The tax treatment of international trade is often viewed as a challenge for the X tax. The primary issue is whether the X tax should be an origin-based tax on domestic production or whether a border adjustment should be adopted to transform it into a destination-based tax on domestic consumption. We conclude that the X tax should be origin based to avoid a large transfer of wealth from Americans to foreigners.
The current income tax system is origin based, meaning that business and personal income taxes are imposed on production inside the United States, including production of goods that are exported to other countries. Conversely, business and personal income taxes are generally not imposed on production outside the United States, even when the resulting goods are imported into the United States.
Early proposals for the flat tax, which, as we discussed in chapter 2, led to the development of the X tax, assumed that the tax would be imposed on an origin basis. The use of the origin basis would promote continuity with the current income tax system.
Because the X tax is a modified VAT, though, one could argue that it, like all of the world's VATs, should be destination based. A destination-based X tax would apply to consumption inside the United States. To place the tax on a destination basis would require a border adjustment, consisting of a tax on imports and a tax rebate for exports. The adjustment would tax imported consumer goods at the 38.8 percent business cash-flow tax rate, deny firms a deduction for imported business purchases against the 38.8 percent cash flow tax, and provide a full 38.8 percent tax rebate on exports. Exporting firms would still deduct all their (domestic) business purchases and their wage payments, even though they would pay no tax on their overseas sales.18
Auerbach (2010) proposes a "modern corporate tax" that is almost a destination-based cash-flow tax. This proposal is similar to the X tax in that the corporate income tax is transformed into a business cash-flow tax where businesses can deduct all purchases from other businesses and employee compensation. Because the plan leaves the individual income tax unchanged, employee wages plus investor returns continue to be taxed at the household level. In his proposal, all cross-border transactions would be ignored by excluding sales abroad from the tax base and not allowing a deduction for purchases or investment abroad. The proposal appears to deviate from a destination-based tax in one respect: foreign consumer purchases are excluded from the tax base, rather than taxed when entering the United States, an approach apparently adopted to preserve the simplicity of ignoring all cross-border transactions.
Many discussions of the X tax argue, or simply assume, that it would be preferable for it to be border adjusted and destination based. Starting from this premise, some analysts see international trade rules as a problem for the X tax. Although the rules allow "direct" taxes to be border adjusted, they prohibit border adjustments for "indirect" taxes. The credit-invoice VATs that prevail in most other countries clearly qualify as direct taxes, but the weight of authority indicates that a flat tax or X tax is likely to be classified as an indirect tax that cannot be border adjusted. Hufbauer (1996, 47–61, 69–70) provides a thorough discussion of these rules.
If desired, it might be possible to overcome this perceived problem. As Weisbach (2000, 623) points out, other nations would probably find it difficult to maintain the current rules if the United States forcefully pressed for modification, particularly because the direct/indirect distinction lacks economic foundation. We conclude, however, that it would be preferable not to border adjust the X tax, which removes the trade rules as a problem.
To begin, we reject an argument that seeks to impose the destination basis as a matter of vocabulary. The argument observes that a destination basis is a true consumption tax, whereas an origin tax is a tax on consumption plus net exports. A destination basis may therefore seem more faithful to this book's call for a "consumption tax." But vocabulary should not control policy. As we explained in chapter 1, the economic advantage of a tax on consumption is that it eliminates the marginal tax burden on new investment. Because an origin-based X tax would retain this advantage, as shown below, it is also a "consumption tax" in the economic sense. The border adjustment question should be decided on the basis of economic consequences, not terminology.
The most common argument for border adjustment is that it would make U.S. producers more "competitive," thereby permanently boosting exports and reducing imports and "improving" the trade balance.19 As we explain below, this argument is invalid because the competitiveness effect would be offset by movements in the real exchange rate. Moreover, a permanent reduction in the trade deficit, even if it could be achieved, would not be an economic improvement.
To be clear, we argue below that border adjustments have no competitiveness effect—positive or negative. If competitiveness were the only relevant factor, we would be neutral or agnostic with respect to the desirability of a border adjustment. But we also explain below that border adjustments would involve a significant transfer of wealth from Americans to foreigners, which, in our view, decisively tips the balance against a border adjustment.
In analyzing border adjustments, economists often emphasize a simple model that makes a number of restrictive assumptions. The model assumes that the border adjustment applies uniformly to all goods and services; the prime example is the addition of a border adjustment to a comprehensive uniform-rate VAT that previously lacked such an adjustment. The model also assumes that the border adjustment is perfectly implemented, with no enforcement problems. It is also simplest to begin with a single-period model in which there is no passage of time. This discussion draws on Viard (2004a, 2008, 2009a).
Irrelevance in Single-Period Model. In this simple model, as set forth below, a border adjustment does not make domestic producers more competitive. Indeed, the model actually yields the more dramatic conclusion that the border adjustment has no real economic effects at all; origin and destination taxes are equivalent. (As will be seen, more realistic models reveal that border adjustment has some real effects, but not the commonly imagined competitiveness effects.) The equivalence of origin-based and destination-based tax systems was first noted by the Tinbergen Report (European Coal and Steel Community, High Authority, 1953). Numerous economists have reiterated this point over the last six decades.20
Because taxing an activity tends to discourage it and subsidizing an activity tends to encourage it, it seems plausible to argue that an import tax reduces imports and that an export subsidy increases exports. The argument is invalid, however, because it ignores the budget constraint linking imports and exports. In the single-period model now being considered, the budget constraint simply requires that imports equal exports.
The terms of trade at which a country buys and sells with respect to the world adjust to keep exports equal to imports in the single-period model. For countries with flexible exchange rates, changes in the terms of trade generally occur through exchange-rate fluctuations. (The "Border Adjustment with Fixed Exchange Rates" box (pages 116–17) examines the case of fixed exchange rates.) The adoption of a border adjustment by the United States triggers an exchange-rate movement that offsets the perceived change in competitiveness.
Suppose that the United States initially has a 38.8 percent X tax with no border adjustment and that it exports 100 goods to Europe at a price of $10 apiece and imports 100 goods from Europe at a price of $10 each. Trade is in balance, with exports and imports both equal to $1,000. Suppose that the initial exchange rate is 1 euro per dollar.
What happens if the United States border-adjusts, imposing a 38.8 percent tax on imports and paying a 38.8 percent rebate on exports? At first glance, the change appears to increase exports and reduce imports. Before the adjustment, American producers charged $10 for each good that they exported to Europe, but cleared only $6.12 after tax. Because exports are now tax free, American exporters need charge only $6.12 to clear $6.12. If the exchange rate remained unchanged, goods from the United States would sell in Europe for $6.12 rather than 10 euros, giving American sellers a competitive edge. Conversely, European imports are now subject to a 38.8 percent U.S. tax, so European sellers must charge 16.34 euros to clear 10 euros (they pay tax of 6.34 euros, 38.8 percent of 16.34 euros). If the exchange rate remained unchanged, goods from Europe would sell in the United States for $16.34 rather than $10, putting European sellers at a disadvantage. In equilibrium, though, these outcomes cannot occur because they violate the budget constraint requiring that imports remain equal to exports.
Instead, the dollar appreciates to 1.634 euros; in other words, the euro declines to 0.612 dollars. Although American producers can now sell goods in Europe for $6.12, that translates into an unchanged price of 10 euros at the new exchange rate. Although foreign producers must now charge 16.34 euros for goods sold in America, that price translates into an unchanged price of $10 at the new exchange rate. Trade flows are unaffected.
The border adjustment also leaves after-tax real wages and other producer incomes unchanged. The border adjustment raises the prices paid by American consumers, as expressed in euros, by 63.4 percent, but also increases the after-tax incomes received by American producers, as expressed in euros, by 63.4 percent. As measured in foreign currency, destination-based taxes raise consumer prices, while origin-based taxes lower disposable income, but both taxes cause the same reduction in real disposable income.
Irrelevance in Simple Multiperiod Model. The fact that the economy lasts for many periods does not, by itself, make border adjustment relevant. The equality between exports and imports continues to hold, although in a slightly different form. For any household or nation, purchases must equal sales in present discounted value over their lives. Purchases are financed by the proceeds of sales, and sales are made to finance desired purchases. For any nation, the present discounted value of exports equals the present discounted value of imports over its entire history. Any policy that permanently reduces imports also reduces exports; any policy that permanently increases exports also increases imports. The quest to permanently increase exports and reduce imports is futile, whether pursued through border adjustment or otherwise. As in the single-period model, the adoption of a border adjustment that applies to the country's entire history causes an exchange-rate appreciation that leaves exports and imports unchanged in each year.
So, a border adjustment would not permanently increase exports and reduce imports. This is a good thing, too, because such an outcome would permanently reduce American living standards, as Viard (2008) emphasizes. We would forever send more goods and services, produced by our toil and natural resources, to foreign consumers while forever receiving fewer goods and services for our own enjoyment in return. Imports are the gain from trade, and exports are the cost of trade; we give up exports to obtain imports. The desire to increase exports and reduce imports reflects the mercantilist fallacy that Adam Smith (1976, book IV, chapter 1) condemned in 1776.
The simple form of the present-value equality rests on the assumption that no cross-border investments have above-normal returns. Moreover, the assumption that the border adjustment applies to the country's entire history implies that no cross-border investments are in place when the adjustment is adopted. As discussed below, the two most dramatic effects of border adjustment concern the treatment of above-normal returns (which is intertwined with the transfer pricing issue) and the transitional impact on existing cross-border investments.
For now, we note that the border adjustment has no impact on the choice of where to locate a new investment that yields only the marginal rate of return. To be sure, a destination-based tax applies to investments made abroad by Americans but not to investments made in the United States by foreigners, while an origin-based tax applies to the latter but not the former. But these distinctions make no difference for new investments earning the marginal rate of return because, as we explained in chapters 1 and 2, the X tax imposes a zero effective tax rate on such investments. Being subject to the X tax therefore poses no disincentives for such investments.
Still another way to explain why border adjustments have no effect is to analogize them to traditional and Roth IRAs. From a national perspective, border adjustment resembles a traditional IRA, as the deduction for exports is similar to the upfront deduction for contributions, while the tax on imports is similar to the tax paid on distributions. In contrast, an origin-based tax resembles a Roth IRA, in which individuals receive no deduction for contributions and pay no tax on distributions; an origin-based tax offers no deduction for exports and imposes no tax on imports. As we explained in chapter 2, the value of the upfront tax deduction exactly offsets the tax on the distribution in present value, provided that the tax rate remains constant over time.
Some Cases in Which Border Adjustment Is Relevant. Before discussing the critical issues of transitional impact and above-normal returns, we briefly examine some other, generally less significant, factors that can cause border adjustment to have real effects.
A border adjustment has real effects if the tax that is being border adjusted does not apply uniformly to all value added. Consider, for example, a value-added tax that exempts food and assume that food is readily tradable across international borders. Border adjusting such a tax does improve the competitiveness of the producers who are subject to the tax (those outside the food industry) and does increase their net exports. But the border adjustment reduces the competitiveness of food producers, who receive no direct benefit from the border adjustment of a tax from which they are exempt while facing competitive pressures from the resulting appreciation of the exchange rate. Although it is impossible to simultaneously make all industries more competitive, it is possible to make some more competitive while making others less competitive, and one way to do that is to border adjust a tax that applies to only some industries.
When tax rates differ across goods and services that are readily tradable across international frontiers, the differentials are primarily reflected in consumer prices under destination-based taxes and are primarily reflected in producer prices under origin-based taxes. In general, tax rate differentials tend to produce more sensible results under destination-based taxes. For example, preferential rates for food are intended to aid consumers, not producers, of food, and a tobacco excise tax is intended to reduce the country's consumption, not its production, of tobacco. To obtain the desired results, such policies should be adopted on a destination basis.
Despite initial appearances, however, this consideration actually supports putting the X tax on an origin basis. The fact that tax rate differentials will not accomplish their intended goals under an origin-based tax is likely to be clear to policy makers, which should dissuade them from adopting such preferences. That outcome is desirable because it allows the business cash flow tax to remain uniform across different types of goods. If a tax preference for medical services is desired, for example, it can and should be provided to consumers of such services through a deduction or credit on the household wage tax returns, where it can be calibrated to the household's circumstances; it should not be provided through a tax reduction for the firms producing medical services. If a tax on tobacco consumption is desired, it can be imposed as a separate destination-based excise tax (which is how today's tobacco tax is imposed) rather than being made part of the X tax system.
The present-value equality between import taxes and export rebates breaks down if the border adjustment faces compliance problems, either the evasion of tax on imports or the filing of refund claims for nonexistent exports. Import tax evasion is likely to become a greater challenge as Internet transactions become more numerous. Fraudulent export refund claims have been a significant problem with European border adjustments, as discussed by Ernst and Young (2010, 2011), Tait (1988, 313–14), and Keen and Smith (2006, 870–72). Also, because the equality of import taxes and export rebates arises from residents' budget constraints, it breaks down if residents are able to avoid the tax by consuming abroad, either as tourists or as retirees in foreign jurisdictions.21 On the whole, these compliance issues appear to offer support for the origin basis.
But two significant differences remain to be considered: transitional wealth effects and the treatment of transfer pricing and above-normal returns. As shown below, the former consideration provides a powerful argument for origin-based taxes, and the latter consideration supports destination-based taxes. We conclude with a recommendation for the origin basis and a proposal to address the potential problems posed by transfer pricing and above-normal returns.
A border adjustment has important effects on revenue and domestic and foreign wealth, due to cross-border asset holdings at the time the adjustment is introduced. The present-value equality between exports and imports must be modified, as follows, at any point in the midstream of a nation's history: The present discounted value of future exports equals the present discounted value of future imports plus the country's net foreign debt (or minus its net foreign assets).
The revenue effects of border adjustment depend on net cross-border asset holdings at the time of reform. According to Bureau of Economic Analysis (BEA) data reported by Nguyen (2011), the United States had a net external debt of $2.5 trillion at the end of 2010. As a result, the present discounted value of future U.S. exports exceeds the present discounted value of future U.S. imports by $2.5 trillion. In other words, today's trade deficits must eventually be followed by trade surpluses, with a $2.5 trillion larger present value, to service the external debt. Under simple assumptions, a 38.8 percent X tax with a border adjustment, relative to a 38.8 percent X tax without a border adjustment, lowers revenue by $0.97 trillion, in present discounted value. This is true even though the border adjustment raises revenue in the short run, due to current trade deficits.
But the revenue impact is the least of the story. A border adjustment also has profound transitional effects on asset values, effects that have drawn surprisingly little attention. The border adjustment brings into the tax base the consumption of Americans financed by their existing foreign assets and removes from the tax base the consumption of foreigners financed by their existing American assets. As a result, Americans' holdings of foreign assets decline in real value, while foreigners' holdings of American assets rise in real value. The appreciation of the dollar reduces the dollar-value of Americans' foreign assets and increases the foreign-currency value of foreigners' American assets.
For two reasons, these wealth effects have more far-reaching policy implications than the revenue impact. First, the revenue impact is merely a transfer among Americans. If the border adjustment causes the tax to raise $0.97 trillion less than it otherwise would, the Treasury suffers a $0.97 trillion loss, but American taxpayers enjoy a $0.97 trillion gain. Changing the tax rate can undo these effects. In contrast, the wealth transfer to foreigners is a net loss for the American people. Second, the wealth effects are vastly larger than the revenue impact. The reason is that, as Viard (2004b; 2008) and Auerbach (2007; 2008) note, the wealth effects depend on the gross holdings of American assets by foreigners and foreign assets by Americans, while the revenue effect depends on the much smaller net difference of those holdings.
The BEA data reveal that Americans held $20.3 trillion of foreign assets at the end of 2010, and foreigners held $22.8 trillion of American assets (the difference is the $2.5 trillion U.S. net external debt, mentioned above). A stylized calculation suggests that a 38.8 percent X tax with a border adjustment, relative to an identical tax without a border adjustment, imposes a $7.88 trillion heavier tax burden on American consumers and an $8.85 trillion lighter tax burden on foreign consumers, in present discounted value. (The difference is the $0.97 trillion net revenue loss, mentioned above.) A wealth transfer to foreigners of $8.85 trillion, roughly seven months of GDP, is a staggering burden on Americans. Viard (2009a) and Auerbach (2007, 45–46; 2008, 19–20) provide similar calculations with older data and different tax rates.
Transition relief is likely to reduce the wealth transfer to some extent. If the tax is not border adjusted, transition relief is likely to be provided at the firm level, as we will recommend in chapter 8, easing the burden on firms' foreign stockholders. In contrast, transition relief for a border-adjusted tax would presumably be provided to Americans at the household level, as we will discuss in chapter 10 for a destination-based VAT, with no benefit to foreigners. Even so, the wealth shift would still be several trillion dollars. Because this transfer would dwarf the efficiency gain from tax reform, the move to a border-adjusted consumption tax would actually be a gift to the world, rather than a gain for the United States.
This discussion reveals that other nations' VATs are gifts to the United States, reflecting the fact that these nations have gone astray in the pursuit of mercantilism. It is ironic that international trade rules thwart the adoption of a border adjustment that would actually benefit the world at Americans' expense. This issue highlights the insidious effects of the competitiveness fallacy. Proponents of the fallacy imagine that a U.S. border adjustment would tax the foreigners who sell goods to us, when it would actually tax the American consumers who buy these products.
Although transition effects provide a strong argument against border adjustment, we must now consider the countervailing implications of above-normal returns and transfer pricing.
The equivalence of origin and destination taxes breaks down in the presence of above-normal returns, just as the equivalence of conventional and Roth IRAs breaks down in the presence of such returns.
The Problem Facing Origin-Based Taxes. As we emphasized in chapters 2 and 3, consumption taxes, including the X tax, impose tax on above-normal returns, such as returns from innovation or the exercise of market power, even though they do not tax marginal returns to investment. The presence or absence of a border adjustment can affect which above-normal returns are taxed and thereby alter incentives for generating above-normal returns in different places, as Avi-Yonah (2010) and others have noted.
Suppose, for example, that an American has an idea that will give rise to $1 million of profits. A destination-based tax is neutral with respect to the location at which the idea is developed, because it taxes consumption done by Americans. No matter where the American develops the idea, his consumption will be taxed under a destination-based tax. In contrast, an origin-based tax appears to have the potential to deter the development of the idea inside the United States, because the tax applies to investment inside the United States but not investment abroad. Moreover, the disincentive may be more severe than under the current tax system under our assumed 38.8 percent cash-flow tax rate, which is higher than the current corporate income tax rate.
As Bradford (2004, 19–20) and Grubert and Newlon (1997, 18–20) explain, the origin-based tax need not always have these dire effects. To begin, there is no disincentive if the American sells or leases his idea to an unrelated foreign firm for $1 million. The American's $1 million receipt is from the export of an intangible asset and is therefore subject to the origin-based tax. After making this payment to the American, the unrelated foreign firm earns only normal returns on its investment. Because the above-normal return is generated by the American's idea, not by the development of the idea overseas, it is subject to the origin-based U.S. tax.
In principle, nothing changes if the American instead develops the idea through a wholly owned foreign firm. The wholly owned firm should make the same $1 million payment to the American, and this amount should still be taxed as an export receipt.
But the transfer pricing problem, the distinctive bane of origin-based taxes, comes to the fore in this situation. Because the American owns the firm, the price he "charges" it is merely an accounting entry rather than a real payment, leaving him free to choose a value that minimizes tax liability. Suppose, for example, that the American charges his wholly owned firm a mere $700,000 for the idea, even though it is actually worth $1 million. Then, the firm appears to generate a $300,000 above-normal return, which escapes tax due to its location abroad. The American ultimately receives this return in dividends or other profit distributions, but these payments are financial flows that are not taxed by the real-based X tax. In cases in which the correct market value cannot be objectively determined, as will often be true in cases involving above-normal returns, the origin-based tax offers an incentive to relocate investment abroad and to misstate prices to prevent the proper amount of tax from being collected in the United States.
In contrast, border adjustment eliminates transfer-pricing problems between firms. Under a destination-based tax, only sales to domestic consumers, whether by domestic or foreign firms, are ultimately taxed. Transactions between firms wash out, with the selling firm's inclusion offset by the buying firm's deduction. In principle, a destination-based tax need not even distinguish domestic and foreign firms, because the net tax liability depends only on the consumer's identity.22
Transfer pricing problems are not unique to an origin-based X tax; they arise today under the origin-based income tax. As Weisbach (2000, 642) notes, the transfer pricing regime of an X tax need not be much more strictly enforced than that employed under the current income tax system. In either case, though, the presence of above-normal returns suggests that an origin-based tax can have undesirable effects on the location of innovation.
It is desirable that the X tax avoid the complexity and administrative costs associated with transfer pricing enforcement under the income tax system. Fortunately, Bradford (2003; 2004, 34–45) proposed a solution to this problem.
Bradford's Solution. As we explained in chapter 2, the basic design of the X tax is real based. By ignoring financial transactions, the X tax generally achieves greater simplicity than the PET, which uses a real-plus-financial approach. As we also noted in chapter 2, though, it is possible, and sometimes desirable, to depart from the basic X tax design in particular areas. Bradford's transfer-pricing solution applies an R+F cash-flow method to cross-border transactions between related parties, the same approach that we recommended in chapter 6 for transactions involving financial intermediation.
The Bradford proposal calls for U.S. firms and households dealing with foreign related parties to include all inflows, both real and financial, from the related parties and to deduct all outflows, both real and financial. This procedure accurately captures the effect of the transactions with the related parties on the consumption of the U.S. firm or household. In the example given above, if the American sets a $700,000 price on the $1 million idea and therefore avoids tax on $300,000 of export receipt, he will be taxed on an additional $300,000 (in expected present discounted value) of dividend inflows from his wholly owned firm.
Ex post, the R+F cash-flow method may not correctly measure the net gain from the taxpayer's real transactions with the foreign related parties. Suppose, for example, that the wholly owned firm invests its $300,000 profit in a risky option that, one year later, has a 50 percent chance of paying nothing and a 50 percent chance of paying $600,000 and that the firm then paid a dividend to the American equal to the proceeds. The ex post dividend payment, either zero or $600,000, has no relationship to the $300,000 net payoff from the idea (above and beyond the $700,000 already booked by the American).
From the relevant ex ante perspective, though, there is no problem. If a security offering this pattern of payoffs sells for $300,000, the market value of these uncertain payoffs must be $300,000. (With risk aversion, the market value may differ from the expected value; the method continues to work, as we explained in chapter 6.) In other words, the deviation of the ex post payoff from the purchase price must have a market value of zero. So, the Bradford solution actually taxes the $300,000 real payoff plus an uncertain component (which ultimately turns out to be either negative $300,000 or positive $300,000) that has zero market value. This approach gives the American the correct ex ante incentives at the time he decides where to develop his idea. This example illustrates, yet again, the principle we discussed in the "Zero Revenue from Taxation of Risky Returns" box in chapter 3 (pages 51–52), namely, that the taxation of risky returns has no real effects on the government or on taxpayers.
The R+F cash-flow method faces the same challenges in this context that we discussed in chapter 6. In particular, it would be necessary to adopt transition arrangements similar to those we recommended in that chapter.
The United States currently maintains withholding taxes on dividends and some forms of interest paid to foreign persons, which are typically reduced as part of the bilateral treaties that the United States has with more than fifty nations. The United States should maintain its withholding taxes as a bargaining chip while seeking their bilateral elimination in treaty negotiations.
Today, the five overseas U.S. possessions (Puerto Rico, Northern Marianas, Guam, the U.S. Virgin Islands, and American Samoa) are largely, though not completely, exempt from U.S. corporate and individual income taxes. Because exempting these areas from the X tax minimizes disruption and maximizes simplicity, the geographical scope of the X tax should be limited to the fifty states and the District of Columbia. Of course, the tax treatment of the possessions may change if and when their political status changes.
Despite broad political support for the destination basis, the origin basis represents better policy because it avoids a massive transfer of wealth from Americans to foreigners with no loss of "competitiveness." Transfer pricing problems can be addressed by applying a real-plus-financial cash-flow method to transactions between U.S. and foreign related parties.
BOX
BORDER ADJUSTMENT WITH FIXED EXCHANGE RATES
Graetz (2008, 81) and others argue that the analysis in the text breaks down when a foreign country pegs its currency against the U.S. dollar. The argument holds that the exchange rate adjustment relied on in the text cannot occur when the foreign currency is pegged. In reality, though, the analysis in the text holds with little change when the exchange rate is pegged, as Viard (2009a) emphasizes.
A pegged exchange rate remains fixed only until the pegging country changes it. If a foreign government that pegs its currency at ten units to the dollar observes a 38.8 percent border adjustment by the United States, it can immediately repeg its currency at 13.88 units to the dollar. That simple step maintains an unaltered real equilibrium with no change in exports or imports, thereby preserving the real advantages that the government perceives from its decision to peg. Unless the foreign government is metaphysically attached to the number ten, it has no apparent reason not to repeg. In short, a border adjustment is an unpromising strategy to alter trade patterns between the United States and a pegging country because it can be defeated by the pegging government's choice of a number.
Consider the impact on traffic speeds if the United States switched from the English to the metric system of measurement. If a state "pegs" its highway speed limit at 65, would a switch to the metric system impose a crippling slowdown on the highways by reducing the allowable pace of travel within the state from 65 miles per hour to 65 kilometers (40.39 miles) per hour? Surely not. The state's choice of a 65-mile-per-hour speed limit reflects a desired balance between traffic safety and transportation efficiency, not a metaphysical attachment to the number 65. After the switch to kilometers, the state would "repeg" its speed limit, now expressed in kilometers per hour, to a number around 104.61, thereby maintaining the desired balance and avoiding any change in traffic speeds.
The only jurisdictions that cannot repeg are those that use the U.S. dollar as their own currency, for whom the exchange rate cannot deviate from one-to-one. Aside from the five overseas U.S. possessions, the British Virgin Islands, East Timor, Ecuador, the Marshall Islands, Palau, Panama, and Turks and Caicos Islands currently use the U.S. dollar. If the foreign government cannot repeg, economic balance is restored through a decline in the pegging country's price level. A 38.8 percent border adjustment would cause prices and nominal wages in these jurisdictions to fall 38.8 percent, thereby eliminating any competitive advantage for American producers.
To be sure, prices and nominal wages do not move as quickly as exchange rates. Before the price decline is complete, the border adjustment does indeed increase U.S. exports and reduce U.S. imports. But the border adjustment thereafter reduces U.S. exports and increases U.S. imports to maintain the present-value equality between exports and imports. The border adjustment therefore delivers only a temporary "improvement," followed by a long-run "deterioration," of the U.S. trade balance, and it does so only for U.S. trade with a small number of tiny dollarized jurisdictions. In the process, it inflicts a wrenching price decline on those jurisdictions, which are either our own overseas possessions, home to four million U.S. citizens and nationals, or friendly foreign jurisdictions whose use of the U.S. dollar brings in revenue for our Federal Reserve.