The O-Ring Theory of Economic Development
The most important paper ever published?
“The O-Ring Theory of Economic Development” is possibly the most important paper of the last forty years. Certainly no paper has so influenced my thinking as it. I will attempt to explain why it is so meaningful to me, and how it explains the world.
The basic model
Production does not take place all in one step; it can take many steps, a failure in any one of which would wreck the whole process. Different workers have different odds of making a catastrophic error, the odds of which we can refer to as “human capital”. This is true to reality, but leads us to several conclusions.
Perfect Matching
There is an enormous gap in productivity and wages between developed and developing countries. The model explains this. Given a number of steps, changes in human capital lead to exponential changes in the number of units produced. Thus, differences in human capital do not need to correspond 1 to 1 to explain differences between countries – they can be much smaller. Just as people get paid more for doing the same tasks in rich countries than in poor countries, so too do workers in productive firms get paid more for doing the same task as in less productive firms. Under perfect matching, the workers will work with people of the same skill level at their respective jobs, and pay will be higher or lower relative to the same tasks. This could very well lead to unemployment. Suppose a firm is looking for some niche task. If the pay to one worker will affect others, firms won’t hire out to marginal product. We should also expect that, as humanity gets more skilled and production processes get longer, we will have increasing income inequality. (The most skilled workers will get even greater returns.) In fact, it can even lead to wages for unskilled workers in a country falling.
Imperfect Matching
But that’s not all! Assume imperfect matching. Then countries can be stuck in a bad equilibrium of insufficient investment into education, because your returns to education depend on what everyone else is doing. What good is it for you to invest in education, when everyone else isn’t? Your talents will be frittered away. He does not cover this in the paper, but it is one of the most astoundingly powerful arguments for immigration. There need be nothing inherently different among the people of different nations for you to not have convergence, because all you need are bad institutions affecting what people’s beliefs are as to the return of education. Someone can move to the higher equilibrium country, like the United States, and overnight earn many times more than they would have in their home country. Them and their kids will seek the level of education which is optimal to the new, higher productivity country, and thus escape a poverty trap. Indeed, it very likely would help the whole country. If people expect that they are going to be able to move to a much richer country, in which education is valued, then the whole country can be lifted into a higher equilibrium. This explains why the Philippines, for example, has so many nurses. That they trained for foreign work means that they acquired more education, and when they came back their skills encourage others to do more for domestic production. The earlier result for wages for unskilled workers falling, by the way, is conditional upon trade barriers being more porous than the barriers to immigrate. Otherwise, people could move. It is yet another reason why immigration restrains the growth of nations.
It explains how discrimination could be persistent too – if everyone knows they are going to be discriminated against they are disincentivized from seeking education, which inevitably causes people to discriminate when information is scarce. Requiring non-discrimination, or subsidizing education, can improve the state of all people from all groups.
Trade
But wait, there’s more! The O-ring model keeps explaining stuff. Quotas are going to be much more harmful to production than tariffs, because the marginal product of any one task is equal to the whole possible product. The losses from foiled trade are going to be much, much larger than would be implied by the normal Harberger Triangle, which assumes away production. Firms will tend to completely vertically integrate when local product quality or contract enforcement is scarce. If we suppose that production is done in sequence, then the most skilled workers will work at the end of the production process. Thus, the developed world specializes in finished goods, and the developing world focuses on primary inputs. Firms will also be larger in more productive countries, and themselves be more productive.
There is no paper which has more profoundly affected my thinking than this. Read it here in full; it is well worth it. We should be less confident that admitting immigrants from poorer countries will make our institutions worse; it is very plausible that we are making their equilibrium better.