Dynamic optimal distinctive configurations enabling manufacturing firms to achieve high green innovation efficiency

When undertaking greening innovation activities, firms often encounter the paradoxical challenge of meeting regulatory compliance requirements while simultaneously pursuing competitive differentiation. Drawing on optimal distinctiveness theory, this study develops an antecedent framework that integrates rule-based drivers (regulatory, normative, and mimetic pressures) with choice-based drivers (executives' green awareness, green technological capability, and management capability) to explain firms' green innovation activities. Using dynamic Qualitative Comparative Analysis (QCA) on panel data from 82 Chinese manufacturing firms between 2017 and 2022, the study identifies multiple configurations leading to high green innovation efficiency and uncovers their temporal evolution. The findings indicate three effective strategic patterns: a market-oriented strategy, in which firms use mimetic pressure as a benchmark to manage uncertainty while pursuing differentiated green practices; a cognition-driven strategy, in which executives' environmental awareness guides sustained innovation in response to long-term transformational pressures; and a capability-enabled strategy, in which regulatory pressure is leveraged as a signal for technological upgrading to build unique green innovation trajectories. By demonstrating how rule-based and choice-based logics interact as complementary mechanisms, this study further extends optimal distinctiveness theory by elucidating the dynamic processes through which firms balance conformity and differentiation in the context of green innovation. • A “pull-push” framework links rule-based drivers and choice-based enablers to green innovation • A dynamic QCA reveals evolving configurations that sustain high green innovation efficiency • Three distinct strategies to high green innovation efficiency: market-oriented, cognition-driven, and capability-enabled approaches • The findings suggest potential boundary conditions regarding the influence of benchmarks