DOI: 10.1108/jsma-01-2026-0040 ISSN: 1755-425X

Board of directors and innovation efficiency: a necessary condition analysis

Giacinto Coniglio, Nicola Cucari, Salvatore Esposito De Falco

Purpose

This study examines whether corporate governance mechanisms constitute necessary conditions for achieving innovation efficiency in the healthcare industry. Building on the view that boards are not merely control bodies but strategic resources that enable the implementation of complex innovation strategies, the study investigates which governance attributes must be present for efficient innovation to be feasible at all.

Design/methodology/approach

We apply necessary condition analysis (NCA) to assess whether board size, board independence, board gender diversity, board tenure and CEO duality represent necessary conditions for innovation efficiency, defined as the conversion of R&D investments into patented innovation outputs. NCA is suited to this purpose because it identifies minimum governance thresholds required for strategy execution rather than estimating average effects. Ceiling line estimation, permutation tests and bottleneck analysis are employed.

Findings

The results identify board independence as the only robust necessary condition for high innovation efficiency. Firms with low levels of independent directors are unable to reach higher innovation efficiency, regardless of their R&D intensity. Board size, gender diversity and board tenure exhibit weaker necessity patterns that do not withstand permutation testing, while CEO duality shows no necessity relationship. Bottleneck analysis reveals sharply increasing minimum thresholds for board independence as innovation efficiency rises.

Originality/value

This study advances corporate governance and innovation research by introducing necessity logic into the analysis of board mechanisms, reconceptualising innovation efficiency as a governance-relevant outcome and empirically demonstrating that board independence operates as a non-compensatory prerequisite rather than an average-effect driver.

More from our Archive