Channels of Power in Production Networks: A Taxonomy for Measuring Technological Dependence¶

Stefane Fermigier (Abilian) · sf@abilian.com

Draft working paper v0.7, 2026-08-04. Complete draft; the operation and structure channels are anchored in measured data; the cognition-channel observables are stated as measurement protocols awaiting their datasets.

Download the PDF

Abstract¶

Measuring technological dependence requires knowing every way one actor in a production network can move another. The measurement traditions each hold a partial list: trade economics sees supply, corporate-control studies see equity, the coercion literature sees law, platform economics sees distribution. The soft channels through which markets are actually steered day to day (analyst gatekeeping, standards authorship, narrative and expectation management, training and certification pipelines, the governance of open-source commons, and lobbying) are studied in separate literatures that the measurement work rarely touches.

This paper assembles that list, aiming at completeness. The claim is inductive; a sixteenth channel would be a finding. Fifteen channels fall into three families, distinguished by what they act on. Operation channels act on the feasibility and cost of running a technology stack today (supply, technology and licensing, operational service, data and identity, distribution, legal reach). Structure channels act on who exists and who controls what (equity, capital, competence supply). Cognition channels act on what buyers consider, believe, and know how to do (evaluation gatekeeping, standards authorship, narrative, epistemic formation, commons governance, rule shaping). Each channel is specified by mechanism, timescale, observables, formal slot and literature anchor.

The thesis is that the cognition family drives a wedge between option sets that measurement must keep apart: the feasible options that exist, the considered options that enter decisions, and the exercised options that revenue keeps alive. If unexercised options decay, as the paper hypothesises, consideration capture today becomes feasibility collapse tomorrow, and soft power turns hard without any actor ever denying anything.

Applied to the European digital stack, the taxonomy shows which channels the existing measurement already covers, which are quantifiable next and how. It predicts that sovereignty audits that ignore the cognition family overstate the breadth Europe actually has, and sets out a test of that prediction on public procurement data, whose first stage finds a median of two bids per lot in 1,657 cloud and productivity award notices of 2024 to 2026, against measured feasible sets of twelve and seven options: the count does not refute the wedge and cannot confirm it. The schema is domain-portable: nothing in it is specific to digital technology.

Keywords: dependence, production networks, soft power, industry analysts, standards, expectations, skills, open source, lobbying, measurement.

JEL: F52, L14, D85, L86, D83, O33.

The fifteen channels of power in three families, shaded by how far the European stack's measurement already covers each. The cognition family is the least measured.

About this paper¶

Cite this paper¶

Fermigier, S. (2026). Channels of Power in Production Networks: A Taxonomy for Measuring Technological Dependence.
Draft working paper v0.7, Abilian Econ Lab.
https://econ.lab.abilian.com/papers/channels/