Measuring tax avoidance: A stochastic frontier perspective on affiliated firms

Corporate tax avoidance poses a critical challenge to fiscal sustainability and corporate accountability. However, accurately identifying this behavior remains difficult because traditional proxies, such as effective tax rates (ETR) and book–tax differences (BTD), often conflate structural firm characteristics with intentional tax planning. This study addresses these limitations by introducing a novel measure of tax avoidance based on Stochastic Frontier Analysis (SFA), which conceptualizes avoidance as tax inefficiency relative to a "best practice" frontier. Using rich firm-level data from Slovakia (2014–2023), we examine how ownership structure and governance shape this latent behavior. Our findings reveal that firm affiliation is generally associated with lower tax avoidance, consistent with the notion that affiliated firms are subject to greater monitoring and face higher reputational costs. This disciplining effect is further amplified among affiliates with foreign ownership. In contrast, purely domestic affiliates exhibit higher levels of avoidance, though this result appears sensitive to model specification and does not hold after applying the control-function correction for endogeneity. Managerial ability emerges as a crucial factor: technically efficient firms systematically avoid taxes less, suggesting that superior management fosters compliance rather than opportunistic optimization. Furthermore, female leadership is associated with higher avoidance, whereas alignment between ownership and management reduces it. These results demonstrate that ownership and managerial efficiency are primary drivers of tax behavior, suggesting a need for enforcement strategies tailored to specific ownership profiles.