Aging America
A Brief Spin Through the New Age Estimates
Census released new age and year-specific estimates of the national population today. So let me walk you through them.
Census made minute revisions to the historic age composition of the population. Even in 2016, the year with the largest revisions, the average age of the American resident population was only revised by about 12 hours. That is to say, the average age was revised down by 0.001357 years.
The time trend for the new revisions shows that we continue to age as a nation, however.
Each year, the average age rises by about 65 to 70 days. So we are getting older.
No surprise there!
Let’s pivot to sex. Did Census revise age-and-sex-specific population ratios?
For most age groups, Census made revisions resulting in relative more men. The magnitudes are not extremely large: we’re talking about a few dozen more men per 100,000 women. And women are still a larger share of the overall population. But, on the margin, Census revisions were male-biased for the whole population, though I do not know exactly why.
What about age generally? Did Census appreciably change their estimate of individual ages, even if it didn’t change average age?
There are changes, particularly increases in estimates of the 25–45 population. But overall, the changes are small:, topping out at around a quarter of one percent.
So the revisions were, on the whole, not that exciting.
Let’s switch from looking at revisions to look at some salient figures from the data.
How Old Is America, Really?
We can straightforwardly compare US population by age in the 2010 Census versus the current round of estimates by just lining up each category and connecting the dots.
The blue line is the Census, the red line is the summer of 2017 estimate. As you can see, today we have a lot more people over age 55 than we had in 2010. We also have more people aged 22–37 than we used to have: that’s the echo-boomer, Millennial generation. Meanwhile, we have fewer people between the ages of 38 and 54 than we had in 2010, and also fewer people ages 14–21.
Mostly you can see that this is just a product of population aging.
But what if we undo that aging and look at the population in the country by specific birth cohorts in 2010 and 2017? Any difference would have to be due to migration or mortality! To ensure I’m not picking up methodology differences, I’ll use the population estimate for 2010, not the Census number.
At the left, with early birth cohorts, you see the effects of mortality as people age. The 2017 estimate is far below the 2010 estimate. At the extreme, 83% of all people who were born in 1918 and alive in 2010 were dead by 2017.
At the younger end of the cohort, we see a different trend: the number of people in a given birth cohort rose! That doesn’t make any sense; you can’t just add to a given birth cohort!
Well, actually, you can! Through immigration! So then the gap between the two lines shows us the cumulative effect of age-specific mortality and migration from 2010 to 2017. Let’s look at that more closely!
Migration has offset mortality for cohorts born 1972 or later. For cohorts born before 1972, the reverse is true. That’s basically as you would expect.
But let’s take a longer timespan. Let’s compare each cohort in 2017 to how many people were born in the US that year, to see cumulative survive complete-life-cycle-to-date impacts of mortality and migration! This is the kind of spicy content you can only get here, folks!
Look at that! First of all, let’s talk about the birth trend. Births rose from 1935, the earliest complete national data to the 1950s. Surviving population also rises over that timespan, partly due to the mortality trend (earlier birth years have a higher rate of attrition due to mortality), but also simply because more people were born. You can see the big spike in 2017-surviving-population for the 1946–1947 birth spike. We are somehow missing the 1940–1943 pre-Boom-Boomers from modern population, maybe they had higher mortality, although the population graph does flatten out a bit for their years.
Then the birth line plummets as fertility fell nationally. The surviving age cohorts have a tight trend with those birth totals. You can see that the trough in US population today around middle-age is a direct result of low births in the 1970s. But, with time, the generation turns, fertility perked back up a bit, and births rose. The very large 1990–1991 birth cohorts (that’s me!) are associated with a similarly large number of 26–27 year olds in the vintage 2017 estimates.
But there is a positive gap for birth cohorts 1958–2005. That’s due to immigrants who were born in those years. This effect is large: for the 1972–1975 birth cohort, total population is about 25% higher than births were in those years. Since some people born 1972–1975 have died, this means that the net effect of migration has been to add more than one immigrant of those age cohorts to the US for every 4 native-born people in the group.
But How Many Immigrants Per Age Group Are There?
Here I’ll pivot to American Community Survey data and Census samples. Here’s the foreign-born share of the population for 2012–2016 ACS, 2000 Census, and 1950 Census.
Let’s compare 2000 to 2012–2016 first. The foreign-born share has risen dramatically for people over the age of 27 or so, especially people around age 40. The gap you can see there is genuinely striking.
But I want to point out a less-striking but perhaps equally-important gap: the foreign-born share of the population under age 27 or so has fallen. Foreigners make up a smaller share of young people today than they did in 2012–2016. That surprised me when I saw it. This is probably due to much lower illegal immigration, and the general shift towards more Asian immigration. Both of these bias older.
Of course, compared to 1950, the foreign born population share has risen for every age from 0 to mid-50s. Because as we know, in golden-age 1950s America…. foreign-born people made up about a quarter of all retirement-age people.
Again, that surprised me. It makes sense, of course: those people came as young immigrants in the period before the 1920s restrictions. But still, I had not seen the numbers myself before making it that clear that foreigners had an extremely high share of retirees.
So let’s look at this another way. Let’s look at the foreign-born-share across time of three age groups: 0 to 25, 26 to 60, and 61 and older. I’ll also do a narrower 30–50 breakout to catch peak working years.
The graph above shows something interesting. On the one hand, the black line shows headline foreign-born population share. In 2016, it is near its highest levels ever, and rising.
But oddly enough, when you look at specific age groups, the story changes. Yes, the foreign-born share of ages 26–60, 30–50, and 60+ are rising. But none are anywhere close to their historic peaks. And when we look at people 25 and younger, not only is the foreign-born share not at a peak, it seems to be falling! So while the national foreign-born share is quite high in aggregate, subgroup shares are somewhat less alarming.
That’s all I’ve got for today. Enjoy Census Data Release Day!
I’m an Advisor at Demographic Intelligence, the nation’s leading producer of rigorous national- and regional birth and marriage forecasts. I’m also a Research Fellow at the Institute for Family Studies, a Senior Contributor at The Federalist, and I write periodically for Vox’s Big Idea column. I’m a native of Wilmore, Kentucky, a graduate of Transylvania University, and also the George Washington University’s Elliott School. My real job is as an economist at USDA’s Foreign Agricultural Service, where I analyze and forecast cotton market conditions. I’m married to a kickass Kentucky woman named Ruth. I am not paid one penny by anybody for this blog post.
Follow me on Twitter to keep up with what I’m writing and reading. Follow my Medium Collection at In a State of Migration if you want updates when I write new posts. And if you’re writing about migration too, feel free to submit a post to the collection!
Check out my Podcast about the history of American migration.
DISCLAIMER: My posts are not endorsed by and do not in any way represent the opinions of the United States government or any branch, department, agency, or division of it. My writing represents exclusively my own opinions. I did not receive any financial support or remuneration from any party for this research.
