They Warned Us: America’s Corporate-Political Oligarchy Is Not New

They Warned Us

America’s Corporate-Political Oligarchy Is Not New

By Jonathan Johnston

Posted on September 19, 2026

Research and drafting assistance by ChatGPT (OpenAI).

How concentrated wealth becomes political power—and why American presidents have been warning about it for generations.

In this essay

“Corporations … are fast becoming the people’s masters.”

That was not Bernie Sanders. President Grover Cleveland said it to Congress in 1888 while discussing trusts, monopoly, concentrated wealth, government favoritism, and the growing power of corporations over ordinary Americans. (American Presidency Project)

“Behind the ostensible government sits enthroned an invisible government.”

That was not Donald Trump describing the “deep state.” It appeared in the 1912 platform of Theodore Roosevelt’s Progressive Party. The next sentence called for the destruction of an “unholy alliance between corrupt business and corrupt politics.” (American Presidency Project)

“Government by organized money is just as dangerous as Government by organized mob.”

Franklin Roosevelt, 1936. (American Presidency Project)

And in 1961, President Dwight Eisenhower—former Supreme Allied Commander in Europe, five-star general, hardly an enemy of the American military—used his farewell address to warn about the “military-industrial complex.”

He did not claim that it was an organization secretly commanding the country. His warning was more subtle. The United States had created an enormous permanent military establishment joined to an enormous permanent arms industry, and the resulting combination possessed “total influence—economic, political, even spiritual” throughout American society.

Eisenhower warned against its acquisition of “unwarranted influence, whether sought or unsought.”

Then, almost immediately, he warned of another institutional coupling: the increasing dependence of scientific research on government money and the danger that public policy could become captive to a “scientific-technological elite.” (National Archives)

These warnings came from different eras and political traditions.

Yet across generations they kept describing versions of the same recurring problem:

Concentrated economic power does not remain merely economic.

It seeks political influence because political decisions affect economic outcomes. Political actors seek economic support because campaigns, institutions, expertise, and organizations require resources. Government seeks private expertise because modern society is complicated. Private interests seek government authority because government possesses something that no corporation, however wealthy, can simply purchase on an ordinary market:

the legal authority to make rules coercive.

Money can persuade you.

Money can employ you.

Money can reward you.

Government can require you.

When concentrated private wealth becomes deeply interwoven with public authority, the distinction between “corporate power” and “government power” becomes much cleaner on an organizational chart than it is in practice.

There is an old word for rule by a wealthy few:

Oligarchy.

The claim here is not that the United States is a pure oligarchic regime, or that its democratic institutions are unreal. It is that a corporate-political oligarchy exists within the broader democratic system: concentrated private economic actors possess political capacities radically greater than those available to ordinary individuals.

That does not mean a secret council controls America.

It does not mean corporations always win.

It does not mean every politician is corrupt, every scientist compromised, or every institution captured.

It does not even mean the powerful agree with one another.

The oligarchy is not an organization.

It is a distribution of power.

The network is how that power is exercised.

What Is Power?

Before asking who has political power, it helps to define power without reference to offices or titles.

At its broadest, social power is the capacity to cause coordinated human action—to alter the probability that other people will act in a desired way.

A president possesses power because a few words can cause thousands of people to begin acting across a continent.

A CEO possesses power because a decision can redirect the work of tens of thousands of employees.

A billionaire possesses power because money can activate lawyers, engineers, researchers, consultants, lobbyists, advertisers, political organizations, and entire companies.

An ideological or religious leader may possess enormous power without commanding either wealth or police because belief can cause millions of people to act voluntarily.

There are three broad mechanisms by which people cause other people to act.

Coercion

Do this, or something undesirable happens.

Fines. Imprisonment. Regulation. Taxation. Seizure. Exclusion. Military force. Physical violence.

The state specializes in coercive power because it claims lawful authority to impose rules even on people who object to them.

Exchange

Do this, and you receive something you value.

Wages. Contracts. Investment. Campaign support. Grants. Consulting fees. Jobs. Promotions. Access. Favors. Future employment.

Money is extraordinarily useful because it is a general-purpose claim on human labor. It allows stored economic value to be converted into organized action almost anywhere in society.

Persuasion

Do this because you believe you should.

Because it is true.

Because it is moral.

Because it is scientific.

Because it is normal.

Because your profession expects it.

Because your political faction supports it.

Because people you trust recommend it.

Persuasion is a mechanism of power regardless of whether what is being communicated is true.

A scientist who convinces the public that smoking causes cancer is exercising persuasive power.

A tobacco company that deliberately creates doubt about that evidence is also exercising persuasive power.

Those acts are not epistemically or morally equivalent. One may be truthful and socially beneficial while the other is deceptive.

But both demonstrate the same important fact:

belief changes behavior.

Education, journalism, science, religion, advertising, ideology, prestige, culture, and professional expertise therefore matter politically because the persuaded person supplies much of his own energy.

He does not need to be paid each time.

He does not need to be threatened each time.

Once people believe a behavior is right, necessary, respectable, or inevitable, they frequently enforce the rule upon themselves and encourage others to do the same.

Legitimacy is therefore among the cheapest and most durable forms of power.

Power Has Multipliers—and Memory

Information tells you where to apply coercion, money, or persuasion.

Prestige makes persuasion cheaper.

Networks make coordination cheaper.

Organizations allow thousands of people to pursue objectives repeatedly rather than assembling from scratch every time.

Institutions add something else:

they can store previous exercises of power.

Suppose an industry spends twenty years obtaining a favorable statute.

Eventually it wins.

The lobbying campaign ends.

The law remains.

The law generates administrative regulations.

The regulations generate procedures.

The procedures generate court precedents.

Companies make investments around them.

Government employees are trained within them.

Universities teach the resulting body of law.

Twenty years later, the people who originally fought for the rule may be dead.

The rule still acts.

Yesterday’s exercise of power becomes today’s normal procedure.

That is embedded power.

And it leads to another distinction that matters enormously:

exercised power is not the same thing as available power.

A corporation may devote little effort to influencing an agency when only $2 million is at stake.

That tells us almost nothing about what happens when $20 billion is at stake.

As the expected value of a political outcome rises, spending more resources to obtain it becomes rational: more lawyers and lobbying, more public relations and research, more trade-association coordination and litigation, more senior executives spending political capital.

Power has a price curve.

Nobody rationally spends a billion dollars obtaining a million-dollar advantage.

The absence of overwhelming pressure in ordinary circumstances therefore tells us little about an actor’s maximum political capacity.

The meaningful question is not simply:

“Does company A control agency B?”

It is:

How much can A change the probability of outcome X, how much would doing so cost, and how valuable would X have to become before A mobilized that capacity?

A company that can move an outcome from a 20 percent probability to 60 percent possesses tremendous power even though it still loses four times out of ten.

Power is not certainty.

Power is leverage.

The Corporate-Political Oligarchy Is a Network, Not a Club

“Corporate” in this essay does not mean the neighborhood bakery because its owner filed articles of incorporation.

The legal form is irrelevant.

A plumber with an LLC and Walmart are both technically business organizations. They do not possess remotely comparable political capacity.

What matters is concentrated economic power: corporations, major owners, financial institutions, trade associations, contractors, and other organizations with enough resources to exert consequential influence.

“Oligarchy” does not require a membership card.

The relevant network changes with the subject.

Defense policy involves defense contractors, the Pentagon, congressional committees, intelligence institutions, consultants, lobbyists, and policy organizations.

Pharmaceutical policy involves manufacturers, FDA officials, hospitals, insurers, physicians, medical societies, researchers, journals, legislators, and lobbyists.

Finance has another configuration.

Energy another.

Technology another.

Some actors repeatedly occupy central positions because they possess highly convertible resources: vast capital, political access, government contracts, control of infrastructure, expertise, prestige, communications reach, or large organizations.

Others are powerful only around particular issues.

There is no fixed roster because oligarchic power is not an identity.

It is a position within a network.

And networks do not require everyone inside them to agree.

The Army and Navy can compete over budgets while both remain part of the Pentagon.

The IRS and FDA rarely need to coordinate for both to remain parts of the federal government.

Congress fights presidents.

Agencies fight each other.

Judges reverse officials.

Political parties compete for control of the state.

Nobody concludes from this that “the government” is imaginary because no single person commands every component.

Why demand a much stricter standard of unity before concentrated private political power can be analyzed as a system?

There can be conspiracies inside such a network.

There need not be one conspiracy of the whole network.

They Were Describing It Long Before Us

Andrew Jackson warned about an organized “moneyed power” in 1837.

His immediate fight concerned the Second Bank of the United States, and his economics should not simply be transplanted into the present. But the political mechanism he feared remains recognizable.

He warned that concentrated financial interests could act “with unity and concert,” support or defeat government measures, and cause government to pass “from the hands of the many to the hands of the few.” (American Presidency Project)

Fifty-one years later, Grover Cleveland told Congress that corporations intended as servants of the public were becoming its masters.

Then came Woodrow Wilson.

Whatever one thinks of Wilson’s presidency, The New Freedom confronts the same conceptual trap that still derails discussions of elite power today.

Wilson described concentrated credit, interlocking corporate boards, legislation influenced by powerful men outside Congress, and economic development increasingly controlled by small groups. Then he said:

“I do not suspect that any man has deliberately planned the system.”

He immediately added that “control of credit also has become dangerously centralized.” Elsewhere in the same book he described the United States as among the most “completely controlled and dominated” governments in the civilized world and spoke of government under the “duress of small groups of dominant men.” (Project Gutenberg, The New Freedom)

Wilson explicitly rejected the necessity of one “malevolent combination” directing the entire arrangement. His argument was that concentration and institutional interdependence could produce domination without a master planner.

That distinction is fundamental.

Company A and Company B can compete fiercely in the marketplace while coordinating through a trade association on regulation.

A billionaire can support one candidate while another billionaire supports his opponent.

A bank can oppose a technology company on one statute while agreeing with it on another.

Aligned incentives can coordinate action without a meeting. Meetings can coordinate action where alignment alone is insufficient. Explicit conspiracies can occur inside particular episodes.

No supreme chairman is required.

In 1912 Theodore Roosevelt’s Progressive movement described the relationship between business and government as an “invisible government.” Its platform also warned that concentrated corporate wealth had placed “enormous, secret, irresponsible power” in a few hands and produced “sinister influences” on public agencies. (American Presidency Project)

Franklin Roosevelt later called concentrated financial and corporate interests “economic royalists” and said it was “natural and perhaps human” that the beneficiaries of these economic dynasties would reach for control over government. (American Presidency Project)

Eisenhower saw another part of the structure emerging after World War II: industrial, military, political, and scientific institutions becoming mutually dependent enough to form a new center of power.

Then, in 1971, we get an unusually valuable view from the other direction.

Corporate attorney Lewis Powell sent a confidential memorandum to Eugene B. Sydnor Jr., chairman of the Education Committee of the U.S. Chamber of Commerce. Less than two months later Richard Nixon nominated Powell to the Supreme Court.

Powell believed American business was losing an ideological and political struggle. His proposed response was not merely better advertising. He discussed sustained organization involving politics, courts, universities, scholarship, textbooks, television, and public communication. The original typescript and printed memorandum survive in the Lewis F. Powell Jr. Papers at Washington and Lee University. (Washington and Lee University)

The Powell memorandum is sometimes mythologized into a master blueprint for everything that followed. That claim is unnecessary and historically difficult to prove. Read conservatively, the document is more valuable: a future Supreme Court justice, privately advising the U.S. Chamber of Commerce, understood that economic resources could be converted deliberately into political, legal, academic, and persuasive power across multiple institutions.

That is not an outside critic speculating about organized business. It is an insider discussing how he believed organized business should wield power.

This Is Not a Partisan Theory

The easiest way to misunderstand this argument is to assume “the oligarchy” means whichever political party one already dislikes. It does not. Democrats and Republicans are not identical; their differences can have enormous consequences, and wealthy people and industries often have genuine ideological preferences.

But institutions expecting to survive dozens of election cycles have incentives different from ordinary partisan voters.

Pfizer’s own report for the 2023–2024 election cycle describes its political-action committee as bipartisan: 56 percent of PAC disbursements went to Republicans and 44 percent to Democrats. Pfizer says its contribution decisions consider candidates’ views on issues affecting the company and its employees. (Pfizer political-contributions report)

Lockheed Martin similarly says its political activity is evaluated according to the interests of the corporation and its stockholders, while its advocacy spans officials of both major parties. (Lockheed Martin political disclosures)

That is rational. If one party controls a committee today and another controls it four years later, a company planning to exist for fifty years has little reason to make permanent enemies of either.

The public often experiences politics differently. Ordinary people possess relatively little political power individually. Their great resource is numbers: voting blocs, movements, unions, associations, consumer action, mass opinion. Millions acting together can overwhelm concentrated money—but they must coordinate first.

A billionaire does not need to persuade fifty million people before deploying a billion dollars. A large corporation already arrives organized: employees assigned, lawyers retained, lobbyists paid, objectives pursued while most citizens are at work.

That gives faction an asymmetric effect.

Faction divides the resource ordinary people possess in greatest abundance: their numbers.

It does not equivalently divide an institution that already possesses concentrated money and permanent organization. This does not mean corporations invented polarization, that partisan disagreements are fake, or that every division benefits the same interests. It means something narrower:

A public divided horizontally into hostile political factions has greater difficulty exercising countervailing power against durable institutions capable of maintaining relationships across those factions.

A defense contractor can work with both parties because either may vote on defense appropriations. A pharmaceutical company will still face regulation after the next election. Banks hire officials from Republican and Democratic administrations; lobbying firms employ people from both.

The faction changes.

The institution remains.

The conflict may be entirely real. Its structural consequence can still benefit actors capable of operating across it.

The Soft Touch

When people imagine political domination, they often imagine Stalin.

The dictator gives an order.

The secret police arrive.

Everyone knows where power resides.

Contemporary American oligarchic influence is obviously not morally equivalent to Stalinism. It is vastly less coercive, less violent, less comprehensive, and more contestable.

The comparison is useful in only one respect:

obvious authoritarian power announces itself as power.

Nobody living under Stalin needed an essay proving Stalin had political authority.

The command itself revealed the hierarchy.

Soft oligarchic power is harder to see precisely because its successful exercise can look like a series of independent, ordinary decisions.

A regulator follows procedure.

A scientist pursues available funding.

A legislator receives expert testimony.

A candidate raises campaign money.

A journalist interviews recognized authorities.

A government employee considers future career opportunities.

A professional follows prevailing standards.

A company obeys a regulation established decades ago.

Nobody feels ordered.

Nobody necessarily thinks he is corrupt.

Sometimes nobody has done anything individually improper. The regulator may have followed procedure, the scientist followed evidence, the politician considered constituents, and the former official simply accepted a job.

Every sentence can be true.

The pattern can still matter.

The policeman’s baton announces itself as power. A career incentive does not.

Neither does a research grant, a campaign contribution, a consulting contract, or an attractive job waiting after public service. Neither does a technically dense policy recommendation supplied to an understaffed agency.

Soft power can be less destructive than brutal coercion while simultaneously being more difficult to perceive.

That obscurity is itself politically important.

People readily resist power they can identify.

A system of distributed incentives can be difficult even to name.

Sometimes Selection Makes Orders Unnecessary

Influence does not always require changing someone’s behavior after he acquires power.

Sometimes the more effective mechanism is determining which kinds of people reach positions of influence in the first place.

Call the first transactional influence.

An actor already holds office, and some outside force changes his incentives.

Call the second selection influence.

The institutional environment can make some kinds of people more likely to acquire, retain, or advance within positions of authority.

Imagine two regulators.

One regards close industry cooperation as sensible, values private-sector experience, and would comfortably take an industry position after public service.

The other distrusts the industry, favors aggressive enforcement, and considers later employment by regulated companies ethically questionable.

Neither person needs to be corrupt.

But they will not encounter identical networks of support, recommendations, job opportunities, and relationships.

The same principle can operate elsewhere.

Research questions attractive to major funders have more potential funding.

Politicians compatible with large donor networks find fundraising easier.

Policy experts compatible with a think tank’s mission are more likely to work there.

Journalists whose work fits an organization’s audience, standards, and business model are more likely to advance within that organization.

This does not predetermine outcomes.

It creates selection pressure.

Over thousands of careers and decades of institutional development, relatively weak selection pressures can produce substantial alignment.

Eventually fewer direct interventions are necessary.

The system increasingly selects people inclined to behave in compatible ways without being ordered.

That is one reason searching only for bribery misses so much of how institutional power works.

Politics Has an Economy

According to the Federal Election Commission, in the 2023–2024 federal election cycle presidential candidates raised about $2 billion, congressional candidates about $3.8 billion, political parties roughly $2.7 billion, and political-action committees roughly $15.7 billion.

Reported independent expenditures totaled approximately $4.4 billion. (Federal Election Commission)

Those categories are not additive—money can move among candidates, parties, and PACs—and they are not synonymous with corporate spending. PACs represent many kinds of organizations and interests, and federal election law imposes important restrictions on direct corporate contributions to candidates.

But the basic observation is unavoidable:

selection and persuasion of political officials occur within an enormous money economy.

Why do sophisticated businesses spend heavily on lobbying?

Presumably because they expect that it sometimes works.

One Harvard Business School working paper examined lobbying around the American Jobs Creation Act of 2004 using a two-stage model. For the median firm in its sample, the authors estimated that an additional $1 million in lobbying was associated with about $32.35 million in tax savings. They noted that even this estimate was almost an order of magnitude below earlier descriptive estimates. (Chen, Gunny & Ramanna, HBS Working Paper 15-050)

That is a context-specific estimate in an unusual tax-policy setting, not a universal exchange rate between lobbying and legislation.

It is enough to illustrate the incentive.

Political influence can have measurable financial returns. If firms can reasonably expect even a fraction of returns like these in some settings, political influence becomes a rational category of investment.

A company declining to pursue legal political advantages may simply surrender them to competitors that will.

This is why the system does not require unusually villainous executives.

The incentives themselves push economic power toward political power.

Concentrated Benefits, Diffuse Costs

Imagine a regulation that produces $2 billion in benefit for one industry while imposing $10 in costs upon each of 200 million people.

For the industry, spending $100 million trying to obtain the rule can be rational.

For an individual citizen, spending even one Saturday understanding and fighting it is irrational.

The benefit is concentrated.

The cost is diffuse.

The industry can employ people whose entire careers concern one subsection of one regulation.

The ordinary person has work, children, bills, dinner, and a few hours in which he is somehow expected to understand the entire government.

The corporation can fight for fifteen years.

The citizen may never hear of the issue.

Political information is expensive.

Expertise is expensive.

Organization is expensive.

Monitoring legislation is expensive.

Concentrated interests can justify those costs because the returns are concentrated too.

This creates enormous advantages even without corruption.

And once political infrastructure exists, using it again becomes cheaper.

A corporation already possessing lobbyists, lawyers, trade associations, government-relations staff, and relationships in Washington can redirect that machinery toward the next issue.

Ordinary citizens often have to organize countervailing power from scratch.

When Private Interests Help Make Public Rules

Sometimes the overlap is explicit.

Consider the American Legislative Exchange Council, ALEC.

Critics often summarize ALEC by saying that corporations write laws. That is too crude.

ALEC’s own materials are more useful.

Its private-sector membership brochure says business leaders can “participate in model policy development.” Each task-force membership includes a private-sector voting representative, and ALEC advertises the opportunity to “debate, discuss and vote on model policy.” (ALEC private-sector membership brochure)

ALEC separately explains its approval process: a state legislator submits a draft; task-force members debate and amend it; then separate majority votes are taken among legislators and private-sector members. Both groups must approve it before it advances to ALEC’s legislator-only board for final approval. (ALEC’s description of its model-policy process)

This does not mean corporations secretly occupy legislative seats.

Something more ordinary happens:

Private-sector representatives and elected lawmakers openly participate together in an organization that develops model laws intended for use by public legislatures.

One can support ALEC’s agenda or oppose it.

The structural fact is the same.

The border between private advocacy and public policymaking is porous.

The Revolving Door Is a Labor Market

The defense sector provides another example.

The Government Accountability Office examined fourteen major defense contractors and their employment of former Department of Defense personnel.

In 2019, those contractors employed 1,718 people who had left DOD between 2014 and 2019 after serving as senior military, senior civilian, or acquisition officials.

They also employed tens of thousands of other former DOD personnel. (GAO)

That does not establish that 1,718 officials sold favors.

It establishes something much more defensible:

There is a substantial labor market connecting the public institution buying military goods with the private firms selling them.

The relationship exists for legitimate reasons.

Former officials possess valuable expertise.

Government benefits when officials understand industry.

Industry benefits from people who understand government.

Those practical advantages create the structural conflict.

The government official knows which companies hire people with his experience.

The contractor knows which officials understand procurement.

Former officials retain relationships with current officials.

Nobody has to sign a corrupt agreement for those expectations to affect behavior at the margin.

More than sixty years after Eisenhower’s farewell address, the personnel connection he warned about is measurable.

Sometimes Government Delegates Part of Government

Aircraft certification requires extraordinary technical expertise.

For practical reasons, federal law allows the Federal Aviation Administration to delegate certain certification, examination, testing, and inspection functions to qualified private persons and organizations. The FAA’s Organization Designation Authorization program can give companies—including aircraft manufacturers—authority to perform specified functions on the agency’s behalf. The FAA emphasizes that this is not formal self-certification: it retains oversight and authority. (FAA: Delegated Organizations; FAA: How Does the FAA Certify Aircraft?)

After the two fatal Boeing 737 MAX crashes, the Department of Transportation Inspector General investigated FAA certification and delegation.

The Inspector General found that weaknesses in FAA guidance and processes limited the agency’s understanding of Boeing’s MCAS safety assessments before the first crash. It also concluded that the Boeing ODA structure did not adequately ensure the independence of personnel exercising delegated authority, and recommended stronger controls where company engineers might both work on a design and make compliance findings for the FAA on that same design. (Department of Transportation Inspector General)

The interesting fact is not that Boeing secretly seized the FAA.

It did not.

Government had legitimate operational reasons to rely on industry expertise. But the resulting structure allowed employees of the regulated company to perform some regulatory functions on behalf of the regulator.

Private and public authority partially overlapped by design.

That is more useful for understanding institutional power than a simple story about bribery.

Modern capture is often a problem of dependency.

“Regulatory Capture” Is Ordinary Governance Vocabulary

The World Bank has used concepts such as regulatory capture and state capture for decades.

A 2000 World Bank report described regulatory agencies as “captured” when they regulate businesses according to the private interests of the regulated rather than the public interest for which the agencies were established. Crucially, it added that regulatory capture “is not limited to influence through corrupt means.” The same report distinguished this broader regulatory concept from its narrower definition of state capture, which focused on illicit, illegitimate, and nontransparent influence over the formation of laws, rules, and decrees. (World Bank, Anticorruption in Transition, Box 1.1)

That distinction should remain.

Businesses should be able to petition government; regulators genuinely need outside expertise; former officials can have valuable private-sector skills; private funding can produce excellent science; and public-private cooperation can solve real problems.

The existence of legitimate influence does not eliminate capture.

It makes the interesting question one of degree.

How much access?

How dependent is the regulator?

How asymmetric are the resources?

How easily can the institution resist?

How expensive is resistance?

The broader corporate-political oligarchy is not identical to regulatory capture.

It is the distribution of resources and relationships that makes repeated capture attempts possible.

Tobacco Is the Example Everyone Recognizes

A 1969 Brown & Williamson strategy document contains perhaps the most famous sentence in the history of corporate influence over science:

“Doubt is our product.”

The document discussed the strategic value of creating public controversy around the growing body of evidence concerning cigarettes. (UCSF Industry Documents Library)

Most people already know the general tobacco story.

Companies funded research, cultivated experts, emphasized uncertainty, attacked adverse evidence, and attempted to shape public understanding because public understanding influenced regulation and profits.

That recognition gives us a useful question:

Was tobacco exceptional?

Did one uniquely corrupt industry accidentally discover that scientific belief can be politically and economically valuable?

Or is influence over knowledge a predictable target wherever enough money is at stake?

Sugar: Influence Can Happen Before Fraud

In the 1960s the Sugar Research Foundation funded Harvard researchers to review evidence about diet and coronary heart disease.

Archival correspondence later showed that the foundation funded the project, communicated its interests, supplied materials, and received drafts. In a November 1966 letter acknowledging receipt of a draft manuscript, SRF vice president and research director John Hickson told one of the researchers that it was “quite what we had in mind.” (original SRF letter; historical analysis and archived correspondence)

The resulting review did not disclose SRF’s financial role.

That nondisclosure should not be judged by today’s journal rules. The New England Journal of Medicine did not announce a stated financial-conflict disclosure policy until 1984. The point is not that the authors violated a later rule retroactively; it is that readers of the 1967 review were not given information that would now be considered relevant to evaluating the sponsor’s relationship to the work. (NEJM, 1984 conflict-of-interest policy)

The surviving documents do not establish that the foundation directly edited the manuscript, and the scientific debate over sugar, fat, and cardiovascular disease was genuinely complicated.

That limitation makes the case more instructive.

Influence over science does not require inventing data.

It can occur earlier.

Which questions receive money?

Which scientists receive support?

Which uncertainties receive attention?

Which lines of inquiry become prestigious?

Which research programs become large enough to generate their own institutional momentum?

Funding affects the landscape in which sincere scientists conduct sincere science.

No order to falsify anything is required.

Purdue Put Marketing Inside the Doctor’s Computer

The Practice Fusion case makes the mechanism almost painfully concrete.

Practice Fusion made electronic health-record software used by physicians.

Purdue Pharma admitted in federal court that it paid Practice Fusion nearly $1 million for a software prompt intended to cause doctors to recommend or order Purdue’s extended-release opioid products. Purdue’s chairman admitted that one purpose of the prompt was to increase opioid sales, and Purdue pleaded guilty to conspiring with Practice Fusion to violate the federal Anti-Kickback Statute. (U.S. Department of Justice: Purdue guilty plea)

The recommendation appeared inside physicians’ clinical software—a place doctors reasonably expected to receive unbiased medical information.

Look at the chain:

money → software intermediary → medical information environment → physician behavior → prescriptions → revenue

Purdue did not need to telephone every doctor and issue an order.

The goal was to alter the information architecture in which doctors independently made decisions.

And Purdue was not the only pharmaceutical relationship at issue. In a separate civil settlement, Practice Fusion resolved federal allegations involving thirteen other sponsored clinical-decision-support arrangements intended to increase pharmaceutical sales. DOJ explicitly noted that those civil claims were allegations and that liability had not been determined. (U.S. Department of Justice: Practice Fusion resolution)

That is soft power in unusually visible form.

NIH Shows Why “Control” Is the Wrong Word

In 2018 the National Institutes of Health terminated the Moderate Alcohol and Cardiovascular Health trial.

The National Institute on Alcohol Abuse and Alcoholism expected to contribute $20 million over ten years. The Foundation for the NIH had raised $67.7 million in private donations from five major alcohol producers: Anheuser-Busch InBev, Carlsberg, Diageo, Heineken, and Pernod Ricard.

NIH found significant process irregularities in development of the funding opportunity, policy violations related to solicitation of outside funding, and study-design concerns serious enough to cast doubt on the trial’s credibility. It terminated the project and announced broader measures to detect inappropriate outside influence. (National Institutes of Health)

This example is important because the institution resisted.

A theory claiming that alcohol companies controlled NIH would have difficulty explaining the outcome.

A theory of disproportionate but contested influence explains it easily.

Private interests possessed enough access for a serious integrity problem to develop. Other people inside the institution possessed enough independence to detect the problem and stop it.

Power is not ownership.

It is leverage exercised inside systems containing countervailing forces.

The Pattern Is Large Enough to Measure

Perhaps lead, tobacco, sugar, Purdue, and alcohol are simply scandals selected because they fit the argument.

That is a reasonable objection.

So move from anecdotes to aggregation.

A Cochrane methodological review examined 75 papers comparing industry-sponsored drug and medical-device research with research funded from other sources.

Across the included comparisons, industry-sponsored studies were more likely to report efficacy results favorable to the sponsor’s product (risk ratio 1.27) and more likely to reach favorable overall conclusions (risk ratio 1.34). Cochrane found that the association was not explained by the standard methodological risk-of-bias measures it examined. (Cochrane)

The evidence quality was not uniform—Cochrane rated the efficacy finding moderate quality and the favorable-conclusion finding low quality—and none of this means industry-funded science is false.

Industry funds enormous quantities of useful research. A well-conducted experiment does not become wrong because a corporation wrote the check.

The result is more interesting:

The source of financial interest is statistically associated with the direction of findings and conclusions across a large body of research.

The mechanism is not confined to famous scandals.

It leaves a population-level footprint.

Finance Gives Us a Large-Scale Stress Test

The Financial Crisis Inquiry Commission was established by Congress to investigate the 2007–2009 financial crisis.

Its commissioners did not agree completely, and the final report included extensive dissenting views. That disagreement should not be hidden.

But the majority’s description of the political economy is unusually direct.

It concluded that the financial industry “played a key role in weakening regulatory constraints on institutions, markets, and products” and said an industry of such wealth and power could be expected to exert pressure on policymakers and regulators. From 1999 through 2008, the report said, the financial sector spent approximately $2.7 billion on reported federal lobbying while individuals and PACs associated with the sector made more than $1 billion in campaign contributions. (Financial Crisis Inquiry Commission report, majority conclusions)

One need not accept the majority’s entire theory of the crash.

The narrower fact is enough:

Actors in and around one of the wealthiest regulated sectors in America spent or contributed billions of dollars through lobbying and campaign channels while government was making decisions with enormous consequences for that sector.

Those expenditures were rational because regulatory outcomes were enormously valuable.

Economic power pursued political power.

Powerful People Do, In Fact, Meet

Elite conferences attract mythology because private meetings invite imagination. The organizations’ own descriptions are more useful.

The Trilateral Commission says it has spent decades bringing together senior policymakers, business leaders, and representatives of media and academe. It describes itself as a venue to “incubate ideas and form relationships across sectors and geographies.” (Trilateral Commission)

Bilderberg says its annual private meeting brings together roughly 130 political leaders and experts from industry, finance, labor, academia, and media under the Chatham House Rule.

It also explicitly says no resolutions are proposed, no votes are taken, and no policy statements are issued.

Include that fact.

Bilderberg’s own description is not evidence of a secret parliament.

It is evidence of a private forum in which unusually powerful people exchange information and have the opportunity to form relationships. That creates channels through which later coordination, trust, or reciprocal help can occur; the meeting itself does not prove that any particular downstream policy was caused there. (Bilderberg Meetings)

Ditchley describes itself as a confidential forum that brings together government, business, and technology leaders, uses invitation-only convenings and smaller backchannel talks, and says its route to impact is engaging people and organizations with “direct influence over outcomes.” (Ditchley Foundation)

The World Economic Forum operates more publicly and on a larger scale. Its 2026 Davos meeting brought together nearly 3,000 participants, including a record 400 top political leaders, nearly 65 heads of state and government, and about 830 CEOs and chairs; WEF described the meeting as a platform for consequential dialogue across government, business, civil society, and academia. (World Economic Forum)

None of this establishes a hidden command structure. It establishes something much less sensational:

Cross-sector networking among unusually powerful people is real, intentional, and openly organized.

These forums create opportunities for information exchange, relationship building, and coordination. What they do not establish by themselves is control over government or the cause of any particular policy outcome.

Competition Does Not Cancel Oligarchy

Banks compete with banks. Drug manufacturers fight over patents. Billionaires fund opposing candidates. That is expected: an oligarchy need not be a cartel.

The question is not whether powerful actors have identical preferences, but whether their capacity to pursue those preferences is distributed remotely equally throughout society. It is not.

Internal competition can constrain oligarchic power. It does not make ordinary citizens institutionally equal to the contestants.

Democracy Can Contain Oligarchy

The United States holds competitive elections. Governments change peacefully. Citizens possess broad speech and associational rights. Courts constrain officials, and public movements sometimes defeat much richer opponents. Those are real forms of popular power.

But democracy and oligarchy need not be mutually exclusive features of a political system. Aristotle’s Politics distinguished oligarchy from democracy chiefly by wealth and poverty, not merely by the number of rulers, and he explicitly discussed constitutions that mixed democratic and oligarchic elements. (Aristotle, Politics III; Politics IV)

I use oligarchic in that dimensional sense. The claim is not that the American constitutional order is a pure classical oligarchy. It is that effective governing power can contain a strong oligarchic concentration even where elections and broad political rights remain real.

Every citizen may receive one ballot. Not every citizen can maintain a permanent Washington office, hire lobbyists, finance litigation for a decade, commission technical research, employ former regulators, fund political campaigns, purchase national advertising, or offer lucrative employment to senior officials leaving government.

So the useful question is not whether America is “really a democracy” or “secretly an oligarchy.” It is:

How democratically is effective political power distributed?

On that measure, one-person, one-vote tells only part of the story.

The Three Forms of Power Become Convertible

Return to coercion, exchange, and persuasion.

The most powerful systems are not those possessing only one.

They are those capable of converting one form into another.

Money finances research.

Research creates expertise.

Expertise influences journalism.

Journalism influences public belief.

Public belief changes electoral incentives.

Political incentives shape laws.

Agencies implement laws.

Courts enforce them.

Or:

Political influence produces a favorable tax provision.

The provision increases profits.

Profits finance lobbying.

Lobbying protects the provision.

The provision generates further profits.

Or:

Government creates a procurement program.

Contracts build an industry.

The industry hires former government officials.

Those relationships and skills improve the industry’s access and indispensability.

Government becomes increasingly dependent upon the industry.

More contracts follow.

The loops reinforce themselves:

wealth → political influence → policy → economic advantage → greater wealth

wealth → persuasion → legitimacy → political authority

political authority → contracts and privileges → private institutional power → political influence

This explains why complete control of any single institution is unnecessary.

Suppose an interested actor possesses only modest leverage over research funding, public relations, political donations, regulatory access, professional organizations, and litigation.

Each influence may be weak individually.

Together they can cascade.

A funded researcher publishes a legitimate study.

A journalist reports it.

A policy organization cites it.

A lobbyist presents it to legislators.

An agency incorporates it into a regulatory record.

A court later reviews an agency decision partly built upon that record.

Most participants may never communicate with the original funder.

The effect still propagates.

This resembles control theory more than monarchy.

Find high-leverage points.

Apply inputs.

Let the system transmit the effect.

It Does Not Require Evil People

A durable political-economic system should not depend on everyone inside it being unusually wicked. This one does not.

A CEO is paid to advance the company’s interests within the law; if lobbying can produce an advantage, refusing to lobby may simply leave it to competitors. A politician must win elections, creating incentives to cultivate people able to finance them. Regulators need technical knowledge, scientists need funding, universities want grants, news organizations need revenue, defense agencies need weapons, and contractors need customers.

Most participants can behave rationally within their local incentives while the aggregate result concentrates power.

That is how durable systems often reproduce themselves: not because bad people constantly force everyone else to behave badly, but because locally rational choices repeatedly reproduce the same structure.

And They Do Not Always Win

The powerful lose. Public outrage can overwhelm money. Competing firms can neutralize one another. Judges intervene, regulators prosecute, scientists and journalists expose manipulation, whistleblowers defect, voters punish politicians, and new technology destroys established industries.

NIH terminated the alcohol trial. DOJ prosecuted Purdue’s misconduct. FAA changed oversight practices following the MAX disasters. Countervailing power is real.

That does not make concentrated power imaginary. A powerful actor is not defined by winning every contest, but by possessing an unusual capacity to change the odds.

Most Recently, They Used the Word

In January 2025, outgoing President Joe Biden warned:

“An oligarchy is taking shape in America of extreme wealth, power, and influence.”

A reader may agree with Biden’s targets or regard the warning as partisan hypocrisy. The more interesting fact is historical: he joined a line of American political figures extending backward through Eisenhower, Roosevelt, Wilson, Cleveland, and Jackson who warned that concentrated private power could become political power. (American Presidency Project)

Different parties, centuries, industries, and political enemies; the underlying incentive survived them all. Economic power seeks political influence because political decisions affect economic outcomes, while government seeks resources and expertise that concentrated private institutions can supply. Once those channels exist, successful influence can become embedded in institutions and outlive the people who originally exercised it.

Look at the System

Viewed individually, every component has an ordinary name.

Campaign finance.

Lobbying.

Government relations.

Model legislation.

Public-private partnerships.