Ungovernable Galaxy: A reflection of the truth, long before you get to the size of an entire galaxy, at least when you’re talking about centralizing-hierarchist structures. As they scale up, they start bottlenecking horribly – there’s a reason why most growth patterns matching this structure stall out well before they get to 100 systems. Hell, a large subset of them crash and burn before they reach one planet.
(The exception that proves the rule is the Voniensa Republic, which claims 8,000 systems – but then, not all of those 8,000 are technically “its”, and the Shell is different from the Core, and so forth. That said, they are perhaps the acknowledged masters of making centralizing-hierarchist structures work on this scale, inefficient and kludgy though they are; just because they insist on being primitives doesn’t mean they’re stupid.)
As for the Empire? It was pursuing alternative approaches long before it hit the one-planet level. If you look over here, you’ll see this:
Peter B. Evans uses Williamson’s control loss model to show that higher efficiencies are possible when the Emperor switches to “multiple hierarchy” systems, such as the dual hierarchy. If the Emperor creates a complete second command hierarchy in parallel with the first, his effectiveness rises by nearly two-thirds. The superiority of dual hierarchies is well-known in business (line-and-staff) and in public administration (especially Communist bureaucracies). Lattice structure systems are a more sophisticated form, involving a complete lattice of hierarchial links providing a startling multiplicity of pathways to the top. Such novel system my not encourage galactic stability, but the opportunities for palace intrigue are legion!
Now, what’s the limiting case of a lattice-structure system?
That’s the strategy the Empire is pursuing – a radically decentralized system, with tremendous local autonomy handed out at each level, based around Symbol, Meme, and Mesh.
You’ve got a nice spectacular center – the Imperial Couple, the Senate, the Curia – who do serve a key function as deciders-of-last-resort, but who work very hard to avoid decisions having to reach their level, and whose main function, along with the trappings of office and capital, is to be the Symbol, the gravity well around which all else orbits.
You have the Meme, the idea of empire, the dream that is Rome, the ideology that guides policy. Which works much better as a control mechanism because it doesn’t need a center. Memes replicate. It’s what they do. There is a minor centralizing element inasmuch as the Meme must be tended, mutations pruned, and so forth, but that itself can be distributed.
And you have the Mesh. Not a single, massive, centralized hierarchy, but a whole team of organizations flying in close formation, orbiting the same point but not directly controlled by it, with each one – like flocking birds – correcting and corrected by the others near it. Exchanging information – flowing in to the center, back out to the edge, and around peripheral routes. Local nodes of distributed AI systems make decisions based on local knowledge but following shared ideas, creating global coordination without need for centralization. Everything is disseminated everywhere. Everything checks everything else.
Will it scale to an entire galaxy?
This, of course, also leaves out the point that it’s easier to govern when you aren’t trying to govern every little thing. Or, to put it another way, they’re not these guys: