New paper: Regulation as pruning of the adjacent possible

My article, Forgone Innovation: Regulation as Pruning of the Adjacent Possible, is now published in the European Economic Review. The paper develops an argument I have been working toward for some time: regulation does not only alter prices, incentives, or compliance costs. It can also alter what people are able to discover.

That distinction matters because innovation is not simply the result of more research and development, more capital, or better incentives. Much innovation arises through recombination. A tool developed for one purpose becomes useful in another. A technology becomes a component in a different system. A business model developed in one setting becomes the basis for experimentation elsewhere. Each successful experiment changes the menu of what can be tried next.

This is the idea behind the adjacent possible. At any moment, some possibilities are close enough to be attempted. They are not yet realized, but they are reachable from the existing stock of tools, knowledge, organizations, practices, and institutions. When one of those possibilities is realized, it opens still more possibilities.

My paper asks what regulation does to that process.

The standard way economists evaluate regulation is to compare costs and benefits within a fixed opportunity set. A rule may raise compliance costs. It may reduce transaction costs. It may protect consumers, standardize quality, prevent harm, or solve a coordination problem. Those effects matter. But they are not the whole story.

If innovation is combinatorial, then regulation can also change the set of combinations available for experimentation. It can remove building blocks. It can make unusual combinations too costly or legally uncertain to attempt. It can change the expected payoff from success. In those cases, the relevant cost is not only what firms spend complying with a rule. It is also the discovery that never takes place.

I call this forgone innovation.

The paper models regulation as operating through three mechanisms.

The first is subtraction. A rule may ban, mandate, or displace some element from the usable set of technologies, contracts, data pathways, or organizational forms. When that happens, we lose more than the element itself. We also lose the combinations that would have used it.

The second is plausibility. Some rules do not formally prohibit experimentation, but they make certain experiments too costly, uncertain, slow, or risky to try. A new arrangement may remain technically legal while becoming commercially implausible.

The third is payoff. Rules can alter the expected return from success. If a new model cannot be reimbursed, certified, scaled, defended in court, or integrated into existing systems, innovators may not pursue it, even when the underlying idea is feasible.

Together, subtraction, plausibility, and payoff help explain why the dynamic costs of regulation can compound. Missing first steps do not open the niches that would have followed. A rule that appears narrow in the present may eliminate branches of future development that no one can fully specify in advance.

This is also why forgone innovation is so difficult to measure. The missing products do not announce themselves. The lost business models do not appear in datasets. The most important losses may occur upstream, in experiments never attempted.

The main case in the paper is the European Union’s USB-C mandate. The policy was adopted to promote consumer convenience and reduce e-waste by standardizing wired charging for many portable electronic devices. Those goals are understandable. Standardization can solve real coordination problems. Anyone with a drawer full of old chargers can see the appeal.

But standardization also changes the innovation frontier. A common-charger mandate does not merely settle today’s connector dispute. It redirects investment, certification, supply chains, accessory markets, and consumer expectations toward one interface. Once a large jurisdiction mandates a standard, global producers may converge on that standard even outside the jurisdiction. A regional mandate can become a de facto global path.

The question then is what kinds of experimentation become less likely. What happens to rival connector designs? What happens to magnetic, optical, ultra-thin, wireless, modular, or still-unimagined alternatives? What downstream products and uses might those alternatives have enabled?

We cannot answer those questions fully. The relevant innovations may never be tried. That is precisely the problem.

The paper also points toward my next project: a forgone-innovation diagnostic for health care. Health care is full of technical possibility, but dense permission structures often determine what can actually be tried. Payment rules, licensing, liability, data governance, reimbursement, and institutional routines shape which combinations of care, technology, organization, and patient control become feasible.

The broader lesson is that regulation shapes innovation not only by changing incentives, but by changing the experiments that appear possible. If we care about long-run discovery, we need to ask different questions. Which building blocks does a rule remove? Which combinations become too costly or uncertain to try? Which successful experiments become difficult to scale? Which forms of problem-solving are crowded out before we can observe them?

Static analysis remains useful, but it is incomplete. In a creative economy, the opportunity set is not fixed. It evolves. Policy can widen or narrow it.

The hidden cost of regulation, then, is not simply that we may get less innovation. It is that we may get a different pattern of innovation: more effort spent adapting to the rule, less effort spent exploring alternatives; more innovation inside the mandated path, less innovation across paths. We may still observe patents, products, investment, and growth. But the branches that were pruned will be missing.

This paper is an effort to make those missing branches analytically visible.

The central claim is simple: regulation should be evaluated not only by what it accomplishes today, but by what it prevents tomorrow’s innovators from discovering.

Article behind this essay:
Forgone Innovation: Regulation as Pruning of the Adjacent Possible
European Economic Review, 2026