Resource context (ESPON House4All)
This resource from ESPON, authored by Franziska Sielker (TU Wien) and Selim Banabak (TU Wien), explains how the ESPON House4All project is using online housing advertisements to build a pan-European housing affordability map at regional and sub-regional scales. It frames affordability as an urgent, Europe-wide issue, with rising housing costs pushing more households toward financial overburden, and argues that better spatial detail is needed to understand where pressures are most acute.
Why more granular affordability evidence is needed
The text describes housing affordability as housing costs relative to household income, where costs may include rent or mortgage payments, maintenance, and sometimes energy and transport. A common “income ratio” benchmark treats households spending roughly 30–40% of income on housing as overburdened. However, applying this approach at regional or city level is difficult because traditional sources (surveys and government records) often lack consistent, fine-grained sub-national data. This limits the ability to assess affordability precisely across diverse regions and urban areas.
Method: repeated web-scraping of housing adverts across Europe
To address data gaps, the House4All project uses repeated web-scraping of adverts between April 2024 and April 2025 across 31 countries in the ESPON space. The approach targets listing platforms such as Nestoria and Properstar and extracts variables including advertised price, geolocation, and dwelling characteristics (e.g., number of rooms and condition). The project then aggregates listings to compute indicators such as average prices and price per square metre for chosen spatial units, enabling comparisons to income measures from conventional statistical sources.
What housing adverts add compared with traditional datasets
The resource highlights that advert data can deliver much higher spatial granularity than survey-based evidence, with flexibility to aggregate from neighbourhood level up to regions. It also emphasises that listings reflect current market conditions and therefore capture the prices faced by newcomers to a city or region, which can be important in fast-changing urban markets. With repeated scraping over a full year, the project expects to differentiate market segments and analyse contrasts between renting and buying/selling offers across locations. Regular collection is also presented as a way to infer which segments and areas are most in demand, for example by identifying adverts with shorter online durations versus those that remain listed longer.
Coverage, comparability, and price-bias limitations
The text also details key pitfalls. Listings cover only the publicly advertised part of the market and may exclude social housing as well as the highest- and lowest-priced segments, meaning the dataset may miss conditions affecting the most vulnerable households (including people relying on benefits). Cross-country comparison requires extensive harmonisation because “listed price” can include different components across countries (e.g., whether land, utilities, or taxes are included). In addition, advertised prices may diverge from final transaction prices: sellers or landlords may test the market or price strategically, and negotiations can produce gaps that vary between renting and selling and between price segments. The resource notes that reflecting these negotiation differences comprehensively is difficult within the project, even though transaction-price data is increasingly becoming available for longer-term improvements.
Income data constraints and implications for indicators
Computing price-to-income ratios depends on income data that is often only available as aggregates rather than for individual households, so analyses rely on averages that can mask within-region variation. Going below NUTS2 can be especially challenging due to limited income statistics. The project is preparing a methodology to adjust income inputs; the text notes that GDP per capita is sometimes used as a proxy but can mislead (e.g., regions with corporate headquarters may show inflated GDP that does not represent household incomes). At the same time, the resource argues that these issues are less problematic for within-region comparisons and local-regional mapping than for cross-country comparisons.
Intended outputs and policy relevance
Overall, the resource positions repeated advert scraping as a promising way to strengthen affordability monitoring across Europe, by producing a comprehensive mapping of housing offers and enabling tracking of price developments over time. It suggests that repeated collection can help reveal the effects of new policies on prices and provide more robust evidence as the dataset accumulates. The authors emphasise that methodological refinement and collaboration between governments, private platforms, and academia are important to improve access, reliability, and representativeness of advert-based affordability indicators.