Charles Davidson: Welcome to another Economy Matters podcast. I'm Charles Davidson, a staff writer with Economy Matters, the Atlanta Fed's digital magazine. Today, we're joined by Ellie Terry. Ellie is a researcher here at the Bank who spends a lot of time and a lot of brainpower thinking about labor markets. Thanks for being here, Ellie.

Ellie Terry: Thanks for having me.

Ellie Terry at the recording of this podcast episode

Ellie Terry at the recording of this episode.

Davidson: We're going to talk about labor force participation in particular today, and some of the reasons why it's remained stubbornly low even since the end of the Great Recession. Ellie, first off, we're going to talk about how health and physical disability appears to be preventing people from working, or even looking for work, to a greater extent than it has historically. But can we set the stage a little bit? Let me ask you—just why is this important in the first place, labor force participation? The rate, as I understand it, is around 62.8 to 63 percent, which is still about 3 percentage points lower than before the Great Recession. Why is that important?

Terry: The labor force participation rate is an important measure of employment opportunities that exist in the economy. This number you mentioned, 63 percent, reflects the share of the population that are either looking for work or are already employed. So all else constant, increases in this number would generally indicate that economic conditions are improving. More people working would then directly contribute to economic growth, and less people on the sidelines would suggest there are enticing enough opportunities to make those people stop what they’re doing and start looking for work.

Davidson: Right. So do we see any demographic factors at work here?

Terry: Yeah. As you can imagine, many demographic factors can affect this statistic. For example, older people are less likely to engage in the labor market than middle-aged people because they’re more likely to be retired. And in addition to age, education, and cultural norms—that differ by race or ethnicity—can also affect a person's decision to engage in the labor market or not. As it turns out, the changing demographic profile of America can help explain about half of the decline in the rate of participation since the Great Recession.

Davidson: We're talking about aging there primarily, right?

Terry: Primarily it's aging, yes—but somewhat education. But education tends to offset the effect of aging to some extent.

Davidson: We mentioned poor health—essentially, physical disability—being a greater factor than it has been historically. How big a factor is that in particular?

Terry: It's a pretty large explanation, actually. And what's kind of interesting about this survey that's done each month by the BLSOff-site link [U.S. Bureau of Labor Statistics] is we actually know what people are doing instead of working. They're asked, "If you’re not working, what's your primary activity?" And then your response can get captured into one of five categories: you can say you're retired, you're in school, taking care of house or family, other, or you're too sick or disabled to work.

So they have a good set of different options, and it captures most everyone there—very few people put the other box, for example. If you're saying you're too sick or disabled to work, that specifically means you cannot do any kind of job for at least six months. So it's a pretty significant issue. You can't just have a cold or have the flu or something like that. Two interesting facts about the share of population that fall into this last category: one, it's the number one reason for not participating in the labor market for 25- to 54-year-old men, and it's a close second for the same age women—of course, behind household and childcare responsibilities.

The second really interesting thing about the share of the population that are saying they're too sick or disabled to work is that—like we alluded to earlier—the share of the population falling into this category is higher now than it was two decades ago, and that's true for every age group under retirement age. So 50-year-olds today are more likely to say they're too sick or disabled to work than two decades ago, but so are 20-year-olds. So it's kind of a broad-based issue across the age spectrum.

Davidson: Now, Ellie, just to clarify: when we talk about people answering the survey, we're talking about the Current Population SurveyOff-site link that the Census Bureau conducts. Is that correct?

Terry: Yes, the BLS and the Census Bureau do it together.

Davidson: Okay. Now, if you say in that survey that you’re too sick to work, or physically disabled, does that mean you’re collecting Social Security Disability Insurance?

Terry: No, it doesn't necessarily mean that. That's going to be something a little bit different. Social Security Disability Insurance is a program that's a social insurance thing, right? So you have to undergo interviews and sometimes medical examinations in order to prove you're unable to engage in what they say is substantial gainful activity, which generally means earning about $1,000 a month. If you can’t do substantial gainful activity, then you may be able to qualify for disability insurance.

These measures are related. Just as the share of the working age population that say "I'm too sick or disabled to work" has risen over the last two decades, the share receiving disability insurance has also grown over this time period. But because the questions and the qualifications differ slightly, there are likely people on disability insurance that don't say they're too sick or disabled to work, and vice versa.

Davidson: Right. So these two are different measures, but they're both higher than they were a couple of decades ago. It could seem a little puzzling, because by some measures the population in general is healthier, right? We're living a little longer—a couple years on average—than two decades past. Smoking is less common now, and Americans are less likely to die from cancer and heart disease. So how do these two things square? They seem to be going in different directions, slightly.

Terry: Yes, it is definitely very interesting that by some measures we seem to be getting healthier. But at the same time, other measures say the population is dealing with more health problems. If you think about high blood pressure, or high cholesterol, diabetes, obesity rates—all of these things are higher than they were two decades ago.

Davidson: Are there are geographical differences here? It's unfortunate, but it tends with these sorts of things that certain states—your less-affluent states—tend to fare worse in a lot of these measures. Is that what you found here?

Terry: Yes, Charles. Actually, that is what we're seeing here, and it's a puzzle why some states may have higher "too sick or disabled to work" rates than other states. But it's also perhaps a clue, because we can use that variation by the states to then look at different health issues and how they vary by state.

Davidson: Okay, so what do you see there?

Terry: Just to give you an idea of how different the states are: people living in Mississippi, Alabama, Kentucky, or West Virginia were three times as likely to say that they're too sick or disabled to work than residents in Utah, North Dakota, Iowa, or Minnesota. So if you imagine a map, maybe you would see the states that are closer to the Appalachian Mountains tend to have a higher share of the population that say they're too sick or disabled to work.

The most significant difference between two states is between Mississippi and Utah. In Mississippi, 13 percent of residents have some sort of health or physical disability issue that prevents them from working, compared to only 3 percent of those living in Utah.

Davidson: The opioid crisis receives a lot of attention these days. Is that condition showing up here?

Terry: In terms of simple correlation, I looked at a variety of health factors. When I looked at the opioid prescription rate, there was not a very strong relationship between the share of people in a population—or the share of prescriptions written per person in a given state—and the share of people that are too sick or disabled to work. There was much stronger correlation with some of the other issues that we’re facing as a nation. For example, high blood pressure and diabetes are very highly correlated with this "too sick to work" measure across states. But of course, we know that correlation is not causation, and there could be something else here that is correlated with obesity or high blood pressure that is also causing people to be too sick or disabled to work.

But getting back to your question about the opioid crisis: that one's kind of tricky, because when I first started thinking about this that was where my head went, too: "maybe it's linked to all these news stories I keep reading about the opioid prescription rates." So I was surprised when I looked into it that the correlation wasn't particularly high. But then I thought maybe it's because there's not that many people that are being prescribed opioids. But it turns out that's not the case—about 19 percent of the population received one or more opioid prescriptions in 2016, compared to 30 percent of adults that are obese.

So a lot of people in the U.S. are affected by opioids and by diabetes both. It's just that for some reason the "too sick or disabled to work" rate is not very highly correlated with the share that are taking opioids, but is very highly correlated with obesity.

Davidson: Yes, a lot of stuff going on here. Let's end on a slightly more positive note here: the percentage of the working age population who say they're too sick or disabled to be in the labor force has actually been going down, right? For about the last year or so?

Terry: Yes, it has been going down.

Davidson: Okay. Now, is that promising? Is that important? What do we make of that?

Terry: Absolutely, while there is a decade-long upward trend in the share of 25- to 54-year-olds that are too sick or disabled to work prior to the Great Recession, that trend accelerated when the unemployment rate peaked and then continued to increase even as the unemployment rate fell.

Davidson: Fell in the wake of the recession?

Terry: Yes, it fell after the recession. Basically, the peak in the "too sick or disabled to work" rate was after the unemployment rate peaked. The good news is that it has been declining for three years now, so this suggests that some of that acceleration that occurred during the recession was likely due to cyclical factors. And the fact that it's still elevated now could just be a hangover, an echo, of these weak economic conditions—it's now fading out.

Davidson: Right, very interesting. I guess it would be interesting to—and this is sort of beyond economics, maybe—but if during the recession, health problems caused by things like unemployment sort of exacerbated all this even more. I suspect there could have been something at work there.

Terry: Yes, there's definitely some research that says these two things are related—right?—that unemployment can cause health problems, especially mental health problems.

Davidson: Yes, all right. Well, Ellie, thanks so much for being here today. It was interesting.

Terry: Yes, thanks for having me.

Davidson: Well, that'll do it for our discussion for today. If you're interested in exploring labor force participation—and many other aspects of the labor force—in more detail, check out the labor force participation dynamics page on the Atlanta Fed's website, which the home page is frbatlanta.org, and you can find your way from there. And please be sure and come back next month. We're going to be talking with Julie Hotchkiss, a research economist here at the Atlanta Fed, who's going to discuss her research on how rising unemployment affects families and their finances. Thank you for listening.