Differences in employment news: ADP, NFP, JOLTS, and jobless claims
New jobs, unemployment rate, job openings, layoffs, and wages are not the same indicator. Read employment releases as a group.
Table of contents
Remember this first
First, 3 things to remember
- 1ADP is private data and is not a definitive forecast for NFP.
- 2The unemployment rate and employment growth use different survey subjects and methodologies.
- 3Connect wages, working hours, participation rates, and revisions.
What does each report measure?
Even if NFP is translated as "employment index" on a calendar, the frequently displayed figure is the net month-over-month change in non-farm payroll jobs. Since one person can hold multiple jobs, it must be distinguished from the number of people.
| Announcement | Key points |
|---|---|
| ADP | Private employment change based on private payroll data |
| BLS NFP | Non-farm payroll employment change from the establishment survey |
| Unemployment rate | The percentage of unemployed people within the labor force from the household survey |
| JOLTS | Labor turnover, including job openings, hires, quits, and layoffs |
| Initial jobless claims | Initial jobless claims, weekly change |
| Average hourly earnings | Average hourly earnings of the target employees |
ADP is not a preview of NFP.
ADP is a private employment measure based on payroll processing data, and its sample and methodology differ from the BLS. ADP itself states that it is not a report intended to forecast NFP. Its scope also differs from the total NFP, which includes government employment.
Even if ADP is weak, NFP can come in strong. Do not use the difference between the two figures as evidence of falsehood or manipulation; instead, check the private vs. government sources, industry composition, and revisions to previous data.
A decline in the unemployment rate does not always mean an increase in employment.
The unemployment rate uses the labor force as the denominator. If people give up job hunting and leave the labor market, the unemployment rate may decrease. Look at the labor force participation rate, employment-to-population ratio, and employment changes by survey together.
Average wages also fluctuate due to changes in the composition of low-wage and high-wage workers. You can better understand household income trends by separating month-over-month and year-over-year wage growth and checking average weekly hours. Do not interpret the wage growth rate as the wage increase rate for every individual.
Job openings and hires, initial and continuing claims
The number of job openings in JOLTS represents vacant positions, which is different from the actual number of hires. If job openings decrease while layoffs remain low, it may indicate a cooling of overheated demand. Voluntary quits are a clue to confidence in changing jobs, but they do not confirm the economic outlook on their own.
Initial jobless claims show new filings, while continuing claims show the flow of people remaining on benefits. The reference weeks for the two series may differ. Since weekly figures are affected by holidays, seasonal adjustments, and specific regional events, also check the 4-week moving average.
Strong employment and rising stock prices are not the same thing.
Robust employment can support consumption and corporate revenue, but strong wage and price pressures can lead to expectations of higher interest rates. Conversely, even if weak employment lowers interest rate expectations, concerns about a severe recession may outweigh expectations for corporate earnings.
Do not just compare the current figure with expectations; record the revisions for the previous two months along with employment and wages by industry. Since reports are based on sample estimates, do not treat a single small discrepancy as a definitive trend.
You should not make such assumptions
“If ADP is below expectations, NFP must also be below expectations.”
The data, subjects, and methods are different. ADP is not a report intended to forecast NFP.
Check what you have read.
- Did you distinguish between the number of jobs and the number of people?
- Have you looked at the unemployment rate and participation rate together?
- Have you checked the previous revisions and wages?
- Did you check the reference week and the average for weekly claims?
Apply what you read to the actual screen
Try it on WAPE
Source and scope of verification
Reviewed · Version 1
- National Employment ReportNew window
ADP Research · Verify 2026-10-20
Private employment data, differences in samples between BLS employment reports, and non-predictive relationships - Employment Situation · Technical NoteNew window
U.S. Bureau of Labor Statistics · Verify 2026-10-20
Establishment survey and household survey, non-farm payrolls, unemployment rate, wages, and working hours · Limitations of samples and revisions - JOLTS Frequently Asked QuestionsNew window
U.S. Bureau of Labor Statistics · Verify 2026-10-20
Scope and definitions of job opening, hiring, and turnover data. - Unemployment Insurance Weekly ClaimsNew window
U.S. Department of Labor · Verify 2026-10-20
Initial and continuing jobless claims and the 4-week average; source path for weekly releases
This is educational material for observation and information comprehension. Hypothetical examples do not represent actual stocks or performance and do not provide buy or sell recommendations.