Scott Sumner  

Further thoughts on corporate income taxes

Money as Coined Liberty... The Growth of the Chinese Levi...

Stephen Williamson and John Cochrane have an interesting discussion of the effects of a corporate income tax. This comment by Cochrane is interesting:

So how do you deduct investment and leave something left over to tax? It rests on two ideas. First, that the tax code can distinguish "real" investments like buying forklifts from "financial" investments like buying stocks and bonds, and only deduct the former.

Second, that there is some pure "profit," some pure "rent," some "unreproducible input" (i.e. something that did not come from a past unmeasured investment), something like the classic "unimproved land" that can be taxed, without distorting any decision. It goes hand in hand with the complaints of greater monopoly.

But I find it hard to find and name a concrete source of profits that, once named, does not distort the decision to undertake some useful activity to make those profits. Starting, organizing, and improving a business, figuring out the intangible organizational capital that makes it a successful competitor, creating a product and a brand name, are all crucial activities for which no investment tax credit will successfully offset a large profits tax. "Intangible capital" is about all most companies have these days.

Then Gideon Magnus left this comment:

Yes with full investment expensing the tax burden as well as the expected present value of tax revenues is essentially zero, and the only reason to levy the tax is to close the loophole of people incorporating and then paying zero taxes on their labor income.

So putting this together, it seems like corporate profits involve a return on:

1. Land
2. Physical capital
3. Human capital
4. Innate talent (cleverness, etc.)

In that case, why not do the following:

1. Tax corporate income as personal income (which would effectively be the top marginal tax rate for big companies--40.8%.)

2. Allow full expensing of physical investment.

3. Do not allow interest expenses to be deducted.

4. Do not tax interest, dividends and capital gains (as part of the personal income tax).


1. You'd like to avoid taxing investment income. This system does that.

2. You'd like to tax labor income being disguised as capital income. This system does that. It also taxes land rents and the innate skills of entrepreneurs.

3. It treats debt and equity equally.

One downside is that it taxes income from investments in human capital. But in practice that's not really much of a problem. Recall that wage taxes also tax income from investments in human capital. So the treatment is equalized between the corporate and non-corporate sectors. More importantly, the acquisition of human capital is already heavily subsidized by the government. Indeed I'm persuaded by Bryan Caplan that it's too heavily subsidized. So if this sort of corporate tax regime slightly discourages the acquisition of human capital, it may actually improve welfare for standard "second best" reasons.

Consider how this would also simplify the taxation of personal income. For the vast majority of taxpayers, i.e. those paid wages and salaries, there would be no need to do taxes each April. The payroll and personal income tax systems could be combined into one---merely taxing wage income. For those claiming that savers are getting a tax break, just point out that they are actually paying taxes at the top rate, but at the corporate income level.

Proprietorship income would also be taxed as personal income, again with full expensing of physical investments.

What's wrong with this idea?

Comments and Sharing


COMMENTS (13 to date)

I don't understand why Cochrane suggests that there would be nothing left to tax. Investment expensing just changes the timing of deductions, not the amount of deductions. Accelerated depreciation methods do the same thing compared to straight-line depreciation. Investment expensing is just an even more accelerated depreciation method.

Expensing reduces future deductions. It seems like you would need to keep expensing amounts far in excess of what is currently being depreciated to eliminate taxable income each year. You might do that with a start-up for a number of years, but mature companies typically don't have capital expenditures significantly greater than their depreciation deductions because they are mainly just replacing fixed assets, not adding to them.

Thaomas writes:

Not taxing investment income is a good idea IF there is a progressive consumption tax. Until there is, let's tax investment income (but only inflation adjusted capital gains) and labor income at the same rates.

How else are we going to get high income people to support a progressive consumption tax?

Phil writes:

What you are describing is a cash flow tax. A cash flow tax does not tax the risk-free rate, but it does tax "economic profit" or inframarginal returns. It does not raise the cost of capital for marginal projects.

It is a very efficient tax, although raises far less revenue than an income tax, because an income tax taxes the normal rate of return.

So Gideon's comment needs an addendum - it does tax economic profit so tax revenues have a positive NPV - the government becomes silent partner in all projects, including ones that make economic profit.

For a great and fun read on some of these ideas, this is a great paper:
David A. Weisbach, "The (Non)Taxation of Risk," 58 Tax Law Review 1 (2004).

Googling that paper will bring it up.

Scott Sumner writes:

Mark, I don't think he'd be saying there is nothing left to tax, but rather that you'd be taxing intangible investments, which can't be deducted.

Thaomas, I think that's a really bad idea. "Inflation adjusted returns" on investment would be an absolute nightmare to calculate. Taxes already make April the cruelest month for me, and you want to make it far worse?

Phil, Is your definition of "economic profit" anything above the risk free rate of return? That seems an odd definition.

Phil writes:

Hi Scott

I am referring to what are variously called “inframarginal returns,” “rents,” or “economic profits.”

That is, higher returns than are otherwise available in the market.

It could be e.g. some slight (or strong) local monopoly power, particular skill in navigating planning laws that makes your developments earn higher risk-adjusted returns, or incredible stock picking ability (Buffett).

All of these economic profits are taxed (with no ensuing distortion) under a cashflow tax, while activities than earn the risk-free rate are not taxed under a cashflow tax.

A cashflow tax is simply one way of implementing a consumption tax, and it works quite well if applied at the corporate level.

Here's Auerbach using the term consumption tax when he could be using "cash flow tax":

Another conclusion of this literature, to which I also return below, is that the income tax
differs from the consumption tax only in its taxation of the safe rate of return on assets. To the extent that an assetís return is simply an excess return that compensates for risk-taking, the investor can undo the income tax by scaling up the position in the risky asset. Only the tax on the safe rate of return cannot be undone by this portfolio shift, and thus only the tax on the safe rate of return is relevant in assessing the distortion of an individualís intertemporal decisions.

Thus, whether investments are safe or risky, excluding from a consumption tax the cash
flows on new, marginal investments would, to a first approximation, have no impact on the
economy. But there are remaining components of the cash flows ) ( ) ( f f R − I + R − I which
6 may be grouped into two categories: the normal returns (including recovery of principal) from earlier investments and supernormal economic rents. It is useful to distinguish these two sources because opinions vary as to how these elements of the cash flow tax base should be treated in a transition from the income tax to the consumption tax. In particular, some view the tax on returns to existing capital as an unfair byproduct of the transition from one tax base to another, as
only the net returns, rather than the recovery of capital, would be taxed under the income tax. Imposing a consumption tax therefore introduces a levy on the sale of existing assets. The taxation of economic rent, on the other hand, is typically not viewed with such concern, because rent would be taxed equally under an income tax or a consumption tax.

More here:

Also, you write:
Mark, I don't think he'd be saying there is nothing left to tax, but rather that you'd be taxing intangible investments, which can't be deducted.

I think that's wrong. Intangible investments can and are deducted. When Apple employs engineers and designers to create intangible intellectual property, the salaries paid are completely deductible. That is, intangible investments *already have* expensing treatment.

Alan Goldhammer writes:

I'm afraid that Scott's column, though well meaning just continues to tinker around the edges. Corporate tax revenues as a percentage of total Federal revenue have been decreasing over the years. As TR Reid points out in a wonderful Wash Post column arguing for ZERO corporate income tax it is now barely 7% of total tax receipts.

Reid has a very good point here. No matter who carefully Congress tries to craft tax legislation there will always be loopholes that good corporate tax lawyers will uncover. Why not just take the easy approach and do away with the whole thing and move to some type of VAT to make up for the lost revenue. The fact that our "legislators" can't even do a simple tax reform bill that does away with all kinds of preferences shows how corrupt the whole process is.

anon writes:

I think this might have a loophole.

Someone has just built a building for $5mm and is hoping to sell it for $6mm. A corporation approaches and says, we want to buy the building for $100mm. We will pay you $6mm in cash and give you a 99 year mortgage for the balance. The note will be structured so that the 2% interest accrues like a zero coupon bond and the $667mm balance is due in 99 years. It's a non-recourse note, secured only by the asset. Under today's tax laws, the seller would reject this offer due to taxes that would be due on this immense capital gain and the phantom income on the accruing interest, but now the seller can accept this offer. The corporation can expense the $100mm and save $40mm in taxes. The corporation can pay for this building as much as necessary so that it pays no taxes at all and can repeat in future years as needed.


Jeremy Goodridge writes:

Hi Scott,

Seems like your proposal is very similar to the one by Hall/Rabushka -- tax income at the source. So business would pay taxes on income, but able to expense physical investment. In addition, no one would pay taxes on dividends, interest income, or capital gains (including hedge funds). It's another way to convert to a consumption tax : Tax income from labor but not from investments. Two other options are a VAT and an individual consumption tax (in which individuals deduct all $$ they use for investment -- like expanding IRAs to infinity and removing any/all rules on how/went the IRAs are spent).

One thing I don't understand : you say, "the vast majority of taxpayers, i.e. those paid wages and salaries, there would be no need to do taxes each April. The payroll and personal income tax systems could be combined into one---merely taxing wage income. " I don't see how this would be. People would still have to figure out what bracket they are in, or deal with deductions, and manage adjustments for couple income vs individual income. Or are you saying there will be one flat rate and no deductions (including charity?) and no joint returns?

Scott Sumner writes:

Thanks Phil, That's helpful.

I had always defined "economic profits" as profits above the risk-adjusted rate of return, not profits above the risk free rate of return.

On the intangible investments, I was thinking in terms of intangible investments made by the owners.

Scott Sumner writes:

Alan, There is something to be said for getting rid of both personal and corporate income taxes. I'd replace them with both a VAT and a payroll tax.

anon, Good question. Perhaps only allow expensing for the actual cash paid to the builder of the investment good. You finance the project somewhere else. Of course the scheme you describe would also be a problem with the current system---does anyone know how the IRS handles it?

Jeremy. Good points about the different ways of doing a consumption tax. I envision it as being like the current FICA system. No deductions and everyone treated like an individual. That system is now regressive, it can easily be made progressive.

zeke5123 writes:

Anon -- the note would need to be respected (i.e., if sham, then the Service could disregard the note as part of the purchase price).

The question then becomes whether obligation to pay the extra 667m in 99 years NPV is greater than the NPV of the cash tax savings. If you assume the whole expensing deduction is used in the first year, the NPV of the expensing would be 41m (100m * 40.8% tax rate).

Break even point seems to be around 2.8%. That tells me there is a flaw with Scott's plan.

Presumably, the Service could disallow expensing for any amounts not cash settled within X years (similar to the AHYDO rules). If the period is short enough (or if there is a requirement that interest be cash paid annually), then this planning doesn't work. Of course, it adds complexity back to the tax code -- something everyone finds out when attempting reform.

anon writes:

It wouldn't be a problem under our current system. The Seller would owe taxes on the huge capital gain and phantom interest.

Justin D writes:

--"Alan, There is something to be said for getting rid of both personal and corporate income taxes. I'd replace them with both a VAT and a payroll tax."--


How would you structure them? Would the VAT be flat and the payroll tax progressive, or would both be flat and be paired with a rebate/basic income grant? If both were flat, most individuals probably would never have to file for taxes at all.

How much larger would you expect RGDP to be 10 years after revenue-neutral tax reform to a VAT/payroll tax combination?


Return to top