Deep in the comments of this characteristically puzzling Elizabeth Anderson post, David Velleman strikes back at those who deny that the state is productive:
This is the fallacy that lies behind so much of the distrust of government in which Mr. Ridgely revels (irresponsibly, in my opinion). The fallacy is to think that because government doesn't produce anything, it doesn't contribute to economic productivity. Go look at the current economic productivity of Iraq, and ask yourself what's most needed before it can be improved. What's needed is government.
How productive would our country be without secure borders, without civil order, without a properly regulated currency and banking system, without a judicial system to enforce contracts and adjudicate disputes, without an interstate highway system, air traffic control, well functioning ports and waterways, orderly division of the electromagnetic spectrum, protection for intellectual property rights, guarantees of the safety of buildings, food, and drugs, a well-educated and healthy work-force … must I go on? All of these government functions make productivity possible. Without them, we would all be less productive and poorer.
I agree that the state can contribute to economic productivity by setting in place a system of sound basic institutions that help create the conditions for efficient market cooperation. But Velleman seems to dip his toes in the opposite fallacy. Markets are limited and feeble without effective political and legal institutions. Therefore assume effective political and legal institutions! And behold their glory! Government makes markets go ZOOM!
But but but wait. Market cooperation is limited in the absence of effective government because of, to make a very complicated story very uncomplicated, problems of assurance and trust. But problems of assurance and trust are exactly what prevents government from being effective, too. In order to fund a government sufficient to do anything worth having a government do, there has to be sufficient wealth, and thus markets that have already solved, to some extent at least, the problems that you need government to help markets solve.
The more effective, efficient, trustworthy, non-corrupt, credible, etc. a government is, the less you will need government. Whatever norms a society has deployed to partially solve the principle/agent & state autonomy & nepotism & predation problems that afflict political institutions everywhere and always, those are norms that will have also been deployed to solve problems of market cooperation.
So here's my little superficially paradoxical rule of thumb, which I hereby dub Wilkinson's Law, until I am told that it has already been dubbed: The more your markets need government, the less your government will be able to do for your markets. Or, equivalently, the more your government is able to do for your markets, the less it will need to do. Pithier still . . . Government: if you need it, it won't be good, and if it's good, you don't need it.
Ceterus paribus, naturally.
This leads me to believe that Velleman's list of the wonders of good government has too much stuff on it. Indeed, his entire list after “adjudicate disputes” is pretty questionable. If you've already got the kinds of markets that give rise to issues about interstate highways, air traffic control, spectrum allocation, food and drug safety, and whatnot, then you've likely already developed the wherewithal to solve market coordination problems using largely market means. And if the market can do it, then the government isn't likely to do it any better. And the government doesn't contribute to productivity if it's crowding out something that could be do the same thing better by other means. And when the government does it, its coercive. And coercion is bad, on the face of it, and raises questions of moral legitimacy that do not arise under institutions of voluntary coordination. Because voluntary cooperation is good, which is why markets are.