Construction & Engineering · Eurostat
Eurostat Construction Statistics & Building Permits
Datadory delivers eurostat construction statistics building permits data covering the European Union's harmonized monthly short-term indicators: building permits in number of dwellings and square metres of floor area (about 425,700 observations from January 1994), plus construction production indices and cost and producer prices for 33 European geographies back to January 1980. Delivered daily, weekly, or hourly as an API, files, or straight into your warehouse.
API, files, or your warehouse. Daily, weekly, or hourly.
- Where it covers
- 33 European geographies: EU-27 aggregate, euro-area aggregates, EU member states plus Norway, Switzerland, the United Kingdom, North Macedonia, Türkiye and Ukraine
- How far back
- Building permits monthly from January 1994 to July 2026; production, cost and price series from January 1980 to June 2026
- How fine
- One observation per month per geography per indicator per adjustment status, by country, CPA 2.1 building type (permits) or NACE Rev.2 division (production); national level only
What is the Eurostat Construction Statistics & Building Permits dataset?
Two tables, two halves of the same question: what were builders allowed to start, and what did they actually build? sts_cobp_m, building permits by type of building, carries the permission side - BPRM_DW counting dwellings permitted and BPRM_SQM counting square metres of useful floor area, each observation classified down the CPA 2.1 product tree into residential buildings (one-dwelling, two-and-more-dwelling, residences for communities), non-residential buildings and office buildings. Roughly 425,700 observations run from January 1994 to July 2026 across EU and euro-area aggregates plus sixteen countries including Belgium, Germany, Greece, Spain, France, Cyprus, Hungary, the Netherlands, Portugal and Romania.
sts_copr_m, production in construction, carries the output side: volume indices for NACE Rev.2 section F and its divisions F41 Construction of buildings, F42 Civil engineering and F43 Specialised construction activities, joined by cost indices and producer prices. It reaches 33 geographic entities back to January 1980 - 558 months of history per geography.
Both sides ship harmonized, which is the part no national statistical office can give you alone: every country's permit count and production index sits on identical dimension codes, so a Lisbon-Helsinki-Warsaw comparison is a filter, not a translation project. Verified readings put the EU27_2020 production index at 101.3 (2021=100) in May 2026, up 1.2% on the previous month.
What do sample rows look like?
One observation per row: a geography, a month, an indicator and the reading. Straight from the archive:
dataset : sts_copr_m (Production in construction - monthly)
geo : EU27_2020 (EU-27 aggregate)
indic_bt : PRD unit : I21 (index, 2021 = 100)
s_adj : NSA time : 2026-05 value: 101.3
dataset : sts_copr_m geo : EU27_2020 indic_bt : PRD
unit : PCH_PRE (change on previous period)
time : 2026-05 value: +1.2%
dataset : sts_cobp_m (Building permits by type of building - monthly)
indic_bt : BPRM_DW (number of dwellings permitted)
cpa2_1 : CPA_F41001 (residential buildings, one-dwelling)
time span : 1994-01 through 2026-07The pairing is the point. The same May 2026 month reads as an index level (101.3) and as movement (+1.2%), because the units dimension ships I21 and I15 index bases next to PCH_PRE period-on-period change and PCH_SM change versus the same month a year earlier - deltas arrive computed rather than derived on your side. Request a sample and rows come back cut to whichever countries, building types and years you name.
Which fields does the dataset include?
Nine fields define every observation, each a documented dimension code rather than free text - which is why joins never drift. Definitions below were verified against live query responses during Datadory's August 2026 research pass, not inferred from documentation.
What does coverage look like across geography, time and granularity?
Geography - 33 European entities: the EU-27 aggregate, euro-area aggregates, EU member states plus Norway, Switzerland, the United Kingdom, North Macedonia, Türkiye and Ukraine. The permit table concentrates detail on sixteen reporting countries; the production table runs the wider 33-entity panel. Aggregates and member states ship side by side, so a Europe-wide chart and a per-country breakdown come from one query shape.
Temporal - monthly throughout. Permits run January 1994 through July 2026; production, cost and producer-price series reach back to January 1980 and forward to June 2026. That is four decades on the output side - long enough to cover several complete European construction cycles, rate regimes included.
Granularity - one observation per month per geography per indicator per adjustment status, crossed with CPA 2.1 building type on permits and NACE Rev.2 division on production. Coverage stops at national level: no regional, city or company detail anywhere in the panel. For sub-national European depth, pair this feed with a national permit registry; the full shortlist sits on our best construction engineering datasets page.
The industry context lives on our construction engineering data hub.
How is the data delivered?
API, files, or your warehouse. Daily, weekly, or hourly.
Name the countries, indicators and date range when you request the sample; the sample ships first either way, with the full field dictionary attached so your engineers can validate the shape before anything recurring starts.
Cadence is yours to set - and to change when your models change.
Who uses this data, and for what?
- Cross-country market ranking - compare permit momentum and production indices across EU27, euro-area and neighbor countries under one definition instead of stitching national releases. See market researchers use cases.
- Cycle modeling - train on a monthly panel reaching to 1994 (permits) and 1980 (output): the SCA variant hands models a clean trend signal while NSA preserves the raw seasonal pattern worth learning.
- Sector rotation signals - read harmonized production indices and cost/producer-price pressure for European builders and materials names through mid-2026. See investors quants use cases.
- Leading-indicator dashboards - permits lead completions; the BPRM_DW and BPRM_SQM pair lets you separate how many homes were authorized from how much floor area, catching a shift toward larger or smaller units before output confirms it.
- Citation-grade research - quote one institution's harmonized numbers for EU construction health rather than reconciling two dozen national press releases. See journalists academics use cases.
Which personas get the most value?
Market researchers and consultants, investors and quant researchers, data scientists and ML engineers, developers and data-product builders and journalists, academics and students all rate this feed at top relevance - one group ranks markets, one trades the cycle, one trains on it, one pipes it, one cites it. Competitive intelligence teams use permit inflections by building type as demand evidence ahead of region-allocation changes, and sales teams rank European territories by permit momentum before committing distribution. Persona-by-persona workflows sit on our construction engineering data hub.
How does it compare within construction engineering data?
It is Europe's harmonized answer in a catalog of mostly national series. New Residential Construction (Building Permits, Housing Starts & Completions) is the American counterpart - permits, starts and completions in units for the US; the head-to-head continues on Eurostat vs New Residential Construction. Value of Construction Put in Place (VIP) prices American work in dollars where this feed indexes European volume, and Building Construction Investment by Type of Structure does the same for Canadian housing by province. FRED Economic Data (Housing Starts & Construction Series) redistributes US series, and OECD Data Explorer (Building Permits & Construction Indicators) widens the country list beyond Europe at shallower depth. Where this dataset stands alone is combining permit counts with floor areas and pipeline/mp1/cost/price indices across 33 European geographies in one dimension-coded schema.
What should I know before requesting a sample?
Four things worth knowing upfront.
First, values are indexed or percentage changes, not euros or unit counts on the production side. A reading of 101.3 on the 2021=100 base means output sits 1.3% above its 2021 average; absolute euro revenue needs a separate value anchor.
Second, the two tables have different horizons: permits run to the July 2026 reference month while production and price series extend to June 2026. Align end dates before computing cross-table ratios.
Third, quarterly and annual sibling datasets exist in the same family but their exact table codes went unconfirmed during research - if your pipeline needs lower-frequency series, say so in the sample request and we pin the codes before delivery.
Fourth, publication trails the reference month, so treat the series as confirmed-history input for modeling and benchmarking rather than a nowcast. All nine field definitions carry verified confidence, so the schema itself is not among the open questions.
Get a sample of this dataset and we route rows scoped to your geographies, indicators and date range, with the field dictionary attached.
Field dictionary
Every field below is documented against real records. The full dictionary ships with the sample.
| field | type | definition | example |
|---|---|---|---|
freq | enum | Time frequency of the observation; M = monthly. | M |
indic_bt | enum | Business trend indicator: BPRM_DW (building permits - number of dwellings), BPRM_SQM (building permits - square metres of useful floor area); PRD (production volume) in the companion table. | BPRM_DW |
cpa2_1 | enum | CPA 2.1 product classification of the permitted building type - buildings overall, residential buildings (one-dwelling, two-and-more-dwelling, residences for communities), non-residential and office buildings. | CPA_F41001 |
nace_r2 | enum | NACE Rev.2 activity classification on the production side: F Construction, F41 Construction of buildings, F42 Civil engineering, F43 Specialised construction activities. | F |
s_adj | enum | Seasonal adjustment flag: NSA (unadjusted) or SCA (seasonally and calendar adjusted). | NSA |
unit | enum | Unit of measure: I21/I15 index bases (2021=100, 2015=100), PCH_PRE (percentage change on previous period), PCH_SM (change versus same period previous year). | I21 |
geo | string | Reporting entity: EU27_2020, euro-area aggregates, individual member states plus Norway, Switzerland, the UK, North Macedonia, Türkiye and Ukraine. | DE |
time | date | Reference period in YYYY-MM form; permits span 1994-01 through 2026-07, production and price series 1980-01 through 2026-06. | 2026-06 |
value | number | Observed value for the dimension combination - an index level, a percentage change, a permit count or a floor area. | 103.6 |
Questions buyers ask
How far back does the Eurostat construction data go?
Production, cost and producer-price series reach back to January 1980 - 558 monthly periods per geography - while the building-permit table runs from January 1994 to July 2026 with roughly 425,700 observations. Four decades of output history covers several complete European construction cycles, recessions included, long enough to train seasonal models on rather than against.
Which countries does the dataset cover?
Thirty-three European entities: the EU-27 aggregate, euro-area aggregates, EU member states plus Norway, Switzerland, the United Kingdom, North Macedonia, Türkiye and Ukraine. Permit detail concentrates on sixteen reporting countries including Germany, France, Spain, the Netherlands, Portugal, Romania, Hungary, Cyprus, Belgium and Greece; the production panel runs the full thirty-three.
What is the difference between the building permits and production indicators?
Permits measure permission: BPRM_DW counts dwellings authorized and BPRM_SQM counts square metres of useful floor area, classified by CPA 2.1 building type. Production measures execution: PRD volume indices for NACE F41, F42 and F43, plus cost indices and producer prices. Reading both separates what regulators approved from what builders delivered - permits lead, output confirms.
Why are there unadjusted and seasonally adjusted variants of every series?
Every series ships as NSA (unadjusted) and SCA (seasonally and calendar adjusted) side by side. Use SCA to read the underlying trend month over month without winter dips masquerading as slowdowns; use NSA when modeling the raw seasonal pattern itself or reconciling against national releases published unadjusted. Swap variants without changing anything else about the query.
Can I track month-over-month and year-over-year change directly?
Yes - the units dimension publishes PCH_PRE (percentage change on the previous period) and PCH_SM (percentage change versus the same period a year earlier) beside the I21 (2021=100) and I15 (2015=100) index bases. Deltas arrive computed and harmonized, so no derived-column maintenance lands on your side.
Is there a quarterly or annual version of the building permits series?
Sibling datasets at quarterly and annual frequency exist in the same Eurostat short-term family, though their exact table codes were not individually confirmed during Datadory's August 2026 research pass. Name the frequency you need when requesting a sample and we verify the specific series before anything reaches your pipeline.
Does the data go below country level?
No - coverage is national-level only, plus EU and euro-area aggregates. There are no regional, city or company breakdowns inside these tables. For sub-national work, pair the harmonized national trend with a municipal permit registry such as Chicago Building Permits or NYC DOB Job Application Filings.
Who uses Eurostat construction statistics data?
Seven Datadory persona packs carry it, five at top relevance: market researchers ranking European markets, investors trading the construction cycle, data scientists modeling on four decades of monthly panels, developers wiring dimension-coded feeds into products, and journalists and academics citing one harmonized source instead of twenty-plus national offices. Competitive intelligence and sales teams use permit momentum as demand evidence.
Notes on this record
- sts_cobp_m alone holds roughly 425,700 observations from January 1994 to July 2026 - one schema covering dwellings and floor areas across EU aggregates and sixteen reporting countries.
- The production, cost and producer-price panel reaches 558 monthly periods per geography - four decades of European construction cycles in one dimension-coded table.
- BPRM_DW and BPRM_SQM authorize future work; NACE F41-F43 production indices record delivered work. Holding both turns a leading indicator into a confirmation loop.
- PCH_PRE and PCH_SM percentage changes sit beside I21 and I15 index bases in the same units dimension, so month-over-month and year-on-year reads need no derived columns.
- NSA and SCA variants exist for every series - trend reads and seasonal-pattern modeling pull from the same query shape rather than two pipelines.
- All nine core definitions were checked against live query responses during Datadory's August 2026 research pass; the dictionary above describes arriving data, not advertised data.
- Coverage stops at countries and aggregates - no regional or company splits - which is exactly what makes 33 geographies comparable on identical definitions.
See the rows before you pay anything.
Name this dataset and we send real records from it — scoped to the fields you asked for.