I’ve always cared a lot about “development”. Even before I knew the word I was always concerned about really poor people over there, even when I was really young and thought the world was divided into Britain, Europe, Centre Parcs, America and Ethiopia.
China has very bad economic, human rights & environmental policies. But can we talk about the first derivative? http://t.co/oQ41F24GBU
— Left Outside (@leftoutside) November 18, 2013
I linked to this Scott Sumner piece on China which picked up something I find really intuitive but which maybe other people don’t get. You can be a really, really poorly run country and still see tremendous economic growth. The majority of modern world history is the history of this fact actually. 
Factories are really, really productive. I think a lot of people now don’t interact with them, but modern factories are really something. Manufacturing itself is rather unique in that even poorly run countries will see their manufacturing sector’s productivity catch up with the developed world. This unconditional convergence has been one of the major drivers of economic growth and convergence for countries like Germany, China, Korea etc.
This doesn’t happen in most sectors though. There really is something unique about the industrial working class and a manufacturing base. However, most of your economy is not manufacturing. Most of any economy will be services and a bit of construction and a little agriculture. This is the problem.
So there is an ideal type out there at any moment: a technological frontier of efficient production and rational organisation. This will oscillate wildly but predictably outwards as we work out new ways to do good things. This is, hypothetically speaking, the toolkit the world has to become wealthier and drag itself out of poverty.
Even with minimally functional institutions and rampant corruption, so long as things are kept broadly predictable manufacturers can upgrade their equipment, staff will refine processes and they can easily ship and trade their product. Manufacturing catches up to the technological productivity frontier, possibly because it happens in one place and you can trade globally.
This doesn’t happen in other sectors and I’m not, and I don’t think most people are, entirely sure why. The main concern of development is to get manufacturing clusters as well established in the developing world as possible and to try to work out how to get the other sectors to pull their weight. 
Beyond a certain level of stability, manufacturing is uniquely able to chase down this moving target. But this isn’t generally true and this is where I see institutions coming in. Institutions set how closely your non-manufacturing sectors can approach this technological frontier. They also set how large your manufacturing base will become relative to other sectors.
Many developing countries have subsidised their manufacturing bases. In days gone by this was because manufacturing was ideologically fashionable and its capitalists relatively strong. Coincidentally they hit upon unconditional convergence and this rent seeking and ideological patchwork worked out.
These days manufacturing’s days of dominance are over. Manufacturing employment is shrinking everywhere because it is just so damn productive. The technological frontier will carry on working its way out as we work out more and more productive ways to put resources to work but the days of catch up might be largely over. It is just a lot harder to get non-manufacturing to catch up with global best practice.
 This is why I’m not so concerned as Tim about the developed world’s regulations stifling growth. I agree they may make it a bit of a bugger but largely the rules of the game are predictable and the returns are possible. Growth won’t stop but it might proceed from a slightly lower base and this may be worth it or it may be not.
Incidentally, this is why I’m a particularly poor leftist when it comes to the West’s declining manufacturing base and attendant job losses. I wave happily goodbye as they sail across the ocean because I’m buggered if I can think of a better way to export prosperity abroad. Taking good institutions to the developing world is really hard, and the developing world’s population are against moving the poor here where we already have good institutions.