The black teen pregnancy rate dropped nearly 50% in 5 years.

In 2010 it was just over 15% of all (non-Hispanic) black births. By 2015, the most recent complete data available, it was 8.6%.

There are several reasons this could be the case. It could be people leaving high-crime urban areas, which are where the bulk of black teen pregnancy is concentrated. It could be part of the increase in older, more educated foreign-born and second-generation non-Hispanic black immigrant mothers making up a bigger share of black births. It could be both, or some other factors. It’s not an increase in abortions, though.

For perspective, here’s a chart with rough estimates since 1970 and percent changes.

% teen pregnancy % change
1970 31 N/A
1975 33 2
1980 26.5 -6.5
1985 23 -3
1990 23 0
1995 23 0
2000 21.5 -1.5
2005 17 -4.5
2010 15.2 -1.8
2015 8.6 -6.6

The 1970s were pretty bad, as were the 1980s, but the last decade has seen a significant and major turnaround in teen pregnancy.  Teen pregnancy for black Americans specifically is particularly “bad” because it’s skewed unmarried for much longer than was the case with white teen pregnancies, starting with under 15 pregnancies being mostly unmarried ones in the 1950s when that wasn’t the case with white pregnancies under 15.  In 1966, not long after the famous Moynihan Report, black teen pregnancies to mothers aged 15-19 crossed 50% unmarried.

Through the 1960s and 1970s, teen pregnancy became completely decoupled from marriage for black teen mothers and heavily decoupled for black mothers in their early 20s, culminating in 1978.  Five years after Roe V. Wade, black births to mothers in their early 20s were now majority unwed and teen births were 80-97% unwed, depending on age.  However, teen births were only about 25% of total births, during the 1960s.  The table shows how things shifted after the collapse in married black motherhood during the 1970s and 1980s.

All birth data used for rough percentages comes from Monthly Vital Statistics Reports or National Vital Statistics Reports.

Itty bitty teeny weeny post about Scandinavian birth patterns.

They are, for the most part, very similar to American ones, but in a way I didn’t expect once I looked into the data.  They have nearly the same percentages of women having 3 or 4 kids as America does (so, about 25% or so combined).  They have a similar pattern of fewer women signing onto the motherhood project, but the ones that remain having 2-3 and a bit less often than in America 4 or more.  It’s not a sea of women having just one and grudgingly two at all.

I have run into a lot of references to having three kids in English-language articles about various Scandinavian countries and it turns out that is partly because a three child family is not actually that uncommon in those countries.

This is interesting.  I tried to see if this was true in non-Scandi Europe (France, Germany, UK), but the data wasn’t laid out for English speakers in a way that made this easy to find, so  I still have no idea if it’s true with them too.  It’s also pretty SWEET that Scandinavian countries put up some pretty elaborate birth data charts ‘n’ graphs in English.

College education and birth numbers, 2007-2015

The CDC has a birth database,, and it is pretty complete.  It’s where the numbers come from.

What I was talking about here  is a little more obvious once the percentages are there.  This is just for black and white non-Hispanic women, the groups you can go back to 1960 or earlier on.

Black women: 53% of births to women with high school completed or less, 47% of births to women with some college education, including completing bachelor’s and advanced degrees.  Nearly 4 million births for that timeframe.

White women: 32% of births to women with high school completed or less, 68% to women with college education.    Nearly 15 million births for them.

These numbers represent all births for this timeframe, first kid, second, fifth, whatever.

Now we look at the impact of education level on marital status for black women:


Race Education Level Percent Married at Birth Percent Unmarried at Birth
Black 8th grade or less 38.49% 61.51%
9-12 grade 9.63% 90.37%
Finished High School 18.30% 81.70%
Some college 27.68% 72.32%
AA degree 44.23% 55.77%
BA/BS 62.28% 37.72%
MA 76.47% 23.53%
PhD or Professional Degree (MD, JD, etc.) 86.00% 14.00%

Though not ideal, the impact of more education is evident, women with even a little college education represent a substantially lower number of out of wedlock births.

Here’s the impact for white women:

Race Education Level Percent Married at Birth Percent Unmarried at Birth
White 8th grade or less 67.41% 32.59%
9-12 grade 30.05% 69.95%
Finished High School 50.39% 49.61%
Some college 63.35% 36.65%
AA degree 79.10% 20.90%
BA/BS 92.69% 7.31%
MA 95.74% 4.26%
PhD or Professional Degree (MD, JD, etc.) 96.06% 3.94%

Amish births distort their 8th grade numbers in favor of marriage, but given the radically different structure of Amish society,  one could argue that this is a more education= more marriage situation too.  It is worth noting that the out of wedlock numbers for white women with BAs or more have remained in the single digits for decades now, despite the group itself being much larger a chunk of the mothers.

Having looked at basic out of wedlock differences by education we come to the question the natalist of any sort is most interested in: what percentage of BIRTHS do the higher-educated moms represent?

Let’s look at the numbers for black women first:

Race Education Level Percent of total married births Percent of total unmarried births
Black 8th grade or less 2.35% 1.48%
9-12 grade 6.04% 22.28%
Finished High School 21.68% 37.99%
Some college 26.26% 26.94%
AA degree 10.04% 4.97%
BA/BS 20.67% 4.92%
MA 10.38% 1.25%
PhD or Professional Degree (MD, JD, etc.) 2.57% 0.16%

And the kicker, percentage of total births:

Race Education Level Percent of all births
Black 8th grade or less 1.72%
9-12 grade 17.70%
Finished High School 33.39%
Some college 26.75%
AA degree 6.40%
BA/BS 9.36%
MA 3.83%
PhD or Professional Degree (MD, JD, etc.) 0.84%

As you can see, the ground gets pretty thin for black women with more education, but they still represent about 20% of total births at the AA+ level, and at that level a comfortable majority of births are in wedlock.  This means most of the 30% or so of black births born in wedlock these days are heavily drawn from that 20% group, despite their smaller separate percentages.  It also means that even including the “some college” group’s high OOW rate, black women have a majority of births in wedlock being born to college educated mothers.

But here’s something interesting with the data for white women:

Race Education Level Percent of total married births Percent of total unmarried births
White 8th grade or less 1.33% 1.54%
9-12 grade 3.41% 18.99%
Finished High School 16.11% 37.99%
Some college 19.53% 27.07%
AA degree 10.42% 6.60%
BA/BS 31.53% 5.96%
MA 13.94% 1.49%
PhD or Professional Degree (MD, JD, etc.) 3.74% 0.37%

If you look at unmarried birth percentages, you see that they are nearly identical in distribution to the black percentages.  They are in fact identical at the “Finished High School” level, with that being the largest chunk of unwed mothers for black women and white women.  What we also see is that for white women, there are MORE college educated mothers having OOW births as a percentage, while they have a smaller group of births in the “Grade 9-12 only” category.

That was surprising.

Here’s the total birth percentages for white women by education level:

Race Education Level Percent of all births
White 8th grade or less 1.39%
9-12 grade 8.00%
Finished High School 22.55%
Some college 21.75%
AA degree 9.30%
BA/BS 24.00%
MA 10.28%
PhD or Professional Degree (MD, JD, etc.) 2.75%


Just like with black women, you get a shrinking of percent from some college to AA and then a strong bump for the BA, but it’s much more significant here.  So even if you don’t care about test scores and all that jazz, the basic cognitive sort is blindingly obvious in both black and white women.  Black women who marry before kids are heavily college attending, and white women in general, but ESPECIALLY those who marry are heavily college attending.

Anyway this post is just putting the numbers up.  These births, as noted in the title, are recent and span nearly a decade and are all births, not just first births.  So these children are almost entirely under 10 years old and this represents some completed families and siblings.  This has interesting implications for the future.

The Marginal Child in 2014

These are heat maps of where people decide to have the marginal third child that breaks the “family of four” paradigm that is reflected even in consumer goods and packaging because it’s become such a core part of post-Vietnam American culture.

For all races, about 30% of births for 2014 were third kid or higher.

Third births and higher, all races

For whites, it was about 25%

Third births and higher, whites only

A starting point for discussion is that while the coasts with good jobs where both parents can potentially earn 75-100k apiece are punching a little below the national average, they are nevertheless putting up third babies in the double digits in many high-cost counties.

Where the babies are, 2014

Heat map of where 4th and higher births are by county for all races.  National average is 12.4% of all births.

Fourth and higher births

Here’s just non-Hispanic whites.  Their national average is 10.2%.

Fourth and higher order births, whites only

The hottest counties have 24-27% of births (all races) and 31-35% of births (white only) as kid #4 or more.

Open for discussion.

American childbearing goes Scandinavian

The latest birth and fertility data are showing what I feared, the only age groups showing an increase in births were the 30-44 age groups.  All groups under age 30 have been showing consistent and steady declines of late, with the 25-29 group finally succumbing after occasional blips upwards half a decade ago.

Lower under-18 births is ok, and that continues (and no it is not via more abortions, but less sex).  But the shift towards having kids in your 30s as the norm continues apace.  This is the modern Scandinavian model.

The plus side is that there’s more married childbearing in all racial groups.  The unwed motherhood trend is going ever so slowly backwards, consistently.  Black births to unwed mothers dropped while the total number of births increased, lowering their percentage of unwed births.

There’s a shift in the last decade where unwed mothers are finally seeing percentage and absolute decreases in their births, but married mothers are having more children.

Fewer women are having children, but the ones who get married first are having more third and up children than they used to five years ago.  This is tricky though, because with women having all their kids in their 30s and 40s more, we can’t be sure we haven’t moved into a different fertility pattern of relative robustness for the women who can navigate waiting that long to marry and pop out 3 or more kids.  It could be the latest dead cat bounce before we get yet another step down in higher order births.

Nearly 10% decrease in number of third and higher births from 1992 to 2014

For white non-Hispanics, a little over 600k births in 1992 were the third live baby or higher.  This represented just under 1/4 of all births.  Over 75% of all births for 1992 were first or second births.

The approximate distribution of the 600k higher birth orders (less than 100% due to rounding) was:

3rd order: 66%

4th order: 22%

5th order: 7%

6th order: 2.6%

7th order: 1%

8th order or higher: 1.1%

In 2014, there were over 50k fewer such births, a bit over 550k and that represented just OVER 25% of all births for that year.

The distribution of these 550k higher-order births over 20 years later was:

3rd: 61%

4th: 23%

5th: 8.4%

6th: 3.4%

7th: 1.6%

8th or higher: 2%

The total births for 1992 were around 2.5 million, while for 2014 they were around 2.1 million.  So people were having fewer children overall, but the ones having many are chugging along pretty impressively.  The problem is that there’s no filtering for how much of that chugging along is in little horse-powered buggies, so there’s that to keep in mind.

To put this distribution of higher-order births in context, here’s the “white” distribution for 1970, ten years after the Pill and IUDs were introduced.

Total white births: approx 3.1 million

Total third or higher order white births: approx 1 million

Percentage distribution of third and higher order births:

3rd: 48%

4th: 25%

5th: 12.5%

6th: 6.4%

7th: 3.4%

8th: 4.5%

Admittedly this includes some Hispanics, but only about 4-5%, not enough to shift the overall pattern.  This pattern from 1970 could be returning at the higher orders, but it’s too soon to tell.