The Jobs Report Explained: NFP, Unemployment and Wages

Key takeaways

Payroll employment counts covered jobs; household employment counts people. Their coverage and monthly changes differ.

Read unemployment alongside participation and employment. A falling unemployment rate can occur without new jobs, as the separate invented example shows.

Keep current-month payroll estimates separate from revisions to earlier months. In the fictional report, +200,000 remains the current estimate despite −70,000 of earlier-month revisions.

Commonly watched average hourly earnings cover private nonfarm employees. Nominal growth and changes in the job mix do not establish a pay increase or purchasing-power gain for every worker.

Read the release as a whole. Yield, dollar and stock-index reactions depend on surprises and context; a report does not guarantee the next candle or an execution price.

Full transcript

Payroll growth beats the forecast. But unemployment rises, and earlier job gains are revised lower.

Is that a strong jobs report?

Before choosing a label, we need to separate what each figure measures. The employment report brings together different surveys, different units and different periods. One headline cannot explain all of them.

This is the Employment Situation report from Nigel Kavor: nonfarm payrolls, unemployment and wages.

We will briefly examine its history, distinguish its main measures, and work through an invented release. Our purpose is to read the evidence more carefully before interpreting a market response.

A brief history

The report grew from the need to measure work and unemployment. In nineteen fifteen, the Bureau of Labor Statistics began collecting employment and payroll data for four manufacturing industries. The Great Depression later intensified the need for reliable unemployment statistics, and the monthly household survey that became the Current Population Survey began in nineteen forty.

In nineteen sixty-six, household unemployment figures and payroll jobs data began appearing together in the Employment Situation release. That was a reporting milestone, not a first trading date. Research using nineteen ninety-one through nineteen ninety-five data documents Treasury-market responses to payroll announcement surprises. Market use was established decades ago; we are not claiming to identify its exact beginning.

The release calendar

Today, the Bureau of Labor Statistics, or BLS, normally publishes the report at eight thirty in the morning, Eastern Time, on its scheduled release day. Check the official calendar rather than assuming every report falls on the same Friday.

Two surveys: jobs and people

Start with two surveys.

The establishment survey collects information from businesses and government agencies. It supplies nonfarm payroll employment, hours and earnings.

The household survey asks people about their work and job-search activities. It supplies the unemployment rate, labor force participation and other measures of people's employment status.

Payroll employment counts jobs. Household employment counts people. Someone holding two covered payroll jobs can be counted twice in the establishment survey, but once as an employed person in the household survey.

Coverage also differs. For example, the household survey includes unincorporated self-employed workers, while the payroll measure excludes them. The two surveys are not interchangeable counts, and their monthly employment changes need not match.

The familiar NFP headline usually describes the change in total nonfarm payroll employment. A gain of two hundred thousand means the estimated number of covered payroll jobs increased by that amount from the previous month. It is a net change, not a count of every person hired or every vacancy filled.

Unemployment and participation

The unemployment rate answers another question: what share of the labor force is unemployed?

In general, someone counted as unemployed has no job, is available for work and has actively looked during the preceding four weeks. There are specific exceptions, including people on temporary layoff expecting recall. Receiving unemployment benefits is not what determines the survey classification.

The labor force combines employed and unemployed people. People who meet neither classification are outside it. The participation rate measures the labor force as a share of the civilian noninstitutional population aged sixteen and over.

That denominator matters.

A separate denominator illustration

Consider a separate invented illustration. A labor force contains one thousand people: nine hundred and sixty employed and forty unemployed. Forty divided by one thousand gives a four percent unemployment rate.

Suppose ten of those unemployed people stop actively searching and no longer meet the unemployment classification. Hold everything else constant. Employment remains nine hundred and sixty. The labor force falls to nine hundred and ninety, with thirty unemployed.

The unemployment rate becomes about three percent, even though nobody in this illustration obtained a job.

That is one possible mechanism, not an explanation for every falling unemployment rate. To understand an actual change, examine employment, unemployment and participation together.

Wages and job mix

Now consider wages. A commonly watched figure is average hourly earnings for all employees on private nonfarm payrolls. It is not the same population as total nonfarm payroll employment, which includes government jobs.

Average hourly earnings are expressed in dollars, with monthly and annual percentage changes also reported. They are not adjusted for consumer-price inflation. A nominal increase does not by itself establish an increase in purchasing power.

The average can also change when the mix of jobs changes. If lower-paid jobs disappear, average earnings can rise even without a pay increase for every remaining worker. Read hours worked and the composition of employment alongside the wage figure.

An invented employment release

Let us open a fictional report. All the following release figures and forecasts are invented. They are not current data, an actual historical event, or my research.

Our calendar shows a forecast of one hundred and fifty thousand additional payroll jobs. The initial estimate is two hundred thousand.

That is fifty thousand above the stated forecast.

But two earlier months have changed. One previously showed one hundred and twenty thousand additional jobs; it now shows eighty thousand. Another previously showed one hundred and forty thousand; it now shows one hundred and ten thousand.

Those revisions reduce the earlier months' combined reported gains by seventy thousand. They do not turn the current month's two hundred thousand into a different current-month estimate. Keep the periods separate.

Payroll estimates are revised as more information arrives and seasonal calculations are updated. There is also an annual benchmarking process. A first estimate is useful information, but it is not an exact, permanent census of every payroll job.

In our fictional household figures, unemployment rises from four percent to four point two percent, against a forecast of four percent.

That is an increase of zero point two percentage points. These figures alone do not reveal whether job losses, changes in participation or other movements explain it. We need the underlying household data.

Meanwhile, monthly average hourly earnings rise zero point two percent. The previous reading and forecast were both zero point three percent. Wage growth is positive, but slower than those comparisons. Average weekly hours also decline from thirty-four point four to thirty-four point two.

Reading a mixed release

What can we conclude?

The new payroll estimate exceeds its stated forecast. Earlier payroll gains are revised lower. Unemployment is higher, monthly wage growth is below forecast, and average weekly hours are lower.

Those observations describe a mixed release. They do not establish the direction of the next futures candle or guarantee the path of interest rates.

Possible market reactions

Stronger-than-expected hiring can lead traders to expect interest rates to stay higher, supporting Treasury yields and the dollar. A weaker report can encourage expectations of rate cuts, putting downward pressure on yields.

For stock-index futures, the interpretation can go either way. Strong hiring may support confidence in company earnings, but higher rate expectations can pressure prices. Weaker hiring may bring relief about rates—or concern about a slowing economy.

These are possible reactions, not instructions for the next candle. Read payrolls, revisions, unemployment and wages together, and ask which concern matters most in the current environment.

A practical reading order

Instead, prepare a short reading order: identify the survey and unit, compare the new estimate with the recorded forecast, inspect revisions to earlier periods, then examine unemployment, participation, earnings and hours together.

Reference periods and uncertainty

Check reference periods and seasonal-adjustment labels. The household survey generally refers to the week containing the twelfth of the month; the establishment survey refers to the pay period containing it. Neither is a live count of conditions at the moment the report is released. Both are statistical estimates with uncertainty.

Preparation and review

If you watched the CPI lesson, the forecast-recording and execution checks still apply. Preserve the provider's estimate before the release, check the actual publication time and your current account rules, and follow the written plan. Understanding a report does not guarantee an entry or exit price.

Afterward, record what was published, what changed from earlier estimates, and what you inferred from the response. Keep those separate.

Payroll jobs, unemployment and wages belong in the same conversation. Understanding their differences is what makes that conversation useful.

Next: FOMC Explained

Connect economic evidence with policy decisions. Examine the target range, expectations, statements, projections and press conferences through an invented meeting.

Keep learning

Revisit inflation, trading orders and the written plan, or explore the full lesson library.

CPI Explained · How Trading Orders Work · Build a Trading Plan · All lessons