Thanks to TomB for this article.
SUMMARY OF ARTICLE:
No Information, No Democracy
| Democracy depends on informed citizens, yet governments and corporations routinely conceal information necessary for public accountability. Without transparency, citizens cannot properly judge those exercising political and economic power. Australia provides numerous examples of secrecy’s consequences. Government grants worth billions frequently lack adequate evaluation, while the Robodebt scandal demonstrated how concealment enabled unlawful conduct. Freedom of information requests are often refused or delayed, weakening public scrutiny. Radical transparency would require public disclosure of government contracts, subsidies, political donations, lobbying activities, corporate ownership and program outcomes. Independent evaluations should determine whether publicly funded programs achieve their objectives. Particularly rigorous disclosure should apply to medicine and artificial intelligence. Pharmaceutical companies should publish all clinical trial results, including failures, while AI developers should disclose safety testing, risks and independent evaluations. Commercial advantage alone should not justify secrecy. Transparency should extend to everyone’s income, assets and taxes, helping expose inequality and discourage tax evasion. Scandinavian experience demonstrates potential benefits, although privacy safeguards remain essential. Legitimate exceptions include national security, health records, family matters and private communications. Such exemptions must be narrowly defined and independently reviewed. Ultimately, democracy requires more than elections. Citizens need reliable information to question decisions, expose wrongdoing and hold powerful institutions accountable. Secrecy protects power; transparency empowers citizens. |
FULL ARTICLE:
A vote is only as good as the information behind it
Democracy asks ordinary people to make informed decisions about how they are governed, who should govern them, and how public money and public power should be used. That promise collapses if the information needed to judge those decisions is held by the very people being judged. Most citizens cannot see how a contract was awarded, what a grant achieved, who lobbied for a regulation, or what a company knew about a risk before it harmed people.
They are asked to vote, comment and complain, but largely in the dark. Participation without information is not participation. It is a ritual. This article argues that transparency across government and business, and over the income and assets of everyone, is not a bureaucratic nicety. It is the precondition for democracy working at all. The exceptions should be narrow: security, and personal data such as health and family life.
Who is telling us what we know?
Almost everything the public learns about government and business arrives through a filter: a minister’s media office, a corporate communications team, a lobbyist’s briefing, an industry-funded report, a selective leak. Each of these sources is entitled to tell its own story. Each also has a reason to tell the version that suits it. This is not mainly a story of villains. It is a story of incentives.
A small group that stands to gain from a decision will spend time and money shaping how it is described. The far larger group that pays for it, as taxpayers, customers or neighbours, has jobs, families and little time to dig. The gap between those two groups is where poor decisions survive. Journalists can narrow the gap, but they too depend on what they are given, and many newsrooms are smaller than they once were.
When documents are withheld or delayed until the story has moved on, even a determined reporter is working with a blindfold on.
The remedy is not to trust a better set of intermediaries. It is to let people see primary information themselves: the contract, the criteria, the outcome, the audit, the meeting. When the raw material is public, anyone can check the story, and anyone telling it knows they can be checked.
Waste that nobody has to explain
Some of the costliest failures in public life are not dramatic. They are programs that spend money, announce a benefit and never have to show a result. Australia hands out an estimated $125 billion a year in grants, according to SmartyGrants.
A white paper from the Institute of Grants Management reviewed a decade of auditor-general reports and found that nearly a third of the problems identified came down to poor data and weak performance evaluation. Its conclusion was that many audited programs could not show whether the money achieved the policy objective it was meant to.
The pattern shows up in individual audits too:
The Australian National Audit Office found that a billion-dollar-scale hospital and health grants program was administered ineffectively and fell short of ethical requirements, and that as of December 2022 no overall evaluation of the program had been done. A centralised grants hub system that cost more than $150 million rarely had its evaluation services used. The problem is not uniquely Australian. An auditor-general in Nova Scotia reached a similar conclusion about provincial grants: departments were not adequately checking whether programs met their goals.
The point is not that every grant is wasted. It is that the public usually cannot tell. If a program must publish its goals, its measures and its results, failure becomes visible and fixable. If it does not, a program that achieves little can run for years because no one is required to say so.
Corruption and concealment need the dark
Corruption, favouritism and cover-ups all depend on the same thing: that the people affected cannot see what is happening. Transparency does not make wrongdoing impossible, but it raises the cost of it, because every decision is made knowing it may be read later. The clearest recent Australian case is Robodebt. The Royal Commission found that the scheme issued more than $1.7 billion in unlawful debt notices and that its net cost to the Commonwealth was about $565 million.
It also found dishonesty and collusion to stop the scheme’s lack of a legal basis from coming to light. Legal scholar Darren O’Donovan notes that cabinet confidentiality and similar exemptions meant political accountability could not be fully explored without a Royal Commission. In March 2026 the National Anti-Corruption Commission found that two of the six people referred to it had engaged in serious corrupt conduct, according to a statement from independent MP Kate Chaney, who also criticised the absence of public hearings.
The tools meant to expose such failures are themselves under strain. An audit reported in May 2026 found major departments, including the Prime Minister’s own, were routinely refusing freedom of information requests. Of 25,211 decisions it examined, a quarter were full refusals and 54 per cent were only partly granted. This is not a problem of one party.
The audit office has criticised grant processes under governments of different persuasions, including a $444 million grant to a small reef charity that it found had not complied with transparency rules. Secrecy is a habit of power, not of any one side of politics.
Business holds public power too
Transparency is not only a duty of governments. Large companies shape what we pay, what we breathe, what is safe and which laws get written. They sell essential goods, win public contracts, receive public money and fund politics. Where they exercise that kind of power, the public has a legitimate interest in seeing how. Consider three examples. Who owns what. Australia still has no public register showing who ultimately owns and controls unlisted companies.
The government has committed to build one, but its own announcement says detailed policy work will not begin until early 2027. Until then, regulators, journalists and competitors often cannot see who is behind a company. What is claimed to customers. In May 2026 the Federal Court found that Coles’ “Down Down” promotions were not genuine discounts, accepting the ACCC’s case. Shoppers could not check the price history themselves. It took a regulator, years and a trial to establish what the pricing record showed.
A parallel ACCC case against Woolworths was still before the court at the time of reporting. Who funds politics. Under current rules, donations are reported through annual returns, so the public may learn who gave money months or more than a year after an election. Federal reforms passed in 2025 lower the threshold and speed up disclosure, but the new scheme only starts on 1 January 2027. None of this requires exposing trade secrets or private lives.
It requires that facts the public needs to judge power, such as ownership, public contracts, lobbying, political donations and verifiable claims to customers, are published in usable form and on time.
Medicine and AI: where secrecy costs the most
Some industries hold information whose concealment does not just waste money. It slows or distorts progress on human health and on technology that will shape everyone’s lives. Drug development and artificial intelligence are the two clearest cases, and the argument here is that they should be held to the strongest transparency standard of any sector. Drugs and treatments When a trial result is never published, doctors, patients and other researchers lose evidence.
A failed approach gets repeated, a harm goes unnoticed, and a promising lead may never be followed. Yet non-publication is common. One study of a random sample of 400 US trials found that 39 per cent had not made results public more than three years after completion, including 40 per cent of trials funded by the National Institutes of Health.
An earlier study of trials submitted to the US drug regulator found that trials with favourable results were more likely to be published, which skews the evidence doctors see towards good news. The common defence is commercial: companies spend heavily on research and must protect their return. But the patent system already offers a trade, a temporary monopoly in exchange for public disclosure of the invention. Withholding trial data, negative results or safety findings goes beyond protecting an invention.
It protects a product’s image at the expense of the people who may take it, and of competitors and academics who could build on the findings. Volunteers who joined a trial also took on risk on the understanding that their contribution would add to knowledge. A strong standard would require:
- registration of every trial before it begins, and publication of full results, including failures, within 12 months, with the data available to independent researchers
- publication of the full clinical study reports submitted to regulators, not just journal summaries
- disclosure of public funding and of who owns the resulting patents and licences
- real penalties, since compliance is weak where enforcement is rare Artificial intelligence AI systems are being built by a handful of companies and deployed into health care, education, hiring, finance and government. The public is asked to trust their safety claims without seeing the tests behind them.
- The principle should be the same as for drugs: before a powerful system is released, the public and independent experts should be able to see what it was trained to do, how it was tested, what went wrong and what the developer knew. The law is starting to move. California’s Transparency in Frontier Artificial Intelligence Act took effect on 1 January 2026.
- It requires large developers to publish safety frameworks and transparency reports and to report serious safety incidents, and New York enacted its own safety-protocol law in December 2025. But California’s law lets companies redact material to protect trade secrets, among other grounds. Critics of such regimes can fairly ask who decides what is a trade secret and what is simply embarrassing.
- The case for going further is the same as with medicine: if a system’s failures can harm the public, the evidence about those failures belongs to the public, including independent evaluations, incident reports, and meaningful information about training data. The limits, again Two cautions keep this honest. Participant privacy matters, so trial data should be shared in de-identified form.
- And some technical details, such as the inner workings of systems that could help someone build biological or cyber weapons, fall under the security exclusion this article already accepts. Outside those narrow areas, the burden should be on the company to justify every secret.
What real transparency looks like
The principle is simple: information about the use of public money and public or economic power should be public by default, and secrecy should have to be justified, not assumed. In practice that means a short list of concrete commitments. Publish by default. Contracts, grants and subsidies should be listed in open, searchable data: who got the money, how much, on what criteria and who decided. No outcome, no renewal.
Every funded program should state measurable goals before it starts, be independently evaluated, and publish the result. Funding should depend on what was achieved, not on what was announced. See who is influencing whom. Lobbying contacts, ministerial diaries and political donations should be disclosed promptly, not months later. Make information requests work. Freedom of information needs adequate staffing, firm deadlines, effective independent review, and a duty to keep proper records so there is something to release. Show who owns companies.
Public registers of beneficial ownership should cover companies and trusts. Protect those who speak up. Whistleblowers inside agencies and firms are often the first to see failure, so they need real legal protection. Write for people. Publish plain-language summaries alongside the data, so access does not depend on having a specialist or a lawyer. Assets and income The strongest version of the argument applies to everyone, not only officials and companies.
Without visible income and assets, the public cannot judge whether tax is fair, whether wealth and political influence are concentrating, or whether someone’s lifestyle matches what they declare. Personal data such as health, family life and private communications is a different matter and stays private. Income, assets and tax paid are financial facts about how a society shares its burdens and rewards. There is a working example. Norway has long published individuals’ income, wealth and tax, and Sweden and Finland publish narrower versions.
Researchers found that when Norway moved the data online, reported business income rose by about 3 per cent where the change made returns much easier to see, suggesting disclosure deters evasion. The same research notes that disclosure also corrected misperceptions about the income distribution, though lower-income people reported lower life satisfaction.
The risks are real and the design matters: Norway found that curiosity was driving much of the traffic. After it stopped allowing anonymous searches in 2014, searches fell by about 90 per cent. Logging who looks is a cheap safeguard. Privacy advocates argue that published wealth data helps criminals pick targets. Some Swedish crypto holders have linked attacks on them to the tax lists.
A sensible system would publish income, assets and tax paid, but not home addresses, birth dates or identity numbers, and would limit bulk downloading. Whether the benefits justify the loss of privacy is a genuine dispute. Countries such as the United States treat individual tax returns as confidential, and many people there and elsewhere regard that as a basic right. The limits Transparency has legitimate boundaries, and a credible case for it names them. This article excludes two areas: Security.
Operational national security, intelligence sources and active law enforcement need confidentiality to work. Personal data. Health records, family life, private communications, home addresses and identity numbers stay private. Income and assets are treated differently, as argued above. A company’s wish to protect its own advantage is not a reason for secrecy. An exception is justified only where disclosure would harm the public, and it should be narrow, time-limited, explained in writing and reviewable by someone independent of the body that wants secrecy.
The objections, and the replies
There are serious arguments against pushing transparency further, and they deserve a straight answer. “Officials need to give frank advice.” Defenders of cabinet and deliberative confidentiality argue that if every draft and email can be released, public servants will stop writing things down or soften their advice. The reply is that the remedy for a chilling effect is rules on timing and narrow exemptions, not blanket secrecy, and that the Royal Commission into Robodebt showed what unrecorded or hidden advice can cost.
“It costs too much and slows government.” Processing requests, publishing data and evaluating programs takes staff and money, and the government has argued that its own freedom of information reforms aim to modernise and streamline the system, though critics from several parties and the crossbench argued they would reduce access.
The reply is that publishing by default is cheaper than answering requests one by one, and that the cost of evaluating a program has to be weighed against years of funding one that no one has checked. “Sensitive information must be protected.” For security and personal data this is accepted, which is why the limits above exist. Commercial advantage alone is not a public interest, and a company’s wish to keep something to itself is not a reason to withhold it.
“Data dumps are not understanding.” Raw information can overwhelm, be gamed, or be misread. The reply is that this argues for better presentation and for independent bodies such as auditors, journalists and researchers to interpret it. It does not argue for withholding it. “Some benefits are hard to measure.” Defenders say research, prevention and defence pay off slowly, unpredictably or only by averting something that never happens.
The reply is that most of these can be measured in general terms: a research portfolio can be judged on its overall results, prevention by comparing groups or trends, and deterrence by assessing threats and risks over time rather than counting equipment used. Imprecision calls for sensible measures and longer time frames, not for exemption. A program that cannot say what success would look like should have to explain why it is funded.
“Companies will stop investing if they cannot protect their data.” Drug and AI developers sometimes argue that disclosure weakens the incentive to invest. The reply is that companies have no real choice. A firm that stops developing products loses its market to competitors, so competition keeps investment going whatever the disclosure rules are. Results kept under the bed earn nothing, and the money is made by selling a product, not by hiding evidence about it.
In any case, a company’s wish to protect its own advantage is not a public interest. The test is social benefit: information is withheld only where disclosing it would harm the public, never merely because it would inconvenience a company. The public also often helps fund the underlying science, so it has its own claim to the results.
Informed participation is the foundation
Democracy is often described as a matter of voting every few years. In fact it is a continuous process of noticing, questioning and correcting, and that process runs on information. When citizens cannot see how decisions were made, what money achieved or who benefited, they are left to rely on the account of whoever has the most to gain from telling it. Transparency will not fix every failure, and it can be done badly.
But the examples above point the same way across governments of different parties and across the private sector: failures last longest where no one outside the institution can see them. Opening the books, with narrow and honest limits for security and private lives, does not weaken government or business. It gives the public the thing it needs to hold either to account.
Sources
SmartyGrants: Institute of Grants Management white paper
ANAO: Community Health and Hospitals Program audit
The Point: FOI audit, May 2026
Treasury: beneficial ownership register
AAP: donation disclosure delays
AEC: funding and disclosure reform
Figures and findings are as reported in these sources as of November 2026 and may have changed.