Panel on Fertility Trends at NBER Summer Institute
The 2019 NBER Summer Institute’s Social Security meeting featured a panel discussion on the economic determinants of fertility decisions. In introducing the panel, NBER President James Poterba called attention to the very recent decline in fertility rates in the U.S., as well as the longer-term downward trend, and noted that fertility is “an incredibly important driver of the long-run financial health of Social Security and other related programs,” making it central to any discussion of these issues.
Karen Glynn, Deputy Chief Actuary of the Social Security Administration (SSA), led off the panel with a discussion of assumptions used by SSA, including in the annual OASDI Trustees Report. She noted that demographic assumptions shape the age distribution for the future, and that birth rates are the most critical factor. There have been significant changes over time in age-specific birth rates – rates have risen for women ages 30 to 34, 35 to 39, and 40 to 44 over the past four decades, while declining for women under age 30, particularly since the Great Recession. Cohort total fertility – the total number of births per woman throughout her lifetime – remained near 2.0 for women born in the 1950s and 1960s. For the 2019 Trustees Report, the Office of the Chief Actuary projected a future cohort total fertility rate of 2.0 for its intermediate assumption (best estimate).
Martha Bailey (University of Michigan and NBER) provided a longer historical perspective in her remarks. She noted that completed childbearing has been remarkably stable across birth cohorts, with strong adherence to a “two-child norm” – about one-third of women born in each cohort between 1940 and 1980 had two children, a much higher share than had zero, one, or three children (about 15-20 percent in each case). Among these women, falling birth rates at younger ages appear to have reflected delay, not foregone childbearing. These population values mask growing differences by education. Bailey found that the mean age at first birth has remained roughly constant across cohorts for women in the lowest quartile of educational attainment, at around age 22, while rising from about age 24.5 to age 27.5 between the 1940 to 1960 birth cohorts for women in the highest quartile.
Kasey Buckles (University of Notre Dame and NBER) focused on the post-2007 fertility decline and its causes and implications. She noted that the total fertility rate, which equals the sum of age-specific rates, has declined by 18 percent since 2007. Fertility declines have occurred exclusively among younger women, with birth rates dropping by 58 percent for teenagers, 36 percent for women ages 20 to 24, and 20 percent for women ages 25 to 29. Beyond the recession, Buckles pointed to significant changes in the economic well-being of low-income women and increased access to contraception as potential explanations for the decline in young women’s fertility. In terms of the effect on lifetime fertility, Buckles pointed out that women born in the mid-to-late 1980s would need to experience near-Baby Boom fertility rates to reach a cohort total fertility rate of 2.0.
Caitlin Myers (Middlebury College) discussed the “mechanical causes” of the fertility decline, including changes in sexual activity, contraception use and effectiveness, and abortion access. She noted that the rising use of long-acting, reversible contraception (LARC) use does not necessarily reduce pregnancy risk – for example, older women are substituting LARC for surgical sterilization. However, increasing contraception use does seem to be a contributor to declining teen birth rates. Myers reported that increased access to abortion after its legalization explains most of the decline in births before age 19 between the 1940 and 1958 birth cohorts. Based on state policy environments and credible estimates of the causal effects of travel distance on abortion use, Myers estimated that a reversal of the Roe v. Wade decision could increase the total fertility rate by up to 0.05.
Steve Goss, Chief Actuary of the Social Security Administration, ended the panel with a discussion of the effect of employment and earnings on births to women ages 20 to 34. He remarked that three out of four births occur to women in this age group, and that steady birth rates among women ages 30 to 34 since the Great Recession have helped to offset fertility declines in the younger groups. He noted that the employment to population ratio has been rising since 2011 and earnings have also begun to rise recently for young women and men. The lack of a reversal in birth rate declines among those under 30 and the continued level (not rising) birth rates at ages 30 to 34 thus suggest that there may be a lag between a change in economic conditions and a change in birth rates, potentially reflecting the need to develop sufficient confidence in the persistence of improved economic conditions.
Video and slides from this panel are available at: