Conspiracy, Capture, or Coincidence? How Power Moves Without a Secret Meeting

Most people imagine a conspiracy as a room.

A heavy door. A dark table. A handful of powerful people leaning over a map while one of them says something that would sound ridiculous if the lighting were better.

That happens sometimes.

But it is not the only way power coordinates.

Power can move through a contract no one bothered to read. Through a regulation written in language only the regulated industry can afford to interpret. Through a revolving door that is perfectly legal. Through a funding structure that rewards one conclusion more reliably than another. Through a rule sold as protection that quietly raises the ladder after the largest companies have already climbed it.

Sometimes there is a conspiracy.

Sometimes there is capture.

Sometimes the incentives are aligned so precisely that nobody needs to make the phone call.

And sometimes two things merely happen at the same time.

The difficult work is telling the difference.

The Wrong Question

When a suspicious outcome appears, the public conversation usually collapses into one question:

Is it a conspiracy?

That question is too blunt.

It forces a complicated system into two boxes:

• Either powerful people secretly planned everything.
• Or nothing improper happened and anyone who noticed the pattern is imagining things.

Real institutions do not work that cleanly.

A policy can benefit a narrow group without every beneficiary helping design it. An agency can drift toward the worldview of the industry it regulates without its employees accepting bribes. A contractor can gain influence through expertise, access, employment networks, and procurement dependence without a cinematic villain issuing orders from underground.

The better question is:

What mechanism produced the outcome?

That question is less dramatic.

It is also far more dangerous to weak explanations.

Four Ways the Same Outcome Can Appear

Imagine a new rule is announced as a public safeguard.

The rule imposes heavy compliance costs. The largest firms can absorb them. Smaller competitors cannot. Within a few years, the market is more concentrated, the established firms are stronger, and the public has fewer choices.

From the outside, the result looks the same regardless of how it happened.

But there are at least four different explanations.

1. Criminal conspiracy

A conspiracy involves actual coordination. Under the general federal conspiracy statute, two or more people must conspire to commit an offense against, or defraud, the United States, and at least one must take an act to advance the object of the agreement.[1]

Other conspiracy laws have different elements, but the essential idea is agreement: people knowingly coordinate toward a concealed objective.

In our example, conspiracy might mean firms secretly agreed with officials to design the rule as a weapon against competitors, while disguising that purpose from the public.

That is a serious allegation. It requires serious evidence.

2. Regulatory or policy capture

Capture is different. It does not always require a prosecutable crime, a bribe, or a secret oath.

George Stigler’s influential theory of economic regulation argued that industries often seek and acquire regulation that benefits them. Later scholars developed more careful definitions. Daniel Carpenter and David Moss describe regulatory capture as a process in which regulation is repeatedly directed away from the public interest and toward the regulated industry through the industry’s own intent and action.[2][3]

The OECD uses the broader term policy capture for public decisions that are consistently or repeatedly redirected from the public interest toward a specific interest.[4]

In our example, capture might mean the largest firms used lobbying, access, technical expertise, trade associations, and sustained pressure to shape a rule that still sounds public-serving but works disproportionately in their favor.

Everything important may have happened in meetings with calendars, nameplates, legal counsel, and catered sandwiches.

The absence of a smoke-filled room does not make the outcome neutral.

3. Aligned incentives

Sometimes nobody coordinates the full outcome.

Each participant simply responds to the incentives in front of them.

The company seeks lower costs and fewer competitors. The regulator wants information, cooperation, and a future career path. The politician wants donations, endorsements, favorable headlines, or a policy achievement before the next election. The consultant wants another contract. The media outlet wants access and attention. The public wants the problem solved without reading 900 pages of administrative language.

Nobody needs to say, “Let us build a system that serves us.”

The system already pays each person to move in roughly the same direction.

This is not necessarily conspiracy or capture. It may be institutional convergence: separate actors producing a coordinated-looking result because they share incentives, assumptions, dependencies, or constraints.

A school of fish does not hold a briefing before it turns.

4. Coincidence, error, or unintended consequence

Sometimes a rule benefits an industry because the public interest and the industry’s interest genuinely overlap.

Sometimes officials make a mistake.

Sometimes the harm appears only later. Sometimes complexity produces consequences nobody predicted. Sometimes a pattern exists in our attention more than it exists in the world.

Research on conspiracy belief has repeatedly found links between uncertainty, perceived lack of control, and the tendency to detect meaningful patterns—including patterns that may not be real.[5]

That does not mean pattern recognition is defective.

It means pattern recognition is a tool.

A rifle can protect a family or put a hole in the wrong wall. The existence of bad shots is not an argument for throwing away marksmanship. It is an argument for discipline.

Legal Is Not the Same as Legitimate

One reason these conversations become confused is that people treat legal as a synonym for good.

It is not.

Law tells us what a system permits, prohibits, rewards, requires, or refuses to punish. It does not automatically tell us whether the arrangement is wise, fair, transparent, or worthy of trust.

Many forms of influence are legal:

• Lobbying
• Campaign support within legal limits
• Trade-association advocacy
• Industry-funded research
• Public-private partnerships
• Government contracting
• Hiring former officials
• Supplying technical expertise to regulators
• Writing model legislation
• Litigating until the rule changes

Each can serve a legitimate purpose.

Each can also become part of a machinery that concentrates power.

The mature position is not that every legal relationship is corrupt.

It is that legality answers only one question.

Sometimes the arrangement is hidden from the law.

Sometimes the arrangement is the law.

Why Small Groups Keep Winning

The public often asks how a policy that imposes small costs on millions of people can survive while producing large benefits for a small group.

The answer may be less mysterious than it looks.

Mancur Olson’s work on collective action described a structural advantage held by smaller, concentrated interests. A group expecting a large benefit has a strong reason to organize, spend money, study the issue, hire specialists, attend hearings, and keep pressure on decision-makers. A large public paying small individual costs has much less reason to do any of those things.[6]

Suppose a rule transfers $20 from each of ten million people to an industry that gains $200 million.

The public may dislike it in the abstract. But no individual citizen can justify taking a month off work to fight over twenty dollars.

The industry can justify hiring a building full of lawyers.

That does not prove a conspiracy.

It explains why the room can remain technically open to everyone while only one side can afford to live there.

Stigler’s theory pushed this insight further: the state can selectively help or burden industries through its power to compel, prohibit, tax, subsidize, license, and restrict entry.[2]

A rule does not need to hand a company a bag of money to create value.

It can raise the cost of entering the market.

It can define which evidence counts.

It can decide which credentials are mandatory.

It can turn compliance itself into a moat.

It can make the public believe the moat was built for their protection.

Sometimes it was.

Sometimes protection was real, but the moat became wider than necessary.

Sometimes the people already inside helped choose the dimensions.

The Network Does Not Need Everyone in the Room

One objection to large-system conspiracy claims is reasonable:

Too many people would have to know. Someone would talk.

Correct—if everyone knew the entire operation.

But large organizations rarely work that way.

In the Marine Corps, the rifleman on the line does not need the commander’s entire scheme of maneuver. He needs his task, his timing, his sector, and enough information about adjacent units not to shoot the wrong people.

Compartmentalization is not proof of conspiracy.

It is proof that coordinated outcomes do not require universal knowledge.

A procurement officer can follow policy. A contractor can pursue profit. A regulator can rely on industry data. A lobbyist can argue for a client. A legislator can vote for a bill staff summarized. A researcher can apply for available funding. A communications office can repeat the approved language.

Each person may believe they are performing an ordinary role.

Together, those roles can still create a durable network of power.

The defense world offers a documented example of how dense those networks can become. Federal law places post-employment restrictions on some former officials, including limits connected to matters they handled while in government.[7] In 2021, the Government Accountability Office reported that 14 major defense contractors employed about 1,700 recent former Department of Defense senior or acquisition officials in 2019, along with more than 35,000 other recent former DOD employees.[8]

Those numbers do not prove that the people involved acted corruptly.

They prove that the revolving door is not imaginary—and that the government itself recognizes enough conflict-of-interest risk to regulate parts of it.

Eisenhower’s warning about the military-industrial complex was not that every service member, civil servant, engineer, or factory worker had joined a secret cabal. He warned that the conjunction of a permanent military establishment and a vast arms industry could acquire “unwarranted influence,” whether that influence was sought or unsought.[9]

That phrase matters.

Sought or unsought.

Sometimes a network is planned.

Sometimes it becomes powerful because the structure keeps feeding it.

Capture Without a Villain

The easiest capture story to understand involves bribery.

The more common and difficult version may involve dependency.

Regulators need technical information. Much of that information lives inside the industry. Agencies need experienced employees. Many qualified candidates come from the industry. Those employees build specialized knowledge. When they leave government, the industry may be where that knowledge is most valuable.

None of this requires evil people.

Scholars of capture have argued that ordinary economic and career incentives can push even well-intentioned regulators toward the perspective of the entities they oversee. The regulated industry becomes the agency’s most informed, organized, persistent, and professionally familiar audience.[10]

Over time, the question can quietly change.

Not:

What protects the public?

But:

What can the industry realistically tolerate?

That may be a responsible question in some circumstances. Regulations that ignore technical reality can destroy useful services, create shortages, or produce worse outcomes.

The problem begins when industry feasibility becomes the outer boundary of public imagination.

Capture does not always look like someone selling the public.

Sometimes it looks like everyone in the room sincerely agreeing about what is “reasonable” because everyone who would define reasonable differently stopped being invited years ago.

The Case Against Seeing Capture Everywhere

Now for the discipline.

A company benefiting from a rule does not prove the company captured the rule.

A campaign contribution does not prove a purchased vote.

A former official taking an industry job does not prove prior betrayal.

An agency agreeing with industry does not prove the agency is controlled by industry.

A pattern of benefit is a reason to investigate.

It is not the completed investigation.

Carpenter and Moss warn that capture is frequently overdiagnosed. Scholars may observe that policy outcomes correlate with industry preferences while failing to show whether the result came from capture, legitimate representation, technical realities, public-interest overlap, bureaucratic error, or coincidence.[3]

Their proposed evidentiary standard is demanding for a reason. A serious capture claim should identify:

1. A defensible account of the public interest.
2. A policy shift away from that interest and toward a special interest.
3. Evidence that the special interest acted intentionally and effectively enough to help cause the shift.[3]

This does not mean that capture is unreal unless somebody leaves a confession beside the copier.

It means the word should describe a mechanism, not merely our dislike of an outcome.

There is also a real danger in treating every regulation as captured. If the diagnosis is wrong, dismantling the rule may hand even more power to the strongest private actors. Capture can produce excessive regulation, but it can also weaken, delay, or hollow out regulation that threatens established interests.[3][4]

The answer to captured government is not automatically no government.

A vacuum is not public sovereignty.

Sometimes it is simply an unregulated acquisition opportunity.

What “No Evidence” Actually Means

A lack of evidence does not make a claim false.

History is full of ideas proposed before the evidence was sufficient to establish them. Records can be classified, destroyed, hidden, never created, or scattered across institutions. Powerful organizations can use legal privilege, secrecy rules, nondisclosure agreements, compartmentalization, and simple complexity to make investigation difficult.

But there is an equal mistake on the other side:

The evidence was hidden, therefore my theory is true.

That turns concealment into an all-purpose substitute for proof.

Research on reasoning from absent information suggests that missing evidence becomes informative when the evidence would reasonably be expected to appear if the claim were true.[11]

So ask:

• What evidence should this claim produce?
• Who would possess it?
• Could it realistically be preserved?
• Could independent traces exist elsewhere?
• Has anyone searched in the places where evidence should appear?
• Does the theory predict anything before it happens?
• What observation would make the theory less likely?

If the alleged operation would require thousands of equally informed participants, decades of perfect silence, no records, no budget traces, no defectors, no contradictory behavior, and no measurable consequence, the lack of evidence matters.

If the alleged action could be completed by six people using classified channels and disposable records, the lack of public evidence means less.

Unproven is not the same as false.

But possible is not the same as probable.

The honest position may remain:

I do not know yet.

That sentence does not weaken discernment.

It keeps discernment from becoming theater.

How to Tell What You Are Looking At

When a suspicious policy, institutional pattern, or coordinated narrative appears, do not begin by choosing the villain.

Begin by identifying the mechanism.

1. State the exact claim

Not “the system is rigged.”

Rigged how?

By whom?

Toward what measurable outcome?

Over what period?

2. Separate the outcome from the explanation

Outcome: Large firms benefited after the rule changed.

Explanation: Large firms secretly designed the rule to eliminate competition.

The first may be documented while the second remains an inference.

Do not smuggle the explanation into the observation.

3. Look for agreement

Is there evidence that actors coordinated?

Emails, meeting records, shared drafts, synchronized actions, witness testimony, contracts, common funding, or explicit strategy documents can help establish coordination.

Shared benefit alone cannot.

4. Identify the mechanism

Did influence move through:

• Lobbying?
• Contracting?
• Regulatory comments?
• Employment networks?
• Funding?
• Litigation?
• Credential requirements?
• Information control?
• Enforcement discretion?
• Public relations?

A theory becomes stronger when it explains how, not merely who gained.

5. Follow concentrated benefits and distributed costs

Who gains enough to organize?

Who loses too little individually to fight?

Who can attend every hearing?

Who discovers the cost only after it has been divided among millions of people?

6. Test the public-interest claim

What was the rule supposed to accomplish?

Did it accomplish that purpose?

Were less burdensome alternatives available?

Did the policy create public benefits even while favoring an industry?

Weak capture and public benefit can coexist.

7. Seek the strongest non-conspiratorial explanation

Could the outcome result from incompetence, path dependence, ordinary politics, public pressure, technical constraints, institutional culture, or unintended consequences?

A serious theory survives contact with the best alternative explanation—not the dumbest one you can defeat in a comment section.

8. Seek the strongest evidence against your preferred theory

Do not merely ask how your theory could be true.

Ask what should be different if it were false.

9. Assign the right evidence level

Use the ladder:

Documented fact → supported interpretation → inference → open question → speculation

Do not promote an inference because it is emotionally satisfying.

Do not demote a documented fact because it is politically inconvenient.

10. Keep the allegation proportional

If the evidence establishes lobbying, say lobbying.

If it establishes capture, say capture.

If it establishes criminal agreement, say conspiracy.

If it establishes only that the outcome is strange, say the outcome is strange.

Accuracy is not cowardice.

It is how you keep your fire on the target.

The Most Important Distinction

Here is the distinction that can save us from both obedience and paranoia:

Intent matters, but structure matters too.

Some harms are planned.

Some are produced by systems that reward the same behavior until planning becomes unnecessary.

Some laws are written to help a narrow group.

Some laws begin with a legitimate public purpose and are bent over time.

Some rules protect the public and protect incumbents simultaneously.

Some apparent patterns dissolve when better evidence arrives.

The world is not divided between one perfect official story and one perfect hidden story.

It is filled with partial knowledge, unequal access, lawful influence, illegal coordination, institutional momentum, competing interests, honest service, opportunism, secrecy, self-deception, and people trying to pay their mortgage without thinking too hard about the machine around them.

That is harder to fit on a meme.

It is also closer to reality.

Thread Note

Sometimes the meeting is secret.

Sometimes it is public.

Sometimes there was no meeting because the incentives did the talking.

Do not call every benefit proof of capture.

Do not call every legal arrangement harmless.

Do not confuse coordination with universal knowledge, or uncertainty with innocence.

Name the mechanism.

Trace the benefit.

Test the alternatives.

Keep the allegation proportional to the evidence.

The first job is not to identify the villain.

It is to understand the machinery.

Research Notes & Further Reading

[1] U.S. House Office of the Law Revision Counsel. “18 U.S.C. §371: Conspiracy to Commit Offense or to Defraud United States.” Current federal statutory text. The URL parameters identify the exact statute and should remain.
https://uscode.house.gov/view.xhtml?req=granuleid:USC-prelim-title18-section371&num=0&edition=prelim

[2] George J. Stigler. “The Theory of Economic Regulation.” The Bell Journal of Economics and Management Science, 1971. Foundational argument that industries may seek regulation that benefits them.
https://bfi.uchicago.edu/wp-content/uploads/2023/02/3003160.pdf

[3] Daniel Carpenter and David A. Moss, editors. “Introduction” and “Detecting and Measuring Capture,” in Preventing Regulatory Capture: Special Interest Influence and How to Limit It, Cambridge University Press, 2013. Provides a disciplined definition, distinguishes strong from weak capture, and warns against inferring capture from benefit alone.
https://tobinproject.org/sites/default/files/assets/Introduction%20from%20Preventing%20Regulatory%20Capture.pdf
https://doi.org/10.1017/CBO9781139565875.006

[4] Organisation for Economic Co-operation and Development. Preventing Policy Capture: Integrity in Public Decision Making, 2017. Defines policy capture and outlines transparency, accountability, stakeholder, and integrity safeguards.
https://doi.org/10.1787/9789264065239-en

[5] Karen M. Douglas and Robbie M. Sutton. “What Are Conspiracy Theories? A Definitional Approach to Their Correlates, Consequences, and Communication.” Annual Review of Psychology, 2023. Reviews definitions, psychological correlates, pattern perception, uncertainty, and the fact that conspiracy claims can in principle be true.
https://doi.org/10.1146/annurev-psych-032420-031329

[6] Mancur Olson. The Logic of Collective Action: Public Goods and the Theory of Groups, 1965. Foundational analysis of why smaller organized interests can hold structural advantages over diffuse publics.
https://users.ssc.wisc.edu/~peoliver/SOC924/Articles/olsonlogic.pdf

[7] U.S. House Office of the Law Revision Counsel. “18 U.S.C. §207: Restrictions on Former Officers, Employees, and Elected Officials.” Current federal statutory text governing certain post-government activities. The URL parameters identify the exact statute and should remain.
https://uscode.house.gov/view.xhtml?req=granuleid:USC-prelim-title18-section207&num=0&edition=prelim

[8] U.S. Government Accountability Office. Post-Government Employment Restrictions: DOD Could Further Enhance Its Compliance Efforts Related to Former Employees Working for Defense Contractors, GAO-21-104311, September 9, 2021. Documents employment networks and compliance concerns without treating post-government employment itself as proof of wrongdoing.
https://www.gao.gov/products/gao-21-104311

[9] Dwight D. Eisenhower. “Farewell Address,” January 17, 1961. National Archives transcript and record of the warning concerning unwarranted influence by the military-industrial complex.
https://www.archives.gov/milestone-documents/president-dwight-d-eisenhowers-farewell-address

[10] Luigi Zingales. “Preventing Economists’ Capture,” in Preventing Regulatory Capture, Cambridge University Press, 2013. Examines information dependency, professional audience, and career incentives as mechanisms that can influence well-intentioned regulators.
https://doi.org/10.1017/CBO9781139565875.010

[11] Anne S. Hsu, Andy Horng, Thomas L. Griffiths, and Nick Chater. “When Absence of Evidence Is Evidence of Absence: Rational Inferences From Absent Data.” Cognitive Science, 2017. Explains why missing evidence becomes more informative when evidence would be expected if a claim were true.
https://doi.org/10.1111/cogs.12356