Labour market and earnings;Labour market and earnings
ledstill, Job vacancies, job offer, industry breakdown (e.g. manufacturing, teaching, health services…).Unemployment , Employment , Labour market and earnings

Job vacancies


Next update

About the statistics

The statistics show the number of job vacancies on a reference date each quarter, the number of job vacancies as a percentage of all positions. The statistics are given for the country as a whole, and are divided into 20 industry groups.


Definitions of the main concepts and variables

Job vacancy

The most important aspect regarding the definition of a job vacancy is that the vacancy must be announced, formally or informally, making it possible for external candidates to apply for it. Formal announcements include for example advertisements in newspapers, on the internet, notifying the public employment services or private employment agencies. Informal announcements include for example advertisements in shop windows or public notice boards, or announcements via employees, friends or family. Job vacancies exclusively open for internal applicants (within the enterprise, organisation or corporation) are not included in this definition of job vacancies. The post must be paid.

We do not divide between part-time and full-time job vacancies, and consequently a vacant part-time post is counted as one job vacancy. Moreover, we do not divide between newly created job vacancies ( non-occupied vacant posts) and posts that are announced as vacant to replace an employee who is going on a leave of absence or quitting ( occupied vacant posts).

The aspect of time is also important to identify a job vacancy. Given the definition above, a post is vacant from the point of time where the information that it is in fact vacant is made accessible for external candidates. For this reason, a post is not vacant if there merely has been made a decision internally to hire a new person without announcing it externally in one way or other. Furthermore, the post is vacant until the point of time where the post has been assigned to a candidate (defined precisely as the point of time where a contract of employment has been signed) or until the point of time where one, for one reason or another, has decided not to employ a new person after all.

Job vacancy rate

The job vacancy rate is the number of job vacancies in per cent of the total number of posts, which consists of occupied posts and job vacancies. We consider occupied posts to be synonymous with employee relationships.

There is a break in the job vacancy rate between the 4th quarter of 2014 and the 1st quarter of 2015. The break is explained below.

Ledige stillinger-en

Employee relationships

Untill 2015 we used to collect information of employees from the aa- register in The Norwegian Labour and Welfare Administration (NLWA). All employers had to register their employees if they were to work at least seven days and the average working time pr week was at least four hours. From the 1st quarter of 2015 we get these information from the "a-meldinga". The number of employees were larger by this change to a large extend because there are no limits in the working time. Therefore, there is a break in the number of employees between the 4th quarter of 2014 and the 1st quarter of 2015. These figures are not published in these statistics, but are included in the denominator in the joc vacancy rate. Therefore, there is a break in the job vacancy rate between the 4th quarter of 2014 and the 1st quarter of 2015.

Reference period

The reference period in regard to all the variables are set to be the end of the middle month of the quarter. We assume that this reference period is representative for the rest of the quarter. We start our data collection for each quarter at this point of time, and with regard to the variable of job vacancies we refer to this period of time in our question. The variables number of employee relationships and industry are gathered from adapted numbers from "a-ordninga" and from the CRE from the same reference period.

Standard classifications

Industry is coded according to the Standard Industrial Classification, SIC2007 .

Administrative information

Name and topic

Name: Job vacancies
Topic: Labour market and earnings

Next release

Responsible division

Division for Labour Market and Wage Statistics

Regional level

Figures are given on a national level.

Frequency and timeliness

This is a quarterly statistics. The figures are released 6-9 weeks after the end of each quarter. Release dates for the next four months are given in the advance release calendar.

International reporting

To Eurostat in accordance with the Regulation (EC) No 453/2008 of the European Parliament and of the Council of the 23rd of April 2008.


Microdata are saved permanently.


Background and purpose

The purpose of the statistics on job vacancies is to give information about the level and the structure of labour demand both on a national level and broken down by industry. This provides knowledge of the development of the labour market and of the economy in general. The statistics are to be comparable with similar statistics from the other European countries affected by the previously mentioned EU Regulation.

There is a need for data on job vacancies broken down by industry to render possible monitoring and analyzing structural imbalances in the labour market.

The survey begun in the 1 st quarter of 2010, and the first release of data was in 2011.

Users and applications

Important users include public authorities, scientists, media, labour and employers’ organisations, enterprises and international organisations.

Equal treatment of users

All external users are given admission to the statistics simultanously at 08.00 on ssb.no after having been announced at least three months in advance in the calendar of the statistics: Statistikkkalenderen.  These are one of the most important principles in SSB to secure that all users are treated equally:lik behandling av brukarane.

Coherence with other statistics

The Norwegian Labour and Welfare Administration (NLWA) produces statistics on job vacancies on a monthly basis, but with another scope and a different definition than the ones stipulated in the EC Regulation mentioned above. The job vacancy statistics of the NLWA includes job vacancies that are reported to the NLWA or announced in the media (newspapers, magazines, etc.). The statistics of the NLWA does not include informal announcements or job vacancies exclusively announced by other means mentioned above (for example announcements on home pages on the internet). As these methods of announcements are pretty common, especially in certain industries, Statistics Norway gets a higher estimate on the number of job vacancies than the NLWA gets. This is mainly due to diverging definitions.

It is possible to compare this statistics with the monthly statistics of the NLWA, keeping in mind that the two definitions are different. A comparison with the figures of the NLWA will show that the figures of Statistics Norway are significantly higher than the figures of the NLWA. In the period from the 1st quarter of 2010 to the 4th quarter of 2014 , the figures of the NLWA amounted to 30 per cent in average of the figures from the job vacancy survey conducted by Statistics Norway.

Legal authority

The Official Statistics and Statistics Norway Act § 10, cf. The employer's reporting of employment and income conditions Act, etc. (the a-opplysnings Act) § 3. 

The Official Statistics and Statistics Norway Act § 20 (compulsory fines)

EEA reference

Regulation (EC) No 453/2008 of the European Parliament and of the Council of 23 April 2008 on quarterly statistics on Community job vacancies.



The statistics cover establishments (with employees) in all industries except industry classifications 97 (Activities of household as employers) and 99 (Activities of extraterritorial organisations and bodies).

Data sources and sampling

The statistics on job vacancies are based on data reported by a sample of establishments.

A random sample of 8000 establishments is drawn from the Central Register of Establishments and Enterprises (CRE). The establishments are divided into groups/strata by industry and size. All the establishments in the population with at least 5 employees are divided into 4 groups depending on the number of employees. The division after industry is done based on the Standard Industrial Classification (SIC2007). This ensures good representation in all industries and among big and small establishments. In each of the 4 size group there are 35 industry groups, so the total number of groups/strata is 140. Within each group/stratum, all the establishments have the same probability of being drawn to the sample.

The probability of being drawn to the sample varies from size stratum to size stratum. The establishments with less than 5 employees are not drawn, while all the big establishments are drawn (more than about 100 employees). In a few industries, all the establishments are drawn. In this way, almost 2 per cent of the establishments have to answer the survey, but they represent more than 25 per cent of the total number of employees in the population.

The biggest establishments take part in the survey all the time, while the others participate for two years at a time. Therefore, we replace almost half the sample each year.

The drawing of the sample is coordinated with several other surveys in Statistics Norway in order to disperse the burden placed on establishments.

Collection of data, editing and estimations

Statistics Norway sends each quarter a letter to inform the establishments that the questionnaire will be ready in Altinn, the offical web-based reporting system. This letter  is sent out in the first week after the middle month of each quarter. It is sent directly to the establishments. The letter about the obligation to participate in the survey is sent earlier to the enterprises. The deadline is appoximately 3 weeks.

Approximately 99 per cent of the establishments which answer the survey do so via the official web-based Altinn reporting system. There, the respondent gets a warning if the reported number of job vacancies is very high or very low in relation to the number of employees. The respondent then can choose to change the reported figure or to ignore the warning and report the figure. In this way some of the typing errors are avoided.

All the replies from the questionnaire are loaded in to a system of editing where the control is activated if the number of reported job vacancies in relation to the number of employees is very low or very high. We e-mail or telephone the relevant establishments to confirm the reported figure. Some errors due to misunderstandings with regard to definition of the establishment (e.g. separating the establishment from the rest of the enterprise) are avoided this way.

In the estimation procedures, there are routines to even further investigate establishments with extreme values. The result of these investigations might be to keep these data out of the estimation procedure.

The estimation of job vacancies is based on the idea that each establishment in the sample is representing several establishments, which means that we can calculate the number of job vacancies for the entire population of establishments. In the estimation we use register information on the number of employee relationships and industry for each establishment in the entire population. The number of employee relationships is gathered from CRE, which for its turn has the relevant information from the "a-melding ". All employers have to report electronically each month to the  "a-melding" all salaries, pensions and other renumerations in the firm. The " a-melding" goes to the Norwegian Tax Administration, the Norwegian Labour and Welfare Administration (NLWA) and SSB. The information on the number of employee relationships is therefore based on the enterprise’s own reporting to the "a-melding".

To inflate the sample data on job vacancies to population level, that is to represent all establishments in the relevant industries in Norway, we use a model-based ratio estimator. In the stratified ratio model we assume that the variable number of employee relationships in the establishment (x) contributes to explain the variable number of job vacancies at the establishment (y) and that the correlation between the two variables is approximately linear within each stratum. Based on the survey data (number of employee relationships and number of job vacancies), we estimate the correlation, represented by the gradient β– between the number of employee relationships and the number of job vacancies. The correlation β is estimated by a weighted least square method in each industry and size stratum. Within each stratum, the total number of job vacancies (y) is estimated by the number of employee relationships (x) in that stratum multiplied by the gradient  β for that particular stratum, plus an individual residual ε :

y =  βx + ε;

By stratum we here mean one of the 35 industry groups within one of the 4 size groups, making a total number of 140 groups.

Since the smallest establishments are not participating in the survey, we make an assumption that the correlation between the number of employee relationships and the number of job vacancies for establishments with 1-4 employees on average is identical with the one for establishments with 5-9 employees.

Seasonal adjustment

Seasonally adjusted figures are calculated by use of the X12-ARIMA method. The number of job vacancies and the number of occupied posts are seasonally adjusted by 20 industry groups. This gives a total of 40 series that are adjusted directly. The figures for occupied posts are not published, but they are parts of the denominator in the job vacancy rate. We get the seasonally adjusted figures for the two totals by aggregating the adjusted component series. Seasonally adjusted figures are also given by 10 industry groups. Five of these groups are adjusted for seasonally variations directly and are also parts of the 20 industry groups, while the other five are indirectly adjusted, i.e. the seasonally adjusted figures for the remaining 15 industry groups are aggregated to five groups. 

X12-ARIMA decides between additive and multiplicative decomposition of the series, and it chooses the best ARIMA model, and seasonal and trend filter. These optimal choices are normally locked for one year. The seasonal components and the parameters in the pre-adjustment regressions are calculated every quarter.

X12-ARIMA pre-adjust the time series by use of regression analysis. Regression variables are pre-specified to handle extreme values, level shifts and changes in the seasonal components. The trend-cycle and the seasonally adjusted figures for occupied posts are also adjusted for the transition to the A-ordningen.


Figures that identify groups of less than three establishments are not released.

Comparability over time and space

Movable holidays will not affect the level of job vacancies due to the chosen reference period (for example, Easter Day can only occur between 22 nd of March and 25 th of April, then Easter will never occur in any of our reference periods).

We compare every quarter with the corresponding quarter(s) the previous year(s). In some industries, it is commonplace to announce job vacancies at certain times of the year. For example, many schools announce job vacancies in the period before the end of the school year. Consequently, the estimated number of job vacancies in education is normally higher in the 2 nd quarter than in the other quarters. Similar seasonal variations may occur in other industries, and it is therefore most reasonable to compare the results of each quarter with the results from the corresponding quarter(s).

As mentioned above the statistics is to be comparable with similar statistics in the other countries in the European Economic Area (EEA). These countries are obliged to produce job vacancy statistics in accordance with EC Regulation.

Changing the register from which we collected the number of employees in 2015 brought quality improvements, resulting in an increase in the collected number of employees. This is explained in more detail above. The number of employees is not published in these statistics, but the figures are included in the denominator in the joc vacancy rate. Therefore, there is a break in the job vacancy rate between the 4th quarter of 2014 and the 1st quarter of 2015.

Accuracy and reliability

Sources of error and uncertainty

Errors may occur in all parts of the data collection. Respondents can do mistakes while reporting. Data gathered manually can be registered incorrectly. Errors in the editing process can also occur. The controls mentioned above helps finding and correcting such errors.

Incorrect interpretations of the term job vacancy can lead to incorrect answers by the respondent. We try to avoid such errors by including a guidance text in the questionnaire, by having a text of instructions to be used by the staff in the response-service of Statistics Norway and by having professional staff available for questions regarding the definition of job vacancies.

Non-response can contribute to skewness in the estimates even though the method of estimation is relatively robust towards distortions in the structure of industry and changes with regard to the size of establishments. As the survey on job vacancies is compulsory, the non-response rate is fairly small. Some natural withdrawal from the sample is unavoidable, for example as consequence of establishments being shut down. During 2014, approximately 500 establishments disappeared from the sample, principally because of natural withdrawal. The gross response rate from 2011-2014 varied between 91 and 96 per cent, while the net response rate varied between 95 and 99 per cent.

All sample surveys are bound to have a certain level of sample uncertainty. The uncertainty of the estimate is measured by an estimated standard of deviation. If the standard of deviation were known, one could find an interval which, with a fixed probability, contained the true value (the value one would find if the total population was surveyed instead of just a sample). This interval is called the confidence interval. If we call the estimated value M and the standard of deviation s, the interval with the borders M ±2will, with a probability of 95 per cent, contain the true value of the estimate. In the fourth quarter of 2014, the estimated total number of job vacancies was 52 700 and the estimated standard of deviation was approximately 2 200. This gave a confidence interval of about 48 200 to 57 100 on a 95 per cent level.

The ratio model is not unbiased, but the bias is low since the sample is big and the non-response rate relatively small.

Since the smallest establishments are excluded from the survey, the uncertainty is higher for the estimates with regard to this group.


At the dissemination of the figures for the 1st quarter of 2017, the figures for the period 2014-2016 was revised. From 2014, there were made changes in some of the input data for the statistics on job vacancies. TThere were made two situational outtakes from the Central Register of Establishments and Enterprises (CRE) each month before 2014, and only one from January 2014 and onwards. Before 2014, we used the number of employees from the second outtake from the CRE, which referred to the middle month of the quarter. After the 2014 change, we should have started using the next situational outtake to continue the practice of using the information on employees referring to the middle month of the quarter. This was not done at the time, but we corrected this in 2017 and disseminated revised figures for the period 2014-2016. In addition, we started (from 2015) to use the figures on employees directly from the same source as the CRE – thereby using the number of employees from the correct month while all the other information from the CRE are collected with reference to the same period as before.

About seasonal adjustment

General information on seasonal adjustment

Monthly and quarterly time series are often characterised by considerable seasonal variations, which might complicate their interpretation. Such time series are therefore subjected to a process of seasonal adjustment in order to remove the effects of these seasonal fluctuations. Once data have been adjusted for seasonal effects by X-12-ARIMA or some other seasonal adjustment tool, a clearer Picture of the time series emerges.

For more information on seasonal adjustment, see http://www.ssb.no/a/english/metadata/methods/seasonal_adjustment.pdf 

Why seasonally adjust these statistics?

Within lots of industries the number of job vacancies and of employees follow patterns that are repeated each year. This is among other Things due to seasonal workers and announcements of these jobs, and various Activity throughout the year. The seasonal variations complicate a direct comparison from one quarter to the next. To adjust for this the time series are seasonally adjusted.

Series that are seasonally adjusted

The number of job vancancies and the number of employees are seasonally adjusted by 20 industries. The 20 seasonally adjusted series are aggregated to get the two seasonally adjusted totals. The figures for employees are not published, but they are included in the job vacancy rate. This means that the job vacancy rate is indirectly seasonally adjusted. In addition, seasonally adjusted figures are published by 10 industries. These are aggregated from the 20 seasonally adjusted series.  


Pre-treatment routines/schemes

Running a detailed pre-treatment. This means using models which are specially adapted for the pre-treatment of the raw data for a given series.

Calendar adjustment

No calendar adjustment of any kind is performed.

Methods for trading/working day adjustment

No correction.

Correction for moving holidays

No correction.

National and EU/euro area calendars

Definition of series not requiring calendar adjustment.

Treatment of outliers

The series are checked for outliers of different types. Once identified, outliers are explained/modelled using all available information. Outliers for which a Clear interpretation exists (strikes, consequences of (government) policy changes etc.) are included as regressors in the model.

Model selection

Pre-treatment requires choosing an ARIMA model, as well as deciding whether the data should be log-transformed or not.

Automatic model selection by established routines in the seasonal adjustment tool.

Decomposition scheme

The decomposition scheme specifies how the various components - basically trend-cycle, seasonal and irregular - combine to form the original series. The most frequently used decomposition schemes are the multiplicative, additive or log additive.

Automatic decomposition scheme selection.

Seasonal adjustment

Choice of seasonal adjustment approach


Consistency between raw and seasonally adjusted data

Do not apply any constraint.

Consistency between aggregate/definition of seasonally adjusted data

In some series, consistency between seasonally adjusted totals and the original series is imposed. For some series there is also a special relationship between the different series, e.g. GDP which Equals Production minus Intermediate consumption.

Do not apply any constraint.

Direct versus indirect approach

Direct seasonal adjustment is performed if all time series, including aggregates, are seasonally adjusted on an individual basis. Indirect seasonal adjustment is performed if the seasonally adjusted estimate for a time series is derived by combining the estimates for two or more directly adjusted series.

Indirect approach where the seasonal adjustment of Components occurs using the same approach and software, and then totals are derived by aggregation of the seasonally adjusted components.

Horizon for estimating the model and the correction factors

When performing seasonal adjustment of a time series, it is possible to choose the period to be used in estimating the model and the correction factors. Correction factors are the factors used in the pre-treatment and seasonal adjustment of the series.

The whole time series is used to estimate the model and the correction factors.   

Audit procedures

Seasonally adjusted data may change due to a revision of the unadjusted (raw) data or the addition of New data. Such changes are called revisions, and there are several ways to deal With the problem of revisions when Publishing the seasonally adjusted statistics. 

Seasonally adjusted data are revised in accordance With well-defined and publicly available revision policy and release calendar.

Concurrent versus current adjustment

Partial concurrent adjustment: The model and the seasonal filters are identified yearly, while all coefficients and seasonal factors are estimated continously as new or revised data become available.

Horizon for published revisions

The period of revisions is defined according to the characteristic features of the series based on information from the seasonal adjustment tool.

Quality of seasonal adjustment

Evaluation of seasonally adjusted data

A detailed set of graphical, non-Parametric and Parametric criteria defined to assess the relevant characteristics of seasonally adjusted data is used.

Quality measures for seasonal adjustment

The full set of Diagnostics and graphical facilities to assess the whole process of seasonal adjustment is used. (Only relevant for some/a few series.) Table of quality measurement for this statistics 

For more information on the quality indicators in the table see: metadata on methods: seasonal adjustment

Special cases

Seasonal adjustment of short time series

All series are sufficiently long to perform an optimal seasonal adjustment.

Treatment of problematic series

Problematic series are treated in a special way only when thay are relevant. The remaining series are treated according to normal procedures.

Posting procedures

Data availability

Raw and seasonally adjusted data are available.

Press releases

For each series, some quality measures of the seasonal adjustment are presented.