Administrative Relations Beyond Organizational Charts

Administrative relations are commonly approached through organizational jurisdictions, coordination arrangements, interorganizational networks, or integration. Each perspective illuminates an important part of administrative order, but none by itself specifies how multiple, non-coincident boundaries and directed administrative relations compose into an end-to-end configuration from policy space to the administrative addressee and back. This article develops a provisional theory model of administrative relations from a long sequence of empirical and conceptual inquiries into jurisdictional overlap, delivery systems, information networks, jurisdictional change, and governance frames. The model treats policy space as an embedding context; allocation, boundary, and relational configuration as mutually constitutive rather than sequential; demarcating, bridging, and integrating (D–B–I) as boundary-related relational operations/configurations rather than an exhaustive typology of interorganizational behavior; and administrative reach as a configuration of multiple directed paths carrying authority, finance, information, execution, supervision, accountability, and access/contact. It further distinguishes delivery from encounter and response, and treats cross-level relation-to-relation translation as analytically irreducible to a simple implementation chain. An internal stress test against earlier cases requires five revisions: boundaries may be graded or contested; formal allocation is not a prerequisite for all relations; D–B–I does not subsume competition, conflict, power, trust, or negotiation; cross-level relational translation cannot be reduced entirely to path; and administrative delivery must be traced through encounter, response, and return flows. The resulting model characterizes administrative order as patterned configuration–movement–reconfiguration rather than as jurisdictional congruence or organizational integration. Its strongest prospective contribution lies in combining boundary-specific relational analysis, multiple directed paths, cross-level relational translation, and end-to-end tracing within a single analytical architecture. The model remains provisional and is offered for empirical testing and subsequent revision.