How accurately does the CCCTB apportionment formula allocate profits? An evaluation of the European Commission proposal

Abstract

The European Commission has been setting the course for an EU-wide reform of the traditional corporate tax system of separate accounting (SA) to a unitary taxation with formula apportionment (FA). The underlying claim of this formula is that it allocates profits more precisely than SA to the country where the value is created, as SA can be distorted by intra-group profit shifting. We analyze and evaluate the proposed CCCTB apportionment formula in terms of how well FA approximates the “undistorted” profit. After identifying potential determinants of misapportionment in the CCCTB formula, we test the formula empirically by randomly grouping pairs of European groups to fictitious metagroups. We allocate profits to the metagroup members according to the formula and compare the apportioned FA profit with the observed SA profit (our proxy for “undistorted” profit). We find huge profit misallocations under FA. Restricting the sample to metagroup years with positive profit of each metagroup member, the mean negative (positive) profit deviation is − 4.2% (6.2%) of total assets. Compared to the mean return on assets of the sample, this results in an average error of − 49.5% or + 74.1%, respectively. The deviation caused by FA is systematic. The main drivers for profit misallocations are the profitability of a firm as well as the within-group heterogeneity in firm size and in profitability. The inclusion of loss years shows that loss firms would suffer from considerable additional tax payments under FA. An adjustment of the apportionment weights would slightly increase the performance of the formula. All in all, the CCCTB apportionment formula performs remarkably badly.