Short answer: No, it does not matter at all in substance, although it evidently matters quite a lot optically.
By the way, Grover Norquist, who is no fool when it comes to assessing optics, understands this, and has apparently criticized 9-9-9 for permitting taxes to look lower because they appear to come in separate bundles. There is an ironic jujitsu to all this: so much of Cain's momentum on this issue reflects the fact that, to the conservative base, 9-9-9 looks lower than 27. But holding constant what the taxes actually are, Norquist is right that, if they look lower, this may grease the wheels for letting them go higher than otherwise.
But enough about the optics - why, as a matter of substance, doesn't it matter if there are three 9% taxes, or an 18% tax and a 9% tax, or one 27% tax?
Let's start with the point that Ed Kleinbard and I have been making, and that others such as Ezra Klein have picked up on, which is that 9-9-9's "business tax" and its sales tax are basically the same tax (a few odd details aside), since they are both flat-rate consumption taxes where the tax base is consumer purchases. So we have at a minimum just two taxes, 18% on consumption plus the possibly separate (or not?) 9% personal income (?) tax.
When I made this point in a previous blog post, a commenter made the point that what I called the business-level VAT is in some ways not that far from an income tax. Basically, all you have to do in order to make it an income tax is the following:
(a) allow wages to be deducted by the business. In the standard VAT setting, deducting wages or not is a wash - it doesn't matter in the end - if (i) they are made includable by the worker only if they are made deductible by the business and (ii) the worker and business tax rate are the same. After all, if my business pays me $100 and both its tax rate and mine are 35%, then deduction and inclusion, instead of neither, saves the business $35 of tax and costs me $35 of tax. By the way, even if I don't own the business, then if the net tax on wage payments is zero one would expect wages to adjust, reflecting real market conditions and people's responding to their actual incentives from after-tax returns, so it wouldn't matter (once everything had time to adjust) whether the wages were includable and deductible or neither - apart from in the situation where the worker's marginal tax rate differs from that of the business.
The Cain plan screws this up, of course, by making wages includable by the worker but not deductible by the business. Is this an evil example of that odious crime, "double taxation"? As an optical matter, I suppose you could argue that it is. But in reality, all it does is cause workers who are paid wages to be taxed, in effect once, at 27%. The problem here, as I have noted in prior posts, is that paying observable wages is tax-discouraged, so owner-employees with sufficient liquidity to fund their personal consumption expenses will tend to avoid it and lower their overall tax rates to 18%.
Okay, that more than covers the first important difference between a business VAT and an income tax. On to the next two:
(b) In a business income tax, financial flows such as interest would tend to be deducted by the business that pays them and included by the recipient. (In the present income tax, of course, dividend payments, unlike interest, are not deductible.) In a VAT, all of these financial flows are typically ignored on both sides of the transaction, leading to problems in figuring out how to tax financial institutions (which embed service fees in the spread between the low interest rates they pay and the high interest rates they charge). Once again, however, leaving aside the double taxation of dividends problem in the existing income tax, both including and deducting or doing neither is a wash if the tax rates on both sides of the transaction are the same.
(c) And now for the big difference, applying even when all cash flows are taken into account (or not) reciprocally and when all tax rates are the same. In an income tax, business outlays that create durable assets or lasting value or income expectations beyond the current year are not deducted. Instead, they are capitalized, creating an asset that has a tax "basis," and are only recovered against gross proceeds, such as through depreciation deductions or by only taxing net gain on an asset sale (amount realized minus basis). An income tax, therefore, unlike a consumption tax style VAT, burdens saving and investment.
Suppose Cain altered the business tax in the 9-9-9 plan to be some sort of an income tax version of the VAT, with business outlays potentially being capitalized rather than immediately deducted. This would make the "business tax" component meaningfully different from the sales tax component. At the cost of economic distortion (i.e., discouraging saving and investment), it would burden the business owners and thus presumably increase progressivity.
But to say that there were now two separate 9s, while in a sense optically true, would in a sense be pure semantics. By making the two taxes a bit more different, we would be burdening some people a bit more and others a bit less. But if rates were adjusted to raise the same revenue either way, there would just as much taxation going on from the two taxes that now were somewhat distinct than there would have been if they were the same.
This brings us to the next element of major confusion in the 9-9-9 plan. The campaign website proudly trumpets that dividends will be deductible, so as to avoid present law's double taxation. But what about interest? Is that deductible as well? It is under the current income tax, but again, a VAT generally ignores it. But of course, to understand what would make sense here, we need to consider the last (or first?) of the three 9's, the tax on individuals.
Surely that must include dividends if deducting them at the company level is necessary to avoid double taxation. But surely interest has to be treated symmetrically as well, and the plan says nothing about it.
What is the base for the individual tax? The website says it's "gross income." But what is that? In the existing Internal Revenue Code, "gross income" includes all receipts (except that basis is deducted from the amount realized, so that only the net gain from an asset sale is included) but has no business deductions whatsoever. But everyone seems to agree that the "personal" tax here will not apply to the gross proceeds of, say, a neighborhood candy store, but is only for salary plus apparently some other gross income items - dividends and (despite the hiccup in the business tax) presumably interest, and perhaps all the other stuff that under present law leads you to get a W-2 or 1099. (Though not capital gain, as the plan purports to exempt this.) By the way, if we are including all this gross income, then I don't entirely understand how Cain is repealing the entire existing Internal Revenue Code, since parts of it are surely needed to figure out what income is.
For example, if interest is included by individuals, and is meant to be deducted by the business even though the plan does not say so, then might we need the original issue discount (OID) rules for interest that has accrued economically but not yet been paid? To be sure, if the treatment is symmetric and all the rates are the same, then perhaps this doesn't matter as much.
What the Cain campaign presumably does mean by "gross income" is that there will be none of the deductions that the Internal Revenue Code allows against "adjusted gross income" or AGI. By AGI, it means net rather than gross income, but without itemized deductions such as home mortgage interest, and without personal exemptions or the standard deduction.
Incidentally, I assume that fringe benefits that aren't currently included would be taxed in the Cain plan. Most would agree that the exclusion for employer-provided health insurance is no less a special tax expenditure than the home mortgage interest deduction. So does he mean to make it nondeductible at the business level as part of the wage (although it's currently deductible), AND includable at the individual level on the same rationale? Inquiring minds would like to know.
Okay, suppose we were to conclude that the personal income tax really is an income tax while the other two parts are consumption taxes. This probably isn't right, as it appears to be more of a wage tax plus perhaps a tax on interest and dividends (which of course are includable in "gross income" under present law). But suppose that one of the 9's is an income tax while the other two 9's are consumption taxes. Does this mean we have two taxes of 18 and 9 rather than one at 27?
This question really has no correct answer, or at least none that is meaningful, as it's a matter of optics and semantics. Let's consider instead substance. Suppose we had a single 27% tax but it was one-third income tax and two-thirds consumption tax. (E.g., suppose returns to capital that were income but not current consumption were two-thirds deductible, or alternatively were taxable, in the manner of capital gains today, at a special rate that happened to be 9% instead of 27%.) This would in a sense be the same as 9-9-9 if one of the bits is an income tax and the other two are consumption taxes, yet it would clearly be one instrument.
Bottom line: counting separate tax instruments is a fool's game. What matters is the overall tax being levied. And here, 9 + 9 + 9 = 27. Since this is an equation, the left side equals the right side, and no one should think it matters which side you are looking at.