Institutions fail in predictable ways. The predictability is the most important thing to understand about institutional failure — because predictable failures are preventable ones.
The Failure Taxonomy
Institutional failures cluster into five categories that are distinguishable analytically and that each require different responses. Capability failures: the institution lacks the resources, the skills, or the infrastructure to perform its function at the required level. Incentive failures: the institution's internal incentive structure rewards behaviour that conflicts with the institution's purpose. Accountability failures: the institution is not subject to adequate external accountability for the quality of its performance. Design failures: the institution's governance structure, its authority boundaries, or its decision-making processes are not aligned with the functions it is supposed to perform. And capture failures: the institution has been captured by the specific interests — internal or external — that the institution is supposed to regulate or serve, and those interests have reshaped the institution's behaviour toward their benefit.
These failure categories are not mutually exclusive — most significant institutional failures involve multiple categories simultaneously. The police department that produces racially disparate outcomes suffers from incentive failures (the accountability structures that reward activity metrics over community safety outcomes), accountability failures (the legal frameworks that limit external oversight), design failures (the governance structure that insulates the department from community accountability), and potentially capture failures (the union contracts that capture disciplinary authority). Addressing any single category without the others produces partial improvement rather than the structural change that addresses the underlying failure condition.
Institutions fail in predictable ways because the conditions that produce failure — inadequate capability, misaligned incentives, insufficient accountability, structural design flaws, and capture by narrow interests — are not random. They are the predictable products of institutional governance choices. The institution whose governance choices create these conditions will fail. The institution whose governance choices prevent them will not. The difference is institutional design, not institutional luck.
Discussion