Planning for Border Adjustments: A Practical Analysis

The destination-based cash flow tax has been supported by some brilliant economists and academics, and this support has played a role in its adoption by the House Republican leadership. These economists and academics have included among their arguments, however, that it would eliminate the need for certain transfer pricing, allocation, and apportionment rules.¹ This was doubtless intended as a limited observation, but it has been translated into a purported advantage of practical simplicity that now permeates the general discourse on the subject.² A recent summary of the proposal in Tax Notes International, for example, noted that "proponents of the destination-based cash flow tax argue that it would be simpler to administer because it would eliminate the need for complex transfer pricing analyses and the need to determine the true location of economic activity."³ Indeed, House Ways and Means Committee Chair Kevin Brady, R-Texas, most recently described the border adjustment plan as "stunningly simple."⁴ Alas, based on my own 32 years of practical experience, I would beg to differ.

For those of you who may not have time to read the rest of this report, the big picture is this: Setting currency arguments aside, border adjustments should greatly increase the incentive for businesses to organize their operations inside the United States (regardless of where their goods are consumed) and so should serve to increase exports, decrease imports, and increase economic growth. Just as it is easier to spend money than to earn it, however, it is easier to grant an exclusion of export proceeds than it is to collect a corresponding import tax. Taxpayers will likely prove more energetic in their pursuit of the exclusion than they will in their efforts to pay the import tax. Indeed, common sense suggests that they will actively seek out the former and avoid the latter, and relevant rules must be written to accommodate this fact. Moreover, the export proceeds exclusion that we will be granting does much more than exempt an exporter's net income from tax. It excludes the exporter's gross sales proceeds while allowing deductions for all of the offsetting costs of production or acquisition, thereby giving rise to very large net tax losses on a stand-alone basis.

This report therefore begins the task of examining, from a practitioner's perspective, what sorts of rules and regulations the government would need in order to defend its revised tax base properly. I say "begins" because I'm sure that it barely scratches the surface of what would be required. But this is an exercise that must be undertaken sooner rather than later. For if the rules that govern our border adjustments are relatively wooden and lacking in fact and situation-specific detail, then maximizing the export proceeds exclusion will be a matter of routine tax planning, while much of the purportedly offsetting import tax will slip through the government's fingers. But if the rules are detailed and complex enough to allow the government to defend its revised tax base adequately, then they will be at least as complicated to apply in practice as our current cross-border rules — and they most definitely will include complex transfer pricing analyses and a need to determine the true location of economic activity. Moreover, given the proposal's novelty, Treasury and the IRS will need to double their resources and their focus to quickly cover a lot of new ground.

The Size and Effect of Border Adjustments

Under the blueprint's proposed border adjustments, we would (1) not allow any business deductions or cost basis for assets purchased from outside the United States, (2) likewise impose tax in respect of a direct purchase by nonbusiness consumers of foreign goods and services,⁵ and (3) most significantly, allow a U.S. exporter to exclude from income the entire gross proceeds from the sale of the exported property while still deducting all of the costs of producing or acquiring the property. The export sale would therefore result in a reduction of the exporter's net income (that is, would result in a tax loss on a stand-alone basis) equal to the entire amount of the proceeds of the sale. (Conversely, in the case of imports, neither deductions nor cost basis would be allowed for any of the costs of acquisition, and the importer's net taxable income would therefore be increased by an amount equal to the entire purchase price of the import.)

I note in this regard that there is still some confusion as to how the proposed export proceeds exclusion would interact with deductions and tax basis under the blueprint.⁶ The blueprint expressly states, however, that it is proposing border adjustments in order to create parity with the border adjustments that are allowed by other countries under their VATs.⁷ Presumably, then, exporters would be entitled to separately deduct all of their costs associated with their manufacture or acquisition of the exported property while simultaneously excluding the proceeds of the sale.

For example, suppose that a U.S. exporter purchases goods from a U.S. manufacturer for $100 million and sells them to foreign consumers for $100 million. If a typical credit-style VAT were in force at a rate of 20 percent of the VAT-inclusive proceeds,⁸ the exporter's $100 million purchase price would have included $20 million of VAT. The exporter would have received a refund of this $20 million of VAT from the government, but it would not have had to collect or remit any VAT in respect of its sale of the goods.⁹ Thus, its $20 million of cash profit would have come directly from the government.

It follows that under the blueprint's proposed border adjustments (which would implement an economically equivalent 'subtraction-style' approach at a rate of 20 percent), the exporter would be entitled to separately deduct the $100 million that it paid for the exported goods (which is economically equivalent to receiving a $20 million tax refund) and then exclude from gross income the entire $100 million of proceeds from the sale of the goods (which is economically equivalent to not having to collect and remit $20 million of VAT on the export). This would allow the exporter to book a $100 million net loss for tax purposes (notwithstanding that it had no loss as an economic matter), which it could then likewise monetize (if not by reducing its own otherwise tax liability by $20 million, or by receiving a $20 million check directly from the government, then by effectively selling the tax loss to other taxpayers — more on this later). Thus the exporter (like the exporter under a VAT) would receive all of its $20 million of profit from the government.

This is to be contrasted with the case in which an exporter purchases goods for, say, $90 million, sells them to foreign persons for $100 million, and the tax law allows the exporter to exclude its resulting $10 million of profit from income (a tax benefit worth $2 million rather than $20 million). The "leverage" associated with allowing full-cost deductions while excluding gross sales proceeds results in a tax benefit that is much, much greater.

The relatively large size of the resulting tax benefit plays a key role in the discussion below. But allow me to start by making an important point. Numerous commentators have asserted that the proposed border adjustments would create "tax parity" by eliminating the incentive for U.S. corporations to organize their operations outside the United States. As compared with current law (and setting currency arguments aside), however, they would in fact create a very large incentive in the opposite direction — that is, to produce goods inside the United States regardless of where they were ultimately consumed.

Consider the following example: A corporation (whether U.S. or foreign) currently manufactures corkscrews in Spain and sells them to people who live in Spain. The corkscrews cost $90 million to manufacture in Spain, and they are sold there for $100 million. (The manufacturer pays $2 million of Spanish income tax on its economic profit, and it also collects $25 million of VAT from the purchasers of its corkscrews and remits it all to the Spanish government.¹⁰) Under the blueprint's proposal, if the manufacturer moves its plant to the United States and exports the corkscrews to Spain, it doesn't just avoid $2 million of Spanish income tax. It also deducts its $90 million of manufacturing costs for U.S. tax purposes while excluding the $100 million of sales proceeds. The resulting $90 million tax loss can either be used to offset $18 million of tax that would otherwise be paid by a U.S. affiliate or effectively be sold to other taxable U.S. persons for more than $10 million, thereby more than doubling the manufacturer's profit (more on this later). That's quite an incentive.

It is true that the manufacturer would still have to collect an additional $25 million of Spanish VAT from its Spanish customers and remit it all to the Spanish government.¹¹ But this would not negate the effect of the U.S. export proceeds exclusion, for Spain would require the manufacturer to do this regardless of where the manufacturing took place and regardless of whether the United States offered the manufacturer an export proceeds exclusion. It is also true that the wages and prices paid by the relocated manufacturer to its U.S. employees and suppliers would reflect the fact that they were subject to U.S. income tax. But Spain also imposes an income tax on its employees and suppliers, at rates similar to the rates proposed by the blueprint.¹²

Conversely, suppose that a business currently manufactures corkscrews in Spain for $90 million and sells them in the United States for $100 million. It is true that Spain refunds all of its VAT when the corkscrews are exported to the United States. But there is no refund of the Spanish income tax, which is imposed at a rate similar to the rate proposed by the blueprint and which therefore already places the Spanish manufacturer in an economic position equivalent to that of a U.S. domestic producer that has been required to pay U.S. income tax (at the corporate and individual levels, taking account of the tax paid by its employees and suppliers) on all of the value added by its U.S. production of corkscrews. The additional imposition of a U.S. border-adjustment import tax would therefore serve to impose U.S. income tax on the manufacturer's value added in addition to Spanish income tax.¹³ To use the numbers set out above, the manufacturer's $10 million of profit would become a $10 million loss after imposition of a $20 million U.S. import tax. The manufacturer would therefore be forced to relocate its operations to the United States in order to pay only one income tax, and the relevant tax base would therefore shift from Spain to the United States.

In other words, if we set aside the argument that all border adjustments are offset by currency adjustments anyway (in which case it's not clear why anyone would be proposing them), the border adjustments proposed by the blueprint would constitute a very large incentive (as compared with current law) for producing within the United States goods and services to be consumed all over the world. What the blueprint asserts is not that the effects of this incentive would be offset by currency adjustments, but rather that this incentive would put us on an even footing with other countries that impose border-adjusted VATs.

Meanwhile, the blueprint's conclusion that border adjustments would raise a considerable amount of revenue appears to be based on the fact that current imports exceed current exports, and so import tax collections should exceed export proceeds exclusions. It doesn't take into account how businesses would react to these very large tax incentives (currency theories notwithstanding). But in theory the resulting inbound relocation of manufacturing for export shouldn't increase tax collections, because the export proceeds exclusion would give up the resulting increase in tax; and conversely, the resulting inbound relocation of manufacturing for import shouldn't decrease tax collections, because the resulting loss of import tax should effectively be replaced by increased income tax collections at the corporate and individual levels. Before we start counting our money, however, we must first address a very important practical question: How — having granted a large export proceeds exclusion to any business that structures its affairs to meet its requirements — are we actually going to collect the offsetting import tax?

The Location of Tangible Goods

One thing that leaps out at you as you begin to consider the practical implications of such border adjustments is that they make the most sense when applied to physical goods that are actually crossing the border. It is reasonable, for example, to allow a manufacturer to exclude all of its proceeds from the sale of flat-screen televisions that it puts on a ship setting sail for Spain (where the purchaser, a Spanish corporation, will presumably distribute them to individuals who actually reside in Spain). And it is probably likewise reasonable to assume in such a case that the revenue lost by granting so large an export proceeds exclusion will be made up for by revenue gained from imposing tax on the import of similar physical goods coming from China and arriving in U.S. ports or airports.

It is probably not a good idea, however, to allow a U.S. manufacturer (or wholesaler) to exclude proceeds from the sale of such televisions to a Spanish corporation that is planning to store them in a warehouse in New York. Why? For the obvious reason that the Spanish corporation will probably distribute those televisions to U.S. wholesalers, retailers, or consumers, and those U.S. wholesalers, retailers, or consumers will probably not realize that they are now "importing" goods as a technical matter (notwithstanding that the goods say "made in the USA") and so deny themselves a tax deduction in respect of their purchases (or in the case of consumers, pay an import duty). It seems unlikely, moreover, that the government would have the resources required to pursue such U.S. wholesalers, retailers, and consumers as a practical matter and recoup the manufacturer's (or wholesaler's) very large export proceeds exclusion through the enforcement of a corresponding import-related tax.

Most VATs deal with this practical concern by incorporating such concepts as the "place of supply" of the relevant merchandise. If the place of supply of the televisions to the Spanish corporation is in the United States, then it is not yet ripe to grant anyone a refund of VAT. The Spanish corporation can claim such a refund if and when it finally loads the televisions onto a ship that actually sets sail for Spain. Bear in mind that the stakes here are enormous, because the seller gets to exclude all of the proceeds of the sale of the property while deducting all of the costs of manufacturing it. Thus, if the seller spends $900 million to manufacture televisions and sells them for $1 billion, the proposed export proceeds exclusion doesn't just allow it to avoid tax on its $100 million of economic profit. Rather, it allows it to report a $900 million loss for tax purposes — a loss that it may be able to "monetize" in the marketplace for as much as $150 million in cash. With incentives so high, could you blame the U.S. manufacturer for always selling the goods in question to a technically foreign purchaser (if that were all that the law required to obtain the export proceeds exclusion)?

A similar practical point can be made about the import of goods that are actually produced outside the United States. If smartphones produced in China are physically entering the United States, it will be possible for import agents to inspect and register them and require that the importer pay an import tax (else the goods will not be released). The Coast Guard can search for and seize any goods that importers are attempting to land on isolated beaches, and if we do build a wall on our southern border, it can serve to keep untaxed goods out of the United States in addition to untaxed people. To the extent that we are planning to forgo such an import tax on the grounds that a business purchaser of the goods will not be entitled to any cost basis in respect of the amount paid for its purchase, we had better be sure that we know who the purchaser is and that we can rely on that purchaser to actually pay a lot of tax later on when it sells the goods in question and then files a U.S. tax return. This may require a fair amount of the IRS in terms of new rules for compliance and enforcement.

Suppose that the importer, a Chinese corporation, does not identify any purchaser, explaining that it has not yet sold any of the goods. It is merely bringing the goods into the United States to store in its own warehouse located in San Francisco. Can we rely on the Chinese corporation to tell its ultimate U.S. purchasers that they are "importing" the goods and must therefore not include any amount that they pay for them in cost basis (or, if they are consumers, that they must file a form and pay an import tax)? Even if we can, can we further rely on those U.S. purchasers to follow through? Probably not. It would therefore seem wiser to treat a U.S. branch of the Chinese corporation as having purchased the goods from its own foreign branch, thereby requiring the Chinese corporation (rather than its U.S. customers) to pay U.S. tax when it sells the goods (given that it would be denied either a cost basis or a deduction for its deemed import purchase). As mentioned above in footnote 11, this is how a VAT would deal with such a case. (The Chinese corporation would pay VAT on its own import and then credit that payment against the VAT it received from its ultimate customers.)

But even assuming we do that, can we rely on this deemed U.S. branch of the Chinese corporation to actually file a U.S. tax return and pay a lot of tax when it later sells the goods to U.S. wholesalers, retailers, or consumers? If the Chinese corporation is big enough, and if its U.S. presence is large enough, the answer could well be yes. But in many cases it will be wiser to collect the tax at the dock and hold the goods up until it has been paid. This is likewise how a VAT would deal with such imports. VAT customs officers would refuse to allow the goods into general circulation until someone had paid the import VAT. This may likewise require a fair amount in terms of new substantive and procedural rules from the IRS.

In thinking through these sorts of practical issues and problems, there are a few things that we should bear in mind. The first is that we should probably not be relying primarily on consumers to help us enforce the import tax. Many consumers will not voluntarily pay an import tax, and it will not be easy to force them to do so. We might consider as a model here the job that the 50 states have done in enforcing and collecting the "use tax" that most of them purport to impose on their residents when they import goods and services from out of state. According to a study done by Copenhagen Economics in 2016 (commissioned by UPS), "shipments sent via national postal operators resulted in a lack of payment of VAT and import duties to the national authorities for more than 60 percent of items purchased online. . . . The resulting loss of public income [for Denmark] is estimated at up to 1.3 billion euros annually."¹⁴

But even if they wanted to comply, most U.S. consumers would have little way of knowing that they had purchased their goods from a technically foreign seller, and they would certainly have no way of knowing this where the goods they had purchased said "Made in the USA." And U.S. retailers might not be in a very different position. Could we really hold to account all of the wholesalers and retailers in the United States and successfully require them to determine whether they had purchased their goods from a technically foreign seller (because someone in the supply chain had obtained an export exclusion)?

In short, imposing and enforcing a destination-based tax in respect of even the most simple tangible goods will require a considerable amount of practical forethought followed by the implementation of detailed substantive and procedural rules, and it will require much more in the case of intangible goods and services, as we shall shortly see.

Transfer Pricing and Recharacterization Rules

There is another reason to limit the export proceeds exclusion to cases where goods physically leave the United States. It is that otherwise, it will be especially easy for taxpayers to sell their NOL carryforwards to other taxpayers. The intention under the blueprint's proposed cash flow tax is that when one business sells assets or goods to another, the buyer deducts the purchase price and the seller includes it in income. But the parties could instead "elect" to do neither if the export proceeds exclusion was not so limited. The seller would simply sell the goods to a foreign corporation (which might be an affiliate of the seller, an affiliate of the buyer, or an independent accommodation party) for "export," and the buyer would then "import" them from this foreign intermediary. Thus, the seller would exclude the proceeds of the sale, and the buyer would forgo its purchase deduction. Needless to say, the seller would be currently taxable, while the buyer would have an NOL carryforward (because it was a net exporter) and so would not have been able to use the forgone deduction anyway.

This point is part of a much broader picture, however, that it might be better to introduce now, before we move on to the harder cases of intangibles and services. As I noted earlier, "Proponents of the destination-based cash flow tax argue that it would be simpler to administer because it would eliminate the need for complex transfer pricing analyses and the need to determine the true location of economic activity."¹⁵ But if we actually do eliminate these things, then "selling" NOL carryforwards to taxable business partners will be like taking candy from a baby.

Let's start with the simple case in which one U.S. corporation (the buyer) makes a payment to another U.S. corporation (the seller) in exchange for a transfer of goods from the seller's foreign subsidiary to the buyer's foreign subsidiary. Arguably, the payment should have been made from the buyer's foreign subsidiary to the seller's foreign subsidiary, where the payment would have been completely outside the U.S. tax system. (The blueprint proposes a territorial system under which profits derived outside the United States by foreign subsidiaries would be exempt from U.S. tax unless and until their goods were imported into the United States.) However, the "misplacement" of the payment allows the buyer's U.S. parent to fully deduct the amount of the payment for U.S. tax purposes. True, the misplacement also results in a corresponding income inclusion for the seller's U.S. parent, but the seller's U.S. parent has an NOL carryforward available to offset the income inclusion.

An equivalent result obtains in the opposite case, where the goods are transferred from one U.S. parent to the other but the payment is booked between their foreign affiliates (which are outside the U.S. tax system). Here the U.S. buyer has NOL carryforwards (and therefore couldn't have used its forgone purchase deduction), while the U.S. seller is currently taxable (and therefore benefits from the fact that it needn't include any sales proceeds in income, even though it has already deducted all of the cost of producing or acquiring the goods that it transferred to the U.S. buyer).

An equivalent result also obtains if the above U.S. seller directs the U.S. buyer to make the payment to one of the U.S. seller's foreign subsidiaries. In that case, the U.S. seller likewise gets to exclude the sales proceeds from income (even though it has already deducted all the costs of producing or acquiring the goods in question), whereas the U.S. buyer doesn't care that it is denied a deduction for its resulting "import," because it has NOL carryforwards. And an equivalent result also obtains in the opposite case where goods are transferred from one foreign subsidiary to another but the parties direct that the payment to the selling foreign affiliate be made by the buyer's U.S. parent. No doubt there are many more variations on this theme.

Surely, then, we do not intend to allow mere cash flows to govern the availability of export exclusions without any economic inquiry into the nature of the transactions that gave rise to them. For if we are going to allow taxpayers to sell their NOLs to each other in this manner without making any effort to stop them, then we might as well allow them to sell them directly to the government (that is, give taxpayers a direct cash payment each year in respect of their NOLs), so that we at least don't experience the transaction costs, economic distortions, and dead-weight loss associated with requiring them to sell their losses to each other.

But one reason why we don't do the latter is that we fear (and rightfully so) that if we did, there might then be a considerable amount of export-related fraud, given the very large size of the tax benefits that we are proposing to grant to exporters. In Europe, the most common tax fraud involves an exporter that claims a VAT rebate and an associated importer that disappears without paying the inbound VAT due. The relevant goods (which tend to be popular high-value goods, such as mobile phones) go round in a circle, which is why this is commonly known as "carousel fraud." HM Revenue & Customs' estimate of the amount of this fraud in the United Kingdom for the tax year 2005-2006 was between £3.5 billion and £4.75 billion.¹⁶ As noted above, we for this reason need a concept of place of supply, so that merchandisers don't routinely transfer their goods to a foreign intermediary and let the government worry about how to collect the offsetting import tax. But even with such a concept, one worries that some merchandisers might purport to send their goods off on a ship only to have them find their way onto the black market for untaxed goods, along with the many smartphones that really had been produced in China but were merely evading the import tax.

This fraud might not be greatly reduced if rather than rebating net losses directly, we allowed taxpayers to monetize them at will by mispricing or misplacing their transactions with taxable parties in the market. It seems to me, therefore, that we must be willing to at least make some effort to audit taxpayers and enquire into the nature of their payments to each other and the economic transactions to which they relate. And if those payments are misplaced or mispriced, then we must have at our disposal adequate transfer pricing and recharacterization rules that would allow the IRS to ignore payments between U.S. companies, or deem them to exist, as appropriate in light of the economic reality. And we had better be planning, in this regard, to beef up our audit and enforcement resources.

The Location of Intangibles

This leads us to the consideration of two important questions: First, how are we planning to deal with intangibles, such as enormously valuable rights to license computer software, to manufacture drugs, or to distribute electronic music, film, and television entertainment? Obviously, these rights do not take a physical form, and they do not enter or leave the United States in such a form — and more importantly, nothing stops them from entering or leaving the United States undetected by government agents. It is therefore difficult to see how we are going to allow, enforce, and defend border adjustments on a very large scale in respect of the import or export of such rights.

How, for example, are we going to collect an import tax when U.S. consumers download video games off the internet from suppliers with no physical location or address? Are we going to chase noncomplying suppliers around cyberspace and attack them with avatars? Or are we going to burst physically into the homes of consumers and demand that they share with us the contents of their clouds? (Lest you think I'm being silly, European revenue agents attempting to enforce a VAT on physical goods routinely follow customers out of clothing stores and demand that they show them their VAT receipts.)

One might therefore conclude that we should not apply border adjustments to intangible goods. But it's not clear that we have the political option to do that. American retailers slapped with a complete disallowance of either cost basis or deductions for the goods that they import from China may not understand why retailers are not subject to the same disallowance when they import Japanese video games. And President Trump — who said "Buy American" in his inaugural address — might not understand why either. Likewise, software manufacturers may complain that when hardworking Americans produce a new video game on a Silicon Valley campus and export it to excited adolescents around the world, they should be entitled to as much of an export proceeds exclusion as any other hardworking American.

That said, it seems to me essential that we limit the application of an exclusion for the export of computer software or electronic entertainment to cases where foreign individuals or businesses are employing them for their own personal use. We mustn't grant a large export proceeds exclusion for selling the wholesale right to license computer software, for if we do, we are bound to discover that such software is ultimately being consumed partly or primarily in the United States, and an offsetting import tax will be impossible to collect. This means that we must also keep all of our rules under sections 367 and 482 in place that deal with the outbound transfer of intangible rights. If such rights are transferred to foreign subsidiaries, they must still be taxed at their fair market values, because we cannot prevent the intangible products they give rise to from reentering the United States on a tax-free basis.

The same logic might not hold true, however, in the case of a sale for export of the right to manufacture drugs outside the United States. The drugs themselves are presumably in physical form, and given that the manufacture and dissemination of drugs is (hopefully) a highly regulated industry, perhaps we can have some confidence in our ability to (1) tax such drugs when they enter our country in physical form, and (2) prevent foreign owners of the license from relicensing it (through intermediaries) to U.S. manufacturers without paying an import duty. If that is the case, then we may need a different rule for exporting rights to license drugs than for exporting rights to license software. Indeed, it may turn out before we are done that we need lots of different rules for lots of different rights — and sorry experience may have to be one of our guides in this regard.

The Location of Services

The second (and sister) question is whether we intend to extend border adjustments to the performance of services. This is another case where one might normally expect the answer to be no, for the benefits would hardly be worth the complexity. After all, most services that are performed in the United States are performed for the benefit of people in the United States, and most services performed outside the United States are performed for the benefit of people outside the United States. But the president made it clear in his inaugural address that the order of the day is not only to buy American, but to hire American. And in that context, it might be difficult as a political matter to disallow deductions for purchasing goods from India but not for outsourcing services to India.

But if we really are going to border-adjust cross-border payments for the performance of services, we are going to have to promulgate a lot of new rules and deal with a lot of new practical problems. Let's start with the difficulty of determining both the place where services are supplied and the place where they are consumed. Suppose, for example, that a U.S. credit card issuer issues a credit card to a German corporation because the German corporation has a customer relationship with the U.S. credit card issuer's German subsidiary, and the card is used by one of the German corporation's officers to buy dinner in New York, and then in Germany. First we must ask whether interchange fees are in fact received in exchange for providing a service, and whether they should therefore be border-adjusted. If so, we must next ask whether these particular interchange fees are received by the U.S. credit card issuer in exchange for exporting this service.

In order to answer that question, we must first determine which company rendered the interchange service. Do we determine this by reference to the company that issued the credit card or the company that had the relationship with the customer? Next we must determine where the service was rendered. Do we determine this by reference to the issuing company's location, to the customer's location, or to the location where the card was used? Next we must determine where the service was consumed. Do we determine this by reference to the customer's location, to the officer's location, or to the location where the card was used?

According to HMRC's VAT manual dealing with the place of supply of services (VATPOSS05000), here are some of the factors that VAT collectors need to take into account in determining where services are supplied or received:

• how the particular services are provided;

• the significance of the activities carried out at each establishment in contributing to the services provided;

• where the necessary human and technical resources (for example, the database, technical equipment, office equipment, and telephones) for actually providing the services are permanently based;

• which establishment appears on the relevant contracts, correspondence, and invoices;

• where the directors or other personnel who entered into the contract are permanently based;

• where decisions are taken and controls are exercised over the performance of the contracts;

• whether reference to the preferred establishment leads to a more appropriate or rational result for tax purposes;

• the contract(s) between the supplier and customer;

• when a business supplies services on behalf of an overseas business as the latter's fixed establishment, the contract(s) between those businesses and the contract(s) between the overseas business and its U.K. customers;

• when there are no written contracts, a written account from the supplier that sets out in detail its understanding of the oral contract;

• oral evidence (in written format) or witness statements obtained from the supplier's staff and the staff of other businesses involved in the supply chain;

• telephone bills, fax printouts, etc. (which may demonstrate the extent of any day-to-day contact with a particular establishment);

• letterheads (although this could indicate a mere postal address rather than a business or fixed establishment);

• invoices or payments for office rental;

• certificates of incorporation and other company details such as common directors;

• job descriptions, performance agreements, etc. (which may be indicative both of the substance and reality of work performed at a particular establishment, and the nature of the supply);

• details of how the business fits into any larger corporate structure;

• company minutes, reports, and other internal correspondence; and

• evidence of any information obtained through mutual assistance.

And this is for plain vanilla businesses.

Whatever the rules may ultimately prove to be under U.S. law, how would both taxpayers and the IRS obtain the necessary information and then organize it to answer such questions? Would we draft detailed allocation and apportionment rules? Would they need to be supported by computer printouts showing the activities that were undertaken by particular employees at particular times of the day? How would the U.S. corporation obtain this information from the German corporation, and how would we force the latter to comply? What hypothetical payments would be proposed to deal with the relevant mischaracterizations, and on the basis of what specific regulatory authorities?

In any case, one thing is clear: There will need to be detailed regulatory answers to these questions, else every taxpayer (and ultimately every revenue agent) will answer them differently. Moreover, the answers will need to be different for different industries and situations. The credit card case is but one small example of what might arise in a modern economy. Even then, the answers will often be uncertain and inconsistent, and they will lead (as they currently do in Europe) to a fair amount of litigation. And compliance with these rules will in any case require a considerable amount of energy, effort, and focus on the part of both taxpayers and the IRS.

Now let us return to the problem of the potential for mispricing or misplacing payments. Consider the case of a U.S. parent corporation and a German parent corporation, each of which has subsidiaries around the world. The U.S. subsidiary of the German parent corporation has substantial NOL carryforwards. The U.S. parent corporation therefore engages it to provide the U.S. corporation and its affiliates with telephone operator services, paying it a fee of $10 million per month. Telephone operators working for both the German corporation's U.S. subsidiary and its various affiliates around the world provide help and information to the customers of the U.S. parent corporation and its affiliates around the world. A centralized computer system apportions incoming customer inquiries to the various worldwide telephone service centers. As a result, the location of the relevant telephone operators, and of the relevant customers, varies with customer inquiry load and service operator availability.

All this sounds rather complicated, and the destination-based cash flow tax is supposed to be simple. We might therefore be tempted to simply follow the cash flows and their nominal recipients. The U.S. parent corporation would therefore be entitled to deduct in full its monthly payment to the German corporation's U.S. subsidiary, and we would ignore the fact that as an economic matter, (1) much of what it was paying for would be for services provided to its U.S. customers by telephone operators residing in India, and (2) much of what it was paying for would be for the provision of services to customers of its foreign subsidiaries, subsidiaries that would not be subject to any U.S. tax. But bear in mind that if the shoe was on the other foot, and it was the U.S. parent corporation that had NOLs (and the U.S. subsidiary of the German corporation that was fully taxable), then the U.S. parent corporation would "import" the relevant services from the German parent corporation by making payments to it directly, and we would ignore the fact that sometimes the U.S. subsidiary of the German corporation was providing services to U.S. customers of the U.S. parent corporation, deducting the full cost of providing these services but not including anything in income.

Nor would the mispricing problems be limited to cases involving unrelated parties, for the blueprint proposes to retain the rules of subpart F. Consider the following case, which is doubtless one of many: A U.S. corporation has an Indian subsidiary that is earning subpart F income subject to U.S. tax. The Indian subsidiary hires a chain of local telephone operators and deducts their salaries, thereby lowering its net earnings (and thus its subpart F inclusions). The Indian telephone operators provide services to the U.S. corporation for free. Obviously, the IRS would have to apply transfer pricing rules to impute a deemed payment for services from the U.S. corporation to its Indian subsidiary (so that it could then disallow the U.S. corporation's deduction while still requiring the Indian subsidiary to include the deemed receipt in its earnings). The United Kingdom certainly relies on such rules to protect its tax base from such mispricing.¹⁷

In short, promulgating and enforcing rules designed to ensure that the sourcing of services giving rise to border adjustments was consistent, accurate, and in accordance with economic reality would be arduous and complex. But the cost of our failure to do so would be relatively high. We'd be better off excluding services from border adjustment.

More on the Sale of Export Tax Benefits

As noted above, the blueprint expressly states that its reason for introducing border adjustments (notwithstanding the enforcement and base erosion problems that they would raise) is to put the United States on an equal footing with other countries that border-adjust their VATs and so encourage exports and discourage imports, as compared with not having such adjustments.¹⁸ I agree with the blueprint. If that is what border adjustments would accomplish for us, then they probably would be worth implementing. More recently, however, some of the blueprint's own proponents have been asserting that the proposed border adjustments would not have these effects because the proposed border adjustments would be completely offset by changes in currency valuations. If that were true, then the blueprint's proposal would be founded on a misimpression, for (1) there would be no need to obtain parity with other countries that were border-adjusting their VATs, since currencies do that automatically, and (2) border-adjusting in response wouldn't accomplish anything anyway, since currency adjustments would reverse the accomplishment.

But it does seem to me that this currency theory is open to question.¹⁹ And more to the point, to the extent that the blueprint's own proponents have concluded that the theory is correct and that the blueprint's proposal for border adjustments therefore was based on a misimpression, then what is the point of continuing to propose them, particularly in light of the very real practical issues and concerns set out above? I shall therefore continue to take the blueprint at its word and assume for the remainder of this report that increasing exports and decreasing imports is indeed the objective of the proposal to implement border adjustments, and that enacting the proposal would indeed accomplish that objective, at least as compared with current law. I would therefore like to express a concern that both requiring and allowing taxpayers to sell their NOL carryforwards to each other could partly undermine that objective.

Now, one of the principal tenets of any consumption tax is that the government be willing to write an annual check to taxpayers whose deductions exceed their gross incomes under the rules in effect. That is how a VAT works, and that is therefore how refunds are implemented under a VAT. Thus, suppose a U.K. exporter purchases goods from a U.K. retailer for £100 and sells them to a U.S. buyer for £100. Assuming that VAT is imposed at a rate of 20 percent, the U.K. exporter receives a check from the Inland Revenue for £20 that purports to be "refunding" the £20 of VAT that has previously been paid on the goods by the U.K. retailer and its previous suppliers (because VAT is not imposed on goods that are ultimately exported).²⁰

In a subtraction-style VAT (which is more analogous to what the blueprint is proposing), the U.K. supplier would not initially receive a £20 refund but would rather be entitled to exclude from income the £100 of proceeds that it received from the sale of the goods to the U.S. buyer. Nevertheless, it would still be entitled to deduct the £100 that it paid to the U.K. retailer. As a result, it would have a tax loss of £100. But here the government would step in and mail the exporter a check for £20, because annual net losses are refunded to the taxpayer at the relevant tax rate.

The blueprint is not proposing to mail any such check to taxpayers. Rather, the blueprint is proposing to require exporters to carry their net losses forward indefinitely. But these losses will never ultimately be used, because the exporter will presumably keep having more losses on account of continuing to exclude the proceeds of their sales of exports from income. In the language of the accountants, such losses would have a 100 percent "valuation allowance." Recall that the exporter is entitled to deduct the entire cost of producing the exported product notwithstanding that the proceeds of sale are completely excluded, and so the resulting net loss is very large, bearing no relation to economic income. Like the U.K. exporter above, therefore, a U.S. exporter would presumably be relying on monetization of these losses as its principal means of profit.

These exporters would therefore need to sell their losses to other taxpayers — most particularly to importers that had been denied any cost basis or deductions for the cost of their imports. Let us assume that the going rate for the sale of a net loss is 50 cents on the dollar, meaning that if the government should have written the exporter a check for $20 under a properly functioning border-adjusted tax, the exporter can instead sell its loss to an importer for $10. We immediately see that the benefit of border adjustments (in respect of encouraging exports and discouraging imports) is cut in half. For the importer is not really paying a $20 tax penalty when it imports $100 worth of goods, as it can purchase a replacement for its lost tax deduction for only $10. And likewise, the exporter is not really receiving a $20 export proceeds exclusion, because it can only sell the export proceeds exclusion for $10. And needless to say, in addition to the reduction in import-export incentives, a lot of unwanted transaction costs accompany the sale of the tax benefits, in the form of serious economic inefficiencies and endless fees for lawyers, accountants, and bankers.

As noted above, the export rules would hopefully incorporate a doctrine of place of supply designed to mitigate fraud. But that rule would not prevent taxpayers from "renting out" each other's import or export status. Thus, suppose an importer (Import Co.) that would otherwise be paying large amounts of U.S. tax (because it cannot deduct its own import purchases) enlists an exporter (Export Co.) that would otherwise have large unusable NOL carryforwards (because it constructs airplanes and exports most of them) to help it to import goods. Export Co. (at Import Co.'s direction) imports large amounts of Chinese consumer goods and then sells them to Import Co. As a result, Import Co. is entitled to a deduction for its purchase of the goods from Export Co., a U.S. person. (Export Co. doesn't mind losing a deduction for its import purchases, since its large NOL carryforwards more than cover its tax liability.)

Meanwhile, Import Co. purchases airplanes from Export Co. and sells them to Export Co.'s customers, deducting the purchase cost while excluding the sales proceeds. Import Co. suffers an economic loss from these transactions, because it effectively splits the value of its net tax deduction with Export Co. by purchasing the planes from Export Co. for more than it sells them for, but, of course, it has an after-tax gain. Might the net tax benefits arising from Import Co.'s export activities be disallowed under the economic substance doctrine, the substance-over-form doctrine, the recharacterization doctrine, or some other common law concept? If so, Import Co. and Export Co. would need to go further. Import Co. could, for example, (at Export Co.'s direction) invest directly in the construction of airplanes (renting out Export Co.'s facilities for the purpose) and export them to selected customers of Export Co., paying Export Co. net management fees that are substantially equivalent to the profit that Export Co. would earn from constructing and selling the airplanes itself. Import Co. would have a significant profit from these activities, as it would be investing a large amount of capital and getting what might approach a lender's return. If the relevant common law required that Import Co. take more risk, then Import Co. would do so, accepting the fact that sometimes it would make more money than expected, and sometimes lose more money than expected, and sometimes it wouldn't know who the buyer would be. This would merely increase the transaction costs associated with Import Co.'s purchase of tax benefits from Export Co. (to everyone's detriment, including the government), but the value of the export proceeds exclusion is so large that Export Co. would still have to sell its tax losses (or risk going out of business, in light of its competitors), and Import Co. would have to buy them (or risk going out of business, in light of its competitors).

Should the governing statutes nevertheless include rules designed to prevent these sorts of transfers, and if so, what would they be? Would they include a rule that "aggregated" Export Co.'s imports and sales to Import Co. into a direct import by Import Co., thereby disallowing Import Co.'s deduction? Would the application of such aggregation turn on timing? Or would the rule rather have something to do with the intent of the parties? And how would such a rule be applied and enforced? Even assuming that we did everything we could in this regard, how would we stop larger transactions, such as the merger of exporters with importers to form a single consolidated tax group? It's hard to see how the government could really stay ahead of the curve in terms of preventing exporters from monetizing their export subsidies.

Other Kinds of Transactions

Presumably we need to consider de novo what the rules would be for the sourcing of other kinds of transactions. For example, under a VAT, real estate is normally sourced by reference to its location, so that sellers neither pay VAT nor receive VAT credits in respect of real estate located outside the country. Such real estate is completely outside the VAT system.

What would this mean for U.S. businesses that recognized economic gains or losses from transactions in foreign real estate? Would they simply be exempt from all taxation in respect of their sales? Or would they be allowed to deduct their purchase prices and required to include their sale proceeds? And if the latter, would they be deemed to be "exporting" such real estate if they sold it to a foreign person, and "importing" it when they bought it from a foreign person?

It is worth considering, in this regard, all the potential for revenue loss where goods are exported, as discussed above. Real estate is the most valuable tangible property of all, and the ability to "export" it merely by passing title might therefore prove problematic. It seems to me, therefore, that even if real estate investments are taxed on a cash flow basis, the place of supply concept should dictate that it is never possible to export or import real estate, because real estate doesn't move.

As for financial transactions, which can be the largest, most malleable and most potentially problematic of all, I deal with them in a separate article and reach a similar conclusion.²¹ But a lot more work needs to be done thinking through from a practical perspective how border adjustments should apply to commodities transactions, private equity, and mergers and acquisitions.

Conclusion

I believe it makes sense to propose border adjustments, provided we do in fact think that they would serve to increase exports, reduce imports, and so increase economic growth. In doing so, however, there are several things that we should keep in mind. First, they are liable to lose some amount of revenue for practical reasons, and this should be taken into account in any relevant revenue estimate. Second, they are likely to involve a fair amount of complexity in practice, and we should not be under any misimpressions in this regard. Third, we should at least consider offering taxpayers a refund in respect of their net losses, weighing in this regard the benefits of increased economic efficacy and efficiency against the increased risk of tax fraud. Finally, in light of these points, proposing border adjustments cannot make sense if, based on currency theories, we no longer believe in the reasons set out in the blueprint for proposing them. But currency theories, however thoughtful, might not bear out fully in practice.