For years, the Dutch government has aimed to build 100,000 homes annually. However, additions to the housing stock from newly constructed homes have struggled to rise above 70,000 a year. To help bridge this gap, converting empty offices, splitting large homes and adding storeys to existing blocks (“topping-up”) have become explicit parts of Dutch housing policy. The existing stock is often presented as having substantial technical potential, seemingly sufficient to close much of the roughly 30,000-home annual gap. However, despite increasing policy efforts, national statistics show net additions from the existing stock falling to just under 11,000 homes in 2025. The Dutch case therefore illustrates a broader challenge for forecasters: headline estimates of housing potential can substantially overstate what is realistically deliverable.
Since 2019, successive Dutch governments’ housing policies have place increasing emphasis on making more efficient use of the existing stock. Following an expansion of subsidy schemes and an increase in budgets, the Dutch government has set high expectations. For conversion of offices into homes, the government set an annual target of 15,000 homes. For topping-up, the housing ministry cited a technical potential of about 100,000 additional homes by 2030, based on a 2023 survey. Splitting is said to have a potential of several hundred thousand homes, with the Chief Government Architect stating that splitting will be more important than newly built homes to meet short-term housing targets.
The realised numbers paint a drastically different picture. The net housing additions from the existing stock fell from 18,000 homes in 2021 to about 10,800 homes in 2025, according to national statistics. The decline is largely driven by conversion, which fell steeply from over 10,000 homes in 2020 to a little over 6,000 in 2025. Splitting, which has long yielded around 2,000 homes annually, netted just over 1,600 homes in 2025. Additionally, EIB estimates that topping-up delivers approximately 900 homes a year. New construction yields about 70,000 homes annually, making the contribution from the existing stock modest by comparison. These figures underline the substantial gap between headline estimates of housing potential and actual delivery.
Scarce locations, costs and current residents limit topping-up and splitting
Topping-up illustrates the gap between technical potential and actual results most clearly. Technically, 100,000 new homes could be constructed by topping-up apartment buildings, usually between 4 and 10 storeys. The technical potential is there, but the decisive constraint here is the people who already live in these buildings. Sitting residents gain nothing from an additional storey, but they have to live on a construction site for 1 to 1.5 years. Therefore, homeowners’ associations and private landlords rarely give consent to these projects. For buildings owned by housing associations, the story is similar. Topping-up seems viable only when buildings are completely vacated for renovation. Such situations are relatively rare, explaining the figure of 900 homes added through topping-up annually.
Splitting mostly runs into location- and cost-related issues. The typical Dutch suburban terraced houses are ill-suited for splitting. They need significant investment to turn one into two homes: an additional front door, an upstairs kitchen, a downstairs bathroom and so on. Moreover, these smaller dwellings are often located in suburban areas, whereas the main target group of young, single-person households tends to prefer inner-city locations. This helps explain why splitting is concentrated in larger dwellings in inner-city areas.
These constraints make it unlikely that additions to the housing stock through this channel will rise significantly above its current level of 1,500 to 2,500 homes annually.
Low-hanging fruit already picked; limited potential in vacant dwellings
Additions from conversions have been on a downward trend for years. This development was to be expected: the easiest-to-convert offices and other buildings have already been converted into homes. Additionally, due to the high level of conversions in previous years, the oversupply in the office sector has diminished, which limits the potential for conversion even further. The office spaces that remain empty are usually either in buildings that are only partially empty or situated in locations ill-suited for residential development. A renewed increase in housing additions from conversion is therefore unlikely.
Another source of overstated housing potential is the cited number of vacant dwellings. Around 190,000 homes are administratively registered as vacant, a number often cited as a potential extra source of housing supply. However, after excluding frictional vacancy and checking whether the remaining dwellings show actual energy use, only around 30,000 appear to be genuinely vacant. Even within this group, some dwellings are undergoing renovation, and others are located in shrinking regions, further limiting how much of this vacancy can realistically help relieve housing shortages.
From headline estimates to realistic housing delivery
For forecasters and policymakers in Europe, the key is to distinguish headline estimates from what can realistically be delivered. Such estimates cannot be translated directly into future housing supply or construction output. Market conditions, costs, location, legal constraints and user preferences all determine how much of the estimated potential can actually materialise and should therefore be reflected in construction forecasts.
*This briefing draws on “Belang van de bestaande voorraad voor het oplossen van de woningnood”, EIB, August 2026, by Marc Boon and Taco van Hoek.
