Business Population Dynamics: The Business Structure Database (BSD)
The Business Structure Database (BSD) is a ‘snapshot’ in time of the Inter-Departmental Business Register (IDBR), which is a live register of firms registered for VAT and/or Pay As You Earn (PAYE) in the UK. The IDBR is estimated to cover approximately 99% of UK economic activity. The ‘snapshot’ of the IDBR used for the BSD is taken around April annually; the reporting period for the firm is generally the financial year although the IDBR data are complimented with data from ONS business surveys and hence reporting periods can vary by firm depending on whether the record has been updated by survey data. The BSD snapshots contain approximately 2 million observations annually. The BSD is divided into two datasets, one covering ‘enterprises’ and the other ‘local units’. An enterprise is the overall business organisation. A local unit is a ‘plant’, such as a factory, shop, branch, etc. In some cases, an enterprise will only have one local unit, and in other cases (such as a bank or supermarket), an enterprise will own many local units. For each company on the BSD dataset, data are available on employment, turnover, foreign ownership, and industrial activity based on Standard Industrial Classification (SIC)92, SIC 2003 or SIC 2007. Year of ‘birth’ (company start-up date) and ‘death’ (termination date) are also included, as well as postcodes for both enterprises and their local units.
1 The UKDS secure data environment is strictly controlled by the UKDS with appropriate login protocols from an accepted HEI (Higher Education Institution) IP address (verified by the individual institution) for an ONS approved academic researcher given specific access to agreed ONS datasets and other imported datasets approved by their owners. All analysis undertaken by the ERC for public dissemination in the UKDS is checked by UKDS statistical staff to ensure the identity of an individual respondent could not be revealed.It is, indeed, a criminal offence to breach these rules.
A longitudinal version of the BSD can be created by linking together the annual snapshots, which are available from 2007 onwards. As a snapshot of the IDBR the BSD contains unique reference numbers .for each firm which can be used to link the data to other ONS business survey datasets that are sampled from the IDBR. Linking to other commercial datasets such as FAME is possible and ERC has undertaken some research projects linking these two datasets. The ERC have put together the annual snapshots of the BSD to produce a longitudinal dataset covering the 1997-2014 period. The dataset is restricted to the private sector and includes employer enterprises only, that is, firms with at least one employee. Birth and death variables have also been re-created, with birth regarded as the first year in which a firm records an employee. It is this dataset which is used to develop metrics relating to the size and structure of the SME population in each LEP area and Local Authority areas outside England.