There has been some confusion about how some parts of the recent G7 agreement on new tax rules for multinational companies might work. The new policies would target the largest and most profitable multinationals and bring in a global minimum tax.

The G7 agreement is subject to further debate and further agreement at the G20 and among more than 130 countries around the world. As debates continue it is important to understand the nuances of what is being discussed.

First is a question about revenue and profitability thresholds—which companies are in and which are out? Will certain large companies be carved in regardless of their profitability? Second, some are wondering about carve-outs and whether countries will continue to provide tax preferences like patent boxes under the global minimum tax or whether some countries would get carved out completely.

On both issues, there are many unsettled questions, but it is worth exploring what has been discussed up to this point and how those questions might be resolved.

One part of the global tax reform discussion, “Pillar 1,” would change the rules for where the largest and most profitable companies pay taxes. The G7 communiqué said that the targeted companies would pay taxes in countries where they have sales on at least 20 percent of profits exceeding a 10 percent profit margin.

For this to work, you need a definition of “largest” and “most profitable.” If a business is “large” but does not have a very high profit margin, then it might not have to comply with the new system. There are some large companies (one in particular) that politicians have identified as needing to be among those that have to comply with the new rules.

This gets to a question of how a company that might not meet the “profitable” metric would be required to play by the new rules. This is where something called “segmentation” comes into play.

For example, if you have a large multinational company that is a conglomerate with multiple business lines, then perhaps the more profitable lines of business would be carved into Pillar 1. Part of the business might be driven by valuable software and have a profit margin of 20 percent while other parts of the business are in manufacturing and distribution earning just a 5 percent profit margin.

If, by itself, the software business met both the “large” and “profitable” thresholds for Pillar 1, then the software portion of the company might have to play by the new rules while the less profitable manufacturing and distribution portions of the business would not.

The challenge for policymakers is how to define the dividing lines between business segments. It’s not too hard to imagine tax authorities gerrymandering businesses in ways that are most beneficial under the Pillar 1 rules.

But there’s probably a simpler (and less political) way to do this. Large companies already provide their audited financial statements to shareholders. Those statements are published in line with multiple rules and regulations and are meant to reflect the business and economic realities of a large multinational. Companies with multiple business lines already report profits by segment, so policymakers wouldn’t necessarily need to design something from scratch.

While it does not make sense for a policy to be designed with one (or more) specific companies in mind, segmentation based on financial statements makes sense in a Pillar 1 proposal that is already complex.

Without segmentation, a company with high profit margins might have an incentive to dilute its profitability by acquiring less profitable business lines. Segmentation provides a way to forestall that sort of merger activity to avoid Pillar 1.

Segmentation is not a new idea. Several pages were committed to segmentation in a blueprint of the policy published last fall. However, the overall approach in that blueprint was incredibly complex, and a lot has likely changed as policymakers have worked toward an agreement.

Basing carve-ins on something like financial reports to shareholders should avoid some of the more difficult political and complexity questions. It might not be a perfect solution, but as Richard Collier (one of the designers of Pillar 1) recently stated, “In Pillar 1, nothing is ever straightforward.”

The other big issue that has received lots of attention is whether there will be carve-outs from the global minimum tax, or “Pillar 2.” Similar to the focus on one company for Pillar 1, there have been questions about one country on the global minimum tax.

Like Pillar 1 and segmentation, the idea of a carve-out for the global minimum tax is not something new. Last October’s blueprint on the global minimum tax considers an option for a deduction for tangible assets and payroll costs when calculating whether the minimum tax applies. In the blueprint, this is labeled a substance-based carve-out.

The idea behind such a carve-out is that if you have real activities (so-called “substance”) even in a low-tax jurisdiction, then the minimum tax should not bite as hard.

The question for ongoing negotiations is whether a deduction for assets and payroll will be sufficient for world leaders. In recent years, countries with preferential tax policies like patent boxes have adopted rules that require some economic substance in order for companies to benefit from lower tax rates.

Let’s say a reasonable amount of economic substance turns off the global minimum tax in a low-tax jurisdiction. That would mean that policies like patent boxes, special economic zones, and related tax preferences would continue to exist so long as businesses are not artificially manipulating their profits to benefit from those preferences.

That would be the most generous form of a substance carve-out, and it would likely benefit many countries that have preferential tax policies.

The least generous form of a substance carve-out is not having any carve-out at all, which is what the Biden administration has proposed for the U.S. version of the minimum tax.

Between the U.S. position and the other end of the spectrum are a lot of alternatives including my own recommendation for designing the minimum tax base.

Whether companies get carved in based on their business segments or activities in low-tax jurisdictions get carved out because of the presence of economic substance, time will tell. There is a lot of work left to be done, but these policy questions should be answered taking into account the need for simplicity and neutrality and an awareness of the trade-offs rather than the politics surrounding certain companies or countries.

Launch U.S. International Tax Reform Resource Center