Policy and Bureaucracy

I had a minor hit yesterday on Twitter with the observation that difficult political questions are usually not about what government action would be “best” (meaning most just, or most efficient), but rather about what policy can actually be implemented in the only way governments can implement policy, by making them directions to a bureaucracy.
The specific example that brought this point to mind was the suggestion from @AvengingRedHand that illegally obtained evidence should be used in court, but the individuals who illegally obtained it should also be punished for their offence.
That is perfectly just. It is also, as far as I know, the law here in Britain. (If you want legal advice, though, ask a lawyer). Not that I actually recall anyone ever being prosecuted for obtaining evidence illegally.
And that’s the point. The US rule that makes the product of illegal searches inadmissible is not directly beneficial to justice, but it has the intended effect that somebody in the whole process has an actual reason to draw attention to illegal searches — specifically, the defense. The suggested rule may be better, if perfectly implemented, but it just isn’t going to be perfectly implemented.
And that’s it, really. Whatever your goals are in politics — justice, efficiency, kittens and rainbows — it’s no good just working out what everybody needs to do to achieve them. It’s not even enough to think also about how you’re going to get and retain power to make people do what they need to do, though that is also necessary. At the same time, you have to understand that governments, like all other large organisations, works through bureaucracy, and what a government actually does is not “enforce laws”, or “redistribute wealth”, but issue commands to a bureaucracy which will then respond to those commands. And it won’t normally respond to them by obeying them totally in letter and in spirit.
We all fall into this error from time to time, but I consider it the fundamental fallacy of progressivism. Take Rawls’ “Theory of Justice” for example. As an anti-progressive reactionary with libertarian tendencies, what do I think of his reasoning? Actually, it’s not bad. I could maybe quibble a bit, but I won’t bother, because it’s a reasonably sensible answer to an absurd question — the question “what should society do to maximise justice?” Whatever the theory, society will do what it damn well wants. The question of politics is, “what instructions should be issued to a bureaucracy to achieve some kind of acceptable standard of life for society?”
Many critiques of progressivism attack the fallacy, but then fail to notice that it applies also to the critique. Libertarianism correctly observes that socialism doesn’t provide the incentives for individuals to act in a cooperative manner, and identifies market forces as a way of providing that incentive, albeit imperfectly. However, it fails on the grounds that the night-watchman or nonexistent state does not provide the incentives for groups to refrain from political activity. They just say, “groups should not gain advantage at the expense of other groups by political activity”. That would be nice, but who’s going to stop them? They’re left talking about a “new man”, just like the utopian socialists.
To borrow a metaphor frequently employed in biology, politics is not about identifying the perfect form of a cake, it is about finding the best recipe for making a cake, that can be made with the tools and ingredients available.