It’s been more than two years since the pandemic trigged massive unemployment in the U.S. Some might say we have a nearly full recovery to pre-pandemic levels, mostly referring to the recovery of the ‘prime age’ labor force participation rate. But the story is different for older workers. Even if the news recognizes the possibility of a “great return” of retirees to work, they seem to interpret not as many older people working as a Great Resignation—implying older people are choosing not to work en masse. The true story is a lot more complicated.
To nail down what happened to older workers in the pandemic, I interviewed Owen Davis, PhD research fellow at the New School’s Schwartz Center for Economic Policy Analysis, who is writing his dissertation on labor flows for people aged 55 and older.
Owen, let’s start from the beginning. How many people would have retired if the pandemic recession in 2020 had not happened?
Comparing the total retired population now to what we expected if pre-pandemic retirement trends had continued, roughly 1.3 million more people age 55 and older are retired today than we would have expected. Here is a graph that shows how many excess retirements we have.
And how do you get that number?
The monthly Current Population Survey (CPS) asks American households about whether they are working and, if they aren’t, why not. If people tell government survey-takers that they are not working because they are retired, we count them as retired (so this definition excludes people who have left their main career job and now work part-time somewhere else).
If we simply track retirement over time using this survey-based measure—what people say they are doing—we see a sharp jump after March 2020 in the share of the older population saying they are retired. In February 2020, 48.1% of people 55 and older were retired; in March, 2022, that number was 49.6%, or 1.8 percentage points higher. That may not seem like a huge jump, but the number of people it represents is measured in the millions.
Of course, we can’t assume that population trends would have been frozen in time after February 2020. The aging of the Baby Boom population would have pushed the retired share up in any case. So we estimated what the retired share—and thus the total retired population—would have been in the absence of Covid-19. If the retired share have increased along a simple straight-line fashion between 2015 and now, we estimate 1.3 million excess retirees.
But Miguel Faria e Castro, senior economist at the St. Louis Federal Reserve, using less conservative assumptions, estimates 2.6 million excess retirees (as of late 2021).
So you are telling me that as we sit here in April 2022 that 1.3 million or even more say they are retired—whether it was voluntary or not—than probably would have been if we hadn’t had the pandemic recession. Isn’t that right?
Yes, there are somewhere between 1.3 and 2.6 million excess retirees in the older population, depending on the assumptions you make (though that latter number may be slightly out of date). I have to add that a routine Census Bureau adjustment may be causing an undercount of excess retirements as of January 2022.
Are people “unretiring,” as economist Nick Bunker of Indeed.com has suggested?
The “unretired” are people who previously said they were retired but now have a job. In the U.S. there are always some people who alter their retirement plans or simply misreport their labor force attachment in one month or the next. The most recent data on unretirement shows people are returning from retirement at pre-pandemic levels or slightly above. So Bunker is right that flows are increasing. But given the longstanding challenges that we know older workers face, including age discrimination, it is worth watching to see if unretirement flows increase to a significant enough level to bring the excess retiree population down to historical trend.
Does the evidence point to these millions retiring because they had enough money and were sick of working under the anxiety of the pandemic? Or were they pushed out? And since the average is never the story, how did retirement trends differ by different groups?
I use CPS data to examine retirement transitions during the pandemic. Since the CPS repeatedly interviews households over a 16-month period, we can see retirement transitions directly in the data to identify just who accounts for the retirement surge.
It appears non-college-educated people were slightly more likely to retire at elevated rates, especially at earlier retirement ages (55-64). There weren’t really any significant differences by race or gender. I was surprised. What best predicted excess retirement was whether older people were working part-time pre-pandemic. And, which is not a surprise, were in occupations that requires close physical contact with others. Workers in these two categories were hit the hardest by the pandemic.
As I said, early job losses were greatest in high-physical-proximity jobs especially. And, of course, the pandemic meant much greater health risks for older workers in close-contact jobs that can’t be done remotely.
But, were older people choosing to make themselves worse off financially because they were scared?
Unfortunately, the data can’t help us distinguish much between workers driven to retirement by health fears or by job loss (or some combination of these factors).
Why were part-timers more likely to retire?
On one hand, part-timers are more likely to work in service jobs of the sort that were jolted by the pandemic. But many older workers also work part-time as a bridge to retirement. Did a large share of these workers decide it was simply time to throw in the towel when the pandemic hit? If so, does this count as an early retirement? We don’t know from this data.
But other evidence makes me think these excess retirements weren’t by choice. As you know Teresa, our colleagues at the New School have found that a significant share of pandemic retirement transitions were unemployed-to-retired transitions, not employed-to-retired transitions. Retiring out of unemployment is what happens when people have run out of options.
We have been following the odd fact that all these retirements have not meant pressure on the Social Security system because claims for Social Security did not go up much more than expected. Please explain?
This is one of the big surprises of the pandemic. Economists, Gopi Shah Goda, Lauren Nicholas and Sarah See Stith recently found, applications for Social Security retirement benefits were largely unaffected by the pandemic, and applications for Social Security disability benefits actually fell (as did Google
What explains the divergence? Two factors are worth considering. One is that many older part-time workers already claim benefits and yet continue to work. This is one of the important findings in your work with Tony Webb and Mike Papadopoulos. So if we saw a large flow of people from working part-time to being retired, it could be that they were already claiming all along, which explains the muted impact on Social Security claims.
Another potential factor is that Social Security field offices were closed for more than two years, only recently reopening to in-person visitors. This could have reduced the number of retirement benefits claims, since not everyone has the resources to claim via the internet.
The first month of 2023 has been relatively kind to the stock market, at least compared with 2022. The…