National Institute of Statistics (ISTAT)

Unit SED (Quick survey on building permits) 

Economic Statistics Directorate – Short Term Statistics Division

Via Tuscolana, 1788

I-00173 Rome

Quarterly indicators of building permits

"Quick Survey on building permits" is a quarterly sample survey; it collects information on new residential or non-residential building projects (even if demolished and entirely rebuilt) or enlargement of pre-existing buildings, with the exception of transformations and renovations. Information is supplied by those who ask for building permits or DIA (start-of-work notification) or SCIA (Start-of-work certified reports) or Public Administrations (for buildings in accordance with article 7 of Presidential Decree n. 380/2001) and is collected by municipalities that monthly validate on line questionnaires to make them available to Istat. A building permit is an authorisation to start to work on a building project. As such, a permit is the final stage of planning and building authorisations, prior to the start of work. Hence, the building permit indices are a good estimation for the future development and trends in construction. It may be the case that a proportion of permits are not used or there is a long time lag between permits and the start of the building project; however, an index based on these permits gives good indications of the workload for building in the near future. Data disseminated are referred to number of dwelling and square meters of useful floor area, communities and non residential activities. Building permits indicator is one of the PEEIs (Principal European Economic Indicators).

Classification of Types of construction (CC)

The coverage is based on the Classification of types of construction (CC) and covers Section 1 (buildings).

List and definition of variables:

1)  Number of new dwellings (total, in one and in two or more dwellings residential buildings)

2)  Square meters of useful floor area of new dwellings (total, in one and in two or more dwellings residential buildings)

3)  Square meters of residencies for communities (new buildings and enlargement of preexisting buildings)

4)  Square meters of non residential activities (total, in office buildings and in other non residential activities) (new buildings and enlargement of preexisting buildings)

Definition of variables is consistent with the European Commission Regulation No 1503/2006.

The target variable is measured as a flow during the calendar quarter.

Reporting unit information is supplied by those who ask for building permits or DIA (start-of-work notification) or SCIA (Start-of-work certified reports) or Public Administrations (for buildings in accordance with article 7 of Presidential Decree n. 380/2001); they fulfill an on line questionnaire. Municipalities monthly validate on line questionnaires to make them available to Istat.

The observation unit(s) are the residential and non residential building projects that have been authorized.

New building projects  for residential; New building projects  and enlargement of pre-existing buildings (for communities and non-residential). The average number is about 13000 per year.

The area covered by the index is the whole national territory.

Data are available since 2000. Questionnaire was last revised in 2010.

Not applicable – building permits data are expressed in absolute values.

The quantity of the residential buildings is measured in cubic meters of volume and square meters, dwellings in habitable floor area, that of non-residential buildings in square meters of area (absolute figures).

Data collection and production process are both montly, while indicators and data dissemination are quarterly.

Legal basis National level: Legislative Decree n. 301/2002 (Bulding permits and DIA); Law 122/2010 (SCIA); Presidential Decree n. 380/2001 and subsequent amendments. European level: Council Regulation (EC) n. 1165/1998 and subsequent modifications and integrations; Regulation of the European Parliament and of the Council n. 1158/2005.

Obligation on units to provide data: This is one of the compulsory surveys fixed every year by a Decree of the President of the Republic (D.P.R.).

No other transmissions to international organisations except Eurostat.

Links to different legal acts on statistics are presented on the website of Sistan - National Statistical System - (http://www.sistan.it/index.php?id=203).

Data trasmitted to Eurostat refer to 10 variables, but only 3 of them are disseminated at national level; so 7 variables are sent to Eurostat by using confidential flags.

Data are only transmitted to Eurostat in according to the Release calendar for euro indicators and disseminated nationally without a predefined release calendar.

Information on release are given via a hhttps://www.istat.it/it/informazioni-e-servizi/per-i-giornalisti/appuntamenti/calendario-diffusioni-ed-eventi

Data are released simultaneously to all interested parties by issuing the press release "Indicatori dei permessi di costruire".

There is no pre-release access

Data are transmitted to Eurostat no later than 90 days after the end of the reference quarter. Data are sent to Eurostat before the National release. These data are considered under embargo until the embargo expires.

Data are transmitted to Eurostat and disseminated to national users quarterly.

"Indicatori dei permessi di costruire" is a quarterly press release, it includes data and a technical explanatory note. In this document graphs and tables ease the data analysis. A short text explains recent evolution for the phenomenon.

• Data are made available to the public through yearly publication Annuario Statistico Italiano; it shows graphs and tables of quarterly indicators relating to the total number and square meters of useful floor area of dwelling in new residential buildings and total square meters of new non residential activities. 
• "Noi Italia - 100 statistics to understand the country we live in" includes an indicator for building permits at regional and national level, with international comparisons.

Time series are included in the ISTAT on-line database under the theme Industry and Construction.

The on-line database I.Stat is open and free, statistics are organised by theme in a two-level hierarchical tree. Data are presented in aggregate form in multidimensional tables; acting on variables, reference periods and the arrangement of heads and sides users can create custom tables and graphs. The system can be searched by keyword, theme and region.
A wide range of standard metadata facilitates the retrieval and understanding of statistics by users.

Before data dissemination, validated microdata are stored in an internal Oracle database.

Data are sent to Eurostat as confidential and used for European aggregates, part of them become free for dissemination at national level once published by Istat, the rest of data continue to be confidential.

A short description of the methodology and characteristics of the survey is contained in the technical explanatory note attached to the press release updated regularly.

Methodology and sources of the index are described in “Estimates based on preliminary data from a specific sub-sample and from respondents not included in the subsample”, (Falorsi, Alleva, Bacchini and Iannacone) Statistical methods and applications, Vol.14, n.1, Springer-Verlag, Feb. 2005.
A description of methodology and sources used in preparing statistics is also contained in the Information System for Survey Documentation and Quality Control (Siqual); it is available on ISTAT's Internet website http://siqual.istat.it/SIQual/index.jsp. This system manages qualitative information related to each single phase of a survey's production process (http://siqual.istat.it/SIQual/lang.do?language=UK)

More information about procedures applied for quality management and quality assessment is contained in the document "Quality Guidelines For Statistical Processes" - December 2012

https://www.istat.it/it/files//2013/04/Linee-Guida-v1.1-Versione-inglese.pdf

Istat quality policy is coherent with the European Statistics Code of Practice adopted in 2005 (and revised in 2011) and available at the link http://ec.europa.eu/eurostat/web/quality/european-statistics-code-of-practice, with Eurostat Quality Definition and with the recommendations of the LEG on Quality. Following the principles of the European Statistics Code of Practice, Italy has adopted the Italian Code of Official Statistics in order to promote quality improvements of the statistics produced by the Italian National Statistical System. Essential points of Istat quality policy are: process and product quality, documentation, respect for respondents, strengthening of statistical literacy and users’ orientation. For more details:https://www.istat.it/en/organisation-and-activity/quality-commitment

.

Quality assurance for this survey, in the framework of the quality policy of Istat, is based on the implementation of various activities, such as

1)      Prevention and monitoring of non-respondent, initiatives to foster survey participation and follow-ups of non-respondent units, penalties for non-respondent municipalities (see point 13.3);

2)      Checking of every paper questionnaire before data entry and submission of on-line questionnaires to some variable compatibility checks; checking of few fields during data entry;

3)      Quality control action based on fully deterministic and mixed procedures for the treatment of deterministic error and outlier detection and imputation; the first procedure is based on deterministic modelling, the second (BANFF procedure) is carried out by clerical reviewers and is based on data from previous editions of the same survey;

4)      Coherence control with previous data of the same survey;

5)      Elaboration of a set of standard quality indicators.

The overall quality assessment can be considered satisfactory as there is a compliance between the production process and Istat quality guidelines. Indicators are relevant with regard to users’s information needs, they are accurate and easily accessible and supported by metadata allowing for a perfect understanding of the data, they allows comparisons over time and respect the timing. A set of standard quality indicators are regularly analysed and shows the following indicators: coverage and unit nonresponse (since 2008), editing and imputation (since 2007), timeliness and punctuality (since 2003), resources (since 2005). They are available on Siqual (Information System on Quality), with graphs and tables that ease the analysis.

In 2014, this survey was submitted to an audit on quality in a testing phase.

Eurostat, local institutions (regions, provinces, municipalities), domain associations and organizations, media, enterprises, reseachers and students, internal users.

Users mainly manage these data for studying evolution and planning policies at national or local level.

Not available.

Eurostat's requirements in terms of time series are fulfilled (some data are flagged as confidential).

All the measures useful to reduce sampling and non sampling errors are implemented, so the overall accuracy of the index can be considered good. Moreover, between the first and the final release there is an average lag of 18 months; these indicators normally show a similar trend. 

In 2013 size sample was modified: from 814 to 1034 units on a total of 7983 municipalities; outlier treatment and adjustments for non-response are included, but an assessment of sampling errors is not available.

Non sampling errors are reduced by applying a set of actions. In the collection phase, for reducing late transmission and increase response rate, postal reminders, fax, email, telephone contact are used; for municipalities that do not comply for the year 2017, it will be applied a financial penalty, in according with the Legislative Decree n. 322/1989 (articles 7 and 11); paper questionnaires are checked before data entry. Moreover the on-line procedure used since july 2010 helps aiming at an increase of the response rate and debugging through a compatibility check on some main variables. Other actions are based on checking procedures of few single fields during data entry and automated error checking procedures after data entry. Response rate is around 71% at the time of the first release to Eurostat and around 77% at the second release (response rates weighted by populations).

The press release is pubblished on average 180 days after the end of the reference period; data are transmitted to Eurostat within 90 days using the confidential clause; data are provisional in the first release, then they are revised and become final in average 18 months after the reference period.

Questionnaires start to be filled in during the reference period, the data collection ends one month after (for on-line questionnaires) and two months after the end of the reference period (for paper ones).

On time releases to Eurostat within 90 days as required. For national dissemination only an internal calendar is fixed.

Data are disseminated at national level complying with the European definitions (CR 1503/2006); the whole national territory is covered by the same definitions and methods.

Annual data on dwellings and square meters of useful floor in new residential buildings per 1,000 households are available by NUTS 2 regions in the online pubblication "Noi Italia - 100 statistics to understand the country we live in".

Data are comparable from year 2000.

Comparisons with index of production in construction: trends are consistent and differences are explained by weight of this survey that accounts for about 40% of total data in the construction sector.

Variables V411 and V412 are coherent for common types of buildings.

There are two types of cost, one referring to 11 persons working full-time for the Index of building permits unit and another one linked to the data-entry; this data entry process is related only to paper questionnaires, used by municipalities that do not use the on-line application.

As regards burden, respondents need about 15 minutes to fill the questionnaire, while municipalities need about 2-3 minutes to validate and send it.

The assessment of these costs and burden covers the statistics for both STS and national requirements.

Revision policy is carried out to consider late incoming data and benchmarking procedures (see point 17.2); a revision calendar is prepared, but only for internal staff. Revisions include seasonal adjustment. Transparency of disseminated data is ensured: major changes in methodology and in others survey features are infrequent but are promptly disseminated in methodological notes attached to data releases or through letters sent to municipalities. Errors are rare but, if occurred, they are promptly corrected: in order to inform users, correct results are disseminated both to Eurostat and on the Istat website. There is an internal database where vintage data are stocked. The same revision policy is applied to data released nationally and transmitted to Eurostat.

Data revision is composed by 3 steps:

- a first revision is carried out after 90 days from the first release, taking into account new data arrived after the first estimate;

- a second revision is carried out in September in the year following the release of the first estimate as a result of a benchmarking procedure based on percentage changes;

- a third revision is carried out in September in the year t+2 of the release of the first estimate as a result of a benchmarking procedure which takes into account data of the yearly survey on building permits.

No revisions are carried out to take into account building permits revoked. A paragraph called "Revisioni" is included in the press release “Indicatori dei permessi di costruire” of each July and it contains differences with previous data disseminated. A methodological note contains further information on revision policy, criteria, periodicity.

Type of source:

1)  Natural or legal person applying for a building permit or a DIA (start-of-work notification) or a SCIA (Start-of-work certified reports) or Public Administrations (for buildings in accordance with article 7 of Presidential Decree n. 380/2001).

2)  Authority issuing the authorization to build (municipalities)

Survey is based on a sample. At the same time, a census survey is carried out also to in order to estimate the grossing-up to population level.

The methodology is based on a two-phase sample and on the hypothesis that the response mechanism is the result of a random processes whose parameters can be suitably estimated. Sample includes 176 self representative units, composed by all municipalities with more than 50.000 inhabitants and all “Capoluogo" municipalities; the 858 remaining municipalities are stratified and selected by a random process.

For stratification there are 4 Geographical areas (North-west, North-east, Centre, South-and-Islands) and Socio-demographic characteristics of Municipalities.

The Frequency of updating the sample is 10 years.

Monthly

The questionnaires used in the survey are separated for residential (Mod. ISTAT/PDC/RE) and non-residential (Mod. ISTAT/PDC/NRE) building. Communication of no building permits in the period from the municipalities is the Mod. ISTAT/PDC/Neg. The deadline for responses is 5 days after the reference month. Late responses are taken into account in the revision process.

Respondents fill an online questionnaire; then,  municipalities monthly validate them so they became available to Istat. If a municipality do not use the on-line application, respodents fill the paper questionnaire attached to the request of building permits and then Municipalities collect and send them to ISTAT by post. For reducing late transmission and increase response rate a set of actions are applied, e.g. postal reminders, fax, email and telephone contacts, penalties for non - respondent municipalities. Each worker periodically monitors and follows up the non response, through informatic procedures.

Data validation is applied following three type of rules:

Validation level 0: format and file structure checks

Validation level 1: intra-dataset checks based at record and cell level

Validation level 2: revision checks and time series checks (coherence control with previous data of the same survey).

Estimates for non-response: For 176 self representative municipalities, in case of PMR (Partial Missing Response) missing values are estimated through the average value available for the same municipality in the 12 previous months; in case of TMR (Total Missing Response) a minimum distance estimation model is used.

Estimates for grossing-up to population levels: For remaining 858 sample municipalities, missing values are estimated through a re-ponderation model based on values monthly availables for respondent units. Moreover, since at the time of the estimation assessment sample coverage is incomplete, additional information from responding non-sample municipalities is used.

Indicators sent to Eurostat are absolute figures.

There are no relevant Method of weighting and chaining.

Raw data are corrected for seasonal effects by TRAMO-SEATS procedure, version 2010 for Linux. Seasonal adjustment series are estimated for all ten series except for residencies for communities (CC113) and for office building (CC122). Adjustment models are completely reviewed once a year, with the release of data referred to the second quarter of the year and in conjunction with the application of the procedure, based on benchmarking and balancing techniques, that makes the quarterly series consistent with those coming from the annual structural survey on building permits. Each quarter, parameters are re-estimated and seasonally adjusted series are revised from the first quarter of 2000.

A Reg-Arima approach is used for seasonal adjustment. No calendar adjustment is performed since no significant effect is found for all series. However, the significance of calendar effects is checked when adjustment model are completely reviewed. 

Outliers (additive outliers, temporary changes, level shift, seasonal outliers) are manually fixed when seasonal adjustment models are revised. The filter length is automatically chosen as normally happens, when diagnostics on residuals are good, with the critical values for the outlier detection. Each series of the data-set follows a log-additive decomposition model and is directly adjusted.