Virtually all economists, aside from David Lereah and his successors, think that the mortgage interest tax deduction should be eliminated, using some sort of sunset combined with a grandfather clause so that we don't suddenly push millions of more American homeowners into foreclosure. Equalizing the treatment of dividends and interest payments is also popular. But I don't usually hear people advocating, as Felix Salmon does, eliminating the deduction for corporate interest payments. In a post titled "What are the arguments for privileging debt?" he says:
The weird thing for me is that when I start banging this particular drum, I always get exactly the same answer: "yes, great idea, not gonna happen". But is there any intellectual justification whatsoever for making corporate interest payments tax-deductible? I can see an argument for a carve-out for highly-regulated banks, since their entire business is based on making profits from the spread between the rate at which they lend and the rate at which they borrow. But banks aside, why should companies pay lots of tax on dividends, and no tax at all on bond coupons?
In a way it's depressing: if this were a real debate and Paul Volcker had a remote chance of making interest taxation happen, then surely there would be no shortage of academics and corporate lobbyists making the case for keeping the status quo. The fact that they're not even bothering is all the evidence we need that this isn't even going to reach trial-balloon status, let alone get signed into law.
But still, the question remains: if they were to start taking this seriously, what arguments would they use? After all, as Surowiecki notes, the likes of Brazil and Belgium seem to do perfectly well without giving debt this artificial advantage
Maybe I was brainwashed by the infamous Chicago School, but I can think of a lot of reasons for the tax treatment of debt. After all, let's think about why corporate debt is deductible, while personal interest largely isn't. Personal income is defined for tax purposes, broadly, as what you were paid this year. But corporate income is defined as what you were paid (revenue) minus what you had to pay others.
That's because government generally assumes that the operating expense and capital requirements for an individual are roughly the same from person to person--you may think you need a 5,000 square foot McMansion and a power boat, but Uncle Sam disagrees. On the other hand, companies differ greatly. Firms like Apple or Kelly Services really do have very different operating and capital structures from Alcoa or Caterpillar for good reason. Heavy industrial companies need more capital to make new investments, and it can make good sense to match the duration of the financing to the expected life of the asset. That's accomplished by borrowing money, not floating a new stock issue or trying to accumulate enough retained earnings to keep up with your competitors. On the other hand, service and software companies basically need some computers and an office lease--their assets are mostly brands, patents or copyrights, and what economists call "firm-specific human capital", which is to say, processes and know-how.
There are two core problems with simply eliminating the tax deduction on interest. The first is conceptual: the corporate income tax is supposed to tax, well, corporate income, which is to say, profits. I doubt Felix would advocate treating debt payments on the financial statements as something other than an expense--replacing net income with EBIT, say. Money that has gone to lenders or the taxman is actually gone; shareholders don't get it. It would be grossly misleading to tell shareholders, "We made $100 million this year" if you've left out the $120 million in debt and tax payments you had to make.
The second is that this would make companies that do use debt finance much more risky. Companies with big capital requirements could find that in a downturn, they suddenly had large debt payments, negative cash flow, and a sizeable tax bill. As a first principle, the tax code should aim to avoid unnecessarily pushing vulnerable firms into bankruptcy.
There's actually a third problem: it's possible to create debt-like instruments that are deductible--complicated lease arrangements, for example. We really don't need any more rules for companies to game.
This seems to me like a better argument for making dividends deductible, or capping the amount of debt that's deductible. But to drag out my regular hobby-horse, it's an even better reason for getting rid of the corporate income tax entirely, along with the special tax rates for capital gains and dividends. Tax income once, when it's distributed to the owners, and then tax that income just like any other kind of income, regardless of source, so that Paris Hilton pays a higher rate on her corporate-derived income than your middle-class grandmother. Then let companies decide which mode of financing makes the most sense for their capital structure, rather than their tax bill.
It's true that this would discourage dodgy LBOs and other temptations to inappropriately load up your company on debt. The problem is, it would also discourage appropriate debt. I think there are better ways to skin this cat.
"Tax income once, when it's distributed to the owners, and then tax that income just like any other kind of income, regardless of source, so that Paris Hilton pays a higher rate on her corporate-derived income than your middle-class grandmother."
The problem with that is that companies would be encouraged to build up cash hordes rather than pay dividends. The cash horde helps build up the stock price. Both Apple and Microsoft have large hordes that help bolster the stock.
Yes. Can you imagine the disaster if all the financial firms, especially, had been encouraged to build up large hoards of cash, rather than debt?
ed-- I would submit that the excessive cash hoard is a drag on their stock prices rather than a benefit.
Having enough cash is a good thing, having more than enough is also a good thing --- having more cash than you could ever invest isnt considered a good thing.
I believe Paris Hilton was disinherited by a grandfather who disapproves of her lifestyle. I heard that most of her money comes from her appearances, not the Hilton fortune. That's going to a foundation run by nuns.
Interesting analysis. Eliminating corporate taxation, while logical from an economic perspective, creates too many problems from the perspective of tax enforcement. Without corporate taxation, you would have a form of tax shelter that simply would be too easy to utilize - everyone in the US would simply incorporate and then you would get into complex enforcement problems to police against potential abuses. In other words, the blogger would become "Megan McCardle, Inc." and she would enter into a contract with The Atlantic (or whoever her employer is) to pay her salary to the corporation. Now, when the corporation distributes the money to MM, it would still be taxable to her personally, but you can rest assured that lots of clever accountants and tax lawyers would be engaged in figuring out ways to defer or eliminate personal taxation in this scheme. The result would be that you would have to file both corporate and individual tax returns anyway, and the burdens that this would impose on IRS auditors would be enormous. The better solution is to simplify corporate taxation by eliminating most deductions and dramatically reduce the rates, which is what most of the rest of the world has done (and establish a VAT as your primary source of revenue). As to the issue of the deductibility of interest, it strikes me that we may not want to provide any special tax breaks to capital-intensive industries that are likely to have higher interest costs. For the most part, these will be older industries that probably don't represent the future of the American economy. I'm not sure that it is good policy to subsidize these industries by giving them an additional tax break that would not be available to more modern service-oriented industries. As I say, I'd be inclined to make both interest and dividends non-deductible, and lower the rates.
Megan,
Keeping in mind that most businesses are smaller and privately held, I have questions. For example, I run my consulting firm as a Sub-chapter S Corp. and my profits are taxed once as personal income. If we eliminated the corporate income tax then I would be able to retain earnings in my business, correct? If I later wanted to sell my business the value would include those retained earnings but the sale would allow me to pay capital gains taxes rather than income taxes?
Thus the idea of taxing income identically regardless of source.
No, because we eliminated the special treatment for capital gains at the same time. Otherwise, I agree, you'd have the mother of all tax shelters.
Almost ever nation taxes capital gains at rates far lower than those for personal income. Germany top income tax rates are 50.5% while capital gains taxes are only 28%, even in Denmark income taxes top out at 61% while capital gains top out at 45%. I would assume this is because it has been found beneficial to encourage business development in such a manner.
How would a reduction in the incentive to accumulate capital gains effect economic growth?
But almost every other nation has a corporate income tax. The former is supposed to offset the double taxation of the latter.
JMO: essentially, yes. But Megan wants to get rid of the capital gains / ordinary income rate distinction, as well, which would reduce your benefit from the reform. Of course, you would still get timing benefits: if you withdrew the earnings from the corporation to make a personal investment, you would have to make the investment with after-tax dollars, then pay tax on the earnings. If you kept the money in the corporation, you could make the investment with pre-tax dollars. You'd still have to pay tax on the earnings, and you'd still have to pay tax on the gains portion of the initial investment, but the tax on the initial investment would be deferred.
I'm not sure why it wouldn't make more sense to tax corporations at personal rates but grant them a dividends-distributed deduction. If nothing else, this would eliminate the more obvious deferral opportunities.
I run my consulting firm as a Sub-chapter S Corp. and my profits are taxed once as personal income. If we eliminated the corporate income tax then I would be able to retain earnings in my business, correct?
If you were willing to pay the accumulated earnings penalty tax that applies to earnings retained in a regular (C) corporation without business purpose. Same as now ("keep income in the business, sell at capital gain rates" is not a new idea.)
There are many other reasons why a small business wouldn't want to be a C corporation as well.
I must admit -- when I first read Salmon's piece on this (maybe a month ago) I was under the impression that he was advocating for making dividends tax deductible -- thus ending the disparate treatment in the tax code of equity and debt financing.
Then it hit me..... he was advocating making interest payments not deductible. I was stunned.
Of course, in Felix's world, such move would simply encourage the same overall level of investment to shift from the debt form to the equity form. Hurray!
It never dawns on him that raising the implicit cost of financing 35% or more would have a destructive effect on the US economy, particularly those industries with heavy capital needs (like manufacturers).
Megan is proposing one of two ideas, though it's unclear which: (1) treat all corporations like partnerships -- i.e., as pass-through entities -- where corporate income is imputed to individual shareholders, or (2) only tax dividends and realized capital gains. The first idea is administratively impossible, and the second idea would result in a tax-planning and tax shelter bonanza.
As to the first idea, in the course of a single tax year, a large corporation might have millions of shareholders of record. Many of those shareholders are day traders, or otherwise holding the shares for very short periods of time. How do you allocate a portion of a company's annual income to such "owners" without overwhelming paperwork? How does an individual trader figure out his or her aggregate distributive share of corporate income from the hundreds of corporations he or she owned for, variously, minutes, hours, days, weeks, or months of the tax year? Can you imagine the resources that would have to be poured into this tax reporting and enforcement? Talk about deadweight loss! And talk about killing liquidity in the financial markets.
As to the second idea, corporations would become the ultimate tax shelters. Earn all your income through a corporate shell, and don't distribute dividends and don't sell your shares. You can defer income taxation indefinitely. Due to the time value of money, tax planners often say that "a tax deferred is a tax avoided." This tax-planning bonanza would soon be met by complex anti-abuse rules from Congress and the IRS. Ultimately, the gaming of the system, the economic distortions, and the complexity that Megan wants to reduce would all be exacerbated.
If you want to equalize the tax treatment of debt and equity, better to exempt dividends and capital gains from the shareholder-level tax. In effect, corporations become withholding agents who remit the tax for the individual - without having to track each individual who owned a share for at least one second of the tax year. The one downside is that the income will not necessarily be taxed at the individual shareholder's tax rate. But I consider this more of an argument for equalizing (and flattening) tax rates than for trying to allocate shares of corporate income to individual shareholders or allowing indefinite deferral of tax through use of the corporate form.
all this can be fixed with one easy change:
make distributed dividends tax deductible to the corporation. If a company wants to hoard cash -- fine, they get no deduction. However, if they distribute dividends out to shareholders, its treated like a deductible expense.
If you wanted to bump up the dividend rate for individuals to ordinary income rates --- while not ideal, that would work.
Here is the non-tax benefits of such:
1. You would eliminate the unintended incentive the tax codes gives to finance with debt.
2. You likely create investor demand for companies to pay out dividends. To pay cash dividends, you need cash earnings and that provides a natural brake on creating earnings with non-cash accounting gymnastics.
George,
I have always thought your last idea was the correct one logistically and philosophically, but, as you point out, it isn't progressive, and this makes it politically impossible in our present environment.
On your other points, I don't really think the taxes can be deferred forever by using corporations as shelters for the simple reason that, at some point, everyone will want to spend their earnings or bequests. As long as you don't allow inheritances to be stepped up or treated in favorable ways by tax rates, all the income will be taxed eventually.
That one's easy to fix: require a substantial number of minority shareholders to avoid pass through, say 45%. No one is going to pass through 45% of their income to defer taxes on the other 55%.
Megan, so how would you treat a closely-held company with five 20% shareholders? If I'm reading you correctly, that type of structure could still game the system. And you would create an incentive for all partnerships and S corps with no majority shareholder to reincorporate as C corps to gain the benefit of deferral.
Maybe a better dividing line would be publicly-traded on a recognized exchange (or OTC) versus privately held? If you are publicly traded you get deferral, but if not you are treated as a pass-through?
This discussion, by the way, highlights another distortion besides the differential tax treatment of debt and equity -- the differential treatment of different business forms. If we were designing a tax system from scratch, we would think it bizarre to tax business profits differently based on the form of the business. Income from partnerships, corporations, LLCs, etc. should all be taxed the same way.
You could also set a minimum number of shareholders--say, 75. These are trivial tweaks; it doesn't invalidate the idea. Moreover, to the extent that it resembles a consumption tax, that might be a good thing.
I'm not sure any number of shareholders fixes this problem. It creates a huge incentive to be a stand-in shareholder for tax reasons.
George,
I believe the wide spread use of tax qualified retirement accounts means that your suggestion won't have much effect in offsetting the advantages of debt financing. Tax qualified investors will see no benefit from the exemption since they are taxed for the full amount withdrawn from the retirement account. They will continue to get substantial benefit from the corporations they own getting to deduct their interest expenses.
When you add to that the fact that most dividends and capital gains are taxed at lower rates than corporate income, even non-tax qualified investors get substantial benefit from the interest deduction compared to tax-free dividends.
@ateamrules: I'm not sure I'm following you. People investing through retirement accounts already obtain a rough form of equalization for their equity investments through either the front-end deduction (traditional IRA) or back-end yield-exemption (Roth IRA). In fact, what I'm proposing is analogous to an unrestricted Roth IRA for dividends and capital gains. (The use of tax-preferred accounts to invest in interest-bearing instruments is more problematic and leads to negative tax rates.) Tax-qualified investors SHOULDN'T see any benefit from the exemption because they're already receiving a benefit through the elimination of double taxation.
Also, I disagree with you that equity investors get a "substantial benefit" from the interest deduction. Corporations can push money out through deductible interest or nondeductible dividends. Because interest is deductible, lenders and bondholders can receive a pre-tax payment, whereas equity owners can only receive an after-tax payment. From $1 of gross profits, a corporation can either pay a lender $1 (because the payment will generate a $1 deduction and therefore zero profits and zero tax) or pay a shareholder 65 cents (because the dividend payment is nondeductible and therefore the $1 in profits leads to 35 cents of tax). The shareholder gets no corresponding benefit to the benefit the interest deduction confers upon the lender, which in my example equals 35 cents.
For the most part, corporations and individuals are on the same footing when it comes to deducting interest expense. Individuals are not allowed to deduct "personal interest" (with the exception of qualifying home mortgage indebtedness), which is interest incurred on non-business and non-investment debt. In theory, corporations shouldn't have such "personal" debt. (If a corporation does have personal debt, it's properly treated for tax purposes as a distribution to the owners or as compensation to the employees, depending on who benefits from the debt.)
Both corporations and individuals are able to deduct business related interest. If someone uses debt to buy a bunch of shoes for his shoe store, the interest is deductible whether the shoe store is a sole proprietorship or a corporation. Since debt has been one of the primary methods of financing business operations for hundreds of years -- long before the income tax existed -- its hard to say business debt exists only because interest expense is deductible.
One other thing making debt not deductible would do is kill smaller business, which often have to float credit to bigger businesses and the government due to slow payments. Smaller businesses typically can't get away with paying for stuff several months after they accept delivery, but this is a standard cash-management approach for bigger businesses and government. So, if you sell stuff into these worlds, you have to have a fairly big line of credit - and you charge these customers enough to cover the debt service, since otherwise you're giving them a fairly long 0% interest loan.
I used to run a small computer sales and service biz that frequently had several hundred thousand in receivables to various businesses and government entities. Cash on hand was typically barely enough to meet payroll, margins were thin, and we had to buy most of our supplies with terms of 15 days or less. Our customers typically demanded - and got - terms of 30 or more for business to six months for government.
Isn't there an easier argument to be made here? We allow corporations to deduct purchases for almost everything. Loans are, in essence, "leasing" money. If we're going to permit deductions for leasing copiers, cars, buildings, etc., what's the distinction to be made between that and leasing money? The only difference I can see is that other leases tend not to be for fungible goods, but in both cases you have to give whatever it was you rented back at the end of the lease.
I mean, sure, we can decide we want companies to spend money on x and not on y, but that's a different consideration altogether and one with a very different set of concerns than the more abstruse tax-geek reasons that are usually floated in such discussions.
The basic tax rule for deducting interest is simple and logical. If interest is incurred for...
[] Business purposes, it is deductible as a business expense. This is so for individuals who are unincorporated, most certainly.
[] Personal purposes, it is not deductible. This is so even for regular (C) corporations.
[] Investment purposes, it is deductible up to the amount of investment income, with the net being taxable.
Now, for individuals, expenses are presumed to be personal so you have to show interest was incurred for a business purpose to deduct it. But if that's so, it's no problem.
(A home mortgage is a personal expense and so wouldn't be deductible if Congress didn't write that deduction into the law, and probably it shouldn't be, but that's Congress gathering votes.)
A regular C corporation is a business entity so its interest expense is presumed to have a business purpose. But if the corporation is found to be incurring interest for the personal benefit of its owners there are going to be problems all around (including an unexpected, unwanted, undeductible taxable dividend to the owners).
Of course, Congress has written all kinds of crazy exceptions and loopholes and preferences into the law, starting with the mortgage interest deducion, to buy votes and keep the tax planning industry fully employed. But the basics are clear enough.
Why, as a matter of principle, would Salmon want to prohibit a deduction for a perfectly real and legitimate business expense when computing net taxable business income?
actually, as I understand it, personal interest expenses of all kinds used to be deductible; they killed the general deduction, but cutting the mortgage deduction was problematic for the same reasons it is today. so the home interest tax deduction is a legacy of the old broader deduction, not something that was created to subsidize homeownership.
I love the mortgage tax deduction. I borrow hundreds of thousands of dollars at an effective after-tax rate of 3% and put it in bonds earning 5-6%. Thanks, Uncle Sam!
But why wouldn't corporate interest payments be deductible? They're just another expense.
Personal income is taxable because the fundamental principle of the 16th Amendment is that all personal income is taxable -- period.
Personal interest isn't deductible because -- well, why would it be? It's just a voluntary use of personal income no different than any other.
The fact that people have different sized homes and boats has nothing to do with it, isn't even the remotest consideration.
"The government generally assumes that the operating expense and capital requirements for an individual are roughly the same from person to person"
Hello? Where'd that come from? That would be a pretty naive assumption! (Congress fully recognizes the differences in personal wealth and expenses, they are reflected in tax rates on net income, personal exemptions, etc. -- which has nothing at all to do with deductibility of interest.)
Business income is taxable as a subset of personal income (as are investment income, rental income, wages, royalties, etc.)
To accurately compute it, total business income must be determined by netting gross business income minus business expenses.
Business interest is deductible for exactly the same reason why cost of goods sold, rent, advertising, and other business expenses are deductible -- they are a cost of producing business income. If no deduction is allowed for them, then business income will be overstated. Taxable income will be miscounted.
The fact that "Kelly Services and Caterpillar have different operating and capital structures" has absolutely nothing to do with it. It is a matter of accurately calculating net income, that is all. The same principles apply if you are operating a candy stand reported on your personal tax return or are Microsoft.
Interest incurred for business reasons is deductible as a necessity of accurately calculating the total amount of business income that is taxable.
Interest incurred for personal reasons isn't deductible, because it's just another personal expenditure like any other.
That is all.
(The favors that Congress throws to Alcoa and Caterpillar etc. to support their capital structure, like those it throws to Microsoft, the housing industry, homeowners, and all the rest, come through the plethora of preferences in the tax code written for reasons we all understand -- not through the basic accounting principle that all legitimate business expenses, including business interest, must be deducted from gross business income to accurately calculate net business income.)
Megan,
Arguing in favor of eliminating the tax on dividends is essentially equivalent to Felix Salmon's point. Both dividends and interest would be taxed at the same rate in your proposal. You just happen to have a different preference for the overall level of taxation.
Insofar as your #2 on making debt finance more risky, you are essentially wrong. It wouldn't make it more risky, it would make it more expensive. It might incentive raising capital from stock over bonds, which might make financing, overall, more risky on average. But the tax liability would be fixed at the moment you take on the debt, so there is no way the tax bill is suddenly going to take a rational business person by surprise.
I think you're missing my point. The point is that a firm that is seriously cash flow negative could nevertheless get hit by a huge tax bill if interest payments were excluded from the income calculation, pushing it from losing money, to insolvency.
No.
Let me try to explain this again.
Current system: I'm a company, I issue some bonds for $x and agree to pay $y per year to bondholders.
Proposed system: I'm a company, I issue some bonds for $x and agree to pay $y to bondholders per year and $z to government per year.
The policy does make debt issue more expensive. If deductible, I pay $y per year, if not deductible I pay $y+$z per year. It doesn't increase risk because the cost is fixed at the time of issue and I'm well aware of it. It is equivalent in terms of solvency to issuing bonds at a higher rate, I'm just paying 2 recipients.
You would only have a case if currently issued debt was not grandfathered in.(so the tax is unexpected)
Even if you are talking about a type of debt where the payment floats based on some independent factor it is still equivalent to issuing that instrument at a higher rate. So all the risk was preexisting in type of debt issue.
It adds an additional liability but not an unexpected liability so cost of debt increases but risk of debt does not.
Hi Megan,
I was thinking a little bit more about why we're talking past each other and realized that I had excluded the possibility that you are making an even more basic mistake.
When you make unqualified, declarative statements like this:
"The second is that this would make companies that do use debt finance much more risky"
I assume you have a good reason for saying it. I now realize that if you were making an accurate statement you would say:
"The second is that this might make a small percentage of companies that use debt finance a little riskier"
Your mistake is to assume that when the cost structure of a company changes that the companies behavior doesn't change at all. Given that you claim to understand economics, you should know that this assumption is in complete disagreement with economic reasoning(of any school, freshwater, saltwater, whatever)
If the cost of debt finance goes up, firm behavior changes. They may pass the cost on to the creditor, to their customers, or to their suppliers. They may change the way they finance, they may finance less. They may decrease their capital investment.
There are a multitude of responses, almost none of them involve ignoring incentives and waiting for risk inherent in their revenue stream to fluctuate below their current costs.
Your statement was false, as you could not have known how companies will respond. (Incidentally it also uses reasoning which completely undermines your normal mode of analysis: law of unintended consequences)
I find it ironic that Megan warns that "maybe she was brainwashed by the infamous Chicago school" and then proceeds to write several paragraphs completely at odds with the teachings of the Chicago school.
Chicago theory being, among other things, that capital structure does not matter in valuing a firm, except (primarily) for tax reasons. So when she says that Apple and Catepillar have different business models, the latter demanding more debt than the former . . . well, she sure isn't being Chicago school.
Generally the idea that equity can't finance capital assets, even heavy industrial capital assets, is nonsense.
It's also silly to bring a company's financial statements into tax discussions. Tax accounting and financial accounting are already treated differently, which is why there exist balance sheet lines called "deferred tax assets" and "deferred tax liabilities." You could very easily treat debt differently for income statement and tax statement purposes.
I agree with the poster who says that Megan and Felix are talking about the same thing, except with different preferences as to the level of the corporate income tax.
Where did I say they made a difference in valuing the firm? The Chicago School teaches that they're the same for valuation purposes, not that they're the same for all purposes. You're also assuming a frictionless universe; in the real world, equity has substantial transaction costs that debt doesn't, which is why you don't float an IPO to pay for car repairs.
I'm not saying that debt and equity are the same. But Chicago school largely does.
If firm value does not depend on capital structure, then there is no reason for a firm to choose one over the other. Apple gains or loses nothing if it chooses debt over equity, and Catepillar nothing if it chooses equity over debt.
This is of course a patently oversimplified model, and subsequent work in financial economics (and law & Economics) has shown the story to be much more complex than that. But that work is not Chicago school. It's sort of nothing, at least no ideology that I know of.
Anyway, it doesn't really matter. It doesn't affect your substantive points. But if you are going to say something is Chicago school, then it shouldn't be sort of the opposite. You'd be better off eschewing the name-dropping and say something like "maybe I was brainwashed in business school . . . "
I suspect you responded to this post because it touches a nerve. That it's just one more example of the type of errors you frequently make when you try to delve into economic theory. You are on much sounder footing when you talk with the voice of a smart, reasonable person who reads a lot and follows issues closely. I wouldn't have read the CMS report had you not brought it to my attention (indeed, I probably won't anyway), and you have smart things to say about it.
I would suggest dropping the "economist" pretense. You aren't an economist; you never will be; and given the plethora of blogs written by actual economists, there is no market need for you to fill. You are best when being a "commentator" as described above, not when trying to talk about price theory or financial economics.
As for mortgage tax deductions. Funny thing. Not all economists think it should be eliminated. About half and half.
http://www.economist.com/blogs/freeexchange/2009/09/what_else_do_economists_disagr.cfm
Muzzy is just factually correct that neoclassical economics (i.e. the chicago school) theorizes that firms can freely choose their capital structure based on the relative tax advantages of interest, dividends, and capital gains. In fact, I'm pretty sure one of them got a nobel prize for saying just that.
Additionally, we already subsidize capital-intensive businesses with depreciation. Buy that factory, and you get to write it off for 20 years. Why do we need to also subsidize it relative to less capital intensive businesses? Doesn't this, ceterus paribus, cause us to overvalue capital as an input relative to labor?
The argument about what-will-shareholders-think neglects that all companies keep two sets of books, one for taxes, and one for accounting. GAAP standards for reporting accounting apply. But tax accounting can be, and often is completely different. Nobody is being confused about the reporting accounting, and people who don't understand financial statements don't play much of a role in setting securities prices.
Your other reason neglects the ex-ante effect of incentives. Firms would not overlever without the tax incentive to do so (or would be less likely to do so). That's the whole point.
Also, because the statutory tax rate in the US is higher than most (but not the effective tax rate, because we let politicians buy tax handouts), we're providing a -35% tax rate on debt, as opposed to rates in the negative teens or twenties in other countries that also subsidize debt.
This isn't a moral or ontological question about "what is income" and how many "times" should we tax a given dollar and during which transactions. We encourage corporations to borrow money at a rate of X%, to pay it out as dividends at Y%, and reinvest capital at Z% (capital gains tax). Policy should optimize F(X,Y,Z) for the public good. It seems very obvious to me that that X is currently too high.