Clear, simple and wrong: Zucman's proposal for reforming international tax

Influential economist Gabriel Zucman has proposed reshaping international tax to stop tax avoidance by multinationals.

https://econfip.org/policy-brief/taxing-multinational-corporations-in-the-21st-century/#

His proposal has the advantage of being clear and simple. But his conclusion that it would prevent tax avoidance is wrong. Here's why.

The idea is that we tax all companies on their consolidated group profit, with each country getting to tax the % of that profit equal to the sales in that country. It's a simple form of unitary taxation.

Let's ignore the considerable practical problems with implementing the proposal consistently worldwide, and say it magically happens. Then take an example:

• Say Apple has $1bn of sales across the EU - and let's simplify things by saying it has no sales anywhere else in the world (this won't change our conclusion).
• Assume Apple has a 40% profit margin.
• So under Zucman's proposal, we take Apple's consolidated profits of $400m and each EU state taxes the proportion that relate to local sales.
• Average EU corporate tax rate is 30%; let's simplify things further by assuming everyone is average.
• So total EU tax is $120m.

Difference from existing system is that Apple's corporate structure is irrelevant. It can have as many tax haven entities in its group as it likes, doing anything and everything - but if 100% of the sales are in the EU then 100% of the profit is taxed in the EU. Location of IP, employees, risk etc all irrelevant.

So the tax result isn't changed by how Apple makes its sales. But it is very dependent on *who* makes the sales.

Say Apple has a rival, Banana:

• Banana has exactly the same sales and profitability as Apple, but Banana sells in a different way. It sells all its products to a third party distributor in the US.
• Say the distributor will make a 2% margin, and cost of sales is $30m. So Banana sells to distributor for $950m. The distributor then ships to Europe and sells for $1bn.
• A bit of the profit previously made by Banana in the US is now made by the distributor in the EU.
• US tax rate is 21%, so Banana pays $380m x 21% = $80m.
• EU average tax rate 30%, so distributor pays $20m x 30% = $6m.

This different sales arrangement means Banana's after tax profits are $300m - Apple's are $280m. And EU's tax take from Apple is a healthy $120m... but its tax take from Banana is a feeble $6m.

The bottom line: sales based apportionment is easily distorted by whether a business makes B2C or B2B sales. 

And it gets worse. Let's hypothetically say that Apple wants to aggressively avoid tax however it can. It copies Banana's setup, but instead of appointing a third party distributor in the US, it appoints one in the Caymans, and so makes all its sales in the Caymans.

• Apple is now taxed on all the $380m of profit in the Caymans, at the rather competitive rate of 0%.
• The distributor is taxed in Europe at 30%, so again $6m.

So we've just spent years completely reshaping the international tax system, only to find we've facilitated avoidance on a spectacular scale.

Of course you could introduce anti-avoidance rules to stop such behaviour. And then new anti-avoidance rules when people employ different avoidance techniques. And then etc etc. Does this sound familiar? 

So the question isn't whether an idealised version of Zucman's proposal is better than the messy status quo. It's whether the messy reality of his proposal, and businesses' rational reaction to it, would be better than the messy status quo. I don't know whether it would. And neither does Zucman.

Coda: this isn't to say taxing on the basis of where sales take place is a bad idea. But there are better ways to do it - but that's the subject for another post. #DBCFT