The Global Tax Devil Is in the Details
NEW YORK – It appears that the international community is moving toward what many are calling a historic agreement to set a global minimum tax rate on multinational corporations (MNCs). It’s about time – but it may not be enough
Under the existing rules, firms can escape paying their fair share of taxes by booking their income in low-tax jurisdictions. In some cases, if the law doesn’t permit them to pretend that enough of their income originates in some tax haven, they have moved some parts of their business to these jurisdictions.
Apple became the poster child of tax avoidance by booking profits made on its European operations to Ireland, and then using another loophole to avoid most of Ireland’s notorious 12.5% tax rate. But Apple was hardly alone in turning the ingenuity behind products we love toward avoidance of taxes on the profits earned from selling them to us. They rightly claimed that they were paying every dollar due; they were simply taking full advantage of what the system offered them.
From this perspective, an agreement to establish a global minimum tax of at least 15% is a major step forward. But the devil is in the details. The current average official rate is considerably higher. It is thus possible, even likely, that the global minimum will become the maximum rate. An initiative that began as an attempt to force multinationals to contribute their fair share of taxes could yield very limited additional revenue, much lower than the $240 billion underpaid annually. And some estimates suggest that developing countries and emerging markets would also see a small fraction of this revenue.
Preventing this outcome depends not just on avoiding a downward global convergence, but also on ensuring a broad and comprehensive definition of corporate profits, such as one that limits deduction for expenses relating to capital expenditures plus interest plus pre-entry losses plus... It would probably be best to agree on standard accounting so that new tax-avoidance techniques do not replace the old ones.
Particularly problematic in the proposals advanced by the OECD is Pillar One, intended to address taxing rights, and applying only to the very largest global firms. The old system of transfer pricing was clearly not up to the challenges of twenty-first-century globalization; multinationals had learned how to manipulate the system to record profits in low-tax jurisdictions. That’s why the United States has adopted an approach whereby profits are allocated among the states by a formula that accounts for sales, employment, and capital.
Developing and developed countries may be affected differently depending on which formula is used: an emphasis on sales will hurt developing countries producing manufactured goods, but may help address some of the inequities associated with the digital giants. And for Big Tech firms, the value of sales must reflect the value of the data that they garner, which is crucial to their business model. The same formula may not work in all industries.
Still, the advances made in current proposals have to be recognized, including moving away from the “physical presence” test for imposing taxes – something that makes no sense in the digital age.
Some view Pillar One as a back-up to the minimum tax, and are thus not concerned about the absence of economic principles guiding its construction. Only a small fraction of profits in excess of a certain threshold are to be allocated – implying that the total share of profits to be allocated is indeed small. But with firms permitted to deduct all production inputs, including capital, the corporate income tax is really a tax on rents or pure profits, and all of those pure profits should be up for allocation. Thus, the demand by some developing countries that a larger share of corporate profits be subject to reallocation is more than reasonable.
There are other troublesome aspects of the proposals, as best as can be found out (there has been less transparency, less public discussion of the details than one would have expected). One concerns dispute resolution, which clearly can’t be conducted using the kinds of arbitration now prevalent in investment agreements; nor should it be left to a corporation’s “home” country (especially with footloose corporations looking for favorable homes). The right answer is a global tax court, with the transparency, standards, and procedures expected of a twenty-first-century judicial process.
Another of the proposed reforms’ problematic features concerns the prohibition of “unilateral measures,” seemingly intended to curb the spread of digital taxes. But the proposed threshold of $20 billion leaves many big MNCs outside the scope of Pillar One, and who knows what loopholes smart tax lawyers will find? Given the risks to a country’s tax base – and with international agreements so difficult to conclude and MNCs so powerful – policymakers may need to resort to unilateral measures.
It makes no sense for countries to give up any of their taxing rights for the limited and arbitrary Pillar One. The commitments asked are incommensurate with the benefits given.
The leaders of the G20 will do well to agree on a global minimum tax of at least 15%. Regardless of the final rate that sets the floor for the 139 countries currently negotiating this reform, it would be better if at least a few countries introduced a higher rate, unilaterally or as a group. The US, for example, is planning on a 21% rate.
It is crucial to address the host of detailed issues required for a global tax agreement, and it is especially important to engage with developing countries and emerging markets, whose voice has not always been heard as clearly as it should be.
Above all, it will be essential to revisit the issue in five years, not seven, as currently proposed. If tax revenues do not increase, as promised, and if the developing and emerging markets fail to garner a greater share of those revenues, the minimum tax will have to be raised and the formulae for allocating “tax rights” readjusted.
Подписывайтесь на нашу страницу в Facebook