Executive Summary
1. INTRODUCTION
The Cooperative Research Centres Association Inc. (CRCA) is pleased to provide feedback to the
Excellence in Research for Australia (ERA) Initiative’s Consultation Paper. This submission is focused on
providing some insights that will address the proposed approach “Indicators of Success in Applied
Research and Translation of Research Outcomes” as presented on Page 8 of that Consultation Paper, and
in particular the reference to
“…measures of excellence in applied research and translation on research outcomes…”.
The Consultation Paper notes that while there are a number of well established indicators available in
relation to research activity and intensity and research quality, developing indicators of excellent applied
research and translation of research outcomes is a more complex task.
The CRCA is acutely aware of the many difficulties associated with impact articulation through its work on
establishing the impacts of the Program and through its involvement with CRC Program application
processes and performance reporting systems. These perspectives and knowledge have been utilised in
the development of this submission.
2. OVERVIEW OF THE CRC ASSOCIATION
The CRCA is the representative body for the organisations operating within the Australian Government’s
Cooperative Research Centres (CRC) Program. The purpose of the CRCA is to promote science in
general, with a particular focus on the future growth of the CRC Program.
The CRCA is an independent body, funded by fees paid through voluntary membership. The CRCA
Constitution states that only bodies classified as “Cooperative Research Centres” by the Australian
Government are eligible to be members of the CRCA. The current membership comprises all 58 CRCs.
3. OVERVIEW OF THE CRC PROGRAM
The CRC Program was established in 1990 by the Hawke Government with the aim of changing the
culture of industry to shift from looking to specific short term problem solving research, to taking a longer
term, strategic approach to investment in research.1 Over the course of its 18 year existence the CRC
Program has met that aim and improved the effectiveness of Australia’s research effort through bringing
together researchers in the public and private sectors with the end users. The CRC Program links
researchers with industry and government with a focus towards research application. The close
interaction between researchers and the end users is the defining characteristic of the Program.
Moreover, it allows end users to help plan the direction of the research as well as to monitor its progress.
Since the commencement of the Program, there have been ten CRC selection rounds, resulting in the
establishment of 168 CRCs over the life of the Program that have operated across Manufacturing, ICT,
Mining & Energy, Agriculture & Rural Based Manufacturing, Environment, and Medical Science &
Technology sectors.
Reflecting its broad areas of activity, the CRC Program draws funding and in-kind resources from a wide
range of sources. Displayed below is the resourcing profile for CRCs in 2006-07.
1 Myers, Rupert. Changing Research Culture, Australia – 1995. Report of the CRC Programme Evaluation Steering Committee, Aust
Gov’t Publishing Service, Jul 1995.
Feedback on the Excellence in Research for Australia Initiative, Consultation Paper CRC Association
3.1 The Performance of the CRC Program
The conventional definition of a CRC is “a company formed through a collaboration of businesses and
researchers. This includes private sector organisations (both large and small enterprises), industry
associations, universities and government research agencies such as the Commonwealth Scientific and
Industrial Research Organisation (CSIRO), and other end users. This team of collaborators undertakes
research and development leading to utilitarian outcomes for public good that have positive social and
economic impacts.” 2 However this definition only tells a part of the story. As the Program has grown
and matured, further benefits have emerged, including:
- CRCs assemble multidisciplinary teams from across research providers to address end user driven research. They collaborate across all sectors (Industry, Academia, State Government, Consumers and Industry Associations) and create a critical mass in their field.
- CRCs provide companies, including multinationals, with the unique and attractive proposition of being able to deal through one organisation (the CRC) that can assemble the best teams in Australia to develop the technology that the company needs, managing the process professionally to deliverables and gearing it with funds from the Commonwealth and research providers who are sharing the risks, and the returns.
- CRCs are managed to deliver impacts not just papers, and are held to account to deliver. The stability of funding provides certainty for the research partners in particular and also for the end-user partners. The overall activities are actively managed by the CRC management team and Board to maximise the national benefits. This includes terminating, redirecting or accelerating projects in a way that is not part of the culture of most other programs.
- CRCs provide a mechanism for realising unanticipated commercial opportunities, i.e. in cases where technologies have applications beyond the interests of the commercial partners, the CRC can pursue these through the creation of spin off companies, licenses etc.
- CRCs play an important role in bridging the gap between discovery research funded by NHMRC and ARC grants and the requirements of industry for commercialisation-ready innovations.
- CRCs encourage innovation through their interaction and reach with SMEs (for example, the CRC for Spatial Information interacts directly with over 55 SMEs).
2 www.crc.gov.au
Feedback on the Excellence in Research for Australia Initiative, Consultation Paper CRC Association
- A CRC is neutral and un-aligned and so can provide a central focus from which grows collaboration.
- CRCs provide research management skills and discipline. This helps ensure the research is managed to a high standard.
- CRCs foster “hands-on” learning. Although they are heavily focused on postgraduate education, and thereby providing training for very highly skilled professionals, CRCs are involved, to differing extents, at all levels of the education and training system.
In the 2006 study on the economic impacts of the CRC Program commissioned by the Australian
Government3, fifty examples were included of economically quantifiable beneficial applications of CRC
research. In these solid, quantified examples, only the clearly measurable components of the outcomes
were included in the calculation of the net economic impact of the Program. Looking only at these clearly
quantifiable impacts, the study showed that as a result of each dollar invested in the CRC Program,
Australian Gross Domestic Product is cumulatively $1.16 higher than it would otherwise have been (had
the money instead been used for tax reductions) and Total Consumption is cumulatively $1.24 higher
than it would otherwise have been (had the money instead been used for tax reductions). It is important
to note that Gross Domestic Product and Total Consumption are two critical indicators of the economic
welfare of the Australian community rather than being measures of the private returns to CRC
participants.
Since its inception the CRC Program has been regularly and meticulously reviewed. The success of the
Program has been recognised not only within Australia but also internationally as the CRC Program has
been researched, emulated and even copied by a number of other nations.
4. ISSUES IN MEASURING THE IMPACT OF APPLIED RESEARCH OUTCOMES
4.1 The Challenges
A threshold issue in the measurement of excellence in applied research and the translation of research
outcomes is whether the assessment will focus on assessing the level of ‘activity’ relating to applied
research and research translation or on assessing the ‘impacts’ of applied research and research
translation.
4.1.1 Measuring ‘Impact’ rather than merely ‘Activity’
Measuring ‘activity’ can be done relatively cheaply and with a heavy reliance on quantitative
performance metrics. If the assessment is focused on process, a range of activity measures can
be used to rate the performance of different research groups. Such measures could include
things such as:
• presentations to end users of research
• exhibitions and/or performances
• commercialisation activity measures such as patenting, licensing of IP and formation
of spin-off companies
• number of research contracts entered into with end users of research.
While these indicators may provide evidence of the extent to which a research group is engaging
in activities designed to promote the beneficial application of research, they do not, however,
3 https://www.crc.gov.au/HTMLDocuments/Documents/PDF/CRC_Economic_Impact_Study_Final_121006.pdf
Feedback on the Excellence in Research for Australia Initiative, Consultation Paper CRC Association
provide evidence of the actual ‘value’ to society that is resulting from the applied research or
research translation activity.
Measuring ‘impacts’, by contrast, is a more complex and costly undertaking and will necessarily
require a greater use of expert judgement in performance assessment.
4.1.2 Measuring the flow of the ‘Impact’
The key challenge in assessing the impact of research is to convey the story of how a research
group’s inputs have flowed through to delivery of an end impact for the community. Establishing
a clear, verifiable, quantifiable and causal link between particular funded R&D and specific final
impacts for the community is an inherently difficult process due to issues such as:
• time-lags involved in the translation of research outputs into final outcomes for
society may be considerable. It often takes time for the true quality and value of
research to become apparent
• difficulty in attributing outcome ‘effects’ to particular research ‘causes’. The quality of
research, the extent to which the knowledge is diffused to those in a position to use
the knowledge to generate impacts, and the ability of research users to extract full
value from it will all influence the final impact of research
• disentangling the contribution of research performed in Australia from research
performed elsewhere when assessing the impact of research
• the fact that outcomes often require many non-research inputs for the ‘value’ to be
realised from the application of research
• the lack of a contractual ‘paper-trail’ in cases where public domain knowledge (such
as that in academic publications) resulting from R&D may be taken up by many end
users
• some research outcomes with commercial potential are kept secret and are hence
both unidentifiable and unmeasurable
• difficulties in attaching a ‘value’ to outcomes in the environmental, health and social
spheres where the outcomes are ends in themselves rather than means to deliver an
economically quantifiable value.
4.2 The Dangers of utilising Proxy Measures
There is always a temptation to base evaluations of research impact on what can be easily measured
rather than on what is important but difficult to measure.
Unfortunately, very few informative proxy indicators are available for either measuring or projecting long
term research impacts. For instance, measures such as patents held, licences executed and spin-off
companies formed are sometimes used as proxies for the likely future economic ‘impact’ of research or
the ‘excellence’ of research translation performance. The use of such measures, or even the use of more
sophisticated measures for commercialisation performance such as the turnover of spin-off companies or
the value of licensing revenue, while perhaps of some predictive value for research that is highly targeted
at commercial outcomes, are however inappropriate as measures of the likely eventual economic, social
and environmental impact of research.
4.2.1 The skewed nature of Volume Measures
Volume measures of research usage, such as the number of patents, licenses or spin-out
companies formed based on research, are not even likely to be accurate indicators of the likely
eventual narrowly economic ‘value’ of the research usage. This is due to the skewed nature of
Feedback on the Excellence in Research for Australia Initiative, Consultation Paper CRC Association
value creation from research. A relatively small number of patents, licenses and spin-out
companies account for a disproportionately high amount of the total economic value from such
commercialisation activities, so simply counting these metrics does not provide a sound indicator
of the future value that will result from them.
An example of the skewed nature of the ‘value’ from commercialisation activities is;sans-serif,pspan style=psociety may be considerable. It often takes time for the true quality and value of provided by
the distribution of royalty and fee income of the University of California (UC). In the United
States, UC is the largest generator of revenue from the licensing of technology. In financial year
2007, UC generated US$97.6 million in royalty and fee income derived from 1,592 inventions. Of
this income, the top 25 commercialised inventions generated US$73.9 million, or 75.7 percent of
total income. The distribution of income amongst the top 25 revenue generating inventions is
further skewed toward the highest earning inventions. In 2007, the top five inventions generated
US$ 47.7 million, or 48.9 percent of total income.4
Another example of the skewed value from commercialisation activity is demonstrated in a 2003
report commissioned by the Australian Institute for Commercialisation which found that only
around one out of every one-hundred high-growth technology-based companies started in
Australia has reached a revenue level of over $50 million per annum.5
A real danger in selecting such volume metrics as proxies for the impact of research translation is
that these will be the metrics that research groups will focus on delivering – even if that is at the
expense of generating actual economic, social or environmental value from the research. For
example, in some cases patenting a new technology may be a less appropriate mechanism for
protecting the intellectual property than keeping it commercial-in-confidence would be, but if
patents are counted and ‘rewarded’ then the tendency may be to patent as many discoveries as
possible. This is just one of many potential perverse incentives that use of inappropriate
intermediate outcome measures can give rise to. Similarly, sometimes the greatest community
value from research will be delivered from wide dissemination of the research to end users to
encourage rapid uptake, rather than from protecting the intellectual property and attempting to
capture commercial returns from it.
4.3 The potential of “Repeat Business” as a single measure of Benefit Delivery
Perhaps the best (although still limited) single measure that a research group can provide as a proxy
indicator for delivery of benefits is evidence of ‘repeat business’. If an end user of research (whether
government or industry) provides ‘repeat business’ to a research group, funding multiple projects over a
period of time or repeatedly using the outputs of the research group’s activities, this is evidence that the
end user believes that the usage of the research group’s activities is delivering a beneficial end impact. If
they did not believe this to be the case they would not provide the ‘repeat business’. In this case, even if
the impact can’t be specifically articulated, a claim can be made that impact is clearly occurring.
The concept of ‘revealed preference’ is also relevant here and suggests that consideration of repeat
business can also allow judgements to be made on the merits of one research group against another
operating in a similar field of research. Revealed preference refers to a situation where people’s
judgement on the merits of one thing against another can be observed through the people’s voluntary
choices and behaviour. If people are free to act in a range of ways, and assuming that people are in
general rational and aware of their own best interests, when people choose one thing over another, this
can be seen as evidence that they believe the thing they have chosen is most likely to best serve their
interests. Hence, if a research user is free to purchase services or use outputs from a range of research
providers in a given field, if the research user regularly chooses to purchase services or use outputs from
a particular research group this indicates a belief that this delivers them a better result than the available
alternative choices.
4 University of California, Technology Transfer Annual Report 2007.
5 Allen Consulting Group, 2003, The economic impact of the commercialisation of publicly funded R&D in Australia, report for AIC.
Feedback on the Excellence in Research for Australia Initiative, Consultation Paper CRC Association
Despite providing some useful evidence of impact performance, it should be noted that the ‘repeat
business’ indicator is not able to provide conclusive evidence of the actual ‘value’ that users place on
research or the extent to which they value the research of one group over other groups.
5. A WAY FORWARD
The lack of simple proxies for research impact (which was a clear finding of the extensive stakeholder
consultations that occurred during the development of the now abandoned Research Quality Framework
proposed by the previous Government) has important implications for how research impact can be
assessed within the ERA initiative.
Given the lack of useful data proxies for past or potential future research impact, a system that is
attempting to assess the impact of applied research and research translation, as opposed to simply
measuring activity, will need to be flexible and rely upon what are in effect ‘evidence based stories’ of the
impacts generated by a research group. This would entail a greater use of expert review in the
assessment of excellent applied research and translation of research outcomes within the ERA initiative
than is required in either the assessment of research activity and intensity or research quality.
5.1 The CRCA Guidebook
The CRCA has recently developed a Guidebook6 for its members in how to articulate the impacts that
their work generates. The key elements of this framework, and the guidance provided to CRCs in using
the framework, are outlined below. If it is desired with the ERA initiative to assess the impacts of applied
research and research translation, this approach may provide a useful framework for broader application
in the ERA initiative. Universities could each present a portfolio of performance for each field of research
in the area of excellent applied research and translation of research outcomes that articulates the impact
of their research.
5.1.1 An Input-to-Impact Approach to Articulation of Impacts of Applied Research
and Translation of Research Outcomes
Research inputs, activities and outputs can often be logically connected to usage levels and final
impacts in the community. Once the logic of the input to impact chain has been articulated (and
it must be recognised that this is often not a simple linear process and that there may be several
research feedback loops before an end impact is generated – not necessarily by the same
research groups that undertook the initial activities), the extent to which the impact statement
will be compelling depends upon the quality of evidence supporting claims relating to each link in
the chain.
6 CRCA, 2007, Impact monitoring and evaluation framework: background and assessment approaches
http://crca.asn.au/sites/default/files/CRCA_Economic_Impact_Guide_1.pdf
Feedback on the Excellence in Research for Australia Initiative, Consultation Paper CRC Association
Figure 1: ‘INPUT TO IMPACT’ TRACKING MODEL
e.g. output gain,
productivity
improvement, water
saving, health
improvement, higher
quality workforce,
etc.
e.g. training
package accessed
by users, process
change
implemented by
companies,
licensing of IP, etc
e.g. publications,
training package,
post-grad student
completions,
prototype, patents,
etc
e.g. research
project, training
course,
communications
program, etc
e.g. cash, in-kind
support, FTE staff
years, background
IP, equipment
access, etc
Key Inputs Key Activities Key Outputs Key Usage Key Impacts
At each of these stages, the relevant performance metrics will be different; some of the key
potential indicators at each stage are listed above. It is also likely that different measures will be
more or less appropriate to each particular research group being considered.
• Inputs: Measuring inputs will be essential in determining how much net benefit arises
from research. Input measures are likely to include information about the resources
committed to the project, including the people, infrastructure or facilities, financing,
and IP.
• Activity: Measuring performance in research activities is likely to involve data
concerning projects, stakeholder engagement, monitoring and evaluation, training
activities and so on. Collecting data about activities will require project managers to
maintain information about appropriate performance measures. With research
projects the key measures of activity will typically be the extent to which project
milestones have been delivered upon.
• Outputs: The appropriate output measure will depend on the nature of the research
question and activity. Output types may include things such as publications, patents,
trials, products, or course completions.
• Usage: Measuring the ways external users apply research outputs may be more
complex than the previous stages, which involve intra-organisation data. A key point
at this stage is that the metrics of usage, and the data required, should be agreed
upon as early as possible, so that the external users are able to keep accurate
records of how research outputs are actually applied. If there has been proper
planning, it is likely to be easier to gather consistent and relevant data from external
users of the research or project outputs. Collecting information about how external
groups have adapted, applied and adopted outputs is likely to require a balance
between gathering enough information to be representative, while minimising the
coordination costs involved. One important factor will be to capture the significant
information – from the large, important or influential users – without expending
unnecessary energy on less useful sources. If the model has been used to specify the
intended impact channels, usage can be measured against this. In addition, it will be
important to gather at least some data about what other resources external users
have incorporated into their use of the project outputs. This is so that there can be a
credible attribution of the outcome to the project or program and so that benefits
can be compared to total costs.
• Impact: Quantifying the final impact of research is necessarily the most uncertain of
the stages. Impact types, which may include productivity gains, industry
development, environmental, health and social benefits, and value from broadening
options, are not always easy to quantify. Many of these are highly contingent on
future contexts. As well, in a real world situation, few of the outcomes likely to be
considered – productivity gains, population health improvements, environmental
benefits, etc – have only one cause. Even if it is possible to identify the other factors
or resources involved in a particular outcome, it is unlikely that exact quantification
will be viable. However, acknowledgement of the issue that other factors will likely
contribute to an outcome, and a transparent attempt to account for it, is important in
any impact evaluation.
At each of these inputs to impact stages, the focus of measurement will be different. This means
that the types of information required are different; it also suggests that the process of collecting
data, and using it in an estimation of effect, will not be the same along the whole input to impact
chain.
In selection of performance measures/indicators, it is important to consider issues of:
• reliability, validity and credibility of the indicators
• cost-effectiveness in terms of cost to collect and process.
It is generally preferable that SMART (Specific, Measurable, Achievable, Relevant and Timebound)
indicators are used for evaluation processes.
To ensure that the right performance measures and indicators are used, some useful checks that
need to be made are:
• linking them to the logic model and the likely time frame for results
• ensuring there are not too many indicators (so that too much time is spent collecting
and not enough time on analysing and using) or too few (so that important aspects
of performance are not included)
• assessing the likely credibility of the measures and indicators to the intended users –
including risks to its accuracy due to differing definitions, difficulties in collecting it, or
disincentives to be accurate.
One fortunate consequence of the skewed nature of impacts from research – with a relatively
small number of research applications accounting for a disproportionate share of the total
impacts of research – is that performance portfolios for any given research field will not need to
include a large number of impacts in order to provide a reasonable sense of the overall impact
that a research group has generated. The focus in the evidence portfolios should be on clearly
articulating the major impacts that have resulted from applied research and research translation
activities rather than on exhaustively documenting all instances of impact.
6. CONCLUSION / RECOMMENDATION
The CRCA believes that it is essential during the establishment of the ERA that a clear decision is taken as
to whether assessment of excellence in applied research and the translation of research outcomes will
focus on assessing the level of ‘activity’ relating to applied research and research translation or on
assessing the ‘impacts’ of applied research and research translation.
If the focus of assessment is to be on ‘impacts’ rather than only ‘activity’, the assessment should be
primarily based on the assessment of evidence portfolios by expert reviewers rather than on the
use of volume based intermediate metrics of activity.
Use of inappropriate metrics would open up the prospect of both misrepresenting the impacts of applied
research and the translation of research outcomes and of encouraging unintended and undesirable
consequences as Universities would ‘manage to the metrics’ rather than manage towards delivery desired
final outcomes for the community.
The CRCA is pleased to offer its Guidebook entitled “Impact Monitoring and Evaluation Framework:
Background and Assessment Approaches” as a resource that the Australian Research Council may utilise
Feedback on the Excellence in Research for Australia Initiative, Consultation Paper CRC Association
in the development of the ERA. A full copy may be found on the CRCA website at :
http://crca.asn.au/sites/default/files/CRCA_Economic_Impact_Guide_1.pdf
Finally, the CRCA believes there is opportunity for the ERA to act as a measure of potential impact by
facilitating the development of strong research groups. The ERA could explicitly reward universities for
forming collaborative research teams that formally involve researchers from the private sector and from
the public sector and other universities. In this way, Australian researchers will be overtly encouraged to
form the optimum combinations of research teams drawn from wherever necessary and therein have the
optimum capacity for ultimate high quality impact.