If the quintessential test for a state is to have its managers identify perfectly with it and with perfect rationality identify its interests, few states have been universally well-served. The fault is hardly with those managers. The interests of the state are seldom self-evident and nearly always subject to debate. Even in absolute monarchies or dictatorships, the sovereign's advisers offer competing policies.

Imperial Japan legally was such an absolute monarchy, with its emperor granted nearly unlimited power, in theory. More, the leaders of Japan's Meiji Restoration had a near tabula rasa on which to design a new state, one that would benefit from their intense study of the West and its institutions. Keenly aware of the West's threat, these leaders – inspired, dedicated, and intelligent all – deliberately set out to build a rational state capable of dealing with that threat. They failed spectacularly.

In reality, Japan was not an absolute monarchy. It was a virtually headless state from 1868 to 1945. For its first forty-five years, a measure of consensus was provided by the commitment of its founding generation to the avoidance of sharp internal disputes. Japan could ill-afford these, menaced by the West as it was. But the construction of that consensus required the construction of a governing apparatus that, ironically, acted to make consensus impossible once the founding generation passed away. In its place arose a structure of autonomous and highly competitive ministries – bureaucracies – that created professional and powerful allegiances to themselves.