INSEE (Institut National de la Statistique et des Etudes Economiques), the French NSI.

INSEE\ DSE Directorate \ DSCT Department \ IPP Division

(in French: Direction des statistiques d'entreprises\Département des statistiques de court terme\Division des indices de prix à la production)

INSEE

Division IPP - timbre E310

88 avenue Verdier

CS 70058

92541 Montrouge cedex

France

The statistical survey on the "New Housing Cost Price” (in French: “Prix de Revient des Logements neufs”, PRLN) summarized in “CCI-NHCP” is conducted by the Statistical Service of the Ministry in charge of Housing.

This survey data are used to assess both the "cost of construction index" and the Producer Price Index for construction of new residential buildings.

The "cost of construction index" (in French: "indice du coût de la construction") is, in spite of its name, about prices. This name was kept for historical reasons. It is used, in combination with other indices, for assessing escalation of commercial buildings' rentals.

Both the Producer Price Index for construction of new residential buildings (for statistical use) and the "CCI" are calculated by Insee with the help of hedonic models.

From the first quarter 2022, the computation of the Cost-of-Construction Index (CCI) makes use of an additional source, the Survey on building lot prices (EPTB) which interrogates individuals or households who receive a permission to build a pure individual dwelling (single-family home).

The new data of the survey replace data from the CCI-NHCP in order to calculate the individual pure sub-indice, while retaining the hedonic method.

This change of source for single-family homes induces no break in series nor revision of the previously disseminated values.

CPA Rev. 2.1, section F.

French Producer Price Index for construction of new residential buildings distinguishes 3 types of dwellings:

- pure individual dwellings (= CC 111);

- grouped individual dwellings (part of CC 112);

- collective dwellings (remaining part of CC 112).

Residences for communities (CC 113) are not observed.

Activities covered: CPA Rev. 2.1, Section F, sub-class 41.00.1: ”residential buildings”, except "41.00.14" Residences for communities.

Size classes covered: All.

The “cost of construction index” (CCI, ICC in French) is the denomination for the index (of price, indeed) to be used for commercial buildings' rentals escalation. It is expressed in reference 100 = 4^th quarter of 1953.

The “producer price index for construction of new residential buildings - base 2015” represents the “statistical” price index in reference 100 = average year 2015.

The “producer price index for construction of new residential buildings - base 2015” and the CCI measure only the development of residential buildings: excluding residences for communities, non-residential buildings, land prices and architect's and other fees. They reflect prices paid by the client to the construction company. Hence, they respect the council regulation n° 1503/2006 on "Output price indices for construction".

CCI includes VAT whereas producer price index for construction of new residential buildings excludes VAT.

Reporting unit: Works owner or prime contractor of a sampled building permit.

Observation unit: Building permit.

Around 100.000 building permits per year.

Metropolitan France, in other words excluding the overseas territories and departments.

Corsica is in the survey scope since 2015.

Overseas departments are not included because construction standards are very different, so that they would require a specific hedonic model. This model could be difficult to estimate, because there should be very few data.

CCI was first created in 1953Q4.

Base (reference) year: The “producer price index for construction of new residential buildings - base 2015” is expressed in reference 100 = 2015. Producer price indeces for construction of new residential buildings shall be dissiminated in base 2021 as of June 2024.

Price index, homogeneous to € / m²

The reference period is the quarter.

Legal basis

This "CCI" is an input for assessing rental escalation for commercial buildings (article 112-2 of the Monetary and Financial code and decree 09-1568 of December 15th, 2009).

Obligation on units to provide data: compulsory.

Data are shared between statistical service of Ministry of Ecology, in charge of the survey, and Insee, in charge of the index relying on a hedonic model.

We are not aware of data sharing with OECD, UN, IMF or other international organization.

Dissemination of terms and conditions under which official statistics are produced, including confidentiality of individual responses

Only total CCI or producer price index for construction of new residential buildings is published.

Not applicable.

CCI and producer price index for construction of new residential buildings are published around T+80.

A monthly calendar is published on Insee website for the four upcoming months:

https://www.insee.fr/en/information/2107811?debut=0

Simultaneous release to all interested parties : Yes

Identification of internal government access to data before release : No

Transmission to Eurostat and further use of the statistics

The transmission to Eurostat is made using the SDMX format, the day before the national release.

Quarterly.

Identification of ministerial commentary on the occasion of statistical releases: none.

The "ICC" index is published in the “Informations rapides” of Insee (https://www.insee.fr/en/statistiques?debut=0&theme=30&conjoncture=50) and in the Journal Officiel about the 20^th of the third month following the quarter under review.

The Producer Price Index for construction of new residential buildings is released on Insee website (https://www.insee.fr/en/recherche/recherche-statistiques?q=Producer%20price%20index%20for%20construction%20of%20new%20residential%20buildings&debut=0&categorie=10), but not through an "Informations Rapides".

The Producer Price Index for construction of new residential buildings is released on Insee website (https://www.insee.fr/en/recherche?q=Producer%20price%20index%20for%20construction%20of%20new%20residential%20buildings&debut=0).

ICC is published simultaneously with the press release on Insee macroeconomic database website: https://www.insee.fr/en/statistiques?debut=0&theme=30&conjoncture=50

Possible download of CSV files.

Not available.

Planned changes in national dissemination methods

The calendar of dissemination was advanced by 20 days in 2014.

The "Producer Price Index for construction of new residential buildings" is transmitted to Eurostat the day before the national release

Dissemination of documentation on methodology and sources used in preparing statistics

Methodology: https://www.insee.fr/en/metadonnees/source/indicateur/p1626/description#consulter-sommaire

Description of standard tables produced

Aggregated index: Price index for construction of new residential buildings.

Not available.

The European statistics code of practice is implemented as far as possible.

The questionnaires for both surveys (ICC-PRLN and EPTB) are updated annually, in order to take account of technical and regulatory developments, and also in order to adapt to the difficulties reported by respondents (new wording of the questions).

The analysis of hedonic models results is used to update models annually, in order to incorporate new variables (with testing them on the past).

Models use moving windows of seven quarters, trade-off between robustness and revision of the past.

Anyway, some difficulties are encountered with:

- regional representativity;

- modification of legal definitions of living area;

- the quality effect linked to upgrading environmental norms.

If there are both a developer of building projects and a builder, the price corresponds to the builder invoice.

Conceptually, one could wonder if the "margin" of the developer of building projects should not be taken into account, consistently with CPA 2008 and SNA 2008.

National Accounts need a deflator on one side (producer price index for construction of new residential buildings) and tenants and landlords an official index for rental escalation on the other side (cost of construction index).

No claims have been received for several years, neither from National Accounts, neither from tenants or landlords.

The coverage of new residential buildings is quite complete, under the restriction that overseas departments are not covered by the "CCI-NHCP" statistical survey, conducted by the Statistical Service of the Ministry of Ecology (in charge of Housing). Overseas departments represent only 4% of the new dwellings and 3.6% of the area of new dwellings in France in 2013.

The individuals or households from overseas departments who received a permission to build a pure individual dwelling are covered by the "EPTB" statistical survey but they are not exploited for the computation of the CCI.

However, only new residential buildings are covered: non-residential buildings are not covered, neither civil works. National Accounts use "Producer Price Index for construction of new residential buildings" or cost of construction indices for these activities/products. But a specific PPI exists for "dwellings maintenance and improvement works" (CPA = 43 \ 43.1).

Not available. Difficult with a hedonic model.

No bias has been detected.

Data of the survey cannot be revised (law 53-321). The hedonic models use data available at the deadline.

Not available. Difficult to assess with a hedonic model.

Response rate of "CCI-NHCP" survey

The rate of 40% is measured between beginning of phase 1 (sample) and end of phase 2 (complete prices and characteristics available for hedonic models). But a lot of conditions have to be fulfilled. During phase 1, the survey separates renovation (out of the scope) from construction of buildings. This explains a part of the low "rate of response", which includes in fact a large proportion of "out of scope".

Actions to speed up or increase the rate of response:

The CCI-NHCP questionnaire collection is done by the statistical office of the Ministry of Ecology. They call back non-respondents.

Response rate of "EPTB" survey

The response rate is measured at 70%. It represents about 17,000 responses each quarter but only 10% concerns the quarter under review. However other responses participate to determine the hedonic model for pure individual dwellings

The EPTB questionnaire collection is done by the statistical office of the Ministry of Ecology. It doesn't call back non-respondents.

Three hedonic models are assessed: for pure individual dwellings (= CC 111), for grouped individual dwellings (part of CC 112), for collective dwellings (remaining part of CC 112). Dozens of variables are tested; nvertheless there may be a lack of variables determining price, which can induce bad estimation for the time component of price.

Publication

80 days after the end of the reference period (quarter).

Timetable of data collection

50 days after the end of the reference period (quarter).

No delay for several years.

The definitions of Commission Regulation 1503/2006 are applied, therefore the comparability between Member States should be established (with same hypotheses on other Member States).

French Producer Price Index for construction of new residential buildings distinguishes 3 types of dwellings:

- pure individual dwellings (= CC 111);

- grouped individual dwellings (part of CC 112);

- collective dwellings (remaining part of CC 112).

Residences for communities (CC 113) are not observed.

In 2009, a hedonic model has been introduced, but should not have affected the comparability over time.

Dissemination of information that support statistical cross-checks and provide assurance of reasonableness: Not available.

Confrontation with other data sets: A quarterly confrontation is made with “BT01 index”, which measures the change in construction cost for activity “41.2 - construction of buildings”. Prices are more reactive than costs; this is usual. Producers reduce first their margin, and costs are more rigid.

Three hedonic models are used: pure individual dwellings (= CC 111), grouped individual dwellings (part of CC 112) and collective dwellings (remaining part of CC 112).
We compare changes in price of the three sub-indices: they are very consistent over the long term, but in the short term prices of individual (pure or grouped) dwellings are more reactive than prices of collective dwellings, because the decision chain is shorter for houses than for dwellings.

The cost for Insee is 0.2 Full Time Equivalent of a category A statistician. The main cost is for the Ministry of Ecology with 7 Full Time Equivalent (1 category A, 5 category B and 1 category C civil servants). External costs represent 500,000 € a year (360,000 € for salaries of surveyors, 140,000 € for reimbursement of incidental expenses).

The statistical burden of the CCI-NHCP survey for each of the 6,000 building permits works owner or prime contractor is about 20 minutes for the first phase and half an hour for the second phase, i.e. (6,000x1/3 + 2,500x1/2) = 3,250 hours a year.

No revision for CCI, because of use for rental escalation.

"Producer Price Index for construction of new residential buildings" (statistical index) could be revised

The revision policy is strictly the same nationally and in transmissions to Eurostat.

Provision of information about revision and advance notice of major changes in methodology: Not available.

Input data and final "CCI" data are never revised.

"Producer Price Index for construction of new residential buildings" (statistical index) could be revised (use of a new hedonic model on the past), but in practice we have not yet used this possibility.

Type of source:

"CCI-NHCP" survey

"CCI-NHCP” statistical survey conducted by the Statistical Service of the Ministry of Ecology, in charge of Housing.

Frame on which the source is based

Sitadel register: the exhaustive source about building permits in France, managed by the Ministry of Ecology.

Sample or census: sample

Criteria for stratification: Type of construction

Initially, about 6,000 building permits are selected every year, of which we expect 2,400 completed with prices and characteristics (some high constraints on dates and characteristics):

- pure individual: ~ 1,200 by year (until the first quarter 2022),

- grouped individual (single house in a subdivision, etc.): ~ 400 by year,

- collective dwellings: ~ 800 by year.

Threshold values and percentages: not applicable.

Frequency of updating the sample: quarterly

"EPTB" survey

"EPTB” statistical survey conducted by the Statistical Service of the Ministry of Ecology, in charge of Housing.

The survey on building lot prices (EPTB) covers the entire territory and questions individuals or households who receive a permission to built a single-family home. It is an exhaustive survey.

The survey on building lot prices produce information (type of acquisition, surface, price…) regarding building lots on which single family homes are to be constructed as well as information on the house itself such as its price, total floor surface, overview of construction progress, heating source or the type of project supervision.

About 5,600 completed or partially completed questionnaires are returned each month. That is roughly 17,000 questionnaires each quarter.

"CCI-NHCP" survey: Quarterly.

"EPTB" survey: Monthly.

Questionnaires used in the survey

"CCI-NHCP" questionnaires:

- 2 types for phase 1 (1 dwelling / 2 dwellings and more);

- 2 types for phase 2 (1 dwelling / 2 dwellings and more).

Planned changes in national questionnaires

Possible slight adaptations every two quarters, in order to take into account legal changes.

Data collection media : Paper.

Planned changes in data collection methodology

The statistical service of the ministry in charge of housing considers modifying the survey to become an online survey.

"EPTB" questionnaire:

One A4 page

Data collection media : Paper.

Filling time : 5-10 minutes

First of all, data are validated at Insee by internal consistency (first detection of outliers).

The hedonic models run twice: the first run detects other outliers (Cook's distance and Student statistics), and then the second run provides the final data, without the outliers.

Estimates for non-response in the "CCI-NHCP" survey

This concept of non-response is not really applicable: most of the 6,000 building permits selected quarterly for the first phase do not fit criteria necessary for the second phase (2,500 building permits usable requested). Moreover, the hedonic model relying on usable building permits of second phase does not need any imputation of building permits given up during the first phase.

Estimates for non-response in the "EPTB" survey

The non-response rate is about 30% each quarter.

Estimates for grossing-up to population levels

Not available.

Type of index

The producer price index for construction of new residential buildings is a chain-linked Paasche index, resulting from the aggregation of three sub-indices in proportion to the importance of each type: pure individual, grouped individual and collective.

Method of weighting and chaining

Chain-linked Paasche.

The weights of the type of construction are estimated by the number of dwellings beginning during the quarter and valuated by the price for a "standard" housing of the sample.

Reference 100 in 2015.

Planned changes in production methods

None

Annexes:
Formula for CCI compilation

WDA, SA and other calculations: None.

None