This paper investigates why Imperial Japan’s promising Taishō-era democracy (1905–32) experienced a gradual descent into militarism that culminated in the collapse of party government and the effective subordination of civilian authority to the military by the late 1930s. Building on North et al.’s (2009) framework, I argue that this democratic reversal was driven by a failure to consolidate civilian control over the instruments of violence. Facing the existential threat of Russian expansion, civilian elites made a deliberate decision to empower the military at a moment when civilian institutional control over the armed forces was weak and the military technology being adopted concentrated coercive power in a professional officer corps insulated from civilian oversight. This combination produced a substantial probability of eventual military takeover, which was realized over the following three decades as each remaining instrument of civilian authority was tested and broken in turn. This trajectory is contrasted with that of Britain, where civilian institutional control over the military had been consolidated two centuries before industrial warfare arrived. Comparable military technologies were later absorbed there into an existing civilian chain of command and kept under civilian direction. I identify a distinct pathway of democratic backsliding and demonstrate that the strength of civilian institutional control and the character of available military technology at the moment of military empowerment are critical determinants of whether democracy endures.
