When Novelty Pays but Doesn’t Persuade: How Evaluative Regimes Shape the Novelty-Value Relationship

Abstract

Technological novelty can create value for firms, yet it can also generate divergent evaluations across audiences. We propose a stakeholder-based framework in which patents acquire economic and scientific value through distinct valuation regimes that resolve novelty-induced uncertainty under different incentives, expertise, and evaluative criteria. Using a two-stage abductive design on 1,946,708 U.S. utility patents granted between 1980 and 2018, we first use machine learning to map the novelty-value relationships and identify opposing, non-linear patterns: novelty is positively associated with market-based patent value and negatively associated with forward patent citations. We then use an independent holdout sample to test the framework’s scope conditions. Technological breadth, domain maturity, and technical complexity systematically moderate these associations and thus specify conditions under which novelty yields valuation divergence versus alignment.