What is the State of Working America and the U.S. economy halfway through 2026?

so much. Um, I think I'm going to pitch the first question to Elise and I think the background is going to be we we did a version of this webinar state of the national economy a few months ago and I think the general overview was the economy was doing okay in the short run but there are some growing weaknesses and the long run outlook was sort of deeply threatened by the current policy trajectory. We'll dig into some of those weaknesses and the long run trajectory going forward. But for now, how's the shortrun state of the labor market in particular? What are today's data points? How do they compare to like a reasonable benchmark of a healthy economy? >> Great, thanks for the question. Uh, so I think it's pretty clear from the data that the labor market of 2025 26, so the last year and a half, it's clearly softer than it was in 2023, 2024. here in this first chart. Hopefully you are seeing that um that employment growth was stronger in um so this chart shows uh changes in pay peril employment growth over the last few years. We had solid and consistently strong employment growth in 2023 and 2024. You see those taller bars over the last two years. Job growth has been in fits and starts with several negative months. Uh the latest data that we got recently showed that there were 57,000 jobs added for June. It's important to note that without the World Cup adding jobs, uh that number would have been far lower as workers and their families cut back on discretionary spending as real wages fell. One particularly troubling um data point that I'm seeing lately is the hires rate on the next figure. The current hires rate resembles what we experienced in 2011 and 2012. So you see how low that hires rate has been dipping. And back in 2011 2012, the unemployment rate was over 9%. So right now we're seeing workers are afraid to leave their jobs and employers are sitting tight with the workers they have. Long-term unemployment is elevated and young workers in particular are having a hard time breaking into the labor market. Okay, so some pretty clear stuff there. Big slowdown in overall employment growth, a hire rate that looks really weak given sort of the rest of the labor market. Um, any data points that are a little puzzling, a little harder to characterize as either clearly strong or clearly weak? >> Yes, absolutely. The one I've been paying attention to is the prime age employment to population ratio. We call it the prime age epop. That has remained pretty strong over the last several months. Over the last year, we did see a dip in the latest data, but it's wait and see on whether or not that weakening for June will hold um in coming months. I'm not sure. It was a pretty big jump. Obviously, the data are volatile, but we'll see where that goes. Um that had been pretty strong. And the overall unemployment rate, we're seeing it's a bit up a bit relative to the 2020s lows. Um and that has mostly been for the wrong reasons as more workers have left the labor force. Uh we're also seeing the employment to population ratio for young and older workers less strong than for this prime age group. Another thing I'd note is that the break even for employment growth, so what is needed to keep up with working age population growth that seems to be heading towards zero. This is consistent with a no immigration economy. So it's a little hard to discern the latest data in that context. Again, we only get those population benchmarks annually. Uh but I think overall, you know, you don't need to see as much job growth because of that slower immigration. I think that another thing to note is that layoffs are low in the official data that we're getting we get from BLS, the JOLTS data um is showing that layoffs rate is still quite low. Even though we're seeing some anecdotal evidence and some reports out of uh from Challenger saying that there are more layoffs, that is not showing up in the official data. But if layoffs do begin to pick up and hires remain depressed, the unemployment rate is likely to spike. >> So maybe one more to you, Elise, that there's been a lot of chatter about young people suffering in the current labor market. Some of that talk moves on to say that AI is clearly the culprit replacing young college grads. Um, and then another strain has sort of noted that the very small number of jobs created this summer for teens, sort of the 16 to 19 year olds. Um, so just generally what should people know about the labor market for young workers in particular? >> Yeah. So when we think about young workers, there are a lot of different young workers depending on education, their age, where they live, uh, their race, and we'll get into that um, a little bit later. For recent college grads, the labor market is clearly softening. Um, and it's softening a bit faster than for overall workers, but it's also softening somewhat for young workers without a college degree. So in this chart you can see that over the long run the overall population um what we see is this narrowing of the gap with the overall unemployment rate and that for young college grads in particular and that's because on the long run the overall population has become more educated um and the democratization of a college degree has meant young college degree holders are less unique. Their opportunities in the labor market aren't as privileged and their college wage premium has beun begun to slide. So yes, it's true that a college degree still yields lower unemployment and higher expected wages, but to a somewhat lesser extent than it has historically. And so while those longerterm trends have moved slowly over time, the unemployment rate for young people with or without a college degree has been rising faster than overall unemployment the last couple years. You see that with that slightly steeper rise in those darker blue lines there. And so while the unemployment rate has been rising for both groups, I think it's important in historical perspective to point out that young people without a college degree are doing somewhat better historically. Um even though there has been this rise and the labor market for young non-ol workers um it's it's not the strong labor market that we saw in the early 2020s, but still relatively strong historically. And that matters. Most people in the United States still do not have a four-year college degree. And when we talk about young people, there's a lot of chatter about what's happening with AI. And I would have to say that there's little evidence that AI is taking over jobs or disproportionately jobs young people. Young college grads, like college grads of all ages, are more likely to work in AI exposed occupations. What you see in this chart is the red mass to the right in the in the let's say the middle graph um shows more exposed occupations. Young college grads in the middle look like college grads at the bottom. that those profiles are pretty similar in terms of more exposure to the right, less exposure to the left. The blue and the red sort of look pretty similar. So, young college graduates are not that different from college graduates overall. Um, and you can see that that young non-ol workers have much less exposed. But at the m same time, we saw that the unemployment rate for young college and non-ol workers has risen similar amounts over the last two to three years. So, I don't really find AI to be a compelling explanation for weaker labor market outcomes for them because of these very different profiles. It's likely something that's impacting young workers generally. Again, I return to the depressed hires rate. I think it's making it harder for new entrance to find jobs. >> Yeah. The other question I'll jump in here that Josh asked was about 16 to 19 year olds. And you know, every year with school being out for the summer, um there are always lots of questions about the labor market for teens. Specifically, there's been a lot of questions this year about whether the number of employed teens is the lowest it's ever been in history. And I think from what we're seeing, the number of teens holding jobs right now is low. um their overall labor force participation rates and their employment to population ratios um haven't changed dramatically though in recent years. So we can pull up that next chart and what you can see here the blue line is showing labor force participation for teens. That green line is the employment to population ratio. The big drop really happened after the 1990s. Um from about 19 the 1970s to 2000 there were typically more than half of 16 to 19 year olds who were engaged in the labor market and about 40 to 50% of them were employed. Now after the great recession we can see that teen labor force participation settles around 35% uh with between 25% and 30% of teens being employed. Um, both of those measures rose in the early 2020s and have fallen a bit since then. So that gets back to my original comment about you haven't seen huge changes in recent years, but definitely over the longer term. Um, you can see there has been a sharp decline. Uh, the other factor at play here would be the decline in immigration since early 2024. um that probably means there are fewer immigrant teens also in the labor force. And if we assume that teen immigrants have a higher than average employment rate, uh then their absence from the labor market could also be pulling down uh teen employment numbers overall. Uh the other important point about teen employment uh is to know that all of these changes really should be looked at in the context of longer running trends and uh particularly uh the long running trend of people staying in school longer and as a result uh we have more people beginning young adulthood while they're still enrolled in school full-time. In that sense, uh the long run decline in 16 to 19 year old labor force participation rates is mostly a positive development. Um if a growing share of teens are able to focus full-time on school and not have to work, um that could be a sign of a rich society getting richer. Now, when I say that, I also have to acknowledge um due to class and race-based inequities, we know that that is not the reality for everyone in terms of everyone having uh equal access to higher education. I think another concern for teen employment especially in that post200 period uh has been the disappearance of summer jobs programs that were an important source of employment for youth and underserved and high uh poverty communities. Uh those summer jobs provided work experience, they provided income and they provided valuable services uh to the communities uh in which they were placed. Uh but in 1998, the Workforce Investment Act eliminated funding for standalone summer jobs. And summer jobs then became one of 10 mandated activities uh that states and localities uh could choose from from just a single pot of funding as opposed to summer jobs having their own pot of funding and really being guaranteed uh to be created. Um those changes were made in favor of more yearround activities, but as we know, most teens are in school at least eight, nine months out of the year, and so then those opportunities for employment without dedicated summer jobs uh funding has really fallen off. >> Thanks, Valerie. Um immigration was mentioned once or twice by now. Can you give a broad overview of what we think trends in immigration and immigration policy have done to the labor market in the US economy over the past year or so? >> Yeah, the most direct effect has been uh slower population growth, slower employment growth overall. Uh the US-born population is trending older, birth rates are declining. Um and as these trends continue into the future, population and labor force growth uh will be further dampened by a reduction in net immigration. Uh but we're also seeing evidence that ICE activity can be clearly linked to damaged local economies. Um, not only have the increased presence and aggression of ICE agents in local communities, uh, been notoriously unwelcome and just disruptive to normal life, um, there's also evidence that they are bad for business. Um, a recent Brookings report estimates that the enforcement surge cost 668,000 jobs across 86 cities uh, where there were the biggest spike uh, in ICE arrests. Um and in fact the job losses in those places far exceeded the number of people arrested. Uh so they estimate that ICE made roughly 52,000 excess arrests and by excess arrest these are defined as arrest of people who are not already in criminal custody. So exactly what we saw just snatching people up off the street raids on workplaces um etc. But each excess uh arrest was associated with 13 jobs lost overall. And while job losses were concentrated in immigrant intensive sectors again due to a lot of the the workplace raids um and those would be sectors like construction and accommodation and food services um the job losses were by no means restricted or limited to those sectors alone. Um, a couple weeks ago, we also hosted a webinar on immigration and one of our speakers, Erin Sojourer, shared results from a report he c co-authored for the Northstar policy uh, action uh, um, group there in Minneapolis and St. Paul. Um, and they showed a 2.8% 8% decline in employees working, almost a 2% decline in hours worked, and 1.7% decline in businesses operating during the DHS surge um in the Minneapolis uh metro area at that time. They also estimated that that surge was associated with about $106 million in lost wages. So considering the magnitude of those effects, it seems reasonable to conclude that immigration trends and immigration policy may also explain, you know, some of the labor market weakness for young workers as well. >> Just want to jump in here. The labor market effects of both actual deportations and advancing in enforcement, they don't just fall on immigrant workers though, right? >> Yeah, absolutely. Um I think the story that the administration wants to tell uh is that by removing immigrants there will be more jobs available for USborn workers but that's just plain wrong. Um this chart um was pulled from a blog post that a couple of our colleagues uh produced uh few months ago and if we just go with what the data show you can clearly see that since Trump took office unemployment among US-born workers has risen. That's what those two lines are showing. The darker blue line is seasonally adjusted. So either way you look at it, we see that rise in unemployment for USORN workers. And I think the reason for this should be obvious to anyone who recognizes that immigrant and US-born workers are actors in the same economy. So, anything that has a major effect on one group will eventually work its way around to the other groups. Um, a few examples of sort of what that looks like. First, again, immigrants are not just workers. They're also consumers who contribute to the growth of the economy. And so, a slowdown in demand um when people are no longer going to work, people are afraid to leave their homes um means fewer jobs overall. The second uh way that this affects uh US workers, US born workers as well is that we know that few workers produce a complete product on their own from start to finish. Uh rather immigrants and US-born workers are compliments to each other in the labor market. So for example uh when immigrant roofers or farmers disappear and sorry or framers disappear and construction slows down um as that construction slows there's also less work available for nativeborn electricians and plumbers. Uh a third example of of the effect that this has is that when we think about child care, uh child care workers are disproportionately immigrant women. And when those women are removed from the labor market through indiscriminate detentions and deportations, uh US-born mothers and maybe even some fathers too work fewer hours to cover uh increased child care responsibilities at home. Just staying on policy influences for a second. Um, and maybe this one to Elise. How big a drag on the labor market has been imposed by the very large, probably illegal firings of federal workers since the beginning of 2025? >> It's huge. Um, the federal workforce today is down 324,000 workers since January 2025. Most of this show up in October. That's where you see this big drop off um when the fork in the road took effect after the end of the fiscal year. Again, this these are huge numbers. All else equally, it would reduce employment by about 0.2%. Uh, federal workers are skilled and hardworking as we know, and so many of them may have found alternative work. Uh, many have not. Uh, but previous work shows on average that those federal workers who were laid off or left um will eventually make more money while working shorter hours in the private sector. And we, the taxpayers who got a huge bargain from their work, will be the ones made worse off. state and local agencies might be just getting se getting seals of super qualified hardworking people. [snorts] uh the federal workforce data that's published by OPM the office of personnel management has a different total count of job losses but clearly enormous job losses as well uh they have a slightly different definition of uh workforce in their data but allows us to see within federal agencies where jobs have actually been added or cut since January 252 for instance the department of education has lost 45% of their employment IRS has lost 29% the EPA has lost 26% over time ICE staffing might make this more of a swap and be muting the continued federal employment losses. So over that same period there was an increase um in 42% in immigration and customs enforcement. And I would argue that's a bad swap and a theme will be coming to later the real damage of the Trump policy is in the long run. The macro economy can absorb a shock of these 342 24,000 jobs um to smart hardworking people likely have labor market success and still keep unemployment about level even if you know we should acknowledge that the shock is cruel and stupid to those people and a bad deal in the long run for taxpayers. But the macroeconomy will be less productive and grow less quickly over time as important federal functions just are not done and there is a cost of not doing let's say monitoring and prevention. So the Doge cuts to those types of um those types of jobs has meant vital government workers weren't assessing risks and mitigating harm. So we didn't learn, let's say, about the screworm until it was too late. And that cost us so much more money in the long run than if we had detected the problem earlier. >> So we've talked about the sort of overall labor market, labor market for young people, effects of immigration. Um maybe a question to Valerie, but what other disagregated analysis of labor market trends um have you been noticing recently? and jump out as interesting. >> Yeah. So, I have done quite a bit of analysis related to trends in uh black employment over the last year, year and a half. But before I talk about that, I do want to emphasize how important it is to have reliable data for doing those kinds of analysis. Um disagregation uh especially when we're talking about smaller racial and ethnic groups uh requires enough observations to produce estimates for those smaller demographic slices of the population and to measure changes in those estimates over time with a reasonable degree of precision. um to get sufficient sample sizes, statistical agencies already over sample certain demographic groups, which means they contact and survey more people from those groups than maybe is representative of their share of the population. Uh but a growing problem with the CPS has been the drop in response rates. It's one thing to ask more people, it's another to actually get responses and to get complete responses to the questions so that you have the data to analyze. Uh CPS response rates have declined steadily over the last 12 years or so. Uh they've gone from about 90% in 2013. Um there obviously was a big dip during the pandemic when everything shut down, but since 2024, um they've been down to less than 70%. So that's at least a 20 percentage point drop compared to 2013, the early part of the 2000s. Now, that's a problem because it makes it more difficult to detect meaningful change uh in labor market measures, especially those estimates that are based on smaller sample sizes like the black uh unemployment rate. But I also want to say that reliability in that context is a different issue than data manipulation, which is a question about the integrity of the data uh themselves. uh the statistical integrity of CPS estimates has remained solid despite declining response rates uh because the people responsible for producing those estimates are professionals. There are expert economists, they're statisticians and civil servants who follow transparent and wellestablished methodologies. So the integrity of the data is intact. The broader issue is the decline in response rates. And BLS and Census are already working to address uh any continued decline in response rates, but there really are just few substitutions for reversing the underinvestment in our statistical agencies um so that there's adequate funding and staffing to maintain and even improve uh data reliability. So, with that little uh PSA uh aside, I'll get on to uh some of the the black employment trends that we've seen. So, we're frequently asked about rising black unemployment as an early indicator of recession, but it is not uncommon to see spikes in black unemployment that are not followed by a recession uh due to more monthly volatility again due to the sample size uh limitations or just more precarious position um of black workers uh at any point in the business cycle. But about halfway through last year, there was a consistent rise in the black unemployment rate where we saw several consecutive months of that rate going up. Um it's been trending down some this year, but it's still slightly above uh what it was at this time last year. uh due to monthly volatility and that black unemployment rate. The previous graph showed the lines and you can see that it just moves up and down a lot from month to month. Um I tend to focus uh my analysis more on long-term trends in the employment to population ratio. Uh so that will be on the next graph with the bars. Um the first quarter of 2025 is is not on this chart. But when we look at the first quarter of 2025 compared to the first quarter of this year, that employment population ratio is down. Uh and it uh dipped down again in the second quarter of this year to 57.8% uh from about 58.2% in the first quarter of the year. Um last year I also did some analysis showing that black women suffered far greater employment losses than black men. Uh notably the largest losses were among black women who were college graduates and public sector workers. Again related to the trend that Elise was just describing in the the huge drop uh in federal employment. This year's data uh however are showing u more weakness among black men. Black men's employment to population ratio specifically for those aged 20 or older decreased by one and a half percentage points. um again from the first quarter of last year to the first quarter of this year. While for black women, we saw that ratio go up about point4 percentage points over the same time. But in the second quarter of this year, we see black men's employment rate up again while black women's employment rate is down again a percentage point. Um, the bottom line, the lower response rates really make it a lot more difficult to determine whether the decline in the black unemployment rate is statistically significant or an artifact of increasing volat volatility, but uh, employment rates are below what they were at this time last year for black workers overall. >> So, that's great. That answers a lot of questions we had about the current labor market experience of different groups of workers. Um, and that essentially boils down to the entire labor market is getting a scoch weaker. Um, particular when compared to very strong early 2020's labor market. So, I'm going to turn to you Josh. Uh, why is the labor market getting weaker? >> So, like the answer to that based in just arithmetic is that there's been a very slight fall-off in GDP growth rates. There's been a very slight pickup in productivity growth. That's how much output you can get with a given amount of labor. And until this recent month, the sort of strong or even slightly rising labor force participation rates. So the combination of all these things means that labor demand has been reduced a bit relative to labor supply. Some of this reduced labor demand has likely been matched by reduced supply stemming from lower levels of immigration um as you two noted before. Um but some of this reduced demand has just spilled over into higher unemployment um across all these different groups like you all showed. I think the economic reason for why demand growth slowed is I think mostly just a bad policy environment for investment. Um policy uncertainty is off the charts when people measure this. Um you just sort of see it over the past year and a half. I think it really began with sort of the Doge effort. Um and then it really sort of spiked off the charts with the Liberation Day tariffs and just sort of the wild and chaotic reversals and swings in those policies. And then every time you start to think policy is maybe settling down, there's another shock. I think there was one in August of last year when the BLS commissioner was fired for reporting inconvenient but accurate data. Um and so basically this is just a policy environment that is really bad for planning um for business investment. And you know the deportation policy agenda is a really key part of this as well. This deportation policy makes both future demand and future labor supply for a bunch of industries incredibly uncertain. And so it makes it very hard to like take the leap of faith to begin sort of a forward-looking investment in a lot of these industries. Um basically if you look at parts of business investment outside hardware, software, data centers, anything not being directly pulled along by the AI tsunami that we'll talk a little bit about um more, you see outright contraction in business investment spending. And I think that explains sort of the weakness in GDP growth. Um I think if you had told me the scale of spending associated with that AI capital buildout and the stock market bubble we're in, they said that would happen in 2025 and 2026 and then asked me to predict what overall growth would have been in those years. I would have been like wow we must be growing really really fast. So to have actually a demand slowdown even with those kind of once in a generation tailwinds from this sort of timing of the AI um boom is just kind of an impressively bad outcome. Um as to why productivity growth picked up, I'd say it's mostly just catchup from sort of the COVID related productivity collapse when you had supply chain snarls all throughout the global economy and this huge reallocation between sectors. I think in a second, there we go. We're gonna have um a figure. This is just the level of productivity. So it's how much output is generated in an hour of work in the economy. Um and basically today we have productivity that's about 1% higher than you might have predicted back in 2019 based on the trend. That that's kind of nice. It's not gamechanging. And if you want to look at the growth rate, that's kind of the slope of these lines. And so actual productivity growth um is the blue line and then the trend productivity growth from before COVID is the red line. in the past two years or so, there's just not a huge trend difference there. So, this is not um a new and sort of unanticipated level of productivity growth. It really is just bounce back from that very weird early 2020s period. >> So, Josh, you just referenced this um a bit in your your previous comments, but you know, the question comes up a lot around AI. um just how reliant on AI spending has the economy become. Can you say a bit more about like the magnitude of that? >> Yeah, I think that's one of the few things about AI we can actually speak with a little confidence on is like the short run macro impact and the short answer is very very reliant. Um basically since like the first quarter of 2025 until now we seem to be running at a GDP growth rate of about 2% annual growth. It ticked up in the first quarter of 2026 for some quirky reasons that are unlikely to be sustained, so call it 2%. This figure shows um basically the contribution to that GDP growth rate from AI spending on capital investment, building data centers, hiring software coders, doing R&D. Um and basically you want to compare the contribution those sectors make to their normal contribution. Like all sectors pretty much in the US economy are growing year after year. So you don't want to set zero as the baseline in this figure. You see that sort of dotted straight line horizontal line at the bottom. That's about 4%. That's their normal contribution of these sectors to growth. It's now about 0.7% higher than that normal contribution. So this is the contribution to overall growth just from capital investment associated with with the big AI buildout. And then on top of that um you've got an extra consumption effect stemming from sort of excess stock market valuations. Um people always consume out of wealth, but we want to know how much of today's consumption is financed out of sort of the the bubbly bit of um current stock wealth, which we define as today's stock market valuation versus what the stock market would be today if normal relationships between profits and stock prices held. And if we could just go ahead one slide, I think I just have um a slide on sort of today's um stock market relative to the normal. This is just the the price to earnings ratio in the stock market. It's basically this uh price of stocks relative to underlying fundamentals and profits. And you can sort of see today's figure of just below 40 is much higher than the historic norm. The only time we've had a more overvalued stock market relative to expected profits was in the internet boom of 2000. That stock market fell a lot and we had a recession because of it. Um but for now, this amount of overvaluation of the stock market is probably adding like another 0.7% to GDP growth. So you add that to the capital investment, 0.7%. More than half of GDP growth is very much associated with sort of the AI investment boom. And I'll just say two quick things on that that are probably obvious to everyone, but worth saying out loud. One, with a tailwind to economic growth this large and not driven by policy, it is again an impressively bad accomplishment to have demand so normal or even a little bit weak in an economy with one engine firing this strongly. It means kind of all the other engines in the economy are not firing very strongly. And again um we can actually go yes to this slide. Thank you. This is sort of contributions um to the economy from business investment outside the AI sectors. And you can see that's actually gotten negative um in the past year and a half. And so that's sort of the example of this is I would argue the fruits of a bad policy agenda. And this is the reason why even with an AI boom, we're not having really really strong economic growth right now. Um, and then I would say, you know, the other issue is that having half or more of growth supported by this sort of narrow and fragile base of AI spending, it just means that um this could deflate really quickly. Like if this main source of economic strength is a single sector that relies at least a bit on some very optimistic projections for the future about what this technology could do. That is a recipe for a very quick evaporation of demand and a recipe that says you know policy makers should be really ready to respond to a deep slowdown in growth or even a recession at some point in the next couple of years. >> Thanks Josh. Just one related question. How reliant has the economy become on spending by the rich? You often hear this stat that 50% of consumer spending is now being done by the top 10% of the income distribution. Is that right? >> I would say that that specific stat is almost certainly not right. Um we know the top 10% have about 50% of overall income, but we also know they have much higher savings rates than everybody else. Basically, the top 10% do almost all the savings of the US household sector. So their income's 50% but they save a lot more. It seems impossible that they're spending 50% uh on consumption. I think this particular stat is probably grouping households by cash income and it's missing a lot of income and consumption that is financed by inind public spending mostly Medicare, Medicaid, SNAP, things like that. This is income and consumption overwhelmingly taking place like in the bottom half of the income distribution. It's a lot of spending. Um, so I think that's probably where that stat is kind of going wrong a little bit. I would just say that the BEA and the BLS have looked at this in the past like what share of consumption spending is done by the top 10 or 20%. They basically have the top 20% of households accounting for about 40% of all spending. So unbalanced but not as unbalanced as is that stat that's move going around. All that said, I do think there's likely an effect of that currently overvalued stock market and pushing up household consumption quite a bit. Um, by definition, the people enjoying the gains from the current stock market boom and increasing their consumption are wealthier than average. I think it's 90% of corporate equities are held by the wealthiest 10 to 15% of families. About a third of all equities are held by the wealthiest 1%. So, if the worry is that the economy is too reliant on a spending flow like consumption out of wealth gains that might quickly reverse if the AI bubble pops, that's 100% fair. I agree with that worry. And if it's also true that this sort of fragile spending flow is currently concentrated among wealthier than normal households. So like the spirit of this concern about too much consumption being driven by top income households seems totally fair enough. I don't think that specific stat is right. >> So you know we rarely just you know people just sit around and just let the economy run and do whatever it's going to do. You know there are like policy decisions that can be made. Um, so if the economy is softening and this is causing the labor market uh to falter, why do you think we didn't see the Fed cut rates at their last meeting? >> Yeah, I would I would say we're in a very complicated time and I have more sympathy for Fed reluctance to cut interest rates than I think I normally would given sort of this constellation of economic data. Basically, even as unemployment has drifted up in the past year or so, which would say to me perhaps time for a rate cut, um the big progress in reducing inflation that we saw in 2023 and 2024 kind of came to a screeching halt even before the Iran um oil price shock. I think some of the inflation measures right now are being pushed up by very weird data quirks. There's we can talk a little bit about that if people care, but basically data quirk or not, the main inflation indicators the Fed has traditionally looked at the deflator for personal consumption expenditures, it's really reacelerated in the past year. And so it's it's tough for them to cut rates in the face of that. Again, I think that's a quirk. If I were in charge of everything, I would not put a lot of weight on that. I would also say though that some of the stalled progress in reducing the high inflation of the early 2020s, it's not a data quirk. It's because of clear policy changes. Um the big disinflation in 2023 and 2024 was really concentrated in good sectors. They were the same sectors that kicked off the inflation of the early 2020s when supply chains broke down. When we untangled those snarls, inflation started to come back down. I think the abrupt end of that disinflation in goods that happened in like mid 2025 just really links up tightly with the big increase in tariffs. Um tariffs raise prices. That's often fine if they come with a strategic purpose to make the higher prices worth it, but they do raise prices and you see it in the data. And then I think other aspects of the administration's policy agenda also stopped disinflation, deportation agenda. Again, I think it led to labor shortages and increased prices in a number of sectors. Just one really salient one people always talk about. Grocery prices in 2024 actually saw inflation under 1% for most of the year. That's a pay slower than pre-COVID growth. the price of groceries relative to wages in that year actually hit near historic lows, but all of this was reversed and by the end of 2025, grocery price inflation is now above 3%. Um, and so, you know, you've got all this sort of building inflation data that I think is really um sort of tying the Fed's hands. And then on top of that, you've got the very large deficit finance tax cut passed in 2025. And so the Fed is thinking, you know, we've got unemployment that's below four and a half% and we just did this kind of maybe fiscal stimulus in the form of the deficit finance tax cut. So I think they uh and we've got all this tariff and deportation supply side effects that may not have filtered all the way through. So I think they feel pretty, you know, reluctant to respond very quickly until they see all the effect of all these shoes dropping. And so I would say like if we had not changed the pre2025 policy path we were on over the past 18 months. Um the Fed would feel much freer to respond more quickly to unemployment that's drifting up. So it's not just that this policy agenda has already done some noticeable damage and we'll do more going forward. I think all of that is true. I also think it has really put some handcuffs on the Fed that is going to make them slow to respond when they eventually do need to respond really sharply. And and that's yet another piece of damage from it. So, like every good economist, there's on the one hand on the other hand. Um, but you know, if you had to decide, you know, what should the Fed cut rates or not? >> Yeah, I would say if if I was in charge of things, I'd personally vote to cut or at least be ready to cut very soon. I would say I'm very doubbish. I think the benefits of very low unemployment dwarf the cost of inflation being a little bit above the inflation's target. Um, I also think a lot of the inflation that we've seen over the past year that up a lot of it is supply shocks that you should look through and hopefully will not be permanent. Um, I get it that most people think that given the public reaction to the inflation of the early 2020s that it's just so politically toxic to ever talk about tolerating even a tiny bit more inflation that you can't do it. But as an economist, I just think the cost of high unemployment is bigger than the cost of a little bit um above target inflation. I will note as well that what sounds like a small drift of the unemployment rate from about 3 and a half% that we reached in 2022 and 2023 to today's 4.2% likely means we've moved pretty decisively off genuine full employment from a macroeconomic perspective. You know, we had sub 4% unemployment for a full year or more pre- pandemic with no overheating at all in labor markets. Um we could probably get back there. And this graph right here basically shows you that when you look across all states since 1979 and you look at the relationship between the unemployment rate and real wage growth, wages adjusted for inflation, anything above 4% unemployment is kind of associated with real wage decline. Um and so we're now at that cusp point where you're going to have a bunch of workers struggling to get any inflationadjusted um raise at all in coming years. And so this is why, you know, I'm pretty doubbish on this. I think you need very low unemployment rates to get any real wage increase at all. This is for median wages um to typical workers. And so this is why I would personally be arguing for cutting. And then of course I talked before about the fragility of the current growth. It's so reliant on a single engine that could really sputter at any time. And so at least the Fed should be ready to pivot very quickly and start thinking about um helping an economy if it really does start to falter. Um, and since we've talked about it, um, a little bit, some of the inflationary pressure, um, all this talk was mostly before the the Iran conflict. All of the influences I talked about were about before the, um, oil price shock from Iran. Um, and so just going to ask you, Elise, what what is the effect of that price shock on the labor market? Yeah. So certainly we've seen a big OB up in the price level that is almost surely going to mean u price growth will beat wage growth for much of um the year uh before it settles down moving forward. So you can see that in this chart here I have uh nominal wage growth and then inflation and and real wage growth. Just a second on nominal wages. I see a several questions in the chat here. Um obviously here we see that prices are growing faster than nominal wages but nominal wage growth has decelerated. um at a pretty slow pace indicating that workers have less leverage to bid up their wages um and there clearly there's no inflationary pressures coming from the labor market. So, um, we see this this slowdown. Again, Josh pointed to the graph showing that that slowdown is is more likely to happen. Uh, as the unemployment rate ticks up, workers don't have that leverage, that kind of bargaining power. We're also seeing this depressed um rate and quits rate. Both of those things together, uh, the best way that people often get uh, wage increases, take another job. as workers sit tight and employers sit tight in the face of uncertainty um that is also I think holding down nominal wage growth. So then we add in prices we see this March in March the prices spiked and um now real wages are below where they were in January 2025. Yes, we don't have the next inflation month but that is surely the case. They were matching it last month um and that has continued to fall down. We've also seen gas prices come down a bit, but it remains the case that price levels remain higher than they were relative to last year um throughout this year, I believe. >> So, at least, do you think there's any danger that the oil price shock um could spark self- sustaining inflation? >> No, I don't think so. Um we have seen the labor market weaken. It is weaker than it had been the last couple of years. Um wage growth is muffled uh rather than amplified oil driven inflation even in 2022. when the labor market was roaring. It's hard to see how much uh today's much softer labor market generates any kind of self- sustaining inflation. I'm not worried about that. Um so that's, you know, a good news story. Wage growth rising in response to inflation is a threat in terms of getting overall inflation back to normal levels, but it's also a crucial protection for workers families income. So the fast wage growth um 2022 that never amplified inflation, but it also made it slightly longerlasting and that was fine. that fast wage growth also get the real income losses of working families in check. So I don't see any self-sustaining risk of inflation, but it has happened before. And so I'm going to turn to you Josh. What's different today about today's oil price shock relative to the ones from the 1970s that were associated with a decadesl long struggle to re in overall inflation? I'll say a couple quick words about that and then we're getting close to the end and I do want to um save some time for like policy responses in case things get noticeably worse um in the next couple of months. So on your answer about you know what is different about today's oil price shock relative to 1970s I think it's a lot of it is related to what you just said. It's basically uh disempowered workers like when a inflation shock begins from outside the labor market because of a big external event like an oil price shock. Whether or not that initial shock is sustained or amplified into higher inflation that goes on a long time really depends on large part about how effectively workers and employers are in trying to protect their own incomes. Like if an oil price shock is met by an empowered workers demanding large wage increases and then employers try to protect their profit margins by passing those higher prices on to customers, you just get this sort of amplified cycle. If instead that external price shock hits disempowered workers, they just get lower real wages once and for all and the economy moves on. And I think between a long run structural disempowerment of the US workforce from the sort of dismantling of unions and other labor standards and the fact that as we've been talking about the labor market just isn't that strong today. I think what's mostly going to happen is the oil price shock hits real wages go down and that sort of snuffs out inflation and like you say it's not like that's a great news story that inflation is stuffed snuffed out quickly. It just means workers take all of the losses. Um I do want to say one word and then I'll um ask you all for some input into this is just we we've been talking about risks to the economic outlook going forward. And so the real question is what what should policy makers be ready to do if we start to see a deterioration in the economy and what would that deterioration look like? How should policy respond? I think I will just sketch out quickly what it what it'll look like and then pitch it to you two for a little more on what people should do, what policy makers should do about it. Basically, I think the story is mostly one where the AI optimism starts to dry up and those big capital expenditures and the stock market boom really start to rain back in and that sucks a lot of aggregate demand out of the economy. What that thing is, who knows what it's going to be. Maybe it's going to be another announcement from like a Chinese AI firm that makes people realize the American firms are losing the race and so enthusiasm dampens. Or maybe there's going to be a big foreclosure of a huge data center. Something's going to make people think maybe this has all gotten a little bubbly. maybe we don't want to be as invested in sort of the AI race. Um, and I think if that happens, given how reliant on those spending flows the current economy is, you're going to see GDP growth really contract. You're going to see unemployment rise. Um, and you're going to see the economy stagnate until policy changes sharply enough to fill in some of those holes in aggregate demand caused by the pullback in AI spending. I think the first policy change that you can do to fill in some of those holes is pretty straightforward, but it's going to be pretty weak. The Fed could cut interest rates. they've got a little room to cut short-term rates. Um they would do that if the economy got really soft. Um my guess is if that happens, we'll see zero short-term rates again and we'll still need further help from policy makers. And I think that'll most be fiscal policy makers. And so I'm going to turn it over to Dall to talk a little bit about what that fiscal policy response um would likely be needed in this scenario. >> I'll I'll start and Valerie can jump in. fiscal policy. Um, just to remind everyone, is changes in public spending and taxes. That has to be done with Congress and the White House working in tandem. Um, recessions really are the only times this happens at all. And so, the general recommendation for fiscal policy in this situation, I think, is easy. One, make sure aid is targeted to alleviate the suffering caused by the downturn. And two, direct more money and income to households who will likely spend it instead of saving it. Luckily, it's easy to hit these goals. So you could ramp up the unemployment insurance system. Both duration and generosity. Similar with SNAP, you could pick up state share of Medicaid spending for some period of time. Uh make the child tax credit larger. All of this can be temporary. I'd love more generous UI and broader public coverage of health insurance and a larger CTC permanently. But the argument to at least do this temporarily seems solid. We saw that in the COVID recession. aid at scale works and we can really use UI more aggressively than we have in the past, even for small periods of time. In the great recession, I think the biggest boost we gave UI benefits was about $50. In the COVID recession, we went to $600. You might think $600 is too high for the type of recession that we're talking about here, but we can surely do better than $50. So, I think those are some of the responses I would I would mention. Yeah, I don't even really have much to add to that. I agree with all of those things. I will just reemphasize the point that the COVID response was really like exemplary both in terms of the magnitude of that intervention at the scale of the problem and the timeliness how quickly we were able to turn things around and turn it around across the board. So, we talked a bit about uh disparities in the labor market, rising black unemployment. Now, that was one of the fastest recoveries for uh black unemployment as a result of a lot of those um interventions. Yep, I think all all that's right. And it would be amazing to have these be automatic and legislated permanently to respond to economic conditions, but that's not going to happen before the next recession. So, we're just going to rely again on ad hoc response from Congress. I will say we have we have very little time, but we also have a bunch some questions. Um, and this one in particular pitched to you all. I think I know the answer, but I want to hear your take on it. It's, uh, what's driving real earnings down in recent months? Uh, inflation is up, but why are average hourly earnings growing more anemically? And is there anything to say about the decline in labor share of income in this story about declining wages? And so, whichever one of you want to jump on that first. >> So, I tried to answer that a little bit in my remarks about why was nominal wage growth slowing. And I think it's, you know, it's a weaker economy. Uh workers don't have the leverage uh to bargain for for wage increases and they're not leaving their jobs uh to try to get at those those either. I think that um that in in some also means that the labor share of income uh the the the share of income going to labor has remained low and we haven't really seen that that climb as well. So, I think that that kind of leverage uh on an individual or economywide level is just simply not there for workers and that's why we're seeing that uh deceleration. >> And I might just answer one last one pretty quickly because we have one more minute. I'm going to squeeze as much as possible. Um someone does say how do how do the recommendations to use fiscal policy aggressively to you know fight an economic downturn. How do they work given large deficits and debt uh in the economy? And I would just say quickly one like I do think deficits are too large right now given that unemployment at the moment is low. Um that to me just says clearly we need more revenue. If you look at sort of the trajectory of revenue and spending over the past quarter century, you see it as revenue that has really been throttled and has been a key driver of deficits outside of sort of the big recessionary crisis responses. And then two, kind of regardless of the level of deficits and debt, whenever the economy falls deeply below full employment, then fiscal policy that is aggressive and uses deficit finance to do needed relief, it just doesn't do any damage. like the damage done by big deficits tends to be high interest rates that crowd out private investment. Interest rates collapse because demand collapses during those times when we need those fiscal interventions. So I think in the long run we want to think hard about more revenue to keep deficits in check, but it is never an excuse to not fight sort of an immediate or sort of downturn when that happens. I feel like we're mostly out of time unless you all want to sneak in one of your uh last questions or answers. >> No, I think this is a good place to wrap it up. >> Excellent. Well, thank you too so much and thank you so much to everybody who's come out to watch us today. I hope any of it was useful and we'll be sharing lots of material going forward.