Is the Labor Market About to Tip Us Into Recession?

Zandi, I think, hit the nail right on the head. Layoffs are the biggest concern. What we're seeing in the labor market is a slowdown of the gears, right? If you think of the back of a watch and you turn it over and you see the gears rotating and the cogs hitting, that's how you should picture the labor market of workers go from job to job to unemployment to job again. Mobility is everything for workers. The ability to move jobs, change jobs, get job offers. We have now three years of data showing just how much those gears have slowed. And it's not something that we typically and historically see in the unemployment rate, but we're not seeing now. You can look at hiring. You can look at quits. You can look at the three-month growth in wages. They all point to a decline in labor market mobility. Layoffs would just dump enough workers into unemployment in a labor market that's already slowly turning and it could lead to a terrible recession quite quickly.

One data point, one jobs report does not make or break the labor market. What's more concerning for me is that the economy shed jobs in four months of last year. And in fact, we've been on track every other month since June, we've lost jobs. So let's see if I can do this on the fly. June, August, October, December, February, we lost jobs. That's, I mean, I don't know how to say this more plainly. That's not good. That's indications of an economy and a labor market that's slowing down. I mean, we cannot come closer to teetering on the edge as we've been over the last 12 months. And we haven't moved really in either direction. But I think one figure that shows this in a really just truly striking way, easy enough to pull up, just the total number of jobs in the economy, payroll employment. Go back to 1939, I believe, when the statistics begin and you can see it's just a solid line going up and then every so often you have a little divot and a little divot and every little divot is a recession. What makes the last five years so remarkable is it's the first time this measure of basic size of our labor market has shown a flat line. We've never just plateaued. You know, we get to the top and we fall. We get to the top and we fall. The idea that we would plateau for one to two years is absolutely historically unprecedented, but that's what we're doing.

Uncertainty in the economy and economic policy. Sometimes you don't have to look that hard for an explanation. Parsing the economy right now is so difficult because the administration has pursued what can only be described as relatively chaotic economic policy. And to the extent that we know what it would do, it's destructive. We can only hit the economy so many times before it falls down. We've been watching it be punched over and over again. It's been resilient, more resilient than a lot of people expected it would be. That doesn't mean that it's impervious to economic policy that's meant to cause harm. Policy like deportation, policy like gutting the federal government, policy like tariffs, and now add to the mix policy like Iran.

Health and education are doing great. Everybody else is doing some type of bad. Hm.

Health and education jobs tend to be educated women. Not necessarily highly educated women, but you know, at least a high school degree. You'll have lots of technical occupations, master's degrees, all the way up to doctors. That is the only sector of the economy that is showing consistent and reliable strength. Everybody else is on some form of decline, either holding steady or outright shedding jobs for the last 12 months. Now, men and women don't hold the same job, and neither do people of different ages or different races. And so a lot of the pain in the economy is proxied through the type of job you're more likely to hold. I saw a Bank of America report. I don't know if y'all have seen this. It was written up in the New York Times of just how kind of ironically enough beneficial the Biden administration was for men's employment given that it had put so much money into manufacturing as opposed to 2025, the first year of the second Trump administration, manufacturing shed a pretty remarkable number of jobs. You see that pattern if you look hard enough at any job. So you're also going to see that, you know, middle-aged Black women in like the DC metro area were going to be hit hard by DOGE versus, you know, younger graduates that have a doctorate are in health or education, you know, they'll do okay. Those patterns hold for a lot of reasons.

I was reading today that one of a very prominent economist was saying that the answer to this puzzle of men's unemployment is to make certain jobs less feminine because they seem to be the ones that are growing for the future. It was an interesting note that kind of reflects what we all see even if we can't describe that men and women don't hold the same jobs.

I think I tend to prioritize growth that is both reflecting underlying economic activity as well as growth that is keeping the gears of the labor market moving. As much as I would have my preferences over the labor market, you know, the ultimate policy is not to mess with it too much, right? It does so much on its own. 170 million people have a job in this country. You've got people who are basically 14 to 95 years old. They have every level of education possible in every state, in every locality, and they've all got jobs. Mostly, it's all working really well. I think what I don't like to see is... I think maybe the better answer to the question is I think the sickness we have in the labor market is not sector specific. The sickness that we're seeing growing and mounting evidence of long term is that we have employer concentration that's freezing mobility and dampening wages. Think of it a monopoly amongst employers. There's mounting evidence that this has been slowly occurring for decades. So my preference would be whichever sector has the most businesses, the less agglomeration, the most competition for workers because that's what will be associated with lots of wage growth and mobility.

To a certain degree, yes. It's convenient and it's financially remunerative to corporations to say that they're laying off because of AI. And it'd be one thing if there was no penalty, but it appears to me there might be a reward for saying that you're utilizing AI. You know, it's not the first time this has happened in the economy. We'd like to think that it's just numbers, but it's really so much narrative as much as it is hard data. And the narrative around, you know, companies blaming something else for layoffs, I mean, that's a tale as old as time, right? We, you know, it's not that we want to cut jobs. It's just that the minimum wage is too high and therefore we need to put all your retail staff on an iPad. It's not that we, you know, don't want to raise your pay. It's that healthcare, you know, Obamacare made it required and now it's too expensive. I mean, if you go back, every decade has a new person to blame, a new story that shifts the blame from employer or rather corporate responsibility onto some other factor in the economy where they say, you know, our hands are tied and this is a blameless way for me to make a decision I'd like to make anyway. Lots of candidates for that over time. You know, we'll never really know the degree to which it's happening. We can only look to see what the data tells us with clear eyes. Those eyes are hard to keep clear.

So I think what it comes down to is we don't know the degree or effectiveness to which AI has been adopted by firms. So we can't trace out its labor market consequences. So it leaves room for basically both of your scenarios. Both that AI is starting to, you know, crowd out certain positions that could be entry-level. So it hasn't led to layoffs, but it's reducing hiring. And at the same time, it could also be a very, very convenient fall guy for otherwise corporate mismanagement during what are unprecedented economic times coming out of a pandemic and going straight into inflation.

Yes. And I think both stories lean on what we think to be absolutely the case, which is that AI will replace jobs in the future. It will cause job loss to the extent that it'll also cause job creation and we don't know what the pace of those two things will be and we don't know how it plays out over time. We know looking back to every other technological development that has hit the US labor market, it takes a long time for it to be absorbed into productivity. I mean, I could show you, you know, per-worker productivity in the United States going back to the 40s and if I covered the X-axis, you'd never find the internet. You'd never find computers. You couldn't point and say that was the year that the internet became mainstream or that's the year the personal computer sales took off. You would never see it because it takes a long time for that productivity to be adopted. So I think where we are right now is in a place where we don't have enough data points to parse out a story. And so we know what's going to happen. We don't know how it's going to happen or how quickly. That leaves room for both of these things.

You know, this is the problem with new technology. It just leaves a lot of room for guessing. I think what we can rest on are certainties and things that we know are happening. The economy is slowing down. The administration has destructive economic policy and we have long-standing weaknesses in our labor market that can be exploited by new technology and would absolutely cause harm without having to guess.

I take your point, but I'll put back to you. What do you need to know? Do you need to know exactly what AI is doing on the job and to workers to respond to weakness in the labor market? I mean, it's not as if the only thing standing between young workers and a job is what private employers are doing or say they are doing or will do with AI. We have a weak economy and a weak labor market with terrible social supports and almost no investment in people that don't have a job. I don't need to know what a company is doing with AI to know that our government does a terrible job in certain aspects of the labor market. So I think your frustration is well-deserved but not necessarily well-placed because there are solutions that don't have to come from just knowing more about AI but have to come from holding our policymakers accountable for economic mismanagement. AI is today's story but there will be another one. Weakness is weakness in the labor market. We don't have great infrastructure for helping people who don't have a job. AI won't change that. It'll just make it look worse. So it's almost... I could see it being so dissatisfying to have to deal with your story, Ed. But there is satisfaction if you go after the things you can change as opposed to the things you can't change or even see.

I think that it's the thing that people care about today. The thing that people are afraid of today. So, you know, we, the economy, we don't work for the economy. The economy works for us. If this is a thing that we think is coming that people are afraid of, we can respond to it. And I can sit in like my big room full of books and be like, well, actually, Ed, people said the same thing in the 1920s. They did. America does not like unemployed people. But that doesn't mean that like we have to do something to correct for the past and me point out that we've never had a great system for unemployment support. If people are scared about AI today, then let's act on AI today. It's going to put you at the same place that it would have put you about, you know, adoption of capital technology in manufacturing plants that reduce the number of manufacturing workers. It would have been the same thing that would have said mechanization of agricultural labor. It's the same problem, but it's today. A lot of people are going to lose their job and they don't know what they'll do after. We have failed this moment before. We don't have to fail at this time. So I don't want to say that you're wrong. I think that if this is the moment and that is the mood, then that's what we should act on. And it doesn't have to be the perfect diagnosis of labor market issues. It doesn't have to be that there's overwhelming evidence that AI's destruction is already clear. If it's something that will make the American economy and labor market better, something that will assuage the fear of American workers, it's the right thing to do. The years will tell us just how right it was.

Yeah, let me start from the beginning. Economy equals people, right? The size of your economy is really well predicted by the number of people in it, but people don't have the same level of productivity or relationship to the economy. You've got children, you've got old people. So really, the more workers you have, the bigger your economy is. And this is the basis for the economic returns for things like child care, for things like paid family leave, that if you maintain a worker in the economy, you've made the economy larger. Immigrants are people. It's not as if the labor market discerns from them in the way that policymakers tend to when they're trying to drum up nativism. If it's a person doing a job, it's not as if the economy is like, 'Nope, this is a separate person or this is a different kind of worker.' It is a worker. They're doing a job. They add to the economy. Immigrants rarely come as children. And they rarely come as the elderly. Most immigrants come in the prime age of their working years to earn money here. And immigrants, given their age distribution, have a relatively high labor force participation rate. So when we lose immigrants, we're not necessarily losing, you know, any random person. We're losing people who are highly attached to the labor force. So we're losing workers. Which means the way that you think about, say, how child care would improve the economy is that it has a similar kind of through-line for immigration: more workers equals bigger economy. Lower the number of workers equals lower economy. What a lot of anti-immigrant policymakers and shills will say is that they're taking jobs. That rests on the notion that there are a fixed number of jobs in our economy and it's slotted to whoever's there. We're not communist. It doesn't work like that. We're a market-based economy. The market is a function of the people in it, not what the government dictates or how those jobs are allocated.

So we said at the top of this that when you look at men and women's employment, they don't hold the same jobs. Immigrants and natives don't hold the same job either. So if I, you know, if I were to tell you my answer for male employment is I'm going to fire half of nurses in the US and then men can take their jobs, you would just be looking at me jaw-dropped like, no, that would never work. You'd have to train those people. They'd have to get the right job and they don't want those jobs anyway. They don't see it as a male job. But if I tell you that we deport immigrants and then Americans will take them, you know, we have such an underlying racism and nativism behind it that you'd fall for that argument, but you wouldn't fall for the male-female one. But from the labor market's perspective, they're relatively similar. Not everyone wants the same job. If you deport someone from a job they're holding, it's not clear that just anyone can slot in. We have these patterns in the labor market because we're human beings. We're not cogs. We're not worker bots. So the underlying economics is when you reduce the number of workers in our economy, you reduce our economy. Full stop.

I haven't seen much yet because I think everyone is hoping that it's quick and transitory. You know, the official line from the Federal Reserve is that they look through, quote-unquote, things like this where they're trying not to... Gosh, I'm so bad at metaphors. Mistake the forest for the trees. Is that it?

Yeah. Okay. Well, it's close enough. You all know what I mean, right? You know, you don't want to think that 3.3% inflation last month is a sign of underlying pressures to raise prices overall in the economy that need to be met with higher interest rates. We know where it's coming from. It's coming from the spike in oil. However, that can translate to overall price increases quite quickly. So it's a very tenuous balance in the labor market. I haven't seen anything directly yet, but that doesn't mean it won't come. When you asked me earlier what I was looking at, I have this like very odd fixation on surveys of manufacturers. Texas has one, the Texas Survey of Manufacturers, and then something like this shows up in the Beige Book of the Fed.