StuW Sonderheft NeSt 2025

StuW Sonderheft NeSt 2025

Abhandlungen – Steuerwissenschaften

S72

Koch / Scheider – The Known Unknown: Tax Avoidance by European Multinationals

ports this to some extent. 8 Specific examples of what effects are captured by these different items can be found in Appendix 3. We are most interested in the effects of intra-group profit dif ferences on the resulting effective tax rates. These effects are most likely captured by the tax reconciliation item DTR that captures all effects related to differences in foreign tax rates. However, the scope of this item depends on the definition of the theoretical tax rate. If TTR is defined with reference to the headquarter ’ s statutory tax rate, as it is the case for large parts of our sample, then DTR reflects both the effects of location choices, i.e. where the multinational invests, and the effects of profitability differences among the affiliates. We therefore rear range, in a second step, the tax reconciliation item DTR to stan dardize the information and split these two influences. To this end, we define STR_HQ as the headquarter ’ s statutory tax rate. TRD is the defined to capture all effects of foreign tax rate dif ferentials, independent of the theoretical tax rate applied in the respective tax rate reconciliation (see equation 1). (1) TRD i ¼ TTR i þ DTR i STRHQ i We then split this variable into the effects of foreign location choices and profit shifting. We calculate TRD_subs , a variable that reflects the effects of tax rate differences due to foreign subsidiaries. In the best case, we would capture the effects of location choice by an asset-weighted average of the statutory tax rates of all affiliates. However, as this information is not en tirely available in databases like ORBIS, this would require ac cess to the firms ’ country-by-country reports. Given the una vailability of this information, we use two different definitions of the average statutory tax rate. For the variable TRD_sub s_unw we use the unweighted average of statutory tax rates of all affiliates (according to equation 2a), whereas we weight each affiliate with its country ’ s GDP for determining TRD_subs_gdp (according to equation 2b). (2a) TRD subs unw i ¼ 1 N i P N i j ¼ 1 ij – STRHQ i (2b) TRD subs gdp i ¼ P N i j ¼ 1 ð ij GDP ij Þ P N i j ¼ 1 GDP ij – STRHQ i with N i being the number of affiliates j of firm i, ij being the statutory tax rate applicable at the location of affiliate j of firm 5 According to IAS 12.81 information may be provided in percentage or nominal terms. If information is provided in nominal terms, we translate them into percentage values by dividing items by pre-tax earnings. 6 The theoretical tax rate can be reported by using the parent company ’ s income tax rate (e.g., British Land Company 2020: „ Taxation on loss on ordinary activities at UK corporation taxation rate of 19% “ ) or by using a weighted tax rate (e.g., Adecco 2020: „ Income taxed at weighted-average tax rate “ ). Companies often do not report on their chosen disclosure op tion in more detail (e.g., Ferrari 2020: „ Theoretical income tax expense “ or Adidas 2020: „ Expected income tax expenses “ ). 7 We have analyzed 50 reports with the largest negative values for TC it for additional explanation in order to better understand the nature of this item. We found additional explanation in 31 of these 50 reports. For 13 of these 31 observations, the item relates to the use of preferential tax re gimes, particularly IP boxes. For the remaining 18 observations, tax cred its seem to be granted on systematic grounds. 8 To do this, we examined the 50 most negative observations of the „ Others “ reported by firms in their annual reports to ensure that remain ing items that we assign to the Miscellaneous do not lead to bias. While additional explanation was found only in 22 cases (44 percent), seven out of these cases referred to tax avoidance through benefitting from lower statutory tax rates.

which may be defined either with reference to the statutory tax burden of the headquarter or the average statutory tax burden of all affiliates, to the GAAP ETR. While all tax reconciliations need to contain the theoretical tax rate as a starting point and the resultant ETR as the bottom line, IAS 12 provides no further direct guidance regarding the structure and definition of other reconciliation items. 5 We use a two-step procedure to standardize the tax reconcilia tion data for our purposes. In a first step, we allocate the het erogenous information observed directly from the tax reconci liations to seven self-defined categories. Zero values are as signed if a category is not presented in the respective tax recon ciliation. The starting point of every tax reconciliation is the theoretical tax rate (TTR) . Firms can use either the statutory tax rate of the parent company or an individually weighted average of the statutory tax rates of the parent company and its subsidiaries. 6 Effects of different tax rates (DTR) captures effec tive tax rate changes that are caused by application of tax rates other than the theoretical tax rate. TC captures the effects of tax credits or similar tax benefits. They may constitute a prefer ential tax regime, an investment incentive or may be granted for systematic reasons, for example to avoid double taxation on corporate dividends. 7 TFINDE captures the effect of tax-free in come and non-deductible expenses. We follow the disclosure practice of many firms in our sample and consider these two influences on an aggregate basis. Similar to tax credits, tax-free income may have various reasons. It may be granted for sys tematic reasons, such as to avoid double taxation in the case of corporate dividends or corporate reorganizations. It may con stitute a preferential tax regime, such as the international REIT regimes. Finally, tax-free income may also be the result of de liberate tax planning, such as the use of hybrid mismatch ar rangements. Non-deductible expenses often relate to tax-ex empt income or may indicate inefficient tax management, for example, if they result from the application of thin capitaliza tion or earnings stripping rules. TLDT considers the effects of tax losses or the recognition and valuation of deferred taxes on the effective tax rate. Tax losses do not affect the effective tax rate if they result in an immediate tax refund, as in the case of a loss carryback, or the recognition of a deferred tax asset. Thus, they increase the effective tax rate only if current-year losses can neither be used nor capitalized as a deferred tax as sets or if previously capitalized tax losses are devalued. Con trastingly, a reduction of the effective tax rate arises if pre viously non-capitalized losses can be offset. Under IAS 12, firms can capitalize tax losses as a deferred tax assets if they ex pect sufficient profits in the near future to offset the underlying tax losses or temporary differences. The relevance of this item for the effective tax rate of U.S. firms is emphasized by Drake et al. (2020) who document for their sample of U.S. firms that declining effective tax rates are largely related to the tax treat ment of prior year losses or the use of tax incentive models. Fi nally, OE is a catch-all item that includes all reconciliation items that do not fit into any of the previous categories and is therefore heterogeneous by definition. It includes items that are classified as „ other “ or „ miscellaneous “ by the reporting com pany itself. This makes it possible that firms may use this item to hide the effects of aggressive tax planning to some extent. A look at the more detailed explanations given by firms in the case of exceptionally large negative values for these items sup