Bulwark Takes - The DOGE Job Cuts Finally Hit and the Jobs Report Is BAD

Episode Date: December 16, 2025

Sam Stein and Catherine Rampell take on a weak jobs report and why the topline numbers may hide more serious economic trouble. They discuss into manufacturing job losses, the Fed’s warning that growth is overstated, and the White House’s shaky spin—plus how AI could soon hammer white-collar workers.

Hey, everyone. It's me, Sam Stein, managing editor at the bulwark, and I'm joined by Catherine Rampel, who's the offer of receipts. You've got to subscribe to it. It's great newsletter on the economy and all things that are going terribly wrong. Hopefully we can start writing things that are going right. Yeah, I would love to include more information about good things happening. Yeah, this is one of those cases where I'm rooting for a good, upbeat newsletter. I don't always feel that way, but in this case, I really wish you could write about more uplifting things, which is not quite there yet. And that's true, I guess, with today's jobs number. So, top line figures, and then I'll let you cook. We had a gain of 64,000 jobs in the past month. The unemployment rate took to 4.6%.

That's the highest in four years. But the real headline was the October revisions, down 105,000. That's primarily, although not exclusively, because of the exit of tens of thousands of federal workers. They took the buyouts and the riffs, and they finally came to fruition in that data set. And then I would just say one last thing, wage growth ticked back down, or not ticked down from where it's been. It's about 3.5% barely beating inflation.

So, you know, we were kind of debating in the Morning Shots, Slack Channel. how to describe this report. We put mixed, and then we said, no, it's not mixed. It's pretty yucky. I don't know. Yucky. I think yucky is a good way. I don't know if that's the word we use, but maybe you should have been in the Slack channel to help us out there. Sorry. It's all right. No, but how would you put it? How do you describe it? Yeah, I think it was not great, particularly, like even if you take it at face value, and I'm going to say in a second, why maybe you shouldn't. The numbers were not great. We had job losses in three of the past six months. now. Beyond that, manufacturing, for example, has been losing months for the, excuse me, losing

jobs for the past seven consecutive months. So manufacturing appears to be in recession, despite all of the promises of a manufacturing renaissance because of Trump's trade wars, which was always BS, but whatever, the numbers do not look great. So there are a bunch of things that are not great in there. And then, like I said, there's what the numbers say. And then maybe how you should look at the numbers, like what gloss you should put on it. All right. So how should we look at the numbers? Yeah.

So Fed Chair J. Powell said last week that his staff at the Federal Reserve think that the numbers are being overstated by about 60,000 jobs each month, meaning that as lackluster and maybe bad-ish as these numbers look on their face, like they're actually overestful. how good the economy is. Can I ask you, why would that be the case? Like, I guess, look, I mean, it doesn't make sense to me that they, I mean, it does, I guess, because there's always revisions, but why would it be overstating versus understating? Yeah. So his staff thinks that they're overstating it by about 60,000 jobs a month. Like, I don't know what the exact number is.

But it's not that unusual, actually, to have big revisions like this or to be expecting big revisions when the economy is turning. And that has to do with the fact that let's say a lot of businesses are closing their doors right now. That messes up the methodology that messes up the survey, right? So like if you're only getting responses from businesses that are still open and you're losing responses from businesses that are shuttered entirely, you have you're going to have selection issues. It's basically just like whenever the economy turns, you have problems like this in the data because you don't exactly know if you're capturing a representative sample anymore. Got you. So it doesn't have anything to do with like Donald Trump cooking the

books, to be clear. I am worried in the long run about Donald Trump cooking the books in this data among others. But in this particular case, this is not about conspiracy theories. It's just like a methodological problem that often happens. I was going to tell my usual joke, which is the only way to solve this issue is to hire EG and Tony to run the BLS, but we already lost that option. Okay, let's say you were, let's say, Catherine, that they gave you the job of being Carolyn Levitt for a day, okay? And you had to like, I'm just saying, please know.

No, you're going to, you're going to play as well. And you had to pick a few data points and you could try to create a narrative that made, put the best gloss possible on these numbers. What would you say? Well, I guess I would say that job growth on net has been positive. Is that true? What is the net, though? It's like, it's like 17,000 jobs or something since the beginning.

I think Justin Wolfer said it was 17,000 jobs a month since April. Yeah, so it's not, not great. They would also argue that, well, we're no longer seeing jobs being taken. and buy immigrants. That's the line that they've primarily been using. Now, it is true that pre-Donald Trump, there were indications that most, if not all of the job growth for some of the past few years, was due to immigrants. But I would say that's a good thing. That was about demographics. It was like old Native-born Americans were retiring and leaving the labor force. And you had an increase of younger working-age immigrants coming in.

It's not, that doesn't mean that younger, native-born Americans were losing their jobs. No, there just weren't enough of them to fill all of the jobs that were going around. So, so it's like they would tell this story as like all of the immigrants were taking jobs. No, there were more immigrants coming in and filling jobs that would have otherwise gone empty. So now they're saying like, oh, there are no, you know, the number of immigrants. in jobs or taking new jobs is but the numbers show that it's not actually i mean the immigration the not the native born unemployment rate has gone from 3.9% to 4.3% so native born americans are actually more likely to be unemployed now than at the beginning yeah i know that's why i'm saying

doesn't even make sense i i'm trying to like tie myself into knots all right why don't we defending defending the data let's play what let's just play what they would say so on this morning Kevin Hassett, one of the top economic advisors in the administration, went on CNBC to talk about this, and let's play a clip, and I'll just get your reaction to it on the other side. You don't think it will cause any disruption shorter term, though? I guess that's the question. And again, we don't know the longer term, because when we have intelligence that exceeds our own, Kevin, I think all bets are off to a certain extent.

But you're not concerned at all about even a short-term dislocation. I mean, the buggy-wit manufacturers certainly had to find something else to do for a period of time. Yeah, they probably were unemployed. No, but think about it this way. The buggy-wip guy had to find another job, and how did he do it? He maybe had a newspaper with one ads in the back, but now we've got a person who might have to find a new job who can search throughout the whole country on all these websites that tell you where the jobs are, and the number of job openings of the country is in the millions and millions. And so moving people from one job to another has never been easier and faster.

and AI and the computer makes that even just more an efficient process. It's easier to find a job because of the internet. Look, I understand the point he is trying to make. Yes, it is probably a more efficient process today than it was 100 years ago to find new jobs. But it's not like losing your job suddenly becomes a lot easier. Like, would he use this same argument when talking with manufacturing workers whose plants, whose plants have closed? No, obviously not. Like, we saw this, including in the industry that this administration has been promising to revive.

They've been promising this manufacturing renaissance. Again, we've seen seven straight months of job losses in manufacturing. And that is just like one great example of how there are. these disproportionate localized effects of job loss of major technological change. I mean, like a part of the reason why we've seen a longer term decline, by the way, in manufacturing is not just trade. It's largely about robots, like people being replaced by robots. AI is a grander version of people being replaced by robots. And it's really hard to retrain people immediately, particularly if, like, your area has been, the one employer in your area has been

depressed, has been, has closed. It's not like it's a seamless transition. You might have, the most charitable way to interpret what he said is that like maybe AI will create jobs. We don't know what those jobs are. Well, I remember Howard Lutnik at one point in the administration she was talking about all the people who are going to have tiny little screwdriver or screwing in the things. But yeah, no, that's dislocated. And we've tried various versions of this before. I mean, the whole sort of thing. Yeah, it's like when the steel business was falling apart and you're going to

retrain people to be, I don't know, tech workers or something like that. And it's just like, in theory, I guess, as an academic exercise, you could do one for one. But like, are you going to relocate your entire family? Are you going to actually, are people going to take pride in those jobs? Are they going to be good at those jobs? I mean, it's not just transferable skill sets. Yeah. Whenever there's a big technological disruption.

And it is true that there is creative destruction, that some jobs are lost and new jobs are created, but the new jobs that are created don't necessarily go to the people who lost their jobs. There are winners and losers. And the mistake we have made as a country over and over and over and over again is not compensating the losers and not helping them get back on their feet, whether we're talking about trade or we're talking about technological disruption. And this is exactly the plan going into what is like.

to be a major disruption by AI from this administration. They're like, oh, well, it'll just solve itself. I was just going to ask you that. That was my next question, which is, okay, let's say, and I think this isn't a controversial assessment, that we're just at the beginning of this AI process, and it's going to be far more disruptive in the months and years ahead. What would you do if you were an administration trying to, I don't know, put up some buffers for the economy around that?

You would at least take seriously the idea of some sort of regulatory framework, right? And I don't detect that at all from this way does. It's sort of like a Wild West approach. Well, I think it's actually a very difficult problem to solve. Like, I'm sympathetic to the idea that there should be some regulatory guardrails, but I don't know what they should be, to be honest. And I think it takes people a lot smarter than I am in this particular area to figure out what we need, what's going to like just inhibit innovation and not actually protect anyone or what's going to protect people and allow a thousand flowers to bloom. I think it's a really, I think it is genuinely a really, like I know it's easy to be flip

and be like, oh, you know, just shut it all down and this is going to destroy the economy. And I think actually we want to make sure that AI brings the benefits that AI could potentially bring, whether it's about productivity or inventing new drugs or whatever. Like there are a lot of great applications that might be in the offing because of AI. Sure. But what would you do with? respect to the labor market. With respect to the labor market, I think it's about making sure that people have access to retraining and education. And you, for all I know, we might want to have some sort of, you know, like we have trade adjustment assistance right now for people who are displaced

by jobs. You may want to have some kind of safety net program. I'm just thinking all out. that helps people get on their feet or pay their bills if they've been displaced by this kind of technological change. I don't know that it needs to be tailored to AI specifically. I don't even know how you would prove that your job was taken by AI. But having a robust safety net so that if people lose their jobs, they're not destitute and that they have the ability to feed their families, I think, is generally a good thing, whatever the cause is.

Start with the podcasters losing their jobs. Star with a podcast. Yeah, I mean, one thing that's going to be different here, most likely, about the job losses due to the AI versus job losses due to other technological disruptions is that it's mostly going to be white-collar jobs, white-collar workers who are displaced as opposed to blue-collar workers. So, you know, it's a different population. It's not a population that this administration is particularly sympathetic to, right? Because white-collar workers more educated. be more liberal. Yep, tend to be more liberal. We've seen this huge bifurcation along educational lines. So I don't know that they're super motivated to help people, but it's going to be very disruptive. And the destruction or the attempted destruction of our entire education system, our higher education system in particular, is probably not going to be complementary to all of this either. Right. Well, I guess to put a bow on this, the reason I was, I kind of veered towards the AI component about the discussion is, well, you can't.

hear from the administration is, you know, and I guess they have to because of the politics that have been a very bullish assessment of what 2026 is going to bring. They say, well, you know, the big beautiful bill, it's really going to kick in. The stimulus from the tax cuts is really going to kick in. We're still sort of the doldrums of the Biden administration will finally be passed us. All the cuts that Elon and Doge did to the, you know, federal workforce are finally baked in. So we don't have to deal with any of that. And 2026 is just going to be gangbusters. And that's, like, obviously quite an optimistic take. I don't know if that's, I honestly don't think so. I think it is

willfully stupid to explain why to close this out. Because there's a lot of fragility in the economy right now, largely caused by Donald Trump's own policies, including the trade wars, the deportations, the attempt to crush the Fed, you know, attempt to crush the statistical agencies. All of those things are bad for the economy. To the extent that we're talking about AI's effects on the economy right now. I think AI is probably going to have big productivity gains in the long run. In the near term, as we've been discussing, there's going to be some severe pain from people who are displaced by AI, who are laid off because they're no longer needed,

graphic designers, podcasters, I don't know, lots of white-collar workers. Pay the podcasters. Yeah, I know. Tiny, tiny violence for us. so there's going to be some pain in the near term that's not about Donald Trump per se there's also potentially going to be some pain if the AI bubble pops and what I mean by that is while there is a lot of potential for AI to improve productivity and maybe living standards in the long run there's so much money chasing after being the big AI winner right now

that that is essentially propping up the stock market. It's propping up investment in the United States. And not all of that is going to pay off. Somebody will win big. Maybe a few somebody will win big and will have, you know, big gains, I hope, that are shared more broadly by the population. I don't know that they will be, but I hope so. In the meantime, all of the companies that turn out to be also RANS that have been investing big in these data centers, you know, they may all, they may simultaneously pull out and say, like, okay, we've been chasing this big win, we lost, and now we're pulling our investment out. If everybody does that simultaneously, that's going to be really painful to the extent that all of that investment has been propping up economic activity.

And the administration doesn't seem to be paying any attention or giving any credence to that big risk. They're like, oh, look, the economy's doing well. Everybody said that the economy is going crash because of tariffs and look it's doing super well the stock market's doing super well it's doing well because of these big investments and by the way the things that are that are being purchased that the inputs that are being purchased for these data centers those are largely exempt from tariffs so they've been insulated from all of this if that sector implodes um because there's one winner and everybody else loses, you could see massive, massive pain ricochishing around the rest of the economy.

Again, this administration is not paying any attention to that either. Well, you're so uplifting. Thank you so much for this. Again, I come back to the fact that if just E.J. and Tony were at the BLS. Oh, no. All this would be fine. No, I'm just joking. And if E.J. is out there.

He doesn't watch his stuff. I kid, E.J. Thank you for your service. Catherine Rampel, author of the Receipts Newsletter. Everyone should subscribe to it. You have a good one coming out. What is it? Tuesday.

You have one coming on Thursday that I'm really excited about. It's not about the numbers, although it has to do with the economy, obviously. Everyone should subscribe to that, subscribe to this feed for a great Bull War Takes. Thank you all for watching. We'll talk to you soon.