WIESBADEN – October 30, 2026 -- The number of employed people living in Germany was unchanged in September 2026 on a seasonally adjusted basis, at about 45.60 million, but sat 211,000 (0.5%) below the level of a year earlier, according to preliminary Destatis figures.
The adjusted monthly change was -4,000 people (0.0%). From January to August 2026, employment had fallen by an average of 19,000 people per month, so the September reading marks a pause in that pace of decline.
Unadjusted employment rises 172,000 as autumn recovery begins
Without seasonal adjustment, the employed population grew by 172,000 (+0.4%) from August. That matches the typical seasonal gain: the September average for 2023 to 2025 was +170,000.
Year-on-year decline extends the downtrend seen since January 2025
The annual rate of change was -0.4% in each month from January to April 2026 and -0.5% in each month from May to August. September stayed at -0.5%, so the year-on-year decline has not narrowed.
Third-quarter employment falls 57,000 on the previous quarter
Averaged over the third quarter of 2026, about 45.67 million people worked in Germany (domestic concept, by place of work). That is 57,000 people, or 0.1%, fewer than in the second quarter on a seasonally adjusted basis. Detailed third-quarter results are due on November 18, 2026.
Adjusted jobless rate holds at 4.1% with 1.77 million unemployed
The Labour Force Survey counted 1.79 million unemployed people in September 2026, down 6,000 (0.3%) from September 2025. The unadjusted unemployment rate was 4.1%, against 4.0% a year earlier.
After adjustment for seasonal and irregular effects, unemployment stood at 1.77 million. That is 7,000 people (0.4%) more than in August. The adjusted rate was unchanged from the previous month at 4.1%.
Destatis revises January-August 2026 levels but leaves annual rates intact
As part of its routine revision of national accounts, Destatis recalculated the preliminary monthly and quarterly employment results for January to August 2026, drawing on all additional sources now available. The absolute levels were adjusted. The previously published year-on-year rates were not.
Employment figures come from the employment accounts, while unemployment figures come from the Labour Force Survey, which is integrated into the microcensus. Both follow the International Labour Organization definition, under which anyone who worked at least one hour for pay in the reference period counts as employed. The resulting unemployment figure therefore differs from the registered unemployment published by the Federal Employment Agency.