Distributive Capture

Changing a service provider can preserve a participant’s task while leaving a wallet condition or common legal rule untouched. Running a local node can change the provider relation while transferring provisioning work to the participant. MetaMask’s documented Ethereum paths make these distinctions visible. They separate provider access conditions, wallet configuration and feature conditions, local provisioning work and optional routing logic. Following the inputs and the term at issue prevents an asserted alternative from silently changing the task or the authority being assessed.

Access theory already distinguishes the ability to benefit from a resource from a right to it; infrastructure scholarship distinguishes allowing access from supplying its means. Platform research explains how enabling an ecosystem can concentrate control. The application here follows those relations through a technically open system. The proposed capture diagnosis concerns a term insulated from practical alternatives and effective institutional constraint, where a claim to review or remedy has an independently justified basis. Prior enabling benefit alone creates no such entitlement. Four official pages and released source support the layered illustration; they supply neither participant outcomes nor an empirical capture classification.

A person can hold the keys needed to authorize an Ethereum transaction while relying on other actors to submit it. A wallet supplies an interface; a hosted service supplies a network connection; the underlying system supplies shared rules. Those contributions make the activity practical in different ways. Authority over one of them does not amount to ownership or control of the whole undertaking.

The relationship also works in the other direction. A service can enable users to act while drawing on a protocol, maintained software and an existing network. Its value need not have been produced by the actor that controls the immediate access path. The research question concerns how authority over continued participation is allocated around this division of provision: who supplies the necessary inputs, who benefits, who sets each term and who can revise it? A service can distribute practical capacities and concentrate some decisions at the same time.

This article concentrates on layered technical systems in which formal permissions or shared resources coexist with separately organized access provision. It asks how practical access relates to an activity that the formal core permits, rather than proposing a general category for every intermediary.

This question does not require a universal sequence in which infrastructure first opens participation and later closes it. An enabling service may have conditions from its inception. Those conditions may change, or another actor may control a different layer throughout. Claims of an actual access expansion still require a baseline and evidence about the affected people. Simultaneous provision and conditional control require their own reconstruction, rather than an invented history of prior openness.

Within this scope, distributive capture is retained as a proposed critical diagnosis of control over a consequential term of participation. Two conclusions must be separated. Structural insulation concerns the absence of effective practical alternatives and countervailing institutions able to change the term. Calling that relation capture in a critical sense additionally requires showing how the insulated term conflicts with a specified, justified claim concerning the participants' or contributors' conditions of access. The claim can concern a continuing access relation or a termination subject to a justified obligation. Insulation alone proves neither misconduct nor an entitlement to service.

The distribution at issue is the provision of a practical capacity; equal ownership, equal benefit and equal governing power do not follow. The enabling relation identifies what participants stand to lose and whose contributions and decisions inquiry must follow. It does not settle whether their claims should prevail. This distinction keeps the description of concentrated power separate from the grounds for criticizing it.

An institutional claim also needs grounds. Supplying a useful service does not by itself create a duty to maintain it indefinitely, waive justified costs or serve every requested purpose. Applicable commitments, public duties, the importance of the task and reasonably induced reliance may supply reasons for review or remedy; their application requires argument and evidence. Moral or political grounds may exceed the rights recognized by current law; a formally available review under a discriminatory standard does not settle whether the standard is justified. Where no relevant claim is justified, the absence of a forum enforcing the participant’s preferred terms cannot establish failed institutional discipline. This qualification matters particularly when a temporary service ends on its declared terms.

Holding an authorization leaves open who can use a resource and on what terms. Ribot and Peluso distinguish access as the ability to benefit from something from property as a right. Their analysis follows access control, the resources and relationships through which actors maintain their own access, and the benefits that flow between actors. Control and maintenance can be divided among cooperating or competing actors (2003, 153–160). Peluso and Ribot’s later account further emphasizes relations with those mediating access and the role of authority (2020, 301–302).

Frischmann makes the corresponding infrastructure distinction explicit: allowing use of a resource differs from supplying access to it. Provision has costs that may fall on users, and charging to recover them need not establish an illegitimate exclusion (2005, 956 n.141). His demand-side account treats infrastructure as an input to varied downstream activities whose social benefits can be difficult to measure (957–959). It supports examining the wider productive arrangement while requiring attention to the resource, its uses and the conditions for commons management. It does not warrant an open-access prescription for every service.

Platform scholarship directly confronts enabling provision and concentrated power. Plantin et al. connect extensibility and user production to the control retained by platform builders as platforms acquire infrastructural roles (2018, accepted manuscript 8–9, 19–21). Blanke and Pybus’s study of mobile software development kits examines monopolization through decentralized technical integration. Their analysis relates services that help developers monetize applications to the expansion of platform infrastructures; its conclusion also emphasizes platforms' interdependence (2020, 2, 11). Enabling an ecosystem and concentrating parts of its infrastructure are already a central problem in this literature. Their mobile-SDK findings do not establish the distribution of power in the wallet path examined here.

The comparison needs two further distinctions. Network gatekeeping already assesses usable alternatives, switching and learning costs, and responsiveness without exit (Barzilai-Nahon 2008, 1500–1505). Capability scholarship distinguishes resources from real opportunities and realized actions, with personal, social and environmental conditions shaping what people can do (Robeyns 2017, §2.6.1, 38–41; §2.6.3, 45–47). A visible configuration option therefore cannot settle a participant’s alternatives. Nor does a person’s decision to remain show that departure is impossible.

The analytical use developed here is an application of these accounts to linked access paths. It distinguishes two errors in evaluating alternatives: transferring provision to the participant without establishing feasibility, and changing a provider while leaving the disputed term elsewhere intact. It asks which layer provides a capacity and which layer controls the disputed term, then tests whether a proposed alternative changes that term while preserving the participant’s substantive task. This can expose a comparison that silently changes its object: from a provider to a wallet, from an access policy to a feature condition, or from hosted provision to the participant undertaking infrastructure work. A new label is unnecessary when the established vocabularies already describe the result adequately. The value lies in whether following the layers changes the diagnosis or the appropriate institutional response.

Every inference remains tied to a specified participant class, task, carrier and period. The task must have a substantive definition justified independently of the disputed carrier and term. Preserving every incumbent feature would make substitution impossible by definition, while preserving bare protocol permission could ignore the work that made participation practical. A comparison should establish which functions, security properties, costs and accessibility conditions the task requires and whose judgments support that choice (Robeyns 2017, §2.6.2, 41–45). Required properties might come from participants' stated objectives, an applicable public standard or a pre-existing workflow. Sensitivity to narrower and broader task definitions should be reported when it changes the result.

The first step is to reconstruct provision: the inputs needed to perform that task, the actors supplying them and the work participants would inherit if one input disappeared. Residual describes a function outside a selected formal core, rather than one outside institutions or law. A protocol can leave connection services to other arrangements without leaving them ungoverned. A necessary function is also distinct from a necessary provider: needing network communication does not establish dependence on one RPC operator.

The next step follows the term at issue. Authority here denotes the practical ability to determine a term, without presuming that the actor is entitled to determine it. A provider may control a commercial condition, implement a public restriction or fail technically. A client developer may control a feature’s configuration requirements. The rule author, the implementer, the client selecting a route and the institution able to review the rule can differ. Control requires evidence of a particular ability to choose, vary, waive or enforce the term; operating the visible interface alone does not establish discretion over it. Substituting a provider tests a provider-specific term only if the alternative is independent with respect to that term. A common legal requirement can survive substitution without proving coordination or independent discretion by the providers.

Exit and voice concern these particular relationships (Hirschman 1970, 4–5). A route outside one provider can preserve a task even when users cannot change that provider’s policy. Conversely, an institution can constrain a term while participants remain dependent on the service. Institutional discipline means a usable route to effective scrutiny and constraint under justified standards. A forum can require reasons, enforce a duty or provide a remedy when warranted; it can also uphold a term through reasoned review. Organized institutions with reliable consequences can qualify without imposing their own legal sanction. A courteous reply alone establishes less than that capacity, and an adverse outcome does not establish its failure.

This separates the proposed capture diagnosis from ordinary specialization and disagreement. Organizing work through an intermediary can be useful precisely because participants cannot reproduce its scale or competence individually (Williamson 1985). Shared contribution to an ecosystem also does not by itself establish a claim on a particular provider’s future revenue or service. A study must show the specific authority insulated from alternatives and justified constraint. Regulatory capture has a different object, involving the displacement of a regulatory institution from a public purpose (Carpenter and Moss 2014). Neither a service restriction nor the implementation of law proves that a provider captured a regulator.

Technical decentralization likewise differs from the distribution of power (De Filippi 2019). Ferreira, Li and Nikolowa’s formal model addresses corporate influence in a proof-of-work industrial ecosystem (2023); it cannot supply empirical findings about this Ethereum wallet path. Increasing returns can stabilize dependence, but a configured default alone demonstrates no such process (Pierson 2000). Architecture, scale and persistence identify questions for investigation, rather than substitutes for the access and authority relations themselves.

The analysis distinguishes questions about provision and power from the grounds for a critical diagnosis:

Consider a stipulated service that makes a task practical for people unable to operate the underlying infrastructure. If an independent provider preserves their task under different terms, the incumbent-specific capture diagnosis fails. The replacement still being an intermediary is no objection: the disputed authority belonged to the incumbent, rather than to the abstract function of intermediation.

If substitution is impractical but an institution can effectively review and constrain the relevant term, dependence remains without satisfying the joint capture condition. This joint condition defines the particular ongoing insulation claim proposed here; an effective remedy or alternative does not erase an earlier wrong or settle every objection to the distribution of power. An institution can reasonably uphold a cost-recovery condition or reject a claim to a continuing subsidy. The fact that an alternative is costly supplies neither a service entitlement nor proof of failed discipline.

Absent a contrary duty or commitment, a declared temporary service provides a sharper negative case. Its ending may return work to users and reduce their opportunities. Those consequences deserve investigation, but they do not establish insulated governing authority over a continuing access path or a violated commitment. A different case arises where an operator continues controlling a necessary input while avoiding applicable review. The two should not be merged merely because both involve the loss of an earlier benefit.

Where practical alternatives and justified institutional constraint both fail, a bounded capture claim may become supportable. It still needs an account of the actor, term, enabling relation and period. An incumbent-specific claim cannot be rescued by changing the carrier to the entire market after a successful switch. A market-wide constraint requires its own evidence about control and alternatives. These contrasts test the diagnosis; they do not establish that the proposed term adds a new causal mechanism to access theory or gatekeeping.

The illustration examines official documentation about a MetaMask user accessing Ethereum through the configured default provider, alongside a bounded check of released routing source. Four MetaMask pages were retrieved with complete response bodies on 17 October 2026. They were selected because they describe the default, an access restriction, connection problems and alternative configurations. The exact Infura FAQ URL also requested that day returned HTTP 500; its error response is retained and supplies no current corroboration. This study contains no original on-chain collection or participant experiment.

The illustration separates three decisions that a general question about decentralization could conflate:

These are analytical comparisons, rather than observed participant outcomes. The common-rule row is conditional: the documentation attributes Infura’s restriction to sanctions but does not establish the rules or implementation of every alternative provider. The regional guidance, feature statement and local-node requirements concern different terms and work. The source-controlled fallback adds a possible route through the provider layer, while leaving its deployed availability and the wallet’s routing decisions unresolved. Following the actual term avoids crediting protocol openness with a practical alternative it has not demonstrated, or treating a provider switch as evidence about every layer of control.

The service requirement changes the interpretation of feature loss. For a stipulated task of basic transaction submission, losing Smart Transaction capabilities does not itself defeat an alternative. If a protection is independently justified as necessary to the task, the comparison must instead establish whether another route supplies an adequate replacement. The documents do not decide which requirement applies to a participant class, or demonstrate successful substitution under either definition.

These distinctions suggest two questions for empirical study. For a specified class and task, does a proposed alternative transfer provisioning work that participants cannot sustain, or does independent shared provision preserve their opportunity? For an identified provider switch, which access and wallet conditions change, and which remain? Sustained task completion would support that route as an alternative; failure linked to the transferred work would support a narrower explanation of its limits. A switch that also changes the supposedly remaining wallet condition would correct the proposed allocation of terms. These results would concern the specified comparison, rather than establish or refute a general mechanism.

A participant study could compare completion, setup work, recurring cost, reliability, privacy, security assumptions and required features across justified alternatives. A retrospective comparison could instead trace a documented change in provision or a disputed term. Either design should fix the class, task, carrier and service requirements before interpreting its outcomes, and explain any revisions. Prospective registration is one safeguard against moving those criteria, rather than a universal condition of institutional research.

Institutional inquiry needs a different body of evidence: the applicable commitments or duties, who can invoke review, the forum’s authority, qualifying demands, responses and remedies. Its observation period and search coverage must be explicit. An unanswered complaint, an absent record, indifference and a rejected claim have different meanings. Establishing that a forum works does not require every complainant to prevail; establishing failure requires more than failing to observe a preferred response.

The article develops a conceptual synthesis and a dated documentary illustration. The pages captured on 17 October 2026 establish the configurations and conditions they describe. The released-source comparison and eleven local predicate checks establish specified routing conditions, rather than an installed extension’s build variables, remote flags or regional behavior. They certify neither actual use nor participant outcomes, a measured access expansion, population dependence, effective substitution or institutional responsiveness.

The conceptual comparison is a purposive reading of access theory, infrastructure economics, platform infrastructure, gatekeeping and capabilities. Each source can explain part of the original question without the proposed term. The comparison uses the versions and sections cited; it is not an exhaustive originality review. Frischmann’s printed journal pages, Ribot and Peluso’s published texts, Blanke and Pybus’s journal pagination and Plantin et al.'s accepted-manuscript pagination remain distinct.

The strongest remaining conceptual test is whether following inputs, terms and decision makers improves an explanation or institutional response beyond the established accounts. A new category earns no status merely by combining their distinctions. Extending the argument to identity services, cloud infrastructure or historical enclosures would require separate actors, histories and sources. None supplies a demonstrated case here.

Practical participation is produced through relations among shared systems, service providers, client software and participants' own work. Authority over its terms can be allocated differently from the work and resources enabling it. Access theory and infrastructure scholarship already explain that relation; the application developed here follows it through distinct technical and institutional layers.

The MetaMask comparison shows why the object of an alternative matters. Changing an RPC provider, preserving a wallet feature and contesting a common rule address different decisions. A claim of capture must identify the relevant decision and establish both the limits of practical alternatives and the failure of justified institutional constraint. The current materials support that analytical distinction while leaving the empirical diagnosis open.

The MetaMask illustration uses four official pages captured on 17 October 2026. A fresh retrieval of the custom-RPC guidance confirms its stated feature condition: selected custom RPC routes disable Smart Transaction capabilities, including MEV protection, until the default route is restored. The documentation supplies a configuration consequence; it does not establish whether a participant needs that feature, can obtain an adequate substitute or successfully switches.

The released-source comparison preserves release commit 217c943356cb96e421fe3f4737d94a36361c106c, the extension lockfile, NetworkController 36.0.0 and eleven targeted function checks. These establish optional routing predicates, rather than an installed extension’s build variables, live flags, successful failover or institutional independence. The October response bodies and source decisions are retained separately; the older July extracts do not preserve the original page bodies.