Jobs report – HubbardOBrienEconomics.com Blog

Surprising Decline in Employment in the February Jobs Report

Image created by ChatGPT

This morning (March 6), the Bureau of Labor Statistics (BLS) released its “Employment Situation” report (often called the “jobs report”) for February. The jobs report for January showed a much stronger than expected increase in employment. Today’s report was a surprise in the opposite direction with employment unexpectedly declining.

The jobs report has two estimates of the change in employment during the month: one estimate from the establishment survey, often referred to as the payroll survey, and one from the household survey. As we discuss in Macroeconomics, Chapter 9, Section 9.1 (Economics, Chapter 19, Section 19.1), many economists and Federal Reserve policymakers believe that employment data from the establishment survey provide a more accurate indicator of the state of the labor market than do the household survey’s employment data and unemployment data. (The groups included in the employment estimates from the two surveys are somewhat different, as we discuss in this post.)

According to the establishment survey, there was a net decrease of 92,000 nonfarm jobs during February. Economists surveyed by the Wall Street Journal had forecast an increase of 50,000 jobs. Economists surveyed by FactSet had a higher forecast of a net increase of 70,000 jobs. The BLS revised downward its previous estimates of employment in December and January by a combined 69,000 jobs. The revised estimate indicates that employment fell in December by 17,000 rather than increasing by 48,000 as in the previous estimate. (The BLS notes that: “Monthly revisions result from additional reports received from businesses and government agencies since the last published estimates and from the recalculation of seasonal factors.”)

The following figure from the jobs report shows the net change in nonfarm payroll employment for each month in the last two years. The current estimates show a net decrease in employment during five of the last nine months.

The unemployment rate, which is calculated from data in the household survey, increased to 4.4 percent in February for 4.3 percent in January. As the following figure shows, the unemployment rate has been remarkably stable over the past year and a half, staying between 4.0 percent and 4.4 percent in each month since May 2024. The Federal Open Market Committee’s current estimate of the natural rate of unemployment—the normal rate of unemployment over the long run—is 4.2 percent. So, unemployment is slightly above that estimate of the natural rate. (We discuss the natural rate of unemployment in Macroeconomics, Chapter 9 and Economics, Chapter 19.)

As the following figure shows, the monthly net change in jobs from the household survey moves much more erratically than does the net change in jobs from the establishment survey. As measured by the household survey, there was a net decrease of 185,000 in February. (Note that because of last year’s shutdown of the federal government, there are no data for October or November.) In any particular month, the story told by the two surveys can be inconsistent. In this case, both surveys indicate a net decline in employment. (In this blog post, we discuss the differences between the employment estimates in the two surveys.)

The household survey has another important labor market indicator: the employment-population ratio for prime age workers—those workers aged 25 to 54. In February the ratio was 80.7 percent, down slightly from 80.8 percent in January. The prime-age population ratio remains above its value for most of the period since 2001. The continued high levels of the prime-age employment-population ratio indicate some continuing strength in the labor market.

The Trump Administration’s layoffs of some federal government workers are clearly shown in the estimate of total federal employment for October, when many federal government employees exhausted their severance pay. (The BLS notes that: “Employees on paid leave or receiving ongoing severance pay are counted as employed in the establishment survey.”) As the following figure shows, there was a decline in federal government employment of 166,000 in October, with additional declines in the following four months. The total decline in federal government employment since the beginning of February 2025 is 327,000. But the decline has been slowing, with a net decrease of 10,000 jobs in February.

The establishment survey also includes data on average hourly earnings (AHE). As we noted in this post, many economists and policymakers believe the employment cost index (ECI) is a better measure of wage pressures in the economy than is the AHE. The AHE does have the important advantage of being available monthly, whereas the ECI is only available quarterly. The following figure shows the percentage change in the AHE from the same month in the previous year. The AHE increased 3.8 percent in February, up slightly from 3.7 percent in January.

The following figure shows wage inflation calculated by compounding the current month’s rate over an entire year. (The figure above shows what is sometimes called 12-month wage inflation, whereas this figure shows 1-month wage inflation.) One-month wage inflation is much more volatile than 12-month wage inflation—note the very large swings in 1-month wage inflation in April and May 2020 during the business closures caused by the Covid pandemic. In February, the 1-month rate of wage inflation was 5.0 percent, unchanged from January. This high rate of wage growth is surprising given the decline in employment. But two month’s data from such a volatile series may not accurately reflect longer-run trends in wage inflation.

What effect is this weak jobs report likely to have on the decisions of the Federal Reserve’s policymaking Federal Open Market Committee at its next meeting on March 17–18? Taken by itself, employment having fallen in five of the last nine months might be expected to cause the committee to cut its target range for the federal funds rate. But disruptions to the world oil market as a result of the U.S. and Israeli bombing campaign in Iraq have caused oil prices to rise, putting upward pressure on the price level. In addition, wage growth in the United States appears higher than is consistent with price inflation returning to the Fed’s 2 percent annual target. These factors make it likely that the committee will keep its target range for the federal funds rate unchanged at its next meeting.

The probability that investors in the federal funds futures market assign to the FOMC keeping its target rate unchanged at its March meeting was largely unchanged this morning at 95.6 percent, only a slight decrease from 96.3 percent yesterday.

How Many Manufacturing Workers Are There in the United States?

Image created by ChatGPT

Every president dating back to at least Ronald Reagan, who took office in January 1981, has promised to increase manufacturing employment. Manufacturing jobs are often seen as making it possible for workers without a college degree to earn a middle-class income. As the following figure shows, though, since 2018, average hourly earnings of workers in manufacturing have actually been less than average hourly earnings of all workers.

If we look at just the wages of production and nonsupervisory workers in manufacturing—like the workers shown in the image above—during the past 20 years, the average hourly earnings of production workers in manufacturing have generally been about 20 percent less than the average hourly earnings of all workers.

The following figure shows the absolute number of all employees in manufacturing (the blue line) and production and nonsupervisory employees in manufacturing monthly since 1939. Employment of production workers peaked in 1943, during World War II. Employment of all employees in manufacturing peaked in 1979. (All employees in manufacturing include, in addition to production workers, managers and other employees with administrative duties, accountants, lawyers, salespeople, and all other employees not directly concerned with production.) The trend in manufacturing employment has generally been downward since 1979 and has been below 13 million every month since December 2008. In January 2026, there were 12.6 million total employees in manufacturing of whom 8.8 million were production workers.

The following figure shows manufacturing employment as a percentage of total employment for each month since 1939. Manufacturing employment peaked as percentage of total employment at 38.7 percent in 1943. It has slowly trended down since that time, being below 10 percent every month since September 2007. In January 2026, manufacturing employment was 7.9 percent of total employment.

All of the data in the figurs shown so far are from the establishment survey (formally, the Current Employment Statistics (CES)). Recently, Adam Ozimek, Benjamin Glasner, and Jiaxin He of the Economic Innovation Group have examined the discrepancy between the number of manufacturing workers as reported in establishment survey and the larger number of manufacturing workers reported in the household survey (formally, the Current Population Survey (CPS).) Each month when the Bureau of Labor Statistics (BLS) releases its “Employment Situation” report, usually referred to as the “jobs report,” attention focuses on two numbers: The change in total employment as calculated from the establishment survey and the unemployment rate as calculated from the household survey.

In addition to the unemployment rate, the BLS releases monthly data on total employment and on employment by industry from the household survey. Most economists, policymakers, and investment analysts pay little attention to the data on employment by industry from the household survey because the employment by industry data from the establishment survey is considered more reliable. In fact, the employment by industry data from the household survey isn’t included among the many macro series available on the FRED site. The following figure reproduces the two establishment survey (CES) data (the blue and red lines) shown in the third figure above along with the household survey (CPS) data (the green line) from the BLS site. (Note that the household survey data is choppier than the data in the other two series because it is not seasonally adjusted.)

Manufacturing employment is consistently larger in the household survey data than in the establishment survey data. For example, in January 2026, total manufacturing employment according to the establishment survey was 12.6 million, whereas total manufacturing employment according to the household survey was 15.4 million—a difference of 2.7 million. Put another way, if the household survey is accurate, manufacturing employment is actually 20 percent higher than it appears from the widely-used establishment survey data.

The establishment survey data is collected by surveying firms, whereas the household survey data is collected from surveying workers. In other words, in January, 2.7 million more workers considered themselves to be in manufacturing than firms reported were actually working in manufacturing. Typically, economists and policymakers consider results from the establishment survey to be more reliable because firms are legally obliged to keep accurate accounts of the number of their employees, whereas the answers from workers responding to surveys are accepted without additional checking.

Ozimek, Glasner, and He note that the persistence of a gap between the establishment and household data on manufacturing employment indicates that there are some establishments that the census considers to be engaged in some activity other than manufacturing but whose workers consider themselves to be in manufacturing. The authors present a careful discussion of the issues involved and the entire piece (linked to above) is worth reading carefully by anyone who is concerned about this issue, but we can mention here one particularly interesting point.

The authors link to a paper by Andrew Bernard and Theresa Fort of Dartmouth College discussing “factoryless goods producing firms,” which are “manufacturing-like as they perform many of the tasks and activities found in manufacturing firms” but that don’t actually manufacture goods. Ozimek, Glasner, and He give as one example Apple’s Elk Grove, California site. They note that at one time Apple assembled computers at that site but that currently “there is no assembly at that location, but thousands of Apple employees work there on logistics, distribution, repair, and customer support.” In other words, the site contributes to manufacturing Apple’s products and, if surveyed, many of its employees might respond that they work in manufacturing, but because no products are actually assembled at the site, the site won’t be considered as engaged in manufacturing by the establishment survey. They conclude that: “These sorts of employees—who work adjacent to manufacturing, but not in categorized establishments—make up a big chunk of the 2.2 to 2.8 million missing manufacturing workers.”

Clearly, an important issue in an accurate count of manufacturing workers is a definition of what we mean by manufacturing. Should a particular site—establishment—be considered as engaged in manufacturing only if products are assembled at that site? Or should a site be considered as engaged in manufacturing if its purpose is to support assembly that is done elsewhere?

Because the number of manufacturing workers and the fraction of the labor force engaged in manufacturing have been important political issues for decades, it’s somewhat surprising how little attention has been devoted to ensuring that we’re actually correctly measuring manufacturing employment.

Surprisingly Strong Jobs Report Accompanied by a Large Downward Annual Benchmark Revision

Image created by ChatGPT

This morning (February 11), the Bureau of Labor Statistics (BLS) released its “Employment Situation” report (often called the “jobs report”) for January. The report was originally scheduled to be released last Friday but was postponed by the brief federal government shutdown. The data in the report show that the labor market was much stronger than expected in January.

According to the establishment survey, there was a net increase of 130,000 nonfarm jobs during January. This increase was well above the increase of 55,000 that economists surveyed by the Wall Street Journal had forecast. Economists surveyed by Bloomberg had a higher forecast of 65,000 net jobs. The BLS revised downward its previous estimates of employment in November and December by a combined 17,000 jobs. (The BLS notes that: “Monthly revisions result from additional reports received from businesses and government agencies since the last published estimates and from the recalculation of seasonal factors.”)

The following figure from the jobs report shows the net change in nonfarm payroll employment for each month in the last two years. The increase in net jobs in January was the largest since December 2024.

The unemployment rate, which is calculated from data in the household survey, fell from 4.4 percent in December to 4.3 percent in January. As the following figure shows, the unemployment rate has been remarkably stable over the past year and a half, staying between 4.0 percent and 4.4 percent in each month since May 2024. The Federal Open Market Committee’s current estimate of the natural rate of unemployment—the normal rate of unemployment over the long run—is 4.2 percent. So, unemployment is slightly above the natural rate. (We discuss the natural rate of unemployment in Macroeconomics, Chapter 9 and Economics, Chapter 19.)

As the following figure shows, the monthly net change in jobs from the household survey moves much more erratically than does the net change in jobs from the establishment survey. As measured by the household survey, there was a net increase of 528,000 in January, far above the increase in jobs from the payroll survey. (Note that because of last year’s shutdown of the federal government, there are no data for October or November.) In any particular month, the story told by the two surveys can be inconsistent. In this case, both surveys indicate unexpectedly strong job growth, with the increase in household employment being particularly strong. (In this blog post, we discuss the differences between the employment estimates in the two surveys.)

The household survey has another important labor market indicator: the employment-population ratio for prime age workers—those workers aged 25 to 54. In January the ratio was 80.9 percent, the highest since September 2024. In addition to matching the recent highs reached in mid-2024, the prime-age employment-population ratio is above what the ratio was in any month since April 2001. The continued high levels of the prime-age employment-population ratio indicates some continuing strength in the labor market.

The Trump Administration’s layoffs of some federal government workers are clearly shown in the estimate of total federal employment for October, when many federal government employees exhausted their severance pay. (The BLS notes that: “Employees on paid leave or receiving ongoing severance pay are counted as employed in the establishment survey.”) As the following figure shows, there was a decline federal government employment of 166,000 in October, with additional declines in the following three months. The total decline in federal government employment since the beginning of February 2025 is 324,000.

The following figure shows wage inflation calculated by compounding the current month’s rate over an entire year. (The figure above shows what is sometimes called 12-month wage inflation, whereas this figure shows 1-month wage inflation.) One-month wage inflation is much more volatile than 12-month wage inflation—note the very large swings in 1-month wage inflation in April and May 2020 during the business closures caused by the Covid pandemic. In January, the 1-month rate of wage inflation was 5.0 percent, up from 0.7 percent in December. This increase in wage growth may be an indication of a strengthening labor market. But one month’s data from such a volatile series may not accurately reflect longer-run trends in wage inflation.

In today’s jobs report, the BLS also included its final annual benchmark revision to the establishment employment data. (We discussed the preliminary annual revision in this blog post last September.) The following table from the jobs report indicates that the revision was quite substantial. The revised estimate of payroll employment is 1,029,000 jobs lower than the original estimate. The increase in total nonfarm employment in 2025 was revised down to only 181,000 from the original estimate of 584,000. Leaving aside the collapse in employment in 2020 during the Covid pandemic, job growth in 2025 was the slowest since 2010 in the immediate aftermath of the Great Recession of 2007–2009.

Despite the large downward revision to job growth in 2025, the strong job growth for January in today’s jobs report makes it unlikely that the Federal Reserve’s policymaking Federal Open Market Committee (FOMC) will lower its target for the federal funds rate at its next meeting on March 17–18. The probability that investors in the federal funds futures market assign to the FOMC keeping its target rate unchanged at that meeting jumped from 79.9 percent yesterday to 92.1 percent after the release of today’s jobs report.

December JOLTS Report Shows Possible Labor Market Weakening

Image created by ChatGPT

Today (February 5), the Bureau of Labor Statistics (BLS) released its “Job Openings and Labor Turnover” (JOLTS) report for December 2025. The report indicated that labor market conditions may be weakening. The following figure shows that the rate of job openings fell to 3.9 percent in December from 4.2 percent in November. The rate was 4.5 percent in October. The job openings rate is the lowest since April 2020, at the start of the Covid pandemic. We should note the usual caveat that the monthly JOLTS data is subject to potentially large revisions as the BLS receives more complete data.

(The BLS defines a job opening as a full-time or part-time job that a firm is advertising and that will start within 30 days. The rate of job openings is the number of job openings divided by the number of job openings plus the number employed workers, multiplied by 100.)

In the following figure, we show a measure of the state of the labor market that economists frequently use: the total number of job openings to the total number of people unemployed. In December there were 0.87 job openings per unemployed person, the lowest value for that measure since March 2021, during the recovery from the pandemic. The value was 1.0 in September. (Note that data for October and November are unavailable because the data weren’t collected during the shutdown of the federal government from October 1 to November 12 last year.) The value for December is well below the 1.21 job openings per employed person in February 2020, just before the pandemic. (Note that, as we discussed in this blog post, the employment-population ratio for prime age workers, which many economists consider a key measure of the state of the labor market, rose in December, putting it above what the ratio was in any month during the period from January 2008 to February 2020.)

The rate at which workers are willing to quit their jobs is an indication of how they perceive the ease of finding a new job. As the following figure shows, the quit rate declined slowly from a peak of 3 percent in late 2021 and early 2022 to 2.0 percent in August 2024, the same value as in December 2025. That rate is below the rate during 2019 and early 2020. By this measure, workers’ perceptions of the state of the labor market have remained remarkably stable over the last year and a half.

Overall, this JOLTS report is consistent with what some economists have labeled a “slow hire, slow fire” labor market. Fed Chair Jerome Powell’s remarks at his press conference following the last meeting of the Federal Open Market Committee (FOMC) indicates that Fed policymakers share this view, which Powell believes complicates monetary policymaking:

“So there are lots of … little places that suggest that the labor market has softened, but part of … payroll job softening is that both the supply and demand for labor has come down … growth in those two have come down. So that makes it a difficult time to read the labor market. So, imagine they both came down a lot, to the point where there is no job growth. Is that full employment? In a sense it is. If demand and supply are … in balance, you could say that’s full employment. At the same time, is it—do we really feel like … that’s a maximum employment economy? It’s a challenging—it’s very challenging and quite unusual situation.”

The BLS was scheduled to release its monthly “Employment Situation” report (often called the “jobs report”) for January 2026 tomorrow. Because of the temporary lapse in funding that began Saturday, the report will instead be released next Wednesday, February 11. That report will provide additional data on the state of the labor market. (Note that the data in the JOLTS report lag the data in the “Employment Situation” report by one month.)

October and November Jobs Data Give Mixed Picture of the Labor Market

Image created by ChatGPT

Because of the federal government shutdown from October 1 to November 12, the regular release by the Bureau of Labor Statistics (BLS) of its monthly “Employment Situation” report (often called the “jobs report”) has been disrupted. The jobs report usually has two estimates of the change in employment during the month: one estimate from the establishment survey, often referred to as the payroll survey, and one from the household survey. As we discuss in Macroeconomics, Chapter 9, Section 9.1 (Economics, Chapter 19, Section 19.1), many economists and Federal Reserve policymakers believe that employment data from the establishment survey provide a more accurate indicator of the state of the labor market than do the household survey’s employment data and unemployment data. (The groups included in the employment estimates from the two surveys are somewhat different, as we discuss in this post.)

Today, the BLS released a jobs report that has data from the payroll survey for both October and November, but data from the household survey only for November. Because of the government shutdown, the household survey for October wasn’t conducted.

According to the establishment survey, there was a net decrease of 105,000 nonfarm jobs in October and a net increase of 64,000 nonfarm jobs in November. The increase for November was above the increase of 40,000 that economists surveyed by FactSet had forecast. Economists surveyed by the Wall Street Journal had forecast a net increase of 45,000 jobs. The BLS revised downward by a combined 33,000 jobs its previous estimates of employment in August and September. (The BLS notes that: “Monthly revisions result from additional reports received from businesses and government agencies since the last published estimates and from the recalculation of seasonal factors.”)

The following figure from the jobs report shows the net change in nonfarm payroll employment for each month in the last two years. The figure illustrates that, as the BLS notes in the report, nonfarm payroll employment “has shown little net change since April.” The Trump administration announced sharp increases in U.S. tariffs on April 2. Media reports indicate that some firms have slowed hiring due to the effects of the tariffs or in anticipation of those effects. In addition, a sharp decline in immigration has slowed growth in the labor force.

The unemployment rate estimate relies on data collected in the household survey, so there id no unemployment estimate for October. As shown in the following figure, the unemployment rate increased from 4.4 percent in September to 4.6 percent in November, the highest rate since September 2021. The unemployment rate is above the 4.4 percent rate economists surveyed by FactSet had forecast. The unemployment rate had been remarkably stable, staying between 4.0 percent and 4.2 percent in each month from May 2024 to July 2025, before breaking out of that range in August. The Federal Open Market Committee’s current estimate of the natural rate of unemployment—the normal rate of unemployment over the long run—is 4.2 percent. So, unemployment is now well above the natural rate. (We discuss the natural rate of unemployment in Macroeconomics, Chapter 9 and Economics, Chapter 19.)

As the following figure shows, the monthly net change in jobs from the household survey moves much more erratically than does the net change in jobs from the establishment survey. As measured by the household survey, there was a net increase of 96,000 jobs from September to November. In the payroll survey, there was a net decrease in of 41,000 jobs from September to November. In any particular month, the story told by the two surveys can be inconsistent. In this case, we are measuring the change in jobs over a two month interval because there is no estimate from the household survey of employment in October. Over that two month period the household survey is showing more strength in the labor market than is the payroll survey. (In this blog post, we discuss the differences between the employment estimates in the two surveys.)

The household survey has another important labor market indicator: the employment-population ratio for prime age workers—those workers aged 25 to 54. In November the ratio was 80.6 percent, down slightly from 80.7 in September. (Again, there is no estimate for October.) The prime-age employment-population ratio is somewhat below the high of 80.9 percent in mid-2024, but is still above what the ratio was in any month during the period from January 2008 to February 2020. The continued high levels of the prime-age employment-population ratio indicates some continuing strength in the labor market.

The Trump Administration’s layoffs of some federal government workers are clearly shown in the estimate of total federal employment for October, when many federal government employees exhausted their severance pay. (The BLS notes that: “Employees on paid leave or receiving ongoing severance pay are counted as employed in the establishment survey.”) As the following figure shows, there was a decline federal government employment of 162,000 in October, with an additional decline of 6,000 In November. The total decline since the beginning of February 2025 is 271,000. At this point, we can say that the decline in federal employment has had a significant effect on the overall labor market and may account for some of the rise in the unemployment rate.

The establishment survey also includes data on average hourly earnings (AHE). As we noted in this post, many economists and policymakers believe the employment cost index (ECI) is a better measure of wage pressures in the economy than is the AHE. The AHE does have the important advantage of being available monthly, whereas the ECI is only available quarterly. The following figure shows the percentage change in the AHE from the same month in the previous year. The AHE increased 3.5 percent in November, down from 3.7 percent in October.

The following figure shows wage inflation calculated by compounding the current month’s rate over an entire year. (The figure above shows what is sometimes called 12-month wage inflation, whereas this figure shows 1-month wage inflation.) One-month wage inflation is much more volatile than 12-month wage inflation—note the very large swings in 1-month wage inflation in April and May 2020 during the business closures caused by the Covid pandemic. In November, the 1-month rate of wage inflation was 1.6 percent, down from 5.4 percent in October. This slowdown in wage growth may be an indication of a weakening labor market. But one month’s data from such a volatile series may not accurately reflect longer-run trends in wage inflation.

What effect might today’s jobs report have on the decisions of the Federal Open Market Committee (FOMC) with respect to setting its target range for the federal funds rate? Today’s jobs report provides a mixed take on the state of the labor market with very slow job growth—although the large decline in federal employment is a confounding factor—a continued high employment-population ratio for prime age workers, and slowing wage growth.

Surprisingly Strong September Jobs Report

Image created by ChatGPT

If not for the shutdown of the federal government, the Bureau of Labor Statistics (BLS) would have already released its “Employment Situation” report (often called the “jobs report”) for September and October by now. The September jobs report was released today based largely on data collected before the shutdown.

Because the household survey wasn’t conducted in October, the data in the October report that relies on the household survey won’t be included when the BLS releases establishment employment data for October on December 16. The data for September released today showed the labor market was stronger than expected in that month.

According to the establishment survey, there was a net increase of 119,00 nonfarm jobs during September. This increase was well above the increase of 50,000 that economists surveyed by FactSet had forecast. Economists surveyed by the Wall Street Journal had also forecast a net increase of 50,000 jobs. The relatively large increase in employment in September was partially offset by the BLS revising downward by a combined 33,000 jobs its previous estimates of employment in July and August. The estimate for August was revised from a net gain of 22,000 to a net loss of 4,000. (The BLS notes that: “Monthly revisions result from additional reports received from businesses and government agencies since the last published estimates and from the recalculation of seasonal factors.”)

The following figure from the jobs report shows the net change in nonfarm payroll employment for each month in the last two years. The figure makes clear the striking deceleration in job growth beginning in May. The Trump administration announced sharp increases in U.S. tariffs on April 2. Media reports indicate that some firms have slowed hiring due to the effects of the tariffs or in anticipation of those effects.

As shown in the following figure, the unemployment rate increased from 4.3 percent in August to 4.4 percent in September, the highest rate since October 2021. The unemployment rate is above the 4.3 percent rate economists surveyed by FactSet had forecast. The unemployment rate had been remarkably stable, staying between 4.0 percent and 4.2 percent in each month from May 2024 to July 2025, before breaking out of that range in August. In September, the members of the Federal Open Market Committee (FOMC) forecast that the unemployment rate during the fourth quarter of 2025 would average 4.5 percent. The FOMC’s current estimate of the natural rate of unemployment—the normal rate of unemployment over the long run—is 4.2 percent. (We discuss the natural rate of unemployment in Macroeconomics, Chapter 9 and Economics, Chapter 19.)

Each month, the Federal Reserve Bank of Atlanta estimates how many net new jobs are required to keep the unemployment rate stable. Given slower growth in the working-age population due to the aging of the U.S. population and a sharp decline in immigration, the Atlanta Fed currently estimates that the economy would have to create 111,878 net new jobs each month to keep the unemployment rate stable at 4.4 percent. If this estimate is accurate, if the average monthly net job increase from May through September of 38,600 were to continue, the result would be a rising unemployment rate.

As the following figure shows, the monthly net change in jobs from the household survey moves much more erratically than does the net change in jobs from the establishment survey. As measured by the household survey, there was a net increase of 251,000 jobs in September, following a net increase of 288,000 jobs in August. As an indication of the volatility in the employment changes in the household survey note the very large swings in net new jobs in January and February. In any particular month, the story told by the two surveys can be inconsistent. as was the case in September with employment increasing much more in the household survey than in the employment survey. (In this blog post, we discuss the differences between the employment estimates in the two surveys.)

The household survey has another important labor market indicator: the employment-population ratio for prime age workers—those aged 25 to 54. In September the ratio was 80.7 percent, the same as in August. The prime-age employment-population ratio is somewhat below the high of 80.9 percent in mid-2024, but is still above what the ratio was in any month during the period from January 2008 to February 2020. The continued high levels of the prime-age employment-population ratio indicates strength in the labor market.

It is still unclear how many federal workers have been laid off since the Trump Administration took office. The establishment survey shows a decline in federal government employment of 3,000 in September and a total decline of 97,000 since the beginning of February 2025. However, the BLS notes that: “Employees on paid leave or receiving ongoing severance pay are counted as employed in the establishment survey.” It’s possible that as more federal employees end their period of receiving severance pay, future jobs reports may report a larger decline in federal employment. To this point, the decline in federal employment has had only a small effect on the overall labor market.

The following figure shows wage inflation calculated by compounding the current month’s rate over an entire year. (The figure above shows what is sometimes called 12-month wage inflation, whereas this figure shows 1-month wage inflation.) One-month wage inflation is much more volatile than 12-month wage inflation—note the very large swings in 1-month wage inflation in April and May 2020 during the business closures caused by the Covid pandemic. In September, the 1-month rate of wage inflation was 3.0 percent, down from 5.1 percent in August. This slowdown in wage growth may be an indication of a weakening labor market. But one month’s data from such a volatile series may not accurately reflect longer-run trends in wage inflation.

What effect might today’s jobs report have on the decisions of the Federal Reserve’s policymaking Federal Open Market Committee (FOMC) with respect to setting its target range for the federal funds rate? The minutes from the FOMC’s last meeting on October 28–29 indicate that committee members had “strongly differing views” over whether to cut the target range by 0.25 percentage point (25 basis points) at its next meeting on December 9–10 or to leave the target range unchanged.

One indication of expectations of future changes in the FOMC’s target for the federal funds rate comes from investors who buy and sell federal funds futures contracts. (We discuss the futures market for federal funds in this blog post.) A month ago, investors assigned a 98.8 percent probability of the committee cutting its target range to 3.50 percent to 3.75 percent at its December meeting. Since that time indications have increased that output and employment growth have continued to be relatively strong and that inflation is stuck above the Fed’s 2 percent annual target. This morning, as the following figure shows, investors assign a probability of 60. 4 percent to the committee keeping its target unchanged at 3.75 percent to 4.00 percent at the December meeting. Committee members will also release their Summary of Economic Projections (SEP) at that meeting. The SEP, along with Fed Chair Powell’s remarks at his press conference following the meeting, should provide additional information on the monetary policy path the committee intends to follow in the coming months.

Fed Has Apparently Lost Early Access to ADP Employment Data

Fed Governor Christopher Wallace on October 21, 2025 at the Fed’s Payment Innovation Conference (photo from federalreserve.gov)

The current partial shutdown of the federal government has delayed the release by the Bureau of Labor Statistics (BLS) of its “Employment Situation” report (often called the “jobs report”). The report had originally been scheduled to be released on October 3. In a recent blog post we discussed how well the employment data collected by the private payroll processing firm Automatic Data Processing (ADP) serves as an alternative measure of the state of the labor market. In that post we showed that ADP data on total private payroll employment tracks fairly well the BLS data on total private employment from its establishment survey (often called the payroll survey) .

An article in today’s Wall Street Journal reports that ADP has stopped providing the Fed with early access to its data. Apparently, as a public service ADP had been providing its data to the Fed a week before the data was publicly released. The article notes that ADP stopped providing the data soon after this speech delievered by Fed Governor Christopher Wallace in late August. In a footnote to the speech Wallace refers to “data that Federal Reserve staff maintains in collaboration with the employment services firm ADP.” The article points out, though, that Waller’s speech was only one of several times since 2019 that a Fed official has publicly mentioned receiving data from ADP.

Losing early access to the ADP data comes at a difficult time for the Fed, given that the BLS employment data are not available. In addition, the labor market has shown signs of weakening even though growth has remained strong in measures of output. If payroll employment has been falling, rather than growing slowly as it was in the August jobs report, that knowledge would affect the deliberations of the Fed’s policymaking Federal Open Market Committee (FOMC) at its next meeting on October 28–29. Serious deterioration in the labor market could lead the FOMC to cut its target for the federal funds rate by more than the expected 0.25 percentage point (25 basis points).

In a speech in 2019, Fed Chair Jerome Powell noted that the Fed staff had used ADP data to develop a new measure of payroll employment. Had that measure been available in 2008, Powell argued, the FOMC would have realized earlier than it did that employment was being severely affected by the deepening of the financial crisis:

“[I]n the first eight months of 2008, as the Great Recession was getting underway, the official monthly employment data showed total job losses of about 750,000. A later benchmark revision told a much bleaker story, with declines of about 1.5 million. Our new measure, had it been available in 2008, would have been much closer to the revised data, alerting us that the job situation might be considerably worse than the official data suggested.”

The Wall Street Journal article notes that Powell has urged ADP to resume sharing its employment data with the Fed.

No BLS Jobs Report Today. Are ADP’s Data a Good Substitute?

Image generated by ChatGPT

Ordinarily, on the first Friday of a month the Bureau of Labor Statistics (BLS) releases its “Employment Situation” report (often called the “jobs report”) containing data on the labor market for the previous month. There was no jobs report today (October 3) because of the federal government shutdown. (We discuss the shutdown in this blog post.)

Economists surveyed had forecast that today’s payroll survey would have shown a net increase of 51,000 jobs in September. When the shutdown ends, the BLS will publish its jobs report for September. Until that happens, employment data collected by the private payroll processing firm Automatic Data Processing (ADP) provides an alternative measure of the state of the labor market. ADP data covers only about 20 percent of total private nonfarm employment, but ADP attempts to make its data more consistent with BLS data by weighting its data to reflect the industry weights used in the BLS data.

How closely does ADP employment data track BLS payroll data? The following figure shows the ADP employment series (blue line) and the BLS payroll employment data (red line) with the values for both series set equal to 100 in January 2010. The two series track well with the exception of April and May 2020 during the worst of the pandemic. The BLS series shows a much larger decline in employment during those months than does the ADP series.

The next figure shows the 12-month percentage changes in the two series. Again, the series track fairly well except for the worst months of the pandemic and—strikingly—the month of April 2021 during the economic recovery. In that month, the ADP series increases by only 0.6 percent, while the BLS series soars by 13.1 percent.

Finally, economists, policymakers, and investors usually focus on the change in payroll employment from the previous month—that is, the net change in jobs—shown in the BLS jobs report. The following figure shows the net change in jobs in the two series, starting in January 2021 to avoid some of the largest fluctuations during the pandemic.

Again, the two series track fairly well, although the BLS data is more volatile. The ADP data show a net decline of 32,000 jobs in September. As noted earlier, economists surveyed were expecting a net increase of 51,000 jobs. During the months from May through August, BLS data show an average monthly net increase in jobs of only 39,250. So, whether the BLS number will turn out to be closer to the ADP number or to the number economists had forecast, the message would be the same: Since May, employment has grown only slowly. And, of course, as we’ve seen this year, whatever the BLS’s initial employment estimate for September turns out to be, it’s likely to be subject to significant revision in coming months. (We discuss why BLS revisions to its initial employment estimates can be substantial in this post.)

Where Did 911,000 Jobs Go?

Image generated by ChatGPT

Today (September 9), the Bureau of Labor Statistics (BLS) issued revised estimates of the increase in employment, as measured by the establishment survey, over the period from April 2024 through March 2025. The BLS had initially estimated that during that period net employment had increased by a total of 1,758,000 or an average of 147,000 jobs per month. The revision lowered this estimate by more than half to a total of 839,000 jobs or an average of only 70,000 net new jobs created per month. The difference between those two monthly averages means that the U.S. economy had generated a total of 919,000 fewer jobs during that period. The revision was larger than the downward revision of 800,000 jobs forecast by economists at Wells Fargo, Comerica Bank, and Pantheon Macroeconomics.

Why does the BLS have to revise its employment estimates? As we discuss in Macroeconomics, Chapter 9, Section 9.1 (Economics, Chapter 19, Section 19.1) the initial estimates that the BLS issues each month in its “Employment Situation” reports are based on a sample of 121,000 businesses and government agencies representing 631,000 worksites or “establishments.” The monthly data also rely on estimates of the number of employees at establishments that opened or closed during the month and on employment changes at establishments that failed to respond to the survey. In August or September of each year, the BLS issues revised employment estimates based on data from the Quarterly Census of Employment and Wages (QCEW), which relies on state unemployment insurance tax records. The unemployment tax records are much more comprehensive than the original sample of establishments because nearly all employers are included.

In today’s report, the BLS cited two likely sources of error in their preliminary estimates:

“First, businesses reported less employment to the QCEW than they reported to the CES survey (response error). Second, businesses who were selected for the CES survey but did not respond reported less employment to the QCEW than those businesses who did respond to the CES survey (nonresponse error).”

The preliminary benchmark estimates the BLS released today will be revised again and the final estimates for these months will be released in February 2026. The difference between the preliminary and final benchmark estimates can be substantial. For example, last year, the BLS’s initially preliminary benchmark estimate indicated that the net employment increase from April 2023 to March 2024 had been overestimated by 818,000 jobs. In February 2025, the final benchmark estimate reduced this number to 598,000 jobs.

Although this year’s revision is particularly large in absolute terms—the largest since at least 2001—it still represents only about 0.56 percent of the more than 159.5 million people employed in the U.S. economy. Still the size of this revision is likely to increase political criticism of the BLS.

How will this revision affect the decision by the Federal Open Market Committee (FOMC) at its next meeting on September 16-17 to cut or maintain its target for the federal funds rate? The members of the committee were probably not surprised by the downward revision in the employment estimates, although they may have anticipated that the revision would be smaller. In six of the past seven years, the BLS has revised its estimates of payroll employment downward in its annual preliminary benchmark revision.

As we noted in this recent post, even before the BLS revised its employment estimates downward, recent monthly net employment increases were well below the increases during the first half of the year. There was already a high likelihood that the FOMC intended to cut its target for the federal funds rate at its meeting on September 16–17. The substantial downward revision in the employment data makes a cut at the September meeting nearly a certainty and increases the likelihood that the FOMC will implement a second cut in its target for the federal funds rate at the committee’s meeting on October 28–29.

Weak Jobs Report Provides Further Evidence of Labor Market Softening

Image generated by ChatGPT

This morning (September 5), the Bureau of Labor Statistics (BLS) released its “Employment Situation” report (often called the “jobs report”) for August. The data in the report show that the labor market was weaker than expected in August.

According to the establishment survey, there was a net increase of only 22,000 nonfarm jobs during August. This increase was well below the increase of 110,000 that economists surveyed by FactSet had forecast. Economists surveyed by the Wall Street Journal had forecast a smaller increase of 75,000 jobs. In addition, the BLS revised downward its previous estimates of employment in June and July by a combined 21,000 jobs. The estimate for June was revised from a net gain of 14,000 to a net loss of 13,000. This was the first month with a net job loss since December 2020. (The BLS notes that: “Monthly revisions result from additional reports received from businesses and government agencies since the last published estimates and from the recalculation of seasonal factors.”)

The following figure from the jobs report shows the net change in nonfarm payroll employment for each month in the last two years. The figure makes clear the striking deceleration in job growth since April. The Trump administration announced sharp increases in U.S. tariffs on April 2. Media reports indicate that some firms have slowed hiring due to the effects of the tariffs or in anticipation of those effects.

The unemployment rate increased from 4.2 percent in July to 4.3 percent in August, the highest rate since October 2021. The unemployment rate is above the 4.2 percent rate economists surveyed by FactSet had forecast. As the following figure shows, the unemployment rate had been remarkably stable over the past year, staying between 4.0 percent and 4.2 percent in each month May 2024 to July 2025 before breaking out of that range in August. In June, the members of the Federal Open Market Committee (FOMC) forecast that the unemployment rate during the fourth quarter of 2025 would average 4.5 percent. The unemployment rate would still have to rise significantly for that forecast to be accurate.

Each month, the Federal Reserve Bank of Atlanta estimates how many net new jobs are required to keep the unemployment rate stable. Given a slowing in the growth of the working-age population due to the aging of the U.S. population and a sharp decline in immigration, the Atlanta Fed currently estimates that the economy would have to create 97,591 net new jobs each month to keep the unemployment rate stable at 4.3 percent. If this estimate is accurate, continuing monthly net job increases of 22,000 would result in a a rising unemployment rate.

As the following figure shows, the monthly net change in jobs from the household survey moves much more erratically than does the net change in jobs from the establishment survey. As measured by the household survey, there was a net increase of 288,000 jobs in August, following a net decrease of 260,000 jobs in July. As an indication of the volatility in the employment changes in the household survey note the very large swings in net new jobs in January and February. In any particular month, the story told by the two surveys can be inconsistent. as was the case this month with employment increasing much more in the household survey than in the employment survey. (In this blog post, we discuss the differences between the employment estimates in the two surveys.)

The household survey has another important labor market indicator: the employment-population ratio for prime age workers—those aged 25 to 54. In August the ratio rose to 80.7 percent from 8.4 percent in July. The prime-age employment-population ratio is somewhat below the high of 80.9 percent in mid-2024, but is still above what the ratio was in any month during the period from January 2008 to February 2020. The increase in the prime-age employment-population ratio is a bright spot in this month’s jobs report.

It is still unclear how many federal workers have been laid off since the Trump Administration took office. The establishment survey shows a decline in federal government employment of 15,000 in August and a total decline of 97,000 since the beginning of February 2025. However, the BLS notes that: “Employees on paid leave or receiving ongoing severance pay are counted as employed in the establishment survey.” It’s possible that as more federal employees end their period of receiving severance pay, future jobs reports may report a larger decline in federal employment. To this point, the decline in federal employment has had a small effect on the overall labor market.

The establishment survey also includes data on average hourly earnings (AHE). As we noted in this post, many economists and policymakers believe the employment cost index (ECI) is a better measure of wage pressures in the economy than is the AHE. The AHE does have the important advantage of being available monthly, whereas the ECI is only available quarterly. The following figure shows the percentage change in the AHE from the same month in the previous year. The AHE increased 3.7 percent in August, down from an increase of 3.9 percent in July.

The following figure shows wage inflation calculated by compounding the current month’s rate over an entire year. (The figure above shows what is sometimes called 12-month wage inflation, whereas this figure shows 1-month wage inflation.) One-month wage inflation is much more volatile than 12-month wage inflation—note the very large swings in 1-month wage inflation in April and May 2020 during the business closures caused by the Covid pandemic. In August, the 1-month rate of wage inflation was 3.3 percent, down from 4.0 percent in July. This slowdown in wage growth may be another indication of a weakening labor market. But one month’s data from such a volatile series may not accurately reflect longer-run trends in wage inflation.

What effect might today’s jobs report have on the decisions of the Federal Reserve’s policymaking Federal Open Market Committee (FOMC) with respect to setting its target for the federal funds rate? One indication of expectations of future changes in the FOMC’s target for the federal funds rate comes from investors who buy and sell federal funds futures contracts. (We discuss the futures market for federal funds in this blog post.) As we’ve noted in earlier blog posts, since the weak July jobs report, investors have assigned a very high probability to the committee cutting its target by 0.25 percentage point (25 basis points) from its current range of 4.25 percent to 4.50 percent at its September 16–17 meeting. This morning, as the following figure shows, investors raised the probability they assign to a 50 basis point reduction at the September meeting from 0 percent to 14.2 percent. Investors are also now assigning a 78.4 percent probability to the committee cutting its target range by at least an additional 25 basis points at its October 28–29 meeting.