Regulation as a Field: Capture, Public Interest, and Evidential Standards

The Two Stories

There are two canonical stories about why regulation exists. In the first, regulation is a public-spirited response to market failure. Markets produce too much pollution, too little safety, too many monopolies, and too much fraud. Government intervenes to correct these failures, acting on behalf of the broad public interest. In the second, regulation is a tool wielded by organized interest groups, particularly the industries being regulated, to protect themselves from competition, raise barriers to entry, and transfer wealth from consumers and taxpayers to the regulated firms and their allies.

The first story, the public interest theory, dominated thinking about regulation from the Progressive Era through the mid-twentieth century. The second story, the capture theory, emerged in the 1960s and 1970s as economists at the University of Chicago applied the tools of price theory to political behavior. The tension between these two stories has shaped regulatory policy, institutional design, and academic research for more than half a century. Neither story is complete on its own, and the field has moved toward a more nuanced understanding that incorporates elements of both while asking a harder question: how do we know if regulation actually works?

The Public Interest Theory

The public interest theory of regulation starts from welfare economics. Markets are efficient under certain conditions: perfect competition, complete information, no externalities, no public goods, no natural monopolies. When these conditions fail, there is a potential role for government intervention to improve on the market outcome. Pollution is the textbook externality: firms impose costs on third parties that are not reflected in prices, leading to too much production and too much pollution. A tax, a cap-and-trade system, or a direct regulation that limits emissions can, in principle, move the outcome closer to the social optimum.

Natural monopoly provides another justification. In industries with large fixed costs and declining average costs, like water distribution, electricity transmission, or railroads, competition may be wasteful or unstable. A single regulated provider, subject to price controls and service obligations, may produce a better outcome than either unregulated monopoly or wasteful duplication.

Information asymmetry provides a third. When sellers know more about product quality than buyers, markets can break down through adverse selection (only low-quality products are offered) or moral hazard (sellers reduce quality once the sale is made). Regulation that mandates disclosure, sets minimum quality standards, or licenses practitioners can improve market functioning.

The public interest theory has an appealing logic, and it describes the stated rationale for much of the regulatory apparatus in modern democracies. Environmental regulation, food and drug safety, financial supervision, occupational licensing, and consumer protection all rest on some version of the market failure argument. But the theory has a crucial weakness: it assumes that the regulator is benevolent and competent, that the regulator actually pursues the public interest rather than its own interests or the interests of politically powerful groups. This assumption is where the capture theory enters.

Capture Theory: Stigler’s Revolution

George Stigler’s 1971 article “The Theory of Economic Regulation” is one of the most influential papers in the Chicago economics tradition. Stigler’s central claim was stark: regulation is not supplied in response to public demand for the correction of market failures. It is demanded by industries that benefit from it and supplied by politicians who exchange regulatory favors for political support, campaign contributions, and post-government employment opportunities.

Stigler argued that industries have strong incentives and superior ability to organize for political action. The benefits of regulation to the regulated industry (higher prices, restricted entry, subsidies, protection from competition) are concentrated on a small number of firms with large individual stakes. The costs of regulation to consumers are diffuse, spread across millions of individuals with small individual stakes. The logic of collective action, as analyzed by Mancur Olson, predicts that the concentrated interests will organize effectively while the diffuse interests will not.

The result, according to Stigler, is that regulatory agencies tend to serve the interests of the industries they are supposed to regulate, not the interests of the consuming public. This “capture” can take many forms: entry restrictions that protect incumbents from competition, price floors that prevent price-cutting, quality standards that are calibrated to the capabilities of existing firms rather than to the needs of consumers, and enforcement patterns that punish new entrants more harshly than established firms.

Stigler’s evidence included the regulation of trucking (entry restrictions that raised prices and profits for incumbent truckers), the regulation of electricity (rate-setting that allowed utilities to earn above-normal returns), and occupational licensing (barriers to entry that raised practitioners’ incomes while restricting supply). In each case, the pattern was the same: regulation benefited the regulated industry at the expense of consumers.

Peltzman’s Extension

Sam Peltzman extended Stigler’s framework in a 1976 paper that made the capture theory more flexible. Where Stigler had implied that regulation was entirely captured by the regulated industry, Peltzman argued that the politician maximizes political support from all affected groups, not just the industry. The equilibrium level of regulation reflects a balance between the interests of producers (who benefit from higher prices and entry restrictions) and the interests of consumers (who benefit from lower prices and wider choice).

Peltzman’s model predicted that regulation would be most favorable to producers when the industry is concentrated and well-organized, and most favorable to consumers when consumer groups are well-organized or when the regulatory issue is politically salient. This allowed for a richer set of outcomes than pure capture, including cases where regulation genuinely served the public interest because the political conditions aligned that way.

The Peltzman extension was important because it moved the capture theory from a blanket condemnation of regulation to a more analytical framework that could explain variation in regulatory outcomes. Some regulations are captured; others are not. The question is not whether regulation is good or bad in general, but what institutional conditions make capture more or less likely.

The Revolving Door

One mechanism of capture that has received extensive attention is the revolving door between regulatory agencies and the industries they regulate. Regulators frequently move to high-paying positions in the industries they oversaw, and industry executives frequently move into regulatory positions. This pattern creates several potential problems.

First, regulators who anticipate future employment in the industry may be reluctant to take actions that antagonize potential employers. This does not require conscious corruption; the mere prospect of a future career in the industry can subtly shape the regulator’s judgment about what is reasonable, what is proportionate, and what the evidence supports.

Second, industry executives who move into regulatory positions bring with them the industry’s perspective, its priorities, its framing of problems, and its definition of what constitutes reasonable regulation. They may be genuinely committed to the public interest, but their understanding of the public interest is inevitably shaped by their professional experience and social networks.

Third, the revolving door creates information asymmetries. Regulators who have worked in the industry have access to knowledge and networks that outsiders lack, which can be valuable. But the same knowledge and networks can make them more sympathetic to the industry’s arguments and less receptive to outside perspectives.

The empirical evidence on the revolving door is mixed. Some studies find that former industry executives who become regulators are more lenient toward the industry; others find that they are actually tougher, perhaps because they understand the industry’s strategies better and are harder to fool. The effects may depend on the specific regulatory context, the institutional design of the agency, and the political environment.

Independent Regulatory Agencies: Design and Incentives

The institutional design of regulatory agencies is a crucial but often neglected dimension of the regulation debate. Independent regulatory agencies, like the Securities and Exchange Commission, the Federal Communications Commission, or the Environmental Protection Agency, are designed to be insulated from short-term political pressure. Commissioners serve fixed terms that overlap with presidential administrations, cannot be removed except for cause, and are supposed to exercise expert judgment rather than respond to political direction.

The theory behind independence is straightforward: if regulators are subject to political pressure, they will make decisions that serve political interests rather than the public interest. Independence allows regulators to take actions that are unpopular in the short run but beneficial in the long run, such as imposing costs on polluters or breaking up monopolies.

But independence also creates accountability problems. If regulators are insulated from political pressure, who ensures that they serve the public interest rather than their own interests or the interests of the regulated industry? The answer, in theory, is a combination of legislative oversight, judicial review, transparency requirements, and professional norms. In practice, these accountability mechanisms are often weak, and the very insulation that protects regulators from political pressure also protects them from correction when they err.

The design of regulatory agencies involves a series of trade-offs: independence versus accountability, expertise versus democratic control, consistency versus flexibility, speed versus deliberation. There is no single best design; the appropriate balance depends on the specific regulatory context, the characteristics of the regulated industry, and the political culture in which the agency operates.

Cost-Benefit Analysis: Sunstein’s Defense and Its Critics

Since the early 1980s, cost-benefit analysis has been the dominant method for evaluating proposed regulations in the United States. Executive orders from Presidents Reagan, Clinton, Obama, and Trump (with varying emphasis and detail) have required regulatory agencies to conduct cost-benefit analyses of major rules and to demonstrate that the benefits justify the costs.

Cass Sunstein, who served as the head of the Office of Information and Regulatory Affairs under President Obama, has been the most prominent academic defender of cost-benefit analysis as a tool for regulatory decision-making. Sunstein argues that cost-benefit analysis promotes rational, transparent, and accountable regulation. It forces regulators to identify the problem they are trying to solve, to quantify (insofar as possible) the costs and benefits of proposed solutions, and to choose the approach that maximizes net benefits. It protects against both over-regulation (imposing costs that exceed benefits) and under-regulation (failing to address genuine market failures).

Critics raise several objections. The most fundamental is the monetization problem: cost-benefit analysis requires putting dollar values on things that are not naturally expressed in dollars, such as human lives, environmental quality, species diversity, and distributional equity. The “value of a statistical life,” currently set at roughly $10 to $12 million in U.S. regulatory analysis, is derived from studies of wage premiums for risky jobs and stated preference surveys. It is a useful fiction, but it is a fiction, and different methodological choices can produce very different numbers with very different policy implications.

A second objection concerns distributional effects. Standard cost-benefit analysis adds up all costs and all benefits without regard to who bears them. A regulation that imposes large costs on poor communities while producing diffuse benefits for wealthy ones might pass a cost-benefit test while exacerbating inequality. Distributional weighting schemes have been proposed but are controversial and rarely used in practice.

A third objection is that cost-benefit analysis systematically favors the status quo. The costs of regulation are typically borne upfront and are relatively easy to measure: compliance costs, administrative costs, price increases. The benefits are often long-term, diffuse, and difficult to quantify: avoided deaths, improved air quality, reduced systemic risk. This asymmetry in measurement ease tends to make costs more salient than benefits, biasing the analysis toward less regulation.

Modern Regulatory Challenges

The frameworks developed in the context of twentieth-century industrial regulation face new challenges in the twenty-first century. Three stand out.

Technology platforms. Companies like Google, Amazon, Meta, and Apple operate platforms that function as marketplaces, communication channels, and information gatekeepers for billions of people. The traditional regulatory toolkit, designed for industries with physical assets and geographic boundaries, is poorly suited to platforms that are global, digital, and characterized by powerful network effects. Should platforms be regulated as common carriers? As media companies? As natural monopolies? As something entirely new? The answers are not obvious, and the regulatory frameworks are still being developed, with the European Union’s Digital Markets Act and Digital Services Act representing the most ambitious attempts to date.

Artificial intelligence. AI systems are increasingly used in high-stakes decisions: credit scoring, hiring, criminal sentencing, medical diagnosis, autonomous vehicles. The potential for algorithmic bias, opacity, and unaccountable decision-making raises regulatory questions that existing frameworks are not well-equipped to handle. How do you regulate a system whose decision-making process is not fully understood even by its creators? How do you balance the benefits of AI (efficiency, accuracy, scalability) against the risks (bias, opacity, job displacement)? The regulatory approaches being developed range from sector-specific rules (like the FDA’s framework for AI in medical devices) to horizontal frameworks (like the EU AI Act) to voluntary industry standards.

Financial innovation. Cryptocurrencies, decentralized finance, algorithmic trading, and new forms of financial intermediation challenge regulatory frameworks designed for banks, brokerages, and insurance companies. The regulators’ toolkit, developed over decades of experience with traditional financial institutions, may not be adequate for a world in which financial intermediation can be conducted by algorithms, smart contracts, and decentralized protocols that have no identifiable operator, no physical location, and no clear jurisdictional home.

The Evidence Question

Perhaps the most important and least discussed question in regulatory economics is the evidence question: how do we know if regulation works? The public interest theory assumes that regulation achieves its stated goals. The capture theory assumes that it serves the regulated industry. But what does the evidence actually show?

The honest answer is that we know less than we should. Rigorous evaluation of regulatory effectiveness is surprisingly rare. Most regulations are adopted without baseline measurement, without control groups, and without pre-specified criteria for success or failure. Post-implementation reviews, when they occur at all, are typically conducted by the same agencies that adopted the regulation, creating obvious conflicts of interest.

The difficulty is partly methodological. Regulations affect entire industries or populations, making controlled experiments difficult. The counterfactual, what would have happened without the regulation, is inherently unobservable. And the effects of regulation often unfold over decades, making timely evaluation impossible.

But the difficulty is also partly institutional. Agencies have incentives to demonstrate that their regulations are effective and to resist evidence that they are not. Industry groups have incentives to exaggerate compliance costs before adoption and to downplay them afterward (if they have been successfully captured) or to exaggerate them persistently (if they are seeking repeal). Consumer and environmental groups have incentives to emphasize benefits and minimize costs. The political environment in which regulatory evidence is produced and consumed is not one that rewards dispassionate analysis.

Improving the evidence base for regulation is not a glamorous cause, but it may be the most important contribution that economics can make to regulatory policy. Better data, better evaluation methods, sunset provisions that force periodic review, independent evaluation bodies, and a culture that treats regulatory effectiveness as an empirical question rather than an ideological one would all help. The alternative is a regulatory debate in which both sides argue from theory and anecdote, and the public has no reliable way to judge whether the regulations ostensibly adopted on their behalf are actually serving their interests.

Where the Field Stands

The economics of regulation has matured considerably since the simple days of “market failure requires regulation” and “regulation is always captured.” The field now recognizes that both outcomes are possible, that the institutional design of regulatory agencies matters enormously, that evidence of regulatory effectiveness is scarce and should be demanded, and that the challenges of regulating a digital, global, and rapidly innovating economy require new thinking.

What the field has not yet achieved is a reliable framework for predicting when regulation will serve the public interest and when it will be captured, or for designing institutions that consistently produce the former rather than the latter. That remains the central challenge, and meeting it will require not just better economics but better institutional design, better evidence, and a more honest conversation about the limits of what regulation can and cannot accomplish.