Demography and the Economy

by John B. Shoven

Overview

Demographics is a vital field of study for understanding social and economic change and it has attracted attention in recent years as concerns have grown over the aging populations of developed nations. Demographic studies help make sense of key aspects of the economy, offering insight into trends in fertility, mortality, immigration, and labor force participation, as well as age, gender, and race specific trends in health and disability.

Demography and the Economy explores the connections between demography and economics, paying special attention to what demographic trends can reveal about the sustainability of traditional social security programs and the larger implications for economic growth. The volume brings together some of the leading scholars working at the border between the two disciplines, and it provides an eclectic overview of both fields. Contributors also offer deeper analysis of a variety of issues such as the impact of greater wealth on choices about marriage and childbearing and the effects of aging populations on housing prices, Social Security, and Medicare.

Product Details

ISBN-13:	9780226754727
Publisher:	University of Chicago Press
Publication date:	01/01/2011
Series:	National Bureau of Economic Research Conference Report
Pages:	448
Product dimensions:	6.10(w) x 9.10(h) x 1.40(d)

About the Author

John B. Shoven is the Charles R. Schwab Professor of Economics at Stanford University, the Wallace R. Hawley Director of the Stanford Institute for Economic Policy Research, and a research associate of the NBER.

Read an Excerpt

Demography and the Economy

---

The University of Chicago Press

Copyright © 2011 National Bureau of Economic Research
All right reserved.
ISBN: 978-0-226-75472-7

---

Chapter One

The Future of American Fertility Samuel H. Preston and Caroline Sten Hartnett

The level of fertility in a population is the principal determinant of the shape of its age structure, which in turn is a critical factor in the terms of trade within a pay-as-you-go system of public pensions. Simulations done by the Social Security Administration (SSA) show that the seventy-five-year actuarial balance of the social security system would be higher by $2.6 trillion in present value if fertility were high (2.3 children/ woman) rather than low (1.7) (compiled from Trustees [2007]). Partly because of their age structural consequences, national fertility levels are considered "too low" by a majority of governments in developed countries (Kohler, Billari, and Ortega 2006).

This chapter reviews the major factors that appear to be affecting fertility levels in the United States, with an eye toward making defensible statements about future directions of fertility. The subject covers a vast disciplinary range including demography, economics, sociology, public health, reproductive biology, evolutionary biology, political science, and psychology. There is no single, widely accepted framework for analyzing the determinants of fertility at the level of a population. In its place, we will pursue an eclectic, inductive approach, surveying the landscape of fertility variation in search of clues about its principal drivers. Our search considers variation over time and space and across individuals.

1.1 Why Do People Have Children in the Twenty-First Century?

It is useful to begin with this provocative question posed by Morgan and King (2001). If there were no compelling answer to the question, we would have to confront the possibility that levels of fertility will approach zero. Clearly, the answer to the question does not lie in the domain of finance, since children are very costly and probably always have been. Early suggestions that children were a net economic asset in hunter-gatherer or subsistence economies appear to have been inaccurate, although children's greater contribution to the family economy in such circumstance reduced their net costs relative to children in the present (Kaplan 1994).

Sociologists have usefully distinguished between childbearing aimed at satisfying social expectations and childbearing aimed at self-fulfillment. Thornton and Young-DeMarco's (2001) review of trends in attitudes about one's own childbearing and that of others shows a huge reduction during the 1960s and 1970s in the degree of "oughtness" regarding fertility. While the desire to satisfy social expectations has not disappeared, people began to perceive less social pressure to bear children and to have less rigid expectations of others' performance. Increasingly, people justified childbearing in terms of its impact on their personal well-being, satisfaction, and happiness. One of the instrumental features of children that several sociologists have stressed is their value in forming social networks (Schoen et al. 1997).

In view of the imperatives of reproduction for the survival of a species, it would be surprising if the rewards from childbearing and child-rearing did not have a deep evolutionary basis imprinted in human biology (Foster 2000). Recent investigations in psychology help to clarify the nature of these rewards. Bartels and Zeki (2004) use fMRI imaging to measure brain activity in mothers when they viewed pictures of their own children and those of acquainted children and adults. Pictures of their own children, but not of others, activated regions of the brain rich in oxytocin and vasopressin receptors—neurohormones associated with pair-bonding—while deactivating regions associated with negative emotions and social judgment. Animal studies confirm the central role of oxytocin and vasopressin in attachment and bonding (Carter et al. 2005).

Mothers are aware of the intense emotions evoked by their children. "The Motherhood Study," a nationally representative telephone survey of 2,009 mothers, found that 93 percent agreed with the statement that "I have an overwhelming love for my children unlike anything I feel for anyone else." Eighty-one percent said that they were very satisfied with their life as a mother and an equal percentage agreed that "being a mother is the most important thing that I do" (Erickson and Aird 2005). The potential rewards of parenthood—presumably social as well as emotional—are acknowledged by high school seniors, three quarters of whom believe that motherhood and fatherhood will be fulfilling. Between 1976 and 1977 and 1997 and 1998, the percentage so reporting rose by eleven points for women and seven points for men (Thornton and Young-DeMarco 2001). The increase was greatest for females and males whose fathers had attended college (Sayer, Wright, and Edin 2003).

It is possible that the rewards and costs of childbearing are not fully appreciated until one has a child. One ethnographic study reports that mothers, in fact, did not anticipate how completely they would fall in love with their offspring (McMahon 1995), which raises the possibility that the motivations for having the first child are systematically different from those of subsequent children. A study in Bulgaria (Buhler 2006) concluded that the principal attitudes predictive of having a first child were beliefs that it would strengthen relations with partner and parents, whereas the principal attitude predictive of a second child for both men and women was the perception that it would bring "increased joy and satisfaction in life." Companionship for the first child is also often cited as a motivation for having a second child (Fawcett 1983). In a careful study of reported happiness among monozygotic twins in Denmark, having one child was found to increase the happiness of young women, but there was no increment in happiness from additional children (Kohler, Behrman, and Skytthe 2005). Once partnership status was controlled, a man's happiness was unaffected by the number of children he had, including the first.

1.2 Recent Trends in American Fertility

The most common measure of fertility is the period total fertility rate (TFR), which indicates how many children would be born to a woman who survived to the end of her reproductive years and experienced at each age the observed age-specific fertility rate of a particular period. The level of the total fertility rate that allows each generation to replace itself exactly is approximately 2.08 children per woman. Figure 1.1 shows the value of the TFR in the United States since 1928. With virtually no interruption except the post-World War II baby boom, the TFR fell continuously from 1820 to 1975 (not shown). Since 1989 it has remained in the narrow range of 1.98 to 2.10. Figure 1.1 also shows the average number of children ever born to cohorts who completed their childbearing and were aged twenty-six during the year shown on the x-axis. Clearly, there has been less volatility in the completed family sizes of actual cohorts than in the period measures based on synthetic cohorts. This relation is also evident in Europe (Bongaarts 2002).

The period TFR is usefully considered to consist of a volume component, measuring the completed family sizes of cohorts then bearing children, and a timing component, indicating when in the course of their lives the cohorts will bear their children. During a period when ages at childbearing are growing older, the period TFR will be systematically lower than the TFR of relevant cohorts because of a "thinning out" of lifetime cohort births. Based upon age-specific rates of childbearing provided by the National Center for Health Statistics, the mean age at childbirth in the United States has risen fairly steadily from 26.00 in 1980 to 27.90 in 2005. Using an adjustment formula developed by Ryder, we find that this delay has reduced period total fertility rates in the United States during this period by about 0.15 children per woman. A more elaborate procedure developed by Bongaarts and Feeney produces a similar reduction averaging 0.14 children per woman over the period 1980 to 1997 (Schoen 2004). Faster delays in Europe have had a slightly bigger impact on period fertility levels there, averaging 0.26 in eighteen countries over the period 1990 to 1997 (Bongaarts 2002). So, the volume components of European and American fertility levels are somewhat more similar than would appear from period TFR measures.

The decline in American fertility is reflected in changes in the distribution of parities (the number of children a woman has borne) among women who have completed childbearing. Figure 1.2 shows that parity two has become the most common destination for women, while parities zero and one have grown steadily in frequency; families of three have become somewhat less common, and families of four or more children have fallen precipitously from being the most common in 1976 (i.e., among mothers of the baby boom) to the least frequent in 2002.

Bearing children is subject to disturbances that can raise or lower the number of births relative to intentions or expectations. Morgan (2003) finds that only 38 percent of women aged twenty-two in the National Longitudinal Study of Youth in 1982 had realized their stated intended parity by age forty. A common form of interference is poor contraception, either through method failure or failure to use any contraception when no conception is wanted. By European standards, Americans have an unusually high incidence of unwanted or mistimed births. Of births during the period 1997 to 2002, 14 percent were retrospectively classified as "unwanted" (i.e., not wanted at any time in the future) at the time of conception by their mother and 21 percent were mistimed (U.S. National Center for Health Statistics 2005b). While a mistimed birth will not necessarily increase a woman's parity above her intentions, an unwanted birth will. Of births to twenty-two to forty-four-year-old women who had not completed high school, 44 percent were classified as unwanted or mistimed, compared to only 15 percent among women who had completed college. The high incidence of unwanted and mistimed births is somewhat surprising in view of the legality of abortion. However, abortion may not be readily available, may be expensive, or may violate personal moral codes. Well-educated women are less likely to have an unwanted or mistimed birth in part because a higher proportion of their unintended conceptions result in an induced abortion.

One factor that can cause fertility to fall short of intentions is subfecundity. Of married women aged fifteen to forty-four in 2002, 7.4 percent were classified as infertile—not practicing contraception and not becoming pregnant for at least one year (U.S. National Center for Health Statistics 2005b). Separation from a partner may also cause women to fall short of childbearing expectations (Quesnel-Vallee and Morgan 2003). The balance of positive and negative forces resulted in slightly fewer births than expected by respondents in the National Longitudinal Survey of Youth; not surprisingly, women who began childbearing late were particularly likely to fall short of targets expressed at an earlier age. Falling somewhat short is the typical, but not universal, cohort pattern (U.S. Census Bureau 2000b; Hagewen and Morgan 2005).

1.3 Women, Men, Partnerships, and Children

By long-standing practices supported by powerful social norms, childbearing and child-rearing in Western countries occurred within marriage. The connection between marriage and childbearing has become more tenuous in the United States:

Of births in 2005, 37 percent were out of wedlock, compared to 5 percent in 1960 and 18 percent in 1980 (U.S. National Center for Health Statistics 2006b; U.S. Bureau of the Census 1979).

Fewer than half of American children aged fifteen live with both natural parents (Kiernan 2004).

Of first births conceived before marriage in 1960 to 1964, 60 percent were "resolved" by marriage, compared to 23 percent in 1990 to 1994 (Ventura and Bachrach 2000).

Two-thirds of adults now disagree with the statement that children are the main rationale for marriage (Thornton and Young-DeMarco 2001).

In short, marriage has become less important as a sanctioning device for childbearing and child-rearing, as well as for sexual expression and cohabitation (Thornton and Young-DeMarco 2001). And marriage itself is changing as husbands and wives are becoming more similar in their household and market activities. Married women are spending less time doing housework while their husbands are spending more time (Bianchi 2000). Of married women aged twenty-five to thirty-four, 68.5 percent participated in the labor force in 2003, compared to 38.8 percent in 1970 (U.S. Census Bureau 2005b). Signaling greater independence of decisions within the family, a married woman's labor force participation has become less responsive to her husband's wage (Blau and Kahn 2005). It has also become less responsive to the presence and ages of her children. The labor force participation rate of women with a child under age one rose from 31 percent in 1976 to 55 percent in 2004 (U.S. Census Bureau 2005a).

It is plausible to argue that the decline in marriage as a social institution and the changes that are occurring within marriage during the last four decades have the same basic sources: greater economic opportunities for women, and vastly improved means of contraception (Chiappori and Oreffice 2008; Lundberg and Pollak 2007; Preston 1987). Both have given women more power in their lives and in their relationships. The advent of the pill and the intrauterine device (IUD) in the early 1960s provided methods that were highly effective in preventing pregnancy, in part because they were independent of any particular act of intercourse and thus required less cooperation from a partner. Marriage became less essential as a precondition for sexual expression. Furthermore, women could invest in their education and in their careers with less threat of disruption from an unwanted pregnancy whether inside a marriage or out (Goldin and Katz 2002). Such investment was also encouraged by the rise in divorce.

If the rise in women's labor force participation had originated exclusively from a supply shift—resulting, for example, from fertility declines induced by contraceptive improvements—it is likely that women's wages would have declined relative to men's. Instead, the median earnings of women working full-time year-round rose from 61 percent of men's in 1960 to 77 percent in 2005 (U.S. Census Bureau 2007b). An important factor in the increase in women's participation and relative wages is probably the rise of service industries in which productivity is not associated with physical strength. Changing norms relating to equity and inequality were probably important as well. The increase in women's labor force participation would not have been as great had they not been able to find acceptable care for their children, and had they not believed that their children were not endangered by such care (Rindfuss, Guzzo, and Morgan 2003).

As Gary Becker (1981) foresaw, the "gains from trade" in the conventional breadwinner/ homemaker marriage eroded as women's opportunities outside the home became more similar to those of men. The reduction in gains was likely abetted by improvements in technology for performing standard household tasks (Greenwood, Seshadri, and Vandenbroucke 2005; see also Isen and Stevenson, chapter three, this volume). What was less foreseeable was that fertility would level off and even rise modestly as the institution of marriage was fundamentally changing. Had bearing children not been a powerful goal of most American women, they would have found ample reason to avoid them by virtue of their increasingly tentative relationships and the growing attractions of work outside the home. Instead, they took advantage of their new powers to maintain a fertility level that is the envy of most other developed countries.

1.4 Individual-Level Characteristics Associated with Fertility in the United States

In this section, we examine fertility variation according to major personal characteristics in order to seek some guidance about future fertility levels. We focus on two variables whose distributions are expected to change in predictable ways and, therefore, might shed light on the future of fertility.

(Continues...)

---

Excerpted from Demography and the Economy Copyright © 2011 by National Bureau of Economic Research. Excerpted by permission of The University of Chicago Press. All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.

---

Table of Contents

Acknowledgments


Introduction
John B. Shoven

1. The Future of American Fertility
Samuel H. Preston and Caroline Sten Hartnett
Comment: Gopi Shāh Goda

2. Fertility Theories: Can They Explain the Negative Fertility- Income Relationship?
Larry E. Jones, Alice Schoonbroodt, and Michèle Tertilt
Comment: Amalia R. Miller

3. Women’s Education and Family Behavior: Trends in Marriage, Divorce, and Fertility
Adam Isen and Betsey Stevenson
Comment: Enrico Moretti

4. Adjusting Government Policies for Age Inflation
John B. Shoven and Gopi Shāh Goda
Comment: Warren C. Sanderson

5. Old Europe Ages: Reforms and Reform Backlashes
Axel Börsch- Supan and Alexander Ludwig
Comment: Alan J. Auerbach

6. The Final Inequality: Variance in Age at Death
Shripad Tuljapurkar
Comment: Victor R. Fuchs

7. Demographic Trends, Housing Equity, and the Financial Security of Future Retirees
James M. Poterba, Steven F. Venti, and David A. Wise
Comment: Thomas Davidoff

8. Aging Populations, Pension Operations, Potential Economic Disappointment, and Its Allocation
Sylvester J. Schieber
Comment: Steven F. Venti

9. Financing Medicare: A General Equilibrium Analysis
Orazio Attanasio, Sagiri Kitao, and Giovanni L. Violante
Comment: Moshe Buchinsky

10. Italians Are Late: Does It Matter?
Francesco C. Billari and Guido Tabellini
Comment: Luigi Pistaferri

Contributors
Author Index
Subject Index