Nontax Use of Tax Havens: Evidence From Captive Insurance

Abstract

Corporate tax avoidance is a recurring focus of policy-makers, the media, activist groups, and researchers. This focus often centers on multinational enterprises' (MNEs) use of tax havens, with a wide body of research utilizing MNEs' tax haven use as evidence of corporate tax avoidance activities. However, the common assumption that MNEs operate in tax havens only for tax avoidance purposes overlooks the role tax havens play as homes for captive insurance entities, which allow firms to secure “self” insurance coverage but do not provide obvious differential federal tax benefits. When we remove the effect of captives on tax haven–based measures, we observe a roughly threefold increase in the magnitude of tax savings specifically associated with haven noncaptive activity. We document that nonfinancial firms' use of captive insurance occurs in approximately 11% of firm-years and spans nearly all Fama–French 49 industries. We construct a haven captive use determinants model, with strong discriminatory power and compelling out-of-sample corroboration tests, that future research can employ to account for firms' use of haven captives. Our findings underscore the importance of separating captive and noncaptive-related haven activities.