# How to Write About the Birth Rate

by Carl Haub, senior demographer, Population Reference Bureau

No demographic subject captures writers’ imaginations like a country’s birth rate, be it baby “booms” or “busts,” or record highs or lows. But what measure should you use when you’re writing about the birth rate? Yes, there’s more than one—there are three: the crude birth rate, the general fertility rate, and the total fertility rate. In this blog post, I want to clear up the confusion.

First, we’ll take the crude birth rate (CBR), which is simply the number of births in a year per 1,000 population. Rates have a numerator (in this case, births) and a denominator (the country’s total population). The fact that the denominator is the total population of all ages is the reason why the CBR is labeled “crude.” In 2011, the CBR in the United States was 12.7 births per 1,000 population. But the CBR can be significantly affected by age structure, so that a population with a high proportion of elderly will tend to have a lower rate than one with a younger population. Why bother to report it? Because the CBR is one of three essential parts of the national population growth rate, since populations grow or decline based on the number of births, deaths, and net immigration—the balance of people moving in and out.

Second, we have the general fertility rate (GFR), which is similar to the CBR, but the denominator includes only females of childbearing age, usually ages 15 to 44 or 49. It is certainly true that the GFR is more “refined” than the CBR and its denominator more appropriate. For year-to-year changes in the birth rate, it is a better indicator since the childbearing-age group changes little but, over the long term, it can be affected by such changes.

But, don’t we really want to know how many children women will end up with when they finish their childbearing years? Annual birth rates can be distorted because women delay having children—so it looks as if they will have fewer children than they ultimately do. No rate for a single year can measure exactly that.

The total fertility rate (TFR) tells us what we are looking for, the one that measures annual trends with no effects of age structure as mentioned above. The TFR shows how many children a woman would have in her lifetime if the pace of childbearing of a given year remained unchanged. It is derived by simply adding up age-specific birth rates, the total being that average number of children. To explain, we take the sum of the rates for women ages 15, 16, 17…up to 44 or 49 to get a precise snapshot of the TFR.

Several recent articles have reported that, in 2011, the U.S. birth rate reached its lowest point in history. But that was in terms of the CBR and GFR. For example, in 2011, the U.S. TFR was reported by the National Center for Health Statistics to be 1.89 children per woman and the GFR, 63.2. But 1976 had, and still has, the lowest TFR ever reported, 1.74. However, the 1976 GFR was 65.0, higher than in 2011. Why? Because the female childbearing population was “younger,” with a higher proportion in their 20s. But they were having children at a slower rate than those in 2011. Not convinced? Consider this: If women in 1976 had the same TFR as women did in 2011, they would have borne 3,265,869 children. But they didn’t. They had 3,167,788. So, when you are trying to find the apples-to apples way to look at a country’s trend in the birth rate, there really is only one choice, the TFR.

## 11 thoughts on “How to Write About the Birth Rate”

2. “But, don’t we really want to know how many children women will end up with when they finish their childbearing years? Annual birth rates can be distorted because women delay having children—so it looks as if they will have fewer children than they ultimately do. No rate for a single year can measure exactly that.”

As you know, which measure to prefer depends upon our objective. The TFR may be what we “really want to know” for some purposes, but the GFR will be for others. If women delay births, they likely will have fewer children ultimately, since fecundity declines with age and other events (e.g., illness or injury) can disrupt plans for later childbearing. (“A bird in the hand …”)

I find it useful to regard the GFR as an average of age-specific fertility rates (ASFR) weighted by the age distribution of women of reproductive age, and the TFR as an unweighted average of AFSR multiplied by the number of years in the reproductive period. That makes clear that both measures are averages of the ASFR. Dividing the TFR (or multiplying the GFR) by 30 or 35 (for ages 15-44 or 15-49) puts the two measures on the same scale for convenient comparison.

• Carl Haub

Dear Dr. Schoenbach,

I read your comments over carefully and I didn’t quite follow the last one. The GFR uses the entire population of women 15-49 (or 15-44) as a denominator so it could not be weighted and does not use ASFRs at all. The TFR does use the sum of the ASFRs and is “blind” to age distribution, so it, too, is not really a weighted average either. Perhaps I missed something!

Best regards,
Carl Haub

• Dear Dr. Haub,

I’ve been preoccupied with weighted averages since I first learned about age standardization in 1978, and I have been known to see them even where they don’t exist 🙂 However, I do find the concept of averages helpful for understanding both GFR and TFR.

Suppose that the age-specific fertility rates in a given year are r1, r2, …, r7 for ages 15-19, 20-24, …, 45-49 years. The TFR is computed by summing the rates (r1+r2+…r7) and multiplying by 5, the length of each age interval. Suppose instead we take the average of the rates and multiply by 35, which is the length of the reproductive period of 15-49 years, inclusive). The average of the rates is (r1+r2+…+r7)/7, and multiplying the result by 35 gives us the same expression as we got for the TFR.

The GFR is the total number of births in a given year divided by the total number of women age 15-49 years (or 15-44 years). If we ignore births to women younger than 15 or older than 49, the total number of births is the sum of the births to women of each age or age interval. The number of births to women of a given age or age interval is the rate for that age multiplied by the corresponding number of women. Letting w1, w2, …, w7 be the number of women (or women-years) in each age interval, the total number of births is then:
Total births = w1(r1) + w2(r2) + … + w7(r7)
The GFR is that result divided by the total number of women (W=w1+w2+…+w7). Since the w’s sum to W, the resulting expression is a linear combination of age-specific fertility rates weighted by proportions that sum to 1.0.

I hope that the above is correct. Most of my knowledge of demography comes from reading PRB publications (acknowledged in my demography lecture for my course), so I hope I haven’t let PRB down!

It’s a privilege to have this exchange with you.

Best regards,
Vic
go.unc.edu/vjs/

3. While the TFR is probably better than the GFR, ultimately the cohort fertility rate is the most accurate. The problem with that rate is you have to wait for women to complete their childbearing. That takes too long so we accept TFRs, which are often predict more or less children than end up being born:

Further discussion of this is here: