The global tax deal that was struck last year continues to move along in slow and uncertain ways. While the global minimum tax gets much attention in the media (and in my own writings, to be honest), there is another significant piece to the deal. The global minimum tax (also known as OECD Pillar Two) would set a floor on effective tax rates globally, and OECD Pillar One would change the rules for where large companies owe taxes.

In changing the rules for where companies owe taxes there are three questions that continue to rise to my attention as the proposals move forward:

1. How does this fit with existing rules?
2. What happens to digital services taxes?
3. Who decides where companies pay?

To answer these questions, it is worth exploring how OECD Pillar One works and its goals.

In recent years, a new type of tax emerged on the global stage: the digital services tax. The European Commission’s proposal for such a tax targeted large and primarily U.S. digital companies with a 3 percent tax on revenues (rather than profits). The proposal was not adopted by the EU, but individual countries quickly began implementing similar taxes. Major countries like France, Italy, and the United Kingdom put the taxes in place.

The United States government viewed the taxes as discriminatory against U.S. companies and threatened to retaliate with tariffs.

In addition to this tax and trade dispute, there has been a growing number of tax disputes between multinationals and governments about how much tax might be owed on activities like local marketing and distribution.

This situation with a tax and trade war and ongoing disputes across the world has been described as “chaotic” by tax professionals.

Enter OECD Pillar One. The big goals behind Pillar One are to change where companies pay taxes and provide carveouts and create a set of procedures and rules to address the chaos.

The rules would initially impact companies with global revenues above €20 billion (US $20.4 billion at today’s exchange rate) and profitability above a 10 percent margin. The revenue threshold would be cut in half after a review in the seventh year of the policy.

It uses a formula to reallocate where companies owe taxes on their profits (referred to as OECD Pillar One Amount A). The formula takes 25 percent of profits above a 10 percent margin and allocates that share to jurisdictions according to the share of sales in jurisdictions. The goal is to shift a portion of taxable profits away from jurisdictions where profits are booked currently—that is, where they are produced—and move them to jurisdictions where sales are made.

In some cases, the sales and the profits are already aligned, and the rules will take this into account with yet another formula that looks at where a company’s assets and employees are located.

Another part of Pillar One (referred to as Amount B) would make it easier to identify how much tax might be owed on marketing and distribution activities in countries.

Throughout the Pillar One proposals there is a desire to create tax certainty in the process of additional rules.

So, now, back to the three questions. How does Pillar One fit with existing rules? The simple answer is that it just sits on top of those existing rules, so this is of course not simple for taxpayers. Companies would continue to calculate their taxes owed under existing rules and file taxes in connection to current rules. In addition, they would begin calculating additional tax owed in some countries and less tax owed in others based on how the Amount A formula works.

And what happens to digital services taxes? The answer here is both positive and negative. The most recent policy document on Pillar One is clear about the need for countries to withdraw digital services taxes. It also clearly says that countries agreeing to Pillar One will be committing to not adopting similar policies. This would be a significant achievement in line with bipartisan concerns from Congress in recent years. However, as with many things in Pillar One, there is still work to do to outline specifically which policies this refers to. Does it include streaming taxes and digital advertising taxes in addition to the standard digital services taxes? That should be the case, but the list is not yet fully developed.

Finally, who decides where companies pay? In general, the design of Pillar One is based on a formula for determining where companies will owe additional tax money (and where they will get an exemption or credit for existing taxes paid). However, when there are disagreements about the numbers, as there surely will be, there needs to be a process and apparatus in place to adjudicate among the competing concerns and issues.

One document outlining how this could work gives an option for a panel of experts to have final authority on disputes. Representatives from the governments involved in the dispute would be involved at other stages, but unresolved disputes over allocation of taxable profits could arrive at the desk of experts who do not represent any government. The panel of experts would have to meet multiple technical and experiential requirements and they would be selected from pools of experts including individuals nominated by countries.

Another option would be to continue to include government officials in the final review. This would keep governments that have a stake in the negotiations involved. Expert advisors would still be involved in the process but without final determination authority.

A third option would have a mix of government officials and experts resolving disputes.

This issue is particularly important for the United States because U.S. companies will likely make up the majority of both companies and profits within the scope of the rules. According to research from economists at the Oxford University Centre for Business Taxation, that U.S. share of Amount A profits would be 64 percent and total $56 billion.

Pillar One is complicated, and it is challenging to identify narrow things to critique because the entire project is incredibly complex, built upon existing rules (which have their own issues), and the promise of addressing the chaos seems small relative to the challenge of complying with the proposed approach.

Trade-offs are everywhere in economic policy. In this case the choices available are continuing with the current level of chaos and uncertainty and adopting a set of rules that introduces new complexities on top of the existing convoluted system, provides authority to some set of experts rather than individuals representing governments, and somehow delivers some certainty at the end of it all.

There may be little compassion among the public for additional regulatory costs and uncertainties faced by large multinationals, but Tax Foundation defines principled tax policy along four lines: simplicity, transparency, neutrality, and stability. The design of Pillar One violates all of these in different ways.

Whether there is another path out of the international tax chaos is unclear, but as University of Virginia tax law professor Ruth Mason recently said at a tax policy conference, “There are limits to human understanding, and this deal may be hitting those limits,” and “rules that no one can understand are not accountable rules.”