In 2010, £16.1 billion was spent on R&D performed within UK businesses, an increase of 3.7 per cent in cash terms (current prices) compared with the 2009 total. Total R&D expenditure in 2010 represented 1.1 per cent of GDP, in line with recent years (Table SB1).
|In cash terms|
|In real terms|
There was more R&D performed in the Pharmaceuticals product group than any other, with expenditure in 2010 at £4.6 billion, 28.8 per cent of all spending. Other product groups reporting significant R&D expenditure were Computer programming and information service activities, which accounted for £1.7 billion (10.3 per cent), Aerospace, which accounted for £1.4 billion (8.8 per cent), Motor vehicles and parts, which accounted for £1.3 billion (7.8 per cent) and Telecommunications, which reported £1.0 billion (6.5 per cent).
The largest source of R&D funding was from businesses’ own funds at £10.4 billion in 2010, 64.5 per cent of the total. Funding of R&D in UK businesses from abroad was £3.6 billion, 22.4 per cent of the total. The UK Government’s funding of businesses’ R&D was £1.4 billion, 8.7 per cent of the total, and predominantly in defence. Table SB6 shows the sources of funds broken down by Civil and Defence.
The number of full time equivalent staff employed on R&D increased by 4,000 between 2009 and 2010 to 155,000. The number of scientists and engineers increased by 1,000 to 84,000 and accounted for 54.3 per cent of all staff employed on R&D. Between 2009 and 2010, the number of technicians increased by 1,000 while administrative staff increased by 2,000.
In real terms, Civil R&D increased by 3.0 per cent and Defence R&D decreased by 13.8 per cent in 2010. Within civil expenditure, Manufacturing accounted for 70.1 per cent of the total and Services accounted for 28.0 per cent. Out of the broad product groups, Mechanical Engineering has replaced Aerospace as the largest contributor to defence R&D expenditure. Aerospace is now the second largest followed by Electrical Machinery.
In 2010, 92.6 per cent of UK business R&D expenditure was carried out in England, up slightly from 92.4 per cent in 2009. The UK regions with the largest R&D expenditure were: the East of England (24.8 per cent of the total); the South East (23.4 per cent) and the North West (12.7 per cent).
The detailed product group titles in Table SB2 have been updated to better reflect current industry classifications. For more information on this and other changes, please see the Background Notes.
This R&D survey is an annual sample survey based on a continually updated register of R&D performers.
The survey of Business Enterprise Research and Development is conducted by the Office for National Statistics (ONS). Approximately 5,000 questionnaires were sent to businesses known to perform R&D; this included around 400 of the largest R&D spenders, which accounted for 80 per cent of the 2010 total R&D expenditure figure. Smaller R&D performers and others believed to be performing R&D were selected using various sampling fractions. Industry product group and business employment size were the stratification variables. Completed questionnaires were returned by 4,672 businesses representing a response rate of 91 per cent.
This Statistical Bulletin reports business R&D performed in the UK irrespective of the residence of the ultimate owner. Overseas activities of affiliates of UK businesses are not included. Gross expenditure of R&D in the UK performed by all sectors of the economy is reported separately.
A Summary Quality Report for this publication is available on the ONS website.
R&D and related concepts follow internationally agreed standards defined by the Organisation for Economic Cooperation and Development (OECD), published in the ‘Frascati’ Manual. R&D is defined as “Creative work undertaken on a systematic basis in order to increase the stock of knowledge, including knowledge of man, culture and society and the use of this stock of knowledge to devise new applications”.
There may be discrepancies between totals and the sum of their independently rounded totals. In some tables, entries have been aggregated to avoid disclosure of figures in which the returns of individual businesses could be identified. Where this happens, footnotes have been added to the tables.
The results from the Business Enterprise Research and Development survey are based on the whole UK economy. This survey excludes R&D funded by UK businesses that is performed overseas or in other sectors of the UK economy (such as higher education; government departments, agencies and non-departmental public bodies; local authorities; and private non-profit organisations). Public corporations are counted as business enterprises.
The 400 largest R&D spenders are asked to select the industry product groups that best describes the type of R&D activities they undertake. In 2010, these product groups were updated to better reflect the new Standard Industrial Classification (SIC 2007) codes. For the smaller R&D performers, no product group data were collected. However, the businesses’ Standard Industrial Classification (SIC) codes are known from the Inter-Departmental Business Register (IDBR) and the assumption is made that the R&D expenditure is for the detailed product group corresponding to that SIC. This approach must be regarded as approximate since, in practice, an individual business can perform R&D for a range of product groups.
The introduction of SIC 2007 in 2009 has resulted in some businesses’ R&D moving to a different product group than previously published. The largest impact has been with businesses with Publishing activities as these have moved out of the manufacturing sector and are now included under Miscellaneous business activities. There has been an increase in the Other manufactured goods product group due to businesses being reclassified from Textiles, clothing and leather products, Pulp and paper products, Rubber and plastic products, Fabricated metal products, Machinery and equipment and Precision instruments and optical products. Also, the Processing of nuclear fuels is now classified with Non-ferrous metals instead of Refined petroleum and coke oven products.
The broad product groups, which consist of aggregations of the detailed product groups, were refined and expanded in 2002 in order to more accurately categorise the data within the manufacturing and service sectors. Services’ broad product group includes R&D consultancy and other services in support of the manufacturing industry.
Employment is provided by respondents on the basis of ‘full-time equivalent’ staff averaged over the year. The categories of employment used are:
Researchers – engaged in the conception or creation of new knowledge, products, methods and systems
Technicians – perform scientific and technical tasks normally under the supervision of researchers
Others – support staff including skilled and unskilled craftsmen, secretarial and clerical staff participating in R&D projects
The sample and survey results cover ‘business enterprises’ as defined in the ‘Frascati’ Manual. This excludes government organisations, higher education establishments and charities.
R&D surveys pose special problems relating to survey design and estimation. R&D takes place in only a small proportion of businesses but a comprehensive list of who these businesses are does not exist. A simple random sample of the business population would not be suitable for an R&D survey because many of the sampled businesses would not undertake R&D, and many significant R&D performers would be missed in such a sample.
The solution is to implement a stratified sample design. The stratification variable is the known level of R&D performance of the businesses. This information is gained from the previous year’s survey and extra information from various sources, for example, the Department for Business, Innovation and Skills (BIS) and filter questions on other ONS surveys. Two different types of questionnaires are used. A ‘long’ version, requesting a full detailed breakdown of R&D activity, was despatched to those business enterprises which in the previous survey had returned or estimated an R&D expenditure figure of greater than £3.3 million, together with businesses newly identified as very large R&D spenders. Businesses identified as smaller R&D performers were sampled. The selected businesses were sent a ‘short’ version of the R&D questionnaire which requests just the R&D expenditure and employment totals. The detailed information for these businesses was estimated by using the data received on the long questionnaires. The totals of the unsampled businesses were estimated using ratio estimation with business employment as the auxiliary variable, a variable held on the IDBR.
The 2010 survey used individual businesses as reporting units. For the purpose of the survey, businesses were grouped into three strata:
Businesses responding to the 2009 Business Enterprise R&D survey who returned or had estimated a total R&D expenditure value of greater than £3.3 million
Businesses newly identified as large R&D spenders
Businesses reporting less than £3.3 million R&D expenditure in the 2009 survey, or 2008 survey if they were not selected for the 2009 survey
Businesses reporting positively to R&D filter questions in other ONS surveys
Other identified potential R&D performers
The remainder of the UK businesses
The businesses making up strata (i) and (ii) formed a register of R&D performers. The businesses in Stratum (iii) were not included as they were assumed to have zero R&D expenditure. The sample for the 2010 survey was then derived from the R&D performers' register.
The businesses in stratum (ii) were sampled using a stratified sampling scheme where the stratification variables were industry product group and size band. Three size bands were defined using business employment as it appears on the IDBR, and each business was assigned to one of 33 distinct product groups. Businesses with employment greater than 399 were all selected and classed as size band 1. The second size band contained those businesses with employment between 100 and 399. The sampling ratio for this size band was approximately 1 in 3. The third size band contained those businesses with employment of less than 100. These were sampled with a 1 in 4 sampling ratio, although conditional on a minimum sample size of ten in each industry group/size band ‘cell’. With three size bands and 33 industry groups, 99 ‘cells’ were created in all. Approximately 4,700 selected businesses in this stratum were sent a short version of the questionnaire.
Estimation is required in two areas:
a. Long questionnaire non-responders
Businesses who received long questionnaires and did not respond, had estimates constructed using information from all those responders in the same product group who returned questionnaires for the current year and previous year. The ratio of the latest period's results to the previous period's results was calculated for these businesses and then applied to the non-responders’ 2009 figure. This gave a 2010 estimate for the non-responders.
b. Non-responding and unsampled businesses
In stratum (ii), non-responding businesses were treated as unsampled businesses. The R&D expenditure total was estimated separately for each cell using ratio estimation with business employment as the auxiliary variable. This involved calculating the ratio of the 2010 responder’s total in-house expenditure to the responder’s total employment and applying it to the sum of the non-responders/unsampled total employment figures.
The long questionnaire detail for the businesses in stratum (ii) was estimated using the 2010 long questionnaire detail returned from the large R&D spenders. This was done by calculating the long questionnaire detailed variable breakdowns as proportions of the totals and applying these proportions to the totals estimated for the businesses in stratum (ii). This methodology means that it has not been necessary to go back to past data when estimating the current period's results.
The 2008 and 2009 figures have been revised where necessary to take account of businesses misreporting and late returns.
One indication of the reliability of the key indicators in this release can be obtained by monitoring the size of revisions. The table below records the size and pattern of revisions that have occurred over the last five years. Please note that these indicators only report summary measures for revisions. (The revised data may be subject to sampling or other sources of error).
|(1) 2010 Value||(2) Mean revision||(3) Average absolute revision|
|Expenditure on R&D|
|performed by UK||16,067||-425||555|
(1) - Value in latest period.
(2) - Average over the last 5 years (mean revision).
(3) - Revisions between first publication and estimates three years later. Average over the last 5 years without regard to sign (average absolute revision).
The revisions table covers estimates of the UK business enterprise R&D expenditure first published from November 2004 (for 2003) to November 2008 (for 2007).
A statistical test has been applied to the average revision to find out if there is bias in the estimates. No statistically significant bias was identified.
Respondents were asked to make a return for the calendar year 2010 or the nearest 12 month period for which figures were available. Data for all years published in this Statistical Bulletin were collected on the same basis.
The survey questionnaire was redesigned from the 2007 survey to better reflect user needs and to address concerns about data quality and difficulty in completion. While the changes are an improvement, they could have an impact on the comparability of the data returned, as businesses better understand what is required and quality improves.
A table giving estimates of R&D expenditure by SMEs has been included (Table 26). The SME definition used is that under the European Commission Recommendation (96/280/EC) of 3 April 1996, in which SMEs are defined as enterprises with less than 250 employees. In addition, a criterion of independence is used to exclude enterprises that are part of a larger enterprise group, so that only true SMEs are evaluated. This criterion is important in the context of R&D estimates, given that R&D activity is often carried out by smaller businesses which form part of larger, sometimes multinational, businesses.
To apply the SME definition, historic information on business ownership has been obtained using that currently held on the IDBR. Caution should therefore be used in making comparisons over time.
The businesses receiving the long questionnaire (stratum (i)) account for approximately 80 per cent of the total expenditure. Each business is asked to name all the sites at which the business performed R&D, and to allocate the total expenditure figure of the business to the sites on a percentage breakdown basis. Regional data for the businesses within stratum (ii), which account for the remaining 20 per cent of total expenditure, all have a value estimated by grossing up over all three size bands using county region codes from the business register of R&D performers. Aggregating is done at broad product group and county level. This procedure is used to produce regional estimates.
It is possible to calculate the standard error for each detailed product group, as a measure of sampling error. To illustrate this, the total expenditure for product group V (Aerospace) is £1,407 million. The calculated standard error, as shown in Table 25, is £44 million.
The 95 per cent confidence interval for this estimate is then given as £1,321 million to £1,493 million i.e. £1,407 million +/- 1.96 x the standard error. A 95 per cent confidence interval means that it is expected that in 95 per cent of samples, this range would contain the true value.
The estimates in this publication are based on a stratified sample drawn from the population of businesses known to be actual or likely R&D performers. As with any sample survey, these estimates are subject to two types of possible errors:
Sampling errors, due to only a sample of the population being surveyed. These errors can be quantified, as shown in Table 25.
Non-sampling errors. These include factors such as population coverage, misreporting and non-response bias. These errors are generally hard to quantify. Because of the difficulty in identifying the population of actual/likely R&D performers and because of the problems ensuring that businesses adhere to Frascati R&D definitions, non-sampling errors could be sizeable for the R&D inquiry.
The following symbols are used in the tables:
‘..’ denotes disclosive figures
‘-’ denotes value too small to display
‘r’ denotes revised figures
‘†’ crosses denote earliest data revision
Details of the policy governing the release of new estimates are available from the Media Relations Office. A list of those given pre-publication access (34.2 Kb Pdf) is available on the ONS website.
National Statistics are produced to high professional standards set out in the Code of Practice for Official Statistics. They undergo regular quality assurance reviews to ensure that they meet customer needs. They are produced free from any political interference.
© Crown copyright 2011.
Details of the policy governing the release of new data are available by visiting www.statisticsauthority.gov.uk/assessment/code-of-practice/index.html or from the Media Relations Office email: email@example.com
|Mark Williams||+44 (0)1633 456728||Office for National Statistics||RandD@ons.gsi.gov.uk|