REPUBLIC OF SLOVENIA ANNUAL QUALITY REPORT FOR THE SURVEY ANNUAL STATISTICAL SURVEY ON CONSTRUCTION (GRAD/L) FOR 213 Prepared by: Špela Kastelic Date: September 214 1/1
Table of Contents Methodological Explanations... 3 1 Relevance... 5 1.1 Rate of Unavailable Statistics... 5 2 Accuracy... 5 2.1 Sampling Errors... 5 2.1.1 Sampling Error... 5 2.2 Coverage Bias... 5 2.2.1 Coverage Bias... 5 2.3 Non-Sampling Errors... 5 2.3.1 Non-Response Errors... 5 2.3.1.1 Unit Nonresponse Rate... 5 2.3.1.2 Item Nonresponse Rate... 5 2.3.1.3 Imputation Rate... 6 2.3.2 Coverage Errors... 6 2.3.2.1 Overcoverage Rate... 6 2.3.3 Measurement Errors... 6 2.3.3.1 Editing Rate... 6 2.3.4 Rate of the Coherence of Data Sources... Error! Bookmark not defined. 3 Timeliness and Punctuality... 6 3.1 Timeliness... 6 3.1.1 Timeliness of the First Results... 6 3.1.2 Timeliness of Final Results... 6 3.2 Punctuality... 7 3.2.1 Punctuality of the First Results... 7 4 Accessibility and Clarity... 7 4.1 Accessibility... 7 4.1.1 Means Used for the Dissemination... 7 4.1.2 Rate of Means Used... Error! Bookmark not defined. 4.2 Clarity... 7 4.2.1 Results Presented... 7 4.2.2 Level (Detail) of Presentation... 7 5 Comparability... 8 5.1 Comparability over Time... 8 5.1.1 Length of Comparable Time Series... 8 5.1.2 Breaks in Time Series... 8 5.2 Geographical Comparability... 8 5.2.1 Comparability with Other Members of the European Statistical System... 8 5.3 Seasonal Adjustment... 8 6 Coherence... 9 6.1 Coherence between Provisional and Final Data... 9 6.1.1 Coherence between Provisional and Final Data... 9 6.2 Coherence with the Results of the Reference Survey... 9 6.2.1 Reference Survey... 9 6.2.2 Coherence with Reference Data... 9 7 Costs and Burdens... 9 7.1 Survey Costs of the Office... 9 7.2 Burden of Reporting Units... 1 2/1
Methodological Explanations Brief Description of the Survey The annual survey on construction provides data on the value of construction put in place at the facility, regardless of whether these facilities are buildings or civil engineering works, and regardless of whether they are a new construction, reconstruction or maintenance work (investment or regular). It also provides information on the value of the organization that implements building projects, i.e. the service to mobilize financial, technical and physical resources for the construction of buildings (residential and non-residential) for subsequent sale. Since 28, in the observation are included - in line with the new version of the Standard Classification of Activities (NACE Rev. 2) - in addition to construction companies also companies dealing with the organization of building projects. Data on the value of construction put in place in buildings or civil engineering works, are presented by the Classification of Types of Constructions (CC-SI) (http://www.stat.si/eng/klasje.asp). Data on construction are important in the analysis of the construction activities and are necessary, among other tasks, also to calculate the gross domestic product (GDP). Legal Basis for the Survey The legal basis for the survey are the National Statistics Act (Official Journal of RS, No. 45/95 and 9/1) and the Annual Programme of Statistical Surveys (LPSR) for 214 (Official Journal of RS, No. 97/213). According to the stated legislation, provision of required data for the selected companies is mandatory. Observation Units The annual statistical survey on construction involves companies which implemented construction works in the observed year or organized the building projects (the service to mobilize financial, technical and physical resources for the construction of residential and non-residential buildings for subsequent sale) as a prime contractor or as a researcher. The main contractor reports the required data for potential subcontractors. Coverage In determining the list of observation units the method of threshold coverage is used. The survey sample includes all construction companies and companies dealing with the organization of building projects (section F - construction after Nace Rev 2, activities 41, 42 and 43), whose value of income for the taxation of value added tax in the reference year reaches a certain threshold (EUR 25, for 213), and the establishments engaged in construction activity or the organization of building projects which have at least 2 employees, and some other companies that perform construction work. The criterion of 74% coverage in the activities F Construction is taken into account. The GRAD/L survey for the reference year 213 included 834 respondents. Key Variables Key variables of the annual statistical survey on construction are: type of investor, type of works, type of construction according to the CC-SI classification, value of construction works put in place, value of organization of building projects. Key Statistics 3/1
Key statistics of the annual statistical survey on construction are: the value of construction put in place by the type of investor, the value of construction put in place by the type of works put in place, the value of construction put in place by the type of construction according to the CC-SI classification. Questionnaire The questionnaire is available on the website: http://www.stat.si/metodologija_vpr_prikaz.asp?vpr_id=2455&pod=19&kon=&leto= Methodological Explanations The methodological explanations are available on the website: http://www.stat.si/eng/metodologija_pojasnila.asp?pod=19, Value of construction put in place, Slovenia. 4/1
1 Relevance 1.1 Rate of Unavailable Statistics The basis for obtaining data on construction works is the LPSR. Data are collected primarily to meet the needs of public institutions and derivative statistics (GDP), as well as of operators in the industry. Therefore all the necessary statistics are calculated and these commitments are met in full, so the percentage of missing statistics equals zero (). 2 Accuracy 2.1 Sampling Errors 2.1.1 Sampling Error The survey is carried out on the basis of a random sample, so the estimates do not contain the sampling error. 2.2 Coverage Bias 2.2.1 Coverage Bias Explicit estimates of bias arising from the threshold of non-response and measurement errors can not be calculated because there are no adequate data from other sources. The sample does not cover the entire population, but it covers the population threshold. In view of the value of income for the taxation of value added tax in the reference year, 74% of the total population is captured. In 213, the revision of the survey was conducted in which we analysed nonresponse and prepared the procedures to minimize the bias due to non-response. Weighting was introduced. In this way the bias due to threshold coverage was reduced. 2.3 Non-Sampling Errors 2.3.1 Non-Response Errors 2.3.1.1 Unit Non-Response Rate Table 2.1: Unit non-response rate Reference period 213 Number of adequate units 695 Number of non-responses 216 Non-response rate 31.1 2.3.1.2 Item Non-Response Rate There is no item non-response in the survey, as in the phase of collecting and compiling the data the questionnaires are checked as to whether they are completed in full and item nonresponse is removed and examined. If not, the unit will be contacted by telephone or via e- mail and from it the missing information is obtained. In the phase of statistical processing 5/1
thus all data of the units which report all the information required are entered and therefore there is no item non-response. 2.3.1.3 Imputation Rate In this survey we do not use the imputation of missing data. Thus, the imputation rate is zero (). 2.3.2 Coverage Errors 2.3.2.1 Overcoverage Rate Table 2.2: Overcoverage Rate Reference year 213 Number of units in the Address book 834 Number of inadequate units 139 The rate of over-coverage (%) 16.7 2.3.3 Measurement Errors 2.3.3.1 Editing Rate Tabela 2.3: Editing Rate for the key variables. 213 Value of construction works put in place 1,3 % value of organization of building projects,3 % The table shows only data that were edited in the process of automatic logical control. Data that were edited before the process of logical control are not shown. 3 Timeliness and Punctuality 3.1 Timeliness 3.1.1 Timeliness of the First Results Table 3.1: Timeliness of the First Results Reference period 213 Date of publishing 24. 7. 214 Time lag (days) T + 25 3.1.2 Timeliness of Final Results Data published in the First Release are also final data, so the timeliness of final data equals the timeliness of the First Release. 6/1
3.2 Punctuality 3.2.1 Punctuality of the First Results Table 3.2: Punctuality of the First Results Reference period 213 Announced date 24. 7. 214 Publishing date 24. 7. 214 Difference 4 Accessibility and Clarity 4.1 Accessibility 4.1.1 Means Used for the Dissemination Table 4.1: Means Used for Dissemination No.: Mean Used Website 1 (e.g. First Release, E-release) YES 2 Publication in the SI-STAT Data Portal YES Publication in the interactive web tools (e.g. Interactive Statistical Atlas of Slovenia, Thematic 3 Cartography) NO 4 Ad hoc prepared data for users according to their specification YES 5 Data available through telephone answering machine NO General printed publications 6 (e.g. Statistical Yearbook, Slovenia in Figures) YES Thematic printed publications 7 (e.g. Rapid Reports, Brochures) NO External databases 8 (e.g. Social Science Data Archives, Eurostat, OECD databases) NO 9 Statistically protected micro data NO 1 Preliminary access to data according to standard protocol NO To demonstrate the results 4 ways of expressing the results were used, which means that the rate of the methods used is 4%. 4.2 Clarity 4.2.1 Results Presented All results are presented as absolute numbers. 4.2.2 Level (Detail) of Presentation Level data disemination are broken down by: Year The type of construction by CC-SI classification to the level of class 7/1
Investor: Investor total Natural person Legal person: - Together - Public Administration - Utilities - Institutions and other non-profit organizations - Businesses and other organizations Type of construction activity: new construction extension reconstruction and conversion Investment maintenance regular maintenance work Standard Classification of Activities NACE Rev 2: NACE Rev 2 total F Construction other activities. 5 Comparability 5.1 Comparability over Time 5.1.1 Length of Comparable Time Series After having improved the methodology of weighting, the data are published from 21 onwards. 5.1.2 Breaks in Time Series There are no breaks in the time series. 5.2 Geographical Comparability 5.2.1 Comparability with Other Members of the European Statistical System The data are not comparable with other members of the European Statistical System because the survey methodology was not determined by the EU. 5.3 Seasonal Adjustment The seasonal adjustment procedures in this survey are not applied. 8/1
6 Coherence 6.1 Coherence between Provisional and Final Data 6.1.1 Coherence between Provisional and Final Data Preliminary data are not disclosed. We publish only final data. 6.2 Coherence with the Results of the Reference Survey 6.2.1 Reference Survey In the sample of the survey GRAD/M for the reference year 212 there are included all construction companies and companies dealing with the organization of building projects (Section F of the NACE Rev 2, activities 41, 42 and 43), whose value of income for the taxation of value added tax in 212 reached at least EUR 78,, and the establishments engaged in construction activity or the organization of building projects that have at least 2 employees, and some other companies that perform construction work. In the address book for the reference year 213 there were included 235 units. The survey takes place monthly on the basis of the data obtained through questionnaires (postal survey). 6.2.2 Coherence with Reference Data Table 6.1 shows the indices of the reported values on construction works for the surveys GRAD/M and GRAD/L in 213, calculated to the average of 21. The differences are mainly due to the differences in the sample of these two surveys. In the sample of the GRAD/L survey more reporting units are included than in the sample of the GRAD/M survey. Since data for GRAD/L are collected for the previous year and for GRAD/M for the current year, when establishing a sample for GRAD/L all changes detected in reporting units (demographic changes, bankruptcies, etc.) in GRAD/M are taken into account. Table 6.1: Coherence with Reference Data 213/Ø21 GRAD/L GRAD/M DIFFERENCE IN INDEX POINTS CC_SI - TOTAL 66, 65, 1, 1 BUILDINGS 58,7 42,6 16,1 2 CIVIL INGINEERING WORKS 72,9 8, -7,2 7 Costs and Burdens 7.1 Survey Costs of the Office The number of reporting units, which should complete the questionnaire, does not comprise the respondents, for whom it was revealed during the implementation of the survey that they were inadequate. The costs are expressed in working hours. To carry out the annual survey GRAD/L for 213, SURS spent 99 hours. Table 7.1: Survey costs at the Statistical Office Reference period 213 9/1
Number of working hours spent 99 Number of reporting units that had to fill in 695 questionnaires Survey period annual Number of questionnaires per year (total) 834 7.2 Burden of Reporting Units The burden of reporting when fulfilling the Annual Questionnaire of Construction put in place (GRAD/L) is largely different. It depends on the number and fragmentation of construction projects to be implemented in the reference year, on the organization of the company (dislocated branch units and establishments), on the number of subcontractors and coperformers as well as the fact that a unit was for the first time included in the survey, or if it has already got accustomed to the survey in the previous years. The average time spent by the company to complete the questionnaire GRAD/L was estimated on a survey filled in by selected small, medium-size and large enterprises that were covered by the survey. The average time for a reporting unit to collect and enter the data that are required with the questionnaire GRAD/L is.5 hours per year. Table 7.2: Burden of the reporting units Reference period 213 Number of reporting units that submitted the data 619 Annual number of questionnaires per unit 1 Time spent to fill in a questionnaire (hours).5 Total time spent (hours) 39.5 1/ 1