BACKING THE WRONG HORSE: SEQUENTIAL TECHNOLOGY CHOICE UNDER INCREASING RETURNS

Of the recent work in the area of competing technologies and increasing returns, virtually all has been concerned with the effectiveness of the market in delivering optimum outcomes in the absence of any sort of centralized decision making. The fact that the market can lead to inefficient outcomes suggests that there may be a place for intervention in the technology choice process.

This thesis develops a model of sequential technology choice in which the choice process is subject to centralized control. The model is a discrete time, dual control model in which the technologies are subject to increasing returns in the form of learning by using, and one technology is employed each period. A central authority controls the technology adoption process.

The model is used to show that in contrast to a free market, a market with intervention is superior with regard to the goal of maximizing the expected present value of the whole future sequence of adoptions. Furthermore, it will obtain that one of the technologies eventually takes all of the market. However, even with complete control of the decision process, a central authority cannot ensure that an inferior technology does not dominate the market.

It is shown that in the absence of increasing returns, there is no time at which a central authority will know that particular technologies have been eliminated from the market. In contrast, if there are increasing returns to the use of technologies, then there is time after which the central authority knows that only one of the technologies will be used thereafter. Two conjectures are supported by simulation evidence: The stronger are the learning effects, the sooner, in expectation, the market locks in to one of the technologies; and the stronger are the learning effects, the higher the probability that an inferior is locked in.

Finally, a case study is presented. There is mounting evidence that the light water nuclear power reactor is an inferior technology. Nonetheless, the light water reactor has come to dominate the power reactor market throughout the world. The history of this illustrative case is documented.