One tenth of a year is not a trend, but the number underneath it is. Eurostat reported on 15 September that young people in the European Union left the parental home at an average age of 26.3 years in 2025, up from 26.2 in 2024. Since 2002 the figure has barely moved from 26, which makes the headline the least interesting thing in the release.
The spread is where the story is. Croatia records the latest departure at 31.5 years, followed by Greece and Slovakia at 30.9 and Spain and Italy at 30.2. At the other end, Finland sits at 21.4 years, Denmark at 21.8, and Estonia and Lithuania at 22.7. That is a gap of more than ten years between one end of the single market and the other, wider than the gap in most headline economic indicators the Union tracks.
Eurostat pairs the figure with youth employment, and the correlation is clean enough to be uncomfortable. The employment rate for people aged 20 to 29 stood at 65.5 per cent across the EU in 2025. In countries where young people move out early, such as the Netherlands, Denmark, Germany and Sweden, youth employment runs above that average. In Greece, Croatia, Spain and Italy, where departure comes late, it runs below.
The methodology deserves a moment, because it is easy to misread. The indicator is not a survey question about when someone moved out. It is derived from the labour force survey by calculating, for each single year of age from 15 to 34, the share of respondents living in households without their parents, and identifying the age at which half the cohort has crossed that line. It is a modelled median, not a recollection, which makes it robust to memory but blunt about what actually happened: a person who left at 19, returned at 24 and left again at 28 appears only as a shifting probability.
Two days later Eurostat published the September edition of its European Statistical Monitor, which reported employment in the EU at a historic high while unemployment and labour market slack held steady and the job vacancy rate eased slightly. Industrial production, retail trade volume and services output each recorded mild monthly declines while still growing year on year.
Read together, the two releases describe a mismatch that employment statistics alone cannot capture. Record employment across the Union coexists with an age of household formation that has not fallen in more than twenty years, and with a ten-year national spread that has proved immune to two decades of convergence policy. A job explains part of the difference. It does not explain the rest, and the release is careful not to claim that it does.
The gap points at what the labour force survey was never built to measure: the cost of a first lease, the availability of the stock, and whether a twenty-six-year-old in Zagreb or Athens is choosing to stay or simply has nowhere affordable to go. The Commission’s housing work has begun to take that question seriously. The statistics have been flagging it for twenty years.
