This page exists so that nothing in the press-integrity argument has to be taken on trust. Each figure is attributed to the study or brief that produced it, and each is presented with the caveats its own authors attached to it.
Where the evidence supports the argument, it is stated. Where it constrains the argument, that is stated too.
Full citation. Mel Bunce and Beth Pearson, The Value of Journalism: Global evidence on why media matters to economies, national security and crises. Evidence brief commissioned by DW Akademie, IFPIM and UNESCO as members of the Research Working Group of the Media Viability Manifesto initiative. Reference CI-2026/IPDC/PI/2.
Where to read it. UNESCO's launch announcement carries the headline figures directly — New report reveals global evidence of independent journalism's economic and security value. A subsequent Member States briefing on the same document is here — UNESCO Spotlight Session, July 2026. The full brief is catalogued under reference CI-2026/IPDC/PI/2.
Authors. Professor Mel Bunce is Director of the Centre for Journalism and Democracy and Professor of International Journalism and Politics at City St George's, University of London. Dr Beth Pearson is a Research Fellow at the same institution.
Nature of the document. It is a synthesis of existing academic literature, not new primary research. Figures below are therefore cited as reported by the brief, with the underlying study named. Where a figure is an estimate with a wide range, the range is given rather than a midpoint.
Standing. The views expressed in the brief are the authors' own and, as the brief itself notes, do not necessarily represent UNESCO, DW Akademie or IFPIM.
Estimated annual global cost of disinformation across financial, social and political markets. The underlying analysis separates financial costs (fraud, scams, market manipulation), social costs (erosion of institutional trust, fragmentation of cohesion, health impacts) and political costs (threats to democratic integrity, foreign influence operations, and the countermeasures deployed against them).
Use it for: establishing that this is a budgeted problem. Do not use it for: precision. It is a modelled estimate with a wide band.
Bunce & Pearson 2026, citing Sopra Steria 2026
Mis- and disinformation in the World Economic Forum's 2026 Global Risks Perception Survey, placing it second among the most pressing challenges facing the world.
Use it for: the security framing. This is a survey of perceived risk among leaders, which is precisely the audience a governance-infrastructure proposal is addressed to.
Bunce & Pearson 2026, citing World Economic Forum 2026
Reduction in circulation of checked disinformation on the largest social platform in a randomised control trial of AFP fact-checking. Audiences who had shared the checked material also became less likely to share disinformation subsequently.
Use it for: the central argument. Correction works, and works modestly. The gap between an eight per cent correction effect and a half-trillion-dollar problem is the space creation-time provenance is designed to occupy.
Bunce & Pearson 2026, citing Cagé et al 2025
Expert estimate of what it would cost to fund healthy public media and secure information environments for citizens worldwide — described in the brief as equivalent to roughly fifteen days of global military spending.
Use it for: the closing ask. Note: this is an advocacy-sector estimate, not a peer-reviewed costing, and should be attributed as such.
Bunce & Pearson 2026, citing Forum for Information and Democracy 2021
Associated reduction in real GDP growth when a country declines on the World Press Freedom Index, across a study of 97 countries — and the brief notes the effect is slow to recover.
Use it for: economic-development audiences. This is the figure that connects press integrity to the language an economic development corporation already speaks.
Bunce & Pearson 2026, citing Nguyen et al 2021
Average increase in municipal deficit associated with newspaper closure in the United States. The same news-desert literature finds corporate financial violations more likely and their penalties higher where no local paper is present, and one study finds companies pollute more after closures.
Caveat the brief itself raises: some scholars warn the news-desert frame risks romanticizing historical local journalism, some of which was poor quality.
Bunce & Pearson 2026, citing Gao et al 2020; Hesse et al 2022; Campa 2018
What people check against when unsure whether an online claim is false, per the 2025 Digital News Report — ranking above official sources such as a government website.
Use it for: the forged-byline argument. If the outlet's identity is the public's verification instrument, then counterfeiting that identity attacks the verification layer itself.
Bunce & Pearson 2026, citing Newman et al 2025
Reported return on investigative journalism through reclaimed funds, improved public services and reduced corruption. The brief cites Hamilton's analysis of 10,000 US investigations, which estimated US$287 in net policy benefits per dollar spent in one probation-system case, and OCCRP's own estimate of $100 reclaimed per $1 of support.
Note: the OCCRP figure is self-reported by the organization.
Bunce & Pearson 2026, citing Hamilton 2016; OCCRP 2026
These are reproduced because a proposal that only cites the evidence flattering to it should not be trusted with a newsroom's chain of custody.
The brief is explicit that journalism's effect is contingent on other functioning democratic institutions: an investigation can expose major corruption and produce no change if no regulator, prosecutor or electorate can act on it. Provenance infrastructure inherits this limit exactly.
The brief notes that most of this literature is drawn from the United States and Europe, that Global South contexts have been neglected, and that findings from single case studies should not be assumed to transfer across media systems and political contexts.
The brief closes by citing research showing that interventions failing to account for local dynamics are frequently ineffective and can cause harm — and elsewhere records that in some conflict-affected communities, distrust plays a protective role.
Any Mirror Protocol document, deck, or outreach message that uses a figure from this page carries the attribution with it. Three rules, applied without exception:
One. Name the source in the sentence, not only in a footnote. "A 2026 UNESCO-commissioned evidence brief estimates…" is defensible under scrutiny; "disinformation costs half a trillion dollars" is not.
Two. Ranges stay ranges. A figure reported as US$355–516 billion is never compressed to a single number for rhetorical convenience.
Three. Estimates are labelled as estimates, and advocacy-sector figures are labelled as advocacy-sector figures. An engineer, a patent examiner, or a policy staffer will check one number; if it holds, the rest of the document is credited, and if it does not, none of it is.
A further UN source addressing the intersection of organized crime, illicit finance and information manipulation is under review for inclusion. Figures will be added here with the same attribution standard applied above once verified against the primary document.
The relevance is direct. Investigative reporting on illicit financial flows is the work most likely to be attacked through fabrication, and the reporters doing it are the ones least able to absorb the cost of an unanswerable accusation. Where the evidence base extends into organized crime, so does the case for creation-time custody.