Previous articleNext article FreeCommentAlan J. AuerbachAlan J. AuerbachUniversity of California, Berkeley and NBER Search for more articles by this author University of California, Berkeley and NBERPDFPDF PLUSFull Text Add to favoritesDownload CitationTrack CitationsPermissionsReprints Share onFacebookTwitterLinked InRedditEmailQR Code SectionsMoreIntroductionThe tax reform process that culminated in the December 2017 enactment of the Tax Cuts and Jobs Act followed an unusual pattern regarding business tax reform. In particular, the original proposal, the “Blueprint” put forward by Republicans in the House of Representatives in June 2016 (Tax Reform Task Force 2016) called for the adoption of an approach that, at the time, was unfamiliar to many in the economics profession, a destination-based cash-flow tax (DBCFT). The DBCFT would have represented a sharp break from current policy, and the general lack of familiarity with it led many business leaders, policy makers, and economists to misinterpret its aims, characteristics, and properties. The paper by Omar Barbiero, Emmanuel Farhi, Gita Gopinath, and Oleg Itskhoki represents part of a small and growing literature seeking to analyze the DBCFT, or at least one of its key components: a border tax adjustment on imports and exports. In reading the paper, one is reminded of the advantages of following the more standard tax reform approach of analyzing new proposals before voting on them. This is not to say that I agree with all the paper’s modeling assumptions or conclusions, because I do not. But without such concrete analysis, it is difficult to identify key points of professional disagreement and, more importantly, to try to resolve them.The paper analyzes the short-run macroeconomic effects of adopting border tax adjustments on their own, although this is not what was being proposed. However, this is equivalent in the model to analyzing adoption of a full DBCFT, that is, a “source-based” cash-flow tax—a tax on domestic producers’ cash flows—plus border adjustment that removes tax on exports and imposes tax on imports. This equivalence follows because in the model, a cash-flow tax without border adjustment is a nondistortionary tax on pure profits—a lump-sum tax that would then be rebated via an equal-size lump-sum transfer. In particular, the channels through which companies might respond to a source-based cash-flow tax by shifting profitable operations or simply reported profits to lower-tax countries (as discussed by Auerbach and Devereux [2018]) are absent. Thus, only the border-adjustment component of the DBCFT has real effects in the model. The paper contrasts the effects of the border adjustment/DBCFT to those of a traditional value-added tax (VAT), which is equivalent to a DBCFT plus a payroll tax at the same rate, both in the model and in the real world.Why Consider the DBCFT?One thing the paper does not do is explain the motivation for proposing the DBCFT in the first place. Here, it is useful to distinguish the effects of the DBCFT from those of a more traditional cash-flow tax without border adjustment, which we might label simply a cash-flow tax. While the DBCFT is a relatively unfamiliar proposal, arguments in favor of the cash-flow tax are more common and of longer standing. A cash-flow tax even without border adjustment would represent a major change from the traditional approach to business taxation by eliminating the distinction between debt and equity finance, removing the tax wedge facing marginal investment decisions, and making tax administration and compliance simpler by not requiring the calculation of income. For reasons such as these, many economists have long argued in favor of business cash-flow taxation (see, e.g., Institute for Fiscal Studies 1978; Hall and Rabushka 1983).Arguments for imposing the cash-flow tax on a destination basis—for the DBCFT—are newer and less familiar.1 First, by eliminating any tax on the foreign-source income of US corporations, it would remove incentives for companies to use corporate inversions to relinquish US residence. Second, by making US tax liability independent of the prices charged for imports and exports, the DBCFT would eliminate the possibility of using internal transfer pricing to reduce US tax liability. The loss of interest deductibility would make it impossible to use the location of borrowing in the United States to shift profits elsewhere. Third, as a tax based on the location of sales, the DBCFT would impose no additional tax on the inframarginal profits from US production, encouraging companies to locate profitable activities in the United States. Fourth, as a tax only on the nonlabor component of value added, the DBCFT would be progressive, particularly because companies would have no incentive to move capital out of the United States to escape taxation. Finally, unlike other attempts to reform the treatment of multinational businesses, the DBCFT would be self-reinforcing among countries; the incentive for US adoption would not require cooperation with other countries, and other countries would have a stronger incentive to follow suit after US adoption. As to why the case for the DBCFT might be more compelling now than in the past, one need look no further than the changing nature of business activity. The increased importance of multinational companies and the stronger reliance on the use of intellectual property in the production process have made it easier for companies to shift production and profits around the globe and more difficult for tax authorities to pin the location of profits and production.The paper acknowledges that such potential advantages exist and sets them to the side, based on the idea that one can separate such advantages as being of a longer-run nature not relevant to the short-run macroeconomic effects that are its major focus. I will return to this issue below.In addition to these economic arguments for the DBCFT, at least two additional characteristics helped generate political appeal during the US debate. First, to many noneconomists for whom the Lerner symmetry theorem is not self-evident or even plausible, border adjustments appeared to be protectionist. Of course, only some saw this apparent proexport bias as an advantage. Second, during the 10-year budget window, the border adjustment—effectively a tax applied to the US trade deficit—was estimated to raise a substantial amount of money, well in excess of $1 trillion at the 20% tax rate initially proposed (Nunns et al. 2016). While this attribute would have made it much easier for legislators to “pay for” tax cuts, it was also the focus of sharp criticism from those who argued that one could not view such a revenue gain as permanent. The basic argument, as laid out in this paper, follows the standard long-run condition that the present value of a country’s trade deficits equals the initial value of its international investment position. With a negative international position, this predicts that the United States would lose money, at its present value, by imposing border adjustments. However, as the paper also notes, this calculation requires adjustment if some of the US trade deficit arises from mismeasurement of trade. Put simply, if some of the reported trade deficit arises from the understatement of net exports to related parties in foreign tax havens, that trade deficit generates offsetting foreign source income and does not occasion any accumulation of net US liabilities that must be serviced through future trade surpluses.2 Given estimates of the large magnitude of such mismeasurement, border adjustments could have generated substantial tax revenue well beyond the official 10-year estimation period.Of course, there were also arguments against the DBCFT. The concerns of importing industries may have been the most important politically, and these are closely related to this paper’s analysis. Additional concern arose about potential conflicts with respect to international tax agreements, notably the World Trade Organization. The key issue here was the extent to which border adjustments, explicitly permitted in the context of the VAT and trade neutral from an economist’s perspective, would somehow violate international trade norms if applied outside the VAT. I return to this issue below when discussing the paper’s modeling approach. Finally, there was concern about the possible wealth transfers associated with exchange rate appreciation, which this paper explicitly discusses and for which it provides an estimate. Although this wealth effect is relevant in evaluating the DBCFT, it requires more context than the paper provides.As discussed by Auerbach (1997), adding border adjustments to a source-based cash-flow tax amounts to imposing a cash-flow tax on net foreign investment positions or, equivalently, imposing a cash-flow tax on outbound foreign investment positions and a cash-flow subsidy on inbound foreign investment positions. While a cash-flow tax generates no revenue in present value on new marginal investments, it does collect revenue on the positive cash flows from past investments and new inframarginal investments. Thus, the cash-flow tax on outbound investment generates tax revenue in present value, representing a net flow from domestic individuals to the government, whereas the cash-flow subsidy on inbound investment represents a transfer to foreign investors. This is, essentially, the wealth transfer to foreigners that the current paper discusses, although there are further complications associated with the nominal currency denomination of assets and liabilities that affect the distributions of gains and losses.However, there are two things to keep in mind about this wealth transfer. First, applied in the context of imposing a DBCFT rather than on its own, the border adjustment is just undoing the imposition of a cash-flow tax on foreign investment. That is, a source-based cash-flow tax applies to domestic cash flows. The border adjustment adds in foreign cash flows of domestic residents and takes out domestic cash flows of foreign residents. Thus, the cash-flow subsidy for inbound investment offsets a tax that otherwise would be imposed on foreign investors by the domestic cash-flow tax. Second, the extent to which one is transferring wealth to foreigners by adopting a border adjustment depends on how much of a burden the tax system without border adjustment would impose on them in the first place, that is, the extent to which the incidence of source-based business taxes falls on foreign owners. There is a general view that, through capital flight as well as tax avoidance, much of the burden of traditional business income taxes falls on domestic fixed factors, notably labor.In summary, one should measure the wealth transfer to foreign investors in net terms, which is not something that can be done using this paper’s modeling approach. All this being said, adopting a DBCFT does mean giving up on attempts at “tax exporting,” and this should be weighed against the associated economic gains, a trade-off modeled by Auerbach and Devereux (2018). For example, if a country’s primary source of domestic business income were from the development of natural resources, fixed in location and relatively easy to measure in value, a border adjustment would provide a windfall to foreign investors with little offsetting economic benefit. As such, it would make sense, if imposing a DBCFT, to maintain some source-based taxation on industries where the distortionary behavior of modern multinationals is not so important a problem (Auerbach et al. 2017).Interpreting the Paper’s Simulation ResultsOne additional argument made in 2017 against adopting the DBCFT was that it represented a major change not only from past US policy but also with respect to what other countries had done, leaving considerable uncertainty about its economic impact, particularly in the short run. Would the large nominal appreciation predicted by basic theory occur? Would there be disruptions to the level and pattern of trade? The heart of this paper is directed to address such questions. After highlighting the conditions needed to guarantee the simple outcome of full dollar appreciation and a neutralization of trade disruptions, the authors develop an elaborate model aimed at incorporating various real-world frictions to assess what might actually happen.Perhaps most important among these frictions is the dominant currency (i.e., dollar) pricing of imports and exports, which makes such prices sticky in dollars in the short run, even if substantial dollar appreciation occurs. This means that US exporters fail to pass along the benefits of border adjustment of exports to their foreigner buyers, and US importers do not see much of a change in the dollar price they pay for imports, making such imports more expensive once the border adjustment is taken into account. Consequently, both exports and imports fall substantially in the short run. These offsetting trade effects still result in strong dollar appreciation but without the trade neutrality that exists in simpler models in the short run or in this model in the longer run. The net impact is a small short-run increase in gross domestic product (GDP; Barbiero and colleagues’ fig. 2), which becomes larger when considering wealth effects (their fig. 3) because of the resulting moderation in import demand. Perhaps most striking is the impact of retaliation by other countries, which is modeled initially by assuming that other countries defend their currencies against the dollar. Why other countries would seek to do this is unclear, given that the result is a stronger short-run improvement in the US trade balance and a much larger positive jolt to the US GDP. Of course, international trade policy is an area where decisions are not always easily explained, so it is hard to predict how other countries might react. However, an alternative response in other countries to US adoption of a DBCFT might have been to follow the US lead, either by adopting a DBCFT as well or by simulating a move in that direction by increasing existing VATs and reducing employment taxes. If countries were to adopt this alternative approach, then one would expect muted effects in the United States with respect to exchange rates, trade, and output. Indeed, this is what the paper finds in modeling a response by the rest of the world to announce adoption of the DBCFT once the United States does. In a parallel universe, important economies might take such a coordinated approach to smooth the path for adopting DBCFT-based reform.The paper’s findings for a VAT are quite different. Whereas adopting the border adjustment alone, which, again, is equivalent to adopting a DBCFT, is mildly stimulative in the short run, adopting a VAT has a sharply negative impact on output. To understand why, note again how these two tax systems differ. The VAT is equivalent to a DBCFT plus a tax at the same rate on domestic labor income, imposed at the employer level. Thus, whereas the DBCFT is a nondistortionary tax (at least when prices are flexible), the VAT imposes a tax wedge on the labor market, whose impact on labor supply is exacerbated in the paper’s model, because with all taxes refunded, the impact on labor supply is through a pure substitution effect. There is no negative income effect to dampen the demand for leisure. However, before concluding that the authors finally arrived at an explanation for why the United States chose not to adopt a VAT, one should keep in mind that, unlike in the paper, VATs have not simply been added to existing tax systems. Rather, VATs have arisen around the world either to replace more distortionary taxes, such as turnover taxes or sales taxes that (like those at the US-state level) miss much of consumption and tax many intermediate products, or to cover funding imbalances, which without the VAT would have eventually required the introduction of other taxes.One other key difference between the effects of the VAT and those of the DBCFT that does make sense, and that I think would be present more generally, is that a price adjustment to the VAT would occur largely domestically, through an increase in the price level rather than an appreciation of the nominal exchange rate. The paper explains this difference in terms of the DBCFT treating foreign and domestic goods differently, but I find much simpler intuition in the fact that for a VAT not to result in a price-level increase would require a fall in the sticky nominal wage rate. Put another way, adjustment to the VAT could occur through dollar appreciation plus a fall in nominal wages, whereas for the DBCFT dollar appreciation alone would suffice.Returning to the results for the DBCFT, a central question is what to assume about the pricing behavior of importers and exporters. The authors argue in favor of dominant currency pricing over, for example, producer currency pricing, which would lead to an immediate adjustment rather than a short-run trade contraction. Their argument is based on empirical evidence of such behavior. I do not dispute the evidence, but I question whether one can treat it as reflecting a structural model of behavior, invariant to the cause of exchange rate fluctuations. Put simply, while it might be quite rational for an optimizing producer to maintain a fixed dollar price in the short run in the face of a fluctuating exchange rate, a large, onetime, permanent change in relative costs induced by a border adjustment would call for a different response. If such changes do not feature prominently in the data used to estimate pricing behavior, then one may need to apply the resulting empirical evidence with considerable caution.As an illustration of the potential pitfalls, consider an alternative method of implementing border adjustment. While the approach proposed in the United States, and the one modeled in the paper, is to implement border adjustment through the US firms involved in cross-border transactions—denying a deduction for imports and exempting export revenues from taxation—one could also impose border adjustments on the foreign parties to these transactions. That is, one could tax companies exporting to the United States and provide tax rebates to purchasers of US exports. Indeed, this alternative approach to the border adjustment of imports is not just a conceptual experiment, it has been suggested as a way of ensuring the compatibility of the tax with WTO rules (Grinberg 2017), and so represents more than just an intellectual exercise. This transfer of tax liability from importers to foreign exporters would be very large—a 20% tax rate would mean a transfer of 20% of the purchase price from seller to buyer. Yet under the paper’s modeling assumptions, it might appear that this shift would have no immediate impact on the dollar price of imports. The authors argue (in n. 9) that an assumption of no pass-through of a 20% tax rate on producers is unrealistic, and I wholeheartedly agree. But I disagree just as strongly with their contention that a permanent 20% tax rate would be passed through fully while a concurrent permanent 20% exchange rate depreciation would have very limited immediate pass-through. Unlike the authors, I don’t see evidence of asymmetric past responses to exchange rate fluctuations and tariffs as being potentially very informative here, given the much closer underlying economic symmetry of the exchange-rate and tax-rate changes in this case, in terms of their effects on relative prices, their size, and their permanence. I cannot cite clear empirical evidence in favor of my argument, because the experiment would represent a sharp break from historical practice, but my own sense is that adoption of a DBCFT with a standard approach to implementing border adjustment would lead to a much faster adjustment of dollar prices than is assumed in the paper’s calibration.Finally, in a more realistic modeling of the macroeconomic effects of the DBCFT (and not simply a border adjustment), one would expect there to be a short-run impetus to invest coming from the removal of source-based taxation on the profits from domestic production. This would likely lead to capital inflows and higher short-run dollar appreciation.Looking AheadThe paper concludes with a call for more research. Papers often end this way, but in this instance the call is clearly warranted. The US proposal for a DBCFT might have fared better or might have been modified to make it more attractive if the proposal and its effects were better understood by policy makers, but especially by economists. In addition to giving further consideration to the modeling issues just discussed (as well as the assumption regarding monetary policy, which the authors highlight in the conclusion), one useful thing to consider would be how a DBCFT might appeal to different countries, acting either unilaterally or together. As the paper notes, the effects would be quite different within a currency union where the implicit fiscal devaluation could not be offset by a nominal exchange rate adjustment. However, there are also differences in exposure to trade, domestic market flexibility, and, of course, the currencies in which prices are set.In light of the failure of the United States to adopt the DBCFT in 2017, a potential researcher might ask whether time would be better spent exploring other research questions. As to this, one can observe that, even though the DBCFT as originally proposed was not part of the Tax Cuts and Job Act, the DBCFT proposal had an impact on the law. The new tax system contains provisions that reflect some of the same motivations as those behind the DBCFT: to make the system less distortionary, to encourage production in the United States, and to limit the ability of multinational companies to shift profits to lower-tax jurisdictions. On the domestic side, the law has introduced (if only temporarily, and applying only to some assets) expensing of investment and a limit on (although not elimination of) interest deductibility by businesses, moving in the direction of cash-flow taxation. Regarding cross-border transactions, the law did not adopt full border adjustment but did introduce two new provisions not found in other countries’ tax systems that effectively implement partial, limited border adjustment. These are a minimum tax on domestic income (the base erosion and antiabuse tax, or BEAT), which allows no deduction for certain imports from related foreign businesses, and a reduced tax rate on income (over a threshold rate of return on tangible assets) attributable to exports (i.e., on foreign-derived intangible income, or FDII). These provisions represent a more modest and therefore potentially less controversial move in the direction of the DBCFT, but at the cost of much greater complexity and a variety of new distortions.3 Given the new law’s built-in instability (because of its many expiring provisions), new opportunities to consider tax reform are clearly in our future. With the same economic factors that drove US consideration of the DBCFT at work in other countries as well, the value of additional research should be high.EndnotesFor acknowledgments, sources of research support, and disclosure of the author’s material financial relationships, if any, please see http://www.nber.org/chapters/c14070.ack.1. See Auerbach (2017) for details regarding these arguments.2. Auerbach (2017) discusses the adjusted calculation in more detail.3. For example, the base erosion and antiabuse tax may lead companies to spin off related foreign parties in order to avoid losing a deduction for imports.ReferencesAuerbach, Alan J. 1997. “The Future of Fundamental Tax Reform.” American Economic Review 87 (2): 143–46.First citation in articleGoogle Scholar———. 2017. “Demystifying the Destination-Based Cash-Flow Tax.” Brookings Papers on Economic Activity 2017 (2): 409–32.First citation in articleCrossrefGoogle ScholarAuerbach, Alan J., and Michael P. Devereux. 2018. “Cash-Flow Taxes in an International Setting.” American Economic Journal: Economic Policy 10 (3): 69–94.First citation in articleCrossrefGoogle ScholarAuerbach, Alan J., Michael P. Devereux, Michael Keen, and John Vella. 2017. “Destination-Based Cash-Flow Taxation.” Working Paper 17/01, Oxford Centre for Business Taxation.First citation in articleGoogle ScholarGrinberg, Itai. 2017. “A Destination-Based Cash Flow Tax Can Be Structured to Comply with World Trade Organization Rules.” National Tax Journal 70 (4): 803–18.First citation in articleLinkGoogle ScholarHall, Robert E., and Alvin Rabushka. 1983. Low Tax, Simple Tax, Flat Tax. New York: McGraw-Hill.First citation in articleGoogle ScholarInstitute for Fiscal Studies. 1978. The Structure and Reform of Direct Taxation. London: George Allen & Unwin.First citation in articleGoogle ScholarNunns, Jim, Len Burman, Ben Page, Jeff Rohaly, and Joe Rosenberg. 2016. An Analysis of the House GOP Tax Plan. Washington, DC: Urban–Brookings Tax Policy Center. http://www.taxpolicycenter.org/publications/analysis-house-gop-tax-plan/full.First citation in articleGoogle ScholarTax Reform Task Force. 2016. A Better Way: Our Vision for a Confident America. Washington, DC: US House of Representatives. https://www.novoco.com/sites/default/files/atoms/files/ryan_a_better_way_policy_paper_062416.pdf.First citation in articleGoogle Scholar Previous articleNext article DetailsFiguresReferencesCited by NBER Macroeconomics Annual Volume 332018 Sponsored by the National Bureau of Economic Research (NBER) Article DOIhttps://doi.org/10.1086/700908 © 2019 by the National Bureau of Economic Research. All rights reserved.PDF download Crossref reports no articles citing this article.