US economic outlook: September payroll weakness is expected to be seasonal, while sticky inflation keeps the Fed on a hawkish path

Overview

This US macro weekly argues that a soft September payroll headline should not be read as evidence of broad labor-market deterioration. The institution sees resilient demand, sticky inflation and a hawkish Fed backdrop, supporting its forecast for further rate increases in October and December.

Core views

Bank of America forecasts September nonfarm payroll growth of 60k, including 50k private-sector jobs, below consensus and down from August’s 162k. It argues that the apparent weakness will largely reflect reversal of unusually favorable August seasonal adjustment rather than a loss of underlying momentum. Non-seasonally adjusted August job growth was lower than a year earlier, but a near-zero seasonal adjustment—versus a 178k subtraction in August 2025—allowed most of the gain to pass into the headline. Benign jobless claims, improving ADP readings and broader labor indicators lead the report to estimate underlying job growth at 100k+. The report expects hiring to broaden across sectors despite the softer headline. Construction is expected to post a seventh consecutive monthly gain, aided by data-center investment; manufacturing momentum is linked to firmer demand, defense spending, reduced tariff uncertainty and AI-related investment. Professional and business services has added jobs for eight straight months after contracting in 10 of 12 months during 2025, which the report interprets as normalization after pandemic-era white-collar overhiring rather than AI displacement alone. Education and health services should remain steady on demographics and slower AI adoption, while trade and transportation may benefit from strong 2Q goods imports. Leisure and hospitality and local government education are expected to slow after contributing disproportionately to August’s seasonally adjusted upside. The labor-market details could be affected by the suspension of TPS-related work authorization for Haitian nationals. Roughly 200k Haitian TPS holders are estimated to have participated in the US workforce, with greatest exposure in food services, healthcare, transport and warehousing, and retail. The report’s base case is a gradual, mild payroll and participation headwind over months, cushioned by alternative legal status, employer adjustment and labor shortages; it sees little unemployment-rate effect. A larger-than-expected September effect remains a downside risk. The unemployment rate is forecast at 4.1% for a third month, although reversal of August’s 569k household-employment jump could round it to 4.2%; the report says even that outcome would be consistent with healthy underlying conditions. Participation is forecast at 61.6%, average weekly hours at 34.4, and average hourly earnings at 0.3% m/m. For inflation, the report forecasts August headline PCE at 0.4% m/m and core PCE at 0.3% m/m. Revisions to the treatment of portfolio management, computer software and accessories, and legal services are expected to reduce both headline and core year-on-year inflation by around 0.2pp, yielding forecasts of 3.6% headline and 3.2% core. The institution does not expect these revisions to alter the near-term policy path because the Fed was likely already accounting for them. It sees underlying inflation near 2.5% with little progress over the past year, while higher energy and food prices, AI-related investment, constrained labor supply, trade-policy uncertainty, the Iran war and weather-related supply shocks leave risks skewed upward. Growth remains resilient in the report’s view. Its 3Q GDP tracking estimate is unchanged at 3.0% q/q SAAR and 2Q tracking remains 1.8%. Lower-than-expected industrial production led to higher tracking for consumption and structures investment but lower equipment-investment tracking; stronger new-home sales and prior-month revisions lifted residential-investment tracking. The report lowered its 4Q/4Q 2026 growth forecast by 20bp to 2.1% because higher oil prices and reduced prospects of fully reopening the Strait of Hormuz should modestly restrain real consumer spending and capital expenditure. Even so, it identifies AI-related investment and consumer spending as key demand drivers and expects solid growth through its forecast horizon. Bank of America card data showed total spending per household up 6.9% y/y in the week ended September 19, while household assets exceeded 10 times liabilities in 2Q for the first time in six decades; even a 20% equity-market decline would leave the ratio at 9.5. The policy conclusion is explicitly hawkish. After the Fed raised rates 25bp in September, the report cites stronger growth, higher inflation, a lower unemployment rate in the SEP, and Chair Warsh’s emphasis on incomplete inflation progress. Markets were pricing nearly 100bp of rate increases over the next year and almost a 75% probability of an October hike. The institution believes September CPI will matter more than payrolls for the October decision and maintains its out-of-consensus forecast for 25bp hikes in October and December. It also notes that federal net interest expense reached a record 3.3% of GDP in 2Q 2026, but argues long-end yields may need to rise materially further before Washington pursues fiscal consolidation.

Analysis framework

The report separates temporary seasonal effects from underlying labor demand by comparing seasonally adjusted and non-seasonally adjusted payroll patterns, claims, ADP data and sector hiring. It then combines inflation forecasts, policy developments, GDP tracking from incoming activity data, spending data and household balance-sheet measures to assess growth and the Fed outlook.

Methodology notes

• Other

  Seasonal adjustment analysis

  The report compares seasonally adjusted and non-seasonally adjusted payroll changes to argue that August’s headline was unusually boosted and September may show a mechanical payback.

• Other

  GDP tracking estimate

  The institution mechanically aggregates incoming data that feed into BEA GDP calculations; it distinguishes this real-time tracking measure from its official GDP forecast.

Key data

• September nonfarm payroll forecast60kBelow consensus; compared with 162k in August.
• Underlying job growth estimate100k+Based on claims, ADP and broader labor-market indicators.
• September unemployment-rate forecast4.1%A rise to 4.2% is a risk if August household-employment strength reverses.
• August PCE inflation forecast0.4% m/m headline; 0.3% m/m coreMethodological revisions are expected to lower year-on-year headline and core readings by around 0.2pp.
• August PCE year-on-year forecast3.6% headline; 3.2% coreReflects expected methodological revisions.
• 3Q 2026 GDP tracking3.0% q/q SAARUnchanged; 2Q tracking remains 1.8% q/q SAAR.
• 4Q/4Q 2026 growth forecast2.1%Lowered by 20bp from the mid-year forecast.
• Federal net interest expense3.3% of GDP in 2Q 2026A record high in data available since 1984.
• Card spending per household6.9% y/yFor the week ending September 19.

Impact & implications

The report argues that a weak September payroll print should be interpreted cautiously because seasonal effects may obscure continued labor-market resilience. With inflation still above target and growth solid, it expects CPI rather than payrolls to be pivotal for the October Fed decision and maintains its forecast for two further 25bp hikes in 2026.

Risks

• TPS-related work-authorization changes could have a larger-than-expected negative effect on September payrolls and labor-force participation.
• The unemployment rate could rise to 4.2% if August’s surge in household employment reverses.
• Inflation risks are skewed upward from energy and food prices, AI-related investment, constrained labor supply, trade-policy uncertainty, the Iran war and weather-related supply shocks.

What to watch

• September payrolls, private payrolls, unemployment, participation, weekly hours and wage growth.
• August PCE inflation and the methodological revisions to PCE components.
• September CPI, which the report considers more important than payrolls for the October Fed decision.
• 2Q GDP revisions, goods trade, inventories, construction spending, ISM manufacturing and consumer data for GDP tracking.