Unintended effects of anti-tax avoidance laws on multinational firms: Evidence from controlled foreign company rules

Peter Egger, Georg Wamser

07 October 2015



Controlled foreign company rules

Action 3 of the Action Plan on Base Erosion and Profit Shifting, published in 2013 by the OECD (see OECD 2013), recommends strengthening of controlled foreign company (CFC) rules to limit tax base erosion and profit-shifting by multinational firms (Action 3: Strengthen CFC rules). Many countries have implemented CFC rules as part of their anti-tax avoidance legislation with the purpose of taxing the foreign income of multinational enterprises (MNEs) associated with profit-shifting. Income of foreign subsidiaries of multinational enterprises is otherwise exempt by most countries from home-country taxation if it arises from active (production) activities. The purpose of CFC rules is to withdraw this exemption privilege if the income is associated with passive income – i.e. it is not generated through the main activity of firms, such as income from rents – in low-tax or tax-haven countries and the shifting of profits among affiliated firms from high-tax to low-tax countries.1 Clearly, the aim of such CFC rules is to incentivise firms to not engage in cross-border profit-shifting but to generate investment with the profits in their main activity. The general presumption is that passive income and investment exist mainly for the purpose of avoiding taxes and that reducing them will spur active investment and additional tax revenues in high-tax jurisdictions. However, in reality it is not necessarily the case that passive investments are substitutes to active investments and that limiting passive income will raise tax revenues in high-tax countries.

Consequences of controlled foreign company rules: New evidence

In new research (Egger and Wamser 2015), we aim to assess the consequences of CFC legislation at the firm level, using census data on German multinational companies and their subsidiaries worldwide. Earlier research suggests that the implementation of such rules indeed reduces passive income and investments. However, our focus is on unintended effects of CFC legislation, and therefore we assess the role of the legislation on firms’ real, active investment activities abroad.

The German CFC legislation is a marvellous example of German law-making. It contains a host of provisions and is tremendously complicated (leading prominent tax lawyers to conclude that it is never properly applied). In particular, the provisions suggest that a multinational firm falls under the law if a number of thresholds are exceeded. In particular, passive income needs to be absolutely high (in euros), it needs to be relatively high (as a ratio of total income), and for it to count, it must be generated in high-tax countries (relative to Germany). Some of these provisions apply for the firm as a whole (i.e. they aggregate over relevant passive income everywhere in the firm), while others address each subsidiary separately. This generates a tax design – determining when foreign tax income is exempted from German taxation – which involves numerous discontinuities associated with those thresholds. These discontinuities may be used to quantify the treatment effect of German CFC legislation on real investments across subsidiaries when comparing ones which are only just not treated (are on the low-tax side of the multidimensional CFC rule but in the neighbourhood of the threshold) with ones that are only just treated (are on the high-tax side of the multidimensional CFC rule but in the neighbourhood of the threshold).

CFC rule treatment significantly changes the taxation of all profits of a foreign subsidiary due to a sharp jump in the general cost of capital. Therefore, such rule treatment may lead to substantial adjustments in general investment behaviour, not only investments in passive assets.


Our results suggest a significant and economically large impact of the German CFC legislation on real activities abroad of multinational enterprises.

  • The benchmark estimates imply that fixed assets at foreign subsidiaries decline, on average, by about €7 million per subsidiary in response to CFC treatment in the neighbourhood of the two most important treatment thresholds regarding the frequency of firms.

Most German-held subsidiaries appear to obtain CFC treatment status due to a high share of passive income in foreign low-tax countries. Figure 1 provides a graphical illustration of the treatment effect in those two dimensions of the law. The vertical axis measures the fixed assets of German foreign subsidiaries. The two horizontal axes measure the frequency of the two most important aspects of the controlled-company rule application. The first is the size of the corporate tax rate abroad minus the German one (only if this is negative does a subsidiary qualify for treatment). The second is the share of passive income measured as passive assets relative to passive plus active assets by subsidiary minus the threshold value of that variable (only if this is positive does a subsidiary qualify for treatment). The figure reports actual subsidiary-level data in the space of these two critical measures, where red and blue dots indicate the actual level of real, active investments of CFC rule treated and untreated subsidiaries on a vertical axis associated with different configurations of the rule measures on the horizontal axes. The figure also displays two smoothed surfaces through the red (treated) and blue (untreated) points, respectively. The figure indicates that subsidiaries whose relative passive assets are marginally too high in countries where the tax rate is marginally too low for non-treatment undertake substantially smaller real active investments (their cost of capital is higher due to treatment) than those that are marginally not treated. This is the case at the edge of the red surface in the back of the figure, where the blue surface is obviously discontinuously higher (lower cost of capital leading to higher real investments). The CFC rule treatment effect of €7 million corresponds to the vertical distance between the blue and red surfaces in the aforementioned corner. 

Figure 1. Fixed assets for treated and untreated foreign subsidiaries

Numerous tests in our paper demonstrate that these findings are not confounded, for instance, by treated subsidiaries being specialised in other investments (e.g. financial investments), so that their lower fixed assets would be due to other aspects than the controlled-company law. These tests suggest the identified impact of the German CFC rule appears to be causal, since it affects subsidiaries only in their treatment status (and the associated outcomes affected by it) but nothing else.

Figure 1 also suggests some heterogeneity in the estimated rule-treatment effects. For example, the (negative) treatment effect increases in the tax-rate differential between Germany and the foreign (low-tax) country as the tax penalty of removing the exemption privilege becomes larger. Indeed, the vertical distance between the surfaces becomes larger when moving along the line where the tax rate becomes more negative and the passive income ratios are fixed close to the threshold value.

Policy implications

Controlled foreign company rules are implemented by countries to prevent adverse tax-revenue effects associated with the profit-shifting activities of multinationals. However, our estimates suggest unintended consequences of such rules for real investment activity.

One interpretation of these findings is that the German CFC legislation renders Germany’s system of tax exemption (of active investment) obsolete and introduces features of a residence-based system of taxation. Tax policymakers often argue that countries should apply tax-exemption systems to guarantee a level playing field in host countries, irrespective of the taxes of parent countries. Accordingly, recommendations on the design of CFC rules should keep an eye on the harmonisation of its legislation provisions across countries in order to not undo the benefits of tax-exemption systems to a large extent.  


Egger, P H and G Wamser (2015), “The impact of controlled foreign company legislation on real investments abroad. A multi-dimensional regression discontinuity design”, Journal of Public Economics 129 (2015) 77–91.

OECD (2013), “Action Plan on Base Erosion and Profit Shifting”, OECD, Paris.


1 Even if the home country runs a tax credit system, foreign-source income is typically exempt upon repatriation. The approach of exempting active income, while taxing passive income, is often referred to as the tainted-income approach. Passive income may, for example, include dividend or interest income from financial investments.



Topics:  Taxation

Tags:  tax avoidance, controlled-foreign-company rules, profit-shifting, multinational enterprises

Professor of Applied Economics, ETH-Zürich; Research Fellow, CEPR

Professor of Economics at the University of Tuebingen