Twelve years, or none

One · The interval

The years nobody counts

Eurostat publishes the age at which young people leave the parental home, and the age at which they first marry. It publishes them separately, in different tables, gathered by different services. Subtract one from the other and you get a quantity nobody reports, which says more about a society than either number alone: how many years a person spends in a home that belongs neither to their parents nor to a marriage.

This interval holds everything we associate with young European adulthood — the shared flat, the moves, the relationships that go nowhere, the year abroad, the bad cooking for one. It is the stretch in which someone learns to be a household by themselves, before becoming half of a joint one.

In one part of Europe the interval runs well over a decade. In another it does not exist. Croatia's is negative: the average woman marries a few months before she leaves her parents' house, passing directly from a household run by someone else into one she founds. Slovakia has two months. Serbia has ten.

Romania sits at 3.1 years, between Poland and Greece. A Romanian woman leaves her parents at 25.3 and marries at 28.4. A Swede leaves at 22.6 and marries at 35.2, and the twelve and a half years between those dates are the longest such stretch in Europe.

The ordering overturns an expectation. Swedish women marry latest in Europe, tied with Spaniards, and Romanian women marry among the earliest. Judging by marriage age alone, you would say Sweden postpones family formation and Romania does not. Add the second column and the picture changes: the Swede had left home three years before the Romanian. She takes her independence early and postpones only the wedding.

Two · The line

What Hajnal saw in 1965

In 1965 the demographer John Hajnal published an essay that changed how Europe reads its own family history. Comparing censuses around 1900, he noticed that the continent's marriage behaviour breaks in two along an imaginary line drawn between Saint Petersburg and Trieste.

West of it, women married in their mid-twenties, and among those aged 20 to 24 only a quarter were married. As many as a third never married at all. East of it, over three quarters of women that age were already married, and fewer than five per cent stayed single for life.

At the heart of it lay a rule of organisation. In north-western Europe a marriage required a new household with its own hearth, and a household required capital. Young people left home around fifteen and entered other people's houses as servants, farmhands or apprentices, where they spent years accumulating money and skill. They married when they could afford a house, which made them late, and those who could not afford one never married. East of the line, marriage required no new house: the bride joined her husband's parents' household, where several related couples already lived. Because nothing had to be saved up beforehand, people married early and almost universally.

The western rule required a marriage to produce a household. The eastern rule required a marriage to fit inside an existing one.A summary of the mechanism Hajnal described, 1965

What matters is that the two things travelled together. Marriage timing and household form were two faces of one rule, and a single line separated both. That is the testable claim, because Eurostat measures those same two quantities today.

Four · The break

The package came apart

The two things Hajnal found together have stopped being the same thing. Marriage timing stayed where it was, on a line running from north-east down to south-west. Household form slid onto a horizontal line dividing north from south. Roughly sixty degrees separate them.

The countries that changed sides make it clearest. Estonia and Lithuania lie deep east of Hajnal's line, and in 1900 they married early and lived in complex households. Today their young people leave home at 22.7, within a year and a half of Finland and ahead of Sweden. Lithuania has the highest share of one-person households in Europe, 55.7%. Running the other way, Italy and Spain sat comfortably in Hajnal's west, and their young people now leave home at 30.2, within eighteen months of Croatia.

The distance itself has not shrunk. A young Croat leaves the parental home at 31.5 and a Finn at 21.3, and those ten years and ten months are the widest measurable spread in Europe on any age indicator. Between 25 and 34, some 62.3% of Croats live with their parents and 3.9% of Danes, a ratio of sixteen to one. Slovakia has households of three or more adults in 22.6% of cases, Finland in 1.4%, sixteen to one again.

Age of leaving the parental home, 2025

Five · The test

It is not the money

The explanation that comes to mind first is that young people in the south and east stay at home because they cannot afford anything else. It has the advantage of being testable, and the test rejects it.

The correlation between latitude and the age of leaving home is −0.76 across the 31 countries with data. Holding gross domestic product per head constant, it moves to −0.73, which leaves it essentially untouched. In the other direction, the correlation between income and leaving home is −0.35, and it collapses to −0.18 once latitude is held constant. Wealth explains little of what geography has not already taken.

The other three material explanations do worse. The market rental sector, the share of owner-occupiers and the overcrowding rate all leave the latitude effect intact, and in fact strengthen it when held constant.

Pearson correlations and partial correlations, computed on the countries with complete data for each pair. A correlation shows agreement between two columns of numbers and does not by itself establish a cause; what the table shows is that the material variables tested do not absorb the geographic effect, which therefore remains to be explained.

One institutional difference runs the same way, though it was not put through this test. Continental European civil codes retain a reciprocal duty of maintenance between relatives in the direct line: article 516 of the Romanian Civil Code establishes it between spouses, direct-line relatives and siblings, and article 519 places the descendant ahead of the ascendant in the order of payment. France has it at article 205, Germany for the whole direct line. Sweden, Denmark and Norway have no equivalent legal duty of adult children towards elderly parents, because their care is a public task funded from taxation. Where the law says the family answers for its members, the household has one more reason to stay together.

Six · Time

The distance is not closing

Twenty years of a common market, mobility programmes and free movement of workers should have brought the two Europes together. The average age of leaving the parental home in countries above 54 degrees north and below 47 degrees shows the opposite.

In 2010 the distance between the two averages was 5.5 years. In 2025 it is 6.3. It climbed to 7.0 years in 2020 under the pandemic closures, and has come down since without returning to where it began. What the chart shows is a gap widening slowly rather than one fading out.

Seven · The meaning

The same roof, two different things

A thirty-year-old man living with his parents is the same thing in Helsinki and in Valletta only if you look at the address. Eurostat also asks what these people do, and the answer separates two societies.

Among Maltese aged 25 to 34 who live with their parents, 87.3% hold a full-time job. In Romania, 72.2%. In Finland, 39.9%, while 22.9% of them are unemployed and 18.5% are students.

In the southern half of the continent, living with your parents is the ordinary condition of a working adult. In the north it is a stage of passage or a sign that something has gone wrong — university, a lost job, a separation. The same box ticked on a survey form, two social positions with nothing in common. For the Finn of thirty, moving back home is news you deliver with embarrassment. For the Maltese of thirty, moving out before the wedding would be the news.

This is why the figures are hard to read from outside. A Nordic journalist writing about the generation that will not leave home is describing a problem, because in that society leaving at twenty-two is the norm and any departure from it demands explanation. A Croatian journalist writing the same piece would be describing the ordinary life of most of their thirty-year-old compatriots.

Living with parents, aged 25–34, 2025

Eight · Romania

The inherited house

Romania has 93.2% owner-occupancy, the second highest rate in Europe after Slovakia and nearly double Germany's 47.2%. It also has one of the shortest intervals of independence on the continent. Each fact explains the other.

In a market where housing is inherited or built by the family's own labour, the moment of founding a household turns on a family decision about who gets the house and when, more than on a salary accumulated over years. That shortens the interval from both ends: no decade of saving is needed before marriage, and housing that can be rented cheaply, briefly and without guarantees barely exists.

The eastern rule Hajnal described has moved to another part of Romanian life. Today's young Romanian no longer brings a bride into his parents-in-law's house, yet he stays in his parents' house until close to the wedding, marries among the earliest in Europe at 28.4, and has a first child at 27.2, the earliest in the Union after Bulgaria. What changed since 1900 is the number of couples under one roof. What stayed is the order of events.

The 3.1 years a Romanian woman has between the parental house and marriage follow from those two thresholds having stayed fused, as they were fused everywhere in eastern Europe a century ago, while in the north they drifted a decade apart. About attachment to family these data say nothing, because they do not measure it.

Years between the parental house and marriage, women

Nine · Method

How the figures were calculated

All values come from Eurostat's public interface, queried on 5 September 2026, at the latest year available for each series: 2025 for household indicators from the EU-SILC survey and 2024 for vital statistics. Each country's coordinates are those of its capital.

The line test works as follows. Each capital's longitude and latitude are converted to kilometres, with longitude multiplied by the cosine of 48 degrees so that the two axes share a scale. For a given line, each country takes a value equal to the projection of its position onto the direction perpendicular to that line, which is its signed distance from it. The correlation between those distances and the indicator's values measures how well the line divides the map. The percentage shown is the ratio between the chosen line's correlation and the highest correlation any line achieves, searched degree by degree.

The interval between leaving home and marriage subtracts the mean age at first marriage for women, from civil registration, from the estimated age of leaving the parental home for women, from EU-SILC. These are two different instruments applied to two overlapping but non-identical populations, so the difference measures the distance between two social thresholds in a country, at that country's scale, rather than any individual's path. The ranking survives that caveat, because measurement discrepancies are far smaller than the thirteen years separating Sweden from Croatia.

The data cover 36 countries, with coverage varying by indicator: 31 countries have the age of leaving home, 30 have household structure, 23 have age at first marriage and 21 have both, and therefore the interval. Hajnal's figures for 1900 are those in his 1965 essay and the literature that followed it; they were not recomputed here.