more children (Clark & Oswald, 2002).

    Even in the long-run, there is mixed evidence on the effect of reproduction on happiness: while Glenn and McLanahan (1981) find a persistent negative effect; Margolis and Myrskyla (2011) argue that the impact of reproduction on happiness varies along the life-cycle: happiness decreases with the number of children during childbearing, flattens and later increases as children grow older, especially for those parents who become sick during their elder years because their children provide them with care and serve as an insurance mechanism. In a more recent paper, Myrskyla and Margolis (2014) find that happiness initially increases prior and during the first two years of the birth, but, on average, it later returns to its pre-birth setpoint. However, these effects vary across group: reproduction exhibits long-term negative effects for young parents and positive long-term effects for better educated and older ones.

    Given the overwhelming evidence that, on average, children are, at best, not detrimental to well-being, it is worth asking why do people reproduce at all. The question seems to be more relevant for women, since they are the ones most negatively affected by reproductive decisions. This paper uses regression analysis to estimate both the preference for children and the observed demand for children for women and men in Costa Rica using the National Surveys on Sexual and Reproductive Health. This allows the identification of the correlates associated with reproductive decisions in a developing country. Previous research has used the data collected in this survey during 2010 to analyze demographic trends and fertility rates (Robles & Gonzalez, 2012a), but to my knowledge the data collected in the 2015 survey has not yet been analyzed. The research also differs from the previous one in that it does not solely focus on the effective demand for children. Instead, it also aims at understanding whether the factors associated with child preferences are the same as those related to the observed demand for children. In order to do this, a double-hurdle model is specified to estimate both the preference and the observed demand for children. Aside from the usual human capital variables, this paper explores whether variables proxying identity, sexuality and contraceptive knowledge are associated with both the preference for and the observed demand for children.

    Indeed, the findings show that the stated preferences and the demand for children respond to different variables. In general, preferences for children seem to be quite fixed around a constant value and are mainly associated with religious beliefs but they are not found to be associated with human capital. In contrast, the observed demand for children is related to an array of social and economic variables. In particular, it is associated with the investment in human capital and sex education. The latter suggest that in an abstract or ideal world, people might carry an innate or unconscious idea about having two offspring, but when this idea is confronted with reality, their economic and educational possibilities as well as their experience gained through age and their identity construction produce an outcome that differs from their imaginariness.


    Despite the negative impact of children on well-being, when asked, most people claim to want to have not one, but two children (Carey & Lopreato, 1995). This is known as the two-child psychology hypothesis, proposed by Lopreato and Yu (1988) and inspired in the Malthusian and Darwinian theories. According to this hypothesis, there is an innate behavioral predisposition according to which women aim at having two surviving offspring. Evolutionary gains are weighed against the costs of reproduction (such as the risks of maternal and child mortality, comfort and happiness) in order to optimize the available resources and continuation of the species. Therefore, as societies develop and infant survival rates approach unity, two-child families become the norm, due to natural selection: that is, if an individual plans to have two surviving children and the probability of survival is high, then she or he will plan to have on average, a little over 2 children. Costa Rica might be an example of this (2). According to the Costa Rican National Surveys on Sexual and Reproductive Health (Centro Centroamericano de Poblacion, 2010, 2015), almost half the population over 15 years of age reports wanting to have two offspring, although it seems they end up having less children than the ones initially desired: Costa Rica's fertility rate reached 1.83 children per women in 2010 (Robles & Gonzalez, 2012a, p.19), which is part of a decreasing trend since the nineteen sixties (Rosero-Bixby & Oberle, 1989; Rosero-Bixby & Casterline, 1995). Aside from this evolutionary explanation, Lopreato and Yu (1988) also argue that participation in the labor market and literacy are the main explanatory factors of fertility, hinting that fertility is not only a response of evolutionary factors, but social ones as well.

    In contrast, Goldstein, Lutz and Testa (2003) argue that, while the ideal number of children has been stable above the replacement rate in Europe (i.e., above 2 children), this might be beginning to change, since it has decreased in time and it is below the replacement rate in Austria and Germany. Further, women also expect to have less children than their ideal number. The reason for this decline is not biological, as in Lopreato and Yu (1988), but rather cultural: as younger cohorts are born and raised in small families and a context of low-fertility norms, their preferences are shaped to favor this type of families, which results in a time trend of declining ideal family size.

    Also drawn from the Malthusian and Darwinian theories is Becker's (1991) quantity-quality model. He argued that his was a more general model because it included the possibility of cultural selection and introduced the cost-benefit analysis at an individual level (rather than just at the genetic one). Becker's (1960; 1991) and Schultz' (1974; 1997) models have become the standard in measuring the demand for children in the study of household economics (3). According to these, a person's utility includes not just the quantity of children wanted, but their quality as well, which is achieved through human capital investment. As a result, the quantity and quality of children are substitute goods. That is, those people whose preferences are inclined towards quality end up having less children because this allows them to invest more human capital in them.

    In its reduced form, these models follow Mincer (1974), so that the quantity of children is a function of education, household income, child mortality, wealth, agricultural work and the region where one lives. Education and income measure the shadow price of time: as people become more educated their opportunity cost of having children (i.e., their labor market earnings) increases, which leads to a reduction in the number of children. At the same time, the increase of educational attainment is associated with an older age of first birth and the acquisition of certain values and attitudes, including a higher attachment to the labor market that might be negatively related to fertility (Neels, Murphy, Bhrolchain & Beaujouan, 2017). The impact of income on the demand for children might also differ due to the household members' specialization, particularly so for women: for those specialized in the labor market, the opportunity cost of a child is higher and, therefore, she would demand less children than a woman who specializes in household production.

    Non-labor income is usually introduced as a proxy for wealth. If children are normal goods, wealthier families would demand more children due to an income effect. However, both income and wealth might also reflect other noneconomic variables. For example, the level of education and norms might influence the demand for children through income (Bagozzi & Van Loo, 1978). These also serve as a signal of family status, which is ultimately exhibited through a proportional higher consumption of status goods. Because this type of goods is more expensive, the opportunity cost of an additional child is higher for high-status than that for low-status people. Therefore, the former would demand less children due to a difference in preferences (Bagozzi & Van Loo, 1978).

    Living in a rural area and working in agriculture are indicative of the cost of...

To continue reading

Request your trial