NYU Tax Policy Colloquium, week 2: Mark Gergen's "How to Tax Global Capital," Part 1

Yesterday at the colloquium, Mark Gergen presented his paper, “How to Tax Global Capital.”  A predecessor paper that provides important background for it, “How to Tax Capital,” is available here.

Herewith some background and thoughts regarding this very interesting paper. Part 1 discusses Gergen’s proposed major tax reform, as applied purely in the domestic setting. Part 2, which I will place in a separate blog entry given the length of Part 1, discusses the international aspect, which was the subject of his new paper (which, however, is still at an early stage of development).

1. DOMESTIC SETTING

One could think of the income tax as falling on labor income and capital income. In practice they often are mixed.  For example, if Mark Zuckerberg sold his Facebook stock, the capital gain might look like capital income, but in true substance would mainly be labor income. The existing tax system generally doesn’t try or need to tell them apart (and if one thinks of the capital asset definition as aiming to do so, it gets this wrong, as in the Zuckerberg). Nonetheless, the distinction is conceptually important, since one might favor different policies towards the two pieces if one could sufficiently effectively distinguish them.

As has frequently been discussed (including at our week 1 colloquium), a system that provided expensing for all capital outlays, or that provided interest on basis at a suitable interest rate, would effectively exempt the normal return piece of capital income, which is the piece that it might be reasonable to want to exempt.  Thus, expensing or interest on basis can be the backbone of a well-designed consumption tax. This, in turn, may be viewed as equivalent to a labor income tax, if you spend what you earn and saving does not alter the present value of the tax liability that would have resulted from spending one’s earnings immediately.

Gergen’s novel proposal involves repealing the income tax, and replacing it with separate instruments in lieu of the labor income tax component and the capital income tax component. While this is not the main focus of his proposal, he’d replace the labor income tax with a VAT or a consumed income tax (i.e., cash flow personal tax that is like an income tax with unlimited IRA deductions for saving and inclusions for dissaving). As for the capital income tax, he’d replace it with a two-part instrument, comprised of a “securities tax” and a “complementary tax.”

The securities tax would apply to all capital represented by a publicly traded security, at its market value. It would be assessed and collected from issuers. The tax rate might be, say, 0.8% annually. (As I further discuss below, the reason the nominal rate is so low is that it is a wealth tax, rather than a capital income tax.)

To illustrate: Suppose that Apple has a current market capitalization of $650 billion (i.e., this is the market value of its equity). Suppose further that Apple has issued publicly traded debt securities that are worth $100 billion, and has non-traded outstanding debt (such as accounts payable) worth $50 billion. Apple would annually remit securities tax of (a) about $5 billion with respect to its stock (at a rate of 0.8%, (b) $800 million with respect to its publicly traded debt, and (c) zero with respect to its other debt – which is not a publicly traded security, and hence is subject to the complementary tax, described next, rather than the securities tax.

When the value of a publicly traded security reflects the issuer’s holding of other publicly traded securities, a tax credit prevents cascading application of the same tax multiple times. For example, if Goldman Sachs holds Apple stock, Goldman Sachs gets a credit for the securities tax paid by Apple. For example, if Goldman Sachs holds $100 million of such stock, it would get a tax credit of $800,000. Absent this feature, when Goldman held Apple stock, the value of such stock would in effect be taxed twice (as it presumably would increase the value of Goldman’s stock by its own value).

The complementary tax would apply to all income-producing assets other than publicly traded securities. In addition to items such as Apple’s non-publicly traded debt, it would generally apply to partnerships, other businesses, bank accounts, gold and cash held by individuals, etc. More below on its applying only to income-producing assets.

The complementary tax would apply the same rate (such as 0.8%) as the securities tax, to what I will call its “basis” although this is meant to be a proxy for its value. An item’s initial basis typically would be its cost basis (presumably, carryover basis for gifts and bequests, unless the gift is a revaluation event, as might be the case if the item had to be valued for gift or estate tax purposes). It would then be increased by a measure of the normal rate of return – say, the federal borrowing rate plus 2%.  The complementary tax would then apply annually to what I am calling basis (i.e., estimated value under this methodology). However, the basis would be changed to a market-based measure of value whenever some arm’s length event, such as the redemption of someone’s interest in a hedge fund, made this feasible. (This market revaluation principle also might apply, say, to one’s gold holdings, and it certainly would apply to one’s bank account.)

To further illustrate the complementary tax, suppose I buy or create a business for $10 million, without its having publicly traded securities. In Year 1, I pay complementary tax of $80,000. Suppose that, for the ensuing year, the deemed rate of return is 4%, and that I take no cash out of the business (which would reduce “basis”). Then, in Year 2 the “basis” is $10.4 million, and the complementary tax (at 0.8%) is $83,200.

It’s crucial to the proposal’s successful operation in practice that the complementary tax be at least a decently good proxy for the securities tax. While a degree of divergence might be tolerable (and is inevitable, given the difficulty of measuring the value of non-publicly traded items) – even though this would create some distortions around decisions to issue publicly-traded securities and to sell other assets – if it gets too great, then a key underlying assumption is undermined. This is the assumption that people won’t be enormously eager to avoid the securities tax, because even if they do the complementary tax will get them to approximately the same place. If the two are close enough, then the main difference between them is administrative – the securities tax collects the tax at the issuer or entity level, while the complementary tax does so at the owner level.

Here are some main issues that I see as raised by the proposal, holding off the very challenging international issues for Part 2 in my next blog post:

1) Constitutional issue – This is basically a wealth tax, although one could try to call it a levy on mandatorily imputed capital income. (The latter characterization seems unlikely to succeed, since the “imputation” based on value is irrebuttable.) Hence, there are major constitutional issues raised under the constitutional requirement that “direct taxes” be apportioned among the states. See discussion here (at pages 46-49) in a paper that I recently coauthored with Joseph Bankman, and that subsequently appeared in the Tax Law Review.

2) Wealth tax vs. capital income tax – The two are identical, with adjusted nominal rates, if assets’ rate of return is completely fixed both across time and between assets. Thus, suppose that all assets always earn a rate of return of exactly 4%. Then a 0.8% wealth tax is effectively equivalent to a 20% capital income tax. For example, an asset worth $100 million (generating $800,000 of 0.8% wealth tax liability) would always generate a return of $4 million (generating $800,000 of 20% capital income tax liability).

Things are not so simple, however, when rates of return change across time or differ between assets. For example, if the rate of return drops to 2%, then a 0.8% wealth tax is instead identical to a 40% capital income tax. (The legislature could presumably keep changing the tax rate under one instrument or the other in order to keep them in equipoise, but this would change the burden of inertia.) Likewise, if Asset A is worth $1 million because it quadrupled in value over the last year, while Asset B is worth $1 million because it lost 80% of its value, this year’s wealth tax treats them the same but this year’s capital income tax treats them very differently.

So, even apart from the constitutional issue, one may want consult one’s own beliefs in order to determine which one seems preferable. Does one want the effective tax rate to be automatically adjusted when normal rates of return change? Does one care about asset value changes in the past year, or only about current value?

3) Flat rate structure – On the labor income piece, the VAT would have a flat rate structure, while the consumed income tax could have a nonlinear rate structure. But this lies outside the proposal’s distinctive features. The securities tax inevitably is a flat rate tax, since it’s collected at the issuer level, and the complementary tax may need a flat rate (and the same rate) as well, both to maintain parity between the two components and since, to apply a nonlinear rate intelligently to it, one really would need to know how much wealth a particular taxpayer had that was subject to the securities tax.

4) Transition issues – Lots to analyze here, although I won’t attempt it at present. These include transition gains from repealing existing taxes, transition losses from enacting the new taxes, and anomalous effects on pre-enactment contractual and other arrangements that reflected assumptions about who – as between, say, an issuer and a holder – would be charged with remitting the tax liability.

5) Liquidity issues posed by both taxes – Some of the taxpayers who were required to remit either tax might face liquidity issues. E.g., suppose (counterfactually) that Apple not only suffers huge losses in a given year, but also is strapped for free cash.

6) “Forced dividend” issues posed by the securities tax – Suppose that the tax liability pertaining to Apple’s $650 billion of stock really would be the same (about $5 billion) whether Apple remitted it under the securities tax, or shareholders remitted it under the complementary tax. (Based on what I’ve said so far, the latter would require that Apple stock not be publicly traded – a rather large and unlikely change – but, as we’ll see in the international piece, Apple might avoid being required to remit the tax if it could avoid being classified as a U.S. issuer.)

Why might Apple’s managers or shareholders care which of them pays the same $5 billion amount? One could think of (a) requiring Apple to pay it as equivalent to (b) requiring the shareholders to pay it, but also requiring Apple to pay out a dividend to the shareholders in the amount of the tax. Not just liquidity issues but also principal-agent issues could make this choice consequential to the parties.

7) Owned-occupied housing and consumer durables – These could easily be subjected to the complementary tax, assuming a political and policy willingness to do so. The main question posed would be how to adjust “basis” from year to year. Should it appreciate even though the owner is presumably generating an imputed return by reason of asset use? Should economic depreciation instead apply? The latter would surely be necessary for short-lived consumer durables. One problem with not taxing this category of items is that one could imagine very rich people avoiding the tax by buying super-expensive sports cars, artworks for their personal collections, etcetera.

8) Consumer and other nontraded debt – Maybe I don’t understand the proposal correctly in this regard, but I wonder if it has a problem dealing with nontraded debt in some settings. Case 1, Goldman holds $X of debt issued by Apple. Under the securities tax, Apple pays tax on the value of the debt, but Goldman gets a credit. Case 2, I borrow $Y from my neighborhood bank. It pays complementary tax on the value of the debt obligation, but I don’t get any sort of credit. Is this a problem? Gergen thinks not, but I’m not entirely sure on what grounds. Happy, however, to be enlightened on this.