Who owns Luxembourg's building land, and why it is not being built on

Luxembourg has 4,294 hectares of vacant building land, enough to absorb several decades of growth without touching a single zoning boundary. It is not being built on. To understand why, you have to look at who holds it, what holding it has earned them, and what the tax on it costs them: almost nothing.

Three thousand people for half the private potential

In July 2023 the Observatoire de l'habitat (the national housing observatory) published its Note 32, an update of the land reserves available for housing. The 4,294 hectares counted in 2022 are plots already zoned for residential or mixed use in the municipalities' general development plans (PAG). They are not waiting for any boundary change. I have set out elsewhere what this land is and where it lies; here, the question is who owns it.

The breakdown by category of holder is known. Private individuals own 63.9% of the potential, private companies around 20%, public and semi-public bodies 14.5% (Note 32, July 2023). The State and the municipalities, which carry the public debate about the housing crisis, therefore hold only one seventh of the land on which homes could be built.

0.5%

It is within private individuals that the concentration becomes striking. According to the same Note 32, 3,447 individuals, or 0.5% of the population, hold half of the privately owned potential, 1,865 hectares. The top hundred owners hold 360 hectares, 13.2% of the land held by individuals. The top thousand hold between 1,135 and 1,288 hectares depending on the counting method, more than 40%. Fewer than 3% of the population own any building plot at all.

On the corporate side, ten groups account for 47.9% of the land held by legal entities. Improof.lu, in an April 2025 analysis of the financialisation of land, puts the reserves of each of the five largest local developers at around 500 million euros, and the building-land holdings of each of the top hundred private individuals at some 30 million euros. These figures are valuations, not cadastral data; they depend on the reference price chosen.

The OECD picked up this work in its Economic Survey of Luxembourg of April 2025. Its summary states that 0.15% of the population holds more than 45% of the building land. The figure describes the same reality as the Observatoire's “top 1,000”, with a slightly different scope: a thousand people set against some 670,000 inhabitants makes 0.15%. The OECD draws direct recommendations from it, which I return to below.

Holder	Share of potential
Breakdown by category (2022 data)
Private individuals	63.9%
Private companies	≈ 20%
Public and semi-public bodies	14.5%
Concentration among private individuals
Top 100 owners	360 ha (13.2%)
Top 1,000 owners	1,135 to 1,288 ha (> 40%)
3,447 people (0.5% of the population)	1,865 ha (50%)
Concentration among companies
Top 10 groups	47.9% of corporate-held land

A rural inheritance that nothing pushes to sell

The Observatoire's data do not say who these 3,447 people are, and this article will not name them. They do say what shape the holdings take. Luxembourg's building land is first and foremost inherited land. The residential zones of the PAGs were drawn, revision after revision, around villages whose land belonged to farms and to families settled there for generations. When the building boundary extends over a meadow, the meadow changes value, not owner. The Observatoire notes, moreover, that some plots zoned for building are still farmed or planted with vines (Note 32). They pass from one generation to the next by succession, and nothing obliges the heir to do anything with them.

In my early working notes I had described Luxembourg as a “basically feudal country”: a few families, most of the land, a long-standing presence in public life. That phrase is a perception, not a data point, and I set it aside. What the figures confirm is more sober and is enough: a small number of old landed estates, inherited rather than bought, with no debt to service and no deadline. An owner who bought on credit has to earn a return; an owner who inherited can wait. And in Luxembourg, waiting has long been the best investment of all.

Holding has paid better than building

+136.5%

Between 2010 and 2021, the price of building plots rose by 136.5%, according to the STATEC series cited in Note 32. Over the same period, construction costs rose by 32.4% and the price of homes under construction by 107.4%. A plot therefore gained more value than the home that could have been built on it, and four times more than it would have cost to build it. For a holder under no financial pressure, the choice was simple: do nothing.

This rent explains the weight of land in the final price. According to Note 32, land accounts on average for 29.5% of the price of a new home in Luxembourg, and 43.8% in Luxembourg City. In the capital, nearly half of what a buyer pays remunerates no construction work at all.

The 2022-2024 correction has dented this rent without erasing it. The Observatoire's analysis report no. 19 (October 2025) measures a 15% fall in plot prices between 2023 and 2024, including around the capital. Homes followed the same slope: the STATEC price index fell by 16.3% between the third quarter of 2022 and the first quarter of 2024. Since then, stabilisation: +0.1% over 2025, +1.7% year on year in the first quarter of 2026 (Logement en chiffres no. 19, March 2026; analysis report no. 25, June 2026). An owner who held their plot in 2010 keeps a capital gain well above 100%. The fall cooled new buyers, not long-standing holders.

The behaviour of the owners of Baulücken, those serviced infill plots in the heart of built-up neighbourhoods, confirms it. They represent about a third of the national potential and are 75% privately owned. Surveyed as part of Raum+, the municipalities estimate that only 60% of their owners wish to build or sell (Observatoire de l'habitat, analysis report on the owners of Baulücken). Four in ten owners of a ready-to-build plot, connected to the networks, sitting between two houses, have no intention of using it. No physical, legal or administrative constraint explains this.

A property tax twenty times below the OECD average

€40m

If holding costs nothing, it is because Luxembourg's property tax rests on unit values set in 1941 and never revalued since, as the government's own dossier “Eng modern Grondsteier” (“A modern property tax”) acknowledges. The firm ATOZ quantified the gap in November 2022: property tax brings in around 40 million euros a year to all the municipalities combined. Aligned with the average weight of property taxes in OECD countries, it would bring in 806 million. The ratio is one to twenty. Holding a vacant plot whose value has more than doubled in ten years costs, each year, only a token sum.

The OECD, in its Economic Survey of April 2025, does not merely recommend raising this tax. It calls for municipalities to be obliged to penalise unused building permits, for permitted residential density to be increased, for the tax deductibility of mortgage interest, which props up demand without creating supply, to be phased out, and for the land purchases of public social landlords to be funded directly. The summary published by the government on 24 April 2025 reproduces these recommendations without comment.

Bill 8082: four years of back-and-forth, entry into force around 2028

The government does have a plan. Tabled in 2022, bill 8082 creates three instruments: a reformed property tax (IFON) based on updated values, a land mobilisation tax (IMOB) levied on building plots left vacant, and a tax on unoccupied homes (INOL). Its journey sums up how difficult the subject is.

• 2022: bill 8082 tabled in the Chamber of Deputies.
• Opinion of the Council of State: 22 formal objections, an unusual number for a single text.
• July 2025: split in committee. The property tax and IMOB part becomes bill 8082A; the vacant homes part (INOL) is detached.
• 12 November 2025: government amendments, including the exemption of land used for existing farming or wine-growing activities.
• 3 February 2026: new opinion of the Council of State.
• 2 July 2026: the text is still in committee, with no vote in plenary. Entry into force is envisaged around 2028.

The IMOB scale is progressive over time, and that is where its ambition can be measured. The example circulated when the text was presented concerns a six-are plot in Mersch.

Time the plot has been held vacant	Annual IMOB
Years 1 to 5	€0
After 5 years	€260
After 11 years	€3,900
After 16 years	€12,900
From 20 years	€23,200

Nothing is owed for five years. The first instalment that actually bites comes only after eleven years. If the text enters into force in 2028, as envisaged, and if the clock starts from that date, no owner will pay a single euro before 2033, and the scale will not reach full effect until the mid-2040s. My working notes mentioned a first tax bill in 2030 and a full rate in 2035; no official document confirms that timetable. In any case, a plot inherited today can stay vacant for another decade without consequence.

The exemptions are the other weak point. The government amendment of 12 November 2025 exempts from IFON and IMOB land used, in the first tax year, for existing farming or wine-growing activities, outside localities of particular interest for spatial planning. Improof.lu (April 2025, then March 2026) sees a major loophole: a significant share of vacant building land is precisely land in agricultural use, and that share would escape the tax for as long as the farming continues. A meadow mown once a year is still a meadow. The think tank also judges the text's progressivity insufficient in view of the rent accumulated. These criticisms are a reading of the text, not an official assessment; I have found no public estimate of the share of the potential that would be exempt.

What is blocking, and for whom

The IDEA foundation summed up the options in March 2026, in an analysis relayed by LesFrontaliers.lu. Compelling owners to build is judged too risky politically. Expropriation would cost too much at current land prices. Property tax is unpopular. Densification is contested by neighbours, in a reflex Anglo-Saxon planners call BANANA, “Build Absolutely Nothing Anywhere Near Anything”. Every lever exists; each one collides with a group that votes.

Developers have no interest in a supply shock either. Five of them hold reserves of around 500 million euros each (Improof.lu, April 2025). Building fast and in volume would push down the selling price of homes and, mechanically, the value of the remaining reserves. The optimal pace for a landholder is the one that keeps prices up, and that is not the one that meets the theoretical need of 6,000 homes a year estimated by the Observatoire de l'habitat. Wherever building has happened on a large scale, rents have fallen or stopped rising; that is precisely what a landholder is trying to avoid.

Then there is the electorate. My notes said “80% homeowners”; the figure is wrong. Eurostat (EU-SILC, indicator ilc_lvho02) measures 63.5% of households owning their home in 2024, against 72.4% in 2022, a drop that probably owes more to a break in the series than to a real shift. Even at 63.5%, a clear majority of resident households own a home whose value depends on scarcity. A government that seriously brought prices down would weaken the wealth of nearly two voters in three. Cross-border workers, who fill around 47% of salaried jobs (STATEC, first quarter 2026) and are the first to be shut out of the market, do not vote in general elections. The electoral arithmetic does not tilt towards supply.

None of this is new. In 2012, the then Prime Minister put it bluntly.

Is it really necessary for those who own land to squeeze the last drop of blood out of those who need it?

Fourteen years later, the property tax has not changed, the bill meant to reform it is in committee, and the vacant plots of 2012 are, for the most part, still vacant.

A shortage that suits those who could end it

Luxembourg does not have a land problem. It has an owner problem. 4,294 buildable hectares, 85% of them mobilisable without physical or legal obstacle according to the Observatoire, enough to build up to 161,500 homes according to LISER, are held for the most part by a few thousand people and a few dozen companies that have no economic reason to part with them. The tax system encourages them to hold on. The reform that ought to force their hand will not enter into force before 2028, will not bite before the mid-2030s, and its exemptions narrow its reach. The housing crisis is not a geographical fate; it is the outcome of an equilibrium in which every actor able to resolve it has an interest in seeing it last.

For anyone looking for a home today, this means that supply will remain constrained in the medium term, and that price movements will come from interest-rate cycles, not from a surge of construction. What I see in Immovalu's data, 33,749 listings for sale and a median asking price of 845,000 €, is a market renegotiating itself one listing at a time. The opportunities come from sellers in a hurry, not from a relaxed market, and the Towns page shows, municipality by municipality, where the gap between asking prices and expected values is widest.