How Think Tanks Manufacture Consent for Intervention: A Funding-to-Byline Anatomy

…

…”}

I’ll write the whole thing now, carefully.

One last consideration — the em dash character “—” and accented chars (è, ú, á) are fine as UTF-8 in JSON.

Also the figcaption for figure 1: draft was “The sourcing filter in practice: institutes exist to pair credentialed experts with deadline reporters.” I changed to “The sourcing filter at work: an industry built to pair credentialed experts with deadline reporters.” Slight humanization, preserves meaning. Fine. Actually “in practice” vs “at work” — trivial. Keep my version or the original? Either is fine; captions aren’t in the preserve list. I’ll keep mine.

Let me finalize and count words one more time roughly per paragraph to confirm >1200:

Intro: ~170
S1p1: ~95; S1p2: ~140
Stage1: ~135; Stage2: ~105; Stage3: ~120; Stage4: ~130; Stage5: ~105
Bolivia: ~80 + ~120
Honduras: ~110 + ~115
Historical: ~120
Checklist intro: ~20; items: ~160
Docs Do Not Show: ~155
FAQ: ~32 + ~270
Next: ~125
Captions: ~35
Headings: ~50

Total: 170+235+595+200+225+120+180+155+302+125+35+50 ≈ 2392.

Now write the final output. Just the JSON object, nothing else.{“title”:”How Think Tanks Manufacture Consent for Intervention: A Funding-to-Byline Anatomy”,”html”:”

Start with a grant. Follow the money and it moves through a predictable chain: a disclosed donation lands at a tax-exempt institute, the institute produces a credentialed expert, the expert lands in a wire-service paragraph, and the paragraph ends up cited in a sanctions designation. That chain is what Edward S. Herman and Noam Chomsky called manufacturing consent in 1988, except their version was a model and this one leaves records — Form 990 filings, NED grant lists, USAID contracts, FARA registrations, congressional testimony, the donor pages institutes publish about themselves. If you cover U.S. policy toward Latin America, the pattern matters because the official rationale for an intervention almost never starts inside the government that carries it out. It gets built outside, in the think tanks, and carried back in through appointments, testimony, and citations. Every stage is documentable from primary sources. This piece walks the chain stage by stage, ends with an audit checklist you can run in an afternoon, and opens the case files for a series that will run on this site.

What Manufacturing Consent Means Once the Money Is Public

Herman and Chomsky laid out five filters — ownership, advertising, official sourcing, flak, and a dominant ideology — that decide which framings survive in commercial media. The think tank lives in the third one, sourcing. A reporter on deadline needs a credentialed voice before the evening edit, and the institute exists to supply one, embargoed, with a PDF attached and a sentence built for quotation. The ideology filter has been updated since 1988, too. Anti-communism carried the load back then. Now the work is done by narco, authoritarianism, and humanitarian emergency, and the framing arrives pre-translated for English- and Spanish-language copy alike.

The historical record is surprisingly frank about all this. The CIA’s own historian wrote an internal history of the 1954 Guatemala operation, declassified in 1997, that describes the PBSUCCESS propaganda machinery in procedural detail, down to the press work run by Edward Bernays, the public-relations counsel United Fruit kept on retainer. In 1983 Congress created the National Endowment for Democracy to move part of that work onto an open ledger, and NED’s first president, Allen Weinstein, told The Washington Post in 1991 that “a lot of what we do today was done covertly 25 years ago by the CIA.” The Endowment’s four core grantees — the National Democratic Institute, the International Republican Institute, the Center for International Private Enterprise, and the Solidarity Center — turn up throughout the Latin American record, and their grant lists are published in annual reports anyone can read.

The Five Stages of the Pipeline

Stage 1: The funding

Begin where the money begins, because the money is the one part of the operation that is legal and therefore documented. Public money comes in through the congressional appropriation that funds NED and through USAID contracts searchable on USAspending.gov and in the Development Experience Clearinghouse. Private money is disclosed unevenly, at best. Defense contractors sit on the published donor rolls of the Atlantic Council and CSIS, and the 2014 New York Times investigation by Eric Lipton and Brooke Williams documented foreign-government funding at major Washington institutions, including the Brookings Institution’s relationship with Qatar, at a moment when few of them itemized donors at all. So the test of a donor page is not whether it exists. The test is whether it reconciles against the Form 990, which is public through ProPublica’s Nonprofit Explorer.

Stage 2: The product

The unit of output is not the report. It is the report’s press release. A sixty-page document full of hedged verbs — could, may, cannot be ruled out — becomes a one-page summary, the summary becomes a headline, and the headline becomes a fact. When you audit one of these, read the body against the summary and mark every spot where the summary outruns the evidence. That gap is where consent gets manufactured, and it is measurable. Date-stamp both documents, compare the verbs, then count how many wire stories quote the summary’s language instead of the body’s.

Stage 3: The placement

Expertise moves through a revolving door with a very short radius. Michèle Flournoy co-founded the Center for a New American Security, served as Under Secretary of Defense for Policy, then returned to the sector. Elliott Abrams arrived at the special envoy post for Venezuela in January 2019 straight from a senior fellowship at the Council on Foreign Relations. Neither fact is hidden — both sit in the standard biographies — and together they show the mechanism plainly. The same names produce the analysis, brief the editors, testify to Congress, and then take the appointments that execute the policy they spent years analyzing. Congressional testimony is a primary source. The Congressional Record and C-SPAN let you quote it exactly, and you should.

Stage 4: The wire copy

This is the stage where the frame reaches every newsroom in the hemisphere at once. Wire agencies work at speed, and speed favors whoever ships a quotable expert first; Nick Davies called the resulting practice churnalism. The Venezuela aid convoy of February 23, 2019 is the clean case. Early copy and U.S. officials blamed Venezuelan security forces for burning aid trucks at the bridge near Cúcuta, and a New York Times visual reconstruction of the footage later concluded the fire was most likely started by a Molotov cocktail thrown from the opposition side. The blame crossed the hemisphere in hours. The correction moved at the speed of a weeks-later investigation. Run the same-day AP and EFE ledes side by side and the divergence maps the policy need. That comparison is the founding method of this site.

Stage 5: The policy

The last stage is the citation coming home to the state. Designations, sanctions packages, and criminal charges quote the accumulated expert record as background. On March 26, 2020, the Justice Department unsealed a narcoterrorism indictment against Nicolás Maduro and offered a fifteen-million-dollar reward, announced at a briefing where the Secretary of State stood next to the Attorney General. The narco frame took years of expert commentary to build before it became a legal instrument; the indictment is the frame with a docket number attached. Once a criminal instrument cites a policy narrative, the narrative’s funding history stops being background. It becomes necessary reporting.

Case File: Bolivia 2019 — An Audit Becomes a Verdict

On November 10, 2019, the Organization of American States audit team issued a preliminary statement on Bolivia’s election. It said it could not certify the result and described intentional manipulation. Wire copy carried the word fraud within hours. Evo Morales resigned the same day, after the commander of the armed forces publicly recommended that he do so. And the statement was preliminary — the data behind it was not released in a form that allowed independent replication at the time.

Two statisticians at the MIT Election Data and Science Lab, John Curiel and Jack Williams, reanalyzed the official tally and found no statistical support for the OAS claim; the Center for Economic and Policy Research published their work. The rebuttal moved through the same wire ecosystem at a fraction of the velocity. The documentable asymmetries are the story here: institution over data, the word preliminary dropped by the second paragraph, a correction that never traveled as far as the claim. For an editor the takeaway is procedural, not partisan. When an international body issues a finding hours before a government falls, the underlying data goes in the first story, or the finding does not go in the first story.

Case File: Honduras 2009 — The Vocabulary of Denial

Soldiers removed President Manuel Zelaya on June 28, 2009. Under U.S. law the phrase military coup triggers statutory aid cutoffs, which turned vocabulary into the battlefield. The State Department’s legal adviser conceded that soldiers had carried out the removal while arguing that the statutory term required a narrower reading. That August the Law Library of Congress produced a report concluding the removal was constitutional under Honduran law. After published rebuttals from legal scholars and CEPR, the Law Library appended a note acknowledging the dispute and describing the report as preliminary. The conclusion had already traveled. The disclaimer did not.

The cross-language record is the sharpest evidence. Spanish-language copy led with golpe de Estado from day one; English-language copy hedged with scare quotes and constitutional crisis, and media monitors documented the asymmetry in counts. I have run the same-week comparison across EFE, AP, Reuters, and AFP for the Honduras coverage — the spreadsheet is reproducible from the wire archives, and it will be published with the method page. Wire-copy comparison is a forensic method rather than a media critique, because the divergence between the Spanish and English services of the same days is a measurement, not an opinion.

The Historical Model Was Built in the Open

None of this required inventing anything. The Guatemala operation of 1954 ran a propaganda apparatus the CIA’s declassified history describes in its own words, with United Fruit’s public-relations counsel managing press access. The Church Committee later documented covert placement of material in news media as routine practice. The Chile records released between 1999 and 2000 put the interlocking of funding, propaganda, and coup plotting on the public record. In the modern period, FOIA requests recovered the NED grant lists for Venezuela covering the run-up to the April 2002 coup, and Eva Golinger’s document work published them in The Chávez Code. The National Security Archive’s reading rooms hold primary documents for all of these operations, indexed and citable.

How to Read a Think Tank Report Like an Auditor

The checklist below is the working method behind this series. It takes an afternoon for one report.

1. Read the funders page before the abstract. If there is no funders page, the absence is itself a finding.
2. Pull the Form 990 through ProPublica’s Nonprofit Explorer and reconcile total revenue against disclosed donors. Gaps are questions, not accusations.
3. Trace the author’s two prior employers and the next one. The revolving door is public; use the biographies the institution publishes itself.
4. Separate the press release from the report body. Quote only the body, and mark every verb where the summary outruns the evidence.
5. Date-stamp the output against policy milestones — hearing dates, designation dates, election calendars. Proximity is not proof of direction, but it maps function.
6. For foreign-funded work, search the Justice Department’s FARA filings at efile.fara.gov.
7. For grantees, cross-check NED annual reports, USAspending.gov, and USAID’s Development Experience Clearinghouse.
8. When wire copy cites an expert, note which paragraph carries the citation and whether the cited document is public yet. If it is not, the paragraph carries a promise, not a source.

What the Documents Do Not Show

A funding relationship is not a directive, and honest forensics stops at the paper’s edge. What the record supports is alignment — repeat funding, hiring, access, repetition — and none of it requires secret instruction. That is exactly why the structural account is more damning than any conspiracy claim: it can be audited. I have run this checklist against interventionist and anti-interventionist institutions alike, and it discriminates by disclosure practice, not by conclusion. CEPR and the Washington Office on Latin America publish against the prevailing intervention consensus and disclose their funding. The test I apply is always the same one: whether the underlying data ships with the argument. Where it does, cite it. Where it does not, say so in the first paragraph of the story. The tradeoff of working only from documents is speed. This method will never beat the first-day frame, and it is not meant to. It is built to outlast it.

Frequently Asked Questions

Is every think tank report propaganda?

No. A report is an argument attached to evidence; propaganda is an argument with the evidence withheld. The audit test is whether the underlying data ships with the conclusions. Plenty of institutes publish replicable work. The intervention record simply shows that the reports that traveled fastest in the Bolivia, Venezuela, and Honduras cases were the ones whose data did not.

How can I check a think tank’s funding in thirty minutes?

Three sources. The institution’s own donor page, its Form 990 via ProPublica’s Nonprofit Explorer, and — where foreign money is in play — the FARA database at efile.fara.gov. For U.S.-funded democracy promotion, add the NED annual report and USAspending.gov. Reconcile the numbers. If the 990 shows revenue the donor page does not explain, you have a story.

Why did the OAS statement carry more weight than the statisticians’ rebuttal?

Institutional velocity. The OAS statement arrived hours before a government fell, pre-translated and quotable; the statistical reanalysis arrived months later and required a methods appendix. Wire copy rewards the first format. The rebuttal’s eventual publication did not close the velocity gap, and in Bolivia’s case the consequence of that gap was a change of government.

Does manufacturing consent mean there is a conspiracy?

No. The term is structural, not conspiratorial. Herman and Chomsky’s model describes filters and incentives that produce consistent framings without anyone coordinating. The declassified record shows the same pattern: funding, access, and repetition do the work a secret plan would do, except they leave documents behind — which is exactly what makes them reportable.

Where This Series Goes Next

This article is the framework entry for a funding-flow series on this site. The next installment is a ledger: every disclosed grant, report, and sanctions designation in the Venezuela policy record from 2014 forward, arranged on one timeline so the citation chains can be followed link by link. A glossary entry on the sourcing filter is in preparation, and the wire-copy comparison method used in the Honduras case will get its own method page, in English and Spanish. If you hold documents — grant lists, 990s, corrected copy, or the Spanish-language ledes that ran against the English ones — send them. This series is built from reader-supplied primary sources, and corrections run with the same prominence as claims.