In 1840, Festy made a distinction between countries with late marriages, over the age of 27 – Scandinavian countries (Denmark, Norway, Sweden), Belgium, the Netherlands, Switzerland – and countries with early marriages, under the age of 26 – Finland, Great Britain, Germany, Austria, Southern Europe (Spain, Italy and Portugal) and North America (Festy 1979, p. 28). He showed the complex relationship between late marriage, low fertility, breastfeeding habits and differences in fecundity between the rich and the poor in 1870 (Festy 1979, p. 42). In addition, “intermediate fertility variables”, which are biological, social and economic as well as cultural, work together in the reduction of fertility (Davis and Blake 1956). Bongaarts calculated that they could be reduced to four determining variables, which alone explain the majority of changes in fertility: the age of the first union, the duration of breastfeeding, the use of contraception methods and abortion (Bongaarts 1978). In Asian and Latin American countries, modern contraception and abortion have played a predominant role in birth limitation, because “the contraceptive revolution” reached all the regions in the world (Leridon and Toulemon 2002). In Africa, the duration of prolonged breastfeeding has a reducing effect on the number of children per couple (Tabutin and Schoumaker 2004).
Coale grouped the main macrosocial causes that explain fertility changes in the RWA model (Coale 1973):
– a first factor is the reasoned choice to give birth to a child. The advantages and disadvantages of an additional birth should be accepted by couples. This factor is called readiness (R). It determines the beginning of controlled fertility, in particular due to economic reasons;
– a second factor is the perception that low fertility is socially and economically beneficial for mothers and fathers. This factor is called willingness (W). It points to the social legitimacy of fertility control within families and a normative/cultural change in values;
– the third factor is the availability of birth control methods and their effective use. This factor is called ability (A). It depends on the diffusion of contraception methods, birth control infrastructure, and health and sanitation policies.
According to Coale, the main factors for the fertility transition to come about must simultaneously be economic, cultural and institutional. The three RWA preconditions act mutually, and if only one is lacking, the decline in fertility will not take place (Coale 1973; Lesthaeghe and Vanderhoeft 2001). As not all individuals change their behavior at the same time, and only pioneer groups adopt them, the diffusion toward the rest of the population is conveyed by means of “social interactions” (Bongaarts and Cotts Watkins 1996). But it can be slowed down or accelerated depending on the cultural, religious and linguistic standards of each region11. Bourgeois-Pichat (1976) stressed the importance of group effects: “Of course, the couple decides, but they do so according to the social criteria and cultural heritage which, although quickly driven out, paradoxically maintain their rights” (p. 1077).
To explain contemporary demographic transitions, Coale’s (1973) diffusion model has been enriched by many authors. McNicoll (1980) analyzed the consequences of political and institutional factors in the behavior of families. Becker (1991) explained the limitation of births within couples due to microeconomic factors. Caldwell (1982) highlighted the role of intergenerational wealth flows: these are transferred from children to parents in high-fertility contexts and their direction changes, from parents to children, when the costs in children’s education and health increase, leading to fertility control. From a psychosocial perspective, Fawcett (1983) proposed to take into account the perceptions of the value of children. Simons (1982) was interested in the impact of religious practice on reproductive behavior.
Another theoretical proposition is that of the second demographic transition, which refers to the diversification of family configurations in post-modern societies, marked by the increase in cohabitation outside marriage, the increase in divorces and new cohabitations after union breakdowns, the delay in age for having the first child, the increase in births outside marriage and childless couples. These changes in nuptiality and fertility are probably derived from the primacy of individual choices, more equitable gender relations, greater autonomy for women, macrostructural factors, microeconomic calculations and new cultural models and values, which are expressed individually and collectively (Lesthaeghe 2010). While these movements may have taken place in some European countries, they have not yet spread to all other parts of the world, where family and gender systems are extremely resistant to change.
1.4.2. Questions
The observation of contemporary demographic evolution raises questions about the universal validity of the three paradigms by Chesnais to explain the demographic transition: first, on the likelihood of the stabilization hypothesis when the transition is complete; then, on the reproductive transition in two phases (limitation of marriages, limitation of births); finally, on the influence of entry into modern economic growth (Chesnais 1986a, p. 1061).
The first question concerning the transition’s completion is based on the observation of recent upward and downward oscillations in contemporary low fertility. In 1979, Bourgeois-Pichat asked: “Is the current decline in fertility in Europe part of the demographic transition model?” To this question, he gave the following answer:
What we are currently observing seems to indicate that this last phase is not characterized by constant fertility, but rather by a succession of waves produced by modifications in family structures, which themselves result from variations in fertility. (Bourgeois-Pichat 1979, pp. 293–294)
Vallin also stressed that “the evolutions observed in the most advanced countries are moving further and further away from the model of demographic transition, and, while convergence is still possible, it is unlikely that it will lead to stabilization” (Vallin 2003, p. 75). Moreover, Myrskylä et al. (2009) have shown that the relationship between fertility in the most developed countries and the high level in the Human Development Index (HDI)12 has become positive, which explains the recovery in fertility levels in countries having reached the higher level of development.
The second question concerns the paradigm of the two-phase reproduction transition module (limitation of marriages, and consequent limitation of births), which is not always confirmed. For example, in Mexico, nuptiality is still early: in the 30 generations between 1951 and 1980, the median age for the first union13 of Mexican women was stable at age 21, despite a significant increase in their education. This did not prevent fertility from decreasing rapidly, and other factors explain this, such as the diffusion of modern contraception methods, including female sterilization (Zavala and Paéz 2016). A comparison between Algeria and Mexico showed that the reduction in fertility was similar between 1970 and 2005: decreasing from 7 to 2.4 children per woman. However, the average age for the first union evolved differently throughout the period: while it remained roughly stable in Mexico (from 21 to 23 years old), it sharply increased in Algeria (from 18 to 30 years old) (Cosio Zavala 2012).
Finally, the other paradigm that does not hold true everywhere is that of economic growth to explain changes in fertility. As a matter of fact, in developing countries, there are two fertility transition models: the first model is that of the urban and educated population, which controls their births, following an improvement in their standard of living. The second model, Malthusian poverty, has its origin in weak economic growth and poor living conditions. This has been the case in Mexico for three decades, in a situation of interminable economic crisis, where families limit the number of children because they lack the means to raise them (Cosio Zavala 1996). These two fertility transition models are also present in sub-Saharan Africa: in Kenya and Côte d’Ivoire (Vimard and Fassassi 2001), Benin (Capo-Chichi 1999), Nigeria (Caldwell et al. 1992) and in a number of countries in the region (Lesthaeghe and Jolly 1995). Therefore, the postulate of a negative relationship between fertility levels and economic growth does not apply to developing countries or to highly developed populations, as mentioned above (Myrskylä et al. 2009). It is, therefore, not a global relationship,
