Wednesday, January 13, 2016

Dean Baker on "A Progressive Way to End Corporate Taxes"

In today's New York Times, Dean Baker floats a corporate tax reform proposal that, as he notes, is well-known in the biz and yet has gotten little attention in recent years:

"Suppose that, instead of taxing corporate profits, we required companies to turn over an amount of stock, in the form of nonvoting shares, to the government.  We can fight over the percentage later ....

"The shares would be nontransferable, except in the case of mergers and buyouts, but they otherwise would be treated just like any other shares.  If the company paid a dividend to its other shareholders, then it would pay the same per share dividend to the government. If it bought back 10 percent of its shares, then it would buy back 10 percent of the government's shares at the same price. In the event of a takeover, the buyer would have to pay the same per-share price to the government as it did to the holders."

Suppose the government's non-voting share percentage is 25%.  If $75 million in dividends are paid to the regular shareholders, the government gets $25 million.  Obviously, this has a lot in common with the case where the government taxes dividends to the shareholders at a 25% rate, in lieu of owning nonvoting shares. In that scenario, the company pays out the same total of $100 million, all to the shareholders, but they pay $25 million over to the government.  Or, you can think in terms of the case where, with an entity-level corporate income tax of 25%, the company earns $100 million, pays $25 million in taxes to the government, and then can pay out the remaining $75 million to the shareholders.

Obviously, this is an integrated corporate tax if we don't otherwise tax dividends.  But if we still do, then it's analogous to the existing two-level corporate income tax. To make life simpler, one could imagine eliminating the shareholder-level tax when the proposal is hypothetically adopted, and simply taking that elimination into account when one decides what share ownership percentage to give the federal government.

The rationale for the proposal is that the corporation can no longer use tax planning to avoid its liability. The government stands in the same boat as shareholders.  Now, this clearly does leave some other games on the table.  As Herwig Schlunk notes, in his article "The Cashless Corporate Tax" that I believe we discussed at the NYU Tax Policy Colloquium around 15 years ago, the games companies might be expected to play include "the use of [financial] instruments that can avoid the designation of equity. This is not a new problem. Under the current corporate tax, the tax base (taxable income) is avoided through the use of debt." These problems might get worse under the new regime, albeit subject to influence by the question of how dividends versus interest are treated at the holder level, but presumably there'd still be less overall scope for tax planning than under present law.

To my mind, one of the biggest problems with the approach, which I haven't seen addressed at length although there might be something out there that I simply don't recall offhand or haven't seen, lies in the international realm. If we apply the proposal to domestically incorporated companies without regard to their mix between domestic source and foreign source income, then we have in effect adopted the equivalent of a worldwide residence-based corporate income tax (with no deferral, and foreign taxes merely being deductible).  If we try to avoid that result for U.S.-incorporated companies, then we are at a minimum bringing back some of the problems that the proposal presumably is meant to avoid.

What about foreign-incorporated companies that have U.S. source income?  Baker says in the op-ed that "we would presumably ... require that foreign companies making a substantial portion of their profit in the United States grant shares."

Without meaning to be too dismissive up-front, I am inclined to say "Ah, there's the rub."  Presumably, we'd have to look at the U.S. source versus foreign source income of foreign companies. That's bad enough to start with, but it's not even clear how income shares for a given year would translate to ownership percentages over a longer period.

The proposal is a non-starter unless something that's good enough can be devised to handle international issues.  I'm pessimistic about this, but admittedly have not studied the question, or read about it recently.


Susan Morse said...

Surely international is the rub not mentioned, though if one is willing to really suspend disbelief, the idea of a collective investment vehicle involving all countries, where the vehicle owns a certain percentage of [equity] of all businesses is kind of intriguing. The terms of the interests could reflect not only whatever completely hypothetical and counterfactual Kumbaya agreement might be assumed about the factors that determined each country's first order share, but also negotiations about what different countries wanted to get out of corporate taxation (time allocation, risk allocation etc).

Daniel Shaviro said...

Yes, agreed that this could be interesting in principle but is likely to remain Kumbaya in practice.

andy grewal said...

I thought this proposal required the payment of the current income tax liability in the form of nonvoting shares, and not a transfer of a percentage interest of the corporate stock that approximates the statutory rate.

That is, if the rate is 35%, and company worth $10B earns $100M, it would transfer $35M worth of shares in lieu of $35M of cash.

Is Baker instead suggesting that the government immediately tax the company for $3.5B worth of nonvoting shares? Would that even be constitutional without apportionment etc?

Daniel Shaviro said...

If I understand it correctly, the idea in the purely domestic setting is simply to travel with the other shareholders without further share transfers.

If a company worth $10B earns $100M, then all of the shares appreciate by 1% (all else equal). If the government owned 35% of the company, its stake would have gained $35M in value, without the need for extra shares. As for cash, the government gets its portion when dividends are paid.

I see that the op-ed write-up by Baker is ambiguous, and could be read Andy's way as well as mine. So I am basing my interpretation on how I have understood this type of proposal to operate, and on the logic expressed in the op-ed. E.g., there would still be an incentive to understate income even if the tax were payable in nonvoting shares rather than in cash.