The problems with mis-specifying proxies (part 1)

At the moment, my thesis research is on identifying ways of modelling the consequences of countries' endowments of natural resources. There is a huge literature on this topic, and so it is quite hard to try to include them all. However, at the very core, the claim which echoes through the literature is that of the so-called `Resource Curse'.

The Resource Curse hypothesis is that countries with large endowments of natural resources will grow slower than resource-poor countries. This counter-intuitive proposition uses several channels to explain itself. They can be divided into the Dutch-Disease literature, the Institutional literature, and the Learning-by-Doing literature.

The Dutch Disease literature claims that countries that export natural resources will see an increase in their real exchange rates, either through currency appreciation or inflation, and this squeezes out producers in other sectors. This argument is often extended to include that natural resource exports also increase labour and credit costs, also increasing the cost of doing business for factories and pubs.

The Institutional literature claims that the presence of natural resources changes the incentive structure of countries, and so to the extent that a country allows itself to become over-run by robber-barons (either domestically-bred or foreign), natural resources cause economic stagnation.

The Learning By Doing literature says that in countries with natural resource abundance, a greater proportion will be induced into working in the resource sector. This results in a smaller proportion of the population working in the services and manufacturing sectors---both of which tend to be associated with higher skills, and continual opportunities to Learn by Doing. The long-term consequence of this is that people spend too little time acquiring skills, and too much time digging holes.