Policy Options for U.S. Export Taxation
CRS Report for Congress
Policy Options for U.S. Export Taxation
David L. Brumbaugh
Specialist in Public Finance
Government and Finance Division
Prior to 2004 the Extraterritorial Income (ETI) provisions of the U.S. tax code
provided a tax benefit for exporters. ETI, like the Foreign Sales Corporation (FSC)
provisions they replaced, was designed to boost U.S. exports. The European Union
(EU), however, filed complaints with the World Trade Organization (WTO), charging
that ETI and FSC were export subsidies and so violated the WTO agreements. In a
succession of rulings, the WTO upheld the EU’s charges, and authorized the EU to levy
retaliatory tariffs on U.S. goods. The EU began a phased-in application of tariffs in
March 2004. Several policy options for the United States were proposed in Congress,
including retaining ETI and accepting whatever tariffs would be imposed; overhauling
the U.S. method of taxing foreign-source income by adopting a “territorial” tax system;
reforming the U.S. tax system in general by adopting a consumption tax similar to the
value-added taxes levied by European countries; attempting negotiations with the EU;
and repealing the ETI provisions while replacing them with tax cuts elsewhere. The last
of these approaches was ultimately taken by Congress in October 2004 with the
enactment of the American Jobs Creation Act (AJCA; P.L. 108-357). Although the EU
lifted its tariffs in January 2005, it also lodged a WTO complaint against transition
provisions in AJCA, which had the effect of rescinding the ETI only gradually in certain
cases. In May 2006, a provision of the Tax Increase Prevention and Reconciliation Act
(P.L. 109-222) repealed the transition rules, apparently bringing an end to the long-
The Economics of the ETI Provisions
To understand the implications of the various policy options that were considered,
it is useful to first look briefly at the basic economic effects of ETI and its predecessor,
FSC. The FSC provisions allowed U.S. firms to exempt between 15% and 30% of export
income from taxation; ETI provides a tax benefit of roughly the same magnitude. The tax
benefits attract investment to the U.S. export sector and, as a consequence, U.S. exports
are probably higher than they would be without the provisions. Beyond this effect,
however, traditional economic theory indicates that the export benefits produce a set of
effects that are perhaps surprising to non-economists. First, because of exchange rate
adjustments, the FSC/ETI-induced increase in exports is diminished, and U.S. imports
Congressional Research Service ˜ The Library of Congress
also are increased; sales of U.S. import-competing industries thus fall. Economic theory
indicates that although the provisions thus increase the overall level of U.S. trade, they
do not change the balance of trade.1
This result is perhaps better seen by stepping back from the exchange rate
adjustments and recognizing that, when a country runs a trade deficit, it is using more
goods and services than it produces. To finance these purchases, it must necessarily
borrow from abroad by importing more foreign investment than it exports. A country’s
trade deficit, in other words, is matched by a deficit on capital account (investment
outflows minus investment inflows). And a country’s trade balance changes only if the
balance on capital account changes. Thus, if we assume that the export benefits do not
change the balance on capital account, they cannot change the trade balance.2
Perhaps more importantly, the export benefits also affect U.S. economic welfare.
Traditional economic analysis indicates that they reduce overall U.S. economic welfare
because at least part of the tax benefit is passed on to foreign consumers in the form of
lower prices. This price reduction can be viewed as a transfer of economic welfare from
U.S. taxpayers in general to foreign consumers. These effects, however, are probably
quite small. According to CRS estimates based on 1996 data, FSC increased the quantity
of U.S. exports by a range of 2-tenths of 1% to 4-tenths of 1% and increased the quantity
of imports by a range of 2-tenths of 1% to 3-tenths of 1%. The shift of economic welfare
to foreign consumers is equal to an estimated 1-tenth of 1% of exports. The impact on
the trade balance was probably negligible. FSC’s cost in terms of forgone tax revenues
is estimated by the Joint Committee on Taxation at $2.7 billion for FY2000. The ETI
provisions were estimated to reduce revenue by between $300 million and $400 million
per year beyond the cost of FSC.
If economic analysts are generally critical of tax benefits like FSC and ETI, support
for them can be found in the business community. A reason for the divergence in views
may be perspectives: economic analysis looks at the benefits’ impact from the perspective
of the economy as a whole, attempting to account for their full range of effects and
adjustments in all markets. Supporters of the provision, however, are frequently
businessmen whose exporting firms would likely face declining sales, profits, and
employment if provisions were to be eliminated. For economists, there is no denying that
FSC and ETI boost employment and increase incomes in certain sectors of the economy.
But they also result in contraction of other parts — for example, firms that compete with
imports — and transfer economic welfare to foreign consumers.
FSC and the ETI provisions have also been defended on the grounds that they
counter subsidies provided to foreign producers by their own governments. A purported
subsidy that is sometimes cited is the practice among European (and other) countries of
1 For a detailed analysis, see CRS Report RL30684, The Foreign Sales Corporation (FSC) Tax
Benefit for Exporting: WTO Issues and an Economic Analysis, by David L. Brumbaugh.
2 ETI probably reduces the net flow of U.S. investment abroad. By definition, an export incentive
encourages only domestic investment, since exports cannot be produced abroad. Also, if ETI’s
reduction in tax revenues is not offset by higher taxes elsewhere, then the provisions likely
increase the U.S. federal budget deficit and drive up real interest rates, attracting capital to the
United States. Thus, when capital flows are considered, ETI may increase the trade deficit.
rebating the value-added taxes (VATs) that would otherwise apply to export sales.
However, from an economic perspective such “border adjustments” do not distort trade.
As described above, the trade balance changes only when capital flows change, and border
adjustments by themselves have no impact on capital flows; exchange rate adjustments
neutralize any impact they might otherwise have on the balance of trade.
Retaining ETI. One option suggested by some was to do nothing: the United States
could leave the ETI provisions in place and face the EU’s tariffs. To roughly gauge the
size of tariffs’ impact, note that total U.S. exports to the EU were $165.1 billion in 2000.
The $4 billion in WTO-authorized tariffs is thus 2.4% of the value of exports to the EU
— a level of tariffs that might be characterized as small but noticeable. But even this
gauge of the tariffs’ likely impact is likely too high. The tariffs might diminish U.S.
exports to the EU, but as with ETI and FSC exchange rate adjustments would occur that
would mitigate the tariffs’ impact. The overall level of trade would shrink, but, as with
ETI and FSC, there would be no change in the trade balance.
The impact of the tariffs on U.S. economic welfare is uncertain but likely to be small.
In isolation, imposition of a tariff could reduce U.S. welfare by effecting a transfer from
U.S. exporters to foreign countries and by shrinking the level of trade below its most
efficient level. However, the tariffs would not be enacted in isolation; they would
supplement the ETI export tax benefit that, as described above, inefficiently expands the
level of trade and transfers welfare to foreign consumers. Thus, tariffs may mitigate part
of the adverse welfare effects of ETI but compound others. If U.S. economic welfare in
general might not be substantially effected by the tariffs, there would still be winners and
losers. Given the shrinkage of the U.S. export sector, some U.S. exporting firms and their
employees could be expected to be adversely affected. At the same time, U.S. industries
that compete with foreign imports would likely indirectly benefit from the tariffs. The
size of these welfare effects, however, would probably be quite small. But acceptance of
tariffs might also lead to a worsening of trade relations in general with the EU and a3
reduction in the economic gains that accompanies the free flow of trade.
Repealing ETI. A second alternative was to repeal ETI but make no other basic
changes in the U.S. tax system. As with tariffs, this would trigger exchange rate effects
that would result in a contraction of trade — both of imports and exports. In this
scenario, employment in export industries would likely fall compared to what would
otherwise occur, and employment would increase in industries that compete with imports.
Also, existing stockholders and owners of export firms would likely incur a windfall loss
as the value of their holdings decline from what would otherwise occur; employees of
ETI-using firms would also likely be made worse off as their employers shrink and reduce
their employment. However, other sectors of the economy would expand as the export
sector shrinks. Owners of firms in those sectors, for example, the import-competing
sector of the economy, would register windfall gains. In the aggregate and in the long run,
the gains to the economy from ETI’s repeal would outweigh the welfare losses, and
3 For further information on the imposition of tariffs, see CRS Report RS21742, European Trade
Retaliation: The FSC-ETI Case, by Raymond J. Ahearn.
overall U.S. economic welfare would be increased. This would be a mirror of the adverse
welfare effects of ETI outlined above.
Provide Alternative Tax Benefits. Transition effects were a concern regarding
ETI’s repeal: there was the possibility of a loss of jobs among the affected exporting
firms and windfall losses by stockholders. These transition effects could be offset by
transition relief; for example, a phased-out reduction in taxes linked to a firm’s past ETI
or Foreign Sales Corporation benefit. Such provisions could structured so as to be “infra-
marginal”; that is, they would be based on exporters’ past behavior, not their future level
of exports. (For example, a tax benefit could be provided whose size would be dependent
on the average level of exports over a specific period in the past.) Such benefits would
not be export subsidies in substantive terms. EU officials, however, have stated that they
would object to an ETI repeal that includes a transition period. (Indeed, upon approval
in October 2004, of an ETI measure including transition provisions, the EU lodged a
complaint with the WTO.)
Another type of alternative tax benefit is one that would be heavily used by firms that
used ETI, but is not linked to exports. (This is the option that was, in fact, contained in
P.L. 108-357.) For example, firms that use ETI necessarily invest in domestic production;
such investment is, by definition, necessary to produce exports. Accordingly, such firms
would likely find a tax benefit linked with domestic investment in general attractive.
Here, however, the tax benefit’s link with exports is severed by making the benefit
available beyond just the export sector, leading to either a large tax revenue loss or a
benefit smaller than ETI on a firm-by-firm basis.
In 2002, Chairman William Thomas of the House Ways and Means Committee
introduced a bill (H.R. 5095) that would simultaneously repeal the ETI provisions and
provide a range of tax reductions for foreign-source income. In doing so, the bill took the
approach of addressing “competitiveness” issues related to overseas business operations
while repealing the export benefit. There was some lack of congruity, however, between
firms who use ETI and those whose foreign operations would receive tax cuts. The bill
was not considered by the full House before the 107th Congress adjourned.
In the 108th Congress, Representatives Crane and Rangel and Senator Hollings
introduced H.R. 1769/S. 970, which proposed to phase out ETI while phasing-in a tax
deduction linked with income from domestic (but not foreign) production. Representative
Thomas introduced H.R. 2896, a proposal that is similar to his earlier proposal, but with
the addition of several domestic investment incentives. In the Senate, S. 1475 (Hatch)
also proposed to replace ETI with a mix of tax benefits for overseas and domestic
investment. In October 2003, the House Ways and Means Committee approved a
somewhat modified version of H.R. 2896, and the Senate Finance Committee approved
S. 1637, containing a different mix of tax benefits for overseas and domestic investment.
Several other Senate bills proposed to repeal ETI while adopting tax benefits exclusively
for domestic investment. These included S. 1688 (Rockefeller), S. 1922 (Smith and
Breaux), and S. 1964 (Stabenow). The full Senate approved a slightly modified version
of the Finance Committee bill on May 11, 2004. On June 17, the full House approved
H.R. 4520, a modified version of H.R. 2896. A conference agreement reconciling the
House and Senate measures was approved in October, and became P.L. 108-357.
Territorial Taxation. There are two basic ways a country can define its tax
jurisdiction in an international setting. First, it can assert the right to tax on the basis of
residence, and the right to tax its citizens and residents on their income, regardless of
where it is earned. Alternatively, a country can define its tax jurisdiction on the basis of
territory or the geographic source of income, and tax only income earned within its
borders. Under a territorial tax system a country’s taxes stop at the border; the foreign-
source income of the country’s citizens and residents is free of home-country taxes.
The current U.S. tax system is generally based on residence; absent special
exemptions, individuals who are U.S. residents and corporations that are chartered in the
United States are subject to U.S. tax on their worldwide income. In contrast, several
European countries, such as France, the Netherlands, and (in some circumstances)
Germany, have territorial systems. Some have argued that the United States could
resolve the FSC/ETI controversy and still maintain favorable tax treatment for exports by
adopting a territorial system of taxation and exempting foreign-source income from U.S.
tax. Indeed, the designers of the ETI provisions have argued that, because U.S. exporters
can use the ETI benefit to exempt a certain amount of income from foreign operations
from tax, the ETI provisions are, in fact, territorial, and hence do not violate the WTO
agreements. (The WTO rulings, however, held otherwise.)
The likely impact of a territorial system on exports is uncertain. If a territorial system
is to provide a tax benefit for exporting, at least part of export income must be assigned
a foreign rather than a domestic source for tax purposes, and international norms for
determining the source of income may pose problems for allocating export income to
foreign sources. The internationally accepted method of determining the source of
income is “arm’s length pricing,” which divides income by assigning hypothetical prices
to transactions between different parts of the same firm; the prices that are assigned are
those that would apply if the parts of the firm were unrelated. Under arm’s length pricing,
income from exports would possibly be required to be allocated to the part of a firm that
adds value to an export, and thus little, if any, export income would be allocated to a
firm’s foreign branches or subsidiaries. Thus, little export income would be exempt under
a territorial system unless income allocation rules that depart from arm’s length pricing
were adopted. Given this requirement, the debate over export subsidies might shift to
controversies about methods for allocating income and “arm’s length pricing.”
Adoption of a territorial system has implications beyond exports. Depending on the
particular form a territorial system takes, it might increase the tax incentive for U.S. firms
to invest abroad rather than the United States — an incentive that would potentially
reduce economic efficiency and U.S. economic welfare. At the same time, at least one
analysis has found that if some restrictions on interest tax deductions are enacted along4
with a territorial system, capital would actually flow into the United States from abroad.
Consumption Taxes. The topic of consumption taxes arose in the FSC/ETI
controversy in the form of a debate over the value-added taxes (VATs, a form of
consumption tax) widely used in Europe and elsewhere. The U.S. export benefits have
occasionally been defended on the grounds that they counter tax subsidies inherent in the
4 Harry Grubert and John Mutti, Taxing International Business Income: Dividend Exemption
Versus the Current System (Washington: American Enterprise Institute, 2001), 67 pp.
practice among VAT-imposing) countries of rebating the VAT that would otherwise
apply to their countries’ export sales. The rebates work as follows: under a VAT, tax is
generally applied at each stage of production, to the value-added by that particular stage.
When a firm in a VAT-imposing country makes an export sale, it receives a rebate of all
the VAT that has been paid with respect to the good at every level of production. At the
same time, when a VAT-imposing country imports an item, it generally applies its VAT
to the full value of the import.
Economists have long held that such “border adjustments” do not, in fact, help a
country improve its balance of trade. In a manner similar to the exchange rate
adjustments that occur in response to ETI and FSC, exchange rate adjustments mitigate
any increase the rebates may cause in the exports while at the same time causing an
increase in imports. While VAT-imposing countries rebate the tax on exports, they also
impose it on imports. In a manner that mirrors the exchange rate effects of the export tax-
rebates, the imposition of VAT on imports, in isolation, causes no improvement in the
balance of trade, but reduces exports and imports alike. In fact, when the combined
impact of export tax-rebates and the imposition of VAT on imports can be considered
roughly offsetting, assuming the VAT applies at the same rate to all commodities. As a
consequence, the distorting impact of the VAT on trade is generally small.
A consumption tax may affect the balance of trade through its impact on capital
flows. If a consumption tax triggers capital inflows, the U.S. trade deficit would increase;
if it causes a net capital outflow, the deficit would shrink. The impact of a consumption
tax on capital flows, however, is uncertain. There would likely be a net inflow to the
United States of equity capital in response to the lower business tax burden. On the other
hand, debt-financed investment would not be tax-favored as under current law; it would
no longer be deductible. There may as a result be a net outflow of debt.5 In any event,
as with adopting a territorial system, the impact of adopting a consumption tax would be
considerably broader than any effect it may have on exports and imports. Much of the
debate over the economic effects of replacing the current tax system with a consumption
tax has centered on its impact on tax fairness in general and long-run economic growth.
Negotiations. Some suggested that an alternative approach to the ETI dispute
might be negotiations, while others argued that such an approach would trigger the EU’s
retaliatory tariffs. It might also be pointed out that a negotiated settlement could not be
strictly bilateral, between the EU and the United States; given the WTO panel rulings,
other countries might file successful complaints against ETI even if a settlement is
reached between the EU and the United States. Accordingly, a long-run negotiated
solution would probably have to include an alteration of the WTO rules themselves.
5 Harry Grubert and T. Scott Newlon, Taxing Consumption in a Global Economy (Washington:
American Enterprise Institute, 1997), 52 pp.