Job seekers wait in a line at a job fair in Southfield, Michigan.

New research shows the significant decline in lifetime income growth for most men in the United States over the past six decades and provides some evidence that the gains for women in the second half of the past century may soon be giving way to declining lifetime income growth, too. This is the main thrust of the paper by economists 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, but they also present data on inequality of lifetime income in the United States.

The results of their research are surprising, given the trend in traditional snapshot figures such as the famous data on inequality compiled from data analyzed by economists Thomas Piketty at the Paris School of Economics and Emmanuel Saez at the University of California, Berkeley. Piketty and Saez show large increases in income inequality since the late 1970s, but this new research demonstrates that lifetime income inequality didn’t change that much, despite the significant increases in one-year income inequality. What could explain the difference between the two measures of income?

Digging into the data a bit more reveals that differences in gender make a bigger difference for lifetime income trends. The overall level of inequality didn’t rise much year on year, but inequality of lifetime income for each gender did. Remember that women across the income distribution saw increases in lifetime income while most men saw declines, which means there was a decrease in the inequality of lifetime income between the genders. This decrease was particularly pronounced for the bottom half of income distribution, as the convergence for men and women’s income was strongest there.

This decline in gender inequality counteracted the increase in inequality within each gender, leading to no increase in overall lifetime income inequality. So, even though the lifetime income of, say, a woman in the top 10 percent of income earners has increased much more than a woman with lifetime earnings at the median (the 50th percentile), and even though the lifetime income of a man among the top 10 percent also outpaced a man at the median, the overall distribution of lifetime income didn’t change much when we look at men and women together.

The income gains for women over the past several decades seem unlikely to be replicated in the future, so looking at the trends within genders may be useful for thinking about the future. The reasons for rising lifetime income inequality differ for men and women. Higher inequality of lifetime income could happen because the inequality of initial incomes has increased (more inequality when a person turns 25), or because there’s inequality in life trajectories (differences in earnings gains over the course of a working life), or some combination of the two.

For men, the rising inequality of lifetime income appears to be due to higher levels of initial inequality. Inequality for men is seemingly locked in for men around the time they turn 25. And that level of initial inequality has increased over the years. Understanding what’s driving the decline in lifetime inequality for men requires understanding why inequality at age 25 has been increasing.

For women, the story is a bit more complicated. Inequality at age 25 hasn’t changed that much, but trends in inequality over a work career have changed. But these trends haven’t been consistent over the several generations examined by Guvenen, Kaplan, Song, and Weidner. For earlier groups, the researchers find that inequality dropped over time, and for later groups inequality fell for a bit and then rose later in life.

The four co-authors admit they don’t have explanations for these trends and the forces behind them. The changing life-cycle trends for women will be particularly interesting and important to understand, as well as the force increasing inequality early in men’s careers. Hopefully some more fruitful research will come these findings.