Most people believe they understand what the term ‘profit’ means. It’s a measure of revenue minus costs. Actually, however, profit is a pretty meaningless term, because there is no clear definition of “cost”.
I was reminded of this when reading a debate in the Financial Times that discussed whether interest payments should continue to be tax deductible for businesses. The person opposed to ending the tax deductibility of interest made the following claim:
No — Tax should be paid on profits, not revenue
It is a fundamental principle in the UK, US and most other jurisdictions that companies pay taxes on their profits, not on revenues, writes Jonathan Blake. It follows that businesses should be able to subtract legitimate expenses when calculating their taxable profits. The deductibility of employee, rental or inventory costs is uncontroversial — why should interest costs be different?
Readers may recall that I believe we should tax consumption, not “profits”. But even if I’m wrong, I see no reason to make interest payments tax deductible.
Both interest and dividends are forms of capital income. Interest is capital income earned on debt, while dividends represent capital income earned on equity. It makes no sense at all to call one “profit” and the other a “cost”. Nor does it make any sense to make one tax deductible but not the other.
If neither interest nor dividends were tax deductible, then we could have lower tax rates and still achieve the same revenue target. Treating the two equally would also lead to relatively more equity and less debt, reducing the severity of financial crises and making our economy more stable.
Again, I’d prefer not taxing profits at all. But if we must tax profits, then let’s do so in the least inefficient way possible. And most of all, let’s not justify our highly flawed tax system by pointing to mythical concepts such as “profit”.
READER COMMENTS
Garrett
Jul 24 2020 at 3:16pm
“Profit” as in net income has actually become more meaningless over time as well, as the biggest revenue-generating assets in the economy have shifted from the tangible, which are capitalized and depreciated over time, to the intangible, which are immediately expensed.
Scott Sumner
Jul 25 2020 at 12:10am
Good point.
mbka
Jul 27 2020 at 1:24am
Well in that vein, the entire notion of capital has long failed to expand to the kinds of capital that actually matter today (the intangibles – human and social capital, brand value etc). Which is one reason why I find Piketty so misleading.
Steve
Jul 24 2020 at 3:28pm
I die a little inside when someone insists that a ‘non-profit’ firm is better (more moral?) than a for-profit firm. The tax code has cemented our misunderstanding of this concept for eternity.
Scott Sumner
Jul 25 2020 at 12:10am
I agree.
Matthias Görgens
Jul 25 2020 at 9:52pm
Especially since non-profits can overcharge customers just as well or badly as anyone else.
Thanks to principal agent problems most businesses are run for the benefit of management, not shareholders. Non-profits are even more susceptible.
Lorenzo from Oz
Jul 28 2020 at 1:52am
It is even worse than that, because tax-paid bureaucracies also get a moral tick. Yet, consider tax-funded health bureaucracies. Do their incomes go up or down if the general health of the citizenry improves or if it gets worse?
Obviously, it goes up as the general health of the citizenry deteriorates because their demand on health services goes up.
You are paid to do what makes your income go up, not what it says on the shingle. We are literally being forced to pay for institutions whose revenue incentive is to have us become chronically sicker. As the general metabolic health of the citizenry has been deteriorating for decades, health departments are delivering what they are actually paid to deliver.
But who worries about incentives? It is the declared intention of the expenditure that counts. Delivering health, is the declared intention. Only whack jobs think that the incentives might be what behaviour is selected for and there might be an institutional reason that the metabolic health of the US populace has deteriorated along with tax-payer funding of heath departments going up.
Of course, the “real” problem is that the American people have become greedy and lazy. But if you view welfare bureaucracies, including health bureaucracies, as, in effect, colonising society for their own profit, colonisers always blame the colonised when legitimising their extractive income flows.
But if it comes from taxes, it is not ‘profit’.
As someone who has lost 25kg (55lbs) over the last couple of years without changing my exercise regime, let me assure you what you eat affects profoundly how much you eat (and how much you move, though that’s less important).
Dietary guidelines and food plates that make little evolutionary sense, have remarkably little serious science behind them (but considerable corporate support), but are very good at generating deteriorating metabolic health in the citizenry are precisely what one would predict, if one assumes that what makes income go up is what people are paid to do.
Joseph
Jul 24 2020 at 3:59pm
Doesn’t it mean that the interest is taxed twice: once on the profit made by the lender, the other – that of the borrower?
Josh
Jul 24 2020 at 4:13pm
Along those lines, I’ve often wondered why I don’t see economists propose a corporate *revenue* tax instead of a profits tax?
Profit is maybe something like 10% of revenue on average so that means a 15% profits tax could be more like a 1.5% revenue tax to collect the same total tax revenue. In general a smaller rate distorts less.
Plus, compared to a higher rate profits tax, a revenue tax rewards efficient suppliers (those whose revenue greatly exceeds costs) and punishes inefficient ones (whose costs equal revenue). It seems like we should prefer that.
It’s also harder to game a revenue tax. Our profits tax needs all kinds of rules about what you can write off as a cost. All of that goes away with a revenue tax.
I’m sure there’s a reason no one who knows what they’re talking about talks about this. But to me (a non-economist) it seems way better than the current system.
Airman Spry Shark
Jul 24 2020 at 4:22pm
A revenue tax should be equivalent to a sales/consumption tax, so you’ve probably just not heard the former term since the latter is more standard.
Brandon Berg
Jul 25 2020 at 8:17am
Sales taxes are generally only paid on final consumption expenditures. When a furniture store buys a table and then sells it to a consumer, sales taxes are paid only once, not twice. Nor does the firm making the table pay sales taxes on the lumber used to make the table.
With a gross revenue tax, the tax would be paid on the lumber and other supplies, again when the retailer buys the finished table, and a third time when the consumer buys the table.
Garrett
Jul 25 2020 at 5:12pm
Right, so a gross revenue tax would incentivize firms to vertically integrate since there would be less transactions to tax.
Thomas Hutcheson
Jul 25 2020 at 7:50am
Why tax either revenue or profits? Tax (progressively) the consumption of owners.
Brandon Berg
Jul 25 2020 at 8:08am
Gross revenue taxes encourage inefficient vertical integration. Suppose that there’s a 10-step process for producing a product, with one firm handling each step and selling the intermediate product to the next firm, until the tenth firm produces the final product. The effective tax rate won’t be 10 x 1.5%, because each firm in the chain adds value, but it will be much greater than 1.5%. Let’s say 8%.
If all ten firms merge into one, the tax rate falls from 8% to 1.5%. If the merger it’s efficient, great. But if it were, the firms would probably have merged already. Suppose, instead, that the merger increases pre-tax production costs by 5%. Then the merger is inefficient, and we don’t want it to happen, but because it reduces the tax rate by more than 5%, it still makes sense to merge.
A related drawback is that a gross revenue tax artificially advantages firms that are large enough to purchase their suppliers, to the disadvantage of smaller firms.
It’s theoretically possible that reductions in tax complexity and compliance costs could make up for this, but I believe that the general consensus among economists is that gross revenue taxes aren’t a great idea. There are some states in the US that have such taxes, though. They’re sometimes called gross receipts taxes.
AJ
Jul 24 2020 at 4:39pm
It’s a measure of revenue minus expenses – which, like you said, most people understand. I think it’s only economists that find this meaningless.
Scott Sumner
Jul 25 2020 at 12:06am
Most people understand “expenses”? Really? I’ve never met an average person who has any idea what a business “expense” is. How about a three martini business lunch? How about “depreciation”? How about dividends?
Thomas Hutcheson
Jul 25 2020 at 7:55am
Yes (though taxed as an externality) Yes (not taxed as a tiny stop toward consumption taxation), No.
AJ
Jul 26 2020 at 10:32am
I think you’re over-complicating this. But even if the average person truly can’t grasp what an expense is, accountants have a technical definition of ‘expense’ and ‘revenue’, so that profit is actually a meaningful term.
Scott Sumner
Jul 26 2020 at 3:02pm
But these accounting definitions are completely arbitrary; it’s not like they are measuring some objective concept.
Of course there are accounting measures of profit—what does that prove?
AJ
Jul 26 2020 at 4:07pm
I guess, but if that’s the case, then economic concepts and definitions can be said to be completely arbitrary and not measure some objective concept as well, such as utility and economic profit.
Overall, I don’t understand the issue you’re raising. People inherently understand what ‘profit’ means. They understand ‘revenue’ from the equation in your original post, so they surely must understand some concept of ‘expense’, as well, even though it was mixed up with ‘cost’. You said there is no clear definition of ‘cost’ (presumably, opportunity cost (subjective, context specific, etc)), but most people don’t wrestle with this. I think it’s only the academic economist that gets ‘lost’ in the technical or precise definition of these types of things.
Brian M McDaniel
Jul 24 2020 at 4:46pm
It seems to me correct that interest is the cost of capital, expressed in the form of debt. To the extent that there is an inconsistency with the treatment of equity dividends, wouldn’t it be more logical to extend deductibility to dividend payments, as the cost of capital expressed in the form of equity?
This would also bring consistency with the growing use of partnerships, LLCs and other “flow-through” entities.
Scott Sumner
Jul 25 2020 at 1:40pm
Yes, moving toward a consumption tax would be much better.
Rajat
Jul 24 2020 at 9:28pm
I accept that both equity and debt are claims on a business and both dividends and interest are claims on the net income (EBIT, if you like) of a business. But equity represents ownership, and the value of that equity moves with the value of the business, whereas debt needs to be paid regardless. In that sense, it shares some similarities to a wages or electricity bill.
In any case, I think any sort of change like this has to be viewed in a more holistic or general equilibrium manner. I agree with you that neither form of capital income should be taxed – and under those conditions I agree with removing deductability – but the fact is that while interest may be tax deductible to a business (or investor in assets), it is currently also taxable in the hands of the recipient. In Australia, dividends are not tax deductible by businesses, but domestic shareholders receive a credit for the corporate tax paid by the business on its profits (this is known as ‘dividend imputation’). This helps even up the incentives businesses face in deciding how to finance themselves. The only problem with dividend imputation is that it biases firms towards paying out dividends rather than reinvesting profits. Of course, this is often a good thing, but it is obviously not a neutral treatment and some commentators blame it for Australia’s lack of technology and other innovative businesses – it’s just easier to invest in boring firms like the major banks or supermarkets who payout 80-90% of their profits.
If we just removed the deductibility of interest without removing tax on interest income, would you agree that it would reduce the demand for savings even further and push equilibrium real rates even further negative? That may not be a problem under an appropriate monetary policy regime, but it would be at the moment.
Scott Sumner
Jul 25 2020 at 12:10am
I was talking about a revenue neutral business tax reform, which would not extract any more revenue than before, but would treat debt and equity on a level playing field.
On your first point, “ownership” is also a slippery concept. Is my house owned by me if a bank has a mortgage on it? Yes and no.
Garrett
Jul 25 2020 at 5:22pm
I disagree with the implication that the value of debt doesn’t “move with the value of the business.” The main theoretical model for valuing the debt of a company is the Merton model, where it is modeled as a risk-free bond plus a short put option on the firm’s assets struck at the par value. The option has “delta,” or sensitivity to the value of the underlying, which increases as the value of the firm falls.
Brandon Berg
Jul 25 2020 at 8:42am
Does the fact that (at the federal level) the combined corporate income tax and personal dividend tax rates ((1 – 0.21) x (1 – .238) = 0.602, so about 40%) is roughly equal to the top federal tax rate on interest income (37%) mitigate the disparate treatment at the corporate level? I’m not really sure how to think about this.
Scott Sumner
Jul 25 2020 at 1:45pm
Good question. I don’t know all the intricacies of the current tax code, perhaps someone else can chime in.
Zeke5123
Jul 25 2020 at 7:38pm
Few points:
1. You correctly included the 3.8% surtax on qualified dividend income (not all income is QDI; some income is taxed at Norma ordinary income rates). But this amount also needs to be added to the 37% rate. Also would need to account for SALT.
2. Holders of debt may be sovereign wealth funds (ie 892 funds) that are exempt from US tax, foreign investors who qualify for 0% US WHT, US tax exempt entities (eg Calpers, Harvard), or US C Corps. Thus, there may not be symmetry on the other side.
3. There are numerous interest denial provisions that need to be considered (eg 163(j) which pre covid limited net interest deductions to 30% of EBITDA, limitations on high yield accrual debt, etc.).
4. Debt is also an effective way of returning capital in a tax free manner. So the math is a bit more complicated.
Mike Sproul
Jul 25 2020 at 11:07am
Profit=producer surplus=the area above the supply curve and below the price.
Cost=the area below the supply curve and left of the quantity produced.
Scott Sumner
Jul 25 2020 at 1:43pm
That’s reasonable, but of course it doesn’t equate to anything we do in the real world, as “cost” in economics is opportunity cost.
That’s going to be picking up mostly rents, isn’t it?
Mike Sproul
Jul 25 2020 at 1:52pm
Economic rent is just another name for producer surplus, which is another name for profit.
All costs are opportunity costs, and those are defined as the area below the supply curve.
Garrett
Jul 25 2020 at 5:27pm
Economic rent and producer surplus are different concepts. You’re thinking of normal profit, which is another term for producer surplus.
Mike Sproul
Jul 25 2020 at 6:38pm
The two terms are the same thing. Historically, rent was thought of as the return on land, but over time, rent came to be defined as the excess payment to a resource, over and above the amount needed to call it into employment. If land were viewed as having a vertical supply curve, then this excess payment would be seen as the rectangular area above the supply curve and below the price. But once we understand that the supply of land slopes up, then this excess payment becomes the triangular area above supply and below the price. We only have to appreciate the unimportance of triangle vs rectangle to understand that rent is just producer surplus, and both are identical to profit.
Garrett
Jul 25 2020 at 8:41pm
You seem knowledgeable. May I suggest editing the Wikipedia article that contradicts what you’re saying?
Mike Sproul
Jul 27 2020 at 1:39pm
Interesting idea, thanks. I’ll look into it once I get some free time.
Scott Sumner
Jul 26 2020 at 3:07pm
As I recall, the basic idea is that if something is actually a rent, it can be taxed without any welfare cost. In contrast, taxing capital goods income is inefficient.
If land can be created, then the optimal tax system taxes existing land and existing capital goods, but does not tax land and capital goods that are newly created after the tax system is first created. Such as recent additions to Manhattan, or the Netherlands.
Is that right?
robc
Jul 26 2020 at 3:24pm
I think its easier to think of the “new land” in the netherlands or manhattan as old land that used to be under water but now isnt (because that is the real case, no land was created, imo). The unimproved value would be low but non-zero.
Mike Sproul
Jul 26 2020 at 9:04pm
If the supply of some good (land, etc) is rising, then a flat tax will create a deadweight loss and reduce production. But if the tax is structured to only capture the triangular producer surplus, then there is no deadweight loss and no effect on output. Hence producer surplus=rent. Land in Manhattan and the Netherlands is in this category, since it has a rising supply curve.
Pre-existing land has a vertical supply curve, and naturally the (rectangular) producer surplus can be taxed without deadweight loss or any effect on output.
robc
Jul 27 2020 at 8:59am
See my post just above yours, this isn’t true. The Netherlands is merely converting low value “North Sea floor” land into high value “No sea here” land. So the supply curve is still vertical. A land value tax would tax it at its value as north sea floor. which is obviously non-zero.
This is probably not true, as that would be the case if the improvements were private improvements by the land owner. I would guess that the Netherlands government is probably responsible, so they could tax is at the unimproved lack-of-sea value (assuming a proper SLT).
Mike Sproul
Jul 27 2020 at 1:47pm
As the price of (dry) land rises, more dry land will be produced in both Manhattan and the Netherlands. That means that the supply curve of dry land slopes up. The total cost of that land is the area below supply and left of the quantity. The producer surplus=rent=the area above supply and below price, and the total revenue received by land sellers is the rectangle bounded by price on top and quantity on the right.
Zeke5123
Jul 25 2020 at 7:24pm
Dividends are discretionary; interest payments are not. If you are basing your income tax system on ability to pay, then that seems like a relevant consideration.
I tend to agree with the poster above who mentioned a dividends paid deduction (which we have — see eg RICs).
Matthias Görgens
Jul 25 2020 at 10:26pm
With a bit of financial engineering you can transform one into the other.
Eg you can pay your interest from subsidiaries that you can let selectively go bankrupt. (This all assumes that your counterparty is in on the game, and the two of you conspire to work out how to make effectively voluntary payments count as interest.)
Or perhaps you can have a clause that says that interest can either be paid in cash or in kind by the company. Here ‘in kind’ means the bond holder just gets more bonds, which are just pieces of paper the company can create at will.
Or you can issue very long running zero coupon bonds that the company can selectively redeem early, perhaps even just by buying them back on the secondary market. Stock buybacks are after all economically equivalent to dividends, so why not bond buybacks?
robc
Jul 26 2020 at 9:17am
Easier way…someone is offering a piece of equipment for $100k with 10% interest, payable in a year OR they offer it for $110k at 0% interest payable in a year.
Obviously, you can do the same calculation with a more regular payment stream too.
This depends on two things, financing coming from the producer and the item not having a fixed list price the IRS can reference.
So there are plenty of ways to game the system to convert the interest into part of the item expense. Which is the real point, the interest is an expense.
Matthias Görgens
Jul 25 2020 at 10:48pm
PS Despite what I wrote earlier, there is something to the voluntary va fixed nature of dividends and buybacks vs interest payments.
As Scott points out, the voluntary nature of dividends and buybacks gives issuers more flexibility and prevents investors from relying on the payment.
A tax system that rewards impairing your ability to pay by lowering your taxes might not be such a great idea.
For things like people’s income tax we might just bite the bullet here. But for a company’s decision on whether to finance themselves with bonds or equity, we might want to encourage more flexibility in the economy?
(You’d still keep taxes in line with some ability to pay by eg basing them on revenue. Unless you go all the way and don’t have companies pay taxes in the first place. And go with eg all consumption taxes and land taxes.)
Matthias Görgens
Jul 25 2020 at 10:40pm
Scott, to be a bit more precise (or pedantic), profit is not so much meaningless as underdefined.
Very similar to eg inflation. The concept of inflation makes sense and is ‘real’, but precise measurements depend on exactly how you operationalise and on et judgement calls about how to deal with hedonic adjustment.
Similar to talk precisely about profit you have to nail down exactly what definition of cost you are using.
So in pseudo mathematical language, ‘profit’ is not a straight up defined term but a function: you feed in your precise definition of costs, and out comes a precise definition of profit.
‘Functions’ might be a useful way to think about this. Because it shows that for all its problems, ‘profit’ is still a much more concrete term than ‘general will’ or Derrida’s différance.
Scott Sumner
Jul 26 2020 at 3:11pm
There are certainly many ways to define “profit”, but I’d say one that includes capital income from equity but not capital income from debt is especially arbitrary and unhelpful.
Similarly, a definition of inflation that included stock prices would be unhelpful.
robc
Jul 26 2020 at 3:08am
Articles like this are a great argument in favor of the single land tax.
Consumption tax has one of the same problems as the income tax… dead weight loss.
Michael Pettengill
Jul 27 2020 at 3:32am
Providing for the general welfare by paying workers directly or indirectly to consume production is a deadweight loss?
Does that mean you want all workers eliminated? Or all consumers? Should babies and children be eliminated? Old and disabled and orphans be eliminated? Aren’t they deadweight losses?
Bob Murphy
Jul 26 2020 at 12:37pm
Scott, I know I asked you this when you said “income” was a meaningless term, but please remind us: Do you think accountants do anything useful?
Scott Sumner
Jul 26 2020 at 3:09pm
Yes, they organize financial data. Much of what they do is to organize it in a way that meshes with the demand of our tax authorities.
Louis Le Marquand
Jul 26 2020 at 6:25pm
On profit:
A business makes a financial profit/loss when its revenues (sales) are more/less than its costs (expenses).
On cost:
The “cost” of a good or service are the values it is exchanged for.
In a division of labour, market society, goods are exchanged for money, which is then later exchanged for other goods or services.
The money cost of a good/service is what it is exchanged for in terms of money.
Michael Pettengill
Jul 27 2020 at 3:28am
Only the US of all industrial economies does not primarily tax consumption.
That’s what the VAT is.
The tax revenue on consumption flows to the source of production, within the nation consuming the production. Production imported and consumed is taxed where the importer is based, and then along the distribution chain. Production exported for consumption elsewhere is not taxed.
This promotes local government maximizing production by training workers, fostering R&D, providing exceptional infrastructure (transport, telecommunications, water, sewer, etc).
Producers are not burdened by general taxes on exports. Which Trump and his economic advisors call cheating.
VAT is a clever renaming of sales tax. Canada, for example calls its VAT a GST, global sales tax.
In either case, the VAT is the ultimate elimination of all tax dodges and lowering the corporate tax rate, because labor costs are the biggest tax dodge of all.
art andeassen
Jul 27 2020 at 9:58am
Michael: Is the VAT applied to imports at the border which are then immediately VATed when the imported good is sold to the next purchaser? If so the VAT is then similar to a tariff isn’t it? You also don’t mention that VATs are refunded to the exporters which is similar to a subsidy. Are imports to the US not subjected to the taxes domestic manufactures have paid and are US exporters not rebated taxes paid in the US?
Scott Sumner
Jul 27 2020 at 12:51pm
art, No, that’s wrong on both counts. The VAT is not like a tariff because it applies equally to foreign and domestic goods. An actual tariff distorts the economy by driving a wedge between the prices of domestic and imported goods.
And a VAT does not subsidize exports.
Both of my claims are completely uncontroversial among experts.
Comments are closed.