We’re used to wondering how incomes have changed over time and over people’s working lives. Since 1984, they’ve been remarkably stable for men (in 2014 inflation-adjusted dollars):
And increased substantially for women, with larger increases for older women.
We could also look at how individuals move within the income distribution over their lifespan. So at each age, a man can be expected to earn at a lower percentile than in 1984, which makes sense if women are earning more:
Similarly, at young ages, women are not earning substantially at a higher percentile of the overall distribution than they were in 1984: it is instead that women at older ages in 1984 were earning at lower percentiles (since they were more likely to be married and out of the full-time workforce, in general.)
Another question might be how much younger workers in general earn relative to older workers, and whether this has changed over time. Younger workers are, on average, at a lower percentile of the earnings distribution than they were in 1984, and older workers are on average at a higher percentile of the earnings distribution:
That is, we can compare how the percentile an average person can expect to be at each age has changed over time, combining both men and women:
Younger workers are earning substantially less (as a percentage of older worker’s earnings) than they were in 1984.
- In 1984, a 56-year-old earned 27% more than a 28-year-old; in 2014, a 56-year-old earned 43% more than a 28-year old.
- In 1984, a 48 year old earned 83% more than a 24 year old; in 2014, a 48 year old earned 133% more than a 24 year old.
Why could that be? A few possibilities:
a) Young workers are starting their careers later, because they’re spending more time in school.
b) Older workers are “hauling up the ladder” after them, creating more barriers to entry (through credentialing requirements, for example) for young workers.
c) Younger workers suffered more from a weak job market now than in the past (although note that a 24-year-old worker in 1984 would have been affected by the deep 1981-1983 recession at least as much as a 24-year-old in 2014 would have been affected by the 2007-2009 recession.)
d) Older workers are less likely to be out of the workforce or working part-time than they were in the past.
e) The relative human capital of young workers as compared to older workers is lower now than in the past.
Note that (e) would tend to be omitted from projections of future growth, as well as projections of government revenue, taxes, and spending. That is, if younger cohorts are lower skilled than older cohorts were at their same age, it will almost certainly affect economic growth and lots of other outcomes for the economy as a whole, but we might not know it for a while.
This should also make us think about the effects of current differential fertility patterns on future growth, spending, and revenue, and whether changes in the age distribution of earnings are telling us anything about future changes still to come. Most kids in the country now are the children of low-income parents: the majority of students in public schools qualify for free-and-reduced-price lunch. If those children’s future adult earnings are influenced by their parents’ earnings as well as by other factors (technological progress, quality of public services, aggregate earnings across the economy, etc), we may well be overestimating future government revenue, fiscal solvency, and economic growth.