Agency in Fertility Decisions in Western Europe Du

Authors: David Sven Reher Glenn Sandström Alberto SanzGimeno Frans W A van Poppel

Publish Date: 2017/01/09

Volume: 54, Issue: 1, Pages: 3-22

PDF Link

Abstract

We use a set of linked reproductive histories taken from Sweden the Netherlands and Spain for the period 1871–1960 to address key issues regarding how reproductive change was linked specifically to mortality and survivorship and more generally to individual agency Using eventhistory analysis this study investigates how the propensity to have additional children was influenced by the number of surviving offspring when reproductive decisions were made The results suggest that couples were continuously regulating their fertility to achieve reproductive goals Families experiencing child fatalities show significant increases in the hazard of additional births In addition the sex composition of the surviving sibset also appears to have influenced reproductive decisions in a significant but changing way The findings offer strong proof of active decisionmaking during the demographic transition and provide an important contribution to the literature on the role of mortality for reproductive changeThis study addresses the role of childhood mortality for reproductive decisionmaking during the demographic transition It is widely held that before the transition reproductive decisions tended to be made at a societal or group level often by means of changes in marriage timing and intensity Mason 1997 However as the transition progressed reproductive decisions became increasingly individual and familybased responding to concrete conditions of families more than to accepted societal norms Reher 2011 From this perspective the demographic transition can be viewed as a key episode in the progress of human agency central to all processes of modernization Showing the existence of active decisionmaking during the transition empirically especially with respect to the importance of mortality change for reproductive decisions however has not been a simple matter until fairly recently this type of decisionmaking was more basic postulate than a proven cornerstone of transition theory In this article we marshal important new empirical evidence enabling us to view this issue from a different perspective and leading to a nuanced view of the importance of agency and how it changed over the transitionThe role of mortality has long been central to the conception of the demographic transition mostly because of the major changes in mortality during this period Inherent in Notestein’s 1945 original formulation of demographic transition theory especially as subsequently modified by Kingsley Davis 1963 and Ansley Coale 1973 1986 is the role of mortality change as a key factor triggering fertility decline in the late nineteenth century both at a societal and a familial level Reher 2004 In Coale’s formulation of this issue the decline of childhood mortality was a necessary precondition for any sustained decline in marital fertility because it led to the perception that reduced fertility was advantageous see Mason 1997446–447 see also Kirk 1996 although at a societal level other more general factors might cloud this intrinsic link Coale 197354 62–66 Coale 1986 Showing that survival outcomes were a key part of reproductive decisionmaking is a way of underscoring the importance of choice and human agency during the transition Underlying this argument is the supposition that couples generally desired a given number of surviving children which tended to be small as evidenced by prevailing growth rates prior to the demographic transition This cornerstone of transition theory subsequently came under severe criticism even as other scholars have insisted on its validity1Mean number of children ever born and indicators of child survival in Sweden Spain and the Netherlands by marriage cohorts 1870–1949 Mean is based on couples observed until the end of their reproductive history Cohorts married in 1930–1949 are aggregated for Spain Source Sweden POPLINK Database Demographic Database Umeå University 2015 Spain Aranjuez Civil Registers Netherlands Historical Sample of the Netherlands HSNWhen addressing the connection between childhood mortality and fertility outcomes it is important to differentiate between several possible mechanisms by which declining mortality can influence fertility Previous literature has often highlighted three types of factors two of which are pertinent for the issues at hand see eg Palloni and Rafalimanana 1999 First increased child survival has an individuallevel biological effect in terms of longer periods of breastfeeding when infants survive If lactation is not interrupted because of infant death postpartum infecundity is prolonged and thus tends to delay conception even in the absence of active contraception Knodel and Van de Walle 1967 This effect would exist independent of any desire for fertility limitation and therefore choice is not involvedSecond individuallevel behavioral effects operate when couples have preferences regarding family size and at least partially effective means of adjusting fertility These preferences are a response to the number of surviving children rather than to the number of children ever born using two types of strategy One is a shortterm strategy designed to replace a child who has just died This sort of child replacement can be found in many populations past and present although in the past it is very difficult to distinguish it from the mechanical biological response to the early cessation of breastfeeding described earlier Beyond this immediate effect if couples had fertility goals and the ability to implement them their fertility decisions would tend to be based on the overall survival status of their sibset rather than solely on the outcome of the previous birth Both child replacement and reactions to the overall number of surviving children are indicators of the existence of fertility goals—although in the first case it would be at least partly in response to shortterm goals together with the biological effects mentioned earlierAnother related mechanism is an insurance or a hoarding effect Preston 1978 see also Alter 1988 When mortality is high and variable prudent couples will try to overshoot their actual target family size to ensure a minimum number of surviving children that eventually reach adulthood When levels of childhood mortality decrease sufficiently the need for this type of insurance behavior decreases and couples can choose to both stop and space their births at lower parities given that they are confident that most or all of their children will survive to adulthoodAlthough the biological or lactation effect should be found mainly in a natural fertility setting both the replacement effect and the hoarding effect should be present when fertility control albeit inefficient is a realistic possibility as long as childhood mortality is high and unstable enough to make hoarding a sensible strategy We can expect that the ability of people to influence their reproductive outcomes efficiently improved as the demographic transition progressed and as stopping behavior became widely adopted Anderton and Bean 1985 Gillis and Tilly 199215 Whenever possible researchers should distinguish between these different forms of shortterm and longterm behavioral responses to childhood mortality

Keywords: