Technological advancement, learning, and the adoption of new technology

This research focuses attention upon three related issues: the persistence of technological progress over time, multiple adoption decisions over a long horizon, and the impact of production based learning on those decisions. We develop a simple economic model for a single product producing firm incorporating continuous technological progress, linear product demand, linear switching costs, and experience based cost reductions. A dynamic programming framework is used to evaluate cases where either one or an unlimited number of adoptions are allowed over an infinite horizon. Both structural and numerical results are presented. Fundamental results serve to explain several counter-intuitive dynamics. In the single adoption case, faster rates of technological progress, as well as growing markets, or a steeply sloping demand curve, serve to delay adoption while an increased ability to learn may accelerate it. When multiple adoptions are allowed, the adoption of any particular technology presents a "window of opportunity" in which future investment will be warranted. If, for whatever reason, this window has passed, then maintaining the older technology becomes more attractive than adoption. Thus, seemingly outdated technologies may remain embedded in some settings.