How Jobless Claims Data Tells You What's Happening in the Labor Market
What jobless claims numbers measure
Jobless claims are the number of people filing for unemployment insurance in a given week. The U.S. Department of Labor publishes two main figures every Thursday: initial claims (people filing for the first time) and continuing claims (people already receiving benefits). These numbers are not the same as the unemployment rate—they count only people actively filing with a state unemployment office, not all unemployed people.
Initial claims spike when layoffs happen. Continuing claims stay elevated when people remain out of work for weeks or months. Together, they form a real-time picture of labor market stress. A single week's number means little; economists watch the trend over four to eight weeks to see whether layoffs are accelerating, stabilizing, or slowing.
The data comes from state unemployment insurance programs, which means the numbers reflect only people who have filed and remain may be able to access. Self-employed workers, gig workers, and people who have exhausted their benefits do not appear in these counts, even though they are unemployed.
Key Takeaways
- Initial claims measure new filings each week; continuing claims measure people still receiving benefits, and the two tell different stories about labor market health.
- The Department of Labor releases both figures every Thursday morning, with data from the prior week, and reports them seasonally adjusted and unadjusted.
- A single week's number can be noisy due to holidays or processing delays, so economists look at the four-week moving average to spot real trends.
- Rising initial claims often signal the start of layoffs; rising continuing claims signal that people are staying out of work longer or that benefits are being extended.
- Jobless claims data does not include self-employed workers, gig workers, or people who have run out of benefits, so it undercounts total unemployment.
Where to find the official numbers
The U.S. Department of Labor Employment and Training Administration (ETA) publishes jobless claims data every Thursday at 8:30 a.m. Eastern time on its website at oui.doleta.gov. The release includes initial claims, continuing claims, and insured unemployment rates for the nation and each state.
The same data appears on the Federal Reserve's economic database (FRED) at fred.stlouisfed.org, where you can read historical series going back decades and create your own charts. The Bureau of Labor Statistics also republishes the figures on its website.
News outlets and financial data services report the numbers within minutes of release, often with commentary on what the trend means. If you want the raw data without interpretation, the Department of Labor site is the source; if you want context, major financial news sites publish analysis within an hour.
Initial claims versus continuing claims
Initial claims count people filing for unemployment benefits for the first time in a given week. A spike in initial claims usually means employers have just laid off workers or that a seasonal industry (construction, retail, agriculture) has entered its slow season. Initial claims are sensitive to current events—a major company closure or a weather event that shuts down work will show up when ready.
Continuing claims count people who filed in a previous week and are still receiving benefits. This number rises more slowly than initial claims because it reflects the cumulative effect of layoffs over time. If initial claims stay high for several weeks, continuing claims will eventually rise. If initial claims fall but continuing claims stay high, it means people are taking longer to find work or that benefits have been extended.
During recessions, continuing claims often peak weeks or months after initial claims peak, because people exhaust their regular benefits and move into extended benefit programs. This lag is why continuing claims are sometimes called a lagging indicator—they tell you about the depth of a downturn after it has already started.
Seasonal adjustment and why it matters
The Department of Labor reports jobless claims in two ways: seasonally adjusted and unadjusted. Seasonal adjustment removes the predictable patterns that happen every year—retail hiring before the holidays, construction layoffs in winter, school hiring in summer. The adjusted figures are what economists and the Federal Reserve watch, because they filter out noise and show real changes in hiring and firing.
Unadjusted numbers are the raw count. They spike every January when holiday retail workers are laid off, and they fall every June when schools hire summer staff. If you see a headline saying claims jumped 50,000 in one week, check whether that is the seasonally adjusted figure (which would be significant) or the unadjusted figure (which might just be the normal January pattern).
The Department of Labor also publishes a four-week moving average of initial claims, which smooths out one-week noise caused by holidays, data processing delays, or unusual events. This average is often more reliable than a single week's number for spotting whether the trend is actually changing.
What rising and falling claims tell you
When initial claims rise for several weeks in a row, it usually signals that employers are beginning to cut payroll. A sustained rise—say, four weeks above the prior month's average—often precedes a recession or a significant slowdown in hiring. Conversely, when initial claims fall and stay low, employers are either not laying off workers or are hiring enough to offset departures.
Continuing claims rising while initial claims fall can mean that people are taking longer to find work, even though new layoffs have slowed. This pattern often appears in the recovery phase after a recession, when hiring picks up but long-term unemployed workers still outnumber new job openings. It can also mean that Congress has extended the duration of benefits, so people who would normally have exhausted their claims are still filing.
During the COVID-19 pandemic, both initial and continuing claims reached historic highs—initial claims exceeded 6 million in a single week in April 2020, and continuing claims topped 24 million. These figures were far outside historical norms and reflected the sudden, economy-wide shutdown. As the economy reopened, both figures fell, but the decline was uneven and varied by state, showing that recovery was not uniform.
State-level variation and what drives it
Jobless claims vary significantly by state because state unemployment insurance programs have different rules, benefit durations, and may be able to access thresholds. A worker in one state might be able to file when ready after a layoff, while a worker in another state might face a waiting period. Some states have higher wage replacement rates (the percentage of lost wages that benefits cover), which can affect how many people file.
Large states like California, Texas, and New York account for a large share of total claims straightforward because they have more workers. Smaller states show more volatility week to week because a single large layoff can move the needle more. The Department of Labor publishes state-by-state breakdowns, which can reveal regional economic patterns—for example, if claims spike in one state but fall nationally, it might signal a local industry shock rather than a national trend.
State data also lags national data by one week, so the Thursday release includes national figures for the most recent week but state figures for the week before. This timing difference is important if you are trying to match national and state trends.
How jobless claims fit into the broader economic picture
Jobless claims are one piece of labor market data, not the whole picture. The monthly employment report from the Bureau of Labor Statistics, released on the first Friday of each month, includes the unemployment rate, total job gains or losses, and hours worked. The unemployment rate is broader than jobless claims because it includes people who are not filing for benefits—people who have exhausted their claims, people who never filed, and people who are looking for work but have not yet applied.
Jobless claims move faster than the unemployment rate because they are reported weekly instead of monthly, and they reflect only people actively filing. This makes them useful for spotting turns in the labor market early. If initial claims start rising in week one, you will not see that reflected in the unemployment rate until the monthly report comes out weeks later.
The Federal Reserve watches jobless claims closely as one signal of labor market health, along with job openings data, wage growth, and labor force participation. A sustained rise in claims can prompt the Fed to reconsider interest rate policy, because it suggests the economy is weakening and unemployment may rise.
Frequently Asked Questions
Why do jobless claims spike in January?
Retail and hospitality workers hired for the holiday season are laid off after the holidays end. This is a predictable seasonal pattern, which is why the Department of Labor publishes seasonally adjusted figures that remove this effect. The unadjusted number will spike, but the adjusted number should stay relatively flat if the layoffs are on schedule.
Can I use jobless claims to predict a recession?
Rising initial claims over several weeks is often an early warning sign of economic slowdown, but it is not a may provide. A single week or even two weeks of higher claims can be noise. Economists typically look at the four-week moving average and compare it to the prior month's average. A sustained rise above historical norms is more predictive than a one-week spike.
Why are jobless claims different from the unemployment rate?
Jobless claims count only people actively filing for unemployment insurance benefits. The unemployment rate counts all people without a job who are looking for work, including people who have exhausted benefits, never filed, or are not may be able to access for insurance. The unemployment rate is broader but published less frequently (monthly instead of weekly).
Do gig workers and self-employed people show up in jobless claims?
Not in the regular state unemployment insurance system. Gig and self-employed workers can file for Pandemic Unemployment information (PUA) during declared emergencies, but that program is separate and not included in the standard jobless claims figures. In normal times, most gig workers do not appear in jobless claims data at all.
What does a four-week moving average tell you that a single week does not?
A single week can be affected by holidays, data processing delays, or one large layoff. The four-week moving average smooths out these one-time events and shows the underlying trend. If the average is rising, it suggests a real shift in employer behavior, not just noise.
This guide is for general information only and is not medical, financial, legal, or other professional advice. For decisions specific to your situation, consult a qualified professional. See our Editorial Policy.