APPENDIX 2 INNOVATION EXPENDITURE
1 The ABS provides annual statistics related to the incidence of innovation by Australian businesses, along with a comprehensive range of business innovation indicators and related practices collected biennially as part of the Business Characteristics Survey (BCS). The BCS draws on the conceptual definitions and guidelines included in the 'Oslo Manual, Guidelines for Collecting and Interpreting Innovation Data, (Third Edition, 2005)'.
2 Statistics for innovation expenditure provide information on the resources allocated by businesses to innovation-related activities. Innovation expenditure statistics provide a more complete picture of business innovation, which greatly enhances the ability of policy makers and researchers to objectively assess the economic impacts of innovation and innovation policies.
DEFINING INNOVATION EXPENDITURE
3 Expenditure on innovation comprises all expenditure incurred by businesses on the development or introduction of all new or significantly improved goods, services, processes or methods over a financial year period. Business expenditure on innovation activities may include: acquisition of machinery, equipment or technology; training specific to innovation; marketing activities undertaken to introduce innovation; internal and external research and experimental development on innovation; design, planning or testing to develop or introduce innovation; and the acquisition of licences, rights, patents or other intellectual property associated with innovation.
HISTORY OF INNOVATION EXPENDITURE
4 Innovation expenditure has proved very problematic to collect, primarily because it is not recorded as a line item in business accounting records, with many businesses not keeping records that allow for easy extraction of these data.
- In the 2003 Innovation Survey, an initial attempt to collect data relating to expenditure on the development and introduction of new goods, services, operational or organisational/managerial processes was undertaken. Beyond the overarching question of total expenditure on innovation, which was collected as a % of total expenditure, providers also were asked to provide actual dollar values for expenditure on a list of expenditure types for goods or service innovation, as well as totals for operational processes and organisational/managerial processes. Non-response to these questions was an issue and the resultant data quality was poor despite the significant and unsustainable quality assurance processes that were undertaken to obtain this information from providers. Lack of previously collected data also made estimation for non-response difficult.
- In 2005, a second attempt at collecting innovation expenditure saw significant rationalisation of the questions asked and the detail required. Providers could now respond in either dollar amounts or as a percentage of total expenditure. Innovation expenditure was only required for the three main types of innovation (goods or services, operational processes, organisational/managerial processes) and the recording of expenditure on specific innovation activities undertaken were now only required as an incidence question (i.e. tick boxes rather than dollar values). Resultant data quality improved slightly although as a consequence of the still burdensome quality assurance process it was found that the 2003 figures required revision. There was a problem of imprecise reporting and the difficulties of estimation for non-response remained.
- The 2006-07 Business Characteristics Survey was the first iteration of the survey with an innovation focus. The question on innovation expenditure was expanded to include expenditure on marketing methods and there were options to answer in dollars or %. Data quality and the high level of quality assurance remained significant problems but the presence of previously collected data improved estimation for non-response methodology. However, the level of estimation for non-response was still problematic and engendered further concerns regarding data quality. Due to reporting difficulties, a high level of non response and concerns over data quality, it was decided to cease collection of innovation expenditure data. As a result, the second iteration of the Business Characteristics Survey with an innovation focus (2008-09 Business Characteristics Survey) did not include questions specifically relating to expenditure on innovation, however the incidence question was retained.
A report published by the Department of Industry, Innovation, Science, Research and Tertiary Education, Firm-level Financial Measures of Innovation: An Exploratory Project to Derive Financial Measures of Innovation Activity from the Accounting Records of Model Australian Firms
, in 2012 highlighted the difficulties in estimating, separating, disaggregating and sourcing innovation expenditure.
Difficulties related to the collection of innovation expenditure are not unique to Australia. Similar problems have been identified by the European Union and the OECD. Through the OECD Working Party of National Experts on Science and Technology, the ABS will continue to collaborate internationally to develop and refine the range of approaches to measuring expenditure on innovation.
REINTRODUCTION OF INNOVATION EXPENDITURE
Notwithstanding the difficulties associated with collecting expenditure on innovation, there remains strong interest for these data. Given the significant issues which existed, an alternative method of collection was developed and implemented for the 2010-11 Business Characteristics Survey.
For the 2010-11 Business Characteristics Survey a decision was made that, rather than asking all providers to report a specific dollar amount (or percentage of total expenditure figure), a number of expenditure ranges would be provided, from which they could choose. The aim was to have a large number of providers using the smaller ranges but contributing only a small percentage of the total estimate, and the remaining small number of providers spending significant amounts on innovation activities still reporting an estimated dollar value.
The implementation of the refined innovation expenditure resulted in a vast improvement to the response rate and a reduction in the amount of quality assurance required. Please note that in cases where "No expenditure on any activities related to the development or introduction of new goods, services, processes or methods" was a legitimate response, these providers were automatically sequenced past the question on the amount of expenditure.
Caution should be exercised when using estimates of innovation expenditure as many businesses do not keep records that allow accurate estimates of these items to be provided.
Amount spent on innovation expenditure(a), by employment size, 2010 - 11
200 or more persons
|No expenditure ||% |
|$1 to less than $50,000 ||% |
|$50,000 to less than $100,000 ||% |
|$100,000 to less than $1,000,000 ||% |
|$1,000,000 to less than $5,000,000 ||% |
|$5,000,000 or more ||% |
|- nil or rounded to zero (including null cells) |
|(a) Proportions are of innovation-active businesses |
During the year ended 30 June 2011, 37% of innovation-active businesses in Australia reported no expenditure on innovation activities and a further 48% reported expenditure of $1 to less than $50,000.
For the three smaller employment size ranges, the most frequently reported amount of expenditure on innovation was in the $1 to less than $50,000 range, whereas for innovation-active businesses with 200 or more persons employed, $100,000 to less than $1 million was the most common response.
Innovation expenditure of $5 million or greater was reported by 6% of innovation-active businesses with 200 or more persons employed, compared with less than 1% of businesses in each of the smaller three employment size ranges.
In order to calculate a total innovation expenditure dollar value for Australian business, three statistical methods were utilised, each with different assumptions about how the true values may be distributed. An estimated total for innovation expenditure was derived by assigning a particular value to each innovation-active business that reported expenditure within the bounded ranges. These data, and the values of those businesses that reported actual dollar values were then weighted to derive an innovation expenditure total. Over one third of the estimated total expenditure was from units that reported actual dollar values of $5 million or greater.
The three methods were:
ESTIMATED VALUE OF INNOVATION EXPENDITURE
- innovation-active businesses in the bounded ranges being randomly assigned a value from within their selected innovation expenditure range.
- analysis of previously collected data suggested expenditure on innovation values were clustered around the one-third point of the range. Based on this, innovation-active businesses in the bounded ranges were assigned a value at the one-third point of their selected expenditure range.
- using previously collected data, an average value of innovation expenditure across innovation-active businesses was derived. This value was assigned to every innovation-active business that indicated expenditure on innovation, regardless of what bounded range they chose.
For Australian businesses which reported expenditure on innovation in 2010-11, estimates were calculated using the three methods outlined above. The estimated total expenditure derived from these methods ranged from $23 billion to $29 billion
Users are advised to exercise caution when using these data.