The Economic Approach to Fertility: A Causal Mediation Analysis

Get Complete Project Material File(s) Now! »

Introduction

This thesis revisits the inconsistency between the empirical negative fertility income eect and standard economic theory. The empirical evidence that fertility andstandars of living are negatively related is inconsistent with standard consumer choice models of household fertility, which postulate that children are normal goods. If children were normal goods then the fertility income eect would be positive. Much study has addressed the determinants of the negative relationship between fertility and income. However, as yet, not a single explanation provided for this puzzle is universally accepted nor is close to being as robust as is commonly claimed (see Jones et al., 2011). For specicity, we illustrate our analysis using fertility and contraceptive calendar data from the 2000, 2004 and 2010 Malawi Demographic Health Surveys (MDHSs).The impact of childbearing on a woman’s earnings has two components. First, raising a child is a time-intensive endevour. In order to devote time to child care,a woman will have to forgo the opportunity to undertake income generating activities. If we assume that the opportunity cost of time is higher for individuals with higher earnings, higher earning women will tend to use more ecient contraceptive techniques in order to avoid their relatively high time cost of child care. As a result higher earning women will tend to have fewer children than their poor counterparts. Second, as noted by Mincer and Polachek (1974) and Mincer and Ofek (1982), work interruptions due to childbearing have a lasting negative eect on the woman’s future earnings, through forgone experience and human capital depreciations. Furthermore, higher earning women also tend to be more educated and have access to a better information network, which put them at an advantage in more eciently applying any specic contraception strategy they may choose. It follows that high income reduces family size through its positive impact on the choice of contraception eciency level. Our work focuses on establishing this positive link.

Analyzing this Puzzle: Some Empirical Commonalities

In general, the interaction between standards of living and fertility can be analyzed using one of the following three approaches: (i) the Malthusian approach, whichconcerns uctuations in fertility in response to uctuations in wages or productivity; (ii) the time series approach, which focuses on the time series properties of average fertility rates within a certain geographical area undergoing some sort of demographic transition; and (iii) the cross-sectional approach, which sets out to document heterogeneity within a cross-section of the population on the relationship between income and fertility (Jones and Tertilt, 2008). Investigations interested in the Malthusian fertility cycles are among the oldest studies in the eld of population economics. In his seminal work, Malthus (1993)
predicted that fertility will remain at relatively high levels, such that rapid population growth would stie improvements in the standard of living. However, this prediction failed to realize, and current versions of the Malthusian model have so far failed to produce a consensus around the direction of the relationship between fertility uctuations and income. While some studies predict a positive correlation between income and fertility over the business cycle (Simon, 1969, 1977), others suggest that fertility may be counter-cyclical (see Butz and Ward, 1979).

Fertility and Wealth

The link between fertility and wealth may be positive or negative. In his wealth ow theory of fertility decline, Caldwell (1976) argues that the fundamental issue in demographic transition is the direction and magnitude of intergenerational wealth ows. When the ow of money, goods, and resources is from the younger to the older generation, an increase in wealth will result in higher fertility rates. But when the ow of wealth is from parents to children, being childless is the most rational economic behaviour. However, for social and psychological reasons, couples would procreate though they have many fewer children. In this situation, wealth is negatively related to fertility.

READ THE IMPLICATIONS OF REPRODUCTIVE NEEDS TO FUTURE HIV COUNSELLING

Fertility and Education

Almost every discussion of fertility dierentials makes reference to the role of education in reproductive decisions. Moreover, the compositional factors proposition,discussed in the previous section, that fertility levels vary between places because dierent people live in dierent settlements, may actually be an indication of dierentials in the levels of education. For many countries, the share of highly educated people is larger in cities than in small towns and rural areas, such that lower fertility in rural areas is mainly the expression of low levels of education (Kulu, 2013). Table 1.3 suggests that educational composition is an important determinant of fertility variation in Malawi. It appears that fertility levels in the country are lowest for highly educated individuals and highest for individuals with no formal education.

Contributions

The study attempts to recast the debate around the fertility-income puzzle into the income, birth control and completed family size space, with the decision of how much to spend on family planning services treated as an investment decision. The treatment of the cost of contraception as an investment decision in human capital, whose returns are subject to uncertainty, is the principal contribution of this work. The interest is in the factors that in uence the time at which a woman chooses to contracept or not; and if she chooses to contracept, what would be her optimal level of contraception eciency. Happel et al. (1984) argue that, in an environment where the expected average number of children is a societal norm, analyzing timing of births is as signicant as investigating completed family sizes.

Acknowledgments
Dedication
1 Introduction
1.1 A Puzzle Brie y Reviewed
1.2 Analyzing this Puzzle: Some Preliminaries
1.3 Analyzing this Puzzle: Some Empirical Commonalities
1.3.1 Fertility and Wealth
1.3.2 Fertility and Settlement
1.3.3 Fertility and Education
1.4 Contributions
1.5 African Exceptionalism
1.6 Organization
2 The Economic Approach to Fertility: A Causal Mediation Analysis
2.1 Introduction
2.2 An Economic Fertility Model
2.2.1 Theoretical Background
2.2.1.1 Fertility and Wages, Income and Wealth
2.2.1.2 Fertility and Education
2.2.1.3 Fertility and Family Planning
2.2.1.4 Fertility and Fecundity
2.3 The Causal Path to Conception or Birth
2.4 Econometric Analysis
2.4.1 The Counterfactual Denition of Causal Eects
2.4.2 Identication
2.4.2.1 Conditional Independence
2.4.2.2 Identication
2.4.3 Estimation
2.4.3.1 Total Eect
2.4.3.2 Price Eect
2.4.3.3 Contraception Eect
2.4.3.4 Direct Income Eect
2.5 Conclusion
3 The Demand for Reproductive Healthcare
3.1 Introduction
3.2 A Full Health Capital Model
3.2.1 Set-Up
3.2.2 Application to Reproductive Healthcare
3.3 Pure Investment Model
3.3.1 The Model
3.3.1.1 Phase Space Analysis
3.3.1.2 The Impact of Health Care Costs and Wage Move-ments
3.3.2 Investment in Reproductive Health
3.4 Empirical Analysis
3.4.1 Econometric Model
3.4.2 Data
3.4.3 Linear Estimation
3.4.4 Nonparametric Estimation
3.5 Conclusion
4 Family Planning and the Probability of Falling Pregnant 116
4.1 Introduction
4.2 The Sequential Nature of Fertility Decisions
4.3 The Childbirth Hazard Rates
4.3.1 The Model
4.3.2 The Hazard Rates of Giving Birth
4.4 The Fertility Eect of Family Planning
4.4.1 The Outcome Equation
4.4.2 Choice of a Contraception Method
4.4.3 Identication of the Fertility Eect of Birth Control
4.4.3.1 Policy Eects
4.4.4 Estimation Strategy
4.5 Application
4.5.1 Data
4.5.2 Estimation Results
4.5.2.1 Estimated Hazards and Probabilities of First Childbirth
4.6 Conclusion
5 Conclusion