Centralization and decentralization for resilient infrastructure and complexity

The capacities of our infrastructure systems to respond to volatile, uncertain, and increasingly complex environments are increasingly recognized as vital for resilience. Pervasive across infrastructure literature and discourse are the concepts of centralized, decentralized, and distributed systems, and there appears to be growing interest in how these configurations support or hinder adaptive and transformative capacities towards resilience. There does not appear to be a concerted effort to align how these concepts are used, and what different configurations mean for infrastructure systems. This is problematic because how infrastructure are structured and governed directly affects their capabilities to respond to increasing complexity. We review framings of centralization, decentralization, and distributed (referred to collectively as de/centralization) across infrastructure sectors, revealing incommensurate usage leading to polysemous framings. De/centralized networks are often characterized by proximity to resources, capacity of distribution, volume of product, and number of connections. De/centralization of governance within infrastructure sectors is characterized by the number of actors who hold decision-making power. Notably, governance structures are often overlooked in infrastructure de/centralization literature. Next, we describe how de/centralization concepts are applied to emerging resilient infrastructure theory, identifying conditions under which they support resilience principles. While centralized systems are dominant in practice and decentralized systems are promoted in resilience literature, all three configurations—centralized, decentralized, and distributed—were found to align with resilience capacities in various contexts of stability and instability. Going forward, we recommend a multi-dimensional framing of de/centralization through a network-governance perspective where capabilities to shift between stability and instability are paramount and information is a critical mediator.