Acemoglu& Zilibotti, 1997, Was Prometheus unbound by chance? Risk,Diversification, and Growth

[Jan 26th; model out of capacity, theories okay]

It basically ponders on the variability of growth, in poor countries and in the early stage of development, say the 10th-19th century. The framework used here to explain the variability is the relationship among market completeness, diversification, risks, and capital accumulation. In early stages of capital accumulation, chance plays a vital role, if not the only deterministic role. Because the amount of capital is limited in the beginning, agents can only invest in certain projects bearing low risks, or demanding low start-up costs; this fact endogenously incurs lower diversification and higher risks of investment. If chances are good, the returns are accidentally high, then largecapital can be accumulated and get injected into more projects, thus raising the degree of diversification and decreasing risks. So these fortunate countries can earn higher productivity, while those unfortunate countries willstay in the vicious cycle.

The paper also puts forward another point,that decentralized competition can't approach Pareto optimal even though all agents are price takers, and no technological spillovers exist. The essential problem embedded here is the pecuniary externality of diversification: the number of projects invested by one agent will affect the total number of market projects and the degree of diversification, therefore the return rate areinfluenced.

As for the direction of capital flow, two converse theories exist. One theory supports that capital flows to poor countries, based on the law of decreasing return to capital. The other oneindicates that capital prefers rich countries where diversity is better ensuredand risks are lower.