Population structure by age
This is the second of two pieces about demography, this one focusing on the population structure rather than its size- specifically the changing composition by age group. Chart 1 shows three age groups – 20-40, 40-60 and 60+, as a proportion of the adult population (I have arbitrarily defined adults as being all people 20 or older, just because of the ease of aggregating the data, and making the cohorts memorably similar). The striking feature of the graph is the convergence of the three lines – from about 2030 on the three groups are roughly similar in size, having been widely divergent up to that point. The obvious question it provokes is what will this be like, and what difference will it make? The glib response is that we don’t know, because we haven’t been there before, although we are only a dozen or so years away, so we are getting close to age group parity.
In 1985 there were twice as many 20-40-year-olds as adults over 60. But of course, it is that same cohort – baby boomers, moving like a pig through a python, who are steadily swelling the ranks of the over 60s bringing their entitled demand for longevity with them. We have become used to cohort labels – Baby Boomers, Gen Xers, Millennials (soon to be followed by Gen Zers, apparently), labels which are largely applied during a generation’s early adulthood, when their peculiarities first become noticed. But whether they point to factors that are unique to a generation and its relation to a historic context, or merely to age effects – i.e. they are young and therefore different from the people applying the labels- is arguable (although boomers were probably the first self-consciously distinct cohort, defined in sharp contradistinction to their parents).
When we consider the future population structure, we can be fairly confident that the Rolling Stones will cease live performances sometime before 2050, but the question is whether the over-60s in 2030 and beyond will maintain their generational cohort distinctiveness. If the question is expressed slightly differently – ‘Will age effects outweigh cohort effects?’ the tentative answer is yes. Recent data shows that as millennials marry and start families, like their parents they leave Brooklyn and move to the suburbs;- it’s just that they are doing these things later than previous generations. More meaningfully, in all but one presidential election since 1976 the proportion of over-60s voting Republican has exceeded the proportion of 20-40 year olds (the one exception was 2000, when the proportions were roughly equal) suggesting that as they age, people become more conservative. This difference has grown markedly in the last three elections (see chart). Of course it’s a lot more complicated than that, as this New York Times article makes clear; but the interactive chart in the article still decisively shows that although events in formative years influence a person’s basic political allegiance, there is a universal rightward drift as people age.
So, interesting times ahead. Hold on to your hats.