The FDA Decoded: How Regulatory Capture Works

In 1995, the FDA approved OxyContin based on Purdue Pharma's claim that the drug carried a low risk of addiction. The evidence behind that claim was thin—remarkably thin for a decision that would shape the next three decades of American life. The FDA examiner who approved the drug, Curtis Wright, left the agency within two years and went to work for Purdue Pharma. Nobody called it corruption at the time. It was just the normal workings of the system: a regulator reviewing an application, making a judgment call, and then moving to the private sector for better pay. Twenty-five years and hundreds of thousands of opioid deaths later, the story looks different. But the structure that produced it hasn't changed. The revolving door still spins. The incentives still point the same direction. And the next OxyContin is, structurally speaking, already on its way through the pipeline.

What Capture Actually Looks Like

Regulatory capture is not bribery. That's the first thing to understand, because the bribery model suggests individual corruption—bad actors who can be identified and removed. The actual mechanism is structural: the regulated industry gradually gains disproportionate influence over the regulator through legitimate, legal, often well-intentioned channels.

The signs are consistent across captured agencies. Regulations burden new entrants more than incumbents—because incumbents helped design the regulations and built compliance into their existing processes. Approval processes favor companies with resources to navigate complexity—because the complexity was shaped by those companies. Enforcement is vigorous against small players but lenient with large ones—because large companies have the legal resources to make enforcement costly and the political connections to make it risky.

Standards are written by "industry experts" who rotate between writing the rules and profiting from them. Ambiguous decisions consistently break in industry's favor—not because of any single corrupt act, but because the cumulative weight of information, relationships, and career incentives all tilt the same direction. In other words, capture doesn't require anyone to be corrupt. It only requires a structure where self-interest and industry interest point the same way.

The Four Mechanisms

The Revolving Door

FDA officials know, consciously or not, that their next job may be in pharmaceutical or biotech companies. This isn't speculation—it's the documented career path. Nine of the last ten FDA commissioners went to work for pharmaceutical or biotech firms after leaving the agency. The pattern extends well beyond commissioners to reviewers, division directors, and senior staff throughout the organization.

The career incentive this creates is powerful and requires no conspiracy to operate. An FDA official doesn't need to be "paid off" to be industry-friendly. They simply need to be a rational human being navigating a career. Aggressive regulation creates enemies in an industry that may employ you next year. Lenient regulation creates friends. The salary differential reinforces the message: government pay versus industry compensation at two to five times the level. You don't need to be corrupt. Self-interest handles it.

The revolving door also works in reverse. Industry experts join the FDA, bringing genuine expertise—but also industry's worldview, assumptions, and professional network. They see the companies they regulate not as adversaries to be policed but as partners to be worked with. This perspective isn't dishonest. It's the natural result of a career spent inside the industry. But it shifts the regulatory posture from guardian of public interest to facilitator of industry goals.

User Fee Dependence

In 1992, Congress passed the Prescription Drug User Fee Act (PDUFA), which allowed the FDA to charge pharmaceutical companies fees to fund the drug review process. The stated goal was to speed up drug approvals by giving the FDA more resources. The structural effect was to make the pharmaceutical industry the FDA's primary customer.

Today, user fees fund a substantial majority of the FDA's drug review operations. The agency depends on the industry it regulates for its operating budget. Review speed became a key performance metric—because fast approvals please the fee-payers. PDUFA included performance goals for review timelines that created institutional pressure to approve quickly.

Post-market surveillance (monitoring drugs after they're approved to catch problems that clinical trials missed) doesn't generate fees and doesn't have the same performance metrics. It got neglected—not through any deliberate decision, but through the structural logic of resource allocation. Money flows toward what generates more money. Pre-approval review generates fees. Post-approval monitoring doesn't. In other words, the FDA developed a structural bias toward getting drugs onto the market and a structural weakness in catching problems after they arrive.

Information Asymmetry

The pharmaceutical industry knows more about its products than the FDA does. This is inherent and unavoidable. Clinical trial data, manufacturing processes, adverse event reports, market dynamics—all of this information is generated and controlled by industry. The FDA reviews what companies submit. It doesn't generate its own data.

This creates dependence. Regulators rely on industry for the expertise needed to evaluate industry's own products. When the FDA needs specialists in a novel drug class or manufacturing technology, those specialists come from industry—because industry is where the expertise lives. Industry provides "education" to regulators through conferences, briefings, and advisory panels staffed by industry consultants.

The information asymmetry means that the regulator is perpetually a step behind the regulated. Industry sets the terms of the conversation by controlling what information exists, how it's framed, and when it's disclosed. This isn't deception, necessarily. It's the natural advantage of the party that generates the data over the party that reviews it.

Concentrated Benefits, Diffuse Costs

Mancur Olson, the economist who formalized the logic of collective action (the structural reasons why small, focused groups outcompete large, diffuse ones in political contests), identified why regulatory capture is the default, not the exception. The pharmaceutical industry has concentrated stakes in every FDA decision—billions of dollars ride on a single approval. The public interest is diffuse—each person is affected slightly by each decision.

Industry shows up to every meeting, funds lobbyists, hires former regulators as consultants, and monitors every rulemaking. Public interest groups are understaffed, underfunded, and stretched across hundreds of issues. The loudest voice wins in the attention economy of regulation. Industry is always the loudest voice, because industry has the most to gain from each individual decision.

This asymmetry means that even well-intentioned regulatory processes tilt toward industry. Advisory panels hear primarily from industry representatives. Comment periods on proposed rules are dominated by industry submissions. The informational environment surrounding every decision is saturated with industry perspective. The public interest is, structurally, a background whisper against a foreground roar.

Case Studies: Structure in Action

OxyContin and the Opioid Crisis

The OxyContin approval in 1995 illustrates every mechanism operating simultaneously. The FDA accepted Purdue Pharma's claim of low addiction risk despite thin evidence—information asymmetry at work, with the agency dependent on the applicant's data. The approval process was shaped by user fee pressures to move quickly. The revolving door delivered Curtis Wright from FDA to Purdue. And the concentrated benefit to Purdue (billions in eventual revenue) overwhelmed the diffuse cost to the public (hundreds of thousands of lives).

No single mechanism was decisive. The stack was. Each layer reinforced the others. The structure produced the outcome that its incentives predicted.

Vioxx

Merck's anti-inflammatory drug Vioxx was approved in 1999. By 2000, internal Merck research showed elevated cardiovascular risk. The drug remained on the market until 2004. An estimated 60,000 deaths resulted.

The structural story: David Graham, an FDA safety reviewer, tried to raise the alarm internally. He was pressured by supervisors to soften his findings. The agency's post-market surveillance infrastructure was underfunded and understaffed—a direct consequence of the user fee model that prioritized pre-approval speed over post-approval monitoring. Whistleblowers faced career consequences, not rewards. The structure didn't just fail to catch the problem. It actively suppressed the people trying to flag it.

GRAS: The Food Additive Loophole

Under the GRAS (Generally Recognized As Safe) system, food manufacturers can designate their own additives as safe without FDA approval. Thousands of substances have entered the food supply through this mechanism with minimal independent review. The FDA lacks the resources to evaluate them independently—and the GRAS system was designed to accommodate that resource gap.

In other words, when the regulator lacks capacity, the regulated regulate themselves. Industry self-certification is industry self-regulation. The label says "FDA" but the process says "industry."

Why Good People Don't Fix Bad Structures

Consider the reality facing an individual FDA employee. The salary is modest by the standards of the expertise required. The pharmaceutical industry offers two to five times the compensation for the same skills. Career advancement depends on not making powerful enemies in an industry that employs your friends, former colleagues, and future employers. Aggressive regulation invites lawsuits, congressional pressure, and career risk. Lenient regulation means a smooth career and an industry job waiting when you're ready.

This is not a description of corrupt people. It's a description of rational people navigating a structure that rewards a specific set of behaviors. Good people inside bad structures produce bad outcomes—not because their character fails, but because the structure selects against the behaviors that would produce good outcomes. The problem is not personnel. The problem is architecture.

What This Means for Individuals

"FDA approved" does not mean "safe." It means approved by a structurally captured regulator operating under the pressures described above. This isn't nihilism—FDA approval is still informative, still better than no review at all. But it requires calibrated skepticism rather than blind trust.

Novel approvals warrant more skepticism than established ones, because post-market surveillance is the weakest link in the system. A drug used safely for decades has a track record that outweighs any pre-approval review. With newly approved drugs, you are, to a meaningful degree, the post-market surveillance—the early adopters who discover the problems that clinical trials missed or concealed.

Independent research—academic studies with transparent funding, international regulatory decisions, Cochrane systematic reviews—provides a partial corrective. No single source escapes all distortions, but triangulating across sources with different distortion patterns approximates truth better than trusting any single captured institution.

The Principle

Health regulators are structurally captured through the revolving door, user fee dependence, information asymmetry, and the logic of concentrated benefits versus diffuse costs. This capture is not a failure of the system. It is the system operating according to its actual incentive structure. The same pattern appears in financial regulation, environmental regulation, telecommunications—wherever a concentrated industry interacts with a resource-constrained regulator, capture is the equilibrium.

If the errors were random, their direction would be random. They aren't random. They consistently favor industry. That directional consistency is the signature of structural capture, not coincidental failure. The structure is the explanation. Everything else is downstream.

How This Was Decoded

This analysis applied the corruption stack framework to health regulation specifically, tracing the same capture dynamics visible in financial regulation, environmental oversight, and telecommunications. The inference paths: incentive analysis predicts capture, historical case studies (OxyContin, Vioxx, GRAS) confirm the predicted pattern, and cross-domain consistency—the same mechanisms producing the same outcomes across unrelated regulatory domains—strengthens the inference. The directionality test is key: if errors were random, they would favor industry and public roughly equally. The consistent industry-favoring direction is the structural fingerprint of capture, not coincidence.