On an imagined workbench, screwdrivers lie beside a keyboard. The picture connects activities that may require very different places in daily life. It is easy to discuss changing work through a broad “we”. One specific statistic helps us become more precise: which people does it count, what period does it describe, and what exactly does working at home mean?
A quarter — of what?
On 22 April 2026, Destatis reported that around 25 percent of employed people in Germany worked from home at least occasionally during 2025. This means doing so at least once in the four weeks preceding the survey. The figure counts people. It does not say a quarter of all working hours were performed at home. That distinction matters: a person with one such workday can belong to the same group as someone working at home every day. [1]
A comparison with an earlier working life
The same release gives a figure of around 13 percent for 2019. The gap between that and 25 percent in 2025 reveals change, but remains a comparison of two dated shares. The years therefore belong with the figures. We are not reading a live counter for October 2026. Nor can the gap alone explain which organisational decisions, personal wishes or other circumstances caused each individual change of workplace. Recognising change and fully explaining its causes are different tasks. [1]
Four shares, two denominators.
All figures come from the same Destatis release. The last two cards count only people using home working.
- All employed people · 2019
13% at least occasionally
Share of people working at home, not share of all working hours. [1]
- All employed people · 2025
25% at least occasionally
At least once in the four weeks preceding the survey. [1]
- Home-working users only · 2025
24% every working day
One frequency group within the population using home working. [1]
- Home-working users only · 2025
46% on fewer than half
This means fewer than half of their own working days. [1]
Rounded figures for Germany. 2019 and 2025 are selected comparison years. The frequency cards are not exhaustive and therefore do not add to 100 percent. [1]
Home working is not one arrangement
Among people using home working in 2025, Destatis reports that 24 percent did so every working day. Meanwhile, 46 percent of this group worked at home on fewer than half their working days. The denominator has changed: these are not 24 and 46 percent of all employed people. They describe different frequencies within the home-working subgroup. They also do not exhaust every frequency category. The familiar phrase already conceals several arrangements rather than one uniform experience. [1]
The industry changes the picture
The release reports around 74 percent using at least occasional home working in information technology services, compared with around 6 percent in food and beverage services. Each percentage concerns the employed population within that industry. This comparison shows why an overall share cannot represent one experience shared by every worker. Nor does it measure the quality of a job. Where work happens is one answerable question; whether people enjoy it or have enough autonomy would require a different investigation. [1]
How a picture like this is produced
The findings come from the German microcensus. Destatis describes an annual household survey, running since 1957 and covering roughly one percent of the population. Its methodological overview explains the redesign from 2020 and points towards further quality information. Established procedures turn many answers into population findings. A table is therefore neither a collection of arbitrary anecdotes nor an immediate window into every workplace. It is one defined statistical approach to an aspect of working life. [2]
What our imagination may leave out
When we hear “work today”, a laptop might immediately come to mind. That private choice of image is not a description of all work. The workbench in our illustration assembles objects for an article; it represents no statistical sample. A mental cross-check can help: which activities around us happen in fixed locations, and which in changing ones? Even this question opens attention to variety, without trying to derive fresh percentages from people we happen to encounter.
The number belongs with its question
The four weeks preceding a survey are not an entire working life. A percentage of people is not a percentage of hours. One industry is not the whole employed population. These distinctions are not statistical small print taking the place of a story. They form part of the story itself. They show how easily a sentence about a measured slice of reality can become a much larger claim that the original measurement no longer supports.
A more precise address for change
The interesting statement is therefore more specific than “we work differently”. In Germany, the share of employed people working at home at least occasionally was higher in 2025 than in 2019, with substantially different patterns. Questions about travel time, cooperation or personal strain require additional suitable data. The initial figure remains useful. It becomes sharper when we recognise the further questions it raises, rather than expecting it to provide every answer at the same time. [1]
The workshop is an artistic arrangement, not a documented business. The cards compare published shares with explicitly different denominators; they do not form an uninterrupted time series.
Destatis · Anteil leicht gestiegen: 25 % der Erwerbstätigen arbeiteten 2025 im Homeoffice ↗
Release N024 dated 22 April 2026, reporting initial 2025 microcensus results. The rounded 25 percent concerns employed people in Germany working at home at least once during the preceding four weeks. The 24 and 46 percent figures concern only home-working users. Industry percentages concern employed people within each industry. Not a measure of all working hours or findings about working arrangements in 2026.
Destatis · Der Mikrozensus seit 2020 ↗
Institutional methodological overview of Germany’s annual household survey covering roughly one percent of the population and running since 1957. Describes the redesign from 2020 and links quality reports. Not an independent second measurement of home-working shares, nor proof that every annual figure is fully comparable.
Sources rechecked in the integrated browser on 24 October 2026. The working arrangements reported concern survey year 2025, published in April 2026. Distinct denominators are identified explicitly.
