(Click on image to enlarge)

ADP Change in employment month-over-month by employer size

Please consider the ADP^® National Employment Report for January 2026.

Private employers added 22,000 jobs in January.

Job creation took a step back in 2025, with private employers adding 398,000 jobs, down from 771,000 in 2024. While we’ve seen a continuous and dramatic slowdown in job creation for the past three years, wage growth has remained stable.

Month-Over-Month Change by Employer Size

• 1-19: +30,000
• 20-49: -30,000
• 50-249: +37,000
• 250-499: +4,000
• 500+: -18,000

Year-Over-Year Change by Employer Size

(Click on image to enlarge)

Year-Over-Year Change Details

• 1-19: +513,000
• 20-49: -160,000
• 50-249: -252,000
• 250-499: +40,000
• 500+: +241,000

Key Weakness Points

• Businesses with 20-49 employees have negative year-over-year growth for 20 straight months and 24 out of the last 25 months.
• Businesses with 50-249 employees have negative year-over-year growth for 13 straight months.
• Businesses with 250-499 employees have gone nowhere (-26,000 to +60,000 year-over-year) for 28 straight months. The average is +13,000 year-over-year, about 1,000 a month.

This is certainly not a strong labor market.

The weakness is where one would expect, small- to mid-sized businesses clobbered by tariffs.

AI is going to start taking a bite out of large employers.

Related Posts

January 27, 2026: Trump Cheers a Plunge of the US Dollar “I Think It’s Great”

“Look at all the business we are doing,” says Trump.

January 31, 2026: Competing Claims, How Much Labor Market Weakness Is There?

Some see “No Signs of Labor Market Weakness”. Others strongly disagree.

For a look at what health care will do to the PCE price index, please see Expect a Big Divergence This Year Between CPI and PCE Inflation

Rent and Healthcare go different ways in 2026. Plus there are huge timing issues.

February 2, 2026: The Fed Has Two Huge Problems Starting Now, Acyclical Inflation and Jobs

The Fed is not in a good spot.

More By This Author: