TRAMA / WORK IN NUMBERS
Technology at work.
People behind the numbers.
Who organises the working day? Does work adapt when health calls for a change? Explore the experiences of workers across Europe.
See the pictureWho is included? Employees who reported a health condition or problem and answered the questions used here. These figures describe this group. About the study ↗
Organising the day
The computer also manages work.
People report computerised systems doing three kinds of management. A person can experience more than one.
Reported to some or a large extent. Select a function to see it in the picture and compare profiles below.
When health calls for change
Does the workplace adapt?
The survey asked whether the employer accommodates an illness or health problem. These are the four answers.
Different working lives
A closer look at people.
An average can hide different experiences. Explore the profiles without treating small differences as a ranking.
Computerised management · Assesses performance
The dot is the estimate. The line shows its uncertainty. Wider lines mean less precision. A dash does not mean zero.
Putting the pieces together
Is there a connection?
We compared people reporting more and less computerised management, taking their social position, health and employment into account.
No adaptation, despite needing it
35 countries
pp = percentage points. Zero means no difference. A line crossing zero is compatible with a difference in either direction.
Read with context
What is behind these charts?
Exploratory TRAMA analysis of Eurofound’s EWCS 2024. This is work in progress, not a Eurofound publication.
Charts only. No individual records.
The website receives only the aggregate estimates needed for the selected charts. There are no data tables or export files.
Who was included?
22,652 employees in 35 countries, including 18,009 in the EU27, eligible for the health questions and with valid answers to workplace adaptation (Q94) and the three computerised management questions (Q56). The six gender × migration profiles include 22,592 people. Other demographic views exclude missing responses only for the selected characteristic.
What do the questions measure?
Computerised management means systems allocating tasks, determining working time or monitoring performance to some or a large extent. It does not identify a particular algorithm or prove the use of AI. Adaptation concerns an employer accommodating an illness or health problem. Partial adaptation is kept separate from no adaptation despite need. The survey does not establish whether adaptations were requested, refused or effective.
Why are some results missing?
We withhold an estimate if the sample is below 100, either outcome count is below 10, or survey degrees of freedom are below 2. If one adaptation state fails, all four are withheld. A dash means unavailable, never zero. Intervals wider than 20 percentage points receive a warning in the selected profile.
How is diversity represented?
Migration refers to the released generation variable, not race, ethnicity or legal status. Gender follows survey self-identification. The “in another way” category is retained, but is too small for separate estimates. Age groups are 16–34, 35–49 and 50–74. Differences between point estimates alone do not establish differences between groups.
Statistical methods and source
Proportions use the calibration weight (calweight), strata and sampling points nested within country, with survey-domain estimation. Logit confidence intervals use survey degrees of freedom; singleton units are handled by adjustment. Pooled figures use the supplied weights and are not equal-country averages. All intervals are pointwise 95% and are not adjusted for multiple comparisons.
The association chart uses model M2: the three management functions, gender × migration, country, age, health conditions and limitations, six health problems, occupation, contract, sector and workplace size. Marginal probabilities are standardised over complete cases: 21,188 in all countries and 16,927 in the EU27. More use means some/a large extent; less use means a little/not at all. These are associations, not causal effects.
EWCS 2024 · study documentation ↗