Where did all the manufacturing workers go?

It’s no secret that the American manufacturing sector has taken a huge hit over the last 50 years or so. In 1970, about a quarter of Americans were employed in manufacturing, according to data from the Bureau of Labor Statistics, compiled by the St. Louis Federal Reserve Bank. Today, that number is much lower. As of 2023, only about 13 million people still work in manufacturing, which makes up less than 10% of those employed.

So, what happened to all those people who lost their jobs when the domestic manufacturing sector declined?

Matt Notowidigdo is a professor of economics and business at the University of Chicago’s Booth School of Business. He joined “Marketplace” host Kai Ryssdal to discuss the manufacturing decline’s impact on workers.  Below is an edited transcript of their conversation.

Kai Ryssdal: With the acknowledgement that this is only a half hour program, what, broadly speaking, happened to American manufacturing in the last like 50-ish years? Just to set the table.

Matt Notowidigdo: Thanks for the reminder that it’s only half an hour. So, I would say the broad consensus among labor economists is that the decline in manufacturing comes primarily from two main forces. The first is technological changes that have resulted in essentially the automation of work. You could think of manufacturing firms replacing workers with industrial robots, and that’s been going on for several decades. And then the second main factor is what we call the “China shock,” which is just simply the sharp increase in Chinese imports that started around the early 2000s when China entered the WTO. So, think of these as being the two main push factors that led to the decline in the demand for manufacturing workers.

Ryssdal: So on that topic, manufacturing workers, which is why we’re here, the BLS, the Bureau of Labor Statistics says, give or take, the American economy has lost 7.5 million manufacturing jobs since 1980. Here’s my question: Where did they go?

Notowidigdo: It’s a good question. A lot of them, sadly, just left the labor market. So you know another fact to keep in mind is that the number of workers, the employment population ratio, has been falling for many decades.

Ryssdal: Layman’s terms, employment population ratio is for those unfamiliar?

Notowidigdo: So the employment population ratio is the number of workers that are employed divided by the total population of workers who could be working, right? And that ratio has gone down by several million over the last couple of decades, and a lot of that’s coming from declining manufacturing.

Ryssdal: So what happened to these people? I mean, some of them left the labor force, which I want to follow up on in a second. Some of them didn’t. What kinds of jobs are they doing now?

Notowidigdo: Well, when workers lose their job in manufacturing, as I said, a lot of them struggle to get back in finding another job, and if they really struggle to find another job, then they leave the labor force. They don’t work. They go to early retirement. In some cases, they collect government benefits, like Social Security and Disability Insurance. For the workers that have higher levels of education, they tend to just find a way to move out of manufacturing and do something else.

Ryssdal: Like?

Notowidigdo: Well, it’s hard to say because we don’t have great data tracking workers over time, in panel data. In some of the research I’ve done, it looks like some of them took jobs in the sectors that were booming. Like during the housing boom that people might remember. So you know, in early 2000s, 2003, 2004, 2005, house prices were booming. Construction was going up a lot. Real Estate jobs were going up a lot. It does look like those jobs provided an opportunity for manufacturing workers that had lost their jobs in the early 2000s.

Ryssdal: It’s been a while since this has been in sort of the ether out there, but there was a point in this economy, I’m gonna say, like 10 to 15 years ago, where there was a lot of talk about job retraining programs and government investment in that. Did that go anywhere? What happened to that?

Notowidigdo: We’ve been trying job training for a really long time, and it’s really hard. It’s hard to come up with training programs that look effective when people are older. You know, I think the bigger adjustment that’s happened is for the workers that are entering the labor force, if they see all these changes that we’re talking about, they can start college with a different mentality. They could go to college in the first place and try to do something else. That seems like that’s been a much bigger factor than retraining, which we don’t have a lot of successes to point to, unfortunately.

Ryssdal: Can I ask you about the towns and the people in the towns where these, mostly men, lost their jobs? You know, much has been made of the Rust Belt and hollowed out American cities and all that. What happened to the to the people in these towns, the families of these again, I’m assuming, mostly men, who lost their jobs?

Notowidigdo: Well, one of the sad consequences is in people’s health outcomes. When workers in manufacturing lose their jobs, suppose there’s mass layoffs from like a plant closure, it looks quite bad for people’s health. So, mortality rates spike up. People in middle age end up being more likely to die prematurely. A kind of silver lining that I’ve seen in more recent research is that when you go and look at the younger generation of people who are observing what’s happening around them, what we found is that in those areas that are most effective, you see people disproportionately increasing the likelihood they go to college. They decide to get a four year college degree, rather than just entering the labor force with a high school education, and that turns out to offset some of the income declines that you might have otherwise observed, because as long as the college education pays off, this is an investment that you’re making for your lifetime, that provides one way that you can try to adjust to the ongoing manufacturing decline.