This project contains code for the analysis of the innovation cycle in the US, conducted in collaboration with Professor Ethan Ilzetzki. The dataset includes text patent data with over 9 million patent document observations. This repository includes some parts of the data analysis; however, the modeling and econometrics sections are not included due to confidentiality restrictions.