Using Total Variation Distance to Study Technological Transitions : The Case of Renewable Energy Generation

This paper proposes a methodology to detect Technological Transitions (TTs) using Total Variation Distance (TVD), which embodies the relationship between information theory and statistics. We use a database of Renewable Energy Generation (REG) patents to exemplify the usefulness of TVD to uncover moments where a ‘big change’ in REG technology happened. To do this, we compare the observed frequency distribution of patents filed between 1973 and 2015 in the US, spread across seven REG subgroups (e.g., wind and tidal). We identify two crucial TTs: one at the beginning of the 1980s and another in the late 1990s and early 2000s. The first one appears a few years after the end of the 1970s’ oil crisis and the second one after signing the Kyoto Protocol in 1997. In this manner, we reconcile qualitative evidence that registers major REG changes with a quantitative measure that reflects them