Skill intensive technologies seem to be adopted more often by rich countries rather than poor ones. Related to that observation, the ratio of wages of skilled to unskilled workers - the skill premium - shows two important features over time and across countries. In the US the skill premium decreased during the first half of the 20th century and increased after 1950, evolving in a U shaped pattern. On the other hand, the same measure across countries around 1990 is hump shaped when countries are ordered by GDP per worker. By modeling the decisions for factor accumulation and technology adoption, this paper gives a systematic explanation as to why we see ever more skill intensive technologies being adopted both over time in the US and across countries. The model developed here endogenously generates predictions for the skill premium that are consistent with both US and international observations under the same set of parameter values.
