McKinsey projects 11 million occupational changes in the United States by 2035

McKinsey projects around 11 million occupational changes in the United States between 2025 and 2035, in a study published on September 29, 2026. Its central scenario concerns workers who could need to leave an occupation in which labor demand would decline.

The estimate, reported the same day by Axios, concerns projected occupational transitions. It is neither a tally of layoffs nor a count of jobs already eliminated by artificial intelligence. The model incorporates automation and broader structural trends.

McKinsey’s summary presents around 6 to 16 million transitions depending on the scenario. The body of the report also mentions more than 16 million under assumptions of faster adoption or automation that reduces labor demand more substantially. These results are used to test the model’s sensitivity; they do not constitute guaranteed bounds or a confidence interval.

Automation affects tasks first

An occupation encompasses several activities involving judgment and coordination. For example, a tool can prepare a document that an employee then checks before handling exceptions and providing advice. Automating preparation therefore leaves other activities to be carried out.

The time freed up can be used to respond to more requests or improve service. Rising demand can offset some of the effects of automation. A reduction in the time needed for a task therefore has different implications depending on the needs the occupation serves.

Automatable hours, labor demand expressed in full-time equivalents, and occupational changes measure different things. A full-time equivalent represents a volume of work corresponding to full-time activity. The number of transitions depends, among other factors, on the share of automation that actually reduces labor demand.

McKinsey models the distribution of jobs

The firm extends its research by mapping American occupational classifications, used notably by the U.S. Bureau of Labor Statistics, to around 1 800 Lightcast occupations. Its model combines labor demand with skills and qualifications, then with training times and wages.

McKinsey says it uses overall employment growth of 3.1 % as an anchor for its calculations. This figure refers to the parameter chosen by the firm for its transition model covering 2025-2035. The study seeks to describe the distribution of employment across occupations, without producing an independent forecast of the total number of jobs.

Training durations are estimated, notably with the help of a large language model that synthesizes publicly available requirements. These estimates allow comparisons between the modeled pathways, without guaranteeing how long each worker’s training will take.

14 % of transitions are classified as direct

McKinsey classifies 14 % of transitions as direct, 41 % as more winding, and 45 % as facing greater obstacles. These proportions concern the roughly 11 million workers who would need to leave a declining occupation in the central scenario, rather than all American employees.

For McKinsey, a direct pathway leads to a growing occupation with substantial overlap in skills. The category also requires wages to be maintained and legally required qualifications to be attainable in less than six months.

Wage comparisons concern the base wages listed in job postings, excluding total compensation. They describe wages in the destination occupation, without measuring the compensation actually obtained after a transition.

Actual access to these occupations would also depend on the time available and local hiring opportunities. Funding for training and income during that period matter independently of wages in the destination occupation. Even a pathway classified as direct guarantees neither employment nor maintained income during training.