Sydney Law Review
Government-funded Health Research Contracts in Australia: A Critical Assessment of Transparency
Nola M Ries and Kypros Kypri†
Australian governments claim to be committed to improving transparency and democratic accountability. Yet they are increasingly contracting out research to external consultants, ‘think tanks’ and universities and the contractual relationships formed can, in fact, promote secrecy and undermine the goals of transparency and public scrutiny of government actions. This article reports on a first-in-kind study of research contracts between Commonwealth and New South Wales Government entities and external providers. Our analysis reveals that ‘control clauses’ are prevalent: contractually, governments can insist on the rights to determine whether, when and how the results of research are publicly disseminated, to claim intellectual property rights over work produced, and to terminate contractual relationships at will and without cause. These findings have troubling implications for government openness and accountability, for academic freedom when university researchers face restrictions on publication, and for evidence-informed policymaking. We propose solutions for proactive information disclosure to ensure that government transparency promises are realised in practice. We advocate for comprehensive public release of contract details and urge governments to publish the findings of contract research in an online repository.
Information is a valuable resource. The right information at the right time can expand knowledge, enable innovation, boost productivity, and even save lives. Unlike other valuable resources information is not diminished by use. Indeed, the value of information can be enhanced when it is openly accessible and reused frequently.
Missing or buried research represents a democratic deficit. It feeds cynicism and has no place in any honest and well-conducted administration.
It is a sad day when Australian researchers talk about suppression of scientific studies and their findings.
A growing international movement seeks to advance openness and transparency in government activities. Globally, the Open Government Partnership program, of which Australia is a member, wants governments to become ‘more transparent, more accountable, and more responsive to their own citizens, with the ultimate goal of improving the quality of governance, as well as the quality of services that citizens receive’. In its first Open Government National Action Plan, released in December 2016, the Government of Australia promised ‘ambitious action’ to improve: transparency, accountability and public sector integrity, especially in contracting practices; access to government information, including more open datasets; and meaningful public participation in policy development and delivery of services. In March 2017, the Government announced its adoption of the International Open Data Charter, a further commitment to improving openness and citizen engagement.
While making these commitments to transparency and access to public sector information, governments are increasingly contracting out the provision of research to external entities, including ‘think tanks’, private sector consultants and university-based researchers. This trend to external procurement is rooted in the ‘New Public Management’ movement, which champions the values of competition and efficiency and aims to reduce the size of government through privatisation and the outsourcing of activities previously seen as the natural domain of government.
The processes by which governments outsource research and advisory services can, in practice, promote secrecy and undermine the goals of transparency and public scrutiny of government actions. A key mechanism of control is the contract that establishes the terms of the relationship between the government purchaser and the external provider of research. Contractually, governments can insist on the rights to control whether, when and how the results of research are publicly disseminated, to claim intellectual property (‘IP’) rights over work produced, and to terminate contractual relationships at will and without cause. Worryingly, these types of contractual clauses — described here as ‘control clauses’ — can limit the degree to which government-purchased research is exposed to external scrutiny and, in turn, diminish the rigour and quality of research. It can also contribute to duplication of effort and wasteful spending if bureaucrats, unaware of previous contract research that never saw the light of day, go to market again to purchase similar work.
This article reports on our investigation into control clauses in health-related research contracts between Commonwealth and New South Wales (‘NSW’) government entities and external providers. We selected health as the focus of this investigation because health expenditure comprises a substantial portion of government spending — over $180 billion in 2016/17 — and openness about the research informing health policies and programs is vital. Our research contributes an applied, ‘law-in-action’ perspective to transparency discourse. We adopt a broad conceptualisation of transparency as the availability of information about an entity that allows external parties to monitor its internal actvities or performance. Information availability can be achieved through active transparency, which requires governments to proactively release information to the public, or through passive means, such as freedom-of-information (‘FOI’) laws that allow people to request access to government-held records.
This article has four main parts. First, we set the context by summarising the origins of the ‘open government’ movement, briefly describing three main government funding sources for health research, and highlighting previous reports of government control over contract research. Second, we present our investigation of Commonwealth and NSW contracts, which reveals that the use of control clauses is routine. Third, we make recommendations to improve government transparency. We advocate for the full disclosure of contracts on public registers and urge governments, in consultation with key stakeholders, to review and update standard contractual terms dealing with publication, IP and termination, to ensure they advance transparency goals. In doing so, the unique role of universities in society must be respected. We also call on governments to publish reports of contract research. Finally, we conclude with an agenda for further research.
Australia’s 2016 Open Government Plan has its origins in the Government 2.0 movement instigated nearly a decade ago. In 2009, a taskforce was established to advise the Australian Government on matters that included improving transparency, enhancing the accessibility of government information, and establishing a ‘pro-disclosure culture.’ The Taskforce’s central recommendation was that the Australian Government should make a declaration of open government. Further, to achieve greater openness, the Government should adopt a default position that public sector information will be freely accessible and should implement a consistent framework to support publication of government information. The Government followed through with a Declaration of Open Government in mid-2010 and established the Office of the Australian Information Commissioner. The Commissioner subsequently promulgated the ‘Principles on Open Public Sector Information’, asserting the primary principle that government-held information should be publicly available unless there is a legal reason for confidentiality: ‘Information held by Australian Government agencies is a valuable national resource. If there is no legal need to protect the information it should be open to public access. Information publication enhances public access.’
Pledges to improve government openness coincide with the evidence-based policymaking movement, which is championed within the public sector and by external advisors. In a critical analysis of evidence-based policymaking, Head observed that ‘it is axiomatic that reliable information and expert knowledge are integral to sound processes for formulating and implementing policy’. Governments should not only be rigorous in their decision-making, but also be open about the sources and types of evidence used to inform their legislation, policies and practices. Policymaking is not simply a technocratic process of translating evidence into policy. It is, by definition, political and influenced by ideology. This fact underscores the need for governments to be transparent about the evidence that has informed policy decisions. As Hawkins and Parkhurst argue, ‘[t]he policies [governments] adopt, the evidence they marshal to support their decisions and the interpretation of that evidence should be transparent, and thus open to contestation by policy actors and citizens.’ Governments should also be open about the findings of program and policy evaluations so that members of the public can learn whether or not government initiatives have met their stated aims, at what costs, and whether they have had unexpected impacts.
In Australia, government-funded health research can be classified into three types. First, government agencies use internal research conducted by public servants. This capacity has diminished substantially in recent decades and research work is increasingly outsourced. Second, governments invest public funds through agencies whose primary purpose is to support research, principally the National Health and Medical Research Council (‘NHMRC’) and the Australian Research Council (‘ARC’). Such research mostly occurs in universities and medical research institutes, to generate new knowledge from basic science, clinical, public health, and health services research. In this category, researchers propose projects that go through competitive peer review processes. Government is at arms-length from the research, typically not seeking to control its conduct or the reporting of findings, and making no claim on the IP generated. Moreover, researchers conduct this work within universities that have statutory obligations to undertake research in the public interest, advance knowledge, and promote critical and free inquiry.
The third category of government-funded health research, and the focus of our study, involves government agencies issuing tenders to purchase research from external providers in order to answer questions relating to their policy and program development, implementation or evaluation. Successful bidders may be universities, medical research institutes and private or quasi-private bodies, including research organisations that are spin-offs from universities and incorporated business entities. The extent of this type of research expenditure is unknown, but likely to be substantial given the large numbers of calls for tenders announced on the Commonwealth AusTender site and each of its state counterparts. In our pilot research, we identified hundreds of projects advertised on these sites with values ranging from tens to hundreds of thousands of dollars per project. The nature of the relationship formed between government and provider is usually different to that in researcher-initiated projects and the mechanisms for promoting research integrity and transparency are far less straightforward. In contract research, the government purchaser may be involved in the framing of research questions, specification of study design, choice of methods, interpretation of results, preparation of reports, and/or the dissemination of findings. The government might require researchers to turn over IP to government and publish results only with its approval, and work may occur under threat that the funder can terminate contracts at will.
Political scientists and policy scholars have, for some time, studied who is supplying research and advice to governments and the scope of government control over the internal and external sources of expertise. They have developed the concept of a policy advisory system to refer to ‘the interlocking sets of actors and organizations, with a unique configuration in each sector and jurisdiction, that provides recommendations for action to policy-makers’. For example, a policy analyst employed in the public service performs research and provides briefings and advice, but must do so within the boundaries of her or his employment relationship and assigned work duties. Entities established by Acts of Parliament can have statutorily-defined advisory roles and legislative protections can minimise the degree of control exerted by the government. External to government, a range of individuals and organisations can provide input and advice to government, with government exerting no control over some and potentially high levels of control over others. For instance, governments do not control the advice offered by independently funded interest groups that make submissions as part of law reform inquiries. In contrast, consultants and researchers retained by government are controlled by the terms and conditions of their contractual arrangements.
Australian social scientists have made the case for an urgent need to investigate how governments procure and use research to inform policy, enabling a view of ‘the inner workings of the policy “black box”’. No published research examines Australian governments’ use of control clauses in research purchased through tenders — a gap addressed by our research. However, a previous survey and a case study reveal evidence of government control over the public dissemination of contract research. In a 2006 survey of approximately 300 public health academics in Australia, Yazahmeidi and Holman found that 21% of the respondents reported having personally experienced a ‘suppression event’ in the preceding 5.5 years. A suppression event is where a government funder had invoked a clause in the contract to ‘sanitize’, ‘delay’ or ‘prohibit’ the publication of findings. According to the survey respondents, their work was targeted because it ‘drew attention to failings in health services (48%), the health status of a vulnerable group (26%), or pointed to a harm in the environment (11%)’. These findings backed perceptions of a growing tendency among government agencies to control the conduct and reporting of policy-relevant research. A case study of an Australian contract negotiation in 2012 raised concerns about control clauses in health research contracts, including provisions that vest IP rights in the Government and allow the Government to limit publication of results and terminate contracts without cause.
Other sources reveal related problems that undermine the goals of transparency and open government. In Australia, a coalition of major research-intensive universities, the Group of Eight (‘Go8’), was candid in its concerns about research contracts in a 2008 submission to the Review of the National Innovation System:
the standard terms sought by the Commonwealth when entering into research agreements with universities misunderstand the role and nature of universities; are unnecessarily onerous and impractical; often cause delay and uncertainty due to their complexity; and serve to stifle knowledge transfer and innovation by restricting the capacity of universities to disseminate the results of the sponsored research for public benefit.
Successive Australian National Audit Office (‘ANAO’) reviews find misuses of confidentiality clauses in a range of government contracts notwithstanding a 2001 Senate Order that confidentiality provisions should not restrict public access to contract information unless there is a strong justification for non-disclosure. A recent audit ‘found that processes to capture information about basic contract details and the reporting of confidentiality provisions were inadequate. Only 19 per cent of contracts sampled were accurately reported in AusTender.’
Internationally, a British charity, Sense about Science, is doing groundbreaking work investigating government practices in delaying and withholding commissioned and internal research findings. The United Kingdom (‘UK’) is ahead of Australia in its open government initiatives — it is already into a third phase of its National Action Plan on Open Government and has been ranked as a global leader in government transparency. Several high-level protocols from the UK Government Office for Science and the Government Economics and Social Research Team require ‘prompt and complete publication of research conducted or commissioned by’ government departments. Yet, the Sedley Report arising from a 2016 inquiry, sponsored by Sense about Science and led by a former Court of Appeal judge, revealed that governmental bodies routinely conceal externally commissioned research from public view. Suppression is deliberate in some cases where research results are considered politically inconvenient or embarrassing. In other cases, benign neglect seems to be the problem, where departments lack processes to keep track of research they have commissioned and therefore fail to make it public. The Sedley Report endorsed the principle that ‘[p]rompt and full publication of government research is a matter not of contract but of public duty. While research contracts will necessarily vary in their provisions, all contracts should spell out this obligation of principle’, which would advance government commitments to openness and transparency.
Our pilot study, undertaken in 2015–16, examined a sample of Commonwealth and NSW contracts. We sought to identify the presence and content of three types of contractual control clauses; namely, clauses dealing with: (a) contract termination rights; (b) ownership of IP; and (c) control of publication. Through an analysis of draft and executed contracts, we sought to determine whether control clauses are modified or removed in the negotiation process between the government and the successful bidder.
We accessed the Commonwealth tender website (https://www.tenders.gov.au) and the NSW tender website (https://tenders.nsw.gov.au/) between March and November 2015. These websites list project descriptions and typically provide a draft, pro forma contract to indicate the expected terms of the relationship between the government and a successful bidder. We searched for tenders for research and evaluation projects, focused principally on health-related topics. Due to the volume of tenders, we limited the Commonwealth search to tenders issued through the Department of Health. For NSW, we included the Ministries of Health, Justice, Education, Family and Community Services, as well as WorkCover and the Motor Accidents Authority.
A research assistant with an undergraduate law degree produced an initial list of potentially relevant tenders produced using broad search terms of ‘research’ and ‘evaluation’ (‘health’ was added as a search term for the NSW tender website). The lead investigators reviewed the list and selected tenders calling for research to inform proposed government initiatives or to evaluate existing health-related programs. For example, the Commonwealth tenders included a call for an evaluation of a national primary health care strategy, as well as a tender for a research study to investigate how flavourings in tobacco products affect smoking behaviours, both of which we included. We excluded tenders that sought contractors to design research instruments (for example, a survey), carry out health surveillance or monitoring programs, or deliver health programs. On the tender websites we were able to search for awarded contracts to identify the successful bidder. We sought to maximise the heterogeneity in the funding agreements we studied in order to permit a variety of comparisons. Accordingly, we included contracts with different types of research providers, including universities, university-owned research entities, and private sector entities. We avoided selection bias by remaining blind to the content of the draft funding agreement when choosing the cases to study. We hypothesised that university-based researchers might be more likely to negotiate to modify or eliminate contractual clauses that restrict the ability to publish their work. Such control clauses might be of less concern to private sector providers.
We accessed draft contracts that were available on the tender websites and made formal requests to the Australian Government Department of Health and NSW Government entities for executed contracts. These requests were made under the Freedom of Information Act 1982 (Cth) (‘FOI Act’) and the Government Information (Public Access) Act 2009 (NSW) (‘GIPA Act’). In requesting the contracts, we explained we were interested in the specific types of clauses noted above. We made it clear that we were not seeking access to details protected under the FOI Act or the GIPA Act, including information that could harm the commercial interests of a service provider (such as details of a novel research method) or the privacy of individual researchers.
We had a modest budget of $1200 from a University of Newcastle internal grant to cover fees, which allowed us to obtain contract documents for 21 tenders issued by NSW government bodies and six tenders from the Australian Government Department of Health. One NSW government body disclosed draft contract pro formas to us, which were already publicly available on the tender website, but refused to disclose the final contracts on the grounds that ‘these matters are still ongoing and the funding agreements have not yet been executed or grants have been awarded but the projects are not yet completed’. We were informed that the ‘final executed agreements rarely vary from the standard clauses in the [pro forma contracts]’. Two NSW tenders resulted in contracts with several different providers. In total, our contract analysis covers 35 projects.
A major finding of our investigation is that one or more types of control clause are present in all the draft and executed contracts we accessed. This reveals that the contractual relationships between government purchasers and external research providers are weighted in favour of government control and permit conduct that is contrary to transparency commitments. A comprehensive summary of the contract documents we analysed and the sources for all quotations reproduced from contracts is on file with the authors and available on request.
1 Termination for Convenience
From a transparency perspective, termination-for-convenience clauses are problematic as they give broad power to the government purchaser to end a contract simply on notice and not for cause. For example, a government agency may want the right to end the contract in case the timing of the planned study becomes politically inconvenient or because embarrassing or unfavourable research results start to emerge. A new government may also choose to terminate contracts awarded by its predecessor.
In the contracts we reviewed, clauses that allow the government to terminate a contract without cause or explanation and simply by giving written notice are most common and rarely negotiated out. All draft and final Commonwealth contracts contain clauses giving the Government ‘unfettered discretion’ to terminate the contract for convenience, which means that it expressly rejects any requirement to exercise the termination power in good faith.
Of all the contracts we obtained, just three NSW contracts do not have termination-for-convenience clauses and only permit termination for cause. In two cases, a draft contract gives the Government sole discretion to terminate for convenience, however the final contract gives the providers the same power of termination. This implies that the providers negotiated the right to end the contract simply by giving a specific period of notice. However, this does nothing to minimise the risk of the purchaser ending the contract at its discretion. The financial benefit gained from the contract means that it would rarely be convenient for the provider to act on a right to end the contract.
2 Intellectual Property
Contractual terms governing the ownership and use of IP have transparency implications as they can either restrict or promote the public dissemination of research outcomes. The contracts we analysed typically had clauses dealing with the existing IP that parties bring to the contract (which remains their own) and the new IP created during the course of conducting the research or evaluation study.
All the Commonwealth draft contracts vested IP in the ‘contract materials’ in the Government, which encompasses all materials created for the purposes of the contract, including final reports. By owning the IP, the Government has the right to control the public release of contract deliverables. As a result, the pro formas do not have clauses concerning publication. However, as discussed below, all the final contracts incorporated specific terms on public dissemination through journal articles, conference presentations and other means. In one final contract, the university provider negotiated a licence to use the contract material, however the contractual terms required that draft and final reports be reviewed by the Government, thus providing an opportunity to the purchaser to shape how the findings are presented.
In NSW, a majority of the draft contracts stated that the IP in materials produced from the contract vested in the Government and this ownership is rarely modified in the final contracts. Several draft contracts included a licensing provision that allows the provider to use the IP in the materials for its non-commercial teaching and research purposes; this type of licence is sometimes negotiated into the final contract. For three projects, the providers negotiated specific wording in the IP licencing clause to cover use and adaptation for publications. Of these, one contract allows publications
only after the completion of the Project and the provision of the final reports, issues papers and other deliverables under this Agreement, and ... only after the Ministry has officially released or published the results of the Project, OR a period of 18 months has elapsed following provision of the final reports, issues papers and other deliverables under this Agreement ...
Three tender contracts from one NSW Ministry provide for IP to be owned jointly between the Government and the service provider. The tender guidelines from another entity provide that ‘[a] fair and equitable agreement as to the rights of respective parties to the intellectual property created as a result of the funded project will be negotiated with successful applicants on a case by case basis’ and the Government intends that the provider ‘will have the full right to publish any results obtained by them’ for academic purposes through an appropriate IP licence.
Just one final contract vests IP rights in the university provider with a licence granted to the Government to use and exploit the contract materials for its purposes.
3 Control of Publication and Dissemination of Results
Contractual terms governing whether, how and when research results can be disseminated publicly, especially through written publications, have important transparency implications. At one extreme, contracts can allow the government purchaser to exert complete control, for example, by stipulating that research outcomes may only be disseminated following formal processes of government review and approval. At the other extreme, the research provider may have full discretion to disseminate results, including through oral presentations and written publications, without any requirement of prior government review or approval. As a middle ground, the contract may oblige the external researcher to submit a draft report for informal feedback from a government representative, but with no obligation to comply with the government’s preferred wording or timing in the dissemination of results. Contracts may specifically require or expect the external provider to produce reports for publication in peer-reviewed journals. This is beneficial in ensuring that purchased research is subject to independent, expert review and makes a contribution beyond government in the advancement and dissemination of new knowledge. This benefit is undermined, however, if governments have the contractual power to direct how the results are expressed.
Compared to the clauses dealing with contract termination and IP, we found the greatest degree of variation between draft and final contracts in their provisions dealing with the publication of results. This indicates that government purchasers and external research providers engage in more negotiation about this aspect of their contractual relationships than concerning contract termination and IP.
As noted above, in the Commonwealth contracts, the ownership of IP permits government control over the publication of outputs. All the final contracts expand on how publication will be handled and require the research provider to submit draft materials for government review. In all but one case, the Government has the right to approve the final version. One contract with a university stated simply that the ‘draft report must be presented to the Department for comment prior to the final version being completed’. A contract for a research project involving a literature review stipulates that the research organisation cannot publish the review without the Government’s written approval. Other contracts require the researcher to submit interim and final reports to the Department of Health for review, make revisions as requested, and obtain approval before sharing reports with any other party. Another project, a partnership between an Australian Government health commission and a university, requires as a project outcome the publication of results and methodologies in peer-reviewed journals. Draft publications require the Commission’s approval:
Prior to submission of the articles to journals, the Contractor will provide the articles to the Commission for review and approval. In the event of a dispute about a draft journal article, the Advisory Group will be asked to provide a final opinion.
According to the contract, this Advisory Group is a body organised by the Commission.
In NSW, most draft contracts prohibit research providers from publicly releasing results without prior written approval from the Government. A minority of draft clauses provide that this approval will not be unreasonably withheld or that the Government should ‘be amenable to negotiation regarding appropriate and acceptable processes and timing for dissemination’. In some cases, non-disclosure is accomplished by defining as confidential any material created or written for the project and requiring the provider to keep secret any confidential information both during and after the term of the contract.
University-based providers seem more likely to negotiate over restrictions on publication. Some final contracts reveal compromises; for example, instead of requiring formal written approval for any publication or presentation, the provider must instead submit a manuscript or abstract to the government for ‘review and comments’. What happens next varies. The researcher may be contractually bound to modify or delete any information the government ‘reasonably believes will harm, prejudice or in any other way injure [its] interests’ or to ‘consider but [not be] obliged to follow’ the comments.
Two multi-year and multi-million dollar university research programs funded by the NSW Department of Health require 10–15 peer-reviewed publications of research findings as project deliverables. However, there are differences in the degree of control the government may exert over the public dissemination of findings. One contract requires the university to submit an ‘unabridged version’ of ‘any presentation, abstract, journal article, media material, conference paper or similar ... to the Ministry no less than 30 business days prior to the proposed due date.’ The Ministry then has the option to either approve the material for public release or
on a reasonable basis, reject the Publication in writing, in which case the contractor must omit the content identified ... as unacceptable from the Publication, provided that this rejection may not be exercised by the Ministry after a period of 18 months has elapsed following provision of the final reports, issues papers and other deliverables under this Agreement.
The other contract provides that the research team ‘has the right to publish findings of their research’, but must submit draft publications to the government for review and comment, then ‘incorporate feedback’ into the final versions. Both contracts include a schedule that details the review process for publications and reports. For articles submitted to scholarly journals, the Ministry must approve any substantial revisions requested during the peer review process. The schedule states that a ‘substantial revision’ includes ‘those that have implications for government policy’.
Two NSW contracts contain an open access clause:
All research papers that have been accepted for publication in peer reviewed journals, and supported in part or in whole with the Funding, must be deposited into an open access institutional repository within a 12 month period from the date of publication.
Such publications must also bear the disclaimer that the views presented are not necessarily those of the Government.
A few contracts, including those with universities, do not have specific provisions on journal publications and state more generally that the research provider must only use contract materials with the government’s written approval. Since a contract is meant to provide evidence of a ‘meeting of the minds’ of the parties to the agreement, it would be better practice to specify the publication terms in the contract. We argue below that such terms should be consistent with the principle of openness and that any restrictions on dissemination should have a strong justification.
One NSW Department made contracts with seven different research providers (six were universities) to undertake work on a common project. Interestingly, the draft contract issued to all bidders required the Government’s prior written approval to publish research results, but this clause was absent from all the final agreements, suggesting a common concern among the research providers with limiting government control over dissemination.
The findings of our contracts analysis have troubling implications. The routine use of control clauses is at odds with government pledges to be more transparent, provide open access to information, and enable the public to be meaningfully engaged in policy debates and decisions. The ideal of evidence-based policymaking is also compromised. As the Productivity Commission Chair has asked: ‘Can data and analysis that are not able to be scrutinised by third parties really be called “evidence”?’ The influence of ideology and politics on the production and use of evidence is hidden when governments have the power to control external researchers and the outcomes of their work. Moreover, academic freedom, central to universities’ mission to advance and disseminate knowledge informed by free inquiry, is undermined when university-based researchers face restrictions on publication. We turn now to recommendations to improve contracting processes and provisions to ensure that government commitments to transparency are realised in practice.
In a consultation document for its First Open Government National Action Plan, the Commonwealth claims to have a ‘long, proud history’ of openness, but acknowledges it ‘needs to become even more open, transparent and accountable, and improve public engagement.’ The National Action Plan’s commitments include improving ‘the discoverability and accessibility of government data and information’ and enhancing ‘public participation in government decision making’. Australian states also champion the benefits of openness. According to the NSW Government, ‘[a] smart government is transparent and accountable, and understands that solutions to policy challenges can come from outside government.’
Despite these public commitments, the starting point for most of the contracts we reviewed is a substantial degree of government control that may only be diminished through negotiation that depends on the skills and values of the external researchers and their legal representatives (and, of course, on the willingness of the government representative to relinquish control). Our findings underscore the importance of scrutinising government practices and we call on all federal, state and territorial governments to review their procurement practices and pro forma contracts to ensure they advance the principles of openness and transparency.
Transparency in government contracting for research raises several key questions, including:
(1) To whom are government contracts awarded?
(2) What are the terms of the contractual relationship between the government purchaser and the external provider of research services? and
(3) Are the findings of government-purchased research reported publicly?
Reforms have been made to mandate transparency in the awarding of contracts, which is important to root out cronyism and other corrupt practices. The Commonwealth Procurement Rules require that Australian Government entities ‘enable appropriate scrutiny of their procurement activity’ and require that tender awards above reporting thresholds be disclosed on the AusTender website. In NSW, the GIPA Act requires a public register of government contracts over $150 000 and similar registers exist in other states and territories (with variation in the dollar value of disclosed contracts). This compulsory disclosure, however, only covers basic details such as the name of the entity awarded the contract, the contract value, and the contract duration. And, as noted earlier, audits of the AusTender website reveal dismal compliance with reporting requirements.
1 Disclosure of Full Contract Details
To achieve an even greater degree of openness, we agree with Transparency International Australia that the ‘[f]ull details of awarded contracts should be disclosed’. We argue for proactive transparency, which could easily be accomplished by posting the contract on tender websites, alongside the contract award notice. Members of the public could then see to whom the contract has been awarded, as well as the full terms of the relationship between the government purchaser and the external provider. Importantly, this would enable public scrutiny of control clauses and governments could be held to account for contractual terms that allow them to control the conduct and dissemination of research. Proactive transparency would obviate the need for interested parties — such as the present authors — to make FOI requests for contracts and save government departments the costs of processing these requests.
Our recommendation is not radical since the Commonwealth Procurement Rules already express a general presumption that the terms of a contract are not confidential: ‘once a contract has been awarded the terms of the contract, including parts of the contract drawn from the supplier’s submission, are not confidential’. An exception applies to details that come within the scope of the government’s procurement confidentiality principles, which state that contractual terms may be considered confidential if they deal with ‘sensitive security information’ or ‘the contract is for a consultant to prepare a confidential report which is expected to deal with sensitive public interest issues’. No further elaboration is provided as to what counts as such sensitive issues, however governments should not apply exceptions to public disclosure in a manner that is inconsistent with principles of openness. In any case, clauses that set out the terms related to publication, IP and termination do not, in themselves, reveal sensitive information and should not be caught by these exceptions.
2 The Substantive Content of Contractual Terms
To advance the goals of transparency, governments should ensure that the contract terms do not exert an unnecessary degree of control over external research providers. Therefore, in addition to advocating for the full disclosure of contracts on public registers, we also recommend that governments, in consultation with stakeholders, review the terms in their standard form contracts and update these pro formas where necessary to advance transparency goals. Representatives of the university sector, including researchers who have experience performing contract research, are key stakeholders. It is imperative that contractual relationships formed between government and university-based researchers do not undermine the public values expressed in the Acts of Parliament that establish Australian universities and diminish trust in and respect for universities as institutions with a unique and important role in society. University-based researchers who believe that contract research compromises their academic freedom may choose to avoid such work. This would be an undesirable outcome as ‘it is clearly not in government’s interest, or in the public interest, if there is a narrowing pool of people who are willing to work on government contracts.’
It is beyond the scope of this article to propose a standard form for contract research, but we synthesise here several core transparency principles and recommendations pertinent to the three types of control clauses we examined in Commonwealth and NSW contracts.
(a) Termination Clauses
Termination-for-convenience clauses give government purchasers a broad power to end contracts at their discretion. The Australian Government has recently sought to streamline contracts for under $200 000 by establishing a core set of contractual terms, including a termination for convenience clause. The Government adopts a rigid stance in warning that these terms ‘cannot be changed and are non-negotiable’. The convenience for government must be weighed against the potential downsides for external researchers who face challenges in establishing research teams and managing resources when they lack certainty about the funding commitment. If government seeks the benefit of termination for convenience, the contract ought to explicitly incorporate a good faith requirement to prevent the unreasonable exercise of this power based on capricious or ulterior motives. For higher value contracts to fund multi-year research projects, it may be unreasonable to expect external researchers to undertake such work with the risk that the contract may be ended solely at the government’s discretion. Termination-for-cause provisions would meet governments’ interests in ensuring the work progresses in accordance with an agreed plan and schedule.
(b) Intellectual Property
Contractual terms governing the handling of IP shape whether and how research findings are publicly disseminated and used. Various recommendations have been put forward for dealing with IP in the context of government-purchased research and public sector information. Internationally, the OECD acknowledges that ‘[t]here is a wide range of ways to deal with copyrights on public sector information’ and encourages ‘institutions and government agencies that fund works from outside sources to find ways to make these works widely accessible to the public’. The 2009 Government 2.0 Taskforce Report recommended that ‘[n]ew contracts or agreements with a third party should endeavour to include a clause clearly stating the Commonwealth’s obligation to publish relevant data’ using Creative Commons licensing.
In September 2016, the Australian Government released new Intellectual Property Principles for Commonwealth Entities (‘IP Principles’) and Guidelines on Licensing Public Sector Information for Australian Government Entities. The IP Principles state that, in their approaches to the ownership, management and use of IP, Commonwealth entities should advance the policy objectives that underpin Government 2.0, which includes open access and the flow of information within and beyond government. In regard to public sector information, entities ‘should encourage public use and easy access to material’ particularly to meet the goals of informing the public on government activities and policies. Creative Commons licences are recommended to promote the free and open use of public sector information.
The Guidelines on Licensing mark a policy shift away from the standard practice of the Commonwealth asserting copyright on publications, which was the default we observed in many of the contracts in our study. Instead, the new guidelines advocate open access of public sector information and state that the ‘author entity ... is best placed to determine the use that others may make of it’. From the university perspective, the Go8 has argued that universities should own new IP developed during the course of contract research as they are best positioned to ensure public dissemination.
Contracts should not empower governments to sanitise or supress the results of externally commissioned research. The publication and dissemination of research findings should be a key goal of contracts between government purchasers of research and the external researchers who provide their expertise. For university-based researchers in particular, the contractual relationship should respect and support the principle that ‘the right to publish the results of all research in a timely manner is a critical tenet of the concept of academic freedom and of the integrity of the research process’. A consultative process of review and feedback between the government purchaser and the researcher may well help improve the accuracy and quality of presentations and publications, but the government should not wield veto power over public dissemination.
The recommendations of the Sedley Report in the UK provide an exemplar for Australian governments by urging that research contracts make a clear commitment to the publication of results, recognising that ‘prompt and full publication of government research is the norm’. This is especially important for Australia as a participant in the Open Government Partnership. We elaborate below on a proposal for a public register of contract research reports. This is not a radical suggestion, but rather a logical progression in steps toward enhancing open government that include Australian governments’ current initiatives to provide open access public sector datasets and to publish the results of program evaluations.
1 The Open Data Movement
In its current Open Government Action Plan, the Australian Government pledges to increase the availability of public data to support research and data driven decision-making and problem solving. This commitment follows the release of a Public Data Policy Statement in December 2015 in which the Government promises ‘to optimise the use and reuse of public data’ and ‘to release non sensitive data as open by default’. Furthermore, government entities will ‘where possible, ensure non-sensitive publicly funded research data is made open for use and reuse’.
These statements are promising, but two important shortcomings are evident. First, the government websites dedicated to publishing datasets currently have meagre content. As of July 2017, 20 healthcare datasets were available on data.gov.au, covering disparate topics such as: statistics on health behaviours of NSW residents; a database of the location of European wasp nests in the Australian Capital Territory (‘ACT’); and a list of venues for yoga, pilates and tai chi classes in Victoria.
Second, it is important to note that data are distinguished from information: ‘data comprises raw, unorganised material’, while information is produced from data that has been ‘processed and presented in context’. Release of datasets is valuable but the Australian Government’s Open Government National Action Plan also promises to improve access to information.
If governments release data, then interpretations of data — ‘information’ — that inform their policy and program development and public spending decisions, ought also to be available. The analyses and interpretations of data should be subject to scrutiny and critique if the conclusions or recommendations made do not follow from the data or if data appear to have been manipulated to reach politically expedient decisions. Further, if governments are committed to public disclosure of research datasets, then externally commissioned researchers will presumably collect data (for example, responses to a survey) on the understanding that the dataset will be made publicly available. Why then should they be expected to work under contractual terms that permit the non-disclosure of research findings (for example, the report that analyses the survey results in the context of a relevant government policy or program)?
Moreover, releasing data while suppressing information can undermine democratic participation and worsen informational inequities:
Without the skills and knowledge to interpret the vast amount of datasets thrown at them, or any context surrounding the data, many citizens will not be able to make sense of these data. They will not benefit from OGD [open government data] and, hence, they will not have full access to the information that can be extracted from these data.
Hence, it will be highly unlikely that the benefits of OGD for transparency, accountability and public participation will actually materialise in this setting. On the contrary, OGD may actually reinforce or enlarge existing inequalities in access to information — the distinction between the information-haves and the information-have-nots ... — and create an elite that can make use of the available datasets and that will hold an important power over other citizens.
2 Reporting of Program Evaluations
Some Australian governments are taking steps to adopt transparency in program evaluation by endorsing the publication of program evaluation reports. This marks some progress toward a 2009 Productivity Commission recommendation for ‘[a] national repository of evaluation reports ... [to] improve the level of scrutiny and increase the range of evidence available.’ In 2016, the NSW Government adopted new Program Evaluation Guidelines that set a general rule that agencies must ‘proactively and publicly release the findings of program evaluations.’ The 2010 ACT Evaluation Policy and Guidelines state that the ‘[c]ommunication of evaluation results helps disseminate key lessons and experience, informs decision making and promotes transparency and accountability’. However, the Guidelines imply that communication in many cases will only be internal within government and only the results of ‘significant evaluations [should be] publically [sic] available’. The Queensland Program Evaluation Guidelines emphasise the importance of ‘clear, transparent reporting’ of key findings, but do not currently require public disclosure of all evaluations. The Guidelines are, however, meant to align with the Government’s overall Performance Management Framework, which asserts that ‘[p]ublishing performance information is essential for accountability, transparency, driving continuous improvement in performance, and influencing trust and confidence in public sector service delivery.’
We go further and recommend that all externally purchased research, not just evaluation reports, should be made available in a public repository established for this purpose.
3 Public Repository of the Results of Contract Research
In the UK, the Sedley Report recommended ‘[a] standardised central register of all externally commissioned government research.’ We endorse this approach for Australia since it most strongly advances proactive transparency. As the Australian Information Commissioner has pointed out: ‘Communicating information to the public is a major tool and activity of government. It is how government reports what it is doing ... shares research ... and facilitates public participation in government decision making and priority setting.’
Reports could be identified by the original tender identification number, with links to the materials archived on tender websites, such as the approach to market and contract documents. This reporting would be separate from the publication of scholarly works in peer-reviewed academic journals. In exceptional circumstances, reports may justifiably be withheld from publication on compelling privacy or security grounds, but the existence of the report should be noted in the register along with the basis on which it is being withheld. This is consistent with the Productivity Commission’s recent recommendation that ‘[p]ublicly funded entities, including all Australian Government agencies, should create comprehensive, easy to access registers of data ... that they fund or hold.’
We anticipate several objections to this recommendation, primarily from government officials. First, governments will fret about ‘how information might be “spun” by the media, their opponents or those with direct commercial interests or an axe to grind.’ These concerns, while genuinely held and understandable, are not legitimate reasons for withholding externally commissioned research results from public view. FOI laws make it clear that the prospect of being embarrassed does not alone justify withholding government-held information from the public. With effective communication, governments can help the media and public understand research results and explain how ‘negative’ findings are important to evidence-informed policy and program decisions. Governments that publish program evaluation reports will have experience with this type of communication. We also emphasise that proactive disclosure on a public register would apply only to externally commissioned research. Internal research and policy advice from bureaucrats would not be included, thus freeing them to provide the ‘frank and fearless’ advice that Ministers seek.
Second, the costs and administrative burdens of maintaining another public register may be a concern. However, governments are now well experienced in designing and running other centralised databases, including the AusTender website and data.gov.au, and a similar database should not demand novel information technology expertise and platforms. Improving proactive disclosure could also reduce the costs of processing FOI requests.
Third, a requirement to publish the results of externally commissioned research may create an incentive for governments to exert greater control over the design and conduct of the research to produce results that cast them in a flattering light. However, reports should provide a convincing rationale for the formulation of the research questions, design and methods to counter such manipulation. The knowledge that reports will be open to external scrutiny would also militate against designing studies that are obviously contrived to produce favourable results.
Our investigation of Commonwealth and NSW contracts has started to answer the question of what types of control clauses are present in government contracts and to propose reform options to advance transparency in externally purchased research. One limitation of our documentary analysis of contracts is that it was a pilot study with a small sample of tenders for health-related research and we cannot draw conclusions about the prevalence of control clauses across all types of research contracts. However, control clauses were prevalent in the contracts we accessed and it is reasonable to assume that they exist in other contracts. It would be enlightening to undertake a study of contracts for research in other areas of contemporary policy controversy, such as environmental and climate science issues.
Further studies are needed to explore: the factors that contribute to the use of such clauses; the experiences of government purchasers and external providers in contract negotiations; the conduct of research that is subject to control clauses; and the broader transparency and accountability implications in this context. Research is needed to investigate the extent to which governments invoke control clauses, especially to suppress the publication of research and to sanitise results and shape the messages that communicate research findings when they are publicised. It would be worthwhile to follow up on Yazahmeidi and Holman’s survey of Australian researchers about their experiences of ‘suppression events’ when doing contract research. This survey is now over 10 years old and Australian governments have since made commitments to improve transparency and access to public sector information. Have researchers’ experiences of suppression decreased or increased? Do formal transparency requirements, such as obligations to post basic contract information online, lead to greater reliance on clauses within contracts to protect the results from disclosure? The Government 2.0 Report hypothesised that professional advisors to government managers and decision-makers, including legal and communications professionals, are generally risk averse and ‘see maximisation of control as a default setting.’ The attitudes and practices of these advisors are important areas for investigation.
Future study should also examine the types of external entities that provide research and advice to governments and identify any differences that arise depending on the provider type, such as university-based researchers, research business spin-offs from universities, or private sector think tanks and consultancies. Such work could investigate how governments construct research briefs and whether they vary for different types of providers. While our analysis focused on contractual control clauses, it is important to note that governments can also exert significant control over research by specifying the design and methods that external providers must use. A recent Australian study interviewed policymakers and researchers involved in evaluations of health programs and policies. The authors learned that ‘[i]n many cases, evaluations are undertaken specifically to demonstrate positive results and participants reported that briefs for external evaluators were written in a way that implies a certain result is expected.’
As a related question, it is worth exploring how governments view the credibility and legitimacy of different research providers. A recent study by Doberstein observed that ‘[a]cademics have historically occupied a privileged position of authority and legitimacy in the public domain as it relates to policy research, but some argue that this is changing with the growth of think tanks and research-based advocacy groups.’ Doberstein recommended critical investigation of the ‘competition of ideas and for influence, as the magnitude and sophistication of non-university-derived research and policy advice continues to grow’. Scholars have observed a trend to an increased politicisation of research and policy advice provided by external think tanks and consultants. Some research suggests that private sector consultants are ‘often seen as more likely to report positively on policies and programs, in a way that is less of a risk’ for the government purchasing the research. Further investigation is needed to find out whether private consultants provide skewed results. If universities insist on academic freedom and more transparency in contract research, governments may respond by engaging private consultants who are perceived as more compliant than academics. Competition for funding and pressure to do research that has ‘impact’ on government policies and program may also persuade university-based researchers to accept briefs that are designed to produce results favourable to the government purchaser. Schneider and colleagues have recently commented on these types of political influences in the context of evaluation research:
Political imperatives not only influenced if, when and with what funding a policy or program might be evaluated, but indirectly determined the focus of evaluations. For example, there may be a preference for measuring cost-utility rather than population health outcomes. In some situations, it may be more important to be immediately seen to be ‘doing something’ rather than having an actual effect.
In short, the politics and practices of participants and stakeholders in externally purchased research are ripe for legal, ethical and socio-cultural investigation.
In 2013, the Office of the Chief Scientist surveyed over 1000 Australians across the country, asking their views on the most pressing issues that science should address. This project culminated in an essay collection, titled The Curious Country, featuring expert commentary on those issues. The introduction to the volume declared that
it is vital that well-trained minds wield the tools of scientific inquiry. ... the most effective way to enhance Australia’s social, economic, physical and intellectual wellbeing is to sort truth from fantasy, attainable dreams from wild conjecture and apply the findings to evidence-based national, regional and local government policy...
The Chief Scientist lamented ‘the increasingly limited information available about science and society through mainstream media’ and emphasised the need for transparency and openness in research.
We applaud the Chief Scientist and other government offices for their commitments, at least in principle, to robust scientific inquiry, evidence- informed policymaking, public engagement and transparency. However, as we have argued here, much more effort and cultural changes are needed to see these principles regularly and consistently enacted in practice.
Stating lofty objectives is not sufficient, and organisational cultures and practices must shift to remove attitudinal barriers to transparency. The need for culture change is identified in a recent Productivity Commission Report on access to public datasets:
Despite recent statements in favour of greater openness, many areas of Australia’s public sector continue to exhibit a reluctance to share or release data.
The entrenched culture of risk aversion ... greatly inhibits data discovery, analysis and use.
By purchasing external research, governments can deliver to the public the benefit of accessing independent expertise, technical knowledge, innovative methods and ideas, and other resources. Governments could be criticised if they relied solely on internal perspectives in the design and evaluation of their policies and programs. Governments have a corollary obligation to citizens to be open and transparent about the research they purchase and to facilitate the public dissemination of the findings unless there is strong justification for not doing so. Moreover, particularly when making research contracts with universities, governments should not undermine the principal function of universities to engage in the ‘dissemination, advancement, development and application of knowledge informed by free inquiry’.
 Associate Professor, Faculty of Law, University of Technology Sydney (‘UTS’), New South Wales, Australia; Member, UTS Law | Health | Justice Research Centre.
† Professor and Senior Brawn Research Fellow, School of Medicine and Public Health, University of Newcastle, New South Wales, Australia.
The authors acknowledge funding support from the University of Newcastle and the valuable research assistance provided by Shelby Houghton, Briony Johnston and Monique Allan. The authors are grateful for feedback on a draft version of the paper provided by Dr Wendy Lipworth, Sydney Health Ethics, University of Sydney, and for comments from the anonymous peer reviewers.
 Office of the Australian Information Commissioner (Cth), Understanding the Value of Public Sector Information in Australia: Issues Paper 2 (November 2011) 2 <http://apo.org.au/system/files/27576/
 Nick Ross, ‘Foreword’ in Sir Stephen Sedley, Missing Evidence: An Inquiry into the Delayed Publication of Government-Commissioned Research (Sense about Science, June 2016) <https://researchinquiry.org/inquiry-report/> (‘Sedley Report’).
 Anne McKenzie et al ‘Response to “A Survey of Suppression of Public Health Information by Australian Governments”’ (2008) 32(1) Australian and New Zealand Journal of Public Health 90.
 Open Government Partnership, Improving Public Services: Guidance for Developing OGP Commitments, 2 <https://www.opengovpartnership.org/sites/default/files/OGP_public-service-guidance.pdf>.
 Department of the Prime Minister and Cabinet (Cth), Australia’s First Open Government National Action Plan 2016–18 (December 2016) 3 <http://ogpau.pmc.gov.au/australias-first-open-government-national-action-plan-2016-18> . Plans are underway for Australia’s Second Open Government National Action Plan 2018–20: <https://ogpau.pmc.gov.au/consultations-australias-second-open-government-national-action-plan-2018-20>.
 Angus Taylor, Assistant Minister for Cities and Digital Transformation (Cth), ‘Australia’s Adoption of the Open Data Charter’ (Statement, 27 March 2017) cited by Samira Hassan, Australia Adopts the International Open Data Charter (7 April 2017) <https://blog.data.gov.au/news-media/blog/
australia-adopts-international-open-data-charter>. The Open Data Charter has six principles: (1) government data should be open by default; (2) access should be provided in a timely and comprehensive way; (3) data should be accessible and usable; (4) data should be comparable and interoperable; (5) open data should improve governance and citizen engagement; (6) open data should support inclusive development and innovation: Open Data Charter Principles <https://opendatacharter.net/principles/>.
 Ariadne Vromen and Patrick Hurley, ‘Consultants, Think Tanks and Public Policy’ in Brian Head and Kate Crowley (eds) Policy Analysis in Australia (Policy Press, 2015) 167. See also Arnošt Veselý, ‘Externalization of Policy Advice: Theory, Methodology and Evidence’ (2013) 32(3) Policy and Society 199, 206 <https://doi.org/10.1016/j.polsoc.2013.07.002>. Veselý hypothesises three broad reasons for outsourcing: (1) lack of internal capacity; (2) political factors; and (3) the context in which the policy advice is provided.
 Brian Head et al, ‘Are Policy-Makers Interested in Social Research? Exploring the Sources and Uses of Valued Information among Public Servants in Australia’ (2014) 33(2) Policy and Society 89. For commentary on the development of the New Public Management movement from the Hawke Government in the mid-1980s, see Judy Johnston, ‘The New Public Management in Australia’ (2000) 22(2) Administrative Theory and Praxis 345.
 Schneider, Milat and Moore report the concern that if results will not be published, the ‘quality of evaluations was generally lower due to the absence of expectations of public scrutiny’: Carmen Huckel Schneider, Andrew J Milat and Gabriel Moore, ‘Barriers and Facilitators to Evaluation of Health Policies and Programs: Policymaker and Researcher Perspectives’ (2016) 58 Evaluation and Program Planning 208, 211–12.
 A 2017 Productivity Commission Report on access to public sector data points out: ‘The lack of public release and data sharing between government entities has contributed to fragmentation and duplication of data collection activities. This ... wastes public and private sector resources’: Productivity Commission (Cth), Data Availability and Use: Productivity Commission Inquiry (Report No 82, March 2017) 34, 153 (Finding 3.5) <http://www.pc.gov.au/inquiries/completed/data-access/report/data-access.pdf> .
 Australian Institute of Health and Welfare (Cth), Health Expenditure Australia 2016–17
(28 September 2018) <https://www.aihw.gov.au/reports/health-welfare-expenditure/health-expenditure
 In addition, the investigators’ own areas of research expertise are in health-related fields, which strengthened our ability to select and review government contracts.
 Christopher Hood and David Heald (eds), Transparency: The Key to Better Governance? (Oxford University Press, 2006).
 Stephan Grimmelikhuijsen Transparency and Trust: An Experimental Study of Online Disclosure and Trust in Government (PhD thesis, Utrecht University, 2012) 55. See also Stephan G Grimmelikhuijsen and Eric W Welch, ‘Developing and Testing a Theoretical Framework for Computer-Mediated Transparency of Local Governments’ (2012) 72(4) Public Administration Review 562.
 Vincent Mabillard and Raphael Zumofen, ‘The Complex Relationships between Transparency and Accountability: A Synthesis and Contribution to Existing Frameworks’ (2017) 32(2) Public Policy and Administration 110.
 Department of Finance (Cth), Government 2.0 <https://www.finance.gov.au/archive/policy-guides-procurement/gov20/>.
 Government 2.0 Taskforce, Engage: Getting on with Government 2.0: Report of the Government 2.0 Taskforce (December 2009) 88 <https://www.finance.gov.au/sites/default/files/Government20Taskforce
Report.pdf?> (‘Government 2.0 Taskforce Report’).
 Ibid xvii, 22.
 The Taskforce adopted an Organisation of Economic Co-operation and Development (‘OECD’) definition of public sector information: ‘information, including information products and services, generated, created, collected, processed, preserved, maintained, disseminated, or funded by or for the government or public institutions, taking into account [relevant] legal requirements and restrictions’: ibid 4 [1.4]. See also OECD, OECD Recommendation of the Council for Enhanced Access and More Effective Use of Public Sector Information, C(2008)36 (17–18 June 2008) 4 n 1 <http://www.oecd.org/
 Government 2.0 Taskforce Report, above n 17, xix, xxi (Recommendations 6, 8).
 Department of Finance (Cth), Declaration of Open Government (July 2010) <https://www.finance. gov.au/blog/2010/07/16/declaration-open-government/>.
 Office of the Australian Information Commissioner (Cth), Principles on Open Public Sector Information, principle 1 <https://www.oaic.gov.au/information-policy/information-policy-resources/
principles-on-open-public-sector-information>. Greater openness is also championed at state and territorial levels. For example, the NSW Office of the Premier and Cabinet issued a Memorandum of Open Government in 2012 that expressed its commitment ‘to the open government principles of transparency, participation, collaboration and innovation’ (emphasis in original): Barry O’Farrell (Premier), M2012-10 Open Government (22 August 2012) Office of the Premier and Cabinet (NSW) <http://arp.nsw.gov.au/m2012-10-open-government> . The Victorian Department of Premier and Cabinet champions these open government principles since they ‘underpin the good government agenda’: Chris Eccles, ‘The “Open Government” Principles of Transparency, Collaboration
and Participation Underpin the Good Government Agenda’ (Speech delivered at the
ANZSOG Conference, Grand Hyatt, Melbourne, 5 August 2015) <http://www.vic.gov.au/
 See, eg, Productivity Commission (Cth), Strengthening Evidence-Based Policy in the Australian Federation: Roundtable Proceedings, vol 1, 9 <http://www.pc.gov.au/research/supporting/
 Brian W Head, ‘Toward More “Evidence-Informed” Policy Making?’ (2015) 76(3) Public Administration Review 472, 472 (citation omitted).
 Benjamin Hawkins and Justin Parkhurst, ‘The “Good Governance” of Evidence in Health Policy’ (2015) 12(4) Evidence and Policy 575, 582 (emphasis altered).
 Veselý, above n 7; Vromen and Hurley, above n 7.
 Both ARC and NHMRC have adopted the ‘National Principles of Intellectual Property Management for Publicly Funded Research’: see NHMRC, National Principles of Intellectual Property Management for Publicly Funded Research (7 June 2018) 2(c) <https://www.nhmrc.gov.au/grants-funding/policy/intellectual-property-management/national-principles-intellectual-property-man>.
 See, eg, University of Sydney Act 1989 (NSW) s 6(2)(b); University of Melbourne Act 2009 (Vic) s 5(e)(iii); Australian National University Act 1991 (Cth) s 5(1)(a).
 Australian Government, ausTENDER <https://www.tenders.gov.au/>.
 See, eg, NSW Government, eTendering <https://tenders.nsw.gov.au/>; State Government of Victoria, Tenders VIC <https://www.tenders.vic.gov.au>. Australia is not alone in the absence of clear quantification of government spending on externally-purchased research. The 2016 Sedley Report noted that ‘[t]he United Kingdom government spends about £2.5 billion a year on research intended to guide, develop, modify and monitor policy on a wide variety of issues’, however ‘[i]t is difficult to tell how much of this was spent on commissioned research.’: Sedley Report, above n 2, 1–2.
 John Halligan, ‘Policy Advice and the Public Service’ in B Guy Peters and Donald J Savoie (eds), Governance in a Changing Environment (McGill-Queen’s University Press, 1995) 138.
 Jonathan Craft and Matt Wilder, ‘Catching a Second Wave: Context and Compatibility in Advisory System Dynamics’ (2017) 45(1) Policy Studies Journal 215, 215 <https://doi.org/10.1111/
 Head et al, above n 8, 90.
 Boshra Yazahmeidi and C D’Arcy Holman, ‘A Survey of Suppression of Public Health Information by Australian Governments’ (2007) 31(6) Australian and New Zealand Journal of Public Health 551, 552–3 <https://onlinelibrary.wiley.com/doi/abs/10.1111/j.1753-6405.2007.00142.x>.
 Ibid 551. See also 553, 555.
 C D’Arcy J Holman, ‘An End to Suppressing Public Health Information’ (2008) 188(8) Medical Journal of Australia 435.
 Kypros Kypri, ‘Suppression Clauses in University Health Research: Case Study of an Australian Government Contract Negotiation’ (2015) 203(2) Medical Journal of Australia 72.
 Go8, In the Interests of Innovation: Time for a New Approach to Negotiating Research Agreements between the Commonwealth and Australian Universities: A Supplementary Submission from the Group of Eight to the Review of the National Innovation System (April 2008) 1 <https://go8.edu.au/sites/default/files/docs/go8_innovation_review_supp_sub_300408.pdf>.
 ANAO, ‘Confidentiality in Government Contracts: Senate Order for Entity Contracts (Calendar Year 2015 Compliance)’ (ANAO Report No 18 of 2016–2017, 27 September 2016) 7  <https://www.anao.gov.au/work/performance-audit/confidentiality-government-contracts-senate-order-2015>.
 Ibid 8 .
 See Sedley Report, above n 2.
 Cabinet Office (UK), ‘UK Open Government National Action Plan 2016–18’ (Policy Paper, 12 May 2016) <https://www.gov.uk/government/publications/uk-open-government-national-action-plan-2016-18/uk-open-government-national-action-plan-2016-18>.
 World Wide Web Foundation, WebIndex <http://thewebindex.org> . The Productivity Commission points out that ‘Australia’s provision of open access to public sector data is below comparable countries with similar governance structures, including the United States, the United Kingdom and New Zealand.’: Productivity Commission, above n 10, 33, 76 (Finding 1.1).
 Sedley Report, above n 2, 8–9 [2.20]–[2.26]. See also Government Office for Science (UK), ‘Principles of Scientific Advice to Government’ (Guidance Paper, 24 March 2010) <https://www.gov.uk/government/publications/scientific-advice-to-government-principles/principles-
of-scientific-advice-to-government>; Government Office for Science (UK), ‘The Government Chief Scientific Adviser’s Guidelines on the Use of Scientific and Engineering Advice in Policy Making’ (Guidance Paper, 1 July 2010) <https://www.gov.uk/government/publications/scientific-and-engineering-advice-guidelines-for-policy-makers>; Government Economic and Social Research Team, HM Treasury (UK), Publishing Research and Analysis in Government: GSR Publication Protocol (May 2015) <https://www.gov.uk/government/uploads/system/uploads/attachment_data/
 Sedley Report, above n 2, 16. The Inquiry looked into nine case studies of research reports being withheld: see Sense about Science, Case Studies <https://researchinquiry.org/case-studies/>.
 Sedley Report, above n 2, 34 [4.22].
 The study was approved by the University of Newcastle Human Research Ethics Committee (approval number H-2015-0188).
 A scan of tender websites in other states revealed inconsistent practices. In Western Australia (‘WA’), for example, draft contracts are not available on the WA Tender website: WA Government, Tenders WA <www.tenders.wa.gov.au>. In Queensland, draft contracts are not accessible for closed and archived tenders: Queensland Government, QTenders <https://www.hpw.qld.gov.au/qtenders/>.
 Government departments may charge for processing FOI requests; eg, the current NSW fee is $30 per hour: Department of Justice (NSW) Access to Information <https://www.justice.nsw.gov.au/
 Notice of Decision Letter from NSW WorkCover to Shelby Houghton, 17 September 2015 (copy on file with authors).
 Termination for cause was also covered in many contracts, but we do not summarise those provisions here as cause-based termination (eg, due to non-performance of agreed activities) does not compromise government transparency goals.
 The standard Commonwealth contract states that the Government agrees to pay for any services properly rendered up to the termination date and any reasonable and unavoidable costs to the contractor arising from the termination. There is uncertainty in Australian jurisprudence about implied contractual duties of good faith. The High Court of Australia has not accepted an implied good faith term in termination-for-convenience clauses. However, even if such a requirement were implied, it can be explicitly removed through wording that permits an unfettered right to terminate the contract. For discussion of termination-for-convenience clauses, see Ruth Loveranes, ‘“Termination for Convenience” Clauses’  UNDAULawRw 6; (2012) 14 University of Notre Dame Australia Law Review 103.
 These were NSW Department of Health contracts issued in 2013 to a small and a large private consultancy and a medical research institute.
 These were contracts between the NSW Department of Education with a university (2012) and the NSW Department of Family and Community Services with a non-governmental organisation (2014).
 NSW Department of Health contract with a university (2013).
 NSW Department of Family and Community Services (2009; 2012; 2014).
 NSW Workcover (2014).
 NSW Department of Health contract with a university (2015).
 Commonwealth contract with a university (2007).
 Commonwealth contract with a small private consultancy (2011).
 Commonwealth contracts with a university and a small private consultancy (2007).
 Commonwealth contract with a university (2008).
 This provision appeared in draft NSW Department of Health contracts.
 This wording appeared in NSW Department of Health and Department of Education contracts.
 NSW Department of Education contract with a university (2012).
 NSW Department of Attorney General and Justice contract with a university (2013).
 NSW Department of Health contract with a university (2013).
 This was a NSW Department of Health contract with a university (2013).
 NSW Department of Health contracts with universities (2013).
 NSW Department of Health contracts with a university (2011) and a medical research institute (2011).
 These were NSW Department of Family and Community Service contracts with universities and a non-governmental organisation (2014).
 Productivity Commission (Cth), above n 23 vol 1, 9.
 Department of the Prime Minister and Cabinet (Cth), Australia’s First Open Government National Action Plan 2016–18: Draft for Consultation (2016) 3 <https://ogpau.pmc.gov.au/sites/default/files/
 Department of the Prime Minister and Cabinet (Cth), above n 5, 41, 59.
 NSW Government, Open Data Policy (2016) <https://www.digital.nsw.gov.au/sites/default/files/
 LobbyLens is an online tool that integrates Commonwealth Lobbyists Register Data with contract notices data from AusTender: LobbyLens, About LobbyLens <http://lobbyists.disclosurelo.gs/
 Department of Finance (Cth), Commonwealth Procurement Rules (March 2017) 15 [7.1] <https://www.finance.gov.au/sites/default/files/commonwealth-procurement-rules.pdf>.
 Ibid 17 [7.16]–[7.17]. The reporting thresholds are $10 000 for non-corporate Commonwealth entities and $400 000 for prescribed corporate Commonwealth entities. A $7.5 million threshold applies for construction service contracts.
 GIPA Act pt 3 div 5 (‘Government contracts with private sector’).
 See, eg, Tenders VIC, Contracts <https://www.tenders.vic.gov.au/tenders/contract/list.do?action=
contract-view>; Department of Corporate and Information Services (NT), Northern Territory Government — Awarded Government Contracts (1 August 2018) <https://nt.gov.au/__data/assets/
 GIPA Act s 29 lists the details that the contract register must include.
 Transparency International Australia, Transparency in Public Procurement (Position Paper No 6, (January 2016) 2 <http://transparency.org.au/tia/wp-content/uploads/2016/10/PP6-Transparency-in-Public-Procurement-Transparency-International-Australia-Jan-2016.pdf> .
 Department of Finance (Cth), above n 79, 18 [7.21] (subject to some exceptions listed in that paragraph).
 Department of Finance (Cth), Confidentiality throughout the Procurement Cycle (10 July 2014) principle 7 <http://www.finance.gov.au/procurement/procurement-policy-and-guidance/buying/
 Office of the Australian Information Commissioner (Cth), above n 1.
 We acknowledge that pro formas need to offer options to take account of interests of the contracting parties that arise in different circumstances.
 See above n 29.
 This point was emphasised by the Go8, above n 39, 2: ‘Public universities occupy a unique place in society. ... [yet the] important function and role of universities is often not recognised by government[s] ... when they seek to commission research at Australian universities.’
 Sedley Report, above n 2, 8 [2.18].
 Department of Finance (Cth) Commonwealth Contracting Suite (CCS) (21 August 2018) <http://www.finance.gov.au/procurement/commonwealth-contracting-suite/> (emphasis in original).
 The Go8 also noted these problems: above n 39, 10 [6.3]. See also the G08 recommendation at
 There is debate in Australian jurisprudence as to the meaning of a good faith provision in termination-for-convenience clauses. It likely implies a duty to act reasonably when invoking the clause (see, eg, Renard Constructions (ME) Pty Ltd v Minister for Public Works (1992) 26 NSWLR 234) and not to act capriciously based on an ulterior or extraneous motive (see, eg, Sundararajah v Teachers Federation Health Ltd  NSWSC 1443 (18 December 2009)).
 OECD, above n 19, 6.
 Government 2.0 Taskforce Report, above n 17, xix [6.6]. For more information on Creative Commons licensing, see Creative Commons Australia, About the Licences <http://creativecommons.org.au/
 Department of Communications and the Arts (Cth), Australian Government Intellectual Property Rules (10 July 2018) <https://www.communications.gov.au/policy/policy-listing/australian-government-intellectual-property-rules>.
 Department of Communications and the Arts (Cth), Intellectual Property Principles for Commonwealth Entities (10 July 2018) 8 [11(a)] <https://www.communications.gov.au/policy/
 Department of Communications and the Arts (Cth), Guidelines on Licensing Public Sector Information for Australian Government Entities (10 July 2018) 5 <https://www.communications.
 Go8, above n 39, 3 (principle 2).
 Ibid 5 [1.4]. Schneider, Milat and Moore add that ‘allowing adequate academic freedom and openness to publication can act as facilitators and increase high quality evaluation research’: above n 9, 214.
 Sedley Report, above n 2, 35 (recommendation III).
 Department of the Prime Minister and Cabinet (Cth), above n 5, 25.
 Australian Government, Australian Government Public Data Policy Statement (7 December 2015) <https://www.pmc.gov.au/sites/default/files/publications/aust_govt_public_data_policy_statement_1.pdf>.
 Ibid (emphasis altered).
 See Australian Government, About <http://data.gov.au/about> . The site is intended to provide ‘an easy way to find, access and reuse public datasets from Government. The main purpose of the site is to encourage public access to and reuse of public data.’
 The Health Behaviours in NSW dataset provides data on the ‘[p]roportion of NSW population over [the] age of 16 that; smoke, engage in risk drinking, are considered overweight or obese.’: Australian Government, Health Behaviours in NSW (12 May 2013) <http://data.gov.au/dataset/health-behaviours-in-nsw> .
 Australian Government, Location of European Wasps Nests (ACT) (7 March 2015) <http://data.gov.au/dataset/location-of-european-wasps-nests> .
 Australian Government, Yoga, Pilates and Tai Chi in Victoria — GFYL (7 March 2015) <http://data.gov.au/dataset/yoga-pilates-and-tai-chi-in-victoria-gfyl> .
 Department of the Prime Minister and Cabinet (Cth), above n 5, 25 n 2.
 Katleen Janssen, ‘Open Government Data and the Right to Information: Opportunities and Obstacles’ (2012) 8(2) Journal of Community Informatics <http://www.ci-journal.net/index.php/ciej/article/
view/952/954> (citations omitted).
 For links to several government policy statements on program evaluation, see WA Government, Program Evaluation across Australia (28 March 2014) <http://programevaluation.wa.gov.au/
 Productivity Commission, above n 23, vol 1, 207.
 Department of Premier and Cabinet (NSW), NSW Government Program Evaluation Guidelines (January 2016) 16 <https://arp.nsw.gov.au/sites/default/files/NSW%20Government%20Program%
20Evaluation%20Guideline%20January%202016_1.pdf>. This policy defines program evaluation as a ‘rigorous, systematic and objective process to assess a program’s effectiveness, efficiency, appropriateness and sustainability’: at 5.
 ACT Government, Evaluation Policy and Guidelines (December 2010) 14 <http://www.cmd.act.
 Ibid 8. The significance of an evaluation is determined by an ‘assessment of materiality, risk and complexity’.
 Queensland Treasury and Trade (Qld), Program Evaluation Guidelines (November 2014) 7 <https://s3.treasury.qld.gov.au/files/qld-government-program-evaluation-guidelines.pdf>. At page 12, the Guidelines state that ‘reporting requirements will depend on the objectives of the evaluation and the intended audience’.
 Department of the Premier and Cabinet (Qld), Queensland Government Performance Management Framework: An Overview (2017) 15 <https://www.forgov.qld.gov.au/sites/default/files/performance
 Sedley Report, above n 2, 34 (recommendation I). The Report goes on further to call for the ‘[r]outine publication of research government has considered in policy formulation with, if appropriate, reasons for rejecting it.’: at 36 (recommendation IV).
 Office of the Australian Information Commissioner (Cth), above n 1, 1. Transparency International Australia also advocates that ‘details of completion of the contract should be published in a timely manner’ and we believe such details include the research findings: above n 84, 2.
 Productivity Commission, above n 10, 243 (Recommendation 6.4). This recommendation continues: ‘Where datasets are held or funded but are not available for access or release, the register should indicate this and the reasons why this is so.’
 European scholars have conducted interviews with public sector officials and data archivists to explore the pros and cons of proactive release of government information: see Anneke Zuiderwijk and Marijn Janssen, ‘Towards Decision Support for Disclosing Data: Closed or Open Data?’ (2015) 20 (2) Information Polity 103.
 Government 2.0 Taskforce Report, above n 17, 50 [5.6.2].
 For example, the GIPA Act s 15(c) states that in determining access requests, ‘[t]he fact that disclosure of information might cause embarrassment to, or a loss of confidence in, the Government is irrelevant and must not be taken into account.’
 On the publication of ‘negative’ results as a component of credible and balanced performance reporting, see Department of the Premier and Cabinet (Qld), Performance Management Framework — Measuring, Monitoring and Reporting Performance: Reference Guide (2017) 16 (citation omitted) <https://www.forgov.qld.gov.au/sites/default/files/measuring-monitoring-reporting-performance.pdf>. The Government 2.0 Taskforce Report urged that ‘[o]fficials should be able to express themselves informally, tentatively and candidly but they must also do so in ways that retain people’s confidence that they are acting fairly, professionally and impartially.’: above 17, 20 [3.3]. The Sedley Report noted ‘strategies that can help government to commission and publish controversial research without fear of how the public or the media might respond’: above n 2, 31 [4.1].
 Henry Belot, ‘Australia’s Top Public Servants Call for FOI Reform to Hide Advice from Public’, The Canberra Times (online) 11 April 2016 <http://www.canberratimes.com.au/national/public-service/australias-top-public-servants-call-for-foi-reform-to-hide-advice-from-public-20160411-go3gxt.html> .
 Some government departments complain of the cost of managing FOI requests, claiming that many requests are from journalists looking for stories. Proactive disclosure could divert at least some of those requests as journalists could search through the public repository to report on research results. High-quality journalistic reporting can only serve to inform the public of government activities and enhance accountability.
 Schneider, Milat and Moore reported some evidence that this already happens: above n 9, 211.
 Yazahmeidi and Holman, above n 35.
 Government 2.0 Taskforce Report, above n 17, 51 [18.104.22.168]:
it is natural for managers seeking to minimise adverse risk to try to control whatever they can. ... [And] specific professions advising management, such as the provision of legal, communications or IT advice and services will typically see maximisation of control as a default setting to minimise adverse risks. ... one cannot be sure that it [information] will not be used or misused in ways that may embarrass an agency. So why release it if one can avoid it?
 Schneider, Milat and Moore, above n 9, 211.
 Carey Doberstein, ‘Whom Do Bureaucrats Believe? A Randomized Controlled Experiment Testing Perceptions of Credibility of Policy Research’ (2017) 45(2) Policy Studies Journal 384, 375 <https://onlinelibrary.wiley.com/doi/pdf/10.1111/psj.12166> (citations omitted).
 Ibid (emphasis altered; citations omitted).
 Vromen and Hurley, above n 7; Jonathan Craft and Michael Howlett, ‘The Dual Dynamics of Policy Advisory Systems: The Impact of Externalization and Politicization on Policy Advice’ (2013) 32(3) Policy and Society 187 <https://doi.org/10.1016/j.polsoc.2013.07.001>; Veselý, above n 7.
 Schneider, Milat and Moore, above n 9, 211.
 Leigh Dayton (ed), The Curious Country (ANU Press, 2013) <http://press.anu.edu.au/node/159/
 Ibid 9.
 Ibid. The book offers a brief exposition of the scientific method, then states (at 10): ‘in practice the science process is complex. Experiments must be as free as possible from personal bias or error. Other scientists must be able to replicate the results.’
 Productivity Commission, above n 10, 34, 153 (Finding 3.5).
 A recent audit of evaluations undertaken internally by government health promotion agencies in Victoria recommended external collaborations with academic partners to strengthen the quality of evaluation projects: Louise J Francis and Ben J Smith, ‘Toward Best Practice in Evaluation: A Study of Australian Health Promotion Agencies’ (2015) 16(5) Health Promotion Practice 715, 722.
 University of Sydney Act 1989 (NSW) s 6(2)(b). See also above n 29.