Claims Fall to 202k, But Continuing Claims Jump 25k: April 2, 2026 Release Leans on Smoothing

The official press release dated April 2, 2026 touts fewer layoffs—seasonally adjusted initial claims fell to 202,000 for the week ending March 28 (down 9,000). But beneath that headline, continuing claims (insured unemployment, SA) rose by 25,000 to 1,841,000 for the week ending March 21, and a cluster of quiet revisions burnishes the “improving” narrative more than the labor market itself.

Here’s what the data actually shows:

• Initial claims (SA) fell to 202,000; prior week revised up by 1,000 (from 210,000 to 211,000), turning a -8,000 move into -9,000.

• The 4-week average of initial claims (SA) dipped 3,000 to 207,750; prior week revised up by 250.

• Insured unemployment (SA) rose by 25,000 to 1,841,000; prior week revised down 3,000 (from 1,819,000 to 1,816,000), amplifying the week-over-week increase.

• The 4-week average of insured unemployment (SA) fell 7,500 to 1,838,750, flagged as the “lowest since September 28, 2024” (1,834,750)—even as the weekly level actually rose and was lower two weeks ago (1,816,000 on March 14).

• Unadjusted initial claims fell by 1,932 to 184,845 when seasonal factors had expected a +6,972 increase—SA improvement is largely a seasonal-factor artifact.

• Unadjusted insured unemployment fell 18,001 to 2,015,936, short of a -45,400 seasonal expectation; that shortfall produced the +25,000 rise in SA continuing claims.

Revisions Add Shine, Not Strength

The release’s clean headline—“initial claims down 9,000”—owes as much to revisions as to momentum. A +1,000 bump to last week’s initial claims and a +250 nudge to the 4-week average make progress look fractionally better. Meanwhile, continuing claims tell a different story: a -3,000 revision to the prior week makes the current +25,000 uptick to 1,841,000 look larger. The insured unemployment rate (SA) stayed at 1.2%, but the level rose and was already lower two weeks earlier (1,816,000).

This is the essential contradiction: separations remain low, yet re-employment appears to be getting slower at the margin.

Seasonals Did the Heavy Lifting

The mechanics of seasonal adjustment did more work than the labor market did this week.

• Initial claims, not seasonally adjusted, fell by 1,932 to 184,845 when the seasonal model expected a rise of roughly +6,972. That gap converts into the headline SA drop.

• Unadjusted insured unemployment fell by 18,001 to 2,015,936, but fell less than the seasonal pattern anticipated (-45,400). The model interprets that as weakness, hence the +25,000 increase in SA continuing claims.

If you only read the SA headline on initial claims, you miss that continuing claims moved higher precisely because the NSA decline underperformed the typical March pattern.

State-Level Stress That Headlines Ignore

National aggregates flatten out real dispersion:

• Highest insured unemployment rates (SA, week ending March 14): Rhode Island 2.8%, Massachusetts 2.7%, New Jersey 2.7%, Washington 2.4%, Minnesota 2.3%, California 2.2%, Illinois 2.0%, New York 2.0%, Montana 1.9%, Oregon 1.9%, Connecticut 1.8%, Michigan 1.8%.

• Largest prior-week initial claims increases (NSA, week ending March 21): Michigan +2,803; Iowa +730; Hawaii +572; Illinois +386; Georgia +374—offset by declines in Kentucky (-3,498) and Ohio (-1,208).

• Latest state swings (NSA): Initial claims rose in Texas (+1,800), Oregon (+1,487), New York (+1,386), Wisconsin (+821), Illinois (+727); fell in Michigan (-2,749), Georgia (-1,218), Pennsylvania (-641), Florida (-462), California (-454).

• Insured unemployment increased in Washington (+3,963), Oregon (+1,895), Texas (+1,715), Kentucky (+1,871), Wisconsin (+1,774); declined in Florida (-4,217), Pennsylvania (-3,751), California (-3,859), New Jersey (-3,275), Ohio (-2,572), Georgia (-2,912).

For a “stable/strong” labor market, those regional pockets of strain look notably unstable.

Trendlines, Not Taglines

Early-2026: Cooler Separations, Stickier Unemployment

• Initial claims (SA) oscillated in a 201–230k range. The 4-week average peaked into February (e.g., 220.25 on Feb 7) and eased to 207,750 by March 28 as those higher prints rolled off.

• Insured unemployment (SA) climbed into late February (1,871,000 on Feb 21), fell into mid-March (1,816,000 on Mar 14), then rebounded to 1,841,000 on Mar 21. The insured unemployment rate (SA) has held at 1.2% throughout 2026 in the table.

Translation: layoffs remain low, but finding the next job is taking a bit longer.

Year-over-Year: Better, But With Friction

• Initial claims (NSA) are 184,845 vs 200,081 in the comparable 2025 week.

• Insured unemployment (NSA) is 2,015,936 vs 2,058,751 a year earlier.

• The unadjusted insured unemployment rate is 1.3%, down from 1.4% last year.

Structurally, the market is tighter than mid-2025, when initial claims often ran 236–259k, the 4-week average peaked around 242.25 (June 2025), and insured unemployment hovered 1.94–1.96 million with a 1.3% insured rate. But the latest week-to-week rise in continuing claims shows cyclical friction hasn’t vanished.

The Smoothing Effect in Black and White

The “lowest since September 28, 2024” claim for the 4-week average looks less compelling when the current weekly level just rose and was already lower on March 14 (1,816,000)—and even on January 17 (1,823,000). This is storytelling by smoothing, not a contemporaneous signal.

What This Means for Markets

• Rates: The mix—lower initial claims, higher continuing claims—leans “slow re-employment” rather than “new firing wave.” That supports a patient policy stance. With the insured unemployment rate steady at 1.2%, there’s no urgent easing impulse, but the rise in continuing claims nudges the balance toward gradual cooling. Duration adds on weakness still make sense; be cautious chasing front-end rallies tied to a single SA initial-claims print.

• Equities: Quality still outperforms if hiring cools at the margin. Watch exposures to states with elevated insured unemployment rates—Rhode Island, Massachusetts, New Jersey, Washington, Minnesota, California, Illinois, New York—where staffing, temp help, and consumer cyclicals may lag. Logistics and transportation face persistent cost friction when re-employment lengthens.

• Credit: Rising continuing claims can precede higher delinquencies in lower-tier borrowers. Favor BB over CCC risk; keep dry powder for spread widening. In munis, state-level labor dispersion argues for selective exposure rather than broad beta—particularly in higher-IUR states—though fundamentals remain state-specific.

• Labor-sensitive sectors: Temp staffing, entry-level retail, and hospitality are most exposed to longer jobless spells; staples and high free-cash-flow compounders remain defensive beneficiaries.

• Positioning and hedges: Maintain a modest duration bias in the 5–10y sector; pair with selective cyclicals where state-level claims are improving (California, New Jersey, Pennsylvania, Florida, Ohio). Consider downside hedges in consumer credit proxies if continuing claims keep drifting higher.

Looking Ahead: Data to Prove the Story

• Watch whether continuing claims (SA) retreat back below 1.83–1.84 million in coming weeks. A sustained move higher would validate the “slower re-employment” thesis.

• Track NSA-to-seasonal gaps through April. If initial claims keep “beating” because seasonals expect rises that don’t show, take SA headlines with caution.

• Monitor state prints: continued increases in Washington, Oregon, Texas, Kentucky, and Wisconsin would signal that the softening is spreading beyond isolated sectors.

• Cross-check with payrolls and wage growth. If wage gains cool while continuing claims rise, margins stabilize for labor-intensive sectors but top-line growth may soften.

The April 2 release celebrates a tidy 202,000 initial-claims number, then quietly papers over a +25,000 rise in continuing claims and the fact that smoothing—not the contemporaneous data—delivered the “lowest since 2024” line. For investors, the message is straightforward: prize the internals over the headline. Favor quality balance sheets, be selective in cyclicals with state exposure, keep a measured duration tilt, and let continuing claims—not a single smoothed average—set your next move.