Curbing rapid population upsurge has become important to policy makers all over the world. This is particularly relevant in developing countries where rapid population growth could result in increased poverty level and reduced wellbeing, in general. On a broader scale, continuing population growth could hinder a country’s economic performance as this raises pressure on already limited public infrastructure and places extra strain on national government budgets [1]. The nexus between population growth, household welfare and economic performance is made clear in the quantity-quality tradeoff theory by Becker and Lewis [3]; Becker and Tomes [4]; Becker and Barro [2] and Willis [26]. The theory posits that there exists a trade-off between having more children and raising quality children. They hypothesize that households with lower family size are more likely to have higher quality of life relative to households with larger family size. For instance, smaller households are more likely to provide better education and health care to its members than larger households, ceteris paribus.
Moreover, households with relatively smaller family sizes are more likely to benefit from demographic dividends. For instance, households with smaller family sizes have less pressure on scarce resources which could be invested in economic ventures and family welfare. Such families are also likely to have higher per-capita income than larger families [19]. This suggests that when population growth is not controlled it could have negative implications for the economy as a whole and the welfare of individual households within the economy. Aside its economic implications, increased fertility rates could have devastating effect on maternal and child health [10, 18].
Attempts to control population growth have over the years focused on meeting contraceptive needs of the population and improving public education on the awareness and proper use of modern contraceptives. In Ghana, population reduction policies over the years include the Contraceptive Social Marketing (CSM) project (1987–1990), the Ghana Family Planning and Health Programme (FPHP) (1990–1996) and the Ghana Population and AIDS Project (GHANAPA) (1996–2000) [14]. A more recent policy effort targets reducing fertility rate to 3.0 by the year 2020 as well as reducing population growth to 1.5% by the same year. While the early policies resulted in a marked decline in fertility by the 1990s, recent data suggest a slight rise in the fertility rate. The total fertility rate (TFR) declined from 6.4 births per woman in 1988 to 5.2 in 1993 then dropped further to 4.0 in 2008. However, in 2014, TFR increased marginally to 4.2 births per woman. This raises concerns about progress towards achieving the 3.0 target [11].
An important step to achieve the set population growth target is to understand the practical determinants, not only of fertility rate but also fertility preferences among women and couples, by extension. Several socioeconomic and demographic factors have been identified in previous empirical studies. Some of these include the number of children ever born, education [24] and infant mortality [21]. Other factors include maternal age, wealth status of the household, and engagement in economic activities, among others [5, 16, 17, 22].
In developing countries such as Ghana, infant and child mortality continue to significantly influence reproductive behavior among both men and women. This is largely due to the high rates of infant and child mortality in these countries [25]. It is therefore important that linkages between childhood mortality and reproductive behavior is well understood. Another relevant factor that has received minimal attention in the empirical literature is the nexus between women’s bargaining power and reproductive health behavior. Again, this is particularly important in developing countries where women are often considered to have no or little influence on fertility or fertility preferences. While a clear understanding of these relationships will be crucial for effective targeting and policy direction, very little evidence exists in developing countries such as Ghana.
Conceptual framework
The conceptual linkages between childhood mortality and fertility preferences hinges on two strands of theoretical literature. These theoretical predictions seek to explain the behavioral responses of individuals or households to childhood mortality [13]. The first hypothesis posits that households choose to have additional children to replace dead children so that marginal changes in net fertility due to child mortality is zero. This implies that households have targeted numbers of children and that reductions below this target generates disutility. The hypothesis is known in the literature as the replacement hypothesis [7, 13]. The second hypothesis assumes that households have or prefer to have more children as precaution when uncertainties about child survival are high. This is particularly pertinent in developing countries where child mortality remains a major public policy concern. This implies that families facing relatively higher child mortality risk will adapt their fertility behavior accordingly. This hypothesis is known as the child survival hypothesis or hoarding motive [6]. In other studies, this hypothesis has been referred to as ‘anticipatory effect’ [13] or insurance effect [7]. While the former hypothesis is predicted to impact only total fertility positively, the latter is expected to affect both total and net fertility positively. In this study, we hypothesize that child mortality and related uncertainties have positive impact on fertility preferences. We further posit that women’s intra-household bargaining power has a role to play in the child mortality – fertility preference nexus. Against this backdrop, our study deviates from existing studies mainly in our focus on fertility preferences instead of actual fertility. While we consider fertility preference to mean the number of children a woman wishes or desires to have, we define fertility as the actual number of children a woman already has. To this end we sought to provide answers to two research questions; (i) do childhood mortality and women’s bargaining power affect fertility preferences? (ii) are there interactive effects between child mortality, women’s bargaining power and fertility preferences?
Empirical review
The literature on fertility preferences and its determinants have evolved over the years. While some researchers focus on just preferences, others have looked at stability in these preferences over time. As indicated earlier, several socioeconomic and demographic variables have been identified in these studies. Broadly, these factors encompass both gender characteristics and household level factors. For instance, Nyarko et al. [21] investigated the link between male child loss and subsequent fertility in Ghana using the 1993 Ghana Demographic and Health Survey (DHS). The authors sought to find out if there is preference for male children. Using parity progression ratios and time hazard models, the findings of the study suggest that while the death of an infant induced mothers to have another child, the death of a male child reduced the birth interval greatly. Similarly, Tawiah [24] also used the 1993 Ghana DHS data to identify socioeconomic and demographic factors that influence child preferences with specific emphasis on male child preference. The findings showed that male child preference was significantly associated with level of education, region of residence, experience of child loss and religion.
In recent years, there has been growing interest in the relationship between child mortality and fertility in developing countries. Bousmah [6] used micro level data from a demographic and health surveillance system in a rural community in Senegal to explore the relationship between child mortality and fertility. Results from the standard Poisson regression model confirmed the child survival hypothesis with a positive effect of child mortality on both total and net fertility. In a related study, Bousmah [7] used longitudinal data from the same surveillance system in Senegal and found that both the replacement hypothesis and hoarding motives were true in the case of Senegal. The author confirmed that morbidity and mortality from malaria jointly affected subsequent fertility choices positively. Following the Indian Ocean Tsunami in 2004, Nobles et al. [20] found that women who lost one or more children are likely to bear additional children.
In a cross-country study, Canning et al. [8] used data from 46 low and middle-income countries to estimate the relationship between child mortality and fertility. It was evident from their study that reducing child mortality will likely reduce the number of children born but increase the number of surviving children and therefore lead to a rising population growth. The authors also found that where an individual’s fertility choices affect the fertility choices of others (they call this interdependent fertility preferences), the net effect of child mortality on population growth rate is zero. Sennott and Yeatman [23] examined the level and direction of changes in fertility preferences among women. Using panel data from Malawi and multinomial logit model, the authors found that having a child, entering a serious relationship and changes in finances of the household were associated with changes in the level and direction of fertility preferences. In a more qualitative analysis, DeRose et al. [9] investigated the role of perceptions of power in reproductive conflict in a woman’s fertility desires. Using focus group discussions among young Ghanaian men and women, the findings showed that young women’s expected influence were limited to situations where their fertility desires conform to normative expectations. In another study to test the stability in individual fertility preference over time, Kodzi et al. [15] used panel data from Ghana. The authors found that about 20% of respondents changed their fertility preference over time.
The discussions of the literature so far suggest previous studies mostly focused on the relationship between child mortality and fertility. Studies that analyzed fertility preference failed to account for the effect of child mortality. Moreover, the potential role of women’s bargaining power in this relationship has been absent in the literature. We therefore contribute to existing studies in developing countries in this regard. We explored the relationship between child mortality, bargaining power and fertility preferences among Ghanaian women in their reproductive age. We also estimated the interactive effect of these relationship.