Statistics Lithuania

Short Term Business Statistics Division

29 Gedimino Ave, LT-01500 Vilnius, Lithuania

Index of Production in Construction (IPC) is calculated to measure the short-term changes in volume of construction work carried out, broken down into building construction and civil engineering.  Indicator is needed to analyse the development of the construction activity in the country. Data on own-account construction work carried out by construction enterprises are collected.  IPC is presented in unadjusted, calendar adjusted and seasonally adjusted forms.

National version (EVRK Rev. 2) of the Statistical Classification of Economic Activities in the European Community (NACE Rev. 2);
Classification of  Types of Constructions (CC).

IPC covers economic activities listed in Section F of NACE Rev. 2 (Construction). All size classes are covered. The breakdown into Building and Civil engineering is based on CC classification.

IPC is the relative measure of the trend in construction work carried out in the reference period.  IPC is calculated based on the volume of own-account construction work carried out by construction enterprises within the country. Construction work carried out refers to the value (VAT excluded) of all kinds of work carried out when building a new structure, reconstructing, repairing (restoring) or demolishing an existing structure for a customer (sale) or for own needs. It excludes  the value of land and design work.

Reporting unit - enterprise, observation unit - KAU

The frame for identifying units for the population is the Business Register, it includes information about enterprises activity code by NACE Rev. 2, state, turnover, number of persons employed, address and others.   

In 2022 the  frame population comprises 10274 observation units of all sizes and types with the main economic activity defined in the section F of NACE Rev.2.

Geographical coverage of the IPC is whole territory of Lithuania. The activities outside the Lithuania are not included.

Quarterly data on IPC are available since I Q 1998, monthly data – since first month 2015.

Year 2015 = 100

Indices are transmitted to Eurostat, nationally data on production in construction are published in absolute figures (thousand EUR), indices and percentage changes. 

 The reference period of the data transmitted to Eurostat is Quarter.

The European Parliament and Council adopted the European Business Statistics (EBS) Regulation (EU) 2019/2152 on 27 November 2019, followed by the Commission Implementing Regulation 2020/1197 laying down technical specifications and arrangements pursuant to the mentioned EBS Regulation (General Implementing Act).

All the non-confidential data are available on the Official Statistics Portal selected data are transmitted to Eurostat. Data are sent to other international organizations on request.

In the process of statistical data collection, processing and analysis and dissemination of statistical information, Statistics Lithuania fully guarantees the confidentiality of the data submitted by respondents (households, enterprises, institutions, organisations and other statistical units), as defined in the Confidentiality Policy Guidelines of Statistics Lithuania.

Statistical Disclosure Control Manual, approved by Order No DĮ-107 of 26 April 2022 of the Director General of Statistics Lithuania;

The State Data Governance Information System Data Security Regulations and Rules for the Secure Management of Electronic Information in the State Data Governance Information System, approved by Order No DĮ-202 of 27 August 2021 of the Director General of Statistics Lithuania.

Statistical information is published in accordance with an Official Statistics Calendar, release dates are announced one year in advance.

Official Statistics Calendar

Statistical information is published on the Official Statistics Portal according to an Official Statistics Calendar, following the Official Statistics Dissemination Policy Guidelines and Statistical Information Preparation and Dissemination Rules.

No users have advance access to the statistical data. National data are disseminated to all users at the same time. Quarterly IPC totals and detailed breakdowns are transmitted to Eurostat and disseminated nationally at the same day.

Data are transmitted quarterly to Eurostat.

Data are disseminated monthly, quarterly and yearly to national users.

A regular monthly press releases linked to the data on production in construction are published on the 45^th day after the end of the reference month at 9 a.m. 

Data on totals and detailed breakdowns are disseminated in absolute figures (thousand EUR), indices and growth rates (unadjusted, calendar or seasonally adjusted) for national users.

The annual statistical information is provided in common electronic publications:  Business in Lithuania, Lithuania in Figures http://osp.stat.gov.lt/en/statistikos-leidiniu-katalogas

Data is accessed in the Database of Indicators (Business statistics -> Construction -> Indicators of work carried out by construction enterprises).

Main tables produced: monthly, quarterly IPC (Total, Building construction and Civil Engineering) compared with base year, with the previous period and compared with the same period of the previous year (unadjusted, seasonal and calendar adjusted data).

Statistics Lithuania, in response to the needs of users of statistical information, provides them with access to open data sets with data on statistical observation units. More information is available on the Official Statistics Portal, at Open data.

Eurostat's database

The survey methodology (in Lithuanian) is presented on the Official Statistics Portal.   

Metadata for IPC is updated every month and published on the Official Statistics Portal.

The quality of statistical information and its production process is ensured by the provisions of the European Statistics Code of Practice and ESS Quality Assurance Framework. 

In 2007, a quality management system, conforming to the requirements of the international quality management system standard ISO 9001, was introduced at Statistics Lithuania. The main trends in activity of Statistics Lithuania aimed at quality management and continuous development in the institution are established in the Quality Policy. Monitoring of the quality indicators of statistical processes and their results and self-evaluation of statistical survey managers is regularly carried out in order to identify the areas which need improvement and to promptly eliminate the shortcomings.

The monitoring of the quality of IPC is based on the ESS quality requirements (relevance, accuracy, timelines and punctuality, accessibility and clarity, coherence and comparability). Quality of data is checked at all stages of data processing.

The inspection of statistical survey, testing of statistical questionnaire and updating survey methodology, self-assessment of survey manager are carried out.
When the survey is closed for a reference period and IPC is calculated the response rate is about 99 per cent.

Strengths:  monthly, quarterly data are comparable with an annual data, length of quarterly time series since 1998, monthly– since 2015.

Deficit: misclassification of main activity of enterprises occurs, when the sample for responding year is done.

The main users of statistical information are State and municipal authorities and agencies, international organisations, the media, research and business communities, students. Data is transmitted to Eurostat.

The users' needs are satisfied without a breach of the confidentiality principle.

From 2005, user opinion surveys have been conducted on a regular basis. Official Statistics Portal traffic is monitored; website visitor opinion polls, general opinion poll on the products and services of Statistics Lithuania, target user group opinion polls and other surveys are conducted.

The compilation of a user satisfaction index is launched. The said surveys are aimed at the assessment of the overall demand for and necessity of statistical information in general and specific statistical indicators in particular.

IPC is compiled for building and for civil engineering, the distinction between building and civil engineering is based on the  CC classification. The contents and level of detail of the indicator comply with the EU requirements.

The frame population in the LT Business Register includes  10276 operating construction enterprises of all sizes and types with the main economic activity defined in the section F of NACE Rev.2. The statistical survey is conducted applying a stratified random sampling method. The sampling frame is based on actual information about enterprises in operation, their economic activity and the number of persons employed from the Statistical Business Register. Enterprises are stratified by number of persons employed: all enterprises with 50 and more persons employed are included in the sample, for other enterprises, a stratified random sampling method is used. A representative sample includes 1000 units.

To calculate the estimates Horvitz-Thompson estimator is using, that is unbiased.

The accuracy of the estimates is reached by eliminating non-sampling errors, such as coverage, non-response, response mistakes and processing errors, also by calculating sampling error.

Coverage errors appear by reasons of misclassification of  units and changes in state of reporting units.

It is compulsory to respond to the survey. In case of non-response, units are contacted by telephone and email.

The sampling error in percent is calculated for main indicators of survey and published monthly.

Usually survey results are revised  in the end of the year if need of revision appears.

A coefficient of variation is used as a measure of the precision of an estimate. The coefficient of variation is calculated for indicators on total construction, building and civil engineering.

Over-coverage error arises from the presence of units not belonging to the section F (construction) and deaths of sampled units.

The survey is compulsory and non-response is insignificant. Response rate at the data published nationality and provided to Eurostat is the same.

Measurement errors occur during data collection. To detect processing or response mistakes the electronic questionnaire and programming technical specification for data checking at the data entry stage are used. 

Statistical information is published on the 45^th day after the end of the reference month, quarter.

Statistical information is published in accordance with an Official Statistics Calendar.

In case of delay, users are notified in advance by indicating the reason and a new date of publication. 

100 per cent of statistical information is released on time.

Statistical information is comparable among the EU countries.

There are no breaks in time series since 1998.

The data are confronted with data from VAT declarations from State Tax Inspectorate of Lithuania. Differences between data exist because data from VAT includes work done by subcontractors.

IPC trend is comparable with the  Structural Business Statistics data and with National Accounts.

The components CC1 and CC2 are coherent with total construction.

The administrative burden for respondents and cost are determined for statistical survey in total and it is not possible to calculate for separate indicators.

Total cost (NSI) – 5560 hours per year.

Total burden (respondents) – 7264 hours per year.

Revisions of statistical indicators are performed in accordance with the General Principles behind the Performance, Analysis and Announcement of Revisions of Statistical Indicators.

Vintage data is not available.

The same revision policy is applied to IPC data released nationally and transmitted to Eurostat.

Long-term planned (major) revisions are conducted after amendments in legal acts, due to changes in the base year, methodology and classifications entail revisions. The time series from 1Q 1998 are back-casted.

Seasonal adjustment models are revised once a year, data are revised for current year and 3 years backwards.  

Unscheduled revisions are performed when significant respondents’ errors detected after statistical indicators had been officially published.  The announced data are marked as revised.

*As compared to the corresponding period of the previous year

** As compared to the previous period

The monthly statistical Survey on Construction Enterprises Activity is carried out applying a sampling method.

The sampling frame is based on information about enterprises in operation, their economic activity and the number of persons employed from the Statistical Business Register.

A representative sample (1000 units) comprising  10.0 percent of all construction enterprises in operating in the country was drawn for the year 2022.

All enterprises with 50 and more persons employed were included in the sample, for other enterprises, a stratified random sampling method is used.

Activity and number of persons employed are used to stratify the population of active construction enterprises. A new sample is drawn every year.

The sample sizes in stratum (persons employed classes):

1-4 : 2.6%

5-9 : 4.7 %

10-19 : 13.1 %

20-49 : 35.2 %

50 and more : 100%

Monthly. The collected data are referring to full reference month

The data are collected using the questionnaire Report on Performance of Construction Works (KS-01) via the Internet, in the electronic statistical data preparation and transmission system  e-Statistics and email.

Telephone contacts and emails or electronic reminders are used to increase the rate of response. Statistical units are obliged to provide the data according to the Law on Statistics. Fines can be given.

Questionnaire (only in Lithuanian).

The first stage for error checking and correcting is the micro data collection stage.

Data is checked and corrected when provided by the respondents via the electronic statistical data preparation and transmission system  e-Statistics. Quality control requirements are presented in the survey programming technical specification, therefore response mistakes and processing errors are checked at the data entry stage. 

Detection of large differences between the response and the values provided by VAT data.

Detection of large differences between the current month and historical data.

Correction of a variable using the sum of other variables.

Statistical data completeness and reliability are checked; the item and unit non-response rates are determined.

The second stage of validation concerns  aggregated data. 

Detection of large differences between the current and past values (growth rates) – the current value should be compared with the previous values and with the corresponding of the previous year.

Before sending to Eurostat data file is validated and converted into SDMX_ML format using SDMX Converter Tool.

The data file is transmitted to Eurostat in STSCONS domain by eDAMIS. 

In estimating the parameters, the missing data related to item non-response and unit non-response are compensated. 

For the estimation of data on unit non-response, administrative data source (State Tax Inspectorate) is used. For other enterprises, that could not be estimated using administrative data source, the corrected weights are used.

For the imputation of item non-response, the previous periods' data submitted by those units  are used. 

The estimates are computed using stratified random sampling formulas. Results are grossed up  from the sample to the population. The grossed up results disseminated nationally on http://osp.stat.gov.lt/en . Production of statistical information process is described in Monthly statistical survey on construction enterprises methodology (only in Lithuanian).

IPC are calculated in comparison with the previous period , the respective period of the previous year and monthly average of base year 2015. 

The volume of production in construction is deflated using construction input price index (2015 = 100). It is carried out separately for residential building, non-residential building (results are summed up as building production) and civil engineering.

The time series are calendar and seasonally adjusted  also trend series are produced, using the TRAMO/SEATS method  in JDemetra+(v. 2.2.2). 
An indirect method is used for the IPC seasonal adjustment.

The model is chosen using a detailed analysis module. Logarithm transformation is tested. Seasonal and calendar factors are estimated, 2 regressors (Working days) are used, outliers are detected and corrected. The global economic crisis outlier  is used. Critical value and automation of outlier detections disabled (using default options).

The national holiday's calendar is used. It is used additive decomposition on seasonal adjustment.
Traditional statistical methods are used for the validation of the models. Diagnostic statistics are used to control the quality of seasonal adjustment. The significance levels, the number of outliers and statistics on residuals (residual normality, residual asymmetry, residual seasonality etc.) in data series are controlled.
Seasonal adjustment models and parameters are fixed for one year and are revised once a year when the data of all reporting months are available. The horizon of revision takes 4 years: current year and 3 years backwards. 

None