The implications of book-tax differences: A meta-analysis

Over the last decade, a large body of tax accounting literature on the association between book-tax conformity (BTC)/book-tax differences (BTD) and firms' opportunistic reporting behavior has emerged. Yet, existing empirical evidence on the questions whether increased book-tax conformity actually reduces Earnings Management (EM) and/or Tax Sheltering (TS) and whether book-tax differences are really indicative of such opportunistic reporting behavior is not yet clear. We therefore conduct a meta-analysis aimed at identifying the sources of heterogeneity in primary studies and at providing a consensus estimate with respect to the sign and the statistical significance level for the examined association. Our qualitative literature review reveals that major sources of heterogeneity in the study design include differences in the proxies for EM and TS and in the measures used to determine BTD and BTC. Our meta-regression results show that BTD are indeed indicative of opportunistic reporting behavior, and even more so of EM. These results are, however, weaker for studies that determine BTD only roughly as the difference between book and estimated taxable income instead of using more specific BTD proxies. Moreover, examining actual BTD computed from tax returns instead of only approximating these from financial statements strongly increases the effects. Hence, efforts taken to accurately determine BTD seem to be worth wile when it comes to the explanatory power for opportunistic reporting. Furthermore, our results suggest a negative association between book-tax conformity and EM/TS, which we interpret as an indicator for higher conformity indeed being effective in reducing aggressive reporting.