SME Statistics: Business Survival Rates
Survival rates of VAT registered enterprises,
Guidance and Methodology
1. The Small Business Service (SBS) has, in recent years, published annual or biennial estimates of the survival rates of new businesses, based on VAT registrations and subsequent de-registrations. This year, estimates for enterprises registering in 2004 have been published for the first time and previous years’ estimates have been revised. These figures can be downloaded from www.sbs.gov.uk/survival.
2. This paper describes the data sources and methodology used to produced the estimates, along with guidance on their accuracy and how to use the figures. It accompanies a key results paper summarising the broad patterns in survival rates and an Excel workbook containing the following tables:
3. A guide to using the data is included at the front of the Excel workbook.
4. Due to resource constraints, the DTI cannot guarantee that any of these figures will continue to be published in future. They will not be published in 2008 so, at the earliest, the next update will be in early 2009.
5. This series is based on data taken from the Office for National Statistics’ (ONS) Inter Departmental Business Register (IDBR) in May 2006.
6. The IDBR is a comprehensive database of UK businesses based on inputs from three administrative sources: traders registered for Value Added Tax (VAT) purposes; employers operating a Pay As You Earn (PAYE) scheme registered with HM Revenue and Customs (HMRC); and incorporated businesses registered at Companies House. For more information on the IDBR extract obtained by the SBS see the guidance note accompanying Business Start-ups and Closures: VAT Registrations and De-registrations, which can be found at www.sbs.gov.uk/vat.
7. VAT registrations and de-registrations are used as a proxy for business start-ups and closures. However, enterprises registering for VAT are not necessarily new businesses, as enterprises are unlikely to register if their turnover falls below the compulsory VAT threshold (which was £58,000 at the start of 2005). Similarly, enterprises that de-register may not necessarily have closed. Only 1.8 million out of an estimated 4.3 million enterprises in the UK at the start of 2005 were registered for VAT. Our estimates of survival rates therefore exclude a large proportion of the very smallest businesses.
8. In this analysis, enterprises that de-register for VAT are not counted as de-registered if they still have an active PAYE record, as this indicates that the business is still trading despite having fallen below the VAT threshold. This is no different from the methodology used in the VAT Registrations publication.
9. As time goes on, the number of VAT registrations recorded for a particular year tends to increase as late registrations come in. However, although late de-registrations are also picked up, in time the overall number of VAT de-registrations tends to decrease due to dormant records that were previously classed as de-registered becoming active again.
10. Data have been adjusted to take into account lags in the reporting of VAT registrations and de-registrations. The eventual number of registrations and de-registrations in each year of the series has been predicted by looking at how the numbers recorded in previous extracts have changed with time. More information on the adjustment factors used can be found in the guidance note accompanying Business Start-ups and Closures: VAT Registrations and De-registrations at www.sbs.gov.uk/vat.
Survival rates estimates
11. An extract was taken from the IDBR listing the number of new VAT registrations in each year from 1995 to 2004 for each Local Authority, region, industry, etc.. The number of these enterprises that de-registered within each three-monthly interval after registration was also listed.
12. These data were lag adjusted using the method described in paragraph 10, above. Enterprises that de-registered within 6 months of registration were assumed to have de-registered in the registration year, while those that de-registered after six months but within 18 months of registration were assumed to have de-registered in the subsequent year, those de-registering after 18 to 30 months the year after that, and so on.
13. The number of new registrations remaining at three-monthly intervals after registration was calculated.
14. Survival rates were calculated by dividing the number remaining at each three-monthly interval by the original number of registrations.
Changes to the series
15. This year survival rates have been shown for each English and Welsh Local Authority, Scottish Council Area and Northern Ireland District Council Area. This is a change on last year’s publication, which presented survival rates by Business Link area in England, Training and Enterprise Council (TEC) in Wales and Local Enterprise Company (LEC) in Scotland. A list of Local Authorities by Business Link area/TEC is given in Annex A, along with a worked example of how to estimate survival rates for composite areas.
16. The lag adjustment factors have been updated since the last publication in February 2006, affecting all of the UK, although the impact on individual Local Authorities is small. The estimated number of VAT de-registrations in 2004 has been revised downwards by almost nine per cent, leading to a noticeable upwards revision of the one-year survival rates for 2003 and the three-year survival rates for 2001.
17. There has been a minor change in the way de-registration years have been assigned. In last year’s publication enterprises de-registering three to six months after registration were assumed to de-register in the following year. In this year’s publication these enterprises are assumed to have de-registered in the same year as registration. This change in assumptions improves slightly the accuracy of the figures, with a magnitude of around 0.1 of a percentage point.
Rural Urban Definition and Index of Multiple Deprivation
18. This year, survival rates figures have been shown broken down by the Rural and Urban Area Classification and Index of Multiple Deprivation (IMD) quintile, for each region.
19. The Rural and Urban Area Classification was introduced in 2004. Areas are assigned to one of four different types based on the size of settlement that the area lies within. In order of decreasing settlement size these are: urban; rural – town and fringe; rural – village; and rural – hamlet and isolated dwelling. Each of these area types can then be subdivided into “sparse” and “less sparse” based on the underlying population density of the broader area, giving an eightfold classification of the type of area that an enterprise is located in. As Scotland and Northern Ireland use their own versions of the Rural Urban Classification, figures are presented only for England and Wales.
More information on the Rural and Urban Area Classification can be found at http://www.statistics.gov.uk/geography/nrudp.asp or at http://www.defra.gov.uk/rural/ruralstats/rural-definition.htm.
20. The Index of Multiple Deprivation 2004 (IMD 2004) combines seven different Domains of deprivation: income deprivation; employment deprivation; health deprivation and disability; education, skills and training deprivation; barriers to housing and services; living environment deprivation; and crime, to arrive at an overall ranking of deprivation at the Super Output Area (SOA) level. This ranking can then be used in each region individually or for England as a whole to determine the deprivation quintile of the area in which an enterprise is based. As other UK countries define their own multiple deprivation indices, figures presented are for England only.
More information on the Index of Multiple Deprivation can be found at http://www.communities.gov.uk/index.asp?id=1128440.
For further information on the publication and the methodology, please contact:
DTI Small Business Service Statistics Team
Tel: 0114 279 4344
E-mail: [email protected]
Annex A – Composite Areas
Previous years’ publications have broken down the survival rates figures by Business Link area in England, Training and Enterprise Council (TEC) in Wales and Local Enterprise Company (LEC) in Scotland.
In this year’s publication survival rates have instead been presented in more detail, broken down by English and Welsh Local Authority, Scottish Council Area and Northern Irish District Council Area.
Tables 2 and 3 on the following pages show which Local Authorities make up the English Business Link areas and Welsh TECs used in former years. The remainder of this Annex shows a worked example of how to estimate survival rates for composite areas.
Worked example: calculation of survival rates for composite areas
1. This section gives a worked example of how to estimate survival rates for Business Link areas or TECs from this year’s Local Authority figures. We will use Bedfordshire as the example Business Link area. This method requires the use of VAT registration figures in Table 1a of the DTI publication Business Start-ups and Closures: VAT Registrations and De-registrations, which can be found at www.sbs.gov.uk/vat.
2. Table 1 shows that Bedfordshire is made up of four Local Authorities: Bedford, Luton, Mid Bedfordshire and South Bedfordshire. The one-year survival rates for each of these areas for businesses registering in 2004 are shown in Table 1 below, alongside the number of VAT registrations:
Table 1 Survival rates and number of VAT registrations for Bedfordshire Local Authorities, businesses registering in 2004
3. As we want the Local Authorities with a higher number of VAT registrations to have a bigger effect on the average, we need to multiply the survival rate in each Local Authority (first column of numbers) by the number of VAT registrations in this authority (second column). The results of this can be seen in the last column of Table 1.
4. We then add the four numbers in the last column of Table 1 together:
41,360 + 39,585 + 48,925 + 39,060 = 168,930
5. We can then calculate the average survival rate for the Business Link area by dividing the total above by the total number of registrations, given on the bottom line of Table 1:
Average survival rate for Bedfordshire = 168,930 ¸ 1,840
= 92 percent
6. This method can be used to calculate average survival rates for any group of Local Authorities.
The boundaries of English Local Authorities map directly on to the boundaries of Business Link areas. Table 2 shows the composition of the Business Link areas used previously.
Table 2 List of English Local Authorities that comprise Business Link areas
The boundaries of Welsh Local Authorities map directly on to the boundaries of the Training and Enterprise Councils in Wales, with the exception of Gwynedd, which is split between Celtec and Mid Wales. A map of TECs including those in Wales can be found at
Table 3 List of Welsh Local Authorities that comprise TECs
The boundaries of Scottish Council Areas do not map neatly on to those of Local Enterprise Companies. More information on the areas covered by Local Enterprise Companies in Scotland can be found at http://www.scottish-enterprise.com/sedotcom_home/about_se/local_enterprise_companies.htm
and at http://www.hie.co.uk/Local-Enterprise-Companies.htm for the Highlands and Islands.
Our previous publication did not provide a breakdown of survival rates by Northern Irish geographies.
 More information on the Index of Multiple Deprivation 2004 can be found at http://www.communities.gov.uk/index.asp?id=1128440.
 Information on the rural and urban classification for Scotland is available from the Scottish Executive at http://www.scotland.gov.uk/Publications/2004/06/19498/38784
and that for Northern Ireland from the Northern Ireland Statistics and Research Agency at http://www.nisra.gov.uk/statistics/financeandpersonnel/DMB/urban_rural.html.