Job seekers join a line of hundreds of people at a job fair sponsored by in New York.

Most measures of income—and therefore income inequality—are snapshots in time. They look at one year and treat everyone who earned the same amount of money in that year as similar. But the economic prospects of a recent college graduate and a person halfway through a career who both may earn the same amount in one year are quite different. Economists have long recognized this and have noted that measures of lifetime income would be helpful. Yet the lack of a dataset that tracks incomes over long periods of time presented problems.

A new paper uses such a dataset to present trends in lifetime income in the United States. The picture painted by the data is, unfortunately, a bleak one. The authors of this new paper—Fatih Guvenen of the University of Minnesota, Greg Kaplan of the University of Chicago, Jae Song of the Social Security Administration, and Justin Weidner of Princeton University—use data on individual incomes from the SSA spanning from 1957 to 2013. Using these data, they can look at how incomes for different cohorts of people (labeled by the year they turned 25) evolved over the years until they were 55. Given the length of time the data cover, the economists can look at 27 different groups with the youngest being those who turned 25 in 1983 and 55 in 2013.

The most striking findings in the paper relate to changes in lifetime income for men. The median lifetime income for men dropped between 10 percent and 19 percent between the group that turned 25 in 1967 and the group that hit the same age in 1983. (The wide range in the income drop is due to the use of different methods to account for inflation.) The lack of income growth wasn’t a problem for only the bottom 50 percent; more than 75 percent of men saw no increase in lifetime income during this period. The vast majority of U.S. men did not see any increase in income over their prime working years for several decades.

The trends for female lifetime incomes are less frightening. The median lifetime income for women increased by about 22 percent to 33 percent from the 1967 cohort to the 1983 group. But, as the authors point out, these increases are from quite low earnings levels for women in earlier periods. And data on the most recent groups (for which a full 31 years of data are not available) show a stagnation in lifetime income for women. The gains of past decades may not be replicable, as continued increases in the share of women working (and working full time) will be difficult to sustain.

The driving factor behind the declines in lifetime income for men seem to be driven by changes in incomes early in a working career. Incomes when men are 25 years old have declined, and income growth during the first decade after turning 25 are, according to the economists, the main reason why lifetime incomes for men have stagnated. And again, the trends in income growth for younger women in the youngest cohorts are starting to look similar to the trends for men.

The results of this paper, assuming they hold up to future scrutiny, would show that concerns about outright income stagnation for many Americans are not overhyped. For most men, in fact, income growth over the course of a working life has been negligible at best. Worries about the past may be the least of our concerns, as the trends appear to be continuing for men and changing in a negative direction for women. Policymakers who ignore such trends put at risk future U.S. economic growth and broad-based prosperity.