Business Register and Employment Survey (BRES)

Annual employment estimates for 2009 from the new Business Register Employment Survey were released at 09:30 on Wednesday 08 Dec 2010. BRES is a new ONS business survey which replace the Annual Business Inquiry (ABI). Access to the BRES dataset requires special authorisation - see the bottom of this article for how to apply.

Users with permission can access the BRES dataset using the advanced/wizard query facility:

• Business Register and Employment Survey   [link to data]

BRES Data on Nomis

Key features:

• Date: New estimates are available for 2009. Employee estimates for 2008 on the new basis, with the discontinuity removed, have also been produced to allow users to identify the movement between 2008 and 2009. Further details on the discontinuity between BRES and the ABI are available in the discontinuity section below.
  
  
• Geography: Estimates are available down to lower layer super output areas / Scottish data zones. Some geographies (including 2001 TTWAs, NUTS areas and 2009 wards) are not available in the initial release on 8th Dec 2010 but will be added shortly afterwards.
  update 8-Feb-2011: TTWAs and 2009 statistical wards are now available. NUTS areas and middle layer SOAs will be added shortly.
  
  
• Gender: no male/female split is available in BRES. Male and female employee splits are not collected in the survey because some businesses find it difficult to provide this information and the quality of the data can be poor.
  
  
• Employment status: Figures are available for full-time, part-time, total employees. Employment figures are also available which are employees plus working owners. Employment figures were not previously available in the ABI dataset.
  
  
• Industry: Annual employment figures for BRES are only available on the 2007 Standard Industrial Classification (SIC 2007) basis. Figures are available down to the 5-digit subclass level - ABI data was only available at the 4-digit class level.
  
  Broad Industry Groups are not available with the initial release of the dataset on 8th Dec 2010. They will be added as soon as possible.
  
  Farm agriculture (subclass 01000) is available at region and country level only. In the ABI farm agriculture was also available at the lower tier local authority level. The BRES 2009 estimates also include a change to the farm agriculture estimates - see the Discontinuity section below for details.
  
  
• Data units / workplace analysis: The workplace analysis is not available under BRES. Users who require information on the count of businesses should refer to the UK Business release on the National Statistics website.
  
  
• Confidentiality rules: Standard ONS primary disclosure rules are applied to BRES data output from Nomis. Disclosive figures are flagged with an exclamation mark (!) and must be removed before passing onto anyone not named on your notice. You must also manually apply secondary suppression as detailed in the BRES User Guide (pdf).
  
  
• Rounding: Figures are shown unrounded in Nomis outputs. Before passing BRES estimates onto anyone not named on your notice or using it in in publications all figures must be rounded to the nearest 100 (after the confidentiality rules described above have been applied). Therefore, figures of less than 50 should not appear.
  
  
• Data quality: BRES is based on a sample survey so estimates are subject to sampling errors which need to be taken into account when interpreting the data. See the Standard Errors section below.

Discontinuity Affecting the Annual Employment Estimates

The Business Register and Employment Survey (BRES) is a new Office for National Statistics (ONS) survey, the aim of which is to maintain the Inter-Departmental Business Register (IDBR) and provide the basis for annual estimates of employment.

As the new basis for producing annual employment estimates, its introduction results in a discontinuity when comparing the estimates produced by BRES to those of its predecessor, the Annual Business Inquiry part 1 (ABI/1) in a common year, 2008. This will also impact on the Workforce Jobs (WFJ) estimates for employees as these figures are benchmarked against the ABI for certain sectors of the economy.

The overall discontinuity was estimated at 317,000 employees in an upward direction. That is, BRES yields an estimate of employment in 2008 that is 317,000 higher than obtained from ABI/1 for the same year. The discontinuity for WFJ benchmarking is estimated to be 315,000. This differs to the overall discontinuity because WFJ is not benchmarked to ABI for all industries.

The BRES 2009 estimates also include a change to the farm agriculture estimates which are collected by the Department for Environment, Food and Rural Affairs (DEFRA). The figures now only include commercial holdings as opposed to ‘all’ holdings previously. This accounts for a change of 6,500 employees in a downward direction. This change is not included within the analysis of the move from ABI/1 to BRES.

Scaling factors can be calculated by using the estimated BRES and ABI/1 2008 estimates to produce a time series on a consistent basis. Further work will be conducted by the ONS to provide additional guidance for users during 2011.

Standard Errors

BRES is a business survey which collects employment information. Users should be aware that the data presented are estimates, subject to both sampling errors (arising from the fact that the BRES is a survey, not a census) and non-sampling errors, for details see the article:

Estimates are subject to standard errors. The lower the level of geography and industry the larger the coefficient of variation. Also there is a modelling error, due to the minimum domain methodology which is used to produce low level estimates. This error increases as the geographical level becomes more detailed, but it is difficult to measure.

The coefficient of variation (CV) is the ratio of the standard error of an estimate to the estimate, expressed in terms of a percentage. The smaller the CV, the higher the quality of the estimate. The CVs at a government office region / country level for the 2009 total employee estimates are shown in the table below:

  Coefficient of variation - total employees

Standard Errors on an industry basis and at a local authority level are available on the National Statistics website.

The standard error and coefficient of variation exclude farm agriculture data.

Caution should be applied when comparing estimates at or below the local authority level especially when further disaggregating by industry. In this circumstance the coefficient of variation is likely to increase, meaning the 95 per cent confidence interval becomes wider. The 95% confidence interval gives the boundaries between which the true value lies with a 95 per cent confidence level.

For example, the total number of employees for Great Britain is 26,206,000. The CV for this estimate is 0.2, with the standard error being 59,500. That means that we are 95 per cent confident that the true value is between 26,087,000 and 26,325,000.

Managed Service Companies

Managed Service Companies (MSCs) are intermediary companies through which the services of a worker are provided to an end client. The worker in a MSC is almost invariably not in business on his own account and is not exercising control over the business. This control lies with the scheme provider. MSC scheme providers are businesses (usually companies) that provide generic company structures and administer the schemes. MSC providers emphasise that the worker is not involved in running the company and will receive payments in a similar way to if they were an employee.

The workers are all registered at the MSC scheme provider's address and are given the same industrial classification, which skews the industrial and geographical breakdown. Therefore the decision has been made to remove all MSCs from ONS business surveys. The Workforce Jobs review provides further information.

For the BRES all identified MSCs have been removed from the 2008 and 2009 estimates. However, there are some MSCs included in the results for both years that were subsequently identified at a later time. The table below estimates the amount of employment attributable to MSCs.

*The estimate of employment of MSCs included within the BRES results is an estimate of those businesses that have since been identified as a MSC but were not identified at the time the BRES was defined for the specific year. It is difficult to give an exact value of employment attributable to MSCs within the BRES due to identifying those that have since ceased trading and those that started trading after the BRES was defined for the year in question.

How to Get Access to BRES data

Free to view figures at local authority level and above are available from the BRES homepage of the National Statistics website. The free to view data are available at the following aggregations on the National Statistics website:

• GB/UK: Broad Industrial Grouping (BIG)/2-digit SIC/3-digit SIC/5-digit SIC with a public/private split (UK not available at the 5-digit SIC level)
• Government Office Region: total with a public/private & urban/rural split
• Government Office Region: 2-digit/3-digit with a public/private split
• Local Authority County: total with a public/private & urban/rural split
• Local Authority County: Broad Industry Group
• Local Authority District: total with a public/private & urban/rural split

Figures presented within the tables on the National Statistics website are subject to disclosure controls - both primary and secondary disclosure. Secondary disclosure suppresses figures to ensure figures that are primary suppressed cannot be derived by deduction.

Note that public/private and urban/rural splits are currently not available through Nomis.

The full BRES and earlier ABI datasets on Nomis gives access at all geographic and industry levels but access is restricted. You must first obtain a Chancellor of the Exchequer's Notice from ONS. Registered users can check whether they already have access by going to the my account section of the web site and selecting the BRES Notice option (BRES data is not available with guest logins). The cost for a new Notice is £125 + vat. To apply please complete the online application at:

Apply for a New Chancellor's Notice

The access will also allow users to view back data from the Annual Business Inquiry, although caution should be applied if comparing these to BRES (please see the discontinuity section above for further information).

For further information please contact the BRES helpline on 01633 456903 or email: annual.employment.figures@ons.gov.uk