This document has been written to satisfy the Office for Statistics Regulation (OSR) request that the Office for National Statistics (ONS) makes some short-term gains towards the overall requirements contained in the OSR report. There are four areas where the OSR requested that the ONS show immediate progress:
requirement 1: begin to engage with users to understand in detail their limitations and restrictions around data access to, and linking of, Inter-Departmental Business Register (IDBR) data
requirement 5: publish its plans for publishing business demography statistics, taking into account the need to make its annual statistics more timely, and publish its plans for developing the recently introduced quarterly experimental statistics
requirement 7: publish at least a narrative account covering what the ONS already knows about the range of data quality issues highlighted in this report, building on the supporting quality information provided with the new quarterly experimental statistics
requirement 8: publish its plans to restart and resource work to develop its business register
The requirements are covered in this document in the order 1, 5, 8, 7.
OSR requirements
OSR requirement 1
The demography statistics team in the Office for National Statistics (ONS) obtained the names of the users who had reported specific issues, and the users have been contacted to discover the exact nature of the problems faced. The ONS has received responses from them and is currently working on resolving their difficulties.
These difficulties principally related to users obtaining data from the Secure Research Service. The ONS has been looking to resolve those issues prior to the publication of the Office for Statistics Regulation (OSR) report.
This work will continue until these issues are fully resolved.
OSR requirement 5
The ONS introduced an experimental quarterly business demography release in response to user needs during the coronavirus (COVID-19) pandemic. We will continue to produce this and will enhance it with more detail as soon as possible. This product does not use the internationally agreed methodology for business demography because the methodology is not suitable for sub-annual estimates. Therefore, there are differences between the annual and quarterly figures.
The introduction of the quarterly product brings into question the need for the annual product, at least in its current form. Therefore, the ONS intends to engage with users to create a new suite of demography products.
The ONS will set up a steering group consisting of users of the business demography statistics, which will set the direction for this work and assess progress.
Our proposed approach would be to:
- further develop the quarterly demography output, by adding more geographic and industrial data
- reduce the coverage of the annual publication so that only business survival, high growth and employer demography remain
- consider how to meet our commitment to produce data for international comparison – for example, to potentially base this on - the aggregated quarterly dataset
The ONS has also recently published a longitudinal study on business dynamism. We will consider how this type of analysis can be used going forward to enhance Inter-Departmental Business Register (IDBR) based demography releases in the future. We will engage with users on these proposals to agree the new business demography products.
Any users of ONS business demography statistics are encouraged to email Mark Williams at mark.williams@ons.gov.uk with their views.
OSR requirement 8
The project to redevelop the business register was suspended in January 2019 to divert resources to other higher priority projects. The ONS has recently started working on a “proof of concept” to assess the most suitable technology for a new business register. As part of this, design and architectural work is currently being carried out. The proof of concept work is expected to be concluded before the end of the financial year. At that point, the ONS will assess priorities and decide how and when to move forward with a full-scale register redevelopment.
OSR requirement 7
Data quality issues for business demography statistics
The following text has been adapted from that included in the quarterly business demography statistical bulletins that the ONS began publishing in August 2020. This text will be developed over time and, in particular, quality issues that are more relevant to the annual release will be introduced and given greater attention.
Time of recording
Business creations and closures are based on the date on which the action occurs on the IDBR. Data for the ONS’s quarterly business demography release are extracted from the IDBR weekly and are allocated to quarters according to the date of the Friday of that week, as most major updates to the IDBR occur at the end of the working week. This may lead to some misallocation between quarters.
The date a business is added to the IDBR is generally on the same day, or within a few days, of the legal creation of the business as a company with Companies House. However, this can be several weeks after the effective birth of the business. For business closures, the registration process can take a little longer, as the death of a business may be long and complex. The “effective” death of a business may occur several months before its actual death from a legal perspective. A business is removed from the IDBR if information from HM Revenue and Customs (HMRC), ONS business surveys, Companies House or The Insolvency Service indicates it is no longer active. The ONS proves deaths by contacting companies if necessary.
Unregistered businesses
The business demography data are mainly based on data received from the Value Added Tax (VAT) and Pay As You Earn (PAYE) tax systems. There are turnover and lower-earning thresholds before a business registers for these taxes. This means that businesses can trade for some time before they register and hence appear as births in the business demography data. It is likely that some very small businesses are created and closed before reaching these thresholds.
The Department for Business, Energy and Industrial Strategy (BEIS) publication, Business population estimates, includes an estimate of the number of unregistered businesses.
Frequency of data
The IDBR is updated from four main sources:
- VAT
- PAYE
- ONS business surveys
- Companies House
The updates occur in various frequencies from daily to annual. Because the important PAYE update is quarterly, a shorter time period analysis of business creations and closures would be very volatile. As such, this source is best suited to quarterly publications.
Turnover and employment data
The turnover data on the IDBR are mostly derived from VAT or ONS business survey records. Employment data on the IDBR are derived from PAYE or ONS business survey records. In some cases, values are imputed from administrative data. The turnover data are updated annually, every September, from available data. Employment data are updated more frequently for some businesses but at least annually for all businesses.
The turnover and employment data for business closures are the stored values at the last update while the business was active on the IDBR, often the last annual update. These figures are not adjusted for inflation, so the average turnover would be expected to rise slowly over time in line with inflation.
For business creations, the value for turnover is usually that estimated by the business upon registration with HMRC for VAT. The employment value is the number of registered employees on their PAYE scheme if they have one, and it is imputed if they do not. This value is revised on the IDBR when more up to date data are received.
Data on turnover and employment on the IDBR should not be used to measure economic growth or the growth of the labour market – other ONS sources are preferred for these purposes. Data points are generally reasonably current for larger businesses, but they are less so for smaller businesses.
Strengths and limitations
Timeliness
In line with international guidance from Eurostat and the Organisation for Economic Co-operation and Development (OECD), the Office for National Statistics (ONS) annual business demography statistics are produced by comparing annual extracts from the business register. The ONS introduced a new quarterly business demography release in August 2020 to provide a more timely indicator of business creations and closures to support understanding of the impact of the coronavirus (COVID-19) pandemic on the UK economy. The quarterly data will not be entirely consistent with our annual publication, and the annual figures represent the superior measure of business demography at this time.
More detail about the Inter-Departmental Business Register (IDBR) in the annual official statistics on business demography is available.
Relationship with other data
The quarterly business demography release provides the first estimates of businesses added to, and removed from, the IDBR. However, there are many other publications drawing on other data sources to provide estimates of either business creation or business closures at different frequencies. These are briefly outlined in this section.
Quarterly official statistics from Companies House include the number of company incorporations, dissolutions and removals from their register. Companies and businesses are not the same in these statistics – companies are legal entities, as registered with Companies House; businesses are statistical entities, arranged by the ONS on the IDBR, which better reflect their economic activity. Some types of company are present with Companies House but not on the IDBR or in Value Added Tax (VAT) returns, such as small single-person limited companies.
We have published new weekly indicators of company incorporations and voluntary dissolutions using data from Companies House in our latest indicators bulletins, but this covers only one route for a business to be removed from the IDBR.
There are many ways for a business to close or die, of which “insolvency” is only one. Data from the Insolvency Service on company insolvencies therefore do not reflect all business closures – insolvencies typically account for a small fraction of business closures, although they tend to be the highest profile.
Data from the Business Impact of Coronavirus (COVID-19) Survey (BICS) include estimates of the proportion of surveyed businesses that have “temporarily ceased trading”. Businesses that are still active but temporarily ceased trading because of the coronavirus pandemic are not removed from the IDBR and therefore are not reflected in the data in the quarterly release. BICS also reports data on the proportion of responding businesses that have permanently ceased trading. This is based on a filter question in the survey to reduce burden for those businesses that have closed down, as the survey is not relevant for them. As such, those estimates are not intended to capture business closures, so the data in the quarterly release are preferred.
The data in the quarterly release are most similar to the concepts of business births and deaths in the annual business demography statistics, which are National Statistics. The calculation of business births and deaths in the annual publication is more robust than in the quarterly publication, and it follows international guidance. The data in the quarterly release have been created rapidly to provide insight on the coronavirus pandemic, and they should therefore be treated with caution.
Multiple registrations at a single postcode
In recent years, the number of multiple registrations observed on the IDBR at a single postcode has increased. This can cause large increases in both births and deaths and distort lower geography data analysis.
There are several reasons why these multiple registrations can occur. For example:
- the increase in the use of management and personal service companies
- the use of formation agents to register a new business with Companies House
- virtual offices
- the presence of the HM Revenue and Customs (HMRC) local offices
- foreign online sellers being registered at the same address
It is not obvious how to deal with the challenges caused by multiple business registrations at the same site since they are “valid” registrations. The ONS intends to look at the issue carefully and take on views from all interested parties.
Reactivations
The Eurostat and OECD guidelines recommend waiting two years following the reference period before counting a business as dead. This is because businesses can be reactivated, for example, by being sold by the administrator. To avoid such a lengthy delay, the ONS estimates the likely number of reactivations, to provide a better measure of deaths. However, this estimate leads to revisions. We will review the methodology around the estimation of reactivations in the context of the change of focus to produce the main demography data on a quarterly basis.