Corporate tax distortions

Today's New York Times had an article by David Leonhardt on corporate tax reform, of interest even though the chance that such reform will be seriously addressed in the next two years strikes me as close to zero. (The one thing that I could see happening is an unfinanced corporate rate cut and/or shift to exempting foreign source income, which would happen immediately if President Obama agreed to it and Congressional Democrats were willing to go along.)

Key sentence in the Leonhardt piece:

"Arguably, the United States now has a corporate tax code that’s the worst of all worlds. The official rate is higher than in almost any other country, which forces companies to devote enormous time and effort to finding loopholes. Yet the government raises less money in corporate taxes than it once did, because of all the loopholes that have been added in recent decades."

The obvious question of interest is: What are these loopholes?

The number one item the article mentions is "maximizing the amount of profit that is officially earned in countries with low tax rates." This mainly reflects, not so much newly enacted loopholes as the forward march of tax planning technologies to exploit income-shifting opportunities.

Secondarily, it mentions "spending large sums on new equipment or buildings. Such spending can often be deducted." Under a pure income tax, capital outlays could not be expensed (i.e., written off in full right away), but instead would be capitalized and only deductible more slowly under cost recovery rules. But, under currently applicable law, much of it can be expensed - for example, equipment and R&D outlays (which also are partly creditable).

A third item, the allowance of net operating loss carryovers from prior taxable years is, as Leonhardt notes, much less of a policy concern. A company that makes no money over time generally shouldn't pay net taxes even if it has income in some years and losses in others.

What should we think about the first two items? Income-shifting by U.S. and other multinationals to low-tax foreign jurisdictions is not listed as a tax expenditure, because the whole point is that we don't know how to source the income accurately. (Indeed, to some extent this question has no clear theoretical answer even with perfect information.)

But we know that the results companies can and do report under existing law lowball the U.S. while putting economically implausible amounts of taxable income in tax havens where nothing actually happens, plus a few low-tax countries where there actually is some economic activity (e.g., Ireland) but the claimed profit shares are implausibly high.

As I noted in an earlier post, repealing deferral for the foreign source income that U.S. companies earn through their foreign subsidiaries is not really the answer, given that by definition this can only address income-shifting by U.S. multinationals, as opposed to those whose corporate parents are incorporated abroad. Plus, they also tax the "true" foreign source income, which is not necessarily optimal especially at the domestic U.S. rate, subject to allowing foreign tax credits, which definitely are not optimal.

So what we really need is better rules for determining the source of income. These rules would have to treat a multinational corporate group, whether the overall parent was a U.S. company or not, as effectively a single entity for purposes of determining what income is U.S. source. So intra-company debt would be ignored, and it wouldn't matter which group member borrowed from third-party lenders. Existing transfer pricing rules, based on the myth of comparable arm's length prices, would be thrown in the trash where they belong, reflecting universal recognition that they are unworkable. In their place, one might use some set of objective factors that the companies would find costly to move for tax reasons. There have also been other suggestions, however, aimed at teasing out "true" inter-company pricing by alternative methods.

What about expensing? This is a consumption tax feature that we have plunked into the income tax at various points. The odd thing about it is that, in a consistent consumption tax environment where (among other distinctions from present law) everything is expensed, it actually creates neutral investment incentives, based purely on pre-tax expected profit, and without the income tax's discouragement of investment and saving.

But shoving it into an income tax causes a couple of problems. First, items that are expensed are tax-preferred relative to those that are not. Second, combining it with income tax treatment of debt financing can have peculiar results.

Perhaps the following simple hypothetical can help to demonstrate. For arithmetic simplicity, suppose the tax rate is 50%, the normal rate of return on investment (including loans) is 10%, and the government always provides an immediate refund (at the 50% tax rate) when you claim a deduction.

I make a $100 investment that I am allowed to expense (even though we otherwise have an income tax). So it costs me only $50 after applying the instantaneous tax refund. I borrow this $50, and thus get to make a $100 investment at zero out-of-pocket cost.

In a year, given the normal 10% return, I get to cash in my investment for $110. But I have to pay back the bank, which gets $55, as it has earned 10% interest on its $50 loan. So I have $55 left in my pocket before considering the income tax consequences.

As it happens, my taxable income is $105 (i.e., $110, given that the $100 outlay has already been expensed, minus $5 interest). At a 50 percent rate, I pay tax of $52.50 and have $2.50 left. So I generate a profit on a zero cash outlay even though I merely earned the normal rate of return.

Even if I had earned only 8%, I would still have come out ahead. This would have left me with $108 before repaying the loan, $53 afterwards, taxable income of $103, a tax liability of $51.50, and $1.50 left in my pocket even though I invested both the bank loan and my tax refund in something offering a below-normal return.

The underlying point is that expensing of investments plus debt financing that gets income tax rather than consumption tax treatment is affirmatively subsidized, not merely exempted (as under a consumption tax that exempts the "normal" rate of return), potentially creating bad incentives.

Admittedly there may be more to the example, permitting one to challenge the diagnosis. In particular, if the lender is taxable on its $5 of interest income, the government gets its money back and breaks even, rather than handing out a net subsidy. Suppose, however, that the lender is effectively tax-exempt. This brings back the problem, but arguably changes the diagnosis to one of how tax rate differences can be exploited (especially, it turns out, when debt and equity are treated differently).

For me the bottom line is that the discussion of how to finance lowering the corporate rate (leaving aside the political reality point that it will either be unfinanced or remain undone) should focus on (1) the source rules for all (i.e., U.S. and non-U.S.) multinationals, (2) income tax preferences that we are willing to address, and (3) interest deductibility, given both its interaction with expensing and the debt versus equity problem.

The unfortunate thing about #2 on this list is that moving closer to a true income tax is not necessarily the right way to go. But shifting to consistent consumption tax treatment of business (and also of individuals) is probably not on the agenda.