Statistical concept

Unemployed persons are all persons 15 to 74 years of age who were not employed during the reference week, had actively sought work during the past four weeks and were ready to begin working immediately or within two weeks. Figures show the actual number of persons unemployed in thousands.

Employed persons are all persons who worked at least one hour for pay or profit during the reference week or were temporarily absent from such work. This variable is needed for the calculation of unemployment rates (see definition below).

For more details see at the bottom of the page Labour market: main indicators (lfsi_base) and Long-term unemployment (tsisc070_sm1).

Definition of indicators

This domain comprises collections of monthly, quarterly and annual averages of unemployed persons and unemployment rates. The relevant definitions are as follows:

The unemployment rate is the number of people unemployed as a percentage of the labour force. The labour force is the total number of people employed and unemployed.

The long-term unemployment rate (Structural indicator) is the share of unemployed persons since 12 months or more in the total number of active persons in the labour market. Active persons comprise both the people who are employed and those who are unemployed.

Classification system and conformity with official standards

The definition of unemployment conforms to the recommendations of the International Labour Organisation (ILO).

Geographical coverage

Harmonised unemployment series are compiled for all European Union member states, Candidate Countries and Norway. European Union aggregates are available for EU15, EU25 and EU27 and Euro Area aggregates are available for EA12, EA13, EA15 and EA. Data for USA and Japan are also disseminated.

Statistical units

Individuals living in private households

Statistical population

It consists of the total population usually residing in Member States, except for persons living in collective or institutional households.

More details can be found in the European Union Labour Force Survey "lfsi_sm1"(see at the bottom of the page).

Reference period

Data on unemployed persons and unemployment rates refer to monthly (Euro-Indicator), quarterly and annual averages (Structural Indicator).

Base period

Not applicable

Recording of transactions

Number of persons and rates

Data sources used

Compilation of European aggregates

EU and Euro area employed and unemployed persons are calculated on summing country data (see below for adjustments on country data). Rates are subsequently calculated from the data expressed in absolute values (i.e. number of persons).

Seasonally-adjusted European aggregates use indirect seasonally-adjustment method: seasonally-adjusted Member States data are summed and the rates calculated subsequently.

Adjustments and weights

This heading refers to calculations of country data. European aggregates are derived as explained above.

The EU LFS is a quarterly survey. The following method is used in order to produce a monthly indicator: for all countries, the non-seasonally adjusted quarterly averages of the monthly series are benchmarked to the quarterly LFS figures. However, the way the figures for the individual months as well as the provisional figures (for the period when LFS data are not yet available) are calculated depends on the availability and specific characteristics of the sources available in individual Member States. Eurostat aims at harmonizing the calculation process as much as possible. Apart from quarterly figures, in some Member States monthly and/or 3 month moving averages are produced from the LFS as well. Registered unemployment data are used for many Member States as auxiliary source. The length of the series and specific correlation with the unemployment figures as measured with the quarterly LFS varies from country to country.

Germany, Finland and Sweden

Monthly estimates are available directly from the LFS. For Germany and Sweden, the time series are too short for seasonal adjustment. Additional information from registered unemployment is used for this purpose.

The Netherlands, United Kingdom and Norway

Three month moving averages are available directly from the LFS, and are published as monthly figures under the middle month. For the Netherlands, the figure for the most recent month is a provisional estimate, based on a State Space Model combining the three month moving averages with underlying monthly figures.

Portugal

Quarterly LFS data are combined with monthly registered unemployment data by using a temporal disaggregation and forecasting Chow Lin model.

Belgium, Bulgaria, the Czech Republic, Denmark, Estonia, Ireland, Spain, France, Latvia, Lithuania, Hungary, Austria, Poland, Slovenia and Slovakia

Quarterly LFS data are combined with monthly registered unemployment data by using a temporal disaggregation Denton model. For the most recent months (for which the LFS data are not yet available), the monthly benchmark factors are forecasted using seasonal ARIMA regression models. The provisional estimates are calculated by multiplying these factors by the available registered unemployment figures. See the Annex for technical details.

Cyprus, Luxembourg and Malta

Benchmarking to moving annual averages of LFS data and linear extrapolation of registered unemployment data

Croatia

Benchmarking to semi-annual LFS data and linear extrapolation of registered unemployment data

Greece, Italy, Romania and Turkey

For the moment only quarterly LFS figures are published for individual months of the quarter.

Validation of statistical data

Quarterly and monthly figures received from all EU countries are processed and validated before updating Eurostat’s reference database.  

Revision policy

Every month new figures from the public employment offices' administrative registers or from the national LFS are added into the process and new estimates are calculated. This might cause a slight revision in the past figures due to the re-execution of the seasonal adjustment procedure. Whenever new LFS data become available, a potentially larger revision takes place from the months of that particular quarter onwards. Seasonal adjustment parameters are reviewed annually. 

Seasonal adjustment

Basic data are expressed in number of persons and rates and presented as:

- Non-seasonally adjusted (NSA) data

- Seasonally adjusted (SA) data

- Trend data         

Comparability

Often questions are asked concerning the comparability between the monthly unemployment rates (or their quarterly and annual averages) and the direct results of the LFS. Some of the differences are due to the different nature of the two data sets, but some of the differences occur just because the transition period preceding the use of the most recent quarterly data is not yet completed:

·         The headline figures published in the monthly News Release are seasonally adjusted unemployment rates, which are different from the LFS non-seasonally adjusted data. The non-seasonally adjusted monthly data are consistent with the published LFS data;

·         In the monthly application, the idea is to keep the time series as comparable in time as possible. It means that possible breaks in the LFS series due to changes in the definitions or in the filtering of the micro data have been adjusted: in 1991/1992 there was general definition precision; the gradual implementation of the ‘new’ unemployment definition following the Regulation (EC) 1897/2000 still leads to backwards revisions while also a general improvement in the micro data filtering of the LFS data from 2001 onwards caused breaks and backwards adjustments. While the original LFS data consists of the raw series as they are recorded at each point of time, the same series are adjusted when they are used as benchmarks for the monthly harmonized time series;

·         Where moving averages of the LFS are used either as a temporary or definitive solution, the monthly data do not match the corresponding quarterly LFS data;

·         For France the two data sets are not consistent. The LFS for the monthly calculations are complemented with the DOM data (départements d’outre-mer) while the original LFS data cover only “France Métropolitaine”. The second exception is Germany. While the unemployment is measured with the LFS, the employment used in the denominator of the rate is derived from the so-called Employment Accounts;

·         In the cases where only the LFS data of the Spring quarters are used, quarterly and annual averages obtained from the monthly unemployment data differ from the corresponding LFS data. This situation gradually improves when the complete results of the LFS become available and are applied. 

• Last update: Date of most recent update of these metadata.
• Last certification: Date of the most recent certification by the author on the accuracy of these metadata.