The most compelling argument against excessive centralisation comes from modern systems theory. A highly centralised nation-state resembles what engineers call a “star topology”—a network in which all nodes connect to a single central hub. Such systems may operate efficiently in stable conditions, but they contain a fatal vulnerability: a single point of failure. If the hub is compromised by error, incompetence, corruption, external attack, or information overload, the entire system can collapse.
By contrast, a federal system functions as a “distributed network.” Authority is dispersed across multiple nodes, redundancy is built into the structure, and failures remain localised. Shocks that would cripple a monolithic system can be absorbed, contained, or bypassed. Federalism, therefore, is not a political accommodation to regional diversity; it is a structural design principle for resilience—the systems equivalent of the old maxim that one should never place all eggs in a single basket.
Origins of distributed networks
The intellectual genealogy of this idea traces back to the Cold War era, when nuclear anxiety was at its peak. The US confronted a grave vulnerability: its command-and-control communications system was highly centralised. A decapitation strike on Washington DC could blind and paralyse the entire system. Paul Baran, an engineer at the RAND Corporation, was tasked with designing a communications network that could survive such an attack.
In his landmark 1964 paper, On Distributed Communications, Baran rejected the centralised “star network” architecture. Instead, he proposed a “distributed mesh network”—a fishnet-like structure in which every node connects to several others and no single node is indispensable. If one node is destroyed, messages would simply route around the damage through other nodes.
This architecture became the blueprint for ARPANET, which eventually evolved into the Internet. The Internet functions not because it has a strong centre, but precisely because it has none. It is a federation of autonomous networks governed by shared protocols. The lesson extends beyond telecommunications. The survival and long-term resilience of a continental polity like India require a distributed, fishnet-like architecture of governance—one in which authority is dispersed across multiple nodes rather than concentrated in a single centre.
Risks of tightly coupled systems
Sociologist Charles Perrow deepened this insight in his influential work Normal Accidents (1984). Perrow argued that failures in tightly coupled systems—systems in which components are rigidly interconnected with little room for delay, discretion, or correction—are not anomalies but inevitable outcomes of complexity. When a tightly coupled system fails, small errors propagate rapidly through the entire structure, escalating into a cascading breakdown.
A highly centralised state resembles such a system. When authority, resources, and decision-making are concentrated at one point, even routine administrative errors—a delayed fund release, a flawed directive, or a data mistake—can rapidly disrupt multiple jurisdictions simultaneously.
Decentralisation loosens this coupling. It functions as a circuit breaker, confining mistakes within limited jurisdictions. If one State pursues a flawed policy, the damage remains local rather than systemic. This compartmentalisation of risk is a defining feature of resilient systems across engineering, finance, and organisational design.
In simple terms, tight coupling associated with centralisation allows local failures to escalate into national crises, whereas loose coupling under decentralisation keeps failures contained within local jurisdictions and therefore manageable.
Insights from military doctrine
The same logic appears in modern military doctrine. The Prussian concept of Auftragstaktik, or mission command, emerged from the recognition that the “fog of war” makes centralised micromanagement impossible.
Instead of issuing detailed instructions for every battlefield contingency, commanders communicate the objective—the commander’s intent—and delegate discretion to subordinate officers. Field units adapt tactics to local conditions while remaining aligned with the overall mission. This decentralised structure ensures that military units remain operational even if communications with central command are disrupted.
Civil governance operates in a similarly complex environment. A Prime Minister in New Delhi cannot micromanage a District Collector responding to floods, epidemics, or crop failures any more than a General can direct every platoon during combat. Decentralisation allows institutions to continue functioning even when central authorities are overwhelmed or temporarily incapacitated.
Governance through antifragility
Nassim Nicholas Taleb extended the argument further through his concept of “antifragility,” elaborated in Antifragile: Things That Gain from Disorder (2012). A resilient system can withstand shocks and return to equilibrium. An antifragile system goes further: it improves because of stress and volatility.
Centralised systems, Taleb argues, tend to be fragile. In the pursuit of control and uniformity, they suppress variation and eliminate small failures. Over time, hidden risks accumulate beneath an appearance of stability until they erupt as catastrophic breakdowns.
Decentralised systems behave differently. By permitting local variation and experimentation, they allow small failures to occur frequently but harmlessly. Each failure generates learning, adaptation, and institutional improvement. Federal systems, therefore, approximate antifragility. Policy errors remain local; successful innovations spread; and the system evolves through continuous experimentation.
Examining Indian examples
A recent illustration of how centralisation can amplify systemic risk is provided by the National Testing Agency (NTA), which since 2018 has centralised several high-stakes national entrance examinations. In 2024, NEET-UG was compromised by confirmed paper leaks across multiple States; UGC–NET had to be cancelled following a security breach; and CUET results were marred by repeated delays.
The Department-Related Parliamentary Standing Committee on Education (371st Report, 2025) concluded that at least five of the fourteen examinations conducted by the NTA suffered from serious deficiencies. Recognising the systemic risks inherent in such centralisation, the Kurian Joseph Committee on Union–State Relations, constituted by the Government of Tamil Nadu, has recommended that the NTA be disbanded and that responsibility for conducting entrance examinations revert to the diverse agencies that administered them prior to 2018.
Decentralisation as insurance policy
The Resilience Argument yields a counter-intuitive insight: an excessively strong Centre can produce a weak nation. India’s continental scale and diversity make it unsuitable to be governed as a “star network.” It must instead operate as a “distributed mesh”—with States as empowered nodes possessing fiscal, administrative, and decision-making capacity. Federalism, in this sense, is an insurance policy. It ensures that when one node fails, the system does not collapse; when one authority falters, others continue to function.
To be concluded
The author is retired IAS officer of Tamil Nadu cadre, and Member, High-Level Committee on Union-State Relations constituted by the Government of Tamil Nadu