Australia allocates around A$9 billion a year of taxpayers’ money for research, but how do we know if that money is being spent wisely?
With the Australian Government threatening to reduce the amount of money allocated to research, it is time for researchers to take a more serious look at how to improve the research funding system.
Despite criticism of the Australian research funding system, there has been a lack of meaningful change in the system. The majority of researchers see the system as arcane, overly bureaucratic and wildly inefficient.
In many other areas of policy, including health, education and labour market policy, we subject government expenditure to rigorous, systematic evaluation. But evaluation of our research investments lags far behind our analysis of other public policy domains. This is a major obstacle for our future economic prosperity.
It’s time then to review the problem and examine how the research funding system could be improved for researchers and provide the taxpayer with more confidence about the way their money is being spent.
Aligning incentives
Innovative research is arduous, risky and long-term. It almost never occurs in a simple, linear fashion. But the incentives to do genuinely innovative research just aren’t in the system.
Instead we reward short-termism and incrementalism. Although the rules encourage “novel” research, it also has to be “achievable”.
Our system needs to recognise that the best ideas are not yet fully worked out and we need to provide incentives for researchers to do the best possible work they are capable of.
Evaluating performance
Researchers are under increasing pressure to publish their research in order to get tenure, promotion, and improve their institution’s Excellence in Research (ERA) ranking.
The upside of this is obvious: researchers work harder in order to demonstrate their suitability to be part of the profession.
The downside of this is also obvious: researchers tackle simpler, bite-sized research projects in order to get some papers published as quickly as possible.
Objectives
Research is all about the creation and transmission of new knowledge. But the system focuses on intermediate outcomes such as the number of publications and the quality of the journals in which they are published.
If you reward publications, you may get only publications. Focusing on these metrics leads to bean counting behaviour, not true scholarship.
What we want is a system that promotes and rewards the discovery of real solutions to real problems which may or may not be reflected in the number of A* publications a scholar produces in a five-year window.
Efficient peer review
Although peer review may be the best approach available for grant applications, the current system has a high degree of randomness in its outcomes.
It has been pointed out before that there is a high degree of randomness in research grant application outcomes: 29% of all applications were sometimes funded, while only 9% were always funded.
This suggests that the current system is doing no better than if we were to simply undertake some triage on all applications to determine the top and bottom x% (which is relatively easy to do) and then randomly allocate funds to the remainder.
Opportunity costs
The resources required to apply for grants, review/ assess grant applications and report on completed grants is huge.
This is an enormous burden on the system and one which takes researchers away from their main function: to do research!
Recent survey estimates indicates that this costs A$66 million in salary costs of the investigators.
Research lags
There is good evidence that the age at which great invention occurs is rising over time. In other words, it is getting harder to reach the scientific frontier.
But this change is not reflected in our funding system.
Apart from a few new schemes there is no recognition that the research process is being drawn out over a longer period of time.
Full cost of research
Many research grants don’t pay the cost of the chief investigator’s time, which disadvantages scholars in research-only environments. Funding agencies assume research will be subsidised from teaching.
If stand-alone (non-medical) research institutes are going to flourish in our university system, we need to recognise their needs and address this shortcoming in the funding rules.
Specialisation in research
There are many domains where equity is an important consideration in the allocation of resources, but research isn’t one of them.
Rather than giving research funds to all, we should recognise that frontier research is best achieved via economies of scale and scope.
This is not to argue that the existing large research-intensive universities should necessarily get larger and more research-intensive. There is a compelling argument to making some smaller, regional universities (such as in Far-North Queensland or Tasmania) centres of research excellence in fields such as mining or marine science.
What now?
If Australia is to continue with it high standards of living, it must have a vibrant innovation system with a strong scientific research base.
The current fiscal environment poses a major threat to this. We need to do a better job at explaining why the scientific research system is valuable. And this involves serious systematic evaluation of the incumbent system.
Unfortunately, we are still quite ignorant about what works – and what doesn’t work – when it comes to effective research funding mechanisms.
We could start to rectify this by committing to build the type of data infrastructure required for evaluation of the effectiveness of different policy interventions, as has been done in the US.
This type of linked dataset provides, for the first time, a unique holistic overview of the way research investments work their way through to the economy - including the students trained, the research support service jobs supported and the technologies transferred to industry - thereby providing a better basis for understanding how to most effectively allocate our precious research dollars.
To be sure, this is a necessary but not sufficient condition for solving the problem. We also need analysts prepared to pore over the data in order to provide evidence of sufficient quality to help guide the policymakers in the right direction.