The economy’s smoke signals are sending a confusing message about its health.
Growth slowed by 0.3 percent in the first three months of 2025. Yet in April, the economy produced a robust 177,000 jobs. Unemployment is steady, but in the first quarter, labor productivity contracted for the first time in three years. The pace of inflation, though slower than it was, is stickier than economists would like.
When the data is mixed, like it is now, economists look to faster-moving signals to gauge the economy’s health. During these times, one closely watched indicator is initial jobless claims.
The initial jobless claims report from the Department of Labor has several attributes that are useful to market watchers. For starters, the data is released weekly, much faster than metrics that are released monthly or quarterly. Initial claims also are based on hard administrative data, as opposed to surveys, and are correlated with other data economists care about, such as GDP.
What are initial jobless claims?
Every week, the Department of Labor reports the number of people who applied for unemployment insurance the previous week.
Because people who voluntarily leave their jobs are ineligible for unemployment benefits, initial claims data is an indicator of layoffs, firings, and other involuntary separations.
Along with initial claims, the department reports the number of people in the United States who are continuing to receive unemployment insurance benefits. Known as continuing claims, this data tells us how long people are remaining unemployed, which can signal the relative ease or difficulty in finding work.
Initial and continuing claims swing back and forth a lot and can be affected by factors such as weather events and holidays. Because week-to-week changes in the data can be volatile, economists typically look at the four-week moving average of the initial claims data.
How high is too high?
When economic data isn’t showing a clear picture, economists and investors pay close attention to initial jobless claims for signs that businesses are starting to shed workers. That could mean that demand for their product or services is weakening.
During the financial crisis, initial jobless claims doubled between 2007 and 2009, to well above 600,000. During the deep and sudden pandemic recession in April 2020, initial claims surged to more than 6 million, by far the highest on record.
Going back to 1967, initial claims have averaged 363,000 a week. In the last two years, these weekly reports have hovered around all-time lows of about 225,000. Even when claims have drifted above this level, they’ve been followed by a quick decrease.
A rule of thumb—one substantiated by academic research—is that when initial jobless claims exceed 400,000, it’s a signal that labor market conditions are deteriorating and employers are shedding workers at levels that aren’t healthy for the economy.
Can initial claims predict the future?
Initial jobless claims are useful high-frequency indicators, and they tend to lead other metrics that economists track. But they don’t perfectly predict future unemployment rates. In fact, initial claims and the unemployment rate move in opposite directions about as often as they move in synch.
Unlike initial claims, which are based on state data, the U.S. unemployment rate is based on a monthly survey of households conducted during the week that includes the 12th day of the month.
Preceding this week, initial claims are just as likely to move in the same direction as unemployment as they are to move in the opposite direction.
That means the direction of initial claims one way or the other isn’t that helpful for forecasting unemployment. Rather, it’s the overall level of initial claims over several weeks and months that economists care about.
My take
Last week, the Federal Reserve held their short-term interest rates at current levels. Though Fed policymakers pointed out a number of economic positives in the committee’s official statement, including the steady labor market and solid economic activity, they also highlighted the elevated risks of high inflation, high unemployment, and high uncertainty.
Initial jobless claims data have yet to reflect these elevated risks, which suggests that employers are hanging onto workers much more often than they’re laying them off.
Let’s see if the trend holds. ADP employment data, which also is collected weekly, showed evidence of a hiring slowdown in April.
While the smoke signals coming from the labor markets most real-time indicator isn’t conclusive, it’s communicating that the labor market, so far, isn’t in danger of a recession.
The week ahead
Tuesday: In a busy data week, the main event arrives early: a read on consumer prices from the Bureau of Labor Statistics. We should learn whether tariffs are contributing to inflation in consumer goods and services.
Wednesday: Today will deliver three snapshots of the economy. Retail sales data from the Census Bureau will show how consumer spending is holding up. And we’ll get a look at business costs and input prices when the Bureau of Labor Statistics releases the Producer Price Index.
Thursday: Any ripple in initial jobless claims could grab attention.
Friday: Rounding out the week, Census data on housing starts and permits will be an important indicator of the housing market’s ability keep up with growing demand and contain the sector’s contribution to inflation. But the bigger number for inflation watchers on Friday will be the University of Michigan’s read on consumer sentiment. Consumer expectations of higher prices can be self-fulfilling.
