Dependency Concentration

Critical systems centralize because the forces that pull them inward are persistent and the forces that pull them outward are episodic. Economies of scale, regulatory cost, specialised expertise, and network effects all reward consolidation continuously. Diversification tends to be argued for only after a disruption, and the argument fades as the disruption recedes.

Consolidation is also the path of least institutional resistance. Procurement frameworks, certification regimes, and operating standards are easier to maintain against a small number of well-known counterparties. Each individual decision to extend an incumbent's role is defensible. The cumulative effect is rarely examined.

Alternatives do not usually disappear through prohibition. They disappear because the conditions that sustained them — workforces, supplier ecosystems, institutional knowledge, financing — quietly cease to exist. Once a parallel capability is no longer in continuous use, it stops being maintained, and once it stops being maintained, the cost of reviving it grows quickly.

What looks at any single moment like a free choice between options is, over decades, a one-way ratchet. The alternative that is not chosen this year is somewhat harder to choose next year, until eventually it is no longer a real option at all.

A concentrated system carries risk that is correlated rather than distributed. A failure inside a single operator, vendor, or jurisdiction can propagate across what looks from the outside like a diverse field of users. The breadth of impact is decoupled from the size of the originating fault.

Concentration also narrows the range of institutional responses. Regulators, customers, and governments find that their leverage depends on the continued cooperation of the very operators they are meant to constrain. Substitution, the ultimate sanction in most markets, ceases to be available.

Concentration is durable in part because it delivers genuine benefits. A small number of well-resourced operators can sustain investment, recruit specialised staff, and standardise practices in ways that fragmented markets cannot. Many of the safest and most reliable critical systems in the world are also among the most concentrated.

The pattern is therefore not one to be condemned, but one to be recognised. The question for governance is not whether concentration exists, but whether the system retains enough optionality, transparency, and reversibility to remain governable as concentration continues.