The United States economy appears to be approaching a recession, as recent jobless claims data—a reliable precursor to the end of a business cycle—portends.
The key is to measure the rate of growth in jobless claims rather than look at the overall weekly number.
Despite the labor market’s resilience and persistently low claims compared to historical norms, the steady increases in claims since late last year are ringing alarm bells.
Instead of focusing solely on the raw number of claims—which were higher in previous business cycles before the recent tightening in labor market conditions—we examine the pace and persistence of the increase. To do this, we construct new advanced measures for both initial and continuing claims.
First, we calculate the three-month moving average for each series to strip away the weekly noise. Next, we determine the percentage difference between the three-month moving average and its lowest point in the previous 12-month period.
Our new measures reveal clear signs of an impending recession.
Initial claims in June so far are 21.1% above their 12-month low, while continuing claims stand at 35.4%.
Historically, the economy has either been months into a recession or close to one whenever percentages like these are recorded.
This is happening even though the weekly claims numbers are well below those that have signaled recessions in the past.
Is this time different?
The pandemic and the ensuing recovery have changed the equation. With solid spending and robust job growth, breaking the historical thresholds is by itself not enough to signal a recession.
June is the first month that the initial claims surpassed the key 18.1% level that was seen in 1984, the highest level that was seen while avoiding a recession since the claims data began to be collected in the 1960s.
So, what level of initial claims signifies a recession? Our estimate is 250,000, much lower than the 300,000 or more that signaled downturns in the past.
If new claims maintain a level of 250,000 per month for the remainder of the year, our new measure will continue to increase, peaking at around 24% between August and September.
Before June, initial jobless claims were averaging around 230,000, well below the threshold. So far this month, claims have increased, and they are likely to surpass 250,000 this year as the Fed raises interest rates and consumers continue to deplete their excess savings—an important driver of the economy recently.
Such an increase in jobless claims would be consistent with the onset of a recession in the second half of the year, albeit a mild one.
The key level for claims is now much lower than what it was in previous cycles—when it routinely surpassed 300,000—except for the outlier that was the pandemic recession in 2020.
The most obvious reason for the reduction in jobless claims is the structural shift within the labor force because of demographic changes as well as the lasting impact unleashed by the pandemic.
Because of those factors, the natural rate of unemployment has fallen from above 5% before the pandemic to around 4% now.
Outlook for the rest of 2023
Jobless claims data is not one of the explicit criteria that the National Bureau of Economic Research uses to identify a recession.
But given how the labor market remains one of the last bulwarks against a recession, there is a strong case for jobless claims to be considered as a key leading indicator for a recession this time.
What has been keeping the labor market running at an impressive pace has been a robust demand for workers, who are in short supply in many sectors.
What has been keeping the labor market running at an impressive pace has been a robust demand for workers, who are in short supply in many sectors.
There remains an imbalance between demand and supply within the labor market; however, such an imbalance cannot help those who stay on jobless benefits to find a job, as the growth in continuing claims shows.
Our results suggest a different timing for the onset of the recession compared to the Federal Reserve’s new economic projections, which also imply a mild downturn but at a later date, around early 2024.
If incoming data on jobless claims, inflation, labor and spending supports our prediction over the Fed’s, we believe July would most likely mark the final rate hike. Our base case forecasts further signs of economic deterioration around the start of the fourth quarter, with a sufficient impact on inflation to deter the Fed from hiking one more time.
Simultaneously, we expect the underlying personal consumption expenditures index—a key measure of inflation for the Fed—to fall to between 3% and 3.5% by the end of the year, which would still be high enough to prevent the Fed from cutting rates this year.
Of course, there are risks associated with our outlook that we cannot ignore. But the risks are greater for the timing of our predictions than for the final outcomes.
The wildcards include service spending during the summer, as well as goods spending during the back-to-school and early holiday seasons.
Spending that exceeds expectations would preserve job gains and spending levels while keeping inflation elevated, thereby pushing the recession date further down the line and increasing the probability of an additional rate hike after July.
After all, it’s always risky to bet against the resilience of the American consumer.
