Making Sense of Weak Job Growth Alongside Solid GDP Growth

0:05Is the U.S. economy still on solid footing? This spring, the big question was whether a recession was ahead. Now the question seems to be, how has the U.S. economy proven so resilient? And what does its continuing strength mean for the Federal Reserve? I'm Alison Nathan, and this is Goldman Sachs Exchanges.

0:25This week, I'm joined by David Miracle, our chief U.S. economist in Goldman Sachs Research. David, welcome back to Exchanges. Thanks, Alison. Great to be back. David, it's been a little while since you've been with us. So why don't we start with you providing us a top line view of what's going on with the U.S. economy right now? We've obviously been lacking a lot of the data we normally get because we have this ongoing government shutdown. But it seems like growth is holding up relatively well, despite fears that we and I think many people had that tariffs would weigh on growth. So what explains that resilience?

0:59Yeah, I think that's right. We came into the year, like most forecasters, expecting modest tariffs and 2 % plus GDP growth. We found out in March and April that we were getting a lot more than we bargained for in the way of tariffs. And we, in most forecasters, lowered our growth forecast from, call it the 2 % plus neighborhood to the 1 % neighborhood. Since then, we are, as you said, a few weeks out of date because of the government shutdown on what exactly is going on. We've missed a number of data releases. But based on what we knew before the shutdown, it looks like actually we're on track for growth this year more in the high ones rather than 1%.

1:33I think it is fair to say that the impact of tariffs has not been as negative as we might have thought back in March and April. My diagnosis of that would be that it's primarily because we've not seen much in the way of foreign retaliation against US exports. We've seen the dollar depreciate instead of appreciate. So again, that doesn't weigh on exports in the way you might have thought. And maybe the biggest issue, We have not seen much of a negative reaction from the stock market over the course of the year, which is a surprise after in 2019, we saw the stock market react very negatively to even very small tariffs.

2:08Now this year, we think we already have about eight times the tariff we had back then. We think we're on track for more, yet the stock market hasn't had the same reaction. And that means that two important channels of transmission from tariffs to the economy, that blow we expected to exports, that tightening in financial conditions we anticipated, just haven't amounted to as much as we would have thought. But we are having an ongoing government shutdown, which I think is lasting longer than most people had anticipated originally. So what bearing is that having on growth? Sure. So we're now up to about a three-week shutdown.

2:41If you look at prediction markets, they're expecting this to take about another three weeks. The mechanical effects of the shutdown for the economic statistics are that for every week that goes by, just because of the contribution from government to GDP, which is measured by looking at how many hours government workers are working, we estimate that this subtracts about one-tenth of a percentage point from quarterly annualized GDP growth per week of shutdown. But then in the subsequent quarter, so Q1 of next year, you would get that one-tenth growth boost back. When you start to think about a shutdown that goes on for say a month or more, then I think you start to take more seriously spillover effects to the rest of the economy.

3:21If you're not paying contractors, do those contractors start laying off workers? They're not getting the revenue. They're not doing the work. Maybe they don't want to continue paying these costs. At that horizon, I think that one-tenth of a per week estimate probably starts to look more like 15 to 20 basis points. the effects start to feel a little bit more real and not just purely mechanical. So I think we are getting into the range where you start wanting to think about those sorts of spillover effects. Let me ask you about the labor market, because at the same time that U.S. growth, GDP growth, has held up pretty well, it looks like the labor market has slowed pretty meaningfully.

3:56So how do you square that? Sure. So that's right. The jobs numbers are certainly quite weak and a lot weaker than we've all gotten used to over the last decade or so. Meanwhile, we think if you average the first few quarters of the year, GDP has probably grown at at least a 2 % pace. Now, the first thing to be said is in moderation, this combination of softer jobs numbers than we're used to, coupled with still solid GDP growth in the 2 % neighborhood, this is just going to be something of a new normal because productivity growth, which is kind of the gap between GDP growth and hours worked growth or labor supply growth, that has rebounded after being depressed last cycle.

4:36Last cycle came in at about 1.5%. This cycle, it has rebounded to the historical average of about 2%. And I would say that's mostly pre-NEFX of AI. So if we do get a further boost from AI, as we expect, we might very well see productivity growth rise above 2%. And that means that you can have still solid GDP growth, even though with an aging population and immigration that we think is now less than half of what it was in a typical year before the pandemic, we're simply not going to be able to create as many jobs as in the past. The issue this year has been that labor demand growth or job growth has slowed down even a little bit more than labor supply growth.

5:18Most of the decline, I think, is just that we don't have as many people to put to work because we've gone in short order from immigration of about 3 million in 2023 to immigration that we think will be about half a million this year. But the evidence that labor demand has slowed even more than was inevitable because of that slowing in labor supply growth is that over the last couple of months, we've seen the unemployment rate come up a tenth of a percentage point in each of the last two months that we have data for. And if you broaden your perspective beyond the unemployment rate and you look at many different indicators of labor market tightness, the message in general is a little bit worse.

5:55We've seen when we compile all of the different measures of labor market strength, a bit more of a softening to a somewhat weaker point than the unemployment rate in isolation would suggest. So how worrying is that? How worried are you that we see a severe deterioration in the labor market that ultimately weighs on growth? Yeah, I think for now, the concern is really more that labor demand is growing a little bit too slowly to keep up with labor supply growth. That should be by its nature, a slow moving problem, where hopefully a monetary policy intervention cutting a little bit more to try to nudge up total demand and nudge up labor demand can be effective.

6:35We are not seeing, and the same debate kind of played out a year ago when there were also worries about the unemployment rate moving higher in 2024, we are not seeing what I think would be more damaging, scarier, harder to cope with as the Fed, namely a big jump in layoffs where you worry that I lose my job, I have no income to spend at your business, you have to lay someone off, and the whole thing just spirals a little too quickly for policymakers to react and counteract it. So, so far I would say this should be a solvable problem, but I do think there are valid questions about whether or not in the years ahead, as more and more companies try to use AI to save on labor costs, whether or not it might just be a little bit more challenging to keep the labor market at this aggressive definition of full employment that we got used to both last cycle and earlier this cycle.

7:27And you do think that the unemployment rate is going to creep a bit higher from here? That's right. We have it rising a little bit higher through year end. Now, next year, we have that process stopping because we think right now that combined policy impulse from all of the policy changes that are happening this year is kind of at its most negative. Over the last two or three quarters, we've increased tariffs substantially. So the tax-like effect of that and any uncertainty effects of that probably should be having their peak negative impact on quarterly annualized GDP growth. I don't think we've yet through Q3 really gotten that much of an offsetting impulse in a positive direction from the fiscal bill, but over the coming four quarters, especially in the first half of 2026, I think that will begin to materialize.

8:13So when you think about the combined effects of those two things, looking ahead, we think that the drag from tariffs on growth should abate. The boost to growth from the fiscal impulse should begin to kick in and total demand growth should start running at a quicker pace with a little more consistency. And so that would be one reason that you might hope that labor demand picks up a bit too. So David, you briefly mentioned AI, but it's so much in focus. So how much of this jobless growth phenomenon really owes to AI? I think so far the impact on both GDP growth and the labor market has maybe been a little bit more limited than people sometimes think.

8:50On GDP growth, We estimate that AI has been contributing something like a tenth of a percentage point to measure GDP growth. There is an idea out there that AI has really been powering the majority of GDP growth, that without the boost from AI, the consequences of the tariffs would be a lot more dire, that the U.S. economy would be barely growing. I don't think that's right. What did happen is in the first half of this year, there was a huge increase in business technology equipment investment. And so a lot of people have looked at that and thought, well, demand for AI investment is booming. That's what's powering our economy.

9:24What we think actually happened was the White House was threatening to do large tariffs on semiconductors and other computers and electronics. Companies wanted to buy these already very expensive semiconductors and other goods without paying an extra 25%. And so there was a certain amount of front-loading of investment in these areas. A lot of that surge in business equipment investment to us looks like front loading ahead of tariffs rather than a true pickup in trend AI investment. On top of that, you have to remember that if you're importing something, that's not an addition to GDP. You can't just count the positive contribution to GDP on the investment side.

10:03You also need to net out the negative contribution on the import side. And once you do that, the impact looks a lot less impressive. This is a bit more arcane, but there is also a measurement problem where a lot of this AI investment is just being missed in the official statistics. So the thought that the economy would be on the precipice without AI investment is not really true. I think, again, we've been growing at about 2 % so far this year. That wouldn't be that much lower without the AI investment. And on the labor side? On the labor market, I would say you do see in the most exposed areas, think for example, of young computer science graduates, that times are definitely more difficult than they were a few years ago.

10:45And I think it's right to say that in applications where AI is pretty ready to go, this is already affecting the labor market. But more broadly, looking across the economy, we don't really see that pattern. If you look, for example, at the share of jobs or share of tasks that could be done by AI, or if you look at AI uptake already across industries and how that correlates with changes in job growth this year, you don't yet see much of a pattern. So I would say this jobless growth thought that it might be more of a challenge going forward to maintain the aggressive idea or aggressive definition of full employment that we've gotten used to earlier this cycle and last because companies might be hesitant to hire workers if they can automate their jobs with AI.

11:34That is still mostly a speculative thought, mostly a hypothetical, but I think it is one that is very worth entertaining because companies are very focused on this. And even if the technology is not here today, the promise that it might be in six to 12 months might very well lead company management teams to think we should be cautious is about hiring people who we might not need in six to 12 months because hiring and firing is costly. Interesting. Let's shift gears for a moment. Let's talk about inflation. It has cooled, but it remains above target. But is it right to think that fears about higher inflation are now behind us?

12:11That would be my view. I think that would be certainly the view of the Fed leadership. There's a range of perspectives on this. My take would be to think about inflation, you probably want to break it into two parts, the tariff effect and then everything else. Now, everything else I would say looks pretty good. For the last three years, inflation has been gradually coming down, and I think it's been coming down for reasons that are unusually straightforward, intuitive, and predictable. In normal circumstances, when the economy is pretty steady, I would be the first person to say, here's my inflation forecast, but take it with a bit of a grain of salt.

12:46Because while there is a moderate cyclical component to inflation that we as forecasters hope to get right, there is also a large idiosyncratic, somewhat erratic component that is sometimes somewhat divorced from the business cycle and hard to predict. But if you think of the last three years, inflation's been coming down because we got past all of the pandemic supply chain disruptions and shortages. We rebalanced what had been the tightest labor market in the history of the United States. And on the back of that, we've seen wage growth come down sharply, exactly as you would expect. And we've seen gradual exhaustion of what we've called catch-up inflation, where some prices adjust with a longer lag.

13:27But once they catch up in level terms, there's no obvious reason why they need to keep growing at a faster rate. Now, those latter two trends, the rebalancing of the labor market and the end of catch-up inflation, I think have a little bit further to go in bringing down the year-on-year rate. Because six to 12 months ago, catch-up inflation was relevant. going forward. I think it'll be less relevant. So I would argue net of tariff effects, core inflation is already running at about two and a half, and I'm pretty confident it will keep falling close enough to the 2 % target that I don't think anyone will need to agonize about this anymore.

14:03Now, the other half of the story though, is that we are seeing some tariff effects. They're not particularly large, but they are there. And I don't think we're done with all of that just yet, in part because we expect the White House to impose further tariff increases going forward. So we think so far tariffs have raised the price level by about 0.4%. We think that there's about another 0.6 % to come. This is going to leave us, we think, running in the high twos or maybe peaking at around 3 % over the course of the next two or three quarters on a year-on-year basis. Nothing like where we were in 2022, but not the 2 % target either.

14:40I think what this means for policymakers is even if most people probably agree on where we are, what's happening and what's going to come, you could imagine different perspectives on whether or not it's appropriate to be cutting. If you feel as we do, as Chair Powell feels, that this is a one-time price level effect, that it's okay to look through it, you might say, we're at about two and a half net of tariffs. It's not 2.0, but it's not terrible. We're still headed in the right direction. We can go ahead and cut interest rates. Other Fed officials, even if they largely agree with that outlook, they might say, you know, let's wait until inflation is closer to cut.

15:19So I think to some degree, this is more a matter of taste than it is differences in forecast. Well, we do have the Fed meeting coming up next week. And there are other factors at work here besides inflation in the sense that we don't have data for some of the important numbers that the Fed usually has at its disposal when it's making these decisions. How will that weigh on their decision making? Right. I think in the absence of the official data, probably the bias is to just sort of stick to the plan. And the plan, as I would interpret it from the September dots and from Chair Powell's guidance, was to deliver a package of risk management or insurance cuts in September, October, and December, 25 basis points each.

15:58I think you'd probably have to have a strong signal from the official data to deviate from that. And in the absence of any official data, it's kind of hard to see how they would deviate from that, certainly at the time of the October meeting. Now, by the December meeting, I would guess that the government will have reopened, that we'll have gotten three months of data. So it's possible things could look different. They could deviate from that path. But I think there's probably a pretty strong presumption, at least in the leadership's mind, that we're going to deliver a 25 basis point cut then as well.

16:28I'm sure Fed officials are doing what we're doing, trying to make the best of a difficult situation by looking at alternative data. And I would say, considering the dual mandate is in terms of employment and inflation, I think you can do decently well with alternative data trying to get at where the labor market is and where inflation is. On the labor market, we have a bunch of different alternative indicators of job growth. When we compile those into a composite index, it looks like job growth probably picked up a little bit from a low rate in the month of September. We also have a bunch of alternative data indicators from surveys and so on that are still being published that give you a sense of how tight the labor market is.

17:08There, unfortunately, it looks like the labor market continued to soften in the month of September. And then we have a handful of alternative data indicators that we use to forecast inflation each month. If you want to know something like GDP growth, I think it's very, very hard to use alternative data as any sort of substitute for the official data. So there are definitely areas where we are more in the dark. Right. But the Fed is not flying blind here on labor market and inflation data. That's right. David, this is always so interesting. Thanks for the update. I'm glad the news is, you know, mostly good.

17:40And we'll talk to you again soon. Thanks very much. Thanks for listening to this episode of Goldman Sachs Exchanges, which was recorded on Monday, October 20th, 2025. I'm Alison Nathan.

17:53The opinions and views expressed herein are as of the date of publication, subject to change without notice, and may not necessarily reflect the institutional views of Goldman Sachs or its affiliates. The material provided is intended for informational purposes only and does not constitute investment advice, a recommendation from any Goldman Sachs entity to take any particular action, or an offer or solicitation to purchase or sell any securities or financial products. This material may contain forward-looking statements. Past performance is not indicative of future results. Neither Goldman Sachs nor any of its affiliates make any representations or warranties expressed or implied as to the accuracy or completeness of the statements or information contained herein, and disclaim any liability whatsoever for reliance on such information for any purpose.

18:29Each name of a third-party organization mentioned is the property of the company to which it relates, is used here strictly for informational and identification purposes only, and is not used to imply any ownership or license rights between any such company and Goldman Sachs. A transcript is provided for convenience and may differ from the original video or audio content. Goldman Sachs is not responsible for any errors in the transcript. This material should not be copied, distributed, published, or reproduced in whole or in part, or disclosed by any recipient to any other person without the express written consent of Goldman Sachs.

18:56Disclosures applicable to research with respect to issuers, if any, mentioned herein, are available through your Goldman Sachs representative or at www.gs.com slash research slash hedge dot html. Goldman Sachs does not endorse any candidate or any political party. Copyright 2025 Goldman Sachs. All rights reserved.