Who should tax a multinational enterprise?

In the last few days I’ve seen people tweet about this great international tax explainer by TPPI and an article about Australia not signing up to some international tax things. And I thought I’d like to write about them.

But before we can chat about them, I thought it would be fun to talk at a high level about the taxation of companies that operate across national borders (multinational enterprises, or MNEs) - and why the rules around such taxation are now being changed.

Why should I care? Tax their income.

Ok, lets start by focusing on income tax on the corporation.

Income tax on employees occurs in the country where the employees are located, and consumption tax on the final product occurs wherever the final good or service is consumed. As a result, the “labour” factor of production is taxed in the jurisdiction where the worker works, and the final consumption value of a product is taxed where it is consumed via the consumption tax.

Here the question is where the income from non-labour factors of production will be taxed.

So imagine we have a company operating in Australia. It employes people here, all of their assets (including IP) are located here, and it sells all its products here. Well any income earned by this company should then be taxed in Australia right.

But what happens if the intellectual property is in Singapore, the headquarters of the business is in the United States, the employees and factory are in Australia, and the products are sold in New Zealand - who gets to tax the corporate income?

To be clear everyone agrees there is an amount of income that reflects the “value-add” associated with non-labour factors of production - this is income that is liable for taxation. The argument is about WHO can tax it. And the owners of the company may try to shift this income to different countries if they are willing to tax it at a lower rate - leading to concerns about profit shifting that is both inefficient and unfair if we believe taxing rights should be allocated a certain way!

Given this, lets think about a couple of alternatives - noting that these should reflect i) a belief that the taxing rights SHOULD be on this basis ii) an evaluation of the risk of avoidance and the cost of compliance and adminstration:

1. Income is generated by the shareholders - and so the tax liability should be apportioned based on the tax residency of the shareholders. This is elegant in theory - a person has generated income, and we tax those people. However:

   1. compliance costs are high 1: working out the tax liability for a large company on the basis of the details of thousands of shareholders that change through the year sounds complex - unless we require a firm to pay out all earnings as dividends each financial year and ban retained earnings,

   2. compliance costs are high 2: individuals come and go over the year making this messier,

   3. this will incentivise avoidance: individuals can adjust their tax residency to avoid tax,

   4. this does not allocate revenue to the use of local amenities/infrastructure: the location of the activities of the corporation and their owners are different - as a result, the tax is not being allocated as a “user charge” to jurisdiction that provide infrastructure.

2. Apportion the tax liability based on an estimate of the “value-added” in a jurisdiction. This tries to attribute the tax based on the economic value of activities that take place within a geographic location. Such an approach reflects current tax rules, and involves transfer pricing that attributes a return to activities that have an “economic substance”.

   1. more compliance costs: calculating and then justifying economic substance involves a lot of accountants.

   2. administratively costly: this process involves a lot of tax officials and auditing - making it difficult for nations with weaker administrative functions to enforce.

3. Tax at the “destination” where the corporate sells. Instead of attributing random “economic value” amounts, just tax the full final value of the product where it is sold.

   1. This has the lowest compliance and administrative costs, and is also the most difficult to avoid!

   2. BUT it shifts the entire corporate tax base into consuming countries rather than producing countries - or from net exporters (from MNEs) to net importers. This may be deemed to be quite unfair.

As you can see, there aren’t exactly perfect solutions when activities cross borders. However, there are two questions mixed in one - one question about the distribution of taxing rights between countries, and another about the behavioural and administrative costs of a given system.

The current MNE tax system is predicated on the idea that it is “fairest” that factors of production are taxed based on the value-add they produce.

But such a system may make it easy to manipulate where tax is paid, generate high administrative costs for tax agencies to enforce, and necessitate large compliance costs for businesses to meet the anti-avoidance rules.

Hasn’t this always been an issue?

Yes. The initial determination of taxing rights based on source in the League of Nations (1923) was a decision made knowing these trade-offs. Concerns about the reallocation of R&D rights, the existence of “marketing intangibles” that should give destination countries a taxing right, and manipulation of corporate residency rules are all old ideas.

Transfer pricing rules were introduced in order to ensure that associated parties couldn’t just completely make up prices - imposing “arms length” rules and distributor margins.

The OECD BEPS process aimed to deals with issues that cropped up from this due to increasing globalisation and the growing scale of MNEs and cross-border financial transactions.

But digitalisation has in many ways made this worse.

Prior to the internet it was necessary to have a physical presence in a country in order to trade. Such a physical presence generated the ability to use transfer pricing to allocate taxing rights between countries when the corporate selling does have a significant presence in the market.

But now there is no need for such a physical presence - and so the ability to deem that a given potential taxpayer along the value chain does have a significant presence is limited. As a result, multinational firms can “reach into” a market without being subject to the same tax rules on value-added within the jurisdiction as domestic firms would be.

Here the issue has always existed - a multinational enterprise can inflate costs when bringing a product into a country to lower their tax liability in that country, and shift those profits to a lower tax jurisdiction. The inability to deem a rate of return to some function that relates to a physical presence just makes this issue more acute.

I have no problem with a bit more market competition - but if MNEs can structure their tax affairs to generally avoid tax, then such behaviour erodes the underlying tax bases of countries that have a corporate tax system - leading to a “race to the bottom” on corporate tax rates.

And this is the sort of behaviour that people at the OECD found, at that they’ve estimated would be in part redressed by a minimum global tax.

What are people suggesting

So the concern here is that MNEs are able to “reach into” markets, making use of the domestic infrastructure and laws without contributing to the revenue needed to maintain them. Furthermore, if they can organise to avoid paying tax at all the proliferation of large MNEs will erode the corporate tax base in general.

So when the OECD talks about “Pillar One and Pillar Two” they are talking about sets of rule changes that attempt to deal with both.

• Pillar One: Redistribution taxing rights to the destination country, including adjustments to distributional tax rules.

• Pillar Two: Set a minimum global rate of tax that must be paid by MNEs on global income.

Australia has legislated the rules for Pillar Two, but not Pillar One - as this looks like it has been delayed due to further global negotiations.

Are these rules a good idea. I don’t know, I don’t work on tax. But once you lay everything out you can see why it was a super important topic for people to think about!

Isn’t there something simplier

Yeah. Look at our three approaches above.

A final income tax on individuals earnings irrespective of source, a user charge on the use of domestic infrastructure, and a tax on consumption (where you have transfer pricing rules around distribution that allow you to adjust the implicit deduction allowed for imports).

Domestic countries can use a mix of these three instruments, and already have the means of mitigating any tax gap in a transaction they are concerned about unilaterally.

If that is so obvious why don’t they do it? Tax treaties - which are important for tax certainty and to prevent double taxation - limit the ability for jurisdictions to do this. Because it is costly to have a lot of bilateral negotiations the OECD process is trying to “coordinate” it for everyone.

With Pillar Two implemented it is part of the way there.

And with Trump reappearing, it is likely a Destination-Based Cash Flow Tax will as well. If that happens there may be little need for Pillar One in the first place.