Zoom Out: Keeping the Latest Jobs Numbers in Perspective

June 3, 2021

When we zoom out from the monthly jobs numbers, there are good reasons to be confident that employment growth will pick up.

Tomorrow, the Bureau of Labor Statistics (BLS) will release May’s Employment Situation, which includes the closely watched change in total employment for the month. As of today, no one knows what those numbers will look like. But one thing we can be sure of: The people who think American workers have it too easy will see the new numbers as confirmation of their views. If the jobs numbers are low, we’ll be told it’s time to cut back on unemployment insurance—it’s limiting labor supply. And if the jobs numbers are high? Also time to cut back on unemployment insurance—it’s no longer needed.

Suppose, though, that you don’t believe working Americans are slackers who need the threat of poverty to keep their noses to the grindstone, or that you think rising wages are an opportunity rather than a crisis. How should you interpret tomorrow’s jobs numbers in that case?

In this post, we suggest three heuristics for thinking about the jobs numbers, for people who are not already sure that austerity is the answer.

  • First, the numbers are noisy. Don’t put much weight on one month’s—or even a couple months’—numbers.
  • Second, look at the big picture. Whatever the May employment numbers look like, there are good reasons to think we’ll experience strong growth over the next several years.
  • Third, don’t get taken in by labor-supply hysteria. Rising wages should be welcomed, not seen as a problem to be solved.

Wait a Minute, or a Month

Remember: These jobs numbers are an estimate—an extrapolation from a sample rather than a straightforward count of the number of US jobs last month. While the BLS statisticians are, in general, very good at their jobs, there is no way to ensure that the sample is representative, especially given the delays with which many businesses submit payroll information. The numbers are always revised as additional data comes in; these revisions tend to be larger at business-cycle turning points and have been exceptionally large during the pandemic.

So far this year, revisions to the monthly jobs numbers are running two to three times higher than their historical averages. For the first two months of the year, the initial reports had the country gaining a total of 428,000 jobs on a seasonally adjusted basis. But the revised numbers, released two months later, showed total job gains of 769,000—nearly twice as high.

This isn’t just an interesting statistical exercise. The popular political narrative created within the hour of the jobs report is vastly different than the more accurate numbers reported later. The original numbers for February and March in 2021 were 379,000 and 916,000, respectively. This created a narrative that job growth would keep accelerating, setting expectations for over a million jobs in April. But the revised numbers show more similar rates of employment growth in February and March. If we had had access to the revised numbers at the time, there would have been no reason to expect a big surge in employment in April. While the initial April jobs number would still be disappointing, in the context of the revised February and March numbers it would have been less of a shock—a bump in a process of steady growth rather than a major reversal.


The Trend is Our Friend

The tendency of reporters and analysts to create immediate narratives after each individual jobs report, without stepping back and observing longer trends, tends to cause overreactions, which are only amplified by social media. This was a regular problem during the Great Recession and its aftermath, when moment-by-moment narratives distracted from the larger picture of a historically slow recovery. The persistent demand shortfall of the 2010s and lack of catch-up to the previous trend were only visible at a scale of years, not of months.

Similarly, when we zoom out from the monthly jobs numbers, there are good reasons to be confident that employment growth will pick up.

The big picture over the past year is a rapid recovery of employment as pandemic-related restrictions have been lifted. This is very different from the last several recessions, which were followed by long “jobless recoveries,” with employment continuing to fall for months or even years after economic growth resumed. Following the dot-com crash, for example, total employment bottomed out in August 2003, a full two years after the end of the recession. This time, by contrast, employment is rising even before the recession is officially over. As of the last month, we are adding an average of 450,000 jobs a month since January of 2021, and 524,000 since February. Despite the month-to-month blips, this is a very different pattern than in previous downturns.

There are several reasons to think that this growth will continue.

First, the pandemic is still significantly limiting employment, especially in blue states that have been slower to fully reopen their economies. Many states maintain significant restrictions on in-person activities, and as these are lifted, it will take time for businesses and consumers to readjust. As economist Arindrajit Dube has observed, job growth in recent months is strongly associated with the extent of reopening. As of April, employment in eight states had already exceeded 2019 levels; with the exception of Oregon, these were all states that imposed few restrictions on businesses and in-person schooling and lifted them early. As coronavirus restrictions are lifted throughout the country—and as people become more comfortable with resuming in-person activities—we should see catch-up employment growth in the states where the pandemic impact was greater.

The second reason for optimism is that the laws of economics have not been suspended. Businesses hire people to produce goods for sale. Consumers make purchases based on their income and balance sheets. While supply constraints may occasionally limit employment in the short run, they also spur investment, which over time creates both new capacity and additional demand.

Thanks to the extraordinary stimulus of the past year, household finances are in better shape than in many years. Personal savings soared during the pandemic. Surveys suggest that about two-thirds of stimulus payments were saved or used to pay down debt; this dampened their initial effect on demand, but has left households in a stronger financial position. According to the New York Fed, the fraction of households facing bankruptcy fell sharply during the pandemic, and is now at its lowest level in decades. Debt service as a share of disposable income is at its lowest levelsince at least 1980. There are good reasons that households have focused on paying down debt over the past year, but going forward, we should expect that in the absence of some new shock, stronger household finances will translate into increased spending. And it is demand that drives employment; that has not changed.

Increased demand for final goods will get an additional boost from investment spending as businesses increase capacity. Already we are seeing stories about rising manufacturing production. As Jason Furman and Wilson Powell III found, the recovery in goods production is quite strong. The recovery in services will pick up with more reopenings, and as businesses scale up. This push-and-pull of an economic awakening is something we’ve never experienced before. As Cecilia Rouse, chair of the Council of Economic Advisers, told the New York Times, “We put the economy in a medically induced coma, and you don’t come out of a coma and run a marathon the next day.”

These longer-term tailwinds are easy to miss when the discussion is focused on the most recent numbers. But emergence from the pandemic, stronger household balance sheets, and capacity-boosting investment will all continue to boost employment growth over the next year, unless they are derailed by some new disaster or a radical policy shift.


What’s Really Holding Back Employment

Unfortunately, discussion of the economy has paid too little attention to the longer-term factors influencing demand, and too much attention to largely imaginary constraints on supply. Critics of today’s more active approach to macroeconomic policy have put people on unemployment insurance in their crosshairs. The belief that workers are refusing good jobs because of over-generous unemployment insurance has led half of US states, all of them headed by Republican governors, to roll back the recent expansion of the program. As such, a bad narrative has distorted not just the analysis of the recovery but the actual way the recovery is unfolding, as billions of dollars meant to stabilize income and speed up the recovery are being eliminated.

The idea that “labor supply” is the cause of the recent job misses has run far ahead of the evidence. Take the inability of restaurant owners to find workers, a central anecdote in coverage of the recovery so far. Unfortunately for that story, the latest jobs numbers for restaurant and hospitality workers were exactly as predicted between 241,000 and 331,000 jobs, depending on how you measure. Not only was that number solid, it was actually an increase in rate over the prior month. It was other areas (especially residential construction) where employment growth fell short of expectations. So whatever is going on, it has nothing to do with the stories about restaurants that have dominated the news.

Second, as Skanda Amarnath of Employ America has shown, the April job misses are not at the low end of the income distribution, which is where we’d expect them to be if a flat-sum increase in unemployment insurance (UI) were responsible. The missing jobs are in the middle of the income distribution, jobs paying between $18.50 to $31.50 an hour. Earnings in this range are well above the maximum available from UI.

More broadly, claims that UI or other labor supply restrictions are directly limiting employment fundamentally misunderstand the way the labor market works in our economy. Businesses don’t purchase labor for its own sake, but in order to carry out production. And in an economy like ours, how much gets produced depends on how much can be sold. (There’s a reason that it is sellers and not buyers who pay for advertising; businesses almost always worry about finding more customers, but people with money very rarely worry about whether they can find anything to spend it on.) And with millions of people unemployed and millions more not officially counted as unemployed but nonetheless available for work, there is obviously not a labor shortage in any absolute sense. Where labor supply matters is not the amount of employment but the terms on which it is offered—wages most obviously, but also things like training and the types of people considered for various jobs.

Critics who focus on wages, rather than employment, as evidence for labor supply constraints are on stronger ground in terms of logic, but they are still wrong on the data and on policy. It is true, for instance, that wage growth has picked up in bars and restaurants, which at first glance is consistent with limited labor supply. But as Josh Bivens of the Economic Policy Institute points out, “since December 2020, the rise in tip income, not an increase in base wages, can likely entirely explain the acceleration of wages for production and nonsupervisory workers in restaurants and bars.” Full-service restaurants and bars workers account for 70 percent of the rise in wages in this sector, and restaurants are finally filling back up with customers who are tipping. Like many things, the unique circumstances of this pandemic and recovery are causing optical illusions in the data, ones that require careful unpacking rather than politically convenient prepackaged narratives.

Still, it is true that wage growth seems to have picked up in recent months. Much if not all of this uptick is compositional, reflecting the fact that lower-wage jobs were the hardest hit by the pandemic. Still, it is plausible that expanded unemployment insurance has contributed to faster wage growth than we would have seen otherwise. From a policy standpoint, it’s not clear why this is a problem. Higher wages, after all, are a good thing. They only become a problem when they are passed through to higher inflation, but there is no reason to think that is an issue right now. The modest uptick in inflation we are seeing is about supply chains, base effects, and a one-time burst of businesses reopening. And going forward, we may soon find, once again, that the problem is inflation that is too low, not too high. The Cleveland Fed’s widely used estimate of inflation expectations—based on surveys and the yield of inflation-indexed bonds—shows expected inflation over the next decade of just 1.57 percent, well below the Fed’s 2 percent target.

More generally there are distributional consequences of trying to prevent this wage growth. If faster or slower wage growth were always passed on one-for-one to inflation, the fraction of national income going to wages would never change. But historically, this is clearly not the case. In the decades after World War II, labor compensation accounted for a stable 63 to 64 percent of national income. But this share fell sharply in the 1980s, recovered in the late 1990s, and then fell even more sharply in the 2000s, to less than 60 percent today.

What this means is that slower wage growth has not translated into lower prices, but into a smaller share of the pie going to workers. If wages are not allowed to increase faster than productivity in a recovery, it will be impossible to reverse this decline. The assumption that labor markets cannot get tight enough to let wages grow faster than productivity is logically equivalent to saying that the labor share of income can decline but can never increase. Given the bad trends on this front over the past generation, why would we make locking in a lower labor share a goal of public policy? Stated this baldly, it sounds ludicrous. But it is just the logical corollary of claims that faster wage growth is “unsustainable” or a sign of labor supply constraints that need to be removed.

On that front, it is encouraging to see President Biden say that strong wage growth is an important goal of the recovery. In a recent speech he noted, “When it comes to the economy we’re building, rising wages aren’t a bug; they’re a feature” and that this “kind of competition in the market doesn’t just give workers more ability to earn a higher wage; it gives them the power to demand to be treated with dignity and respect in the workplace.” This is exactly what our country needs. The good news is that we’re on the path to getting it.

The biggest danger right now is not a month or two of surprisingly low (or high, for that matter) employment growth, but a panicked overreaction that pulls away support for the recovery—by, for instance, cutting unemployment benefits—before it is fully underway.


J.W. Mason is a fellow and economist with the Roosevelt Institute’s Macroeconomic Analysis team.

Mike Konczal is the Director of Roosevelt’s Macroeconomic Analysis and Progressive Thought teams.