When Retirees Return to Work Is It a Sign of a Strong Labor Force – or a Recession? – Center for Retirement Research

What “not earning” can tell us about the job market.

These days, economists find themselves in a fog regarding the labor market.

First, there is the fact that the September jobs report was released a month late and the October report was canceled entirely. But, beyond that, the existing data provide mixed signals. The number of jobs being added to the economy has decreased to the point of increasing. But, the unemployment rate has remained low by historical standards. Job vacancies and attrition – which measure how easy it is to find a job and how comfortable people feel about leaving it – are falling rapidly (bad) but also in line with historical rates (good). What’s going on?

In these cases, I turn to any measure of labor market health I can find. As a retirement researcher, one of my personal favorites is the unemployment rate. The attrition rate is exactly the same: the share of workers who say they retire at a certain time and then report that they are working a year later.

Going unpaid is a way for older workers to solve both financial and non-financial problems. Some retirees face financial stress due to low savings rates, high medical costs, and longer life expectancy; and others find that retirement is not what they expected and may miss the social connections and sense of purpose that come with work.

One might think that high non-retirement rates are a bad sign for the economy, for example because the stock market is low and retirees need to work to recoup their losses. But, in general, research suggests that the stock market does not continue to behave in relation to retirement. Instead, workers tend to retire when it’s easy to do so — that is, when the job market is tight. Therefore, looking at the unemployment rate can tell us something about the broader job market and where it is headed. If the unemployment rate is high, the labor market is likely to be good.

To check where the unemployment rate stands, I went to the same survey where those monthly employment reports (usually) come from – the Current Population Survey (CPS). I identified workers age 55 and older who said they were retired when they were interviewed by CPS. I then used the CPS longitudinal feature to track those same individuals when they were interviewed one year later. Those who said they were employed at the time were said not to have taken the job.

Figure 1 shows the share of retirees who quit just before COVID until the latest available data. (The time on the graph indicates when the person was first interviewed – that is, someone who said they were retired in August 2024 and working in August 2025 contributes to the August 2024 unemployment rate). The graph highlights in red the period when people would not have retired from the COVID labor market – that is, people retired between March 2019 and December 2020 – and those who would not have retired from March 2020 to December 2021.

This figure does not paint a very good picture of the current job market. Just before the COVID-19 pandemic, the unemployment rate was around 3 percent. That period was at the end of a long economic boom, so unemployment rates at that time could be considered a sign of a “good” job market. For people who would not have retired in the COVID labor market, that number averaged 2.2 percent and decreased by 1.8 percent. With the latest data available, the rate looks more like a COVID rate than an expansion rate. The latest available unemployment rate sits at just 1.9 percent.

At a time when labor market data is sparse and conflicting, any measure can help. Today’s low unemployment rate suggests that older workers find it difficult to re-enter the labor market and thus remain on the sidelines. In the coming months, I look to see if all labor market data starts to look similarly bleak.