Everyone Is at the Table. What Changed?

SALVATOR MUNDI TIMES | GLOBAL EDITION | MEDIA & COMMUNICATION

Analysis · Series: The Political Economy of Information, IV

Councils, think tanks, broadcasters and communication programmes across the world exist to shape how people understand things. The ones examined here can usually tell you what they produced. Far fewer can demonstrate what changed because of it — and the two clearest exceptions found are less flattering than they look.

By Dr Metin Oktay, Chairman, The Global Council on Media and Communication, Salvator Mundi Global

A question I have reason to avoid

Across countries, there are organisations that call themselves councils, institutes, foundations, task forces, working groups or think tanks. They employ credentialed people, convene conferences, publish reports, brief governments, issue recommendations and cultivate access to decision-makers. Their members are not usefully dismissed as fools or idlers. None of that answers the only question that actually matters: what exists because the organisation exists, that would not exist without it?

I ask partly because I chair one.

That is not a confession offered for effect. It is the reason the question in this piece has to survive contact with my own position before it can be used against anyone else's. Governments fund this kind of activity, and so do foundations, corporations and private donors, and nothing that follows treats government money as more suspect than private money, or the reverse. The question is the same regardless of who is paying: not whether the people involved are intelligent or well connected, which they usually are, but whether their existence changed anything that their absence would not have changed just as well.

What these organisations are actually for

Strip away the institutional vocabulary and the organisations examined here claim some blend of five functions. They produce expertise: original analysis nobody else has done. They transmit knowledge: carrying evidence that already exists to people positioned to act on it. They seek influence: trying to change a decision, a vote, a policy or a purchase. They convene: bringing people into a room, physical or otherwise, who would not otherwise have met. And they confer legitimacy: lending the appearance of independent, credentialed endorsement to a position someone already held.

These are different businesses, and each should be judged against a different counterfactual. Expertise should be judged on originality nobody else supplied. Transmission should be judged on whether knowledge reached people who could use it and would not otherwise have had it. Influence should be judged on whether a vote, a policy, a purchase or a behaviour actually changed. Convening should be judged on what became possible that otherwise would not have. And legitimacy, people we trust say this is sensible, should simply be named as what it is, because that is not the same as any of the other four and is rarely audited as its own category at all.

Almost every organisation examined for this piece operates in more than one of these categories at once, and each one reports the numbers from whichever category is easiest to measure, regardless of which one it is actually claiming credit for. That mismatch is the pattern this piece keeps finding.

Reach is not the currency it is spent as

Consider one of the most literal cases: an organisation whose entire purpose is communication, funded to project one country's account of events into environments where it competes directly with others. The U.S. Agency for Global Media, which runs Voice of America, Radio Free Europe/Radio Liberty, Radio Free Asia and related networks, reported a record 427 million weekly measured global audience for fiscal year 2024, nearly double what it reported a decade earlier, on a budget that has run between roughly $800 million and $950 million a year across recent fiscal years.

Reach at that scale is a genuine achievement, and it is not nothing: audiences exist that would otherwise have no access to this content, particularly in countries the agency itself notes are among the most restrictive media environments in the world. But reach is also the easiest number in this field to produce and the hardest to translate into anything else. Nothing in the agency's own reporting establishes what audiences did differently, believed differently or decided differently as a result of that exposure, and its own methodology admits it cannot even measure some of its most strategically important audiences at all, including in China, North Korea, Tibet and the Uyghur-majority regions of Xinjiang, for reasons of state restriction rather than agency failure.

This is not a new observation. The Government Accountability Office made a nearly identical finding about the State Department's wider public diplomacy effort in four separate reports issued between 2003 and 2010, each time noting that performance measurement relied on anecdote and activity counts rather than evidence of any actual shift in foreign audiences' understanding or opinion.

A separate and genuinely distinct problem surfaced in the agency's own fiscal year 2025 financial statements: the independent auditors issued a disclaimer of opinion, meaning they could not obtain sufficient evidence to form a judgment on whether the accounts were accurate at all. In fairness, those accounts were produced while the agency was being pushed toward what its own report calls a statutory minimum posture, with net cost falling by nearly $145 million as programmes and contracts were cut; a disclaimer produced amid that kind of upheaval is not the same finding as one produced in ordinary conditions, though it remains a finding either way. It is also a finding about financial control, not about communications effectiveness, and the two should not be blurred into each other. What both share is the same gap: this agency can tell you, with precision, how many people it reached. This investigation identified no published estimate of what those audiences would have believed or done in its absence, which is the only number that would actually answer the question this piece is asking.

The contested middle

A step further from broadcasting sits democracy assistance, where the National Endowment for Democracy is a prominent U.S. example: a nominally independent nonprofit that received a $315 million congressional appropriation in 2023, working through four core grantee institutes in more than ninety countries, before the administration's Department of Government Efficiency froze its funding in 2025. NED's own account describes itself as one of the most cost-effective instruments of American foreign policy. A 1991 Government Accountability Office review found the opposite kind of problem to USAGM's: not an inability to measure impact, but weak grantee financial monitoring that had allowed some funds to be misused or mismanaged. One long-standing characterisation of the field, offered decades ago in a congressional research account of the debate over NED's funding, holds that demonstrating success in democracy assistance is close to impossible by the nature of the activity, not through any particular institution's failing.

What distinguishes this sector from broadcasting is that the impossibility has actually been tested rather than merely asserted. A systematic review identified forty-six separate studies using instrumental-variable methods to try to isolate the causal effect of democracy aid from the reverse problem that donors often send more assistance to countries already showing signs of democratising, which would make aid and democratisation move together with no causal relationship between them at all. Among studies specifically examining U.S. aid, ten found a positive effect and five found a negative one.

Contested evidence and absent evidence are not the same finding, even when both produce the same verdict of ‘we don’t know.’

That is not a resolved debate. It is, however, a quantified and contested one, which puts democracy assistance in a different evidentiary category from broadcasting reach, where this piece found no comparable attempt to identify a causal effect at all.

Governments trying to grade themselves

Three governments illustrate the same difficulty from the inside, in each case while demonstrating that they know how to evaluate other things perfectly well.

The UK's Foreign, Commonwealth and Development Office published a new evaluation strategy in March 2026 that names diplomatic influencing explicitly as one of its persistent evidence gaps, alongside non-aid spending and corporate activity: a rare instance of an institution naming, in its own strategic planning document, the exact category of activity this piece is examining. The same organisation is capable of rigorous evaluation elsewhere — it published 34 country-level, seven regional and eighteen global evaluations in the three years to March 2025 — and its own account describes an evaluation of a £54 million global health research portfolio that led it to narrow subsequent funding toward programmes more likely to translate research into use. The capacity clearly exists inside the organisation. Its own strategy says plainly that evaluation coverage has not yet extended evenly to diplomatic influencing.

The British Council, the UK's cultural relations body, commissioned an external evaluation of its influencing work in 2025 and 2026 using a combination of outcome harvesting, contribution analysis and appreciative inquiry across four case studies: methods specifically designed for situations where a direct causal claim is judged infeasible, because too many other actors and events are moving at once to isolate one organisation's contribution. That is a serious, honest response to a real measurement problem. It is evidence about how the British Council is trying to know whether its influence work matters, not yet evidence that it has been shown to.

And the European Union's East StratCom Task Force, built in 2015 to counter disinformation and best known for its EUvsDisinfo database, was reviewed by the European Court of Auditors in 2021. The auditors credited EUvsDisinfo with raising public awareness of disinformation, and recommended that the task forces build a proper function to evaluate their own operational effectiveness beyond communication campaigns, a recommendation rather than evidence that the function already existed. The same audit found something more specific and more telling: the EU's Rapid Alert System, built to allow coordinated identification and response to disinformation campaigns across member states, had never actually issued an alert in the manner its own design intended.

A mechanism can be built, funded, and never used for its stated purpose — and the auditors were the ones who noticed.

Three governments, three different structures, and in each case the same shape of gap around the same kind of activity. That is not enough evidence to call it a law of the field; three cases establish a pattern worth testing further, not a structural finding about how communication institutions behave everywhere. But it is a large enough coincidence, across enough different bureaucracies, to take seriously rather than explain away.

Money that is well documented, and influence that isn't

The money is particularly well documented for privately and internationally funded think tanks; the claimed influence is not.

An investigative accounting from the Center for International Policy found roughly $174 million in foreign government funding flowing into the fifty largest American think tanks between 2014 and 2018, from eighty separate foreign governments; the same report found the largest recipients over that window to be the World Resources Institute, the Center for Global Development, the Brookings Institution, the Atlantic Council and the Aspen Institute, and the largest donor governments to be Norway, the United Kingdom, the United Arab Emirates, Germany and Sweden. A separate and more recent tracker from the Quincy Institute, covering a different and more recent five-year window, found over $110 million in foreign government funding to the same top fifty institutions, rated more than a third of them fully opaque about their funding, and found the Atlantic Council, Brookings and the German Marshall Fund the largest recipients in that later period. The two counts measure different years using different methods and should not be added together as one continuous total, but both independently found a large, only partly disclosed river of foreign government money running through the same handful of institutions.

Qatar's relationship with Brookings is one of the cases examined here in the most detail, and it is more interesting for what both sides said than for who was right. Qatar's own foreign ministry, announcing its 2012 memorandum of understanding with the institution, described the partnership as delivering a public-relations outlet that projected the bright image of Qatar in American media: a foreign government stating its own communications objective in its own words. Brookings says it ended its Qatar funding in 2017 and closed the Doha Center the relationship had supported, that foreign government money currently makes up under five per cent of its budget, and that it maintains internal policies against any activity that would compromise its research independence or require registration as a foreign agent. Those policies were tested five years later in a way Brookings did not choose: in 2022, its own president was placed on administrative leave and subsequently accused by federal prosecutors of illegally lobbying on Qatar's behalf while still a senior fellow at the institution, the precise activity Brookings' own rules prohibit. Two Senate offices pursued the matter through formal oversight letters.

Whether the safeguards were sufficient depended on an individual, not a policy document — which may be the most honest thing this investigation found about how institutional independence survives contact with money and access.

Underneath the funding question sits a harder one that money cannot resolve either way: does research from a think tank change policy? Methodological work examined for this piece emphasises how difficult direct policy influence is to identify with confidence. Policy emerges from many actors moving at once; a citation, a meeting or a paragraph appearing in draft legislation are all consistent with genuine influence and equally consistent with a think tank having said something a policymaker already believed and would have acted on regardless. One recent reframing of the field, offered by people who study think tanks rather than fund or criticise them, put it plainly: influence may be the wrong thing to claim credit for at all, and usefulness, supplying analysis decision-makers would otherwise have had to produce themselves, is the more honest and more measurable standard. That is not a comfortable finding for an industry that raises money by claiming the stronger word.

What survives a counterfactual test

None of this should be read as an argument that communication is inherently unmeasurable, because two results in the material gathered for this piece show plainly that it is not, and both are more unsettling than flattering to the institutions that would like to claim their spirit.

In Rwanda, in the aftermath of the genocide, with survivors and the accused living side by side, the psychologist Elizabeth Levy Paluck ran a genuine randomised experiment: twelve communities were assigned, at random, to listen weekly either to a reconciliation-themed radio drama or to a health-themed drama used as the control. After a year, listeners to the reconciliation drama showed measurably greater tolerance of intermarriage across ethnic lines, more willingness to voice open dissent, and more openness to discussing personal trauma, compared with the control group. The result that should trouble anyone in the communication business, rather than flatter it, is what the drama did not do: it changed what people said and did far more than it changed what they privately believed.

Norms and behaviour moved. Belief barely did.

A field experiment published this year in the American Economic Review makes a related and sharper point about a completely different kind of communication: getting evidence in front of policymakers. Researchers in Spain sent an identical, nonpartisan, low-cost, evidence-backed policy recommendation to local governments, varying only who delivered it: an ideologically aligned think tank or newspaper, an ideologically opposed one, or a prestigious but ideologically neutral research institution. When the messenger shared the recipient's ideological alignment, adoption of the recommended policy rose by more than sixty-five per cent compared with municipalities that received no information at all. When the messenger was ideologically opposed, the effect was statistically indistinguishable from sending nothing. And the neutral, prestigious institution, the messenger every council, institute and think tank implicitly claims to be, produced roughly half the effect of the aligned messenger, and could not be shown to outperform sending no information at all with any confidence.

Within this experiment, what changed was not the evidence, which was held constant across every arm, but who delivered it. Ideological alignment strongly affected whether the same evidence translated into policy adoption; credentials and neutrality did not come close to matching that effect. If that generalises even partially beyond one Spanish field experiment, the actual product many of these institutions may be selling is not knowledge. It is trusted permission to act on knowledge you already wanted to act on, which is a real and sometimes valuable service, but a different one from the one usually claimed, and it should be evaluated as what it is rather than dressed in the vocabulary of expertise.

The test, applied to the chair I'm sitting in

So the test that opened this piece is the only one worth keeping: not who is in the room, how credentialed they are, or how the organisation describes itself, but what exists because it exists that would not exist without it. Applied honestly, most of what this piece examined cannot yet answer that question, including, I will say plainly, the council I chair. Its one output I can actually assess without guessing is this series of articles, built over several months with extensive adversarial verification, and even that has not yet been asked the only question that matters: whether anyone outside the people who produced it reads it, and whether anything changes because they did. I don't know the answer yet. Neither, on the evidence gathered here, do most of the organisations that would confidently tell you theirs is different.

None of this means the people inside these institutions are wasting their time, or that the work is worthless, or that the money is stolen. Rwanda and Spain both show that communication, designed and tested properly, can move something real. What the rest of the evidence shows is narrower and more uncomfortable: most of what this piece examined cannot answer that question, and little of it shows any sign of having been asked to.

Everyone is at the table.

The honest question was never who they are. It is what the room would look like with the chairs empty — and how few of the organisations examined here have tried to find out.

Fourth in a series on the political economy of information. The first, 'When Meaning Becomes Scarce', examined information integrity as economic infrastructure; the second, 'When Both Sides Use the Same AI', examined machine-mediated diplomacy; the third, 'The Unmeasured Dividend', examined the economic returns claimed for cultural influence.

References

'National Endowment for Democracy: Policy and Funding Issues', congressional research account of NED funding debates.

Brookings Institution, 'Administrative updates to Brookings leadership', news releases, 10 June 2022 and 24 June 2023.

Brookings Institution, 'FAQs about Brookings' approach to foreign funding', brookings.edu.

Center for International Policy, Foreign Influence Transparency Initiative, foreign government funding of the top 50 US think tanks, 2014–2018, reported via Responsible Statecraft, 29 January 2020.

European Court of Auditors (2021) Special Report 09/2021: Disinformation affecting the EU: tackled but not tamed. Luxembourg: ECA.

Garcia-Hombrados, J., Jansen, M., Martínez, Á., Özcan, B., Rey-Biel, P. and Roldán-Monés, A. (2024/2026) 'Ideological alignment and evidence-based policy adoption', American Economic Review (forthcoming); IZA Discussion Paper No. 17007.

Grassley, C. et al., letter to the US Department of Justice regarding FARA compliance of the Brookings Institution, 2022.

Investigative Project on Terrorism, 'Qatar's insidious influence on the Brookings Institution', quoting a 2012 statement by Qatar's Ministry of Foreign Affairs, 28 October 2014.

National Endowment for Democracy (2025) Fact Sheet: NED and the 2026 Discretionary Budget Request. Washington, DC: NED.

On Think Tanks, 'Not influential but useful: rethinking how we assess and support think tanks', 27 March 2025.

Paluck, E.L. (2009) 'Reducing intergroup prejudice and conflict using the media: a field experiment in Rwanda', Journal of Personality and Social Psychology, 96(3), pp. 574–587.

Quincy Institute for Responsible Statecraft (2025) 'Big Ideas and Big Money: Think Tank Funding in America', Think Tank Funding Tracker.

Responsible Statecraft, 'Feds accuse Brookings president Gen. Allen of illegally lobbying for Qatar', 8 June 2022.

Samarin, A. (2021) Does Aid Support Democracy? A Systematic Review of the Literature. WIDER Working Paper 2021/14. Helsinki: UNU-WIDER.

United Kingdom, Foreign, Commonwealth and Development Office (2026) FCDO Evaluation Strategy 2026 to 2030. London: FCDO, 10 March 2026.

United States Agency for Global Media (2025) FY 2024 Agency Performance Report and Audience and Impact Overview. Washington, DC: USAGM.

United States Agency for Global Media (2026) Agency Financial Report Fiscal Year 2025. Washington, DC: USAGM.

United States Government Accountability Office (1991) Promoting Democracy: National Endowment for Democracy's Management of Grants Needs Improvement, NSIAD-91-162. Washington, DC: GAO.

United States Government Accountability Office, public diplomacy performance-measurement findings, reports issued 2003, 2006, 2009 and 2010. Washington, DC: GAO.

Warren, E., letters to the Brookings Institution regarding foreign funding disclosure, July and October 2022.

Note on sourcing: this piece separates documented facts (funding figures, audit findings, published evaluation strategies) from causally identified findings (the Rwanda and Spain experiments, and the instrumental-variable literature on democracy aid) and from contested characterisations and institutional self-description, which are named as such throughout. Two separate think-tank foreign-funding datasets, covering different and non-overlapping years, are reported side by side and are not additive. The British Council evaluation is described from secondary reporting rather than the underlying report, which has not been independently examined. Where a detail is disputed, such as the precise timing of Brookings' Qatar funding across its own public statements, the piece states what is agreed and flags what is not, rather than resolving it.

About the Author

Dr. Metin Oktay

Chairman of the Global Council on Media and Communication, Salvator Mundi Global

Founder of M3TIN Records and Chairman of M&H Logistics, an international freight forwarding company. His career spans media relations and the entertainment sector, built on senior executive positions in aerospace, energy and mining.

https://www.linkedin.com/in/metinoktay/