Research and Experimental Development, Businesses, Australia methodology

1 The statistics presented in this release have been compiled from data collected in the Survey of Research and Experimental Development (R&D), Businesses for 2017-18.

2 The Survey of R&D, Businesses is a biennial survey. The change to the collection frequency from annual to biennial was made after the 2011-12 survey.

3 R&D, as collected by the ABS, is defined in accordance with the Organisation for Economic Co-operation and Development (OECD) standard as "creative and systematic work undertaken in order to increase the stock of knowledge - including knowledge of humankind, culture and society - and to devise new applications of available knowledge".

4 For a more comprehensive interpretation of the definition of R&D, see the Australian and New Zealand Standard Research Classification (ANZSRC), 2008 or refer to the OECD publication The Measurement of Scientific and Technological Activities: Proposed Standard Practice for Surveys of Research and Experimental Development - Frascati Manual 2015.

5 Data providers report and self-classify R&D survey information based on their interpretation of OECD and ABS definitions and classifications. The ABS makes every effort to ensure correct and consistent interpretation and reporting of these data by applying consistent processing methodologies. See also the Non-Sampling Error sections of the Technical Note under the Methodology page.

6 The Economics Unit Model is used by the ABS to describe the structure of Australian businesses and other organisations. The model consists of four types of units:

• Enterprise Group (EG)
• Type of Activity Unit (TAU)
• Legal Entity (LE)
• Location Unit.

7 Businesses contributing to the estimates in this publication are sourced from the ABS Business Register (ABSBR), and are selected at either the Australian Business Number (ABN) unit or the Type of Activity Unit (TAU) level, as described below.

8 In the Survey of R&D, Businesses, the statistical unit used to represent the majority of businesses, and for which statistics are reported, is the ABN unit. The ABN unit is the business unit which has registered for an ABN, and appears on the Australian Tax Office (ATO) administered Australian Business Register (ABR). This unit is suitable for ABS statistical needs when the business is simple in structure. In these instances, one ABN equates to one statistical unit. These units are collectively referred to as the non-profiled population.

9 For more significant and diverse businesses where the ABN unit is not suitable for ABS statistical needs, the statistical unit used is the TAU, which comprises one or more business entities, sub-entities or branches of a business entity within an Enterprise Group that can report production and employment activities. When a minimum set of data items is available, a TAU is created which covers all the operations within an industry subdivision. The TAU is classified to the relevant subdivision of the Australian and New Zealand Standard Industrial Classification (ANZSIC). These units are collectively referred to as the profiled population.

10 Further information on the ABSBR and ABS economic units model can be found in Australian Bureau of Statistics Business Register.

11 The statistics in this release are classified to industry in accordance with the Australian and New Zealand Standard Industrial Classification (ANZSIC), 2006.

12 Each ABN unit/TAU is classified by the ABS to the industry in which it mainly operates. For the Survey of R&D, Businesses, where an Enterprise Group sets up a dedicated research unit, that unit is classified to the predominant industry of the group rather than to Scientific research services (ANZSIC 6910), in accordance with standards set out in the OECD Frascati Manual 2015.

13 From 2005-06, the survey scope was adjusted to:

• include businesses classified to Division A (Agriculture, Forestry and Fishing) of the Australian and New Zealand Standard Industrial Classification (ANZSIC); and
• exclude businesses with intramural expected expenditure on R&D of less than $100,000 in the reference period (i.e. introduction of an expenditure based scope cut-off).
   

14 Intramural expenditure is defined as expenditure for R&D performed by the statistical unit regardless of the source of funds. Expenditures made outside the statistical unit but in support of intramural R&D are included; for example, payments for analytical work, engineering or specialised services which form part of an R&D project performed by the statistical unit. R&D funded by the statistical unit but performed wholly by another organisation on their behalf (extramural R&D) is excluded. R&D performed overseas by Australian businesses is included. For further information, refer to the OECD Frascati Manual 2015.

15 Prior to the 2005-06 cycle, the Survey of R&D, Businesses included all Australian businesses performing R&D (regardless of the expenditure value) with the exception of businesses mainly engaged in agriculture, forestry and fishing activities.

16 The ABS identifies businesses for inclusion in the Survey of R&D, Businesses frame if the business:

• reported expenditure on R&D in previous surveys;
• applied for an AusIndustry administered R&D Tax Incentive and/or grant for industry R&D;
• were identified through other sources such as newspapers, journals, research compendia, etc. as likely to have expenditure on R&D.
   

17 Collection of data included in this release was undertaken based on a stratified random sample of 4,196 businesses sourced from the Survey of R&D, Businesses frame. The sample was stratified by industry Division and/or Subdivision, and Size group based on their expected R&D expenditure. All businesses on the Survey of R&D, Businesses frame within certain strata were included in the sample if they met industry specific thresholds. Thresholds were based on industry characteristics, expected expenditure and number of businesses in the strata. While the sample design excludes businesses with intramural expected expenditure on R&D of less than $100,000, data of selected businesses which reported expenditure below this threshold are included in the estimates.

18 For the 2017-18 reference period, the survey has been conducted predominantly via online forms. The survey achieved a response rate of 90%.

19 Prior to 2010-11, data for each sector were collected separately for Gross Expenditure on R&D (GERD). Results were published in the Research and Experimental Development, Australia, All Sector Summary (last released in October 2010).

20 From 2010-11, estimates of GERD could not be produced in the same manner as the frequency and timing of some collections used to produce components of GERD had changed. As a result, gross expenditure on R&D has been modelled since 2010-11.

21 Until 2015-16 a predictive model was used to estimate GERD. This utilised a combination of budget papers and annual reports as input variables for modelling the expenditure for the Government, Private Non-Profit and Higher Education sectors.

22 The 2017-18 estimates of GERD applied an ARIMA (1,1,0) time series modelling method to produce estimates for the Government, Private Non-Profit and Higher Education sectors. Expenditure for the Business sector uses the directly collected information for 2017-18.

23 Whilst estimates produced by the two methods are comparable, the introduction of time series modelling improves efficiency, being the most recommended and widely used technique in the ABS for time series analysis and forecasting.

24 Businesses were asked to report the number of persons working for the business during the last pay period in June 2018. For output purposes, businesses are classified to employment size groups based on data reported in the survey.

25 Location of expenditure relates to the region(s) in which the business reported having performed R&D during the reference period. This may not be the head office location of the business.

26 Type of Activity, Fields of Research and Socio-economic Objective statistics presented in this release have been collected and compiled based on the Australian and New Zealand Standard Research Classification (ANZSRC), 2008.

27 The most recent GDP and GSP values available were used to calculate the R&D expenditure/GDP and R&D expenditure/GSP ratios presented in this issue. These values are referenced in the tables below and have been revised from those used to calculate ratios in previous issues.

28 GDP is estimated by the ABS according to the international standards contained in the System of National Accounts, 2008 (2008 SNA) and is not directly comparable to GDP from countries where these standards have not been applied.

29 The chain volume measures appearing in this release are annually reweighted chain Laspeyres indexes referenced to the current price values in a chosen reference year (currently 2017-18). They can be thought of as current price values re-expressed in (i.e. based on) the prices of the previous year and linked together to form continuous time series. They are formed in a multi-stage process of which the major steps are described in Section 15 of the Information Paper: Australian National Accounts, Introduction of Chain Volume Measures and Price Indexes.

30 With the implementation of 2008 SNA, deflators used to calculate the chain volume measure of expenditure on R&D were revised to better capture changes in the unit value of labour used in the production of R&D services, and to increase and refine the number of products included in the deflators. The revised deflators have been used for Business R&D statistics from the 2007-08 survey cycle.

31 Where figures have been rounded, discrepancies may occur between the sum of the component items and totals.

32 Relevant OECD publications include:

Main Science and Technology Indicators Database 

33 The ABS acknowledges the ongoing contribution made by the Department of Industry, Innovation and Science in providing R&D tax incentive and grants lists.

34 ABS publications draw extensively on information provided freely by individuals, businesses, governments and other organisations. Their continued cooperation is very much appreciated; without it, the wide range of statistics published by the ABS would not be available. Information received by the ABS is treated in strict confidence as required by the Census and Statistics Act 1905.

35 As well as the statistics included in this and related publications, the ABS may have other relevant data available on request. Inquiries should be made to the National Information and Referral Service on 1300 135 070.

1 Non-sampling errors may arise as a result of errors in the reporting, recording or processing of data. These errors can be introduced through inadequacies in the questionnaire, treatment of non-response, inaccurate reporting by data providers, errors in the application of survey procedures, incorrect recording of answers and errors in data capture and processing.

2 The extent to which non-sampling error affects the results is difficult to measure. Every effort is made to minimise non-sampling error by careful design and testing of the collection instrument, the use of efficient operating procedures and systems, and the use of appropriate methodologies.

3 When interpreting the statistics in this release, the reliability and comparability of the estimates may be affected by the following specific non-sampling errors:

• Many businesses provided estimates due to a lack of separately recorded data on R&D activity.
• Some businesses may not have reported data as per the definition of R&D used in this survey. This is potentially a result of slight differences in the survey definition of R&D and those used in industry R&D schemes for the allocation of grants, and the AusIndustry administered R&D Tax Incentive scheme for tax deductibility for specific R&D activities.
• Data were self-classified by businesses to Type of Activity, Fields of Research and Socio-economic Objective at the time of reporting. Some businesses may have experienced difficulty in classifying their R&D projects. The ABS makes every effort to ensure correct and consistent interpretation and reporting of these data by applying consistent processing methodologies.
• The estimation method for R&D related overhead costs varied across businesses and reference periods.
   

4 As the estimates in this publication are based on information relating to a sample of businesses, they are subject to sampling variability, that is, they may differ from the estimates that would have been produced if the information had been obtained from all businesses. 

5 The difference between estimates obtained from a sample of businesses, and the estimates that would have been produced if the information had been obtained from all businesses, is called sampling error. This should not be confused with inaccuracy that may occur because of imperfections in reporting by respondents or in processing by the ABS. Please see the section on Non-Sampling Error for more detail regarding these types of errors. The expected magnitude of the sampling error associated with any estimate can be estimated from the sample results. One measure of sampling error is given by the standard error (SE), which indicates the degree to which an estimate may vary from the value that would have been obtained from a full enumeration (the 'true' figure). There are about two chances in three that a sample estimate differs from the true value by less than one standard error, and about nineteen chances in twenty that the difference will be less than two standard errors. 

6 An example of the use of a standard error is as follows (all figures shown are in $'000). From the publication, the estimated total expenditure on R&D was $17,437,585 with a standard error of $301,670. There would be about two chances in three that a full enumeration would have given an estimate in the range $17,135,915 to $17,739,255 and about nineteen chances in twenty that it would be in the range $16,834,245 to $18,040,925.

7 In this publication, indications of sampling variability are measured by relative standard errors (RSEs). The relative standard error is a useful measure in that it provides an immediate indication of the percentage errors likely to have occurred due to sampling, and thus avoids the need to refer to the size of the estimate. RSEs are obtained using the formula: RSE = SE/estimate x 100. RSEs are shown in the Relative Standard Error tables in this section. RSEs for all data included in this release (including data cube content) are available upon request. 

8 Estimates with RSEs between 25% and 50% are annotated to indicate they are subject to high sample variability and should be used with caution. In addition, estimates with RSEs greater than 50% have been included and annotated to indicate they are considered too unreliable for general use. All cells in the data cubes with RSEs greater than 25% contain a comment indicating the size of the RSE. These cells can be identified by a red indicator in the corner of the cell. The comment appears when the mouse pointer hovers over the cell.

9 The comparability of estimates over time may be affected by the following changes in classifications:

• Employment size classification groups are defined on data reported in the reference period, and as such businesses may be categorised to different employment size groups across different reference periods.
• Businesses can also be classified to different industry divisions across survey reference periods as a consequence of structural change. Refer to the Methodology page for more information.