1(5) - Quality description: Job vacancy survey 1. Relevance of statistical information 2. Methodological description of the survey 3. Correctness and accuracy of data 4. Timeliness and promptness of published data 5. Accessibility and transparency/clarity of data 6. Comparability of statistics 7. Coherence and consistency/uniformity 1. Relevance of statistical information The Job Vacancy Survey provides information on the development and changes on the labour market from the perspective of the employer. For the statistics, establishments selected for the sample are asked about the number of job vacancies during a certain time period. The reference period is the first calendar day of the last month of each quarter (1 March, 1 June, 1 September, 1 December). As the reference period remains the same from one year to another, a situation report on job vacancies in every quarter is created based on the responses given by employers. A job vacancy can be a completely new job or an already existing job without a holder. By definition, the position must be open for application by those outside the establishment, and the employer must have been active in finding a suitable employee. Active measures are such as reporting a job vacancy to the labour exchange office, other advertisement channels, and use of contact networks. Job vacancies also include jobs opening in near future reported by employers as open. Job vacancies do not include jobs that are filled by an unpaid trainee or an outside subcontractor or contractor. Job vacancies are not jobs that are filled by an employee returning from paid or unpaid leave. Short-term replacements of less than one month in duration are not included in vacancies either. The data collection includes some 2,500 different establishments in every quarter. Based on the data collection, a picture is formed of job vacancies in the whole of Finland. Data on the number of personnel, form of ownership, the number of job vacancies and their qualitative features, that is, whether they are part-time, fixed-term and hard-to-fill, are collected from the establishments. The survey is also concerned with how many of the job vacancies are without a holder and how many of the jobs without a holder are meant to be filled later. The initiative for the survey came from the Employment, Social Affairs and Inclusion Department of the European Commission and from the European Central Bank. The basis for the survey was the need for reliable, regular, up-to-date and comparable data from member states on the number and structure of job openings. The statistics are based on Regulation (EC) No 453/2008 of the European Parliament and of the Council. 2. Methodological description of the survey The population of the survey comprises all private or public sector establishments located in Finland employing paid employees representing at least one staff-year. An establishment is a unit that is located in one place, and mainly produces one type of goods or services. The branches of industry that 09_12_01 2(5) - are represented do not include activities of households as employers, activities of extraterritorial organisations and bodies, and defence. The sampling frame used for the survey is the Register of Enterprises and Establishments maintained by Statistics Finland, which contains some 150,000 active establishments in which the number of wage and salary earners measured in staff-years is at least one. Since 2013 the survey has been based on the new sampling design. An annual sample is selected from the sampling frame with systematic random sampling by strata. The size category of the establishment based on its number of personnel (1 to 4, 5 to 9, 10 to 49, at least 50) and the industry group of the establishment (A, B-E, F, G-I, J, K-L, M-N, O-Q, R-S) are used as strata. In all, some 10,000 establishments are selected for the annual sample from the sampling frame so that they are divided into size categories with fixed proportions (2,000, 3,000, 3,000, 2,000) and they represent the division of the industry strata of the sample frame within their size category. The sampling frame has been grouped based on the location of the establishment so the annual sample is regionally self-weighting. The annual sample is divided into quarterly samples based on random numbers. One-half of the establishments in the quarterly sample are replaced by new ones every year. An individual establishment is included in the survey independent of the size category at most in two consecutive years and always in the same quarter. The data collection is conducted by the CATI centre that belongs to Statistics Finland's Data Collection Department. Since the first quarter of 2006, the respondents have also been able to answer the inquiry with a web questionnaire. Thirty-two per cent of all interviews were conducted through the web questionnaire in 2014. In 2014, the net response rate for data collections was 87 per cent. In each establishment, the person with best knowledge about recruitment there is sought to answer the survey. In establishments with a lower number of personnel, the respondent is usually the entrepreneur or the director. In the largest establishments, personnel administration is usually contacted. On average, it takes five to ten minutes to reply to the inquiry but dispersion is large due to the differences in industry and size category. In connection with the data collection for the Job Vacancy Survey, data for the Ministry of Employment and the Economy's survey on employment services are also collected. These data are used for developing the activity of labour administration and surveying recruitment problems. The collected data are weighted by stratum to correspond with the number of establishments in the frame. The estimates thus formed describe the number of job vacancies in the whole of Finland in each reference period. In addition to estimates that describe the situation in the entire country, results based on the size category, form of ownership, major region and industry are also compiled. 3. Correctness and accuracy of data Due to the nature of the sample survey, the estimates always include sampling errors caused by the sampling design. The size of this random variation is assessed with the help of standard errors of estimates. With the help of the standard error, a confidence interval is calculated for the estimate, which at certain probability includes the real value of the parameter to be estimated. The sample survey may also include a bias independent of the sample design, which is caused, for example, by non-response or over-coverage or under-coverage of the sampling frame. In 2014 the sampling frame contained less establishments than previously, which may in part be due to increased undercoverage. 09_12_01 3(5) - During the editing and validation of response data, the observation values are checked and, if necessary, corrected partially programmatically but also manually in order to generate the best quality results. This way, the mutual logicality and compatibility of the received responses are ensured. The effect of clearly deviating values on the results is limited by lowering the weight defined for the response in question. The results of the statistics are not seasonally adjusted. A considerable variation caused by the reference period can be seen in the number of job vacancies, because of which it is not possible to compare the figures of successive quarters to make conclusions about the trend of labour demand and possible changes in it. To describe the trend of job vacancies, one should compare only estimates of the corresponding quarters of successive years. Table 1. An example of the accuracy of quarterly estimates: Job vacancies 2014, 1st quarter. Length of the 95 Coefficient Standard per cent Estimate of variation error confidence (%) interval Total Job vacancies ± 16,100 46,000 8,200 17.8 Private enterprise ± 15,500 32,300 7,900 24.4 Form of ownership Municipality or joint municipal Number of personnel Major region (2012) Industry (TOL 2008) board Central government 5,600 1,000 700 200 Organisation, association or fund 2,300 1,000 Other 5,000 1,800 1 to 4 persons 14,900 6,700 5 to 9 persons 5,200 1,000 10 to 49 persons 10,700 1,600 50 persons or more 15,200 4,300 Helsinki-Uusimaa 11,900 2,800 Southern Finland 8,500 1,800 Western Finland 12,800 3,800 Northern and Eastern Finland 12,200 6,500 200 200 B-E 9,300 4,100 F 1,200 700 G-I 9,300 1,800 J 1,500 500 K, L 1,200 500 M, N 12,200 6,500 O-Q 6,600 900 R, S 4,400 1,700 A ± 1,900 ± 500 ± 1,900 ± 3,500 ± 13,200 ± 2,000 ± 3,100 ± 8,400 ± 5,500 ± 3,600 ± 7,400 ± 12,700 ± 300 ± 8,100 ± 1,400 ± 3,600 ± 1,000 ± 1,000 ± 12,700 ± 1,700 ± 3,400 4. Timeliness and promptness of published data The results of the Job Vacancy Survey are published quarterly some 45 days from the end of the statistical reference quarter. No preliminary data are compiled, as the released data are final. 09_12_01 17.5 32.5 41.5 36.1 45.2 19.6 15.0 28.3 23.7 21.8 29.6 53.2 83.7 44.4 56.0 19.5 32.0 41.7 53.2 13.5 39.3 4(5) - 5. Accessibility and transparency/clarity of data Data collected for statistical purposes must be kept confidential by virtue of Section 24 of the Act on the Openness of Government Activities (621/1999). The response data are only used for statistical purposes. The research data are protected in accordance with the data protection regulations of Statistics Finland and responses given by individual establishments cannot be distinguished from the statistical tables. According to Section 13 of the Statistics Act (280/2004), Statistics Finland may, on the basis of a separate application for licence to use statistical data, release data for scientific studies and statistical surveys without data enabling direct identification. The Statistics Act prohibits the use of data collected for statistical purposes in an investigation, surveillance, legal proceedings, administrative decisionmaking or other similar handling of a matter concerning the enterprise. 6. Comparability of statistics The Job Vacancy Survey has been compiled since the first quarter of 2002 but the statistical data were not public until 2003. In 2006, data suppliers were given the option to respond to the survey by web questionnaire, in addition to telephone interviews. Based on a methodological comparison made, the time series of the statistics are comparable despite the revision in data collection. From 2009 onwards, as recommended by Eurostat, an average annual estimate for job vacancies has been calculated as an average of quarterly estimates. Annual averages in the Statfin database comply with this calculation method. For the annual estimates of the Job Vacancy Survey prior to 2009, new inflating coefficients have been calculated based on the combined data. The difference to the average of the quarterly estimates has been one to five per cent. The sample design of the Job Vacancy Survey was changed starting from the statistical reference year 2013 so that the target samples and strata better correspond with the information needs of data users. The population of the survey remains the same regardless of the changes but the turnover of the sample objects has been sped up and, in addition to the size category and location of the establishment, also its industry, which is the key data required by data users is taken into account in the sampling. The implemented reforms have a considerable effect on the results and they cannot be compared to previously published data. Data based on the new design in the Job Vacancy Survey are published from the first quarter of 2013 onwards. The previously produced time series are no longer updated. 7. Coherence and consistency/uniformity In addition to Statistics Finland's sample survey, the Ministry of Employment and the Economy publishes register data on the number of job vacancies by occupational group and ELY area. Due to methodological, definition and production differences, the data published by the Ministry and Statistics Finland are not, however, comparable. The main difference between the statistics is that the Employment Service Statistics of the Ministry of Employment and the Economy only include those job vacancies that employers have reported to employment offices. Statistics Finland's data, in turn, are based on a sample survey and cover to some extent both job vacancies that have been reported to employment offices and vacancies that have not been reported. 09_12_01 5(5) - The Job Vacancy Survey does not include all job vacancies that have been open due to the structure of the Register of Enterprise and Establishments used as the basis of the sampling frame and owing to the data collection method. The Job Vacancy Survey does not take into consideration job vacancies of household employers nor of extraterritorial organisations or bodies. In addition, the criterion of one staff-year set for the number of wage and salary earners and leaving out short-term replacements from job vacancies at an establishment will lead to that part of the open jobs may be reported to employment offices, even though they are excluded from the Job Vacancy Survey. The Job Vacancy Survey is not trying to replace the Employment Service Statistics. These two different data sources should be viewed as mutually complementary indicators for labour demand. 09_12_01