Breakthrough Innovations: How to find them

This is a brilliant paper (albeit on the policy side). I am including the PDF and the summary below. I hope you find it useful.

The PDF appears here

Capponi, Giovanna, Arianna Martinelli, and Alessandro Nuvolari. 2022. “Breakthrough Innovations and Where to Find Them.” Research Policy 51 (1): 104376.

[embeddoc url=””%5D

The summary is below:

[embeddoc url=””%5D


  • Breakthrough or radical innovations are generally regarded as rup­ tures along specific technological trajectories, possibly leading to shifts or transformations in the prevailing technological paradigm (Dosi, 1982)
  • They play a crucial role in the “creative destruction” process that characterizes the long-run dynamics of technological evolution (Ahuja and Lampert, 2001)
  • The literature refers to contin­ uous or incremental innovations when the outcome of the innovation process is an improvement of the existing technology (Garcia and Cal­ antone, 2002)
  • In the welter of large patent samples, how can we identify those covering breakthroughs? The increasing appreciation of the high variability in the value of patents has led to the emergence of a lively stream of research investigating whether the in­ formation contained in patent documents would be suitable to construct reliable indicators of patent quality (Harhoff et al, 2003)
  • Several proxy indicators of patent quality have been tested for their ability to capture the heterogeneity in the quality of the underlying inventions and reproduce the skewed distributions associated with it
  • Exploiting the names of the companies and the time references reported in both data­ set, we considered a match to be successful if the innovator’s name was the same and the time difference between the introduction of the innovation in the SPRU Database and the QAI recognition was at most five years
  • We found 144 matching observations, suggesting that 34% of the prizes awarded between 1966 and 1983 were awarded to in­ novations listed in the SPRU Database
  • To implement our method and identify breakthrough innovations, we only focus on the applications filed at the USPTO
  • 12 The construction of the in-sample requires access to the complete list of the patents associated with a firm, we only considered the winning companies that we successfully matched on AMADEUS
  • We examine the effect of each predicted breakthrough patent on its technological domain for the ten years after their filing, as compared to the previous ten
  • We have introduced a method to identify breakthrough innovations on a large scale starting from award-winning innovations
  • The output of our procedure is a dataset comprising 138,467 patents filed over 37 years, of which 17,176 are classified as breakthroughs
  • Acknowledging the outcome of this exploratory search and the eligibility criteria of the QAI, we considered a patent’s DOCDB a possible match if the earliest filing date was at most the first year of financial figures to be submitted

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.