News & Views item - September 2011 |
Brian Martin: ERA: Adverse Consequences. (September 17, 2011)
The following opinion piece is reprinted from vol. 53, no. 2, 2011 of the Australian Universities Review.
Brian Martin*
ERA: adverse consequences
Excellence
in Research for Australia has a number of limitations: inputs are counted as
outputs, time is wasted, disciplinary research is favoured and public engagement
is discouraged. Most importantly, by focusing on measurement and emphasising
competition, ERA may actually undermine the cooperation and intrinsic motivation
that underpin research performance.
Excellence in Research for Australia (ERA) has the laudable aim of improving the quality of Australian research. Its approach is straightforward: measure the quality and quantity of research in different fields, with the prospect of funding attached to good outcomes, and this will stimulate better outcomes. However, this approach has many adverse consequences.
Misleading journal rankings
In the first ERA round, assessments of the quality of research teams were based, in part, on the quality of articles published in journals, assumed to correlate with the journal rankings of A*, A, B or C. On 30 May 2011, Senator Kim Carr announced that these rankings would be dropped and replaced by ‘journal quality profiles.’ How ERA panels will use these profiles is not clear. In any case, it is worth reviewing shortcomings of journal rankings.
On the surface, it seems sensible to judge the quality of research by the journals it is published in. However, trouble arises in the steps between journal rankings and the quality of research.
The first step is to establish a ranking for each journal, with expert panels relying on input from people in relevant disciplines. Inevitably, subjective factors are involved. For example, panel members might be inclined to rank highly a journal in which they had published and not so favourably inclined towards an unfamiliar journal.
Then there is the assumption of a unitary ranking of a journal. Many journals are less than regular in their treatment of submissions, due to invited articles (sometimes published without refereeing), guest editors and special issues filled with papers from conferences.
The reputation of a journal often depends on its impact on the field, which in turn is due to a small number of articles that are widely known and cited. Other articles in the journal may be unexceptional. Another problem is that impact factors can be manipulated (Arnold and Fowler 2011).
Even if journal rankings were accurate, this doesn’t translate into accurate quality ratings of articles, because journal standards only set minimums. An article’s quality does not go down just because it is submitted to a C or unranked journal rather than an A* journal. Judging quality by where an article appears is like judging a person’s wealth by their address. Moving to a lower-status suburb doesn’t reduce one’s income.
Simon Cooper and Anna Poletti (2011) argue that ERA’s journal-ranking process actually undermined the production of high quality research, by discouraging collegiality and international networking and by not recognising the way academics access materials digitally.
Many academics saw journal rankings as the most objectionable feature of ERA. Although dropping the rankings may give the impression that the rest of ERA is acceptable, there are plenty of other problems, some of them just as serious.
Inputs counted as outputs
In Australia, grant successes seem to be treated as a measure of research
success more than most other countries (Allen 2010). Peer review of grant
applications is one measure of quality, but grant monies themselves are inputs
to research, not outputs. ERA continues the emphasis on grants.
An alternative would be to look at output/input ratios. Imagine a scholar who
spends one-third of their time on research, valued at $30,000. A scholar who has
a $30,000 grant then should be expected to produce twice the outputs, or much
higher quality outputs. But this is not how the game is played. Big grants are
seen as more prestigious, even when there are no more outputs.
Time wasted
Preparing and assessing ERA submissions is time-intensive. It involves many of each university’s most productive researchers who are diverted into ERA administration rather than doing more of their own work.
Disciplines dominant
ERA categories are built primarily around disciplines. Interdisciplinary
researchers often publish in a range of journals. Their outputs are spread over
several different research codes, thus weakening a university’s claim to have
concentrations of excellent research. The result is that more narrowly
specialised research is encouraged at the expense of cross-disciplinary
innovation.
Many of today’s most pressing issues cut across traditional academic boundaries.
By sending a signal that interdisciplinary research is less valued, ERA
encourages a retreat from engaging with real-world problems.
Misleading narratives
ERA rewards the existence of groups of researchers in nominated fields. This
provides an incentive to create, on paper, artificial groupings of researchers
whose outputs collectively seem significant. Then, to fit ERA expectations, a
narrative needs to be composed about how the research of these groupings fits
together in a coherent package. Many of these narratives are largely fiction,
especially in fields like the humanities where researchers seldom work in teams.
The narratives serve the interests of the ARC. Groups are expected to show
high-quality outputs from ARC grants, so outputs are attributed to grant support
even when they might have happened anyway. Researchers without grants are
downgraded. The result is a self-fulfilling process: in essence, the ARC sets
the expectations for ERA reporting that shows how wonderfully effective ARC
funding is for research.
Many researchers give misleading pictures of their own research — on their CVs
and grant applications — for example by claiming more credit for their work than
deserved. ERA institutionalises incentives to create misleading narratives about
research groups and concentrations. Creative research managers might be tempted
to deceptively reclassify outputs, for example by dumping articles in
lower-status journals into a ‘reject’ category in order to boost rankings in
other categories.
Peers, not the public
Because the benchmark for research quality is what impresses other researchers,
there is an incentive to be more inward-looking. By default, applied research
and public engagement are discouraged (Brett 2011; Shergold 2011).
Public engagement — including writing articles for newspapers, blogs and other
online forums — requires a different style than the usual academic journal.
Value is placed on accessibility and relevance. Jargon is to be avoided. Public
engagement is a vital contribution to society, but is given little or no credit
in ERA.
Similarly, applied research useful to outside groups — government, industry or
community — receives less kudos than research pitched to peers. ERA gives no
formal attention to social impact, which might favour applied research.
Susceptibility to misuse
ERA is supposed to be used to measure the performance of institutions and
research groups, not individuals. However, it did not take long before
university managers began enforcing ERA-related measures on individual
academics, for example by rewarding those who published in A and A* journals or
brought in research grants. Academics are at risk of missing out on appointments
or promotions, or even losing their jobs, if their performance falls short in
ERA measures, no matter how outstanding they might be otherwise. The
psychological effect on those whose outputs are deemed irrelevant to ERA
performance can be severe. ERA may inspire better work by some but at the cost
of demoralisation of many others.
University managers could be blamed for inappropriate use of ERA measures. On the other hand, the conception of ERA itself is part of the problem, because it is so susceptible to abuse.
Competition
The ERA system is competitive, with every university and research unit trying to
do better than others. However, no one has presented evidence that competition
is the most effective way of boosting research quality and output. Alfie Kohn
(1986) in his classic book No Contest found competition is the guiding
philosophy in education and work despite a lack of supporting evidence.
Competition stimulates some undesirable behaviours. Universities, in their race
for status, put considerable effort into bidding for top performers, yet this
does not increase overall output in the system. In a highly competitive system,
researchers are more likely to hide or disguise their ideas to prevent others
from obtaining an advantage. Universities emphasise protecting intellectual
property rather than contributing to the public domain, even though few
universities make much money from intellectual property. Competition puts an
enormous strain on researchers and can lead to excessive and damaging work
practices, a type of self-exploitation (Redden 2008).
The alternative is cooperation, well known to be a stimulus for research in
collaborations and research teams. Cooperation in producing software has
generated some of the highest quality products in the world, such as the Linux
operating system. Online tools now enable easy collaboration across continents.
MIT has put its course materials on the web, leading a move towards sharing
rather than hoarding intellectual outputs.
Measurement not improvement
The massive effort involved in ERA ends up with assessments of research
excellence. That is all very well, but does measurement actually improve either
the quality or quantity of research? There is no evidence that it does.
The effort and attention given to ERA might be better spent on programmes directly designed to improve research. Collectively, the Australian academic community has immense knowledge and experience concerning research. Sharing this knowledge and experience could be promoted through training and mentoring schemes.
Research suggests that the key attribute of successful researchers is
persistence, not intelligence (Hermanowicz 2006). Stories of continued effort
despite failure would provide motivation for junior researchers. However, senior
researchers seldom tell the full story of their struggles — including rejections
of their work — as this might detract from their lustre (Hall 2002). In a more
cooperative, supportive research environment, such lessons would be easier to
provide.
Most experienced researchers are driven by intrinsic motivation, including intellectual challenge, fascination in developing new understandings, and satisfaction in working on something worthwhile. Intrinsic motivation can be undermined by offering external sticks and carrots, which is exactly what ERA does. Too many rules and external incentives can be counterproductive. Barry Schwartz and Kenneth Sharpe in their book Practical Wisdom describe how this can happen in law, education and medicine. They say ‘Rules are set up to establish and maintain high standards of performance, and to allow the lessons learned by some to be shared by all. But if they are too strict or too detailed or too numerous, they can be immobilizing, counterproductive, and even destructive.’ (Schwartz and Sharpe 2010: 255).
Schwartz and Sharpe (2010) say that people need opportunities to exercise discretion, balancing rules and circumstances to wisely help achieve the goals of the activity. Arguably, one of the reasons for the vocal opposition to journal rankings was that they removed discretion from academics for deciding where best to publish their research. Although journal rankings have been dropped, the basic incentive system remains. It would be paradoxical if ERA’s apparatus for measuring output and providing incentives for particular types of output actually sabotaged the very thing it is supposed to improve.
What to do?
Some academics have accepted ERA as a fact of life and seek to comply with
directives of university managers, for example to submit papers only to the most
prestigious journals. Others, though, think ERA is so flawed that they must
resist, either individually or collectively.
One option is to carry on with research as before, ignoring ERA imperatives, for
example submitting papers to the most appropriate journals, whatever their
academic status. This option is easiest for those who have opted out of the
struggle for promotions and status through the research game, or who are senior
enough so there is no need to impress others.
Another option is to refuse to participate in ERA exercises, for example
declining to lead panels, do peer assessments or contribute statements and
publication lists to ERA panel leaders. These forms of individual resistance
make a statement but have limited impact unless they become widespread.
A different sort of response is voicing dissent against ERA. This includes
careful deconstructions showing its damaging effects and vocal complaints to
anyone who will listen, including letters and articles in newspapers and blogs.
Academics know a lot of people from different walks of life, which means that
informal complaints to friends and critiques in professional forums will filter
through to politicians and other decision-makers. As well as rigorous critiques,
criticism of ERA can take the form of humour: creativity is needed to generate
the most powerful forms of satire. (I wrote this paragraph before journal
rankings were dropped from ERA, a change directly reflecting the power of
complaint).
Another response is to set up alternative systems for promoting research and
assessing performance, systems that address ERA’s shortcomings. This is a big
challenge but definitely worth the effort. Critique is all very well, but
critics need an answer to the question ‘If not ERA, then what?’
Conclusion
Some of ERA’s limitations are matters of design, for example counting grants as
outputs rather than inputs. Others are matters of orientation, notably the
emphasis on disciplinary research. Yet others are deeper: ERA assumes that
competition and measurement are worthwhile, though both are questionable.
ERA is all about promoting research, but curiously enough there is little
research available to justify the approaches adopted by ERA itself. It is not
evidence-based; indeed, there seems to have been no systematic comparison with
alternatives. Rather than the government imposing a competitive measurement
scheme, a different approach would be to open up space for diverse proposals to
improve research.
Acknowledgements
For useful comments and discussion, I thank John Braithwaite, Judith Brett, Don Eldridge, Anne-Wil Harzing, Tim Mazzarol, Anna Poletti, Guy Redden and others who prefer to remain unnamed.
*Brian Martin is
Professor of Social Sciences at the University of Wollongong, Australia.
__________________________
References
Allen, J. (2010). Down under exceptionalism.
University of Queensland Law Journal, 29(1), 143–154.
Arnold, D.N. & Fowler, K.K. (2011). Maths matters: nefarious numbers.
Gazette of the Australian Mathematical Society, 38(1), 9–16.
Brett, J. (2011). Results below par in social sciences. The Australian,
9 February, 36.
Cooper, S. & Poletti, A. (2011). The new ERA of journal ranking. Australian
Universities’ Review, 53(1), 57–65.
Hall, D.E. (2002). The Academic Self: An Owner’s Manual, Ohio State
University Press, Columbus, OH.
Hermanowicz, J.C. (2006). What does it take to be
successful? Science, Technology, & Human Values, 31, 135–152.
Kohn, A. (1986). No Contest: The Case Against Competition, Houghton Mifflin,
Boston, MA.
Redden, G. (2008). From RAE to ERA: research evaluation at work in the corporate
university. Australian Humanities Review, 45, 7–26.
Schwartz, B. & Sharpe, K. (2010). Practical Wisdom: The Right Way to Do the
Right Thing, Riverhead, New York.
Shergold, P. (2011). Seen but not heard. Australian Literary Review, 4
May, 3–4.