Ask HN: Why do we tax profits and not losses?

4 points by daenz ↗ HN
It seems backwards to me to tax profits and not losses. We want to incentivize profits, not losses, so why are we taxing them to incentivize the opposite? Aren't we encouraging shady accounting practices by incentivizing companies to disguise profits as losses?

Would it mean that companies take fewer risks if the bill for a failed project would be higher? Does that stifle innovation?

To preemptively qualify: I don't think taxes on profits should go away completely, but perhaps the tax burden could be shifted, so that some % comes from profits, and some % comes from losses, such that the net revenue to the government is the same, while still employing this new incentivization structure.

14 comments

[ 5.2 ms ] story [ 22.4 ms ] thread
You don’t want to compound losses by taxing losses.

Profit taxing presumably takes what can be spared.

>You don’t want to compound losses by taxing losses.

From an incentivization point of view, I am making the argument that you do. And businesses, knowing this, would build it into the projected risk of a project.

You’d further punish the risk of innovation, I’m not sure why you’d want to do that.

Also, how do you calculate it?

I just don’t get it I guess.

Here's a quote[1] from Adam Smith about the nature of effective taxes:

"Good taxes meet four major criteria. They are (1) proportionate to incomes or abilities to pay (2) certain rather than arbitrary (3) payable at times and in ways convenient to the taxpayers and (4) cheap to administer and collect."

Is it profit or revenue that is taxed, generally?

[1] - https://en.wikipedia.org/wiki/Theories_of_taxation

Not only are losses not taxed, but they are often deductible against other taxable (profit) income. I believe this is mostly out of a sense of fairness, and also because profitable business activities often start out as unprofitable, and we don't want to discourage that start-up phase, or we might not ever get profitable businesses.

edit: losses are not deductible in perpetuity, however. For small businesses, after several years of losses, the business is presumed to be a not-for-profit ("hobby") activity for which losses are not deductible. And larger businesses can only continue to generate losses by investors dumping more money in only to lose it. (U.S tax system)

>We want to incentivize profits

There is already plenty of incentive without any government intervention.

And as others have pointed out, taxes are heavily based on a ability to pay. For example, while cancelation of debt (forgiven loans) result in taxable income (U.S), there are exceptions for people who are insolvent (and therefore obviously can't really pay the tax).

The point about the start-up phase is a great argument against my question. I mentioned in another post, but for this idea to not impact startups very negatively, I suppose there would have to be an exemption for new businesses.

>There is already plenty of incentive without any government intervention.

That's true, but there is also an incentive to hide profits to avoid taxes. I'm not very knowledgeable in this, but my understanding is that there are a number of accounting tricks that many different industries can use to disguise profits as losses, which feeds directly into corruption.

> I'm not very knowledgeable in this, but my understanding is that there are a number of accounting tricks that many different industries can use to disguise profits as losses, which feeds directly into corruption.

I'm not knowledgeable about this either but I think you're confused. It sounds like you're saying you believe it's easy with "accounting tricks" to flip the sign on net income so that a profit of $X looks like a loss of $X, more so than to make a profit of $X look like a profit of $0 (which is what your proposal incentivizes). I find this very doubtful. There's always incentive to hide profits, because they have to be shared. Then you could imagine an "incentive structure" that pays companies a bonus based on profits. Obviously, that would incentivize exaggerating profits, and people would look for accounting tricks to do that.

Companies wouldn't take risks at all if you're artificially increasing the financial impact of failure. And no one would want to give loans to businesses because if the business fails there would even less chance of getting paid anything back.
I disagree with the first sentence, in particular "at all." Companies take risks because they seek profits. If their profits are not sufficient, they must take risks. If their profits are sufficient and they are risk adverse, they wouldn't take the risk in the first place, regardless of some tax on the loss.

To your point though, I agree companies could be discouraged from higher risk projects, if the overall cost (and therefore tax bill) is too high for failure.

Wrt to loans, I agree loans would be more risky. There would have to be an exception for taxes on new businesses, because they're going to fail at a lot of things on the road to profitability, and we should try not to let that kill them.

Because we don't want to send the message to future entrepreneurs "If you take a risk and it goes badly, we will crucify you,"?
Yes, it is backwards.

But most economists agree that companies should not be taxed. Well, they should be taxed for things like carbon emitted, waste, resources used, value-added, etc.., but their income and profits shouldn’t be taxed.

Economists see the taxes paid by companies as being paid by their owners. So, if a poor person owns a share and you tax the company, the poor person pays taxes. Better to have the company disperse profits to their owners and then tax the rich owners.

https://www.npr.org/sections/money/2012/07/19/157047211/six-...

If we did tax losses, Donald Trump might actually have to pay
Losses are already disincentivized by virtue of their being losses. Your proposal would merely create additional risk for any action. I suppose the extreme would be to forcibly liquidate any company that ever loses money. If you can come up with an argument against doing that, I think the same argument would most likely refute your proposal without modification.