Understanding Governance: Policy Networks, Governance, Reflexivity and Accountability

Rhodes argues that the Westminster model—with its focus on parliamentary sovereignty, strong cabinet government, and ministerial accountability—has become an inadequate description of how British government operates. He rejects the assumption of unified hierarchical control, proposing instead that governance now occurs through self-organizing interorganizational networks that are substantially autonomous from the state. This shift is not a normative preference but an empirical claim about institutional change: power has dispersed upward to the European Union, downward to special-purpose bodies, and sideways to executive agencies.

The book introduced the 'hollowing out of the state' thesis and defined governance as 'self-organizing, interorganizational networks' characterized by interdependence, resource exchange, negotiated rules, and significant autonomy from the state. It also proposed the 'differentiated polity' as an alternative narrative to the Westminster model. These concepts became foundational to what is now called the Anglo-Governance School, influencing public administration research across Europe, and remain central to debates about state capacity and network governance.

Both reject traditional bureaucratic hierarchy, but they diverge in focus. NPM emphasizes market mechanisms—competition, contracting, performance measurement—as tools for making government more efficient and responsive to consumers. Rhodes's governance approach emphasizes the empirical emergence of networks as a distinct coordinating mechanism alongside markets and hierarchies. Where NPM prescribes managerial reforms, Rhodes describes a structural transformation in which the state has lost capacity to steer hierarchically. The hollowing out thesis is about reduced state capacity, not about better management of public services.

Rhodes himself revisited the argument twenty years later, concluding that the trend intensified rather than reversed. The growth of contracting-out and what he calls the 'franchise state'—where large corporations deliver public services under government contract—extended hollowing out into new territory. However, Rhodes acknowledged a limitation: his original account understated how differing party beliefs shape outcomes. New Labour accepted parts of the neoliberal agenda while retaining a more active central state, whereas Conservative-led governments accelerated outsourcing and austerity-driven retrenchment.

No—the distinction is central to the book's argument. Government refers to formal state institutions and hierarchical authority. Governance, as Rhodes defines it, refers specifically to self-organizing networks that coordinate activity through negotiation, trust, and resource exchange rather than command. These networks can include state actors, but they are not controlled by the state in the way government implies. This is a common misreading: treating 'governance' as a fashionable substitute for 'government' misses the specific claim that networks operate with substantial autonomy.

Rhodes argues that traditional British political science relies on positivist assumptions and an institutional focus that cannot capture the complexity of network governance. He calls for a reflexive approach that interrogates how the researcher's own assumptions shape analysis. This involves attending to the 'postmodern challenge'—the recognition that governance is constituted through diverse beliefs and practices rather than a single objective structure. This methodological strand developed further in his later work with Mark Bevir on decentred theory and interpretive political science.