More Thoughts on the Illusion of Declining Birth Rates

My earlier post on birth rates generated some interest and discussion and therefore deserves a bit more attention and clarification.  First, the average birth rate referenced in news headlines is simply the number of births divided by the number of women age 15-44.  This average birth rate conflates age-specific changes in birth rates with changes in the age distribution of the population.  The average birth rate is lower in 2011 in part because there are so many women age 15-19 and age 40-44, when birth rates are low.

Demographers and labor economists use the total fertility rate to study changes in birth rates over time.  The total fertility rate gives equal weight to women of each age from 15 to 44 and therefore is invariant to changes in the age distribution of the population.  My second point is that the total fertility rate in 2011 is higher than it was from 1975 to 1985 and therefore is not the “lowest ever.”  Finally, the total fertility rate understates the expected number of children born to women from a given birth cohort when women are choosing to delay childbirth.  Delayed childbirth will show up as an initial decline in the total fertility rate but not the number of children per woman.

By following a given birth cohort of women from the ages of 15 to 44 I calculate that:

  • Women born between 1956 and 1960 had an average of 2.03 children
  • Women born between 1961 and 1965 had an average of 2.06 children
  • Women born between 1966 and 1970 had an average of 2.12 children
  • Women born between 1971 and 1975 had an average of 2.11 children

The youngest of the women born between 1971 and 1975 are age 37 and are likely to have more children.  Thus 2.11 is an underestimate of the average number of children born per woman for the 1971-1975 birth cohort.  There is no evidence of declining birth rates among women who are at or near the end of their childbearing years.  For women born more recently (1976-1980 or 1981-1985) there are still many more years of possible childbearing.

More recent cohorts of women are having fewer and fewer children before the age of 25.  They are also likely to have more children in their 30’s than their older sisters, mothers and grandmothers.  It is far too early to predict whether lower birth rates for women under the age of 25 in 2011 will translate to lower birth rates over their lifetimes.  What does seem clear is that we are transitioning to a society where the 30’s become the prime childbearing ages for most women.  It is already true that there are more births (per capita) to women age 30-34 than age 20-24.

I do not believe that the delay in family formation we have seen thus far is due to decadence.  Couples are likely delaying children because of the weak economy and the high costs of raising children. Columnists and commentators should write and talk about the dramatic changes that will occur in our society when the majority of first-time parents are in their thirties.  I am skeptical, however, that one of the changes will be a decline in the U.S. population.

Tyler Cowen and Ross Douthat Have Confused Delayed Childbearing with Declining Fertility

Ross Douthat’s recent New York Times column is based on the premise that “American fertility plunged with the stock market in 2008, and it hasn’t recovered.”  Douthat cited a Pew Research Center report that “the U.S. birth rate dipped in 2011 to the lowest ever recorded” because it fell to 63.2 per 1,000 women age 15 to 44 compared to 71 per 1,000 women in 1990.

Unfortunately, because of changes in the age distribution the average birth rate across all age groups from 15 to 44 is a poor way to compare birth rates over time; fertility rates vary among women of different ages.  For example, women age 15-19 and 40-44 have the lowest fertility rates.  In 2011 there are relatively more women in these age groups than in earlier years: 13% more than in 1990.  Changes in the age distribution distort comparisons of average birth rates between 2011 and earlier years. 

Demographers tend to use the “total fertility rate” to describe trends in birth rates over time.  The total fertility rate sums fertility rates across all age groups and therefore puts an equal weight on all ages from 15 to 44, regardless of changes in the age distribution.  The total fertility rate in 2011 is not the lowest rate ever recorded and is, in fact, higher than it was for the entire decade from 1975 to 1985.  This is why Carl Haub of the Population Reference Bureau wrote that reports claiming that the average birth rate in 2011 was the lowest ever are misleading.

Some demographers use the total fertility rate as a proxy for the expected number of children born to women over their lifetimes.  This approximation will only be accurate if women age 15-19 in 2011 expect to have the same fertility rate 15 to 20 years from now as women in their thirties have in 2011.  The approximation will be inaccurate if women in 2011 are delaying childbirth longer than women from earlier birth cohorts.

In recent years the biggest declines in fertility occurred for women under the age of 25.  In the past 20 years births to women age 15-19 are down 49% and births to women age 20-24 are down 26%.  At the same time fertility rates for women in their thirties and forties are higher than ever.  Since 2003 the fertility rate for women age 35-39 has been higher than for women age 15-19 and the gap is widening each year.  Similarly, beginning in 2009 the fertility rate for women age 30-34 increased above the rate for women age 20-24.  There is every reason to expect this gap to increase as women outnumber men in colleges and universities and account for 47% of the labor force.

Although it is likely that couples have delayed childbearing because of the economic downturn, there is no reason to assume that this will lead to a permanent decline in the number of children ever born to women over their lifetimes.  The total fertility rate has consistently underestimated the number of children ever born at times when young women are delaying childbearing longer than women from earlier birth cohorts.

The following charts illustrate actual birth rates for women born from the 1956-1960, 1961-1965 and 1966-1970 birth cohorts and the projected birth rates implicit in total fertility rate calculations made in 1975, 1980 and 1985 respectively (when women from these birth cohorts were 15-19).  For each cohort actual birth rates exceeded the projected rates based on older women in the corresponding cross-sections because younger women were delaying childbirth longer than women from earlier cohorts.  Birth rate projections from cross-section extrapolations are 10.9% to 13.3% lower than the actual number of children ever born to women over their lifetimes.  Birth_Rates1



Although women age 15-19 and 20-24 today may have lower fertility rates than their older sisters and mothers had at the same age, they are also likely to give birth to more children in their thirties and forties than their older sisters and mothers did.

Nonetheless, Tyler Cowen writes that he basically agrees with Douthat’s thesis that:

“The retreat from child rearing is, at some level, a symptom of late modern exhaustion – a decadence that first arose in the West but now haunts rich societies around the globe.  It’s a spirit that privileges the present over the future, chooses stagnation over innovation, prefers what already exists over what might be.”

An alternative view is that couples delay child rearing because they recognize that education and human capital investments are expensive and the economic downturn and weak recovery forestalled the important decision to start a family.  I believe the decision of a couple to wait until they can better afford the education, housing and health care costs for their children is a noble sacrifice rather than an expression of decadence or a choice of stagnation over what might be.

%d bloggers like this: