On this page
Sources and hierarchy
I use official sources first: INE, Eurostat, Banco de Portugal, AIMA, Diário da República and IMF reports. News reports serve as secondary sources, when they give access to original data or studies, and are identified as such.
Each source has a publication date, a consultation date and a verification status in the source library.
Levels of evidence
- Official data published by an official body, as published.
- Own calculation derived from official data with a formula described on this page.
- Third-party estimate from associations, studies or the press.
- Author’s analysis: interpretation or assessment, not a measured fact.
- Awaiting verification: never shown as a headline figure.
Calculations
Averages by period and by government. Simple arithmetic mean of the annual values.
average 2000–2010 = 875,206 / 11 = 79,564 average 2015–2025 = 179,118 / 11 = 16,283
Counterfactual. Difference between the 2000–2010 total and the 2015–2025 total. It is a comparison with an earlier pace, not a measure of a deficit.
875,206 − 179,118 = 696,088
Cumulative change in costs. Ratio between annual averages of the INE index, always stating the base year.
(2025 average / 2020 average) − 1 = 125.25 / 93.88 − 1 = +33.4%
Prices against earnings. Eurostat price index (2015 = 100) and an average earnings index derived from INE. The 2015 value is deduced from the change published up to 2023 and is still to be confirmed in the INE table.
earnings 2015 ≈ 1,505 / 1.277 ≈ €1,179 · 2025 = €1,694 (+43.7%) prices 2025 / 2015 = 263.85 / 100 (+163.9%) · 2.639 / 1.437 − 1 ≈ +84%
Illustrative households. Population growth divided by average household size. It is not a measurement of homes needed.
824,914 / 2.5 ≈ 330,000
Scenarios 2026–2035. Annual assumptions set out on the conclusions page. 2026 is estimated from 2024 licences multiplied by the recent ratio between completions and licences two years earlier (≈ 0.8).
Denominators
No percentage is published without saying what total it is a share of. The most important example:
- 5.0%: buyers with tax residence abroad, across all 169,812 transactions in 2025.
- 27.6%: foreign nationals, across the 148,632 purchases made by households.
- “Resident” in these statistics means tax residence, not a residence permit.
Series and breaks
- Population: since 2021 INE has used a series based on administrative records. It is not compared with the 2021 Census or with estimates up to 2020.
- Rents: the new series starts in 2020 and covers about 50% more contracts. INE states that levels do not change significantly.
- Licences: I use the values revised by INE, which differ from those quoted in the press.
- Homes completed in 2025: INE has two figures, 27,301 (annual count, used) and 26,714 (quarterly estimates).
- Prices: +180% compares Q4 2025 with the 2015 average; +164% compares annual averages.
Governments and decisions
- The government bands in the charts show the government in office for most of each year. They do not show who caused the value.
- Each decision separates the publication date, the state of implementation and the measured effect.
- When only the year is certain, I show only the year. Laws whose number or date is still to be confirmed are flagged.
- Governments appear in neutral grey, never in party colours.
Formats
- Portuguese: thousands with a thin space (125 708) and decimals with a comma (17,6%).
- English: 125,708 and 17.6%.
- Dates: 25 Oct 2026 in the text and 25-10-2026 in tables.
What I do not do
- I do not interpolate missing annual values. When only some years are known, I show only those points.
- I do not use two vertical axes in the same chart.
- I do not present scenarios as forecasts, or counterfactuals as measured deficits.
- I do not attribute precise effects to a policy decision without a study to support it.