Major shoutout to Noah for making sense of the question: Does cutting government make it more efficient? His thought exercise:
Let’s imagine a slightly more concrete example. Suppose that there are three types of government units:
A) Parasites: These units have figured out how to game the political process in order to make themselves quite secure and mostly untouchable, despite the fact that they mostly just extract rents from society without creating any value. They are entrenched vested interests of the type that Mancur Olson talks about.
B) Clunkers: These units are not politically well-connected, but they are inefficient. Maybe bureaucrats or voters thought that these things would be worth doing, but actually they’re not.
C) Whiz Kids: These units produce more than what the taxpayer pays for them. Perhaps they are courts that stop crime and protect property rights. Perhaps they are infrastructure agencies that build needed highways. Perhaps they are R&D agencies that do basic research that no company will do. Etc. And suppose they do it for a reasonable price tag.
(If you’re either a starry-eyed anarcho-capitalist or someone who likes to spout anti-government rhetoric without thinking about it too hard, maybe you think that category (C) just doesn’t exist at all, and that every single thing the government does would be done more efficiently in the absence of government. But you’re wrong.)
Now suppose that someone like Grover Norquist comes along and decides to start hacking away at government indiscriminately. Which type of unit will get cut? Not the Parasites, certainly. They’re too entrenched. It will be the Clunkers and Whiz Kids that get cut. If there are a lot more Clunkers than Whiz Kids, then this is OK, but if there are more Whiz Kids than Clunkers, then you’re in trouble. And either way, cutting government increases the share of the horrible Parasites, and hence their relative power within government.
This isn’t to say that Noah suggests not cutting government at all (which some commenters seem to miss entirely). Instead, he’s quite clear:
Perhaps instead of thinking about how to indiscriminately cut the size of the government, we should be focusing our efforts on how to tell the difference between efficient government units and inefficient ones, and – most importantly of all – how to excise the nasty Parasites.
Ultimately, Noah is right – there is much space for exploration on this topic still. But his simple framework suggests just how complex such exploration could be.