Chang's eighteenth is the regulation can sometimes be good for business, even if business doesn't think so, therefore we should not rail against the regulation of business. This is certainly true to some extent: regulations against trading while insolvent are most certainly helpful to clearing out the deadwood of zombie enterprises from the economy, to the benefit of all other businesses. But that isn't of course, quite what Chang means. His meaning rather is that a benevolent government will stop business doing things that might be extremely worthwhile in the short term but will have bad long term effects. An example of his is that companies in poor countries should be barred from importing old technologies. As this could lock them into a sub-optimal development path, perhaps they should be forced into importing more expensive, more up to date, technology even at that short term cost.
This is an argument to which there are a number of answers. The first and most obvious being, well, what evidence do we have that any random group of bureaucrats or politicians are capable of recognising a technology, let alone an appropriate one? This isn't a skill that is notable in the currently advanced economies after all: vide any and all governmental computing projects. We'll not dwell too long on the way in which in currently poor countries what is an appropriate technology, or supplier of, tends to be influenced by how fine the new car of the licensing bureaucrat's mistress is either. Nor who paid for it. And a third objection would of course be what the hell's it got to do with the government what technology a private company installs?
But this is Chang's point of course: that it is indeed the government's business whether I use (in my current project) wet or dry gravitational separation and that there is indeed a bureaucrat or politician somewhere who knows the answer. I think not: and no doubt I am near blasphemous given that my basic project is to recreate a process that the Socialist Government of this country I'm in closed down in 1952.
What Chang's insistence is missing though is that we've only actually got one process that reveals to us whether a technology is appropriate or not: that's its interaction with the rest of the world through the free market. As before, we are uncertain about the future. We simply do not know what future conditions are going to be. The only method we have of dealing with that uncertainty is to try lots of things and see what does survive in the conditions that happen. Which really does mean that we cannot centrally plan what we do, detail which technologies who should use from the Ministeriums, simply because that will not allow enough variety.
There's a possibility of redemption in this comment of Chang's:
The story of GM teaches us some salutary lessons about the potential conflicts between corporate and national interests - what is good for a company, how important it may be, may not be good for the country.
This is most certainly true: but Chang then manages to get it entirely the wrong way around. Leave aside the now ritual complaint that we shouldn't be talking about the country, or the nation, but the economy, which is a very different thing. Chang avers that government planning and co-management of those large firms like GM will help them to survive in the long term. Which is entirely true, large firms can indeed co-manage government to aid in ensuring the survival of them. The way they do this is by manipulating the desire of the politicians and the bureaucrats to regulate. The burden of regulation is much easier for a large firm to carry than the same burden is for the small and snappy competitor chasing at its heels. Regulation thus becomes a wall keeping out that free market competition which is the gale behind capitalism's creative destruction.
And that's really what is wrong with the idea of extensive regulation of industry. That it is actually bad for the economy in and of itself as it protects the corporate dinosaurs who can manipulate it at the expense of new competitors and the consumers.
Another way of putting this: even if all of those who implemented regulation were the omniscient philosopher kings that Chang assumes inhabit our ministries (rather than the bumblers who do), regulation is still bad for the economy as a whole over time. For regulation protects the incumbents. And the important thing we know about this capitalist and market system is that advances come not from incumbents developing but from their being replaced. It is market entrance and exit that drives the system, not the development of the current market players. Thus a system of regulation such as Chang proposes, one that protects the incumbents at the expense of the newcomers is going to be bad for the economy in the longer term. Even, dare I say it, for the nation.