Is the Rise in Wage Inequality the Price to Pay for Innovation and Growth

This article presents an analysis of the models that have been developed in the literature to account for the rise in wage inequality during the 1980s and the 1990s. These models build upon the assumption of an acceleration in the rate of technological progress associated with the diffusion of the new information and communication technologies. What is at stake in this literature is not to justify the existence of skill-biased technical per se, but rather to explain the main and recent dimensions of such a bias on relative wages. A first dimension lies in the extent of wage inequality, namely the aggravation of inequality since the 1980s in developed economies. The aim is to explain why such a bias has occurred in the 1980s, while technical change has diffused all along the twentieth century, precisely during a period where skilled labour supply increased. A second dimension of wage inequality concerns its fractal nature, that is the fact that it persists within education groups. Furthermore, within groups inequality features a strong transitory dimension, and differs in nature according to the level of education considered. Depending on the determinants - exogenous or endogenous - and the consequences on wage inequality - permanent or transitory - that such models explain through the recent acceleration in the rate of technological change, two categories of models can be distinguished. In the first one, the technological bias is analysed by focusing on the impact of technical change on the returns to human capital. Human capital embeds for example individual ability, the ability to absorb and use new technologies, assets, luck, specific technological competencies or general education. The second category of models highlight the mechanisms likely to explain the interaction between technological progress, supply and demand for skilled workers. These approaches however explain only a restrictive part of the evolution of wage inequality. In the long run on the one hand, and when considering dimensions such as work organization on the other hand, other categories of models are necessary.