Italian National Institute of Statistics

Integrated system on economic conditions and consumer prices unit

Via C. Balbo, 16 - 00184 Rome (Italy)

The House Price Index (HPI) measures the evolution of market prices of all residential properties that are purchased by households (flats, detached houses, terraced houses, etc.), both new and existing, independently if bought for own-occupancy or as an investment. The price of land is included in prices and weights.

The HPI is released quarterly and is composed by the following two sub-indices:

- Price index for newly built dwellings;

- Price index for existing dwellings.

The HPI compilation follows methodological standards set out in the Technical Manual on Owner-Occupied Housing (OOH) for HICP provided by Eurostat to ensure comparability of HPI indices across Member States.

Index number and weights are made available at the level of detail established by the Commission Implementing Regulation (EU) 2023/1470 of 17 July 2023. In particular the classification system for HPI is the following:

H.1. Purchases of dwellings;

H.1.1. Purchases of newly built dwellings;

H.1.2. Purchases of existing dwellings.

The HPI covers the whole household sector in particular the expenditure on purchases of dwellings carried out by households over the whole national territory.

The HPI covers purchases of newly built dwellings and existing dwellings.

The HPI compilation is based on administrative data; in particular, prices of dwellings are gathered from notarial deeds of sales data provided by Tax Office.

The HPI compilation is based on final market prices that are paid by households (VAT included); non-market prices are ruled out from the scope of the HPI (for example, the calculated final price for a dwelling that was developed by a step-by-step self-building is excluded).

The HPI are calculated using a chained Laspeyres formula. The reference base year for all indices is 2015=100.

The statistical unit of the HPI is the household who purchases a residential property.

The target population is the set of households who purchase residential properties, both newly built and existing, independently for own-occupancy or as an investment.

HPI covers expenditure on purchases of dwellings carried out by households over the whole national territory, excluding two provinces (Trento and Bolzano) and few municipalities situated in Veneto (nr. 2), Lombardia (nr. 2) and Friuli (nr. 40) that are regulated by a different cadastral system (totalizing about 2.6% of Italian population).

HPIs for geographical areas have been disseminated starting from October 2018. National HPI is a weighted average of the regional HPIs.

HPI data are available from the third quarter 2007; at the moment indices are published starting from the first quarter of 2010.

In the past House Sales indicators have been produced for a few years and only on experimental base. These experimental indicators have not been disseminated.

The reference base year is 2010 (2015=100).

Following units are used in HPI indicators:

• Index figures with reference to the year 2015;
• Percentage change on the same period of the previous year (rates);
• Percentage change on the previous period (rates);
• Percentage share of the total (weights);
• Total value of transactions expressed in national currency.

In the past House Sales indicators have been produced for a few years and only on experimental base. These experimental indicators have not been disseminated.

Quarter (indices and rates).

* Regulation (EU) 2016/792 of the European Parliament and of the Council of 11 May 2016 on harmonised indices of consumer prices and the house price index.

* Commission Implementing Regulation (EU) 2023/1470 of 17 July 2023 laying down the methodological and technical specifications in accordance with Regulation (EU) 2016/792 of the European Parliament and of the Council as regards the house price index and the owner-occupied housing price index)

*2020-2022 Italian National Statistical Programme - 2022 update, entered into force with the decree of the President of the Republic of 11 July 2023 and published in the Official Journal -general series – nr. 201 of 29 August 2023.

Istat is the unique national Agency responsible for the Italian HPI.

There are not national regulations which deal specifically with HPI confidentiality.

Data treated are confidential and may be accessed only by the staff responsible for the HPI when indicated in internal documents listing people working on personal data.

No procedures in protecting confidentiality have been adopted.

There is a published release calendar. It is available at https://www4.istat.it/en/information/journalists/press-room-calendar

The release calendar is available on Istat website on December of each year (it gives the precise release dates for the coming year). The calendar is distributed to the press and is available free of charge to interested parties.

Data are disseminated to all users through Istat website. No user has pre-publication access to the data.

HPI indices are disseminated quarterly; HPI weights are disseminated annually.

About House Sales indicators, in the past they have been produced for a few years and only on experimental base but these experimental indicators have not been disseminated.

Regular on-line press release at quarterly intervals available on https://www.istat.it/en/archivio/house+prices

Not available.

Time series starting from 2010 are available on the data warehouse I.Stat (http://dati.istat.it)

Micro-data are not disseminated because forbidden by the agreement between Tax Office and Istat.

HPI data are additionally published in the annual Italian Statistical Yearbook (ASI) and in "Noi Italia. 100 statistics to understand the country we live in".

Methodological note attached to each press release of the index (italian version more detailed at https://www.istat.it/it/archivio/prezzi+abitazioni; english version at https://www.istat.it/en/archivio/house+prices).

Not available at the moment.

Continuous efforts to strengthen and improve the procedures of data control checks and adjustment for quality changes are carried out according to the guidelines and standard set out in the Technical manual on Owner-occupied housing for HICP..

Istat monitors the quality of data set provided by Tax Office, the production process and carries out dissemination according to guidelines and standards set out in the Technical manual on Owner-Occupied Housing for HICP..

First of all, HPI respond to the needs of Commission which releases the report on the Macroeconomic Imbalance Procedure (MIP); in fact HPI (deflated by consumption deflator - the annual growth rate) is one of MIP indicators.

HPI is also used by the Commission for international comparison.

Moreover HPI is one of the FSIs (Financial Soundness Indicators) of International Monetary Fund. Finally HPI responds to the needs of users interested in the behaviour of the domestic house price market.

To focus on user needs Istat has set up, starting from December 2011, the Committee of statistical information users (CUIS), that is the place where ISTAT deals with users’ needs in order to improve its supply of statistical information.

HPI fully meets the recommendations included in the Manual; all expenditure categories defined by Commission Implementing Regulation (EU) 2023/1470 of 17 July 2023 are are covered.

HPI is perceived to be accurate and to reflect the overall behaviour of the housing market in a good manner.

HPI is not subject to sampling error because index calculation is based on administrative data that cover the universe of dwelling transactions (except only two provinces) (Trento and Bolzano) and few municipalities situated in Veneto (nr. 2), Lombardia (nr. 2) and Friuli (nr. 40) that are regulated by a different cadastral system).

Administrative data provided by Tax Office are used minimizing coverage errors (administrative data cover expenditure on purchases of dwellings carried out by households over the whole national territory, excluding two provinces (Trento and Bolzano) and few municipalities in Veneto, Lombardia and Friuli Venezia Giulia that are regulated by a different cadastral system and that represent about 2.6% of Italian population).

Provisional indices are transmitted to Eurostat 85 days after the end of reference period; final ones are transmitted in the following quarter.

National HPI is published approximately 80/85 days after the end of the quarter.

Indices are transmitted to Eurostat according to its own calendar.

Data are comparable across NUTS II regions (South and Islands are aggregated).

Data are comparable over time.

Not applicable.

Data are checked for internal consistencies. In particular the net weight of the OOHPI, covering the purchase of newly built dwellings expenditures, is checked annually against its gross HPI weight counterpart.

Costs relate to human resources involved in HPI production process (for the time being two persons working 0.3 FTE who are responsible for data collection, processing, analysis (occasionally assisted by an IT expert) and press releases). Concerning burden on respondents, administrative sources are used to produce HPI indicators.

HPI released each quarter are provisional and they are subject to revision with new information available the following quarter. Final indices relating to the previous quarter are published in the following quarter.

About House Sales indicators, in the past they have been produced for a few years and only on experimental base but these experimental indicators have not been disseminated. In general, House Sales indicators revision policy will be the same as HPI.

HPI released each quarter are provisional and they are subject to revision when new revised dataset are available (the following quarter).

Final indices relating to the quarter t are published in the quarter t+1. Users are informed about revision in the press releases. The text of the press release contains a note explaining which one is a preliminary HPI estimate subject to revision and which one is definitive. No actions are available for the moment to prevent the need for such revisions in the future.

See below.

HPI compilation is based on administrative data (for both newly built and existing dwellings).

HPI weights are based on the same database used for index actual dwelling transaction values. The partition of the total HPI weight into newly built and existing dwellings is done using information available from the same source as used for prices and statistical information. In particular VAT/fiscal code of the sellers are combined with a statistical archive developed by Istat that is the archive of active enterprises with the aim to individuate the economic activity of the seller. This information is helpful for separating between newly built and existing dwellings.

In the past House Sales indicators have been produced for a few years and only on experimental base. These experimental indicators have not been disseminated. The main source  used has been notarial deeds of sales data provided by Tax Office (the same as HPI).

Data are collected quarterly.

Data are encrypted by Tax Office and transferred to Istat using a secure transfer protocol.

Data validation principally consists in outlier detection. This is made:

• Using thresholds for surface and price per square meters pre-defined on the base of the Observatory of Real Estate Market data and Land Office experts suggestion;
• data cleaning is done to remove cases of obviously erroneous data;
• minimum and maximum values are investigated and removed if believed to be suspect.
  A regression model is used to detect any remaining outliers.

See below.

See below.

HPI is a Laspeyres-type price index.

The elementary index is a Jevons index adjusted by a quality change index calculated from a vector of regression coefficients obtained by the estimation on a chosen reference set of price observations. Elementary indices are then aggregated to obtain the two sub-indices (newly built and existing dwellings) and the overall index.

The general aggregate index is calculated as a weighted arithmetic average of stratum elementary indices where the weight of its stratum is proportional to its importance on the total expenditure of households to buy houses (weights are updated yearly and are estimated on the basis of the same data base used for processing price information).

HPI weights are based on actual dwelling transaction values. The partition of the total HPI weight into new and existing dwellings is done using information available from the same source as used for prices and statistical information. In particular  VAT/fiscal code of the sellers are combined with a statistical archive developed by Istat that is the archive of active enterprises with the aim to individuate the economic activity of the seller. This information is helpful for separating between newly built and existing dwellings.

The weight reference period is always the previous year. Elementary aggregate weights are price updated (quantity are updated using mean prices detected during the fourth quarter of the year t-1).

The chaining of the series is carried out by multiplying the quarterly indices (base fourth quarter t-1=100, 6 decimal places) of the current year for the index of the fourth quarter of the preceding year, expressed in reference base (2015=100, 6 decimal places).
The chained indices are then rounded to the first decimal place for publication.

HPI are calculated using stratification combined with hedonic method to take into account of quality changes.

Re-pricing methods imply the use of a hedonic function to clean prices from the influence of qualitative features and geographical locations. Hedonic function is estimated, usually, for each year y, during the indices re-basement, through a regression model that uses the data of the previous year (y-1): estimated regression coefficients are kept constant for the whole reference year (y).

However, for indices of the year 2023, regression coefficients estimated in 2021 were used.

A model is estimated for 11 geographical area (North-west, North-east, Centre, South and Islands) and for 4 big cities (Rome, Milan, Turin and Palermo) and for newly built and existing houses separately (in order to guarantee a minimum amount of observations in each stratum, total strata are 104, whereof 75 are referred to the existing ones and 29 to the new ones).

In the model, the main explicative variables are the following ones: a) natural logarithms of the surface, b) percentage of the surface of the house appurtenance, c) house typology, d) the floor; the dependent variable is the natural logarithms of the price for square metre.

No SPAR method is used.

Each stratum is defined on the base of territorial variables, demographic variables and dwelling size. The criteria adopted for the stratification differs between new and existing dwellings and guarantees a minimum number of observations per stratum. In total, the identified stratum are 104, of which 29 refer to new dwellings and 75 to existing dwellings.

See below.

Transaction price is the price paid when the final deed is signed and the ownership rights are transferred. Prices are included in the quarter when final deed of sale is signed.

VAT is included in the price of newly built dwellings.

Taxes other than VAT are not included in the price of dwellings.

Government subsidies are not taken into account.

The price of land is included in both prices and weights.

Housing cooperatives are covered but they are not detectable at the moment.

Purchases of a share of a dwelling property at reduced prices (for example a part of an inherited dwelling) are examples of non-market price transactions; they are excluded from index calculation but included in weights calculation.

Transactions including more than three properties are exluded from calculation.

Transactions of part of dwellings (part of inheritance) are included in weights calculation for HPI; the same database is used for house sales indicators (still esperimental).

Outlier detection is made:

• Using thresholds for surface and price per square meters pre-defined on the base of the Observatory of Real Estate Market data and Land Office experts suggestion;
• data cleaning is done to remove cases of obviously erroneous data;
• minimum and maximum values are investigated and removed if believed to be suspect.
  A regression model is used to detect any remaining outliers.

Not implemented.

No adjustments other than quality adjustment are applied to the HPI data.

None.