This thesis aims to extend prior research by examining how using tax havens relates to UK public and private firms’ engagement in AEM, REM, and classification shifting. Several widely used proxies of AEM (Dechow et al., 1995; Dechow and Dichev, 2002; Kothari et al., 2005), REM (Roychowdhury, 2006; Haga et al., 2018), and classification shifting (McVay, 2006; Zalata and Roberts, 2017) are adopted for robustness and comparability with previous studies. The broad findings of the thesis are that overall, UK firms' use of tax havens associates with less AEM but more REM and classification shifting. The association is more pronounced for public firms than private firms in the cases of AEM and classification shifting. The association between tax haven usage and AEM and REM is also affected by ownership structure.
This thesis consists of three empirical chapters: The first empirical chapter (Chapter 2) documents that the use of tax havens is linked with reduced AEM but increased REM since the latter is less regulated and harder to detect. This supports the argument that tax haven usage attracts greater scrutiny. In addition, public firms demonstrate reduced AEM and sales manipulation when utilising tax havens than private firms, which is consistent with public firms being subject to greater scrutiny and market demand for higher-quality accounting information. Key findings in the second empirical chapter (Chapter 3) are that managerial ownership and ownership concentration associate with more AEM and sales manipulation in tax haven firms relative to non-tax haven firms, whereas ownership concentration associates with less discretionary expenses and production levels manipulation by tax haven firms relative to non-tax haven firms. In addition, institutional investors are more likely to constrain AEM and sales manipulation of tax haven firms than non-tax haven firms. The third empirical chapter (Chapter 4) shows that firms using tax havens engage in more classification shifting relative to those that do not. Results also show that public firms rely more on classification shifting when using tax havens than when they do not. In contrast, private firms’ use of classification shifting appears unaffected by tax haven use. This is consistent with classification shifting being a less costly alternative for public firms compared to AEM and REM. The findings of this study suggest that classification shifting and the value-destroying REM require higher levels of regulation by UK policymakers. The findings also indicate that the high institutional environment quality of the home country (i.e., the UK) may have a more substantial impact on earnings management behaviour than tax havens.