ENCS
Tuesday, 11 August 2026
Your paper, your pace.

Housing: what has actually worked, and what it cost

In 2025, 7.7% of people in the EU spent more than 40% of their income on housing. Five countries tried to move that number. Two of the attempts went backwards.

Checked on

All ten problems

In 2025, 7.7% of people in the European Union lived in a household that spent more than 40% of its disposable income on housing. That is Eurostat’s housing cost overburden rate, and it is the number this whole page is argued against. It runs from 2.4% in Cyprus to 26.4% in Greece. You can look it up yourself: the dataset is ilc_lvho07a and it is free.

Two warnings about the number before we use it. A low score often means a country full of people who inherited a paid-off flat, not a country where new housing is cheap. And Denmark’s 2025 figure of 23.4% is flagged by Eurostat as a break in the series — the survey changed, not the country — which is why Denmark is not on the board below.

Housing is unusual among big problems because almost nobody disputes the mechanism. A home needs land, permission, money and time. Change any of the four and the price moves. The argument is over who may change which one, and what breaks when they do. Below are five places that changed something and measured what happened, including a rent freeze that a constitutional court erased and a subsidy that turned into a price.

How this is measured

One number, named and sourced. Everything on this page is argued against it, so you can check us.

Housing cost overburden rate — share of people living in a household that spends more than 40% of its disposable income on housing %

Eurostat, EU-SILC, dataset ilc_lvho07a, 2025 reference year, dataset updated 8 June 2026

Who is actually ahead

Ranked on the number above — not on reputation.

Housing cost overburden rate — share of people living in a household that spends more than 40% of its disposable income on housing
Country Result Why
Cyprus 2.4 % Lowest in the EU for a second year. Most Cypriot households own their home outright, so there is no rent and no mortgage to overburden them.
Slovenia 3.5 % Down from 3.8% in 2024. Slovenia sold most of its public flats to sitting tenants in the 1990s; the debt was cleared a generation ago.
Poland 4.1 % Fell from 5.2% in a single year. High outright ownership again, plus incomes rising faster than housing costs.
Finland 4.7 % The best result in the EU that is not mostly explained by inherited flats. Finland houses a large share of low-income tenants in non-profit stock.
Austria 6.1 % Below the EU average every year since 2018. Roughly a fifth of Austria's entire housing stock belongs to limited-profit associations.
Czechia 8.2 % Above the EU average, improving from 9.2% in 2024. Czechia has high ownership but one of the EU's least affordable new-build markets.
Germany 11.2 % Germany is a nation of tenants — around half of households rent — so rent rises land directly in this indicator.
Greece 26.4 % More than three times the EU average of 7.7%. One Greek in four is over the 40% line, a decade after the debt crisis cut incomes.

What has actually been tried

Real countries, real policies, and what the numbers did afterwards — including where it went wrong.

Austria

Austria never stopped building non-market housing. Limited-profit associations, which are legally barred from distributing profits and charge a regulated cost rent, hold over a million dwellings — about 20% of the national stock — and build around 25% of all new homes each year. In Vienna the city itself owns about 220,000 flats and limited-profit associations run about 200,000 more.

What happened. Statistics Austria measured average rent including running costs at €8.5 per square metre in municipal flats in the first quarter of 2026, €9.0 in co-operative flats and €12.5 in private tenancies. Austria's overburden rate was 6.1% in 2025 against an EU average of 7.7%.

The catch. It is not a shield against inflation. The national average rent still rose 4.8% in the year to the first quarter of 2026, to €695.1 a month. The model also rests on land the city and the associations bought decades ago, which a country starting today would have to buy at today's prices.

Statistics Austria, Housing Q1 2026 (published 18 June 2026); City of Vienna; socialhousing.wien

New Zealand

Auckland's Unitary Plan, adopted in 2016, upzoned roughly three quarters of the city's residential land at once — allowing terraces and small apartment blocks on plots that had been restricted to one detached house. It was a change to what was legal to build, not a subsidy.

What happened. Greenaway-McGrevy and Phillips estimated 21,808 extra consents after five years; a later paper by Greenaway-McGrevy put it at 43,500 extra consents within six years, about 9% of Auckland's dwelling stock. Greenaway-McGrevy and So found rents six years after the plan were 28% lower than in comparable cities that did not upzone.

The catch. This is the most contested case on the page. Murray and Helm argue the building surge was a national credit cycle, not the zoning change, and call the Auckland result a myth. Donovan and Maltman reviewed the three main critiques in 2024 and rejected them. Both sides are economists working from the same consent data.

Donovan and Maltman, 'Dispelling myths: reviewing the evidence on zoning reforms in Auckland', Motu Working Paper 24-07, 2024

Germany

Berlin froze rents. The Mietendeckel, in force from 23 February 2020, capped rents on roughly 1.5 million dwellings built before 2014 for five years, and from November 2020 forced reductions on flats deemed overpriced. Rents in buildings completed from 2014 onwards were exempt.

What happened. The ifo Institute compared Berlin with other German cities above 500,000 people and found the supply of advertised rental flats fell by up to 60%. In the regulated stock, rents rose 11 percentage points less per quarter than in comparison cities. In the exempt stock, they rose 5 percentage points more. On 25 March 2021 the Federal Constitutional Court declared the law void.

The catch. The court struck the law down on competence, not on economics: rent price law is settled federally in sections 556 to 561 of the Civil Code, so a state may not legislate it (2 BvF 1/20, published 15 April 2021). Tenants who had paid the reduced rent owed the difference back. The ifo figures come from immowelt.de advertisements, so they measure the market a mover sees, not signed contracts.

ifo Institute, 'Berlin's rent cap drastically shrank supply of rental properties', 12 April 2022; Federal Constitutional Court press release 28/2021

United Kingdom

Help to Buy, running from 2013 to 2023, lent buyers of new-build homes up to 20% of the purchase price interest-free for five years, and up to 40% in London. The aim was to let people with small deposits buy, and to get builders building.

What happened. Carozzi, Hilber and Yu at the London School of Economics found that inside Greater London the scheme raised new-build prices by about 4.5% to 6.5% — roughly £23,682 more per home — with no appreciable effect on construction or on aggregate private mortgage lending. Near the England–Wales border, where land was easier to build on, construction did rise, but new homes shrank by about 7%.

The catch. This is the shape of a demand subsidy meeting a fixed supply: where you cannot build, the money becomes price. The study identifies effects at policy borders and covers the scheme's early years, not the full decade to 2023. It was original research by a university, published in 2019, not an official evaluation.

Carozzi, Hilber and Yu, 'The economic impacts of Help to Buy', London School of Economics, September 2019

Singapore

Singapore's Housing and Development Board has built more than a million flats across 24 towns since 1960 and sells them on 99-year leases, with grants scaled to income. The state acquired most of the land compulsorily under a 1966 law, which is the part of the model that rarely travels.

What happened. In 2023 HDB flats housed 77.8% of resident households, and 89.7% of resident households owned their home, according to the Department of Statistics. But the HDB Resale Price Index reached 203.4 in the first quarter of 2026 against a base of 100 in the first quarter of 2009 — resale prices have roughly doubled in seventeen years.

The catch. The index fell 0.1% in the first quarter of 2026, its first quarterly dip in nearly seven years, so the direction may be changing. And a 99-year lease is not freehold: the flat's value runs down to zero at the end of it, which is a live and unresolved argument in Singapore about what the older stock is worth.

Singapore Department of Statistics, households latest data (2023); Housing and Development Board, 1st Quarter 2026 Resale Price Index

What a machine would optimise for

This is arithmetic, not advice and not a prediction. We name the single number being maximised and follow it wherever it goes. The point is to see the shape of the answer a calculator gives.

The number being maximised

The single number is the housing cost overburden rate: the share of people in households paying more than 40% of disposable income on housing. Drive it down, and nothing else counts.

  1. Build where the overburdened people already are, not where land is cheap. The measure counts heads, so one new flat in a city with a 26% rate removes more overburden than one in a region with a 3% rate — the same brick, a different denominator.
  2. Legalise the cheapest square metre before subsidising the expensive one. Auckland changed what was permitted on existing plots and got an estimated 43,500 extra consents in six years; Help to Buy paid buyers inside London and moved £23,682 per home into the price instead.
  3. Work the bottom of the fraction. The indicator is housing cost divided by disposable income, so a cash housing allowance lowers it without a single home being built — and the calculator, which cannot see a building, is indifferent between the two.
  4. Convert marginal renters into outright owners and stop there. Cyprus at 2.4%, Slovenia at 3.5% and Poland at 4.1% sit at the top of the table largely because their households finished paying long ago, not because their new homes are cheap.

The arithmetic

EU27, 2025: 7.7% of about 449 million people is roughly 34.6 million over the 40% line. Greece is 26.4% of about 10.4 million, or about 2.7 million of them. Bring Greece alone to the EU average and about 1.9 million people leave the total — 5.6% of the EU-wide problem, from a country holding 2.3% of the EU's population. A calculator reading only this indicator would spend its entire budget in Athens and Thessaloniki and nothing anywhere else.

What the machine would miss

The most important part of this page. A number that goes up can still be paid for by somebody, and some things never make it into the number at all.

Who pays

In Berlin the bill landed on tenants twice. Sitting tenants paid less for fourteen months, then owed the difference back when the Constitutional Court voided the law on 15 April 2021. People looking for a flat paid throughout: ifo measured advertised supply down by as much as 60%, which is not a price anyone is billed for, but it is the reason a nurse moving to Berlin in 2020 could not find somewhere to live.

What the number flattens

The overburden rate is a threshold, so it counts people, not distance. A household paying 41% of its income and a household paying 75% are the same single tick. It also rewards inheritance: Croatia scores 3.2% and Cyprus 2.4% largely because so many homes were paid off long ago, which tells a 28-year-old in Zagreb with no family flat almost nothing about her own chances.

What a spreadsheet cannot see

Security. A Vienna Gemeindebau tenancy runs indefinitely and can pass to a household member; a Singapore HDB flat runs out after 99 years. Both are housing costs of a few hundred euros or dollars, identical in the ratio, and utterly different if you are deciding whether to have a child, take a job in another city, or repaint a wall you may not own in twenty years.

Where does your country stand?

Open my country

Where the numbers come from

We are a newspaper, not a government and not an adviser. Nothing on this page is a recommendation to you or to anyone in office. It is what was tried, what it measured, and what a calculator would say if you let it loose on the same problem.

One good piece of thinking a day

The day's most worthwhile story, and the question underneath it. No spam, one click to leave.

One click to leave. We never sell or share your address.