How good are generational boundaries at actually capturing generations?

I wrote a book about the baby boom generation, leading to a surfeit of questions about that generation and about generations in particular.

Fairly early in such conversations, I mention that the baby boom is the only generation considered sufficiently demographically identifiable that the Census Bureau recognizes it as an actual generation. Most of the uncertainty people can have about their own generational identities stems from the vague, subjective way in which the boundaries of other generations are drawn.

Why is the baby boom different? At the risk of stating the obvious, it’s because there were a lot of babies born. A number of babies greater than half of the country’s population in 1945 was born over the next 19 years, the proportionate equivalent to the United States seeing 166 million babies born between now and 2041. It wasn’t just love-starved G.I.s.

When making this point, I often found myself repeating a fact about the duration of the baby boom: Some members of the baby boom were born to other members of the baby boom. This was necessarily true, given that women gave birth at younger ages 60 years ago and that the boom lasted 19 years. But I didn’t know how true it was. How many boomers were born to boomers?

That leads to another interesting question: How good are generational boundaries at actually capturing generational divides. The idea of a “generation,” of course, is that it demarcates the difference between a parent and a child. Do our generational identifiers do that? Is a member of Gen X the child of a boomer (in cases besides myself)?

It occurred to me that there is a way to evaluate this. The Centers for Disease Control and Prevention, tasked with collecting information on U.S. births and deaths, publishes data on the ages of mothers when babies are born. Overlapping those ages with generational boundaries should allow us to see the generations of mothers and children — and to answer our questions.

Some caveats. First, since generational boundaries are subjective, we need to choose some lines. I defer to Pew Research Center’s lines, since they are both thoughtful and generally agreed-upon elsewhere. Second, the annual data on births is bucketed into age ranges for mothers, meaning that we know how many babies were born (for example) to women ages 20 to 24 in a year. In circumstances like that, I simply divided the total for the age range among the years. This introduces a potential inaccuracy! But presumably not a terribly large one and, since age clusters are often in the same generation anyway, sometimes there is no inaccuracy at all.

So, without further ado, here’s the breakdown from 1933 (the earliest year for which I found easy-to-parse data) to 2021. That covers much of the silent generation (preceding the boom) and some of the post-Gen Z-aka-lockdown-generation group of babies, often called “alpha.” The color-coding is consistent: Blue represents the range of silent babies at left, and it represents the percentage of babies born to silent generation mothers in the various rows.

I also highlighted (using black outlines) those occasions in a year where a percentage of babies was born to members of that same generation. In 1964, the last year of the baby boom, about 1 in 9 babies was born to a mother who was also a boomer, which is pretty wild.

But, again, there were a lot of babies born in the baby boom. Below, I aggregated the annual totals and created overall percentages for the generations. I also scaled each row to show a relative comparison of the number of babies born in each generation. The height of the boomer row dwarfs the silent generation (though we’re only using the 1933-on portion of that generation). It’s also bigger than either the millennial or Z/lockdown rows.

But here we also get the answer to our other question. How good are generational boundaries at estimating generations? Mediocre. Gen X was mostly born to boomer parents (there were about 34.6 million of us), but so were most millennials. Most Z/lockdown babies were born to members of Gen X.

Boomers were the most likely to be born to other boomers; about 1.6 percent of the generation, 1.2 million babies, were born to boomer parents. The next most likely were members of the silent generation. The percentages here are skewed, since we don’t have data for the pre-1933 portion of the generation. But there were far more members of the silent generation born to members of the silent generation than X-X or millennial-millennial births combined. A main reason, of course, is that women had children at a younger age 90 years ago.

These are estimates, but they do provide a good sense of how our generational boundaries are sort of scattershot. And it lets me more accurately make my point: 1 in 60 baby boomers was born to a baby boomer.

It was a big generation.