How many jobs were added in August 2026?

Total nonfarm payroll employment rose by 162,000 in August 2026, well above the prior 12-month average of 31,000, and the unemployment rate held at 4.1 percent. However, two categories account for 101,000 of that gain: food services and drinking places added 59,000 against a 12-month average of 12,000, and local government education added 42,000, largely offsetting a 50,000 decline in July. Excluding both, underlying job growth was roughly 61,000.

What is functional unemployment and the True Rate of Unemployment?

The True Rate of Unemployment (TRU) is a measure published by the Ludwig Institute for Shared Economic Prosperity that counts the share of the labor force that is functionally unemployed: people with no job, people who lack full-time work but want it, and people earning below roughly 26,000 dollars per year. TRU stood at 24.9 percent in July 2026 against a headline unemployment rate of 4.1 percent, and it has risen for four consecutive months while the official rate fell.

Why is long-term unemployment rising in 2026?

Long-term unemployment, defined as being jobless for 27 weeks or more, rose by 159,000 in August 2026 to 1.93 million, or 27.0 percent of all unemployed people. The ratio of newly unemployed to long-term unemployed has fallen to 1.04 to 1, compared with about 1.6 to 1 in 2019. Hiring in August was concentrated among workers who had recently left the labor force rather than those unemployed for extended periods, so people who lose work are taking progressively longer to return.

Which sectors have the highest and lowest wage growth in 2026?

Average hourly earnings rose 3.1 percent over the year to August 2026, but growth is highly uneven by sector. Utilities led at 6.8 percent and information at 5.2 percent, sectors employing roughly 620,000 and 2.9 million workers respectively. Private education and health services, which employs about 32 million people, had the lowest wage growth in the economy at 1.8 percent, followed by professional and business services at 2.4 percent with roughly 25 million workers.

Were the July 2026 jobs numbers revised?

Yes. The August 2026 release revised July upward by 44,000, from a reported loss of 23,000 jobs to a gain of 21,000, and revised June upward by 11,000, from 31,000 to 42,000 in combined terms — 55,000 more jobs than previously reported across the two months. This was the first upward revision cycle since spring 2026, following several months in which prior estimates were revised down.

Are the monthly jobs report numbers reliable?

First estimates carry substantial uncertainty. The BLS establishment survey has a 90 percent confidence interval of plus or minus 122,000 on a one-month change, meaning May, June, and July 2026 were all statistically indistinguishable from zero job growth. Monthly figures are also revised twice: May 2026 fell by 109,000 across two revisions while July 2026 rose by 44,000. Separately, the August 2026 preliminary benchmark revision, measured against employer tax records, found private employment overstated by 178,000 through March 2026.

Returns, Raises, and Who It Didn’t Quite Reach. Also Known As: the August Jobs Report

Jillian O’Malior

Sep 4

7 min read

By Jillian O'Malior, Founder & Principal Strategist, New World Labs

Let me start by giving August its due, because it earned some.

The economy added 162,000 jobs. That’s five times the prior 12-month average of 31,000, and the first monthly figure all summer large enough to clear the survey’s own margin of error. Unemployment held at 4.1%. Participation rose, quits rose. Involuntary part-time work fell by 414,000. Prior months got revised up for the first time in half a year.

I’ve spent most of this year writing that the labor market was contracting in ways the headlines weren’t reporting. But in August, some of that reversed; I’m not going to bury that.

But two different things happened in the same report, and likely only one of them is getting covered.

Two Labor Markets, One Report

Here’s what improved:

People came back. The labor force grew by 683,000 in August; the first meaningful re-entry all year. Employment rose 569,000. The number of Americans outside the labor force fell by 551,000. Participation ticked up to 61.6%.

The behavioral signals moved too. Job leavers (people who quit voluntarily, betting something better exists) jumped 121,000 to 914,000. Discouraged and marginally attached workers fell.

That’s real, positive shifts. People who drifted to the edges of the target recently got pulled back in.

Now here’s what didn’t move:

Long-term unemployment rose by 159,000 to 1.93 million—27.0% of everyone unemployed. The ratio of newly employed to long-term unemployed is now 1.04 times to 1. That is effectively parity. In 2019 the ratio was 1.6. Not only that, nearly 44% of unemployed Americans have been out fifteen weeks or longer.

So the labor market reabsorbed the people who left recently and did not reach the people who have been out the longest. August made clear a story that those struggling for awhile already know: recovery has a doorway, but they’re at the back of the line to get through.

45 to 54

Every age group gained employment in August except one.

Workers 25 to 34: up 239,000.

Workers 55 and over: up 354,000.

Workers 45 to 54: down 230,000.

That cohort has now declined steadily five months in a row—31.86 million in April, 31.80, 31.74, 31.66, now 31.43 in August. Over twelve months it is down 954,000, and that number has been getting worse, not better, over the course of the year.

Peak earning years. Two decades of experience. Mortgages, tuition, growing children, aging parents, all at once. This is the population I’ve been describing for awhile now as being met mid-life with erasure, and it is the only group in this economy that this particular “good” month didn’t touch.

What the 162,000 Is Made Of

Two line items:

Food services and drinking places: +59,000, against a 12-month average of 12,000. Five times trend. We have seen this model before: it’s what leisure and hospitality did in May before the World Cup unwound and gave it all back.

Local government education: +42,000, which BLS says “largely offsets” July’s 50,000 decline. Meaning both months were seasonal-adjustment artifacts, not events.

Those two categories are 101,000 of the 162,000. Strip them out and the underlying number is roughly 61,000—right back in the range of where the rest of the summer has lived.

Who is Actually Getting Raises

Average hourly earnings hit $37.75, up 3.1% over the year. That’s the third straight month of deceleration: 3.5%, then 3.2%, now 3.1%.

But the average is hiding a bit part of the wage story. Here is wage growth by sector, next to how many people each sector employs:

Sector	Wage growth, Aug 2025 > Aug 2026	Workers
Utilities	+6.8%	~620,000
Information	+5.2%	~2.9 million
Financial activities	+4.3%	~9.4 million
Construction	+4.2%	~7.8 million
Manufacturing	+3.8%	~13.4 million
Leisure & hospitality	+3.6%	~16.5 million
Retail trade	+3.1%	~15.5 million
Professional & business services	+2.4%	~25.3 million
Private education & health services	+1.8%	~32.0 million

Read the two ends of that table together.

The sectors handing out 5 to 7% raises employ about 3.5 million Americans. The two sectors at the bottom employ 57 million between them.

Private education and health services; 32 million workers, the sector every analyst including me has called the durable engine of this labor market delivered the smallest raise in this economy at 1.8%, while its hiring decelerated for the third consecutive month.

August’s job mix said the same thing in miniature: 59,000 jobs added in food services at $603 a week, on an implied 25-hour schedule. 23,000 jobs lost in information at $2,066 a week.

So, sure: wage growth is real. It’s happening almost exactly where the fewest people are employed.

What 4.1% Actually Measures

I’ll be honest about where I’ve landed mentally after a year of writing these reports.

I don’t think the headline number is dishonest necessarily. I think it’s narrow, and I think it’s far less reliable than the confidence with which it gets reported.

Last week’s benchmark revision, measured against employer tax records rather than a survey, found private employment overstated by 178,000 through March. In 2026 the first monthly print has been revised by 109,000 in one direction (May) and 44,000 in the other (July). The issue here isn’t bias, it’s that first prints are fundamentally not load-bearing. And the establishment survey’s margin of error on a monthly change is +122,000, which meant May, June, and July were all statistically indistinguishable from zero.

Then there’s the narrowness. The Ludwig Institute published a True Rate of Unemployment that counts anyone in the labor force who lacks full-time work but wants it, has no job at all, or earns below $26,000 a year. By that measure, functional unemployment in July was 24.9%. The BLS headline that month was 4.1%.

A 20.8 gap on the surface could look like a scandal, but ultimately it’s two different measures counting two different things on purpose. What matters more is the direction with the two measures: TRU has risen four consecutive months. Over that same stretch, the official rate fell.

The breakdowns are worth sitting with: women at 31.0% versus men at 19.5%. Americans without a high school diploma at 50.3%. And people with some college at 29.5%; worse than people with a high school diploma alone.

Two caveats, stated plainly: TRU is measuring something far broader than joblessness by design, and LISEP publishes a month behind BLS, so that 24.9% is July against August’s 4.1%. Neither weakens the point though. When the narrow measure improves and the broad measure worsens for four straight months, the chasm between them is the finding.

Two Things I Need to Update

In July I wrote that full-time unemployment was down 1.29 million over twelve months and that the economy was converting full-time jobs into part-time ones. Full-time employment rose 735,000 in August, and the year-over-year gap narrowed to 180,000. Part-time is now down 502,000 over the year. That argument isn’t holding this month, so I’m retiring it rather than defending it.

Multiple jobholding needs a smaller fix. That level keeps climbing (8.81 million, up 448,000 since March) but August 2025 was already elevated, so the year-over-year framing I used doesn’t work on this base. The six-month trend is the more realistic version.

The Engine and the Exit

Healthcare added 13,000 against a 12-month average of 32,000. That is the third consecutive month below trend and the slowest yet. The sector everyone treats as recession-proof is decelerating in plain sight.

Information lost 23,000 against average monthly losses of 8,000, roughly three times its own trend. Computing infrastructure and data processing shed 8,000, publishing 7,000, broadcasting 5,000. Notable because last week’s benchmark revised that sector’s employment level up by 87,000. The base was higher than we knew, but the decline is accelerating anyway.

Reading the Revisions

For the first time since spring, they went up. June moved from +20,000 to +31,000. July moved from -23,000 to +21,000; from a month the economy lost jobs to a month it didn’t.

I built a good deal of last month’s analysis on that -23,000; it no longer exists. That’s the deal with tracking revisions: you report them when they cut against you the same week you report them when they don’t.

The three-month average now stands at 71,000. Better than the 20,000 I reported in August. Still roughly half of what it takes to keep pace with population growth.

What I’m Watching

The final benchmark revision lands in February 2027, and it will re-anchor everything above the tax records.

Before that: whether the 45 to 54 cohort breaks a five-month losing streak, and whether food services gives back its 59,000 the way leisure and hospitality gave back the World Cup.

August was a better month. Roughly 683,000 people can back to look for work and most of them found it, which is the most encouraging sentence I’ve been able to write in this series all year.

But it wasn’t a better month for the 1.93 million people who have been out of work for more than six months. Or the 32 million working in the sector with the smallest raise in the economy. Or the 954,000 mid-career workers who left the employment count in the last twelve months and haven’t come back.

One labor market got a recovery in August. The other is still waiting for its silver lining.

Key Data At-A-Glance

Indicator	Aug 2026	Aug 2025	Direction
Nonfarm payrolls (monthly)	+162k	—	↑↑
Unemployment rate (U-3)	4.1%	4.3%	↓
3-month average (Jun-Aug)	+71k/mo	—	↑
True Rate of Unemployment (LISEP, July)	24.9%	—	↑ 4 mo.
Labor force participation	61.6%	62.3%	↓YoY, ↑m/m
Labor force (monthly change)	+683k	—	↑↑
Not in labor force	105.6M	103.3M	↑
LT unemployed (% of total)	27.0%	26.1%	↑
New-to-long-term unemployed ratio	1.04 : 1	—	↓↓
Employed, ages 45-54	31.43M	32.38M	↓↓
Job leavers (voluntary quits)	914k	785k	↑↑
Part-time for economic reasons	4.39M	4.76M	↓↓
Multiple jobholders	8.81M / 5.4%	8.80M / 5.4%	→
Avg hourly earnings (all private)	$37.75	$36.62	+3.1% YoY
Private ed & health wage growth	1.8%	—	↓
Prior-month revisions	Jun +11k, Jul +44k	—	↑

Sources: BLS Employment Situation, August 2026 (USDL-26-1435), released September 4, 2026; Summary Table A, Tables A-9, A-12, B-3. True Rate of Unemployment: Ludwig Institute for Shared Economic Prosperity, July 2026. Benchmark figures: CES Preliminary Benchmark, USDL-26-1425.