NYU Tax Policy Colloquium, week 6: Steve Shay (et al), "'A Better Way' Business Tax Reform: Theory and Practical Complications"

Yesterday at the colloquium, Steve Shay presented an early draft of an article (being coauthored with Cliff Fleming and Bob Peroni) that critically reviews the House GOP proposal to replace the existing income tax on corporate and individual business income with what many call the "DBCFT" (destination-based cash flow tax). The draft is still preliminary, so I'm not posting a link or even the abstract. But here are some thoughts on how one might parse the main issues that the paper discusses: I won't try to offer a full assessment of the DBCFT here - that would require a very long article.

As it happens, many others are writing such articles (including Shay et al). Given how well it's already being covered, the only way in which I plan to enter this fray via conventional publication is from somewhat of an angle. On June 1, I am giving a talk in Haikko Borga, Finland at the annual meeting of the Nordic Tax Research Counsel, in which I'll discuss how the U.S. ended up considering going this way, how the plan relates to what other countries have done, in particular with VATs, what light it sheds on where we and they might want to get in the long run,etcetera. This piece I believe will end up being published in a Nordic tax journal, but also perhaps in the US if it develops in a direction that I deem compatible with this. It differs from what people have been writing these days in that it might involve looking at the thing from a distance and with an effort at historical perspective, rather than focusing on how it might work or whether, taking all the parts together as given, it's on the whole good or bad.

Okay, enough of that detour; back to some of the main issues that Shay, Fleming, and Peroni are evaluating:

(1) Distributional effects – The House GOP tax plan as a whole is extremely regressive, and meant to be so, especially when one adds in the financing for the tax cuts that, while not currently being provided for, the Speaker no doubt intends. He wants to massively slash tax rates on rich people and pay for it by massively slashing benefits for poor people. But let's hold all that to one side - the DBCFT itself raises more interesting, complex, and potentially ambiguous distributional issues than this clear underlying intent might have led one to expect.

Since it's a compound proposal, I think one has to analyze it distributionally in multiple pieces. And for each piece, one must analyze both the transition effects of repeal and the long-term effects. These can be quite different. E.g., even if one were to conclude that the corporate income tax is mainly borne by labor rather than capital, a surprise overnight repeal would offer a huge transition gain to shareholders.

Anyway: here is one way of organizing the pieces one needs to think about, in terms of both the transition and the long-term incidence. BTW, I'm just looking here at the tax side. Obviously, if net revenues change, one must also look at what changes end up resulting on the spending side.

(a) repeal the corporate income tax (and other business income taxes insofar as they are replaced by this).

(b) enact an origin-basis VAT. I put this here as a separate stage to distinguish between (a) the shift to expensing, and an R-based system in which financial flows are ignored, and (b) the destination basis piece, which could end up being called off.

(c) convert the VAT from origin basis to destination basis,

(d) add a wage deduction to the VAT. People have been calling this a wage subsidy, which is fair enough, but since wages are being taxed under the individual income tax one really needs to think about them as a whole. Note that in, say, David Bradford's X-tax, there was a much tighter connection than there seems to be here between the wage subsidy and the wage tax. There the idea was to match the business rate to the top individual rate, which only applied to wages; hence a straightforward net wage subsidy applied until one reached the top bracket (at which point it became a wash).

(e) replace the exclusion of financial flows under a true VAT with a rule apparently including them, subject to disallowing net interest deductions.

(2) Expensing doesn’t automatically yield neutrality – Expensing is commonly lauded as permitting the tax system to be neutral between assets, and also between businesses. But the proponents don't plan to make losses refundable, as they are in a VAT. Instead, they'll offer interest, which is fine (any cash flow problems aside) if one is going to have enough net "income" eventually, but otherwise may disadvantage small and medium sized businesses. Net losses should be especially common here, in contrast to under a VAT, due to the wage deduction plus, in early years, expensing for new assets plus (one presumes, as a transition rule) continued depreciation deductions for pre-enactment assets. There may be a lot of distortionary transactions that end up being used to provide effective refundability.

(3) WTO illegality – It seems to be little disputed these days that the system would be immediately and successfully challenged as illegal under the WTO, even though an equivalent instrument (in which the wage subsidy was handled separately) would presumably be legal. Passing a law that one knows is going to have to be changed or repealed in a few years is pretty crazy. But if one wanted to look at it in a positive light, the idea might be: as soon as it's held to be illegal (which will presumably take a couple of years at least), and before any sanctions start taking effect, change the system then. Only, we'd have to feel good about the feasibility of doing this. I don't feel so good about this given the low health level of our political system, but it could conceivably become a path to having a conventional VAT.

(4) Hard to tax inbound – We in the US are late to the game, and hence are inclined to think that VATs are much better administratively than our various tax instruments. But while this is probably still true, VATs' workability has been undermined by the rise of e-commerce and global service industries. So finding and taxing inbound consumer items, by notionally policing the "border," is much more problematic today than it was when most major VATs came into force. While the OECD and various countries are trying to address this, it's unclear how successful they'll be. Would the DBCFT eventually need to include a use tax that was collected directly from households? Our income tax treaties may impede requiring exporters to the US, if they lack a U.S. permanent establishment, 

(5) Hard to trace "true" outbound – This is the flip side of the inbound problem.

(6) Multiple tax rates – Most prior proposal in the DBCFT's ballpark had a single business rate. Here you have instead a bit of a mess, with different rates for passthroughs and C corporations, continued income taxation at the individual level, etc.  The multiple rates can raise a host of problems - relating, for example, to related party transactions and, for that matter, to exactly how the currency adjustment would play out.  The proponents would also need to figure out and execute rules pertaining to tax-exempts.

(7) Treatment of interest expense – A true R-based system is so much simpler and cleaner than taxing financial income but disallowing net interest deductions. For example, the DBCFT would presumably cause taxpayers to prefer interest income to other income, since it can be used to soak up interest deductions. We have various such problems under present law, but they're hardly a good thing. Note also the need to figure out how to tax financial institutions - more urgent here than under VATs, because the countries with VATs also have corporate income taxes that apply to financial institutions, whereas we would be jettisoning that backup.

(8) How would other countries respond? – Apart from filing WTO claims, via tax competition? Emulation? Taxing US companies more on a source basis, given that treaty reciprocity had ceased to matter and the US had abandoned the outbound-taxation field? If the thing is enacted, which I consider unlikely, we will certainly find out.

(9) No tax on domestic but "outbound' rents - Countries with significant natural resources often like to tax domestic extraction for export. Also, U.S. companies that develop valuable IP may create rents that they realize through foreign sales. One should always want to tax outbound rents domestically, for two reasons. If they're earned by foreigners, it's efficient and the foreigners will actually bear the incidence of the tax. If they're earned by domestic individuals, it's still efficient and inclusion may be desirable from a distributional standpoint. This point tends to be conceded by DBCFT advocates, whom I have heard say (a) we could have a special rule for natural resources, and (b) we're not currently doing a great job with the likes of Apple and Google anyway. But why give up altogether? That's a clear defect, even if outweighed by other features that one considered clear advantages of the DBCFT.

One irony here, by the way, is that after all the huffing and puffing in the Treasury White Paper about how only the U.S. can tax the profits that Apple purported to stash in Ireland, because it's U.S. source income and the Europeans should keep their hands off, the DBCFT would actually define that income (which pertained to EU sales) as foreign source. So much for self-righteous vehemence.

(10) A very tough transition in multiple dimensions – The rush to enact this thing - if it gets enacted - is bound, I think, to end badly. The IRS isn't ready for it, state and local governments aren't ready for it, taxpayers are desperately scrambling to get ready for it, there's unlikely to be any stress-testing or a multi-stage process like that which gave rise to the Tax Reform Act of 1986, last-minute dark deals on obscure aspects are likely to occur, there may not be enough staff with enough time (and treated with enough respect) to get enough of it right, technical corrections may be politically impossible, etcetera. So even if this thing can in principle be done well enough to be worth doing, don't count on its actually happening that way, if it happens.