Topological Politics: Some Implementation Details
The problem of reaching a new, large-scale political order from an old one was, for longer than people expect, a generally, cross-culturally unsolved problem. Many of the earliest tolerable states were tolerable because they did not, and could not, make substantial changes to a deep underlying order. The king of Israel could not change the law. Germanic kings could not change the law. The Caliph could not change the law. The Pope could not change the law. This meant, in general, that systems, when they adapted, adapted by legitimizing new interpretations of old patterns, rather than by imposing new patterns themselves.
Systems for changing the political order are ancient, but -- for a very long time -- didn't scale. For bands, consensus was often enough; for tribes, direct democracy worked well. Beyond that scale threshold, mechanisms like markets, judges, ritual, and regional delegation expanded the scope of public consent, but -- in general -- all forms of large-scale organization have relied on the use of public force alongside consent.
Force is easy. It requires only the consent of men with guns, and the compliance of people on the wrong end of those guns. With enough of a force disparity, you can get precisely the outcome you want, precisely when you want it. In the lead-up to the use of force, it often seems like a useful tool to cross some unpleasant utility valley which couldn't be bridged by consent alone.
Unfortunately, even force requires consent. The more force you intend to apply, and the greater the risk to your gun-wielding elites, the deeper the consent you need from those elites. You could buy them off -- which requires paying them from spoils drawn from the people you're using force against -- or you could make them fanatically devoted to your ideals, which makes them dangerous when, inevitably, you have to change course. Over long timescales, levels of systemic force tend to degenerate into stationary banditry: because relaxation of force would result in the collapse of the system, politico-military elites tend to increase in importance as the amount of force required to sustain the system increases.
Which leaves mass consent.
Obtaining mass consent is more difficult. Mere democracy won't get you there: you need a family of non-interfering norms and laws which maintain consent even in the face of elites or majorities willing to expend that consent for short-term or personal goals. In particular, you need some mechanism to restrain public force, some mechanism to restrain private force, some mechanism to prevent elites from taking strongly consent-depleting actions, some mechanism for swapping elites without and some mechanism to implement decisions within the system more-or-less deterministically.
More or less, you need posse comitatus, legal procedure, meaningful enforcement of laws and norms against private parties, civil rights, an independent judiciary, democracy, and a bureaucracy. There may or may not better mechanisms than these -- but these are the ones most commonly used.
The structure of mass consent, especially mass consent in extremis, gives political elites a limited control surface: they can either respect the structures which allowed rotation of political elites to begin with, or they can attempt to supplement flagging consent with force or quasi-legitimate, unilateral action. This, again, is perfectly good (in a morally neutral sense) for chasing short-term goals, but both (a) renders the system of consent brittle, and (b) risks the development of counter-norms (and, in the American system, law) opposed to long-term objectives.
Which means, basically, that when people tell me, "WHY AREN'T YOU PULLING THIS POLICY LEVER SO HARD IT SNAPS OFF IN YOUR HAND?", it's not that I disagree that the problem is important. It's that I think the problem is either (a) not important enough to risk breaking a control surface to achieve it, or (b) requires sufficiently long-term action that attempting to act now would foreclose the possibility of acting in the future.