Title:Leveraging latent persistency in United States patent and trademark applications to gain insight into the evolution of an innovation-driven economy

Abstract:Objective: An understanding of when one or more external factors may influence the evolution of innovation tracking indices (such as US patent and trademark applications (PTA)) is an important aspect of examining economic progress/regress. Using exploratory statistics, the analysis uses a novel tool to leverage the long-range dependency (LRD) intrinsic to PTA to resolve when such factor(s) may have caused significant disruptions in the evolution of the indices, and thus give insight into substantive economic growth dynamics. Approach: This paper explores the use of the Chronological Hurst Exponent (CHE) to explore the LRD using overlapping time windows to quantify long-memory dynamics in the monthly PTA time-series spanning 1977 to 2016. Results/Discussion: The CHE is found to increase in a clear S-curve pattern, achieving persistence (H~1) from non-persistence (H~0.5). For patents, the inflection occurred over a span of 10 years (1980-1990), while it was much sharper (3 years) for trademarks (1977-1980). Conclusions/Originality/Value: This analysis suggests (in part) that the rapid augmentation in R&D expenditure and the introduction of the various patent directed policy acts (e.g., Bayh-Dole, Stevenson-Wydler) are the key impetuses behind persistency, latent in PTA. The post-1990s exogenic factors seem to be simply maintaining the high degree and consistency of the persistency metric. These findings suggest investigators should consider latent persistency when using these data and the CHE may be an important tool to investigate the impact of substantive exogenous variables on growth dynamics.