Taking Stock
This chapter introduced the basic models of directed technological change. These approaches differ from the endogenous technological change models of the previous two chapters because they not only determine the rate of aggregate technological change, but also endogenize the direction and bias of technological change.
The bias of technological change is potentially important for the distributional consequences of the introduction of new technologies (who will be the losers and who will be the winners from technological progress?). Therefore, the bias of technological change will play an important role in the study of political economy of growth.Equally important, models of directed technological change enable us to investigate a range of new questions. These include the sources of skill-biased technological change over the past 100 years, the causes of acceleration in skill-biased technological change during more recent decades, the causes of unskilled-biased technological developments during the 19th century, the impact of international trade on the direction of technological change, the relationship between labor market institutions and the types of technologies that are developed and adopted, and last but not least, an investigation of why technological change in neoclassical-type models may be largely labor-augmenting.
A relatively simple class of directed technological change models can shed light on all of these questions. These models are quite tractable and allow closed-form solutions for equilibrium relative technologies and long-run growth rates. Their implications for the empirical questions mentioned above stem from two important, and perhaps at first surprising, results, which we can refer to as weak equilibrium bias and strong equilibrium bias results. The first states that under fairly weak assumptions an increase in the relative supply of a factor always induces endogenous changes in technology that are relatively biased towards that factor.
Consequently, any increase in the ratio of skilled to unskilled workers or in the capital-labor ratio will have major implications for the relative productivity of these factors. The more surprising result is the strong equilibrium bias one, which states that contrary to 594basic producer theory, (relative) demand curves can slope up. In particular, if the elasticity of substitution between factors is sufficiently high, a greater relative supply of a factor causes sufficiently strong induced technological change to make the resulting relative price of the more abundant factor increase. In other words, the long run (endogenous-technology) relative demand curve becomes upward-sloping. The possibility that relative demand curves may be upward-sloping not only has a range of important empirical implications, but also illustrates the strength of endogenous technological change models, since such a result is not possible in the basic producer theory with exogenous technology.
The chapter has concluded with a number of applications of these ideas to a range of empirically important areas. Models of directed technological change are very much in their infancy and there are many theoretical dimensions in which further developments are possible. Perhaps more importantly, there are also numerous applications of these ideas.
Finally, this chapter has also been an important step in the investigation of the causes of cross-country income differences and sources of modern economic growth. Its main lesson is in clarifying the determinants of the nature of technological progress. Technology should not be thought of as a blackbox, but the outcome of decisions by firms, individuals and other agents in the economy. This implies that profit incentives will play a major role in both the aggregate rate of technological progress and also in the biases of the technologies that are being developed and adopted. Models of directed technological change illustrate this reasoning in a sharp way and show a range of its implications.
15.10.