US Hiring Rebounded by 162,000 Jobs—but Information Roles Still Fell

U.S. nonfarm payrolls rose by 162,000 in August, unemployment held at 4.1%, and information employment fell by 23,000 in the September 4, 2026 account of the federal jobs report. The combination makes the headline rebound real but uneven: nationwide hiring strengthened while a technology-adjacent industry continued to shed positions.

The report also separates payroll growth from broader labor-market security. More jobs appeared on employer payrolls, but that does not mean demand improved in every industry, that each added position went to an unemployed person or that workers gained purchasing power.

The rebound followed a weak and revised summer

August represented a marked acceleration from early summer, but the preceding months were also stronger than initially estimated: the revised payroll and participation figures raised June and July employment by a combined 55,000, changed July from a loss of 23,000 to a gain of 21,000 and put labor-force participation at 61.6% in August. Those revisions reduce the appearance of an abrupt collapse before the rebound, though they do not establish a sustained hiring acceleration.

The latest payroll estimate remains preliminary and can change as more employers respond and seasonal factors are recalculated. That makes August a strong monthly result rather than definitive evidence of a new labor-market trend.

Payroll employment and unemployment also come from different surveys. The establishment survey counts jobs at nonfarm employers, so a person holding multiple jobs may appear more than once. The household survey classifies people by employment and labor-force status, which is why a rise in payroll positions is not equivalent to the same number of unemployed people finding work.

The gains bypassed information businesses

The worker-focused sector map is sharply divided: the BLS industry and earnings details show gains of 59,000 jobs at food services and drinking places, 42,000 in local government education, 16,000 in manufacturing and 13,000 in health care, against a 23,000 decline in information that included losses of 8,000 in computing infrastructure, data processing, web hosting and related services, 7,000 in publishing and 5,000 among broadcasting and content providers; average private-sector hourly earnings rose by 10 cents, or 0.3%, to $37.75, leaving annual growth at 3.1%, while production and nonsupervisory workers’ average rose by 11 cents to $32.53.

That composition matters for workers because restaurants, public schools and factories have different staffing needs from publishers, hosting providers, broadcasters and data-processing companies. Hiring can improve nationally without producing more openings for people seeking comparable work inside information businesses.

“Information” is an industry classification, not a list of occupations. It covers technical, editorial, sales, administrative and operational employees at businesses assigned to that industry. A software developer employed by a bank, manufacturer or hospital is instead counted under the employer’s industry.

The contraction therefore gives a direct signal about employment at information companies but only a partial view of the wider technology workforce. It also does not identify the reason jobs disappeared: the payroll survey measures where employment changed, not whether weaker demand, restructuring, automation or another factor caused the decline. Research on AI adoption and work similarly cautions against treating evidence of changing tasks as proof that a particular technology eliminated whole jobs.

Slower wage growth weakens the headline’s reassurance

Average pay continued to increase, but the pace was modest compared with the hiring headline. The earnings figures are nominal, meaning they measure changes in pay before accounting for inflation. Whether workers gained purchasing power depends on how prices changed over the same period.

Average hourly earnings are also affected by the composition of employment. Strong hiring in relatively higher- or lower-paying industries can move the aggregate average even when an individual worker’s wage does not change. The measure describes the direction of economy-wide pay, not a raise received uniformly across the workforce.

For workers negotiating compensation or considering a move, slower aggregate wage growth suggests less momentum in pay-setting than the payroll increase alone might imply. It does not prove that every occupation has lost bargaining power, but it limits how much reassurance can be drawn from the national hiring total.

Stable unemployment did not mean equal security

An unchanged unemployment rate describes the labor market as a whole. It does not measure how quickly someone leaving a contracting information employer can find a role with similar pay, responsibilities or working conditions.

The distinction is especially important when hiring is concentrated. Workers attached to expanding industries may see more vacancies even as people in information businesses encounter fewer openings and greater competition. Both experiences can coexist with the same national unemployment rate.

The household and payroll figures also capture different forms of security. Someone can remain employed while facing reduced leverage, fewer external opportunities or concern about future cuts. Conversely, an expanding payroll total can reflect additional jobs without showing whether those positions are full time, durable or matched to the skills of people currently searching.

The recovery signal remains provisional and uneven

The clearest reading is that August hiring recovered at the national level while information employment moved in the opposite direction. The report is more encouraging than the first estimates for early summer, but it does not establish an economy-wide hiring boom or broad security for technology and knowledge workers.

Future payroll revisions, subsequent industry estimates and inflation data will show whether the acceleration persisted and whether nominal wage gains improved purchasing power. For now, the headline and the sector breakdown need to be read together: employers added jobs overall, while an important cluster of information businesses continued to contract.

Also read:

• AI Use Is Spreading, but the Evidence Still Measures Tasks—not Vanishing Jobs
• Private Hiring Slowed to 44,000—Job Switchers Still Won 7% Pay Growth
• Podcasting Gives Startups Mass Reach—but Downloads Still Cannot Prove Growth