The U.S. economy is in a weird place right now on multiple fronts. One immediate problem is that policymakers are flying somewhat blind because the government shutdown has delayed the September employment report. According to the last report available (which was for August), unemployment is relatively low by historical standards. But another source of weirdness is that many people feel very bad about the economy: Consumer sentiment is much weaker than it was pre-Covid, in fact comparable to its level at the depths of the 2008-2009 financial crisis.
So are we in what Phil Gramm — remember him? — once called a “mental recession,” a sort of mass delusion that the economy is bad? It’s likely that some of Americans’ sour mood is driven by political unease. Huge and ever-changing tariffs, masked agents grabbing people off the street, assassinations, vindictive prosecutions, rising measles cases, Trump’s false claims that cities are “war zones” as pretext for sending in the National Guard, and more. Increasingly unhinged statements from the administration feed a general sense of destructive instability. Next thing you know they’ll start demolishing the White House itself to make room for some vanity project. Oh, wait.
Yet it’s not only about political unease. There are some objective, measurable reasons to say that the US economy, which appears OK by the most commonly used measures, is definitely not OK once you look under the hood. One essential aspect of this weirdness is the economy is strongly bifurcated: AI is booming, but the rest of the economy isn’t. Another aspect is that in many ways the economy feels “frozen”: while there have been no mass layoffs so far, people who have lost their jobs or are just entering the work force are finding it very hard to get new jobs. Third, while the economy is growing thanks to AI spending, it’s a K-shaped expansion: People who were already affluent are becoming more so, but the less well-off are under severe pressure. For example, there are clear signs that middle-to-low income consumers are struggling: car loan and credit card delinquencies are rising, and grocers report that shoppers are buying cheaper varieties of food. At the same time, the affluent are spending freely: the top 10% of the income distribution now accounts for nearly half of all consumer spending.
What’s going on? I would argue that Trump’s wildly erratic policies are creating huge uncertainty which is deterring many companies – essentially those that are not in the AI sector or a sector catering to the affluent – from making investments. And those forgone investments include hiring new workers. The result is that much of the economy is frozen — companies aren’t hiring or investing. This freeze, in turn, explains both worker anxiety and rising inequality. Without the AI boom/bubble spending, we might very well have fallen into a recession, as some economists like Mark Zandi have claimed. And despite the AI boom, times for many workers are tough.
Let’s look at the data to understand why what looks on the surface like a fairly benign economy is actually hurting workers.
First, as I said, we haven’t (yet?) seen mass layoffs — except from the federal government! — but the rate at which businesses are hiring is very low by historical standards, not far above its level during the 2008-2009 financial crisis:
These data aren’t currently being updated because of the government shutdown, but we get a more up-to-date picture from private sources, such as job postings from Indeed.com, and they suggest that the picture is if anything getting worse:
Another source of information comes from surveys that ask people about the state of the job market. In particular, the influential monthly consumer survey conducted by the Conference Board — second in influence only to the Michigan Survey — asks people whether jobs are “plentiful” or “hard to get.”
The spread between these numbers is always positive — we are an inherently optimistic nation — but varies a lot and is a good indicator of how people feel about the labor market. In late 2019, on the eve of Covid, almost half of respondents said that jobs were plentiful, versus around 10 percent saying hard to get — a spread of around 40 points. In the latest Conference Board survey the spread was only 8 points, 27 versus 19. This tells us that American workers are very worried that if they should happen to lose their job, they’ll have a hard time finding another.
And they’re right. Overall unemployment hasn’t risen that much, but the number of long-term unemployed — would-be workers who have been jobless for more than 6 months — had soared as of August, and has probably continued to rise since then:
Another important indicator of a troubled labor market is Black unemployment. After all these years, Black workers still tend to be “last hired, first fired.” And while the overall unemployment rate (dashed green line) hasn’t risen much so far, the Black unemployment rate (blue line) has soared, presumably because Black workers are finding it especially hard to find jobs in this frozen economy:
Again, we have yet to see mass layoffs, so most workers still have their jobs. But workers believe, rightly, that if they should happen to lose their current job they will have a hard time finding another. This obviously means that workers have much less bargaining power than they did when the job market was tight. Employers don’t have to give workers big wage increases to hang on to them; they can impose onerous conditions, like ending remote work, without fearing that employees will quit, because they have no place to go.
Historically, strong demand for labor has been especially good for lower-paid workers, while weak demand has hit them hard. The post-Covid expansion, during which labor was scarce, was marked by big gains at the bottom and a surprisingly large fall in wage inequality, what David Autor, Arindrajit Dube and Annie McGrew have called the “unexpected compression.”
Incidentally, all through the Biden-era expansion I kept hearing people say that the economic recovery was only benefiting an affluent minority, that ordinary workers were being left behind. This wasn’t at all true at the time. But it is true now. The Atlanta Fed has a wage tracker that, among other things, estimates the rate of wage growth at different parts of the wage distribution. During the Biden years wage growth for the bottom fourth of the wage distribution (blue line) was consistently higher than wage growth for the top fourth (red line). Now that equalizing process has gone into reverse:
And then there’s the stock market. Investors seem to have decided that the wonders of AI matter more than Trump’s tariff madness, so we’re seeing a stock market surge dominated by technology companies. Aside from the question of whether this is a bubble, it’s important to be aware that the top 10 percent of households own 87 percent of equities, while the bottom half own almost no stock at all and gain nothing from a rising market. Hence the following chart, highlighted by the Michigan Consumer Survey:
Many economists — actually, all the economists I know — are worried about a potential downturn. The AI boom is troublingly reminiscent of the 90s tech bubble. After the sudden bankruptcies first of a subprime auto lender, then an auto parts supplier built on hidden loans, JPMorgan’s Jamie Dimon suggested parallels between bad lending in the private credit market and the bad subprime lending that brought on the 2008 crisis. To quote Dimon: “I probably shouldn’t say this, but when you see one cockroach, there are probably more.”
But I’ll try to assess these concerns another day. My point for now is that even though we haven’t had a recession yet, the frozen state of the U.S. economy has already made life much worse for many workers.
MUSICAL CODA