Kai Heimpel I6038659, 29.11.2012. Course: ICT, Organization and Income. Summary of Section 3 and 4 of „Technical Change, Inequality and the Labor Market” by D. Acemoglu (2002) In Sections 3 and 4 of this paper, Acemoglu seeks to explain wage inequality and its development over the course of the second half of the 20th century based on a framework of supply and demand of skilled and un-skilled workers, as measured by the level of their education. Section 3 is concerned with how the skill-biased nature of technical change impacts the skill premium given different values of elasticity of substitution (σ) between high- and low-skill labor. Using on a CES production function, the author demonstrates that this impact can be turned into its opposite by changing σ and points out that for this reason, a technology itself cannot be labeled as generally replacing (un-)skilled labor. He then walks us through a number of situations that involve different levels demand and supply of high skill versus low skill labor (i.e. relative demand and supply, H/L) and different levels of skill bias and σ. These are illustrated in a graph plotting the inelastic short run supply and downward sloping demand in a coordinate system of skill premium (ratio of high and low skill wages) and relative employment (H/L). Although the author considers the case of σ <1 (e.g. perfectly complementary inputs), in which skill-biased technical change creates an overcapacity of skilled productivity and decreases the skill premium, he stresses the fact that “conventional wisdom” (p. 21) assumes the opposite case (i.e. elasticity of substitution greater than one). Regardless of σ, an increase in H/L supply is met with a decrease in the skill premium. With respect to the average wage, Acemoglu shows that any increase in productivity through technical change always increases it. Building on previous research and a short examination of within-industry changes of employment of skilled labor, Acemoglu concludes the existence of skill-biased technical change since WW2. Building on this conclusion, Section 4 investigates the development of skill-biased technical change for the period it is assumed to characterize. It supplies two hypotheses. The Steady-Demand Hypothesis assumes the contribution to productivity of technical change and the shift of the H/L demand curve over time to be constant, leaving changes in the supply of high and low skill labor to explain variability of the skill premium. In support of this hypothesis, the author points out how the skill premium decreased in the 70s and increased in the 80s, consistent with the growth rate H/L supply being high in the first and lower in the latter decade. But overall, there is more significant evidence provided against this hypothesis. Secondly, the Acceleration Hypothesis states that the degree of skill-bias varies over time. Significant evidence is provided for this hypothesis, building on the increase of the computer wage premium, a correlation of the growth rate of skilled labor demand and use of computers, the declining cost of (low skill replacing) equipment capital, discontinuities in the development of labor market prices and, lastly, differences in the increase of the skill premium from 1940-1970 versus 1940-1995 that are, given similar supply growth, professedly inexplicable assuming steady increase of H/L demand. Acemoglu therefore concludes that an increase of high-skilled labor demand in the 1970s or 80s is likely.