Institutions often expand what people can do. They coordinate activity, accumulate knowledge, establish infrastructure, reduce uncertainty, and make increasingly sophisticated cooperation possible.
Those achievements can have consequences beyond their intended purposes. Investments become difficult to reverse. Expertise becomes concentrated. Participants acquire dependencies. Routines become entrenched. Institutions created to solve particular problems can acquire growing influence over which alternatives remain feasible.
The hypothesis is not that success necessarily produces domination, or that dependence is inherently objectionable. Complex societies depend on shared infrastructure, specialized expertise, durable commitments, and coordinated institutions. The question is whether those arrangements preserve effective possibilities for their own transformation.
An institution’s success at expanding present capabilities can, under certain conditions, diminish society’s independent capacity to revise, replace, or transcend the arrangements through which those capabilities are provided.
The literature supports several mechanisms through which those possibilities can narrow. It provides less evidence for treating them as one general developmental process. That distinction matters: the framework should name something the evidence could disconfirm, not redescribe every durable institution as a case of foreclosure.
What the evidence supports
Path dependence
Early adoption can generate cumulative advantages, and the cost of changing an arrangement is partly produced by the history of operating within it. W. Brian Arthur’s model of increasing returns and Paul Pierson’s account of path dependence in politics make the mechanism clear.
Gregory Unruh’s carbon lock-in is the strongest substantive example, though it is primarily a conceptual explanation rather than a definitive causal test. The QWERTY case is weaker than it is often made to seem; Liebowitz and Margolis challenged the claim that QWERTY persists because it is inferior.
Organizational learning
Success can narrow what an organization learns to do next. Levinthal and March’s competency traps describe how proficiency in familiar activities can crowd out exploration. Historical institutionalists add that change often happens through layering, drift, and conversion rather than wholesale replacement. Persistence is not itself evidence of foreclosure; the task is to distinguish continuity that preserves useful coordination from continuity that restricts alternatives.
Accountability and administrative burden
Formal access to a procedure does not ensure comparable influence over its outcomes or rules. Galanter’s repeat-player advantage and Herd and Moynihan’s administrative burden show how rights can exist on paper while remaining practically unusable. Transparency is neither sufficient nor irrelevant: the Montreal Protocol and the leaded-petrol phaseout show that information can contribute to structural correction when it is connected to organization, resources, enforceable authority, and feasible alternatives.
Platform economics
Network effects and switching costs can make a platform the intermediary for dependent activity. Katz and Shapiro’s work on compatibility, and Adner and Kapoor’s work on innovation ecosystems, explain why competition between products cannot be understood apart from the systems around them. The EU’s Digital Markets Act is an ongoing test, not a settled success; emerging openings and possible circumvention have both been documented.
Recursive correction
Ostrom’s commons governance and the constitutional transitions in South Africa and Chile show that higher-order rules can be contested and revised. They do not show that every affected person possesses equivalent practical authority over those rules. Disability research adds a further qualification: the expansion of productive capability is not an unqualified good if standing and security become conditional on conforming to an institution’s preferred model of capacity.
Second-order consequences
Microcredit evaluations and the One Laptop per Child trial show the limits of using inputs or technical access as proxies for developmental progress. They do not show that the interventions failed because recipients lacked institutional authority. That would require separate evidence. Ecological analogies such as Holling’s adaptive cycle are useful for constructing hypotheses, not for demonstrating that social institutions behave like ecosystems.
What the evidence does not establish
Most of this validates a mechanism, not the full sequence from capability expansion to developmental foreclosure. High switching costs may constrain alternatives without determining who has the power to change them. Institutional persistence may reflect useful coordination rather than entrenchment. A platform may create substantial social value while making independent activity more dependent on it, and those two facts need to be weighed rather than collapsed.
The framework’s contribution is therefore not the discovery that institutions resist change. It is the proposed distinction between expanding capabilities through an institution and expanding society’s effective authority over the future of that institution.
What would test it
A strong test compares arrangements that perform similar functions while differing in how they distribute dependence and corrective authority.
Comparative institutional studies compare jurisdictions using similar technologies under different rules for portability, interoperability, public provision, ownership, and participation. They measure not only adoption and service quality but the cost of creating or adopting alternatives.
Longitudinal studies of institutional success follow organizations as their capabilities increase, and ask whether gains in independent expertise, exit feasibility, and affected people’s ability to secure revisions move with them. The key question is whether success predicts a narrowing of external options, and under what conditions that relationship reverses.
Studies of contested correction identify cases where affected people exposed a failure, demanded a change, and met an explicit decision process. They distinguish correction of a particular decision from revision of rules, resource distribution, and governance.
The indicators are independent capability, switching and replacement costs, developmental optionality, correction depth, distribution of authority, counterfactual social value, and temporal durability. No single measure establishes collective self-government.
What would challenge the hypothesis
The strongest challenges are institutions that become more successful while expanding independent alternatives; concentrated systems that remain substantively corrigible; decentralized systems that restrict transformation; and replacements that destroy valuable coordination without producing greater authority.
Preserving the capacity for transformation does not mean every possible transformation should be undertaken. Rejecting an established arrangement is not automatically progress. The framework has to distinguish the value of having an effective choice from the value of exercising it.
Where the full review lives
The complete review, with the citations, qualifications, and the cases that cut against the hypothesis, is in Research Grounding at The Crumple Zone.