ABSTRACT
In an influential paper, Hornstein and Krusell (1996) advanced the hypothesis that learning may have been a possible trigger of the slowdown in productivity growth that started in the 1970s in many industrialized countries. 1 Their claim is based on the notion of investment-specific technological change. According to their view, those years witnessed a remarkable increase in capital-embodied technological improvements (think, for example, of the introduction of new and more powerful computers into the production process and of the robotization of assembly lines, or simply of the invention of more efficient means of telecommunication) whose most important consequence was a generalized increase in the level of mechanization of many tasks. Once one of these technological breakthroughs had taken place, their argument goes, in order for the same task to be performed properly, a given learning period was generally required during which labor-productivity could ultimately be lower than before. In other words, a higher rate of embodied technological change, by implying that new technologies with which workers have less experience are introduced at a faster rate, can eventually lead (at least temporarily) to lower output growth. 2
