There are some misunderstandings in the comments that seem to stem from not having read the section, so I thought it was worth referencing the actual text [0]. It's quite short and easy to read.
The most important bits:
* Subsection (a) requires amortizing "Specified research or experimental expenditures" over 5 years (paragraph (2)) instead of deducting them (paragraph (1))
* Paragraph (c)(3) is a Special Rule that requires that all software development expenses be counted as a "research or experimental expenditure".
That's it. All software expenses must be treated as research and experimental expenses, and no research and experimental expense can be deducted instead of amortized. Ergo, all software expenses must be amortized over 5 years.
I strongly recommend reading the section before forming an opinion. It really is quite unambiguous and is unambiguously bad for anyone who builds software and especially for companies that aren't yet thoroughly established in their space (i.e. startups).
Also note that this makes Software a special case of R&D. It's the only form of R&D that Section 174 requires you to categorize as such and therefore amortize.
It had a huge impact on my personally, I'm a small R&D shop and basically I have had to end all risky long-term research projects.
In addition to the research costs, I'd also have to pay taxes on the research costs mostly up-front. Significantly, if the project doesn't work out, I'm still out of pocket for the tax money. It's a penalty for taking a risk, and it kneecaps American innovators in a globally competitive technology race.
The rules are even worse than the article notes because it double-dings open source developers. See Section 6.4 of https://www.irs.gov/pub/irs-drop/n-23-63.pdf. The relevant bit is here:
> "However, even if the research provider does not bear financial risk under the terms of the contract with the research recipient, if the research provider has a right to use any resulting SRE product ... costs paid or incurred by the research provider that are incident to the SRE activities performed by the research provider under the contract are SRE expenditures of the research provider for which no deduction is allowed ..."
The rule as written means contractors who write Windows drivers could deduct their expenses (as they would have no residual rights to a closed-source work product), but contractors who write Linux drivers may not (as they would have some rights to open-source Linux drivers).
It makes software temporarily 16.7% more expensive in year one if you’re operating a profitable company, but you do eventually get to deduct that over time. Pay 8% on a 4 year loan and that drops to ~4%.
As has been said repeatedly in this thread, this change is purely a boon for existing big tech companies that now have even less to worry about from startups. It takes a startup 5 years before they'll be playing on an even field with big tech.
> if you’re operating a profitable company
You keep saying this across this thread, and keep ignoring that Section 174 has now redefined "profitable" for tax purposes to include companies who:
* Are in year 1 with no history of expenses to draw on.
* Have spent <900% of their year 1 revenue on software development expenses.
i.e., a startup that earns $1mil and spent $8mil in software dev expenses is only able to deduct 10% * 8mil = $800k of expenses, which means that as far as the government is concerned they made a profit of $200k and owe taxes on that on top of their already-net-loss of $7mil.
You can keep ignoring this fact, but ignoring it doesn't help your case. If you want to argue that this is fine and dandy you need to explain why the above math doesn't prevent new companies from competing on fair terms with big tech.
> The rule as written means contractors who write Windows drivers could deduct their expenses (as they would have no residual rights to a closed-source work product), but contractors who write Linux drivers may not (as they would have some rights to open-source Linux drivers).
Is it just me or are you conflating two orthogonal things?
An open-source Windows driver would have the same issue, no? And a closed-source proprietary Linux driver privately written for some company wouldn't have this issue either, right?
I could see it being inferred that way but, the way I read it, they are not meant as unilateral facts. Rather, they serve as rhetorical examples of where you might find contractors doing similar work, but where the one more in service of "public good" is taxed higher because it's open source. Strictly speaking, Windows bits are not all closed source and there exist closed source Linux bits. But it's not a point that really matters in the context of the conversation.
I think it's fair to use Windows and Linux as stand-ins for closed vs open source because it's a very accessible example. And knowing the technicalities clearly doesn't undermine the argument.
> I think it's fair to use Windows and Linux as stand-ins for closed vs open source because it's a very accessible example
We're talking about businesses here that would struggle with these tax rules. Which I guess is, mainly, contractors or startups. How common is it for them to write open-source drivers vs. closed-source ones? I would've imagined the majority of drivers in such cases are closed-source, on every platform. But I would find it interesting to hear if things are somehow different on Linux.
You are correct. I picked this example under the general assumption that the Windows driver would be closed-source, but you are correct that it doesn't necessarily have to be closed source.
The problem goes with the license, not with the OS.
You're right, the law text doesn't specifically call out the Windows operating system or the Linux operating system. The example you gave of Open Source Windows drivers is valid.
The Grandparent's point about that "it double-dings open source developers" is still correct and poignant even with this clarification.
Work-for-hire open source contributions often already bear a copyright holder of the entity paying for the work. The problem isn't who is the copyright holder.
The problem is that the license assigned says that anyone is free to use the code. Anyone is a set of people that includes the contributor, which then triggers the interpretation that the research is incrementally in the contributor's benefit and thus disqualified from preferential tax treatment.
You'd need a custom license where everyone in the world could use the results except for the contributor, and then like, a source control system that hides the source files from the contributor's view of the repository.
The USA hasn’t managed to completely impose their idea of intellectual property on everyone yet. Some countries you can’t sign away authorship even if you can commercial rights.
We are small and so have been on a hiring freeze since 2022. I’d like to hire but the upfront cost is high.
For those around when this went into effect many business owners were surprised. Our accountants told us they seriously thought congress would fix this before it went into effect.
Often the way this works is that some time bomb is added to the tax code so that forecasts for future tax revenue will be higher (justifying more spending in the short term) and congress then needs to remember to remove the time bomb before it blows up. So it isn’t that uncommon for these things to be reverted.
Everyone knows it’s a time bomb but the only way we can get an omnibus funding bill through Congress is reconciliation. That’s how our government has been functioning since like 1980 and almost nothing has been “reverted” since then except for the time bombs required to get through the reconciliation process.
Or in this case, not justifying extra spending, but justifying extra tax cuts for incomes above a half million dollars per year, inheritances of more than 5 million dollars, corporate profits, private jet flights, and a variety of other special tax cuts, mostly aimed at wealthy GOP donors.
There's another word for "educating" people until they reach the decision you want them to.
Brainwashing.
You know it's completely possible that people have a different outlook or opinion or perspective on things and that is why they disagree with you, not necessarily a lack of education?
Some people think these are good ideas and they vote or welcome them. Some people think they are bad ideas and they vote or oppose them.
No, education is not brainwashing. These terms have different definitions. One is by definition good, one bad. Words matter, definitions of words matter.
I’d love an education system that only teaches scientific consensus, and leaves moral conclusions to parents and households. However I’m sure you’d appreciate that’s not what we have.
Fortunately our system is setup such that passionate folks like you can work to effect change. Go do it - volunteer for your local PTA, run for school board positions, show up to public hearings. Be the change you want to see in the world. God Speed my friend.
Having the ideology of the majority taught in schools is the outcome of a strictly democratic process like the one you’re describing. I’m suggesting that the separation between church and state be extended to any ideological teaching.
I’m directly responding to your point. Ideological education is brainwashing. In fact “brainwashing” is usually just another way of saying “an ideological education I disagree with”, also known as “indoctrination”.
For example: What’s the right way to live? How to relate to others? On what basis do we cooperate? What are/are not the overall goals of society? What’s the meaning of life? What is it to be a good person?
This is one of the worst things MAGA has done. Tech startups are the source of so much of our wealth, and this makes it very challenging to ever build one.
I can’t believe this still exists, and no one has changed it. We truly are governed by morons
for anyone curious, this wasn’t specifically trump, but it was indeed a republican congress bill. texas republican was the initial sponsor and then republicans lined up to cosponsor.
this was done to fuel their tax cuts to a small group of a certain people.
as we are seeing now on a number of issues, sure things can be rolled back, but that doesn't mean a return to normal.
Yes, you can kick over a bee hive, then pick it up and set it back upright, but you are not going to put all the bees back in immediately. There are long term consequences.
The way this is "fixed" right now, every five years we need another round of republican government to make things great again. If only enough democrats cared to fix this.
A dead post below says it was biden and somehow Obama, but sibling reply link says it passed into law in 2017, not 2022, the first year it went into effect I think.
The article is pretty clear that the law was from 2017 had a scheduled start date of 2022. Although probably not the actual intended effect, it does have the effect of confusing who would be responsible for it by those who don't read past the effective date.
> Although probably not the actual intended effect
This is actually a really common intention in laws like this. Get the tax cuts during your term, and then kick the can down the road so your successor's term is marred by the bad law. If your successor wants to fix it, they need to pass a different tax to recoup the costs, and incur the publicity of "raising taxes".
I think it’s partly that and partly the fact that tax bills tend to be scored on a ten year time window. Note that this law doesn’t actually change the amount of tax that software companies can deduct, it just requires them to spread the deduction over several years. So if you’re scoring your new tax bill on a ten year window and five years into the bill this thing kicks in, then it looks like more tax is being collected in years 5-10. But that’s just an illusion because all the deductions are still there, they’re just being pushed out beyond the end of the window where they don’t “count”. At least this is my understanding.
This is 100% the intended effect of leaving a bomb like this unaddressed then stalling all legislation halfway through your opponent’s term. You make them look incompetent and then you complain about how they never fix anything, even though they never get anything done because you’ve stopped cooperating. And even though you’ve stopped you can keep complaining about how they always include things you don’t like in legislation so that’s why you never cooperate. The solution is simple in your eyes: just do exactly what you want and only what you want and then you’ll cooperate. This is how a lot of a certain party was talking on the news from 2021 to 2025 when they were interviewed.
As I understand accounting, this means that reported profits would be higher, and therefore incur more corporate income tax liability. Cash flow isn't effected besides tax.
A startup isn't likely to be making a profit yet, under either accounting rule. Is there a benefit to reporting a larger loss?
My first thought is that this effects Google and suchlike, not startups. But... assuming steady state "r&d" expenditure... it's not that much. Everything gets deducted within 5 years anyway.
So... maybe this hinders more modestly profitable, and fast growing companies most. Those that can't afford to carry 5 years worth of paper profits as easily.
Otoh... I am curious about how the difference between r&d expenses and operational ones are determined irl.
This should be quantifiable. How much extra assets are software companies actually booking?
It seems questionable that this "silent killer" had actually affected employment so much.
> A startup isn't likely to be making a profit yet, under either accounting rule. Is there a benefit to reporting a larger loss?
As an example, A two person software startup; both drawing a salary, each making $100,000 per year. Each doing things related to software development.
Startup brings in 200,000K in revenue.
Under pre Section 174 changes, the profit is zero. Both salaries are expensible in the year they were incurred.
Post Section 174, the profit is now $160,000 each year. Now they pay taxes on $160,000, even though they literally have no money left over because revenues equaled expenditures.
At 25% tax rate, that’s $40,000 in taxes, for a business that made literally no money.
That’s why this is so devastating to small software businesses; unless you’re highly profitable and have cash reserves, this change hits hard.
This example only really has the emotional impact it does because of the specific numbers used, but doesn't really generalize for an arbitrary N.
Clearly if two software engineers build a product that brings in $10M, and each pay themselves $5M, it doesn't seem so outrageous that the can't really claim they're running "a business that made literally no money." Clearly in this second example the problem is that the engineers are paying themselves way too high given the return on their efforts.
What this means is that software engineers will be required to bring in more value to justify their high pay. In your example, it simply means that a software engineer that brings in $100,000 of value to the company, probably shouldn't be paid $100,000.
This seems entirely reasonable to me, and doubly so when I consider how many large corporate teams (who I think will ultimately be impacted more than startups) has huge numbers of highly paid engineers not doing all that much.
In most startups I've worked in it was pretty common for engineers to be delivering multiples of their cost in value, and in every big company I've worked in, it was very common to be delivering fractions of one's cost in value.
> Post Section 174, the profit is now $160,000 each year. Now they pay taxes on $160,000, even though they literally have no money left over because revenues equaled expenditures.
They have the $200k they pulled from their startup, far more than what most people earn. If you make enough to pay yourself $100k then you make enough to pay taxes.
That was my first thought as well, but on second thought I can see how this might cause problems:
For established profitable software companies there was a cliff edge in 2022 when this change kicked in. Staff costs for previous years had already been fully expensed while only 20% of the current year's costs could be deducted.
Second, any sudden increase in research expenditures is now discouraged. This could make companies less nimble.
For unprofitable startups it could cause issues during a phase of very high revenue growth. They could suddenly be liable to pay corporation tax in spite of the fact that they are not profitable in any reasonable sense of the word. It would smooth out later, but that may be too late for some.
What I do not believe for a second is that this is causing major job losses. Companies like Microsoft or Meta do not reduce research or software development just because there is a temporary tax hit. It could be an extra incentive for an efficiency drive I guess.
> For unprofitable startups it could cause issues during a phase of very high revenue growth.
So I guess my most question is "how this work irl?"
Say a new startup raises money and hires 20 people. Pays $5m in salaries, office space and such. All 20 people are developing a software product. Are 100% of this startups expenses amorotized?
Then they sell the product. They receive $2m in revenue. What does the P&L look like.
If they hire 20 devs in their first year paying $5m in salaries, only $1m or $500k (if the mid-year convention applies) would count as a business expense in that first year.
If their revenue was $2m, that would leave them with $1m (or $1.5m) of taxable profits unless that was eaten by other costs.
It doesn't have to be a problem, but if revenue grows fast and they go on another hiring spree in the following year then it could become a problem.
That said, if revenue grows so fast, it seems likely that they would have huge marketing and sales costs that could be expensed immediately. So maybe this isn't really a problem for many startups. I'm not sure.
1 million profit, while they have 3 million negative cashflow, that's exactly the problem. They can only take 20% of that 5 million in R&D investment as depreciation in the first year.
Your analysis is correct, but most software companies were mostly profitable or fast-growing. For every Google, there’s 1000 wordpress vendors you’ve never heard of.
In another year the initial shock will stabilize, but any growth now has a 5-year tax hit attached. And even Facebook doesn’t want to pay that if it doesn’t have to.
Google was reportedly amortizing (by choice) long before this was in effect, so while it might “affect them”, in practice it’s likely business as usual.
It depends on the department. My salary (in a mature product) was already amortized - I suspect the same is true of all their other mature products like Search, Maps, GMail, Chrome, YouTube, etc. But I think they were deducting salaries in the more research-like areas like Gemini, Jax, Assistant, etc. So there is net still a fairly large charge related to it, even if it isn't as large as it could be.
I'm not an accountant, but as I understand it, you don't pay taxes on profits, but on revenue.
So previously, some 20% of all revenue would be owned as corporate income tax, and startups would deduct it all as they're spending much more on R&D than they owe in corporate income tax. But with this tax change, the deduction would be much lower (80% lower IIUC).
Yes, but the main thing here is that ALL software development is now "profit" in the short term. In theory you've developed a capital good that benefits you over time, hence the amortization.
Simplified 2021 example before 174:
100k Revenue
100k Software Dev Costs
No profit or tax
Simplified 2022 example after 174:
100k Revenue
100k Software Dev Costs
90k "profit"
18.9k taxes
Above example is year one of suddenly having these taxes, because if your software costs are the same or lower over time it gets easier. It's just extremely painful for smaller and especially fast growing companies like startups without a lot of cash, especially when interest rates are so high.
Accountants: If I am wrong about the above, please correct me
Large companies always find a way to not pay taxes. It's the little guys that end up paying (a lot!) more, to the extend that it cripples and kills them. But transformative innovation happens with the little guys. As a result, this tax change cements monopolies for megacorps. They will be fine and still pay nothing.
The little guy always pays all taxes. Corporate tax is just a way to palatably shift tax burden to the low and middle classes and away from the owner class. It is pure double speak.
That can't be right. It definitely isn't in my country.
If own a car dealership, and I sell a car for $50,000 that I bought from the manufacturer for $40,000, surely I would pay tax on the $10,000 profit? The tax on the the full $50,000 revenue might exceed my profit!
We don't talk about this enough. International R&D is not offshoring of call-centers to India. International R&D is the IP for the next generation of global communication standards being owned by US-based or foreign corporations, because international (e.g. Canadian, European) standards experts/developers become un-affordable for US-based corporations and are forced to work for our "adversaries" instead. Crazy.
I’m not sure this is exactly true. If your foreign workers are a service contract then those are services expenses immediately deductible. Same if you are using local service contracts. My understanding is this creates a drag for companies that want to hire f/t.
Foreign workers are to my knowledge effectively always a service contract, since it's pretty complicated (if even possible) to hire FTEs across borders without subsidiaries, which are expensive to maintain.
I'm curious if contract work is really exempt, would look like a major loophole to me.
> Foreign workers are to my knowledge effectively always a service contract, since it's pretty complicated (if even possible) to hire FTEs across borders without subsidiaries, which are expensive to maintain.
It's impossible (yes, I'm being absolute) to hire an employee who lives in or outside the US who is not a citizen or doesn't have a green card. All employees must have an SSN and go through i9 verification, which requires in person verification of legal ability to work in the US.
The foreign developers I'm talking about are not US citizens and do not have green cards.
Their work is subject to 15 year amortization per section 174. Period.
Not if they are contractors. That's the point the parent commenter was making. All the reasons you list make it so they need to do so, instead of "hiring" them directly.
There are definitely gray areas to the law, but in my decades of experience dealing with lawyers, they won't steer you to do something very far over the line. I think the companies that do step far over the line, to game the system, are doing so knowing full well they are breaking the law, but they believe they are unlikely to be caught. And they're very likely right. You could separate CEO's/CFO's in to two camps: stay legal and do what you need to do to make the almighty $$. In the 80's the phrase "greed is good" was born, but the last 2 decades have really upped the ante on this.
It's not really targeted at tech, insomuch as at Democrats.
Everyone assumed it was a traditional accounting hack. But given the timing and the reinitialization, it's clearly political, not economic.
The code is a strategic time-bomb designed to cause a high-profile economic downturn during a presidential election cycle, specifically when the following president is a Democrat and Republicans have a house majority.
It was used to harm Biden's economy, and it will happen again in 2030 if the next president is a Democrat. While deferred, it will be spun as a major Trump "economic achievement" for the midterms, because companies will be able to afford to hire again.
The tech industry is merely high-profile fodder for extreme politics. It really is that petty.
The Democrats had control of the presidency and the house in 2022 when this provision first went into effect but had 2 fewer senators (1 fewer if you count the tie-breaking VP). Why didn't they try to change it? Is there some reason a change in the tax code like this can't be modified or repealed once its in place?
Generally, in tax bills they try to keep them "neutral" where any tax cuts or tax breaks are coupled with tax increases elsewhere BUT they tend to report the 10-year affect for whatever reason. This bill provided a ~30% cut in corporate tax on profits, with a delayed increase in tax cost on Software R&D pushed to the next term.
If the next party wants to reverse it, they'd have to find the money with an increase in tax - directly undoing it would be a ~50% increase in corporate tax rate, which (I guess?) would be a tough sell politically. Meanwhile, the tax code on software engineering sounds too niche to expend political capital on.
Either way, its another example of how corporate America is trading long-term growth (R&D, product development) for short term gain (lower taxes today).
> As a progressive, it seems like the Democrats always have Senate spoilers...
With Republicans usually being dominant in a number of states, if Democrats have a Senate majority, it is usually both narrow and dependent on a very small number of Democratic and/or Dem-leading moderate independent Senators from Republican-majority states who vote with the party on leadership, but are soft (or firmly opposed to the progressive preference) on a number of issues important to progressives.
If the US were approximately an equal democracy, this might be less of an issue.
No, the reason the "there is always an in-party Senate spoiler" effect (when they have a Senate majority) seems to be more true of Democrats is because it is more true of Democrats, and the reason is that when the two parties in rough balance by popular support (or even rough balance in Presidential electoral prospects, which has the same directional bias as the Senate but of lesser magnitude), the Republican Party has a systematic edge in dominance of states, which translates into a systematic advantage in the Senate, which means that when the Democrats have a Senate majority, it tends to have a decisive segment in red-state Democratic Senators who are unreliable on key priorities.
The issue being discussed in the Senate is not a symmetric issue resulting from near balance in support between the parties.
It’s also because republicans politically punish dissent, while it is more tolerated in the Democratic Party. The consequences of “disloyalty” are higher in the Republican Party.
This might change. After party leadership got 20% of democratic senators to vote for trump’s procedural blank check, the party’s approval rating dropped to 27%.
If it doesn’t change, I suspect the party will split.
Equal in voting rights. Gerrymandering has been perfected by Republicans. Through that they manage to dilute votes of the opposition. Other measures discourage voters likely to vote against them, like people who cannot easily take time off to vote in person or who have changed their name. Blocking rank choice and maintaining first past the post also disenfranchise third parties, and reinforces the power of incumbents.
Trump himself admitted it's better for Republicans when fewer people vote.
> Equal in voting rights. Gerrymandering has been perfected by Republicans. Through that they manage to dilute votes of the opposition.
This thread is talking about the Senate. The senate isn't gerrymandered. Both senators are state-wide races.
If you want to view it that way, you can view the senate as "pre-gerrymandered". But the last time that was an option was in 1959, and both of those are just "the entire area the US owned, but wasn't a state yet. To get senate gerrymandering, you have to go back to 1912 and the admission of New Mexico/Arizona.
> If you want to view it that way, you can view the senate as "pre-gerrymandered".
That is quite explicitly the history of the US Senate (and House), FWIW.
The Connecticut Compromise was reached to give low-populations states outsized legislative power in the senate. This is the main reason the senate exists.
Building on that, the 3/5th compromise was reached as part of this to give slave states outsized legislative power in the house.
The state of Maine used to be part of Massachusetts, but it was later set up as an independent state in order to increase the number of anti-slavery states in the senate (the Missouri compromise).
Gerrymandering can affect voter sentiment and trigger polling location changes during redistricting, both of which can affect voter turnout[1][2][3] (though the research doesn't seem conclusive on the effect).
And thinking about it more, though I haven't seen if there are studies on it: there are probably manpower/fundraising effects from gerrymandering.
If you're able to protect your political power in one area that probably better enables you to amass resources to use in the area you can't gerrymander.
But all that said, both parties practice gerrymandering and I don't think there's strong evidence of a significant advantage over a major party from current gerrymandering at the national level.
> On a percentage basis, over three times as many districts were competitive in states where independent commissions drew maps as in states where Republicans drew maps.
That’s just confusing cause and effect. If your seats are safe, you have no reason to agree to forming an independent commission. The same is true in both heavily blue and heavily red states. Are districts more competitive in states where Democrats draw maps? I don’t think so.
I don’t even know which group you mean, but “my group has good values and motivations, but the enemy group just values winning at any cost” is exactly what a total partisan who values winning at any cost would say.
The evidence is that independent commissions drawing maps makes for more competitive districts. Which party is most opposed to such commissions? Which party is gleefully dismantling all accountability and oversight positions and departments? Which party is openly inviting corruption and pardoning those they should be prosecuting?
No, this is a misunderstanding of the kind of taxation policy progressives tend to favor. Taxation on profit for businesses should be high, and taxation on upper tiers of individual income should be high, but taxation on funds businesses use to reinvest should be exempted or deductable. Basically the taxation we had in place after WW2 and on, with a steep corporate tax rate and more or less a maximum income for individuals. The R&D exemption removed in the 2017 bill, and discussed in the article, is key to that, because it encourages corporations to reinvest their income in building new products and paying workers rather than taking it directly as profit-- after all, at least they could reap the rewards (in growth and revenue) of the R&D later, instead of just giving the money to the government as taxes.
At first glance I support ... "social and economic equality" and "reforms to improve human conditions, combat corruption, and reduce inequality". Am I progressive?
If you ask me "should corporations pay more taxes?" I will say, yes. Famously so does Warren Buffet, is he also a progressive?
If you ask me, "hey should we gut tax incentives for R&D spending in the USA?" I will say, uhhh no? probably a bad choice?
Recently the progressives have latched on to culture war agendas against the wealthy, educated, white, male, straight and/or over the age of 35 crowd.
In other words, they have a popular agenda, but are political morons that are going to eventually wonder why they can’t break out of solidly blue districts.
I think that is a misrepresentation of the fundamental progressive position, which is to make progress but never at the cost of the marginalized. Because we historically make most progress at the cost of the marginalized it can feel limiting or even discriminatory when we make sure they don’t beat the brunt of continued progress.
There is nothing against the group you mention except that it might be the group that most fights against progress toward equality.
This tax is far more consequential for small companies than for large ones. It probably actually benefits larger companies because it hobbles competition.
This time bomb was created because the bill slashed the corporate tax rate from 35% to 21%. Maintaining the status quo would mean taxing big corporate America more than this bill does.
Providing spoilers was the explicitly designed purpose of the US Senate. It's not a one-sided problem - Senate spoilers are also why the Affordable Care Act didn't get repealed in 2017.
US Senator was an office initially designed to be selected by state legislatures rather than by direct popular election like the representatives. To a populist or a party boss, that might count as a spoiler to the will of the people or to the will of those in DC, or to both. But I may misinterpret GP's point.
I assume the person you're replying to is talking about the Filibuster and supermajority requirements not the direct election history. The filibuster is a senate rule not a constitutional design, so it wasn't part of the "design". Maybe they're both different ways of adding veto points to the same effect, but I think spoilers as "explicit design" is probably not how I'd describe it.
Not parent but the founders were like folks writing smart contract code, thinking about various exploits and vulnerabilities (that might reduce the wealth of their class) so many of the seemingly dysfunctional elements of the system turn out to be designed deliberately to be dysfunctional. Feature not bug.
They were not thinking about various exploits and vulnerabilities but rather making whatever compromises were necessary in order to form the union. It was negotiation, not planning.
Why should they? Why did we allow a president to put in tax raise for the future. Replicants were playing politics from the start. Pass a bad bill, and then hope to get about it when the bad parts kick in when the other side woo be in power
If this was passed in 2017 to go into effect during the next presidential term, wouldn't that only work as a time bomb for Biden's presidency if Trump didn't expect to win a second consecutive term?
Given the history of prior presidents winning 2 consecutive terms, it seems like Trump could have reasonably expected a 2022/2023 tax change to be his problem.
if you retain power, you can fix it. the US government currently has the significant problem that one party campaigns on the government being dysfunctional, so they do their best to make it so.
But what would trump have done if he retained the presidency and lost congress? That's also been pretty common over the last few decades if I'm mistaken, a president with one or both sides of Congress is reelected but Congress flips to the opposition party.
> He would do nothing because his supporters believe misinformation and worship him.
Interesting, that hasn't been my experience.
I live in a very red part of the country and most people I know are Trump supporters, including some family members have been very MAGA since 2016.
I've been hearing more and more complaints over missed promises: no Epstein files, raising budgets, RFK is starting to water down his promises, no end to the Ukraine or Gaza wars, etc.
He missed effectively every promise from 2016. Why did these people vote for him 2 more times, especially after an attempted coup? Maybe these "complaints" are just an attempt to dodge personal responsibility for having supported a catastrophe.
This civic control correlation can simply have more to do with the most-white-supremacist Democrats switching to the GOP en masse and also simultaneously leaving multiethnic cities and school districts en masse after the 1960s. That self-selection left Republicans not a competitive amount of credibility or voter pool behind to work with. Your implication that policy dysfunction has ensued on that account rather than because of fiscal drain -- that's a separate topic. Individual states and individual cities have too many fiscal policy similarities and differences, overlapping, to responsibly compare in any online discussion.
> New York is a better governed state than Florida
Yes, New York is significantly more successful than Florida in almost every way: Better education, better healthcare, longer life expectancy, less pollution, lower crime, more productivity, higher wages, more amenities, better transportation infrastructure, less poverty, happier residents, and so on.
I think quality of governance is a major reason, yes. When my parents immigrated to this country, they moved to a deep red state (Virginia) instead of the deep blue state next door (Maryland). Why? A focus on good schools, low crime, and low taxes, instead of a focus on economic redistribution.
... because nobody moves to Florida for (what they perceive of) the weather, right? Especially not retirees tired of the idea of one more winter in NY.
A government that runs the richest city in the country (SF trades this spot with NY every few years) and makes it look the way it does is the definition of dysfunction.
And Detroit... well, I guess now that they've bulldozed all the abandoned buildings it looks less like a post apocalyptic hellscape and more just abandoned. An improvement I suppose.
California also has easily solvable housing, education, transportation and mental health crises that are entirely driven by mismanagement by the state government. They haven’t done anything meaningful to address these issues in 25-40 years depending on the issue.
Heck, they ignored the water crisis for twenty years, and what they’re doing now for aquifer replenishment is still less than what makes sense.
I say they are easily addressed because simply reverting to California’s policies from ~ 1975 would greatly improve the current situation.
The biggest root problem with California governmental structure is harmful constitutional features added by public referendum, especially Proposition 13 (1978). I guess you can blame "government" for that, but it doesn't seem like quite the right target.
The water crisis is a difficult problem because water rights are complicated and central valley farmers are an influential political group very focused on short-term preservation of water access and not as concerned with long-term sustainability.
> easily solvable housing, education, transportation and mental health crises
I submit that these are much less "easily solvable" than you claim. (What have you personally done to work on these problems, if they are so "easy"?) Legislators don't get to wave a magic wand, but need support of a wide variety of stakeholders who have contradictory demands and expectations (some of which are fairly unrealistic, but anyway..).
Education for example has competing goals of local funding vs. inter-city equity. Should the wealthiest towns get to spend arbitrarily much local property tax money on their own children's public schools while the poorer town next door is running out of toilet paper, or should the state try to equalize funding between schools to give every child the best opportunity? There's not really a "correct" answer to this, and every possible choice has some serious disadvantages.
But more seiously: you are saying that there's this poor little party that can't manage to vote for any laws because big bad Republicans keep coming in and aww shucks. Do we really want such incompetence in government?
You know, I wonder if that's a sad but valuable trait for a politician.
Public opinion can change daily, and external events can appear with no warning. These things can make a prior path of action vanish, or even make it madness to pursue.
If you try to plan everything long term, I bet you hit a lot of disappoint as a politician. If you only see today, then you're not fighting for things that are now not possible.
I imagine one would be far less stressed as a result. And maybe more popular than otherwise.
Some problems require years or even decades to address. Consider how quickly a COVID vaccine was developed, yet it depended upon many years of quietly studying SARS and R&D around MRNA. Or consider trying to address developing or maintaining infrastructure.
A chaotic politican whose mind is changed by the last person they spoke with won't do well facing serious long term problems.
It gets worse if the only things they consistently stand for is their own power, personal wealth, their sycophants, and their grade-school-level (mis)understanding of complex matters.
He seems to care immensely about being viewed as the "winner" and the "best" at everything.
I also have to assume that anyone interested in slapping their name in big gold letters on as many buildings as possible is interested in the perception of legacy.
This is just wrong. It was passed in 2017 (during Trump’s presidency). It was to go into effect in 2020 (a presidential election year during Trump’s presidency). He hoped to be re-elected.
No. It went into effect in 2022 [0], which means the timeline absolutely does track with OP's theory. That gives a hypothetical Trump term 2 a full year to fix it but also imposes enough of a time crunch that they could plan on sabotaging attempts to fix it by another party.
I'm not saying it's the actual story, but the timeline does track.
I don’t think it really changes the narrative. 2020 was also a congressional election year, even had Trump won (as he appeared to want desperately) he could not have been assured of a Republican congress.
My argument is simple: Occam’s Razor
The Republicans in congress put the provision in solely as a gimmick to get past the CBO.
Frankly I don’t think legislators in either party are competent enough to have foreseen the consequences and even if they had been they wouldn’t have put a bomb like this in that would be more likely than not to backfire and affect them.
I just think that too often people interpret incompetence as malice, especially nowadays when things are so polarized that it’s fashionable to hate people who differ with one’s political opinions.
I think you’re wrong that legislators are incompetent. They’re human, and they’ve spent much of their lives learning how to get votes, but they’re not incompetent. A lot of them aren’t malicious, but there seems to be a small group of people outside of the legislative branch who are hell bent on taking control at all costs. And if you want to keep the votes rolling in you have to work with those people or get primaried. The dysfunction follows from fear more than incompetence these days.
I have no doubt it's bad, but I can't believe that it's both fueling mass layoffs and also almost nobody has heard of it. Those are unlikely to both be true.
I'm not sure why TFA makes it sound like almost no one has heard of it, but it was extensively discussed on HN in early 2023 as being a primary cause of layoffs, before it was cool to blame AI.
Why the big tech firms that suddenly laid off a bunch of people the instant they started looking at their 2022 tax bill didn't tell everyone explicitly that that's what was happening I can't say, but it's not like this has been happening in secret.
Obviously interest rates also play a role, and probably a larger one. But this is objectively a very very bad contributing factor, far worse than the impact of coding LLMs.
At least one person at every company impacted knows why they're firing people, right? Likely several people who are in a decision making capacity. At every company.
Those companies have R&D for a reason. A company _wants_ to make things, right? If this is impacting their ability to make things, wouldn't it be in a company's best interest to advocate openly against the tax code, rather than be silent about the reason, fire their staff, and just not make things?
It doesn't make sense to me how so many people are aware of this to the point that many many companies are all doing the same thing for the same reason, but seemingly nobody was talking about it before this post. That doesn't make a lot of sense to me.
It's like this:
Company wants to do R&D, they have a budget, they do math that says they can afford to pay X number of R&D workers with Y budget.
Government changes tax laws in unexpected way, that changes the math so that Y budget only can support X-A R&D workers because the "A" goes to taxes now.
Also important to note, the tech R&D space is a very small part of the overall economy. We exist in a little thought bubble here on HN.
They didn't know the change was coming. Then it happened, now they have to revisit their budgets. Some thought it was going to be changed back before it went into effect, it wasnt.
No, not in this case. 250k is an expense for the company. Company had to amortize this expense over 5 or 15 years. (15 for software engineers outside of the US)
It might be "quite short" but it's full of click bait style text. This tax law will change everything, but we won't say what it is for 4 or 5 paragraphs, nor what changed for another 3
Edit : sorry I just realised you meant the tax law is short. The article itself is very annoyingly written
Agreed. I think the article's text was responsible for the confusion in the comments prior to me posting this. They could have been much more clear and straightforward.
I think the purpose of the change was to "increase revenue":
> Requiring that certain research or experimental expenditures be amortized over a five-year period or longer, starting in 2023, would increase revenues by $109 billion over the period from 2023 to 2027.
> I think the purpose of the change was to "increase revenue"
Yes, but in a specific way: they were trying to offset the tax cuts they wanted so they could pass it via the reconciliation process and avoid the Senate filibuster. They didn't actually care about this revenue and the assumption from most people was that the specific carve-out would disappear in some future bill.
And now with their attempts to keep the tax cuts around, they've just decided to ignore the rule entirely and pretend that extending a temporary tax cut counts as not costing anything. Of course, there's nothing that would stop them from getting rid of the filibuster entirely either, but that honestly just makes it weirder to pretend that this somehow fulfills the requirements rather than just is taking advantage of the rules being only self-enforced.
Idk but it was under trump. And the meta issue was balancing the budget after all his tax cuts so he needed to find more tax revenues. Which this accomplishes pretty handily
I think partially dismissing the question due to the bill happening "under trump" doesn't help the conversation here. If the bill was sponsored by particular reps/senators, then it's worth identifying those, so their voters can factor this bill in to their decision to vote for/against in the future, etc.
> All software expenses must be treated as research and experimental expenses
From what I've read, not for software fixes to ongoing products, but for new products and I can't remember for new feature work. Also if you contract for someone else I heard you can still write off expenses without amortization.
OK, but you've changed the topic from tax accounting to financial accounting/reporting.
In the US, it remains the case that programmers salaries must be treated as an expense (i.e., cannot be amortized) when calculating the company's income statement, balance sheet, etc. Not following that rule will get the accounting firm signing off on those financial reports in trouble (with the SEC, the Public Company Accounting Oversight Board, and maybe even the Justice Department if the purpose of the violation was to defraud investors).
Yes, that is tremendously important aspect here - the US tech would look better on paper - higher paper profits due lower paper expenses - while getting increased cash flow stress due to decreased deductability of the salaries which are among the main expenses in software dev business.
I don’t get the big hoopla. Here in Germany I’m opting to turn development costs into assets (I simplify a bit). I need assets on the balance sheet, otherwise we’re over-indebted. As long as the development costs are much higher than income (I.e. as long as you’re not profitable), then it shouldn’t matter. And once you are profitable, you pay some more corporate taxes, but aren’t they kind of not too high in the us anyway?
the big hoopla is this: you're a newish startup. You have $300k/yr revenue and $$270k/yr expenses of which $250k is paying your programmers.
prior to this rule change, what you pay your programmers is just a deductible expense, so you owe taxes (in this very simplified example with no other expenses etc. etc.) on just $50k.
after the rule change, you can deduct only $50k of the labor cost (in this year), so now you owe tax on $250k.
there is a very good chance you do not have the cash available to make this payment.
of course, after 5 years, things all balance out and are effectively "back to normal". but you have to get through those 5 years first.
How does this work, in your first year, to have 300K revenue on 270K development cost. It sounds like some very specific boot strapping case, and even then after 70K deductions its 230K profit, so like 50K in taxes on 30K profit. Sure it’s annoying, but it doesnt sound like total doom. Its more likely youll have 250K programmer expense in the first year, 100K in revenue, need 200K in outside investment, and pay very little taxes. After a couple of years it evens out anyway.
Consider some sort of logistics company. They might pay 250K for their hardware (e.g. vehicles), 50K in expenses, get 300K in revenue. Theyll be taxed the same as the startup building up their IT assets.
What i like about US is that compare to other countries (like for example Russia where i'm originally from) there is almost no lying and cheating here. Instead there is a respect of the law and an army of talented creative accountants and lawyers. Remember that stale "multi-used" sandwich served with the drink which by virtue of its existence converted drinking establishment into a food serving restaurant? Not being an accountant, i'd just speculate, out of sheer fantasy, that some hardware chip/gadget added to your software may similarly convert your software development into hardware/gadget one.
I'm not sure I buy into this. Sure, compared with Russia it's probably a lot less, at least in terms of being something everyday people engage in. But in terms of comparing with countries like Germany or Sweden I don't know.
Here's some food for thought:
* Global financial crises: Banks were paying (bribing) ratings agencies to rate junk bonds AAA.
* Bernie Madoff: Ran the largest Ponzi scheme ever, with an estimated fraud total of $65B raking in $17.5B in invested cash.
* Enron: straight up accounting fraud sprinkled with intentionally causing brownouts in California to pad their pockets with a side bonus of making Gov Davis unpopular & get him recalled (Enron was closely aligned with the Bush administration).
* Nixon straight up using psy-ops against Democrats & finally trying to burgal the DNC offices.
In terms of stats, the FBI does a few hundred bribery and corruption cases annually. Are they good at catching white collar crime? Well such crimes regularly take more than 5 years to investigate.
And hell, some things that are basically lying and cheating are straight up legal. Usury is legal with minimal to no regulation of payday loans. Pyramid schemes are legal as long as you call it multi-level marketing.
I can’t tell if they’re trying to refute the point (ie the list being long means that it’s definitionally not being hid) or support (ie the list I made is finite and therefore the corruption is hid).
But I think of it like trying to estimate the size of an iceberg by observing the tip. Just because you know a little bit doesn’t mean you actually know about the scale. And there’s every reason to believe it’s quite extensive given how easily money flows from corrupt countries through USD and US persons and companies (eg the major bank that’s constantly getting fined for laundering terrorist and drug cartel money - either they’re the only ones and they’re making a killing providing this service anyway or they’re the only ones anyone is bothering to investigate, but that business is clearly lucrative to continue to engage in).
> What i like about US is that compare to other countries (like for example Russia where i'm originally from) there is almost no lying and cheating here. Instead there is a respect of the law and an army of talented creative accountants and lawyers
I thought you were being sarcastic here at first because, good lord, there is plenty of corruption here in the US (though those doing it used to care more about hiding it). The US, especially in its current state, is certainly not a place I'd describe with "almost no lying or cheating". I do understand that Russia is on another level, though, given the open assassinations and doing things like what was done to Navalny.
> I thought you were being sarcastic here at first because,
You've never been in Russia. There is no clear law abiding business there. That opens a lot of opportunities for those with some power. Corruption is one of them, selective punishment is another. I'm sure in most 3d world situation is not better, but they at least don't have laws to cheat and bribing isn't a crime.
Breaking the law is still breaking the law even if you don't hide it. If anything, not getting in trouble for breaking the law noticeably often means that there's also corruption from the ones who should be holding the corrupt accountable.
Where in that definition does it say that it can’t be done in the open?
See for example Trump’s shenanigans, which are done in plain sight for all to see, but with few if any repercussions (a very brief selection: having foreign dignitaries stay at his hotel in DC while he’s in office; having the Secret Service stay at his resorts when he goes golfing; scamming the public with his family’s meme coins; etc)
Because this is SO much better..../s The only difference in style is that the American billionaire will corrupt everything and still say it is for your own good.
No, he's saying that people respect the law, which they do. It's all about finding loopholes, and sticking to the letter of the law while working around the law to do whatever the law prohibits but doing it in a way that remains legal. This entire situation came up because of loopholes. A great way to offshore money was to spend it on software developed by overseas subsidiaries.
If you’ve never lived outside the US you have zero idea how bad it gets. It literally is an alternate world.
The amount of daily activities in the US that just work 99.999999% of the time that would have a corruption aspect in some other countries is mind boggling.
The closest analogy I can come up with is imagine if every money transaction involved cash tipping the parties involved. And that’s just the beginning.
Or are cynical Americans living in an alternate world, blind to how much better the rule of law is here than most other countries? The commenter's comparison was to Russia. When was the last time Putin lost an election?
I'd say we're slightly behind western Europe as far as rule of law goes, not really sure about the advanced east (Japan, Korea), and miles ahead of just about everywhere else (eastern Europe, Russia, Africa, China, etc). Yes, even with Trump in office, though he really makes me worry.
I mean, the sitting President was shilling cars on the White House lawn and runs an active meme coin bribery slush fund.
This is not slightly behind Western Europe. This is miles behind any developed country. China may be corrupt, but Xi Jinping hasn’t yet sold beans or cars via press conference.
The rule of law gets down to nitty-gritty levels, too, not just a reality show at the highest altitudes: trust the police don't extort you, the ability to gain relief in court (small claims or civil), trust things you build won't be looted overnight, trust in your neighborhood to walk at night or leave something unlocked, trust in your bank to wire things, trust in your title companies, trust in your package deliveries, etc.
It's not perfect, but you could do so, so much worse.
I often couch my arguments in soft language like a conversation would be in order to have a discussion. The idea that the US is miles behind developed nations is nonsense.
Why are these alternatives? I believe it is true that the situation in the US is better than many other countries (not most), and also that "almost no corruption" is false.
Being better than others really isn't the only thing that matters.
But where an established company invests steadily in software, whether it is amoritized or deducted year to year is a wash. Rather than harm tech, this would seem to protect established US companies at the expense of startups. Thats probably great for shareholders in publicly traded companies. It seems just another querk of taxation meant to maintain the established order
It's only a wash if they've been amortizing all along. There's been no advantage to doing so, so established ones have all been deducting, and will have the same five year window of increased taxable income that startups will.
Startups have to face that five year window every time they start up.
Each and every startup will have a year 0 where they're spending more than they earn, and under the new Section 174 they will only get to deduct 10% of their employee's salaries that year. In year two they get to deduct 20% of year 1's salaries and 10% of year 2's salaries, which is still 30% of what the established players will be able to deduct. By year 4, if they make it that far (which most startups don't) they'll finally be at 90% of a full deduction.
Add to that the fact that startups also by definition have a much higher rate of growth than established companies and you'll find that a startup almost definitionally will be paying substantially more in taxes as long as it remains a startup, because they only get to deduct an average of the last 5 years of expenses from this year's revenue in order to calculate this year's profit. That's fine when your last five years are more or less similar to this one, but it's terrible when you've been growing.
The net effect of this change can only be to disincentive startups and cement big, slow established players.
Increasing taxes does not always increase tax revenue. It’s easy to do enough damage to the economy that total tax revenues fall. Past that point is easy for the economy and government revenue to fall into a dwindling spiral.
Not if it limits growth to a commensurate extent (or more)
A big part of why America is as rich as it is in 2025 is Big Tech. If laws and regulations had prevented that industry from taking off by stifling the now-giants back when they were starting up, you may have been more equal today (you’d have fewer billionaires), but there would also have been a lot less wealth to go around, even for the working class
The OBBBA (“Big Beautiful Bill”) suspends amortization requirements for domestic R&D expenditure, and explicitly allows domestic software development as an R&D expenditure eligible for immediate expensing.
The new rules would apply from 2025 to Dec 31, 2029:
Repealing SB174 has bipartisan support. The house already passed its repeal but it died in Senate because a separate took (that also repealed it) took its place but that separate bill stalled out.
174 is so small it can't go through both chambers on its own so it needs to get attached a larger bill like OBBA.
It's unfortunate because it appears both sides want this repealed to allow immediate amortization of domestic R&D expenses.
If they could be required to craft single issue bills, this wouldn't be as big an issue. Instead we get the clusters of good and bad that inevitably die or sometimes worse, pass.
That's still not true. As long as a group within "everyone" (or multiple groups) decide that their support is required to pass the bill, they can suddenly demand concessions and the bill now gets complicated with good and bad.
> As long as a group within "everyone" (or multiple groups) decide that their support is required to pass the bill, they can suddenly demand concessions
Well, yes, but then everyone doesn't really want it, do they? Someone wants something else, and wants that something else enough that it is worth jeopardizing the supposedly universal goal for it.
If you've ever negotiated, I bet you've done the same thing of jeopardizing something you want in order to get something else you want. If you never do that, you'll make a lot of deals where you're riding the edge of just barely acceptable and the other person is taking advantage of you. But in this case, with a standalone law, doing it gets pretty rude and we'd be better off if nobody did it.
It's almost funny that small code reviews are preferred in software engineering (https://google.github.io/eng-practices/review/developer/smal...), but in Congress we have these stupid "big beautiful bill(s)" that are sometimes thousand pages long, and sometimes only released hours before a formal vote. Almost like these bills are intended to fool constituents, cripple opponents' plans, and created just in the hope that they get passed and signed into law without anyone looking carefully.
The short timeline actually makes this an excellent opportunity for LLM analysis if you have a model which could digest the entire bill without reaching token limits.
Arguably, that's the whole of politics: why should I give you something if you don't give me something?
The people involved are, generally, not deep thinkers, aren't aren't thinking much beyond their direct short-term advantage. The system selects against that.
No, the system selects for people who are brilliant for using it to maximize their own individual benefit. Never take the bait that these people are in anyway stupid.
Anything that's not a budget reconciliation bill can just get filibustered in the senate by the minority party. That's why they're attaching everything to the OBBBA.
It's perhaps noteworthy that OBBBA is not the first bill to attempt to revert this tax law. It's simply the latest. There have been other attempts to revert section 174.
Other attempts that come to mind:
1. Tax Relief for American Families and Workers Act of 2024 (H.R. 7024)
2. American Innovation and R&D Competitiveness Act of 2025 (H.R. 1990)
That isn't the reason. They sunset in the bill so it has a lower CBO score (which calculates costs out to 10 years). If you sunset in the bill after 5 years, even if you know it will get renewed, the apparent cost goes down. Get it?
If anything it has been the opposite problem, with modern congresses having been more than happy to delegate away their powers. You might have heard the recent tariff news for example.
Modern being what? The main tariff laws in question happened 50+ years ago. The republicans in congress have been going out of their way to not interfere, but that's significantly different from creating new delegations.
That isn’t how legislation is passed. If anything, it needs a section about acceptable tar shingle application standards for roofs within 6 nautical miles of any heliport operated in a subarctic area on the west cost. Then it’s looking like a bill.
Just last year, Congress snapped to attention and wrote and quickly passed a bill to ban the eminent national security threat of a video-sharing app. That bill doesn't do anything else.
Just a reminder that Congress, even now, can rapidly act on a laser focus when it is sufficiently motivated.
Perhaps not the best example to choose given that the president managed to fully ignore that law. Tik-Tok remains unbanned to this day despite there being no sale.
Is there a good summary of that episode somewhere? I've tried to read up on it as I don't really understand how it was an eminent (imminent?) security threat.
There's a little of this, but more so, you only get one reconciliation bill per year. And anything that's not a reconciliation bill has to be bipartisan.
This bill is goated for upper middle class and tech and defense sector
And I’m tired of pretending like we aren’t going to be beneficiaries
Every Congress increases the debt, we can acknowledge that the cuts they picked are going to wreck the lower class especially with the medicaid, we can acknowledge that it won’t meet its goals of cuts
but are you guys just scared to acknowledge its going to super charge things that you are a beneficiary of too? so busy saying it just benefits billionaires as if we’re trying to avoid guillotines. not gonna happen and many people here are going to try to take advantage of new programs
You don't want to live in a society where an increasingly large percentage of the population have nothing to lose.
Regardless of whether it benefits our industry or socioeconomic status, it'd be incredibly shortsighted to just do all of that at the expense of the lower classes.
Also someday you might find yourself needing things like medicaid. I once met a former SE who was on disability because he lost his sight. Protecting those things for others also protects them for yourself.
I am very much in support of Section 174A no matter who does it, what riders goes alongside it, or what it was a rider to
Congress can always pass anything else at any speed. This slow motion filibuster thing is a choice, and the powerlessness of doing anything about that choice just means everyone else should have a single they care about too to correct the laws and riders that shouldn’t have passed.
I have to say that you sound very biased in the way you're talking about the bill. If a primary goal of this administration is to cut and make government more efficient, this bill is about as big of a failure as one could conceive. Any reasonable person would have to admit that.
I'm fine admitting that I would benefit greatly from this bill. I also hope to heaven it doesn't pass because an additional trillion dollars to suit me sounds asinine. I don't need help.
Amortization is bad policy when it comes software. Software is inherently high risk. Every piece of software is unique and does not guarantee steady income over 5 years. Most startups won't survive 5 years to fully realize the deductions. This is the end of US software dominance.
> Selling you code, lol -- code is more of a liability really :)
It's important to consider that lawmakers (who are not well informed or downright stupid) might think code has intrinsic value because of media married with a lack of real-world experience.
Lawmaker is a misleading word. The people who actually make the law and lobby for it probably know quite a bit. The representatives are law voters not makers. They don't design the laws literally. They vote because they are told to.
I remember the day I mentioned this in my high school^ honors sociology class and the eventual valedictorian exclaimed that I was stupid to think that. The system has been broken for longer than I have been alive, but the indoctrination has been working to make up for it.
Amortization is bad policy, period. If cost is actually incurred, it should be fully deductible immediately. No matter if it's a piece of equipment or software.
i'd disagree heavily with that... let's say you have an expense of an insurance policy that covers you for the next 10 years. You're paying for 10 years of service, that should be amortized over 10 years.
Yeah but if the insurance policy requires me to pay upfront, I'm out the entire ten years' worth of insurance premium. Amortization forces it to be divorced from actual cash flow.
What is that? Software sold by companies that have HQ in the US? Or software created by someone in the US? Because if it is only the first, good riddance.
> Is US software dominance because of our startups? Or because of the giant trillion dollar monopolies we have?
Most likely neither: It is its massive trade deficit, the one it strangely wants to get rid of now, that has allowed US consumers to consume more than they produce (i.e. you can take something with no real expectation of having to give anything back in return). Which, as it relates to tech, has enabled offering services for what is effectively free to dominate the market. Nobody else in the world can compete with that.
> Didn't AAPL, GOOG and FB all create products _before_ they had any taxable income?
Wouldn't you say they had no taxable income because of it? If Facebook brought in $100,000, and paid $100,000 to developers, then there would be no taxable income under normal regimes. But if the developers were not tax deductible, then that $100,000 in revenue would be taxable, even though the bank account is empty. This isn't nearly so simple, but it has changed the calculus in a similar way. The business models of old no longer work because of it.
Well, presumably the claim would be that a factor in their not having taxable income was the fact that they didn't have to amortize their development cost.
Yeah; start-ups will start paying tax much sooner since salaries are the main expense in software development, and only a fraction can be deducted per year. The tax change must make things marginally more difficult for young companies that have some revenue, aren't cash-flow positive, and have a short horizon.
It started with small and nimble innovators. Then it was shifted to Big Tech with the squeeze of patent trolling in the 2000's applied. It was capturing massive created value into the hands of few, connected, corrupt shitbags.
I’m not familiar enough with the very early days of Apple which started out as a hardware company to rebut you; but perhaps you mean the current Apple that has re-invented itself?
This impacts deductable expenses, not profits directly. The labor you pay for internally owned IP related to software must be amortorized. This screwed up an enormous number of business plans because software has more risk than many other endeavors. For small businesses, you basically can't do your own software.
It applies to things like configuring your internal tools too. Good luck at audit time.
I don’t know how it works in the U.S., but we had HMRC in the U.K. write us a cheque every year, as if you have a greater R&D claim than your tax bill, you get a rebate.
It's worth noting that FB was quite possibly being secretly funded with taxpayer money by national intelligence interests at inception, which would have substantially reduced or eliminated commercial pressure early on.
DARPA was working on Project LifeLog starting in 2003, was to be "an ontology-based (sub)system that captures, stores, and makes accessible the flow of one person's experience in and interactions with the world in order to support a broad spectrum of associates/assistants and other system capabilities". The objective of the LifeLog concept was "to be able to trace the 'threads' of an individual's life in terms of events, states, and relationships", and it has the ability to "take in all of a subject's experience, from phone numbers dialed and e-mail messages viewed to every breath taken, step made and place gone".
The program, at least officially and publicly, was cancelled on February 4th, 2004, the exact same day that Facebook was founded.
You can call it a coincidence if you want, I just tend to be very skeptical of "coincidences" where massive, powerful, unaccountable, immoral, unethical institutions like the US intelligence community get exactly what they want at the expense of our civil liberties.
I often wonder if national intelligence interests are behind or have taken control of major corporate players like Microsoft, Google and Apple. There was an article [0] back in 2015 that brought forth the proposition that google was created by the CIA. It would explain the current enshitification of these companies and the lengths they are going to take away choice.
This is insane, how does it make sense? Employee salary expenses are no different from other expenses to run your business. Imagine they did this for raw material instead, a restaurant could only expense 20% of the food that they sell. If they purchased $100 worth of food, but could only sell $50 worth of it, they have to pay tax on that even when making a net loss overall. It just does not make any sense. There would've been a huge uproar if this was done for cost of goods. Why are employee salary expenses any different?
It makes sense when you consider that there is no minimum tax rate on businesses.
Given the choice, Amazon would rather spend 100% of its profits on itself than allow any of its profits to be paid out in taxes. Section 174 was implemented without a minimum tax on corporate profits before voluntary deductions such as research. Therefore, it’s exploitable and all companies ought to hire and fire staff to ensure their profits show as 0%.
This tax code defect is now closed by accident, but could have been done much more intelligently than it was. Oh well.
(EDIT: My first sentence is potentially confusing when I reread it later. To restate: section 174 was defective as implemented due to the uncapped 100% deduction, but the concept of a significant research exemption is still excellent. Just need to close the effective 0% corporate tax rate loophole.)
The company already pays payroll taxes on those salaries, and the employees pay income taxes. And the people hurt by this aren't the shareholders or top executives, it's the rank and file workers getting laid off, losing benefits, and being asked to work more for the same pay.
What this change effectively did was make software developers significantly more expensive, without increasing the amount those developers get paid.
Software developers are already too expensive in US, so this applies some downward pressure on those salaries. Frankly the economy will be much better off when tech salaries equalize across geos, thus avoiding the deep whole US manufacturing is in (for example, manufacturing wages in Vietname are one tenth of US manufacturing wages, and thus it is better to open new plants there).
If you want equalized poverty, feel free to move to the EU. Say goodbye to owning a nice house, or building any kind of wealth - that's reserved for the old money class.
In the US, software is one of the few remaining ways to achieve the American dream. I came to this country to work hard and earn money.
If you look at happiness and indexes versus taxation rates - yes, making everybody poorer does tend to solve things. Not too soon in the growth curve - but certainly not never.
Those two scenarios are only comparable if you isolate happiness and taxation and completely ignore things like social services and inequality.
I think you're referring to Nordic countries which consistently rank as the happiest countries and also have relatively high tax rates (4 of 5 Nordic countries rank in the top 11 tax rates globally. Norway has oil.) The high taxes that "make everybody poorer" also fund extensive social services that contribute to happiness.
However, this conversation is about making (a class of) workers poorer by using tax policy that puts downward pressure on their salaries. Tax revenues will stay the same, so social services will not be increased. Economic inequality increases because the workers became poorer, the C-Suite and Board Members don't.
Don’t forget the other stakeholder - the general public.
Yes it sucks for developers, but does it make any difference for any other employee? Why does Joe’s plumbing have to pay those taxes, but Jane’s AdTech company doesn’t?
Sure, there are benefits to investing in R&D in general, and tech has fueled a lot of growth, so incentivizing it has likely paid off for the whole economy. But will that forever be true? Maybe?
In some parts of the world we have a sales tax which is a form of minimum tax on business outputs. The consumers of plumbing and software pay 10% regardless on a businesses profitability.
Yeah, VAT would help tremendously in alternative here, but for gestures at United States sociopolitics reasons the existing U.S. taxation methods can’t keep up and won’t be repaired any time soon. I could boil the ocean on this down to bedrock (citizens should be taxed on [redacted] in excess of threshold, services and goods should be VATed) but I stand by “section 174 with a sub-100% cap” as what at minimum would have balanced research and taxation.
Joe's plumbing doesn't have to pay those taxes. Operational costs, including paying employees for normal operations, is deductable.
But with the change, the cost of R&D employees is now only partially deductible (right now, you can eventually deduct the full amount over the course of several years), and software development has to be considered R&D.
after 5 years then every year is deducting a whole year's worth of R&D - as long as that investment is not too lumpy from year to year you are back where you started
> Given the choice, Amazon would rather spend 100% of its profits on itself
And why is this bad, exactly? Money will be spent and will go back into the economy. Amazon will have to use the funds to build new offices, datacenters, do research, whatever.
And even if execs give themselves $10^11 USD in bonuses, they will be taxed as personal income, at even higher rates than corporate income.
It is complex - is it better for the money to go back into the economy by paying high salaries to a specific group of highly-educated people? Or is it better for the money to go back into the economy through taxes, then disbursing the benefits to lower-income benefit programs?
I’m not sure what the answer is. The former is likely to drive some innovation, which I’m sure varies by company. Where the latter could also unlock innovation by giving the bottom-quartile of earners a chance to improve their situation.
The answer is simple: it's the biggest growth generator in USA.
Growth has its own problems of course (I don't want to estimate the health impact of Coca Cola), but it's a prerequisite of a country not falling behind others.
At that point, do we need to fundamentally rethink political donations by companies (outright ban them) and SuperPACs? No representation without taxation.
> It is complex - is it better for the money to go back into the economy by paying high salaries to a specific group of highly-educated people?
Yes. Also, the salary will not go _only_ to highly-educated people. For example, if Amazon decides to build a new distribution center, it will employ blue-collar workers to build it, not software engineers.
> Or is it better for the money to go back into the economy through taxes, then disbursing the benefits to lower-income benefit programs?
No.
> I’m not sure what the answer is.
The answer is pretty clear: invest money into the private sector, rather than divert it into the Federal budget. Private actors are more efficient at allocating funds than the government.
I'm not against social spending, it's a necessary evil for any real state. Pure libertarianism leads to dystopian outcomes. But it should be understood that it's a very real artificial inefficiency that is imposed on the economy.
There are also situations where additional social spending is necessary, but they are VERY easy to detect: when your interest rate is near zero.
Then you misunderstand, the markets and economies of the past 5 decades have been two children playing Candyland. Saying it's not is a No True Scotsman fallacy, because clearly since I labeled it as Candyland economy it must be so.
because a high percentage people on HN fall into the group that benefits more from neoliberal economics than the larger group of people within those economies who don't benefit.
I used to think like you, until I saw what the lack of neoliberalism does to countries. And before I witnessed the magic of market economy that adapts to changes far, far, far better than anything else.
If you want a static economy that supports gradual decline (preferably with a mineral-based income stream), then a lot of state spending is fine.
It’s the same for movies, other intangible assets that are valuable and produce income over several years. And it’s done for many tangible goods, like servers in a datacenter, the kitchen equipment in a restaurant.
I think you may misunderstand. For most of those, you get the choice to amortize if you prefer. In this instance, you must amortize, which is a big problem for startups.
Employee salary cost isn't always 100% an expense.
Imagine you are BigCarCo, you make cars. The salary for your factory workers that build cars to be sold is an expense, incurred in that year, to be matched against the revenues earned by selling those cars. But the cost to build the factory needs to be amortized over the lifetime of the factory - and that's true whether you buy a factory from BigFactoryCo or hire a bunch of people to build it.
Now, I'd argue that a) most software dev work is closer to the factory worker than the factory builder and b) the lifetime for most software is less than 5 years, but the idea that some cost of developing software should be amortizable is pretty reasonable.
Actually, if the company isn't selling the software they build, what their software devs do is closer to building a factory rather than working in it.
Mostly developing software is about automating things that are expensive and slow to do manually. So, to stick with the factory analogy, it makes the factory a bit better and more efficient. If you stop doing that because it is too expensive, you fall behind with your factory.
Of course the whole issue in the US is that it outsourced much of what happens in factories to China and software has become one of the main things the country runs on.
Now imagine that a restaurant buys 100 tables, 500 chairs, kitchen equipment, cutlery for 800 people, signage, a security system, and does a remodeling before opening. (Or an airline buys an airplane. Or a hotel chain builds a hotel.)
Should they be able to expense all of those items that provide value for multiple years in a single year?
Does software development provide value exclusively in the year it's done? Or over multiple years?
The reason that we require you to deduct an expense over years for some things is because they have a resale value that needs to be accounted for. It's not a pure expense because you have an asset with real value that came out of the purchase. Employee time has no resale value. Once used it's gone, so employee salaries are expenses, not investments.
The only possible justification for the Section 174 R&D changes is that employees working in R&D theoretically are producing something which does have a resale value, so there's a small tax dodge enabled by direct-expensing your R&D costs but then ending up with an infinitely-copyable asset that came out of it.
If that's what you're saying, then I'd reply to that argument by saying that paying humans to design new things has historically been a business strategy that the government has wanted to incentivize in a way that buying and holding physical assets has not been. I've seen no justification for the government deciding that from 2022 on we should actively discourage R&D, it just seems to be a mistake.
Software is like Art, it doesn't have value until sold or can be used. If they sell services based on the software, they are generating revenue and then taxation on that revenue can occur.
Same as if they sell the software, either as a copy or ownership.
But not being able to take salary as a business expense seems like as thing that would happen if software in and of itself has value, which is largely does not.
> But not being able to take salary as a business expense seems like as thing that would happen if software in and of itself has value, which is largely does not.
To me it seems like a thing that just wouldn't happen. Forget software.
Say you own a McDonald's, and as part of your operations you have some people on staff to take orders, prepare food, and clean the bathrooms. Why are their wages not a deductible business expense?
If the answer is "they are, don't be stupid", then... what exactly was the R&D tax break?
> I've seen no justification for the government deciding that from 2022 on we should actively discourage R&D, it just seems to be a mistake.
Removing a specific tax exemption to create a level playing field isn’t discouraging R&D.
That’s the thing, every year such exemptions exist the US taxpayers are handing out money. Just because we subsidize say EV’s or Corn doesn’t mean that’s the baseline forever more.
> Removing a specific tax exemption to create a level playing field isn’t discouraging R&D.
If the end result of removing this exemption is that there is less R&D done in the US, then yes, empirically, removing the exemption discourages R&D. Assuming the mass layoffs were indeed fueled by the removal of this exemption (I don't know if the article is correct or not), then it is reasonable to assert that it is true that removing the exemption has reduced the amount of R&D done.
Or, you could also say that the "default state" is some low level of R&D, and the tax exemption encouraged and incentivized more of it.
Either way you slice it, though, the status quo prior to 2022 was some level of encouraged/incentivized R&D. That status quo changed to encourage/incentivize less R&D, and companies have followed these lack of incentives and have fired a lot of their R&D staff. Is that a good thing for the US? I can't see how it could be.
> empirically, removing the exemption discourages R&D.
Not clearing a road means fewer people use it, but you not going out with a shovel to clear a public roads isn’t you discouraging their use nor is you canceling your plans to clear said roads.
It didn’t create a level playing field, it just discouraged a very specific type of R&D while ignoring all others. All other types of employee salaries follow certain rules and some can optionally follow R&D rules. Software is now the only one required to follow 5 year R&D amortization so the deck is now stacked against software.
The default situation is whatever was yesterday. I’d be astonished to learn that even a single significant civilization functioned without subsidies or patronage of priorities held by a society’s leaders.
Those subsidies lasted a long time, but just as with a TV they didn’t last forever.
So if your argument is some subsidy will probably happen next year sure, but individual subsidies change over time. No specific subsidy is the default.
Level playing field for whom? Who does incentivizing R&D disadvantage?
Restaurants weren't competing with R&D-heavy corporations in any way. R&D-heavy corporations competed with each other, on a level playing field where all of them can build new stuff without having to pay taxes on negative income in their early years.
The only change this has made is un-level the playing field in favor of old, established corporations that already have the revenue streams in place to fund their new R&D projects.
Taxpayers who end up with the bill and every company is competing for workers, office space, etc. Incentives across decades shift what people study, what business get created, etc. R&D sounds great abstractly, but it’s not some panacea where unlimited funding results in pure gains.
The economy is generally more efficient without central planning, and dumping money into anything that can be classified as R&D is simply inefficient.
> every company is competing for workers, office space, etc
My company is all-remote and none of us would work for a company that isn't doing R&D. Most of an entire profession now has to be amortized over 5 years.
> The economy is generally more efficient without central planning
The old tax code isn't "central planning", it just had the very reasonable property that the government wouldn't force you to pay taxes on a loss.
This scenario [0] is now possible. It wasn't before. That is a catastrophic level of stupidity, and you can't justify it with invisible-hand nonsense.
> none of us would work for a company that isn't doing R&D
So you’d just be unemployed for the rest of your lives? That’s a possible edge case not worth adjusting the tax code for, but it seems unlikely.
> wouldn't force you to pay taxes on a loss.
R&D is an investment, you only pay taxes if the rest of the company is profitable.
If your company is spending 1M / year on R&D and not adding 800k in long term value then in theory you’d be correct. But at that point you either aren’t doing R&D, or are doing such a poor job of it that the government shouldn’t be encouraging that activity.
The problem here is that all software development (excepting that done for hire) is classified as R&D. The software developer working on your Wordpress or Magento site (and arguably the accountant building a spreadsheet, to take the statute at face value) isn't an operational expense, they're now an R&D expense that has to be amortized and can't be taken as an expense against revenue. Previously, this was an optional choice (and many large and mature companies were amortizing anyway), but under the current tax treatment it's required, which essentially turns early-stage startups into cash bonfires, given how many small companies don't make it to year five.
As a practical measure it’s really not. The transition is difficult for existing companies, but a future startup is going to be minimally impacted.
Year 0 you’re unlikely to have any profits, future years you have multiple years of R&D to offset with.
But let’s assume the worst case. Taxes are 21% of profits and at minimum deduction 20% of R&D so the theoretical maximum distribution is 0.8 * 0.21 = 16.8% increase in R&D expenses if profits = R&D year 0. But that maximum case is only year 0, you’d be able to fund R&D with those same profits and easily be profitable after that.
If profits where say 40% of R&D in year 0 you’d have to pay 16.8% of 40% so an increase is only 6.72% hardly likely to tank the business if it’s already generating that kind of income year 0, and again after that point you’ll deduct for multiple years.
More realistic numbers are going to be really low multiples here, more importantly they represent significant investments not operating expenses.
> Year 0 you’re unlikely to have any profits, future years you have multiple years of R&D to offset with.
You're only unlikely to have no profits if you have no revenue. And you only get to break even 5 years in, which most startups will never reach.
In practice what is likely going to happen is that we'll see more and more startups deliberately avoid revenue in the early days. More and more free tiers followed by rug pulls when revenue actually becomes an asset rather than a liability.
There is no unplanned economy, only different outcomes from better or worse plans. And I'm having a hard time imagining a worse plan than one that intentionally disincentivizes businesses from adopting a sustainable business model early in their lifetime.
> unlikely to have no profits if you have no revenue.
It’s much easier to have revenue than profits, set the price lower and suddenly zero profit. Some company avoiding profits because of the 21% tax on profit like that would be mathematically dumb.
> There is no unplanned economy, only different outcomes from better or worse plans. And I'm having a hard time imagining a worse plan than one that intentionally disincentivizes businesses from adopting a sustainable business model early in their lifetime.
There’s zero advantage to avoiding revenue or profit here. You’re tilting at windmills.
You simply need less investor money for R&D when other parts of the company are profitable. As to central panning, the mistake you just made is mitigated when many people are all independently making plans. Governments always need to get it right, the market is fine if some people get it right and therefore can reinvest in their success.
It sounds like you’re talking about government funding of research? This is about private companies funding the costs of making product ideas into actual sellable products.
Are you asserting that software and other labor-heavy startups should raise additional private capital so that they can pay taxes before they’ve established themselves in the marketplace? I’m not sure what you mean to say exactly.
I’m saying investers should pay the full cost of R&D without assistance from taxpayers.
When the non R&D portion of the business is profitable they should start paying taxes. Assuming a company isn’t miss classifying operations as R&D it shouldn’t be a major issue.
This will of course discourage “riskier” startups and dampen innovation and give more power to profitable incumbents who will have less incentive to innovate. (Perhaps the result of this looks like Europe?)
What specifically do you disagree with? That R&D is an investment? I mean outside of the tax code that’s what it means to do R&D.
As to my other point, the highest risk category of startup has zero customers for years they also have zero revenue, zero profit, and zero taxes to pay here. On the 5th year they can deduct R&D from each of those years making the net effect on them minimal vs a startup with profits on year 0.
> The economy is generally more efficient without central planning
Big fat "citation needed" there. I know you chose the term "central planning" to try to invoke the communism boogeyman, but overall, free markets do not exist, and have never existed. Governments constantly use various levers (taxation being one of them) to encourage or discourage certain kinds of business activity. This is nothing new, and I find it laughable to suggest that this kind of thing should be done away with entirely.
There’s a lot of evidence for this outside of communism. Housing markets for example are a clear example of economic inefficiency created by subsides. But you also see problems with farm subsidies, flood insurance, and a host of other related issues.
Markets operate on revealed preferences, which is just a massive advantage in terms of giving people what they want. There’s definitely a role for governments in economies around information asymmetry, safety, etc, but allocation of resources specifically doesn’t work well.
What about construction worker and other labor time to build a factory? That’s the analogy being made here by the tax code: Software whose development is a capital expense with value returned over time.
From a quick search it appears to me like construction labor is deductible as an expense in the year it is incurred. Do you have evidence that says otherwise?
My reading of § 1.263A-1 is that construction labor must be capitalized.
§ 1.263A-1.a.3.A indicates that it's in scope: Real property and tangible personal property produced by the taxpayer
§ 1.263A-1.e.2 specifies that Direct Costs are subject to capitalization: Producers. Producers must capitalize direct material costs and direct labor costs.
(I'm just a taxpayer, not a tax lawyer or even an EA or CPA.)
What tax code references or treasury regulations did you find to support your belief that construction labor can be expensed in the year performed?
> Dear ChatGPT, is construction labor deductible as an expense in the year it is incurred according to GAAP? Please answer in a few lines.
Under GAAP, construction labor is not immediately deductible as an expense in the year it is incurred if it relates to the construction of a long-term asset (like a building). Instead, it is capitalized as part of the asset's cost and then expensed over time through depreciation. Only labor costs not tied to asset creation (e.g., routine maintenance) are expensed as incurred.
Unfortunately my understanding of the R&D expensing rule is that it is lifted directly from GAAP, which means private companies have to adhere to those (heavyweight) rules to comply.
Fair point. I changed the question to "according to the tax code" and it told me that
Construction labor is generally not deductible as an expense in the year incurred if it is related to the construction or improvement of a capital asset (like a building). Instead, under the U.S. tax code (IRC §263A), these costs must usually be capitalized and recovered through depreciation over time. Exceptions may apply for certain small taxpayers or repairs.
It's only shifting what year the government gets its revenue. The government should simply let the company choose how to do it, but if they choose anything other than year 1 interest will be payable at government bond rates.
It's also massively shifting the companies' cash flows. The company paid $X for R&D this year, but for tax purposes 80% of that $X expense is moved to next four years. So for this year's tax purposes, the company R&D expenses are much lower than what the company paid.
Ironically I think they would want to claim that over multiple years unless they have other profitable operations under the same company. E.g. other restaurants.
Imagine a restaurant spends money on employees to build 100 tables, 500 chairs, etc. Those tangible goods would be capital assets, so the labor costs of building them would also be capitalized.
This change to the tax code is just bringing the tax treatment of software development in line with how every other industry is treated. IOW, it was closing a loophole. A very valuable loophole, whose beneficiaries used it to get filthy rich, and bragged about how their industry was so much more valuable than everything else, even though a lot of that value was due to the exception software was getting in the tax code.
Notably, in the current version of the budget as of 6/6, the loophole is temporarily coming back, though given the Musk-Trump feud, it's very possible it will get pulled again to try to mollify the hardline deficit caucus.
If the restaurant buys e.g. a fancy oven or a delivery truck, it can't expense 100% of that cost in year 1, it has to spread that cost over the lifetime of the oven or truck.
Labor that operates the business day-to-day would be an expense, labor that creates a capital asset is more complicated.
I happen to think most employee time in software dev is more on the day-to-day operation side, and should be expensed, but I can see an argument that some should (or could) be amortized.
> If the restaurant buys e.g. a fancy oven or a delivery truck, it can't expense 100% of that cost in year 1, it has to spread that cost over the lifetime of the oven or truck.
The difference of course is that you'll have a truck or oven that can be sold. If you could count the full value in the first year then you could sell and buy one each year to reduce your taxes without actually changing anything.
Thus if we want to go that route for software the salary of the R&D employees should be counted against the value of the software they created (As in, the value were it to be sold wholesale to another company). The time spent by the employees is not an asset, once you pay the employees for their time it's gone even if they generated nothing of value. The actual value is that of the software, but that's obviously not easily assigned a value.
Worth noting: the version of the Big Beautiful Bill passed by the House ends this particular change, starting in tax year 2025. We'll have to see if this provision makes it through the Senate, and in what form.
That's crazy. We're 3 years into a 5 year depreciation cycle, and now they "change their minds". Sure convenient when you know you are in power to supercharge growth and leave a time bomb for the next admin.
That is correct. Some historical context is much appreciated in this thread.
> tl;dr on Section 174, Research & Experimentation costs went from being fully deductible in the year incurred to being deductible over a 5 year period.
Larger tax bills and a tightening on what roles/activities are deductible as R&E are likely what OP is pointing at with his comment.
To the best of my non-inside baseball research, Section 174 changes were simply one part of a package of revenue generating measures to offset the large tax cuts from the broader tax act they were a part of.
The changes came from The Tax Cuts & Jobs Act of 2017 that was introduced to the House of Representatives by Congressman Kevin Brady (R) Texas. The bill passed both houses of Congress along party lines. Then President Trump signed the bill into law. Section 174 changes did not take effect until 2021.
The title of this article implies that it is a major or even the only cause for mass tech layoffs, which I strongly doubt.
For example, rising interest rates I'm sure also independently contributed. I would be interested to if anyone has gotten a sense of exactly how much this has contributed.
agreed, the interest rates and the overestimation on the stickiness of the pandemic’s increase in internet usage post-pandemic are the primary other contributing factors that, imo, represent the lion’s share; even allowing that the tax changes are tertiary is a stretch much less as the primary/secondary reason
I only recently learned about the Section 174 change, and honestly didn’t expect it to have such a big impact.
I used to work at a small startup, and most of our spending went toward engineers’ salaries. If we had to amortize that over several years back then, I don’t think we would’ve made it.
It’s surprising how a single line in the tax code can quietly make it harder for small teams to hire. Makes me wonder how many other policies are silently shaping things behind the scenes.
As a non-American, it seems strange to me that the cost of regular software development, i.e. that is neither “research” nor “experimental” in a conventional sense, would be deductible in the first place (amortized or not). Isn’t that subsidizing a whole business sector? Maybe I’m misunderstanding something.
Which means that if you are in Year 1 of operation, your values for YearlyExpenses[-2:-5] are all 0 and you only get to deduct 1/5 of your actual operating costs for the year from your "profit". So you can be in the hole but still owe taxes on your "profit" for the year because what you spent money on was classified as R&D.
Why should money spent on software _development_ not have to be deprecated over time like other money spent on _development_?
I get that it sucks from a cash flow standpoint but the same is going to be true of other R&D expenses. It's just that we're more exposed to this specific R&D expenditure and not others.
> Originally enacted in 1954, Sec. 174 has historically allowed taxpayers to deduct SRE expenditures in the year incurred. Its original aim was to level the playing field for small businesses, those without dedicated research teams, that may be unable to deduct product development expenses under Sec. 162 because the costs were not ordinary and necessary expenses paid or incurred in carrying on a trade or business
Straight-up, any deviation in the tax code for a special group is always a subsidy.
I mean, yes, it will be true for other R&D types. But that's also new and also broken for the same reason: it means new R&D companies are at a massive disadvantage in their first few years compared to the established players who have lots of expenses queued up to deduct. It's wealth redistribution from young startups to established players who have 5 years of past expenses to use in their favor, and that is going to be a very bad thing for the health and vibrancy of our economy.
And, as a sibling points out (and as I pointed out in a comment at the top level), software is in this regime singled out from all other possible R&D expenses, making it particularly vulnerable. A skilled accountant/lawyer can probably turn big chunks of other R&D expenses into something that doesn't fall under 174. No amount of skill can do that for software, because we're singled out.
Because you slinging a React component or Vibecoding a security pile requires no Technology Readiness Level assessment nor does it have development liability. Rather, what we call Software Development is more appropriately labeled Software Engineering.
Salaries in general (not just of software developers) are tax deductible in many countries. This is desirable because we do not want companies to be paying taxes on revenue.
In the US, unless you are a C Corp then you probably also pay taxes on net income of some form. C Corps have some different accounting, where dividends are double taxed unfortunately.
Small business owners are very impacted by the R&D schedule.
Businesses are taxed on profits, not revenue. Paying people to write code is an expense, so you'd normally deduct that expense (plus all your other expenses) from your revenue to arrive at an amount that should be taxed.
That's the rub. Is it an operational expense, like rent or a capital expense, like buying machinery?
It is sort of between the two in my view and is highly dependant on what the software engineer does each day.
Are they fixing a bug, helping a customer, refactoring? I think that is operational.
Are they building out a new feature? That is capital. But it is not quite like buying equipment because it adds no value to the books. So depreciation seems off.
But the same issue applies to other roles. Is a sales persons day trying to land a sale, or trying to develop the business.
It all comes down to "intangible assets" and whether you are making them.
I think it is easier to just say if you are paying someone to work then you can deduct. There must be better ways to claw it back.
The whole reason for most business to exist is to use operations (operational costs) as a lever to increase the growth and intangible value of the business.
The answer is that it's an operational experience when it's a salaried employee and a capital expense when it's a contractor. Like not in a theoretical sense, this is how it's classified right now.
It stems from the difference in treatment of capital gains and income. Either way it’s deductible, the difference being when it is deductible and how much tax is saved. Capital deductions are typically done later since they require a taxable event.
It’s a fudge to make projections look better to allow congress to pass a budget neutral reconciliation bill with the intent that congress would remove the fudge before the consequences triggered.
Governments in general are pushing for capital gains tax normalization where instead of requiring a taxation event the capital gains tax would be levied yearly. In such a scenario the only difference remaining would stem from the difference taxation rates.
> Governments in general are pushing for capital gains tax normalization where instead of requiring a taxation event the capital gains tax would be levied yearly.
> Because taxing unrealised gains are wealth taxes.
No, wealth taxes are a tax on retained wealth (a stock). Taxing unrealized gains is a tax on income (a flow), it just changes the point at which taxation attaches from a realization event to the actual gain.
Yes, you have. You have an asset of greater value which you can leverage in a number of ways without liquidating it and "realizing" the gains. That's a real gain, with real value.
> you could be taxed over and over again
Only if you make new unrealized gains.
> and if the stick drops or hits zero then what?
Then you have a negative unrealized gain, or, equivalently, an unrealized loss. If you are taxing unrealized gains instead of taxing gains when realized, then the natural assumption would be, just as is done with taxing gains at realization, that negative unrealized gains are either offset against current income or against future unrealized gains, and so effectively create (considered on their own) negative (current or future) taxes. The simplest form of this is to offset only against future gains, by the simple mechanism that when gains are recognized for tax purposes, they adjust the basis value of the asset, and when unrealized losses occur, they don't effect the basis value at all, so you don't have a taxable unrealized gain again until the market value exceeds the basis value established at the prior peak.
More complex versions would allow you to offset some or all of the unrealized loss from the prior basis value against current income of other forms, but the amount of that offset would reduce the basis value of the asset.
The unrealized value is notional, not actual. This is a very important distinction. The notional value is often not remotely realizable. In many cases, the realizable value can be a tiny fraction of the notional value.
Most laypeople grossly conflate notional and real value. Taxing notional value massively inflates the adverse impact of tax incidence on expected returns relative to people’s casual intuition based on the relative tax rates for realized and unrealized gains.
A tax on unrealized gains is in effect a way of laundering a steep tax rate so that it looks “small” and therefore reasonable to the unsophisticated.
No, its an actual thing, measurable by some mechanism. Otherwise, this would be a non-discussion, as taxing it would be impossible, not a possible thing that we can argue about the merits of.
> The notional value is often not remotely realizable.
Whether it is or is not immediately realizable is immaterial to the desirability of taxing it; it may be material to designing the forms of taxation that should be acceptable. E.g., if the difficulty of realizing the value is, across the tax base, likely to making collecting the tax in cash or equivalents difficult, it would argue for permitting a fallback option for the tax to be collected in-kind, e.g., by the taxing jurisdiction acquiring a proportional interest in the asset equal to the share of the value of the asset represented by the taxes not paid by other means.
> A tax on unrealized gains is in effect a way of laundering a steep tax rate so that it looks “small” and therefore reasonable to the unsophisticated.
If you allow carry forwarded losses, even just by the simple method of adjusting basis values, and include taxes on realized gains (and carry forward, offsetting against current income with perhaps a negative net, etc., for realized losses), then taxing unrealized gains is identical to taxing realized gains if the gains are eventually realized, but simply avoids the ability to find maneuvers to benefit from leveraging the value of the asset without paying taxes by avoiding realization. It doesn't make a "steep" tax rate look small, it makes the tax rate look like exactly what it actually is, unlike taxing only realized gains, which makes an effectively non-existent tax on capital gains look like something more, when people can benefit from assets without realizing the gains.
For many assets, like real estate, there are liquid markets with market prices. There are a number of US states that already tax based on real estate value, you can dispute the assessed value but that impacts other things like insured value.
Being difficult to assess value is a problem they’ll make you pay an accountant for and punish you if you get it wrong, it’s not going to stop them.
In the US, most recent studies of asset portfolios suggest that 60-70% of notional asset value has no liquid market. We already generate fictitious valuations for compliance purposes in many cases (e.g. 409A) that no one confuses with being representative of actual value. Tax policy based on overt fiction is bad policy.
Even in the case of real estate, a large amount of value is locked up in extremely non-liquid markets. You might get a vaguely representative market-clearing transaction once per decade, with high price volatility that makes it nearly impossible to predict what the next market clearing transaction will look like. I’ve owned assets in these types of non-liquid markets; differences in subjective valuations can vary by an order of magnitude and there is no evidence from the market to support any of those values.
If you only include extremely liquid markets for tax purposes in order to make valuations vaguely plausible, assets will be made non-liquid such that they are excluded from consideration. Ultimately this is why taxes on unrealized gains have been a challenging proposition in practice. We have no way to accurately model realizable value for the majority of assets and current simple approaches produce extremely wrong estimates a substantial percentage of the time.
They’ll make up a number and make you spend money proving otherwise. The government won’t care about your inconvenience when they need the money.
Of course this is a prediction of something that hasn’t happened before but looking at the chess prices move this does appear to be an intended destination.
That isn’t the asset valuation, that is the range of assessments interested parties make with respect to asset value. Because market clearing transactions are rare in many asset markets, those extremely high variance estimates of asset value are all you have to work with. It is only marginally better than no information at all. Too make matters worse, the rare transactions in these markets frequently have a lot of complicated structure such that the nominal price is not reflective of the underlying value.
tl;dr: Many assets have no meaningfully assessable fair market value. These are investments with extremely long and indefinite time horizons before the asset value can be assessed in a reasonable way. You can look at it as a peculiar type of risk capital portfolio with an extraordinarily long time horizon.
imo, it's in the best interest of the market for people to have to realize their gains otherwise the price of an item is pretty imaginary if it's never realized.
Gains are frequently not realizable as a matter of law and/or contract, for good reason. Additionally, there are many assets with notional value conditional on not liquidating them, which makes them de facto not realizable. And of course, the majority of assets have no liquidity, so realizability is a practical fiction.
The unrealized values are a fiction. There is significant value in treating values as unknowable when they are, in fact, unknowable. Forcing people to make up a fake valuation creates a lot of adverse incentives.
Then instead of taxing the gains, you'd accept the government nationalizing the assets by eminent domain and paying fair compensation that was significantly less than the "fictional" unrealized value?
Or if someone unlawfully deprived you of the asset, you'd accept as restitution or seek as civil damages for the loss something significantly less than the "fictional" value?
Or, when it was no longer an excuse to avoid fair taxation, would that "fiction" suddenly be a lot more real to you?
It would be much better to tax the benefits of the unrealized gain that a person realizes.
It’s much easier to do because there is no disputing the assessment since the person implicitly agrees to the valuation. And it allows people to forgo realizing any benefit from the unrealized value at all to avoid taxation.
Say take x% of the top of the money lent to someone who uses their unrealized gain to secure a loan. Make the money paid count against any tax they owe if they sell the asset later.
Have you taken out a secured loan in the last year over x amount?
Take the value of the asset assessed by the bank and the price paid for the asset to find the total value that is unrealized gain.
Divide that by the total value to get percent of collateral that is unrealized gain. Multiply that by the loan value. Then multiply that by the tax percentage.
All you need is for banks to report secured loans to the IRS and it’s easy.
But if those skyrocket in price from tax, they'll be more subtle about convincing banks they're good for the money and pay a slightly higher rate for unsecured loans.
Or maybe they'll just treat the asset securing the loan as having the pre-gains price. Get the bank to agree it's worth at least what you paid, with no further analysis.
If you try to plug those loopholes you lose the "much easier to do because there is no disputing the assessment since the person implicitly agrees to the valuation" factor.
>treat the asset securing the loan as having the pre-gains price
That’s no different than if the asset had no unrealized gain at all.
>they’ll be more subtle
It only takes a small rise in interest rates before it’s cheaper to pay the tax—assuming the tax isn’t outrageous.
Unsecured are much riskier because of the way unsecured creditors are treated in bankruptcy, so they already have higher interest rates.
It would be very easy to tweak bankruptcy laws to make unsecured loans over a certain amount a bit riskier to increase the delta even more.
We also already have regulations governing how banks assess creditworthiness, and the percent of their capital they can lend unsecured based on risk. As well as the amount of unsecured loans they can make to signal individual. If necessary tweak those values.
Another easy way is to add a surcharge to large unsecured loans where the loan amount exceeds the taxpayer’s assets based on acquisition price by some large margin.
None of those impact implicitly agreeing to the valuation and they are all pretty easy to do.
> That’s no different than if the asset had no unrealized gain at all.
It lets you get loans based on 100% of your pre-gain money with zero taxes paid, which I think is too generous. You wouldn't do that if it was actually all your money. It mostly fits the idea of only taxing the "used" money, but not entirely, and I don't really favor that idea in the first place.
> It only takes a small rise in interest rates before it’s cheaper to pay the tax—assuming the tax isn’t outrageous.
15% tax is pretty big. And if we put capital gains back in line with income tax it would be double that.
> Unsecured are much riskier because of the way unsecured creditors are treated in bankruptcy, so they already have higher interest rates.
If the unsecured loans only go up to 90% of the post-gain asset value, that's not much riskier, is it?
> We also already have regulations governing how banks assess creditworthiness, and the percent of their capital they can lend unsecured based on risk. As well as the amount of unsecured loans they can make to signal individual. If necessary tweak those values.
Yeah okay we could stop unsecured loans from happening. That seems awkward though.
> Another easy way is to add a surcharge to large unsecured loans where the loan amount exceeds the taxpayer’s assets based on acquisition price by some large margin.
You are basically making my case for me. It is widely recognized that replacement value differs materially from notional value. In such cases you may be required to pay much more than notional value because you have to pay for liquidity costs that only exist when transactions are forced to occur. The act of transacting can intrinsically change the asset value, sometimes by a large factor.
Are you oblivious to the extensive litigation that occurs in cases like eminent domain because there are substantial differences of opinion on even the notional value, never mind the realizable value?
Notional valuations are fiction, everywhere and at all times. Treating them as some kind of objective reality is just enabling a lot of abuse and motivated reasoning.
If pegged to inflation then they are not, but I think they generally will not be pegged. People who might think this is great should understand that the government makes more money increasing wealth inequality aligning the interest of the government and the ultra rich.
So if company A pays company B to develop some software, that revenue for company B (or rather, its profit) is still taxable? Then it makes sense I guess.
No, you have it right. Software was getting a special exception from the normal rule that salaries spent on creating capitalized assets are capitalized (which is the general rule for most industries, as well as for software development in most of the EU).
If your payroll ends up being about the same, after 5 years it all evens out in the sense that you will be expensing 100% of your payroll each year (but the expensing will be 20% from each of the prior 5 years).
If your payroll is quickly growing You experience the problem on all payroll growth.
If your payroll is decreasing, you get a tax benefit. Your outgoing cash is less, but you are getting deductions from prior year expenses.
There is definitely a lot of misunderstanding here.
This provision can and does lead companies to owe significantly more in taxes than they make.
The only reason it hasn't been bigger news, is because most companies are pretending it doesn't exist and just sweeping it under the rug, hoping it will get fixed before enforcement gets serious.
I think the real reason it isn't bigger news is because the second you talk about tax code people start to tune out. It's easier to wind people up over AI taking jobs than it is to try and explain what amortization means.
> The only reason it hasn't been bigger news, is because most companies are pretending it doesn't exist and just sweeping it under the rug, hoping it will get fixed before enforcement gets serious.
Why pretend that it doesn’t exist? Why not vocally lobby for a change in the tax code?
There is bipartisan support to repeal the change. Meanwhile, further changes to the tax code are being prepared by the administration, very probably containing further such time-delayed footguns that will be the problem of the next administration to clean up, making them look like they raise taxes.
My reaction to learning about this is that it is good news: this explains the weakening of demand for programmers, but unlike the AI explanation, this explanation does not come with a large risk of the demand becoming much weaker than it is now.
Also, finally programmers with the right to live and work in the US catch a break: salaries for US-based programmers can be amortized over only 5 years as opposed to the 15 years of non-US programmers.
I mean, the link not many people have made is that executives want to replace programmers with AI because of Section 174.
It has effectively become a lot more expensive and difficult to employ a programmer. Once this change went into effect we started to see hundreds of thousands of layoffs.
Then tech executives started aggressively talking up how you could use AI to write code instead of having humans write it.
Now of course reducing headcount and the associated expenses and replacing them with a bot sounds tempting to executives no matter what. But it sounds REALLY tempting when you've been on a hiring freeze since 2022 due to the fact that you can no longer deduct employee salaries in the year you pay them out.
Bear in mind that both Republicans and Democrats say they want to fix this and haven't done so due simply to gridlock and government incompetence.
I think most software businesses are taking a wait and see approach. Don't hire until this thing gets fixed. In the meantime, double down as hard as you can on automating those programmer jobs out of existence, in case the law never gets fixed.
> For cash-strapped companies, especially those not yet profitable, the result was a painful tax bill just as venture funding dried up and interest rates soared
Can someone explain this? What taxes do unprofitable US businesses owe that this would be deducted against?
Here's a toy example that hopefully makes this clear:
In 2024, your business has $1m in revenue and has $2m in expenses. 100% of these expenses are R&D salaries (engineers you hire.)
Your company loses $1m/year. (You brought in $1m and spent $2m.)
Under the old rules, you'd owe no tax because you were unprofitable.
After Sec 174, what the IRS now says is:
You had revenues of $1m. But you only had $400k in expenses (because you now have to spread that $2m in R&D expense over 5 years).
So actually you had a profit of $600k! And you owe tax on that $600k profit (~$120k)
So you now have an additional $120k tax expense, making your business even more cash-flow negative.
.
Amusingly, if you're pre-revenue, none of this matters (you have no income at all, so it doesn't matter what your expenses are.) You get hardest hit by this change when you have some revenue and when you do a fair bit of R&D.
They're still expenses, they just now need to be amortized.
Buying a truck is an expense, as is buying gas for the truck. But the former you have to amortize over x years, the latter you can expense immediately.
The law used to be "employee salaries for software are like buying gas" and now it's "employee salaries for software are like buying a truck".
The critical difference is that the business owns the truck but not the employee. The amortization assumes that the asset can be sold for value. An employee can quit at any time for any reason. You don’t retain the right to their labor for five years.
If they're producing a capital asset, you do retain the right to the fruits of their labor, even if they quit.
The rationale behind amortization isn't exactly the idea that the asset can be sold, it's that the asset is producing revenue over multiple years. For software, the asset is the codebase.
Let's say you hire a single software dev, for one year, and they write Excel++, which you can sell for the next ten years. It would be entirely appropriate to amortize the cost of creating that software over those ten years, based on the matching principle (a fundamental idea of accounting, matching expenses with revenue).
The issue in the real world is that's not how the software industry actually works, 99% of the time.
As anyone that has ever sold software IP knows, most of the value is vested in the person that wrote the code, not the code itself. The code is not a factory, it is the output of the factory.
> ...most of the value is vested in the person that wrote the code, not the code itself.
You must have misphrased what you intended to say. If what you wrote was true, a software company's most valuable asset would be the specific programmers in its employ. If true, average tenure of a programmer would be way longer than 1.5->2 years as companies worked really, really hard to keep their most valuable assets from walking out the door into the doors of another company just to get reasonable pay increases.
Perhaps your opinion is influenced by doing post-collapse-sales of a whole bunch of software houses that built just plain bad software? I can't see why else you'd be selling "software IP" independently of the rest of the business.
Anyway. Given that information, how should you have phrased what you wanted to say?
Most programmers do approximately zero work that is R&D. The most you lose if they walk out the door is institutional knowledge.
On the other hand, I've worked almost exclusively on software R&D for decades and seen the loss of a single person effectively end a project even when the software was essentially finished. Software R&D is about developing abstract knowledge, concrete implementation code is just a useful byproduct of that since R&D is typically motivated by a specific novel requirement.
If the software IP that results from R&D is not core to your business or a competitive risk, there is money to be made by licensing it. I've licensed this type of IP to big tech companies a number of times. If you are not actually doing software R&D, you are unlikely to be in a position where this is a possibility.
In almost every IP sale and licensing deal for software R&D I've seen, the value of any code is almost entirely conditional on retaining the services of person(s) that designed and wrote it. The entire "acquihire" phenomenon is an explicit admission of this. This is true even when the code is in a mostly finished form. Companies are buying capability, not revenue, so the code can't be a black box to their engineers. Companies usually spend more to acquire people with the code knowledge than the actual code.
The practical reality is that it is difficult to reverse engineer abstract knowledge from a concrete implementation. No one wants your code per se, they want to adapt your code to a different application that requires having a deep understanding of the domain the code represents -- they don't know what they don't know.
If you are just grinding out software that could be vibe coded then there is minimal asset value being created in the software artifacts. Anyone else would be better off reimplementing it themselves.
So yes, almost all of the value of code produced by software R&D vests in the people that wrote it. This is evident across many software IP transactions.
Based on my exchange with wdaher, who seems to understand this well, it's a bit more subtle than that:
The salaries are of course expenses, but they are exactly offset by the value of the IP created by the R&D activities.
It's a bit as if you spent money on buying some materials. As long as the material doesn't degrade, the cash is gone but the value is the same and therefore won't reduce your taxes.
If that IP is amortized over a single year, it does not contribute to taxation, but it does if it is amortized over a longer period.
They are expenses, but amortized over 5 years. So if you spent $2m on employee salaries, you would then deduct $400k from your revenue every year for 5 years.
If your employee expenses remained constant, then by year 5 you would be deducting $2m from your revenue since you'd be accumulating the deductions from the previous four years.
So in steady state it wouldn't necessarily be a big problem. But for a startup which is hiring many new employees and whose revenue is growing it's a huge problem.
Buildings have to be depreciated, so probably? If you have to depreciate a building if you buy it, why should you get a free pass just because you built it yourself?
What other cost do you think goes into software development? Companies are not spending that much money on IDE licenses. The vast, vast majority of software/R&D costs are labor
Elsewhere in the world (under IFRS accounting rules) capitalization of R&D costs has been a firm requirement for a while. The US has been somewhat unique in allowing them to be expensed instead, until recently.
Yeah, seems I was wrong about that. Apparently most IFRS countries allow expensing R&D for tax purposes, regardless of accounting. Many even have an R&D superdeduction nowadays.
You can deduct 100% of salaries paid 5 years ago, but only 20% of salaries last year (etc.), and since companies tend to hire more people over time, most of your expenses will have been in the last few years that are still amortizing. You might have enough losses to carry forward in your first year of revenue, but 6 years in that could run out. It depends on the exact circumstances.
No, that's literally the Section 174 change. You now must count them as R&D.
The relevant paragraph from Section 174:
>
(3) Software development
> For purposes of this section, any amount paid or incurred in connection with the development of any software shall be treated as a research or experimental expenditure.
What "in connection with" means is vague. I think a reasonably competent tax attorney could probably argue that the costs of running your production cloud serving existing customers don't count, but IANAL.
If the business has some revenue, but is not yet profitable after deducting development costs, it can become profitable on paper (and owe tax) if R&D is capitalized instead.
That is kind of strangely worded, but I think I see what they're getting at.
Say you would have been exactly not-profitable ($0) if you could expense all of your R&D as in the old system, therefore avoiding tax. Now with the new rules you may be on-paper profitable because you can only deduct 20% of the R&D as an expense this year. The remaining 80% of that expense tips you over, becomes profit, and that's taxable.
Right. With concrete numbers, say your main expense is $1 million in developer salaries and you have $500k in revenue. Going by the previous rules, you have a loss of $500k and don't owe income tax. With the new rules, you can only deduct $200k of expenses which gives you a "profit" of $300k, on which you'll owe $62k in taxes.
What taxes do unprofitable US businesses owe that this would be deducted against?
An unprofitable business doesn't pay income taxes. Businesses are taxed on their net income (i.e., profit).
People are railing against this as the cause of tech's recent underperformance, but it was a non-factor for the vast majority of tech companies, because most tech companies aren't profitable and wouldn't have paid taxes anyway.
From what I understand, this does not actually affect Google. They were already amortizing their R and D expenses.
Over long time scales (and big company revenue streams), this is sort of a wash. I think this hurts startups a bit more due to the long timescales involved which eats up much needed cash in the short term.
For me the worst things is that they treat all software as R&D. I understand in maybe some situation it could be abused but imagine established company having non innovative software that keeps engineers only for bug fixing and security patching and basic maintenance. In true spirit this is not research for sure. It's equivalent of someone having a hotel and suddenly telling that their cleaners, security, gardeners, receptionist qualify as R&D which would be nuts.
AFAIK it was also affecting more freelancers outside of US since amortisation is 15 years. For EU citizen IMHO this is equivalent of US putting tariffs on outside world. I wish EU at least try to fight back and revenge on US Tech by increasing taxes or also making all US tech bought by EU companies to be 15 years amortised so they have taste of their medicine.
It seems that there is quite a bit of confusion about this. What this does is that it reduce your deductible cost in the tax year.
First you have to make a profit (tax is on profits).
Secondly, what this does is to limit your software development expenses for tax purposes in the current year because the development cost is seen as a capital cost that will be amortized over five years opposed to operating expenditure in the same year.
If you are a startup and not make profits, then the loss will be less in the current year, but either way, your tax liability is the same: $ 0.
So software development is moved from opex to capex.
I can see why it would affect startups not making a profit but why would it dramatically affect FAANG (e.g. some of the most profitable companies in the world that have been running for decades)? The article contributes all these large layoffs in FAANG, in part, to this tax rule.
There's a difference of $24 but I have $1200 in cash reserves. And I make up the difference later. Oh no! Guess I have to lay off 10% of my employees now.
I made one of the original posts on HN about this years ago after hearing about it from my CPA. Both then and now these changes make zero sense to me as a matter of good policy. I am also still surprised at the number of people in tech who either haven’t heard about this or are willfully ignoring it and likely filing their taxes incorrectly.
Also, the 10+ years before the layoffs started tech companies were on a hiring binge. Much of big tech was hiring to keep people off the market and off their competitors payrolls (this is from friends of friends in FANG HR departments). These were high paying jobs, too.
In knowledge-based white-collar work, there has been a significant increase in productivity in recent times. Tasks that once took days or even weeks—such as research, content creation, and visual generation—can now be completed within minutes. At the same time, both the speed of production and the quality of output are continuously improving. The outcomes are unavoidable.
It should be illegal to post graphs that start in 2020 when talking about tech hiring trends. The relevant comparison is probably the 2014-2019 era, not the peak pandemic craziness.
The most fascinating question is not "How did a single line in the tax code help trigger a tsunami of mass layoffs?" but how did a single line in the US tax code help trigger a tsunami of mass layoffs in other countries?
Yes, I think there are two similar but subtly different flaws in that always.
1) Accounting rules are to match revenue with the expenses responsible for them, which I think is a good principle. If your workers make something now that provides revenue for 5 years, it makes sense to spread that expense over 5 years too. In many cases, you would want to do that as a business, makes it more clear how your business is profitable vs not.
2) Decisions whether to "build vs buy" a capital asset should not have massive tax implications. If I buy CoolSoftwareProduct from someone and resell it for the next 5 years, I'd have to amortize that. Should be similar if I hire a coder to write CoolSoftwareProduct instead.
(This doesn't mean that "salaries should always be amortized" is the right answer, of course, I think it's a very silly law)
I would say, yes there is a flaw there, because salaries are often a huge chunk of R&D expenses, and for the sake of long term growth, we want to disproportionately incentivize R&D spending
How about the salaries of employees being paid to create a new line of business? Say, the business runs restaurants and you decide to break into the tax software business. Can you avoid taxes on restaurant profits right now in order to build your new unrelated venture. ISFICR, tax law allowed outright deducting of costs of the current business, but not costs for starting a new one. Not deducting the new costs against the old profits.
After that, we can nitpick: should the development costs of new software be encouraged the same as maintenance costs of existing software. If you want to encourage startups, then yes they should. If you want to discourage startups or very temporarily increase tax collection, then no.
Discouraging starting new businesses would be unconstitutional. All freedom in the US is derived from being able to participate in controlling capital.
>The delayed change to Section 174 — from immediate expensing of R&D to mandatory amortization, meaning that companies must spread the deduction out in smaller chunks over five or even 15-year periods.
Doesn't this just amortize out to be roughly the same amount of deduction over the long term?
All the big companies mentioned should be relatively unaffected over an N>5 year time period. Also this was something that's been in the works for years so their accountants should have been planning for it so it wasn't a financial shock (and company financials seem to indicate no such shock).
If you look at the time value of money[1], a $1,00,000 deduction this year is worth more than $200,000 deductions over the next 5 years.
But more importantly, the article claims it was used as a tax shield to grow.
"Basically, as long as spending counted as R&D, companies could report losses to investors while owing almost nothing to the IRS."
"Once those same expenses had to be spread out, or amortized, over multiple years, the tax shield vanished. Companies that were still burning cash suddenly looked profitable on paper, triggering real tax bills on imaginary gains."
Sure, but that doesn't account for the allegedly apocalyptic layoffs from companies that don't fit into the "real taxes on imaginary gains" mold.
I get that this is bad for the VC monopoly bucks scene, but they were already down for the most part. If the changes are as the article alleges than all these big tech companies that are posting huge layoffs should mostly be fine because it's not a serious change from status quo for them.
My company was affected. The amount of money paid in taxes more than quadrupled from one year to the next.
It hurt small businesses that were slightly profitable. No one else.
VC shops aren’t profitable anyway, so no taxes to pay.
Microsoft took a 4 or 5 billion dollar write off, but they can literally write a 5 billion dollar check.
The issue is that the IRS wants you to pay them today on profits and cash that literally don’t exist. You make $1M in revenue and pay 5 developers 200k/year? You have no money left at the end of the year, but you pay taxes as if you profited about 900k.
From the horror stories I read from other founders, the bank wouldn’t loan on those numbers.
When we asked our accountants what they were seeing from other companies, the answer was “mortgage their house.” That assumes they had enough equity to mortgage.
Interest rates are bigger motivator of the layoffs than these changes. When interest rates are high that means investors far more heavily prioritize profits today over profits tomorrow.
> Doesn't this just amortize out to be roughly the same amount of deduction over the long term?
With steady enough employment numbers, sure. Google has a weird one-time cost where they get hit with extra taxes at 80%, 60%, 40% and 20% of their employee's salaries for five-years and then it's all balanced. You can turn the money Google needs to borrow (or not invest) at some interest rate into a known number.
Any startup that is cash poor and especially one that is growing struggles. In year 3 you get to write off 20% of year 1's salaries, 20% of year 2's salaries and 20% of year 3's salaries.
> Doesn't this just amortize out to be roughly the same amount of deduction over the long term?
Yes, but if your business is not yet profitable, having to pay tax on money you don't actually have in the bank (because expenses exceeded revenue during the year) will cut into your runway, perhaps to the point that your company might not exist in five years... or even two or three.
Google, Facebook, Microsoft and many other of those old big companies are profitable though and they dont go anywhere in next 5 years (even if first 2 bleed out users)
Yes, if you are a profitable company operating at a steady state and your investors have a time horizon of (in other words, are locked in for) a decade or more.
Most companies in question don't fit these criteria. They are either large public companies subject to the reactions of the market to quarterly earnings, or small private startups that have limited cash (a runway of far less than 5 years) and are facing a perfect storm of a historic rise in the cost of capital coinciding with this change.
In either case, their cost of labor just went up by a lot and will continue to cause layoffs, labor market shrinkage, and diminished ability to develop new products.
Yes, big tech companies are lesser affected and they were already amortizing their expenses as “cap labor”.
It’s a pretty bad article general and to blame the law change on this is all kinds of disingenuous:
“It’s no coincidence that Meta announced its ‘Year of Efficiency’ immediately after.”
If this article is accurate it doesn't sound like it. The change was a political tactic to make the tax bill it was part of comply with Senate budget rules on paper. Apparently this is a common tactic with tax bills, with the expectation that the changes will be repealed or altered in a later bill. There is a movement to repeal this change, but the effects have already been felt.
> Apparently this is a common tactic with tax bills, with the expectation that the changes will be repealed or altered in a later bill.
None of this adds up. You're saying, the legislators were trying to cheat and because it's a "common tactic" that kind of cheating is somehow good, but it's bad when the cheating doesn't go through?
On the other hand, being a common tactic implies that the possibility of it remaining in the books was well understood, and the declared "expectations" carry zero weight as evidence, even less than zero when coming from politicians.
Legislation like that has far reaching consequences and pretend "surprise" just confirms the intent behind it. It's only prudent to assume that we have a common tactic case of throwing sheet at the wall to see for how long it'll stick. If there's no backlash the "tactic" will remain there forever.
As another example of the same common tactic, consider the fact that all popular browsers have been used as Trojan horses into the users' local networks for like forever. At some point back in 2015 somebody objected so the browser makers started talking about fixing the problem but then stopped talking without fixing it because public opinion moved on to other areas affected by abundant sticky materials... Thus, that particular sheet remained on the wall for another 10 years and counting, and the story may repeat itself again.
When using bill reconciliation in order to avoid Senate filibusters to pass a budget, certain conditions must be met otherwise regular Senate rules and the need for 60 Senators to be onboard to avoid a filibuster come into play.
It's not cheating, it's playing by different rules to get most of what you want/need done and then sometimes those that played and gambled were intending to, or hoped to, make the changes later that require rules. Their hope is that 60+ Senators would be onboard for those changes because they (those that gambled and pushed the budget bill thru) managed to get what they wanted at the expense of #$%#ing something up that most others would then be willing to fix/address.
Agreed. If the members of the majority party can compromise within their single-party system, and play by certain rules, everyone else is powerless to amend or block the legislation.
In tax policy, every single change looks reasonable to one interest group, and like a cheat to a different interest group. That is just the nature of tax policy. Any change hurts some people, harms others.
Changes to Section 174 happen rarely and are not a “common tactic.” Changes to tax policy in general are common, especially in the reconciliation process. They can have unforeseen side effects. As well as side effects that are foreseen but considered more acceptable than other side effects.
> You're saying, the legislators were trying to cheat and because it's a "common tactic" that kind of cheating is somehow good, but it's bad when the cheating doesn't go through?
I don't think GP made any kind of value judgment either way; they were just stating how things seem to usually work.
Around ~2010 I still had a lot of coworkers who claimed tech was basically incapable of defending its interests against other sectors. Maybe a bit different than today. I don't doubt that they thought they would get this repealed, but I would suspect the risk of the live grenade went to the sector with the least lobbying competence per revenue for the tax equations.
Bloomstink has a short article on R&D expenses/tax credits as does Reuters on some of the back and current history.
But just as an accounting note: R&D expense has nothing to do with the company having revenues for an existing product, which already is allowed to deduct cost of goods sold, selling and admin expense. It is a cost related to future business and in that regard, it is not crazy to say it should be amortized. That in the past this did not happen, or that accelerated depreciation for other assets is in the IRS code is a function of the government wanting to effectively subsidize business investment.
The chef doesn't create a meal once that you can sell for the next 10 years though. You pay him for time X, he makes a meal, you sell that meal.
That's fundamentally different from regular software development outside of agencies where there is no direct relationship. Software development is closer to an investment than an expense.
Amortization sucks in general, yes, because the money is gone and it doesn't affect your taxes to the same amount, but that's not different for any company doing manufacturing or anyone needing specialized tools or vehicles that cost significant amounts.
I think there is an argument that both could be valid. I wonder why we cant let the company decide to pick one over another and not be so fixated on one tax code to rule them all.
That is production, not R&D. Basic research into, for instance, semiconductors may one day lead to the production of a new product. The R&D costs would be amortized, the eventual production costs would not.
Deferring depreciation and deductions decreases their value as inflation happens. So it is a double whammy-not only is your profit reduced this year (and the impact that has on your stock price or return to your private equity investors) butthe value of that deduction decreases over time. So it is not a 'wash'.
I was part of a small R&D company that had a promising product (can't say more, NDA) and we had to shut down because of this. Thankfully the founders were able to get us acqui-hired or I'd be in a much worse position. But that IP is just lost to history AFAIK, in spite of significant investment of US research $.
Honest question, is there a community / grassroots effort I can participate in so that this this section 174 change can be reverted to its pre-2022 state?
I'm wondering, if such a movement doesn't doesn't exist already, do I need to start it myself?
- Gather up about 10 million dollars (more will help)
- Bribe the right people
I hate to provide such a cynical and lazy response but we've got until midterms (maybe) before you really have a shot at 'democratically' influencing the system. For the time being you'll have to work with the mafia that's currently running things and outbid whoever wanted this to happen in the first place.
Ironically the debate/discourse here is healthier than anything we see from US Congress. Presidents have a limited number of terms they can serve, but there is no limit on the House and Senate and change is hard if not next to impossible because of this. Term limits would be a good start to introducing positive changes, but good luck finding the necessary majority to vote against their power and very cushy and comfortable lifestyles.
Question, if you have to amortise it over 5 years, and you can survive the initial 4, does it break even in year 5 (assuming stable employment)? Ie you amortise the previous 5 years (20% each) which works out to 100% anyway?
This US American idea of paying taxes = giving money away is weird. Especially coming from tech people, as what is preventing other nations from taxing the hell of big/US tech is the US government and its threats.
> “I work on these tax write-offs and still hadn’t heard about this,” a chief operating officer at a private-equity-backed tech company told Quartz. “It’s just been so weirdly silent.”
Hasn't Ben Thompson of Stratechery spoken about this a number of times? I'm aware of this 'feature' and I'm not even in the USA, let alone a COO at a private-equity-backed yada yada.
I think people are misinformed about how to deal with 174. That said, yes, a repeal would be a good idea. Hopefully that happens through BBB in the next month or two.
You do not have to amortize 100% of your engineering costs. Not even close.
Here's the key:
Development costs incurred to remove uncertainty are amortized.
All other costs are deductible during the tax year where they are incurred.
How does this work?
You are going to design a new robot arm.
In January, you spend $100K to "remove uncertainty". In rough strokes, this means discovering all the things you don't know and need to know for this robot arm to become a product. This amount will be amortized over five years under 174.
Now, with uncertainty removed, you spend an additional $1.1MM from January until December for engineering implementation. No uncertainty being removed. Just building a product. This is 100% deductible that tax year.
Analogy: You want to build a new brick wall with specific properties. You spend $100K to develop a new type of brick and $1.1MM to build the wall using that brick. The $100K is amortized, the $1.1MM is deductible in one shot.
BTW, at year 6 the amortization schedule reaches steady-state and you are amortizing the full $100K every year. In other words, the impact of 174, if treated intelligently, is the time value of money until steady state is reached for the engineering costs incurred to remove uncertainty.
If anyone cares about combatting government propaganda in the US, do this:
Query any search engine for "are US income taxes direct or indirect taxes"
Every one will tell you that they are direct taxes. This is false. The supreme court has exclusively held that income taxes have always been indirect taxes (excises specifically, read about what an excise is in any authoritative source on tax law) in a constitutional sense. (See Brushaber v Union Pacific RR Co. 1916, Moore v U.S. 2024)
The sixteenth amendment did not give congress the power to directly tax citizens (or domestic corporations) and the complexity of the tax code is an attempt to obfuscate this fact, but the code is not inscrutable, it has rules.
Unsure of why this matters? Look up the difference between direct and indirect taxes in US law. None of these deductions matter unless you are a foreign corporation. I have tried commenting about this in other income tax related threads (this is my alt account), but people here don't like the idea that there is government propaganda in the US, or that most people are wrong and blindly accept the socialization about taxation without verifying what the law says.
I realize this is a disturbing truth to accept, not least because it involves accepting that most people who have been prosecuted for income tax crimes are only guilty of ignorance of the true legal purpose of the forms they signed. You can easily verify that most accountants and tax attorneys do not know what they are talking about by asking them this simple question about direct vs indirect taxation.
Note that Trump's Big Beautiful Bill as it passed the House of Reps would bring back 100% expensing of R&D expenses including software development costs/salaries.
If the article is to be believed, though, the damage is already done. Companies have already laid off large portions of their R&D staff, and have canceled lots of forward-looking technical work. Re-hiring those people and restarting those projects can take years, and that's if companies feel confident enough that the exemption will stick around, and not get removed again in a few years.
Meta: god articles like this drive me crazy. What ever happened to the inverted pyramid hierarchy of information? I have to read 3 paragraphs of unmitigated filler before they actually tell me what's changed. It's not just this article, it seems like every article on newer media sites is like this. I understand why, but fuck them very much and the incentives that drive this behavior.
So, we want incredibly profitable companies like Google, Microsoft, and Apple to take their software development costs and subtract that from their tax bill?
These are the same companies that file patents so nobody else can use the ideas that they developed at the expense of public services.
How about making it a tax break only for small and medium sized companies?
Oh interesting. In 2022, the company I worked for folded because a primary investor spontaneously pulled out. We were nearly 100% R&D and the sudden change in relationship was surprising.
Let me guess - the keyword here is "Section 174", just from the title alone :)
Dealing with Section 174 amortization in those first one to three years is a real headache (and your tax bill ends up higher than if it didn’t apply). Once your startup survives that the first few years of doing Section 174, things do get easier... but, sadly, most don't make it that far.
Its so funny to me that people freak out about amortization when I spent several years at a public company having to document my work as being R&D to amortize it to make our EBITDA look better.
How this impacted our business is that when you are doing next year planning, and the goal is to grow the business, it made ads and other marketing investments more appealing versus tech hiring to expand product capabilities.
bit of a self-own it seems. Start-ups and early stage companies might simply decide to start in a more friendly tax jurisdiction. E.g.: Switzerland offers a 135% deduction on R&D-related salaries in the year they are incurred, making it an attractive location for tech development
EU provides a large pool of experienced developers seeking new opportunities on salaries well below SV. Why pay 500K for a burnt out "rockstar" who spends more time on twitter than doing actual work when you can hire highly skilled people in Eastern-EU (or even in Berlin).
Section 174 seems unlikely to progress unless attached to broader legislation.
> "More promising is the Tax Relief for American Families and Workers Act of 2024 (H.R. 7024), which proposes restoring immediate expensing for U.S.-based R&D investments through the end of 2025. " -- https://www.pwc.com/us/en/services/tax/library/tax-committee...
That's a good policy but Switzerland is awful for startups: expensive, strict labour laws, few funding opportunities, risk-averse customers, fragmented European market.
If I could start anywhere in the World, Switzerland would be above all the war-torn and crime-ridden places, but business-wise it's no good for a tech startup.
Investors and stock holders should be extremely outraged that all of these businesses are knee capping their future profitability. Can't make all those future pension payments if all your investments can't stay relevant in the market.
Some people will point out that AI will fix this, no it won't:
1) The real cost is higher than anything you'd pay for a person an there is not likely any real change there.
2) AI will be lies like Actual Indians that won't scale
3) Here's the kicker: If AI does succeed, now these multi-billion dollar firms will have to compete with multi-billion dollar single person businesses, that eat their lunch
Its a race to the bottom right? That means you need to invest in the business and all these layoffs are exactly not that, and will leave companies unprepared for the next 10 years.
Indeed. This is only a problem because software companies try to classify most of their expenses as "R&D" only to look more profitable than they really are, but at the same time they don't want to pay taxes on those supposed profits. For honest companies which don't capitalise their wage expenses, nothing substantial changes.
What really changed things was the end of ZIRP [1] and even then it was opportunistic. Labor costs are a massive cost for tech companies. They have continually tried to suppress wages. In the 2000s, it was the anti-poaching agreement between Steve Jobs, Eric Schmidt and others. In the 2010s, high growth ahnd zero interest meant labor costs continued to balloon.
But then Covid came along and was a massive opportunity. A few companies may have needed to do layoffs but that created the opportunity for everyone else. Big Tech just went full Corporate America with a page straight out of Jack Welch: fire the bottom 5-10% every year. Call it "layoffs". It's a direct pay decrease for those who remain (who get assigned the work). Those are still there won't be asking for raises because they're now afraid of their jobs.
Very little of this was ever necessary. None of the big tech companies ever came close to making a loss. They've remaining insanely profitable, in total and on a per-worker basis. At different times Google's per-worker profit has approached or exceeded $1 million.
The other factor is these companies eventually reached their size limits where antitrust stopped them making any more significant acquisitions.
Consider the timing: this change came in 2017. Where were the mass layoffs in 2018? 2019?
Also, the 2017 tax cuts contained a massive tax holiday for the repatriation of foreign profits.
Mass layoffs are simply wage suppression. It's the end state for any company that can't keep growing the way the market demands: eventually it comes down to cutting costs to make those quarterly profit targets. And in that, they sow the seeds of their own demise.
Big tech companies are both doing mass layoffs AND hiring. How does this fit the narrative that the tax change is at least in part responsible? The new hires still have the same deduction issue, right? So what impact does this really have?
Think of it this way: if this passes, will the layoffs end? Or reduce? Absolutely not. All this does is give line the pockets of shareholders. That's it.
I'm a big fan of tying certain benefits to NOT doing layoffs. This can include:
1. You get this deduction only if you've fired fewer than 1% of your workforce in the last calendar year;
2. You don't get to sponsor for an H1B if you've conducted ANY layoffs in the last calendar year; and
3. The tax deduction only applies to unionized workers.
And while we're at it, let's roll back this ridiculous tax structure where IP can be "sold" to a subsidiary in Ireland and then royalties paid.
That's a solvable problem and probably already solved. Fire more than a certain threshold of your employees over a certain period for any reason and it's a layoff in effect, say 3% over 12 months.
> Big Tech just went full Corporate America with a page straight out of Jack Welch: fire the bottom 5-10% every year
Plenty of "big tech" already did it. Microsoft could not be more famous for stack ranking dating back to the 90s. Amazon have long had that kind of culture too.
This doesn't quite fit into the article and is probably too inside baseball for a general business audience, but as I see it, there’s a real and serious argument to be made here about how Section 174 changes restructured the cost architecture of tech employment (yes, even for big, cash-rich companies). When salaries could be fully expensed, the effective marginal cost of headcount was lower. Amortization means the same engineer now triggers a significantly bigger near-term tax bill. At scale, that’s a serious shift in how labor costs flow through the P&L… functionally, op-ex becomes capex, and cash flow implications for big players run into the billions. But maybe it’s me!
Isn't this literally the content of the article? What you just wrote down is basically this paragraph from TFA:
> And so, on schedule in 2022, the change to Section 174 went into effect. Companies filed their 2022 tax returns under the new rules in early 2023. And suddenly, R&D wasn’t a full, immediate write-off anymore. The tax benefits of salaries for engineers, product and project managers, data scientists, and even some user experience and marketing staff — all of which had previously reduced taxable income in year one — now had to be spread out over five- or 15-year periods.
Only external LLM use is ‘always deductable as op-ex’. If you build your own server farm and/or developer your own LLM, those are capital expenses which must be depreciated.
Contractors licking their lips at the prospect of being a clients op-ex. I think you’re right and hence the slow down in hiring top talent for top dollar.
so will this incentivize a return to revenue/profit-driven business models? will we start to see a reduction in venture capital burning money on revenue-negative startups?
I have to say I do have some sympathy for what's happening in the US at the moment.
Trump is just getting started. By the time he is finished, your economy will be shot to pieces. The US dollar will no longer be the reserve currency for global trade.
This doesn't explain the mass tech layoffs. According to the article, the rule applies to R&D. The vast majority of tech workers laid off in the last two years didn't work in research and development. They wrote regular software for sale, like games, for example.
The games industry, while hugely profitable and bigger than TV, movies, and music combined, laid off tens of thousands of people. It's unmitigated greed is all it is.
> For almost 70 years, American companies could deduct 100% of qualified research and development spending in the year they incurred the costs. Salaries, software, contractor payments — if it contributed to creating or improving a product, it came off the top of a firm’s taxable income.
According to the article, as long as the tech workers contribute to improving or creating a product (be it games or apps), they count as R&D cost.
I worked in games 2 years before the studio shutdown. It wasn't because of "R&D" tax breaks. None of the recent layoffs or studio closures are explained by that. Nor are the Microsoft, Dell, or Intel layoffs which aren't game-related.
To qualify for R&D tax breaks, IIRC having identified qualifying work for a segment of my firm, there must be elements of hypothesis, experimentation, results, etc that I would consider more science-y 'Research' than just turn the crank software 'Development.' It has to be both. And that has to be documented. And offshore research+development doesn't get you a tax break. The irony is that the R+D tax actually discourages onshore pure development as a 'trade' and encourages a split of onshore R+D and offshore D.
This sort of thing appears to be self-reported; I don't know if it ever gets audited. I don't know if big tech lies or creatively interprets what counts and that has contributed to the issue. But this article sort of over-represents what qualifies as R&D for US tax purposes.
Which makes sense. Software is functionally a capital asset, so really it should be depreciated across the length of the copyright term (unless the company wants to release it to the public domain to fully depreciate it early).
Maybe software should be a capital asset, but these depreciation rules don't fix that issue.
The rule says if you pay someone $200k to develop software: then you now have a $200k asset that then devalues to value of $0 over 5 years (starting midyear). That's just plain weird.
For our example a depreciation table might look like:
The final effect of the 174 rule change is that you still finally end up with a software asset worth $0. However you now have taxable income of $200k in year one and expenses equalling $200k spread over 5 years. The taxes paid could be a lot: although the taxation money is really just being lent to the government for a few years at 0%. The actual financial costs are fucking complicated.
Understanding accounting and taxes are two absolutely essential skills if you ever wish to be a founder (and useful anyways).
Finding a solution to dealing with the valuation of assets is difficult. The historical solution of depreciation is broken for software, intellectual property and goodwill. In theory, taxes on dividends and capital gains taxation already deal with the issue (company taxation at x% kinda ends up at $0 because the shareholder pays y% and claims back the x% through imputation).
Right, that weirdness is why it should be depreciated over the length of the copyright term. You spend $200k this year, and now you have a useful asset for the next 95 years (or 120 years if you never publish it).
If it turns out it's not useful, we could then allow companies publish the source and release it into the public domain to immediately "destroy" the asset (the copyright) and claim their deduction. So failed r&d projects would be deductible right away as long as the public gets them, and ones that result in a useful asset get depreciated based on how long they actually last, which is currently potentially multiple lifetimes.
I don't think copyright term is a good rubric/measure here. For SaaS, a company can keep the software locked up indefinitely, regardless of copyright term. Employees can be contractually obligated not to publish source code, even if the copyright has expired.
Amortizing development cost over the useful life of the software is maybe a reasonable thing to do (I don't think it is, but let's for a minute say I agree), but determining "useful life" is not simple.
I get your thinking here but copyright isn’t the only relevant intellectual property constraint.
Software built by a business is a trade secret independent of its copyrightability. Even after the expiry of copyright a business can continue to exploit it as a proprietary asset.
At the last couple of companies we worked at, they just sent out surveys on what time went to different activities. We couldn't possibly fill that out honestly, as that wasn't tracked.
Which, I think is an overlooked part of this. They must constantly have gotten feedback that people were lying to them.
> The vast majority of tech workers laid off in the last two years didn't work in research and development.
I bet they were classified as R&D for accounting purposes. Product development largely falls into R&D - it doesn't matter what the product being developed is for.
Every job I had at a megacorp was classified as R&D, and I know because I had to track hours against such.
> I bet they were classified as R&D for accounting purposes.
It's not just that. Section 174 now explicitly calls out Software as always being an R&D expense:
> (3) Software development
> For purposes of this section, any amount paid or incurred in connection with the development of any software shall be treated as a research or experimental expenditure.
Greed is too easy as a target.. industry space has shifted because of slower innovation and less growth, so cost cutting being more a focus would be a reasonable strategy
If your company is already profitable to the tune of billions annually, "cost cutting" isn't necessary. You're just cutting people out of jobs and out of economic participation in society -- which affects a far larger group than just themselves when those folks can't spend their salaries in other businesses.
There is no justification for "cost cutting" when it hurts the larger economy. If the company were losing money, that would be different, but these mass layoffs are all from firms that make obscene, enviable levels of profit. It's greed.
If a company is making $10 billion in annual profits, and they discover that they're spending $100 million on a useless project, are they morally obligated to continue that spending indefinitely?
There is no justification for "cost cutting" when it hurts the larger economy.
It is not good for the economy to have people doing work that doesn't produce value.
> It is not good for the economy to have people doing work that doesn't produce value.
This is a political statement, not an economic one. What is or isn't good for the economy is up to the goals of that economy. In Japan, for example, they've more or less adopted the opposite principle as an important plank in their political system. Or perhaps it would be better to say that they've adopted the idea that people having jobs is more important than those jobs having a direct connection to some measure of productivity.
1. I start “Facebook for dogs” It’s gonna be massive. For the first year me and five guys code away in the garage and I use my savings / credit card / family trust fund to pay them 100k each. Expenses are 500k, revenue is, amazingly, 1.5M and taxes owed is 500k.
At this point turning round and saying the development was R&D, and claiming 500k of tax breaks is just (to me) ripping off the American Taxpayer.
And I’m not even an American Taxpayer.
If the revenue was zero would anyone suggest that the taxpayer give me 500k to help ? (Ok I would because I like free money but most people won’t)
Everything you just said but imagine revenue is $500K, and you spent $500K on salaries for the team.
You can only expense $100K of the salary costs this year, so even though you're break-even, you pay taxes on $400K in income.
Or, even worse, imagine revenue is $250K, and you spent $500K on salaries for the team.
You can only expense $100K of the salary costs this year, you're already -$250K on the year, and now you're paying taxes on $400K in income. You're destroyed.
VC-backed startups aren't designed to get profitable quickly, and I don't see that as a problem for the American taxpayer, and nobody is saying the taxpayer is giving money or helping. A business losing money should not have to pay taxes on income, as if it's not losing money.
the idea is that normal business expense are deductible.
in this case, your taxable income is $1.0M, and cash flow is $500k ($1.5M - $500k salaries - $500k taxes).
now you have to amoritize it over 5 years. so your taxable income is $1.4M ($1.5-500k*20%), taxes are 700k, and cash flow is $300k.
Uncle Sam just reduced your cash flows by 40% by a simple tax change. You eventually make up the difference, but for fast growing tech companies, that's a large shift in current flows and significantly changes their investment strategy.
There are some misunderstandings in the comments that seem to stem from not having read the section, so I thought it was worth referencing the actual text [0]. It's quite short and easy to read.
The most important bits:
* Subsection (a) requires amortizing "Specified research or experimental expenditures" over 5 years (paragraph (2)) instead of deducting them (paragraph (1))
* Paragraph (c)(3) is a Special Rule that requires that all software development expenses be counted as a "research or experimental expenditure".
That's it. All software expenses must be treated as research and experimental expenses, and no research and experimental expense can be deducted instead of amortized. Ergo, all software expenses must be amortized over 5 years.
I strongly recommend reading the section before forming an opinion. It really is quite unambiguous and is unambiguously bad for anyone who builds software and especially for companies that aren't yet thoroughly established in their space (i.e. startups).
Also note that this makes Software a special case of R&D. It's the only form of R&D that Section 174 requires you to categorize as such and therefore amortize.
[0] https://www.law.cornell.edu/uscode/text/26/174
It's pretty bad.
It had a huge impact on my personally, I'm a small R&D shop and basically I have had to end all risky long-term research projects.
In addition to the research costs, I'd also have to pay taxes on the research costs mostly up-front. Significantly, if the project doesn't work out, I'm still out of pocket for the tax money. It's a penalty for taking a risk, and it kneecaps American innovators in a globally competitive technology race.
The rules are even worse than the article notes because it double-dings open source developers. See Section 6.4 of https://www.irs.gov/pub/irs-drop/n-23-63.pdf. The relevant bit is here:
> "However, even if the research provider does not bear financial risk under the terms of the contract with the research recipient, if the research provider has a right to use any resulting SRE product ... costs paid or incurred by the research provider that are incident to the SRE activities performed by the research provider under the contract are SRE expenditures of the research provider for which no deduction is allowed ..."
The rule as written means contractors who write Windows drivers could deduct their expenses (as they would have no residual rights to a closed-source work product), but contractors who write Linux drivers may not (as they would have some rights to open-source Linux drivers).
You’re likely overreacting.
It makes software temporarily 16.7% more expensive in year one if you’re operating a profitable company, but you do eventually get to deduct that over time. Pay 8% on a 4 year loan and that drops to ~4%.
Eventually, as long as you survive that long.
As has been said repeatedly in this thread, this change is purely a boon for existing big tech companies that now have even less to worry about from startups. It takes a startup 5 years before they'll be playing on an even field with big tech.
> if you’re operating a profitable company
You keep saying this across this thread, and keep ignoring that Section 174 has now redefined "profitable" for tax purposes to include companies who:
* Are in year 1 with no history of expenses to draw on.
* Have spent <900% of their year 1 revenue on software development expenses.
i.e., a startup that earns $1mil and spent $8mil in software dev expenses is only able to deduct 10% * 8mil = $800k of expenses, which means that as far as the government is concerned they made a profit of $200k and owe taxes on that on top of their already-net-loss of $7mil.
You can keep ignoring this fact, but ignoring it doesn't help your case. If you want to argue that this is fine and dandy you need to explain why the above math doesn't prevent new companies from competing on fair terms with big tech.
> The rule as written means contractors who write Windows drivers could deduct their expenses (as they would have no residual rights to a closed-source work product), but contractors who write Linux drivers may not (as they would have some rights to open-source Linux drivers).
Is it just me or are you conflating two orthogonal things?
An open-source Windows driver would have the same issue, no? And a closed-source proprietary Linux driver privately written for some company wouldn't have this issue either, right?
I could see it being inferred that way but, the way I read it, they are not meant as unilateral facts. Rather, they serve as rhetorical examples of where you might find contractors doing similar work, but where the one more in service of "public good" is taxed higher because it's open source. Strictly speaking, Windows bits are not all closed source and there exist closed source Linux bits. But it's not a point that really matters in the context of the conversation.
I think it's fair to use Windows and Linux as stand-ins for closed vs open source because it's a very accessible example. And knowing the technicalities clearly doesn't undermine the argument.
> I think it's fair to use Windows and Linux as stand-ins for closed vs open source because it's a very accessible example
We're talking about businesses here that would struggle with these tax rules. Which I guess is, mainly, contractors or startups. How common is it for them to write open-source drivers vs. closed-source ones? I would've imagined the majority of drivers in such cases are closed-source, on every platform. But I would find it interesting to hear if things are somehow different on Linux.
You are correct. I picked this example under the general assumption that the Windows driver would be closed-source, but you are correct that it doesn't necessarily have to be closed source.
The problem goes with the license, not with the OS.
You're right, the law text doesn't specifically call out the Windows operating system or the Linux operating system. The example you gave of Open Source Windows drivers is valid.
The Grandparent's point about that "it double-dings open source developers" is still correct and poignant even with this clarification.
Just pledge copyright of contributions to eff or something.
Work-for-hire open source contributions often already bear a copyright holder of the entity paying for the work. The problem isn't who is the copyright holder.
The problem is that the license assigned says that anyone is free to use the code. Anyone is a set of people that includes the contributor, which then triggers the interpretation that the research is incrementally in the contributor's benefit and thus disqualified from preferential tax treatment.
You'd need a custom license where everyone in the world could use the results except for the contributor, and then like, a source control system that hides the source files from the contributor's view of the repository.
> a source control system that hides the source files from the contributor's view of the repository.
How would that work?
> You'd need a custom license where everyone in the world could use the results except for the contributor
That one is incompatible with copyright laws in many countries outside USA.
The point is that it's a ridiculous and impractical workaround that makes no sense
^^ this
> That one is incompatible with copyright laws in many countries outside USA.
How so? You can't sign away your interest in a copywrighted work?
The USA hasn’t managed to completely impose their idea of intellectual property on everyone yet. Some countries you can’t sign away authorship even if you can commercial rights.
Parent objected to:
> You'd need a custom license where everyone in the world could use the results except for the contributor
> That one is incompatible with copyright laws in many countries outside USA.
Does authorship confer usage rights?
Create a shell company with an X$ investment.
Have the shell pay write code. (Or more risky pay your company to write code as a work for higher.)
Devolve the shell contributing its assets to OSS.
Take a X$ loss in that year.
Argue that it is perfectly legal to the IRS.
We are small and so have been on a hiring freeze since 2022. I’d like to hire but the upfront cost is high.
For those around when this went into effect many business owners were surprised. Our accountants told us they seriously thought congress would fix this before it went into effect.
[flagged]
Often the way this works is that some time bomb is added to the tax code so that forecasts for future tax revenue will be higher (justifying more spending in the short term) and congress then needs to remember to remove the time bomb before it blows up. So it isn’t that uncommon for these things to be reverted.
Everyone knows it’s a time bomb but the only way we can get an omnibus funding bill through Congress is reconciliation. That’s how our government has been functioning since like 1980 and almost nothing has been “reverted” since then except for the time bombs required to get through the reconciliation process.
Or in this case, not justifying extra spending, but justifying extra tax cuts for incomes above a half million dollars per year, inheritances of more than 5 million dollars, corporate profits, private jet flights, and a variety of other special tax cuts, mostly aimed at wealthy GOP donors.
It's actually not uncommon for something stupid to make it to the official tax code and then get reverted. They knew the history.
Get a better government.
Get better citizens.
Get better education for kids that grow to become citizens
This. We are seeing the ROI for "education" in America.
There's another word for "educating" people until they reach the decision you want them to.
Brainwashing.
You know it's completely possible that people have a different outlook or opinion or perspective on things and that is why they disagree with you, not necessarily a lack of education?
Some people think these are good ideas and they vote or welcome them. Some people think they are bad ideas and they vote or oppose them.
No, education is not brainwashing. These terms have different definitions. One is by definition good, one bad. Words matter, definitions of words matter.
I’d love an education system that only teaches scientific consensus, and leaves moral conclusions to parents and households. However I’m sure you’d appreciate that’s not what we have.
Fortunately our system is setup such that passionate folks like you can work to effect change. Go do it - volunteer for your local PTA, run for school board positions, show up to public hearings. Be the change you want to see in the world. God Speed my friend.
Having the ideology of the majority taught in schools is the outcome of a strictly democratic process like the one you’re describing. I’m suggesting that the separation between church and state be extended to any ideological teaching.
I get it, I'm not trying to chat about that.
I am saying if you can convince enough other people that this makes sense you can effect change. And im wishing you good luck with that.
> No, education is not brainwashing.
I’m directly responding to your point. Ideological education is brainwashing. In fact “brainwashing” is usually just another way of saying “an ideological education I disagree with”, also known as “indoctrination”.
what moral conclusions?
For example: What’s the right way to live? How to relate to others? On what basis do we cooperate? What are/are not the overall goals of society? What’s the meaning of life? What is it to be a good person?
This is one of the worst things MAGA has done. Tech startups are the source of so much of our wealth, and this makes it very challenging to ever build one.
I can’t believe this still exists, and no one has changed it. We truly are governed by morons
for anyone curious, this wasn’t specifically trump, but it was indeed a republican congress bill. texas republican was the initial sponsor and then republicans lined up to cosponsor.
this was done to fuel their tax cuts to a small group of a certain people.
you can see all of the sponsors here: https://www.congress.gov/bill/115th-congress/house-bill/1/co...
It was indeed passed by Republicans in Congress (with Democrats mostly voting Nay), but it was signed by Obama (12/22/2017). Why did he sign it?
I encourage you to look closer at the "in office" section here.
https://en.wikipedia.org/wiki/Barack_Obama
Thanks Obama! singing terrible bills even after he left office.
It is being roled back partly https://news.ycombinator.com/item?id=44028106
Until is is rolled back, it isn't rolled back.
too little too late?
as we are seeing now on a number of issues, sure things can be rolled back, but that doesn't mean a return to normal.
Yes, you can kick over a bee hive, then pick it up and set it back upright, but you are not going to put all the bees back in immediately. There are long term consequences.
But we are great now. What else do we need?
The way this is "fixed" right now, every five years we need another round of republican government to make things great again. If only enough democrats cared to fix this.
A dead post below says it was biden and somehow Obama, but sibling reply link says it passed into law in 2017, not 2022, the first year it went into effect I think.
The article is pretty clear that the law was from 2017 had a scheduled start date of 2022. Although probably not the actual intended effect, it does have the effect of confusing who would be responsible for it by those who don't read past the effective date.
> Although probably not the actual intended effect
This is actually a really common intention in laws like this. Get the tax cuts during your term, and then kick the can down the road so your successor's term is marred by the bad law. If your successor wants to fix it, they need to pass a different tax to recoup the costs, and incur the publicity of "raising taxes".
I think it’s partly that and partly the fact that tax bills tend to be scored on a ten year time window. Note that this law doesn’t actually change the amount of tax that software companies can deduct, it just requires them to spread the deduction over several years. So if you’re scoring your new tax bill on a ten year window and five years into the bill this thing kicks in, then it looks like more tax is being collected in years 5-10. But that’s just an illusion because all the deductions are still there, they’re just being pushed out beyond the end of the window where they don’t “count”. At least this is my understanding.
This is 100% the intended effect of leaving a bomb like this unaddressed then stalling all legislation halfway through your opponent’s term. You make them look incompetent and then you complain about how they never fix anything, even though they never get anything done because you’ve stopped cooperating. And even though you’ve stopped you can keep complaining about how they always include things you don’t like in legislation so that’s why you never cooperate. The solution is simple in your eyes: just do exactly what you want and only what you want and then you’ll cooperate. This is how a lot of a certain party was talking on the news from 2021 to 2025 when they were interviewed.
[flagged]
Exactly!
The change is very simple. And the predictable impact of the change is very clear.
It shouldn’t impact large companies that are already profitable. But it’s devastating for software companies that are not profitable yet.
And that’s without even getting into the philosophical issues with it.
What are the implications of this.
As I understand accounting, this means that reported profits would be higher, and therefore incur more corporate income tax liability. Cash flow isn't effected besides tax.
A startup isn't likely to be making a profit yet, under either accounting rule. Is there a benefit to reporting a larger loss?
My first thought is that this effects Google and suchlike, not startups. But... assuming steady state "r&d" expenditure... it's not that much. Everything gets deducted within 5 years anyway.
So... maybe this hinders more modestly profitable, and fast growing companies most. Those that can't afford to carry 5 years worth of paper profits as easily.
Otoh... I am curious about how the difference between r&d expenses and operational ones are determined irl.
This should be quantifiable. How much extra assets are software companies actually booking?
It seems questionable that this "silent killer" had actually affected employment so much.
> A startup isn't likely to be making a profit yet, under either accounting rule. Is there a benefit to reporting a larger loss?
As an example, A two person software startup; both drawing a salary, each making $100,000 per year. Each doing things related to software development.
Startup brings in 200,000K in revenue.
Under pre Section 174 changes, the profit is zero. Both salaries are expensible in the year they were incurred.
Post Section 174, the profit is now $160,000 each year. Now they pay taxes on $160,000, even though they literally have no money left over because revenues equaled expenditures.
At 25% tax rate, that’s $40,000 in taxes, for a business that made literally no money.
That’s why this is so devastating to small software businesses; unless you’re highly profitable and have cash reserves, this change hits hard.
This example only really has the emotional impact it does because of the specific numbers used, but doesn't really generalize for an arbitrary N.
Clearly if two software engineers build a product that brings in $10M, and each pay themselves $5M, it doesn't seem so outrageous that the can't really claim they're running "a business that made literally no money." Clearly in this second example the problem is that the engineers are paying themselves way too high given the return on their efforts.
What this means is that software engineers will be required to bring in more value to justify their high pay. In your example, it simply means that a software engineer that brings in $100,000 of value to the company, probably shouldn't be paid $100,000.
This seems entirely reasonable to me, and doubly so when I consider how many large corporate teams (who I think will ultimately be impacted more than startups) has huge numbers of highly paid engineers not doing all that much.
In most startups I've worked in it was pretty common for engineers to be delivering multiples of their cost in value, and in every big company I've worked in, it was very common to be delivering fractions of one's cost in value.
> Post Section 174, the profit is now $160,000 each year. Now they pay taxes on $160,000, even though they literally have no money left over because revenues equaled expenditures.
They have the $200k they pulled from their startup, far more than what most people earn. If you make enough to pay yourself $100k then you make enough to pay taxes.
They do pay taxes. They each pay personal income tax on their $100k.
Still plenty left to pay their business taxes.
And if it’s employees? Do you ask them to contribute to the company’s taxes as well? After they’ve paid their own?
Yeah, this provision is a complete fuck up.
That was my first thought as well, but on second thought I can see how this might cause problems:
For established profitable software companies there was a cliff edge in 2022 when this change kicked in. Staff costs for previous years had already been fully expensed while only 20% of the current year's costs could be deducted.
Second, any sudden increase in research expenditures is now discouraged. This could make companies less nimble.
For unprofitable startups it could cause issues during a phase of very high revenue growth. They could suddenly be liable to pay corporation tax in spite of the fact that they are not profitable in any reasonable sense of the word. It would smooth out later, but that may be too late for some.
What I do not believe for a second is that this is causing major job losses. Companies like Microsoft or Meta do not reduce research or software development just because there is a temporary tax hit. It could be an extra incentive for an efficiency drive I guess.
> For unprofitable startups it could cause issues during a phase of very high revenue growth.
So I guess my most question is "how this work irl?"
Say a new startup raises money and hires 20 people. Pays $5m in salaries, office space and such. All 20 people are developing a software product. Are 100% of this startups expenses amorotized?
Then they sell the product. They receive $2m in revenue. What does the P&L look like.
If they hire 20 devs in their first year paying $5m in salaries, only $1m or $500k (if the mid-year convention applies) would count as a business expense in that first year.
If their revenue was $2m, that would leave them with $1m (or $1.5m) of taxable profits unless that was eaten by other costs.
It doesn't have to be a problem, but if revenue grows fast and they go on another hiring spree in the following year then it could become a problem.
That said, if revenue grows so fast, it seems likely that they would have huge marketing and sales costs that could be expensed immediately. So maybe this isn't really a problem for many startups. I'm not sure.
1 million profit, while they have 3 million negative cashflow, that's exactly the problem. They can only take 20% of that 5 million in R&D investment as depreciation in the first year.
Your analysis is correct, but most software companies were mostly profitable or fast-growing. For every Google, there’s 1000 wordpress vendors you’ve never heard of.
In another year the initial shock will stabilize, but any growth now has a 5-year tax hit attached. And even Facebook doesn’t want to pay that if it doesn’t have to.
Google was reportedly amortizing (by choice) long before this was in effect, so while it might “affect them”, in practice it’s likely business as usual.
It depends on the department. My salary (in a mature product) was already amortized - I suspect the same is true of all their other mature products like Search, Maps, GMail, Chrome, YouTube, etc. But I think they were deducting salaries in the more research-like areas like Gemini, Jax, Assistant, etc. So there is net still a fairly large charge related to it, even if it isn't as large as it could be.
The R&D credits are deducted from Payroll taxes, so they impact pre-revenue startups as well.
I'm not an accountant, but as I understand it, you don't pay taxes on profits, but on revenue.
So previously, some 20% of all revenue would be owned as corporate income tax, and startups would deduct it all as they're spending much more on R&D than they owe in corporate income tax. But with this tax change, the deduction would be much lower (80% lower IIUC).
No, you pay taxes on profits. What this does is reduce your upfront deduction.
Yes, but the main thing here is that ALL software development is now "profit" in the short term. In theory you've developed a capital good that benefits you over time, hence the amortization.
Simplified 2021 example before 174:
Simplified 2022 example after 174: Above example is year one of suddenly having these taxes, because if your software costs are the same or lower over time it gets easier. It's just extremely painful for smaller and especially fast growing companies like startups without a lot of cash, especially when interest rates are so high.Accountants: If I am wrong about the above, please correct me
If companies paid tax on revenue the US budget would be perfectly fine.
Large companies always find a way to not pay taxes. It's the little guys that end up paying (a lot!) more, to the extend that it cripples and kills them. But transformative innovation happens with the little guys. As a result, this tax change cements monopolies for megacorps. They will be fine and still pay nothing.
The little guy always pays all taxes. Corporate tax is just a way to palatably shift tax burden to the low and middle classes and away from the owner class. It is pure double speak.
> you don't pay taxes on profits, but on revenue.
That can't be right. It definitely isn't in my country.
If own a car dealership, and I sell a car for $50,000 that I bought from the manufacturer for $40,000, surely I would pay tax on the $10,000 profit? The tax on the the full $50,000 revenue might exceed my profit!
Welcome to the Democrat version of taxes. In Michigan, restaurant owners had to pay a tax on revenue and not profit around 2008 or so.
lots of retaurants went out of business overnight.
Sales tax (which most US states collect) specifically is a tax on revenue, but it is the exception.
And if the R&D uses foreign workers, because you can't afford to pay US wages, then the 5 years goes to 15 years!
This hurts small companies (like mine) that were priced out of the US developer market.
We don't talk about this enough. International R&D is not offshoring of call-centers to India. International R&D is the IP for the next generation of global communication standards being owned by US-based or foreign corporations, because international (e.g. Canadian, European) standards experts/developers become un-affordable for US-based corporations and are forced to work for our "adversaries" instead. Crazy.
I’m not sure this is exactly true. If your foreign workers are a service contract then those are services expenses immediately deductible. Same if you are using local service contracts. My understanding is this creates a drag for companies that want to hire f/t.
Foreign workers are to my knowledge effectively always a service contract, since it's pretty complicated (if even possible) to hire FTEs across borders without subsidiaries, which are expensive to maintain.
I'm curious if contract work is really exempt, would look like a major loophole to me.
> Foreign workers are to my knowledge effectively always a service contract, since it's pretty complicated (if even possible) to hire FTEs across borders without subsidiaries, which are expensive to maintain.
It's impossible (yes, I'm being absolute) to hire an employee who lives in or outside the US who is not a citizen or doesn't have a green card. All employees must have an SSN and go through i9 verification, which requires in person verification of legal ability to work in the US.
The foreign developers I'm talking about are not US citizens and do not have green cards.
Their work is subject to 15 year amortization per section 174. Period.
Not if they are contractors. That's the point the parent commenter was making. All the reasons you list make it so they need to do so, instead of "hiring" them directly.
Tax law is full of major loopholes. It’s highly political law, so doing one thing while saying you are doing another is a feature, not a bug.
There are definitely gray areas to the law, but in my decades of experience dealing with lawyers, they won't steer you to do something very far over the line. I think the companies that do step far over the line, to game the system, are doing so knowing full well they are breaking the law, but they believe they are unlikely to be caught. And they're very likely right. You could separate CEO's/CFO's in to two camps: stay legal and do what you need to do to make the almighty $$. In the 80's the phrase "greed is good" was born, but the last 2 decades have really upped the ante on this.
This is simply not true. Says my lawyer and CPA. And every other CEO/CFO I've talked with.
It's not really targeted at tech, insomuch as at Democrats.
Everyone assumed it was a traditional accounting hack. But given the timing and the reinitialization, it's clearly political, not economic.
The code is a strategic time-bomb designed to cause a high-profile economic downturn during a presidential election cycle, specifically when the following president is a Democrat and Republicans have a house majority.
It was used to harm Biden's economy, and it will happen again in 2030 if the next president is a Democrat. While deferred, it will be spun as a major Trump "economic achievement" for the midterms, because companies will be able to afford to hire again.
The tech industry is merely high-profile fodder for extreme politics. It really is that petty.
The Democrats had control of the presidency and the house in 2022 when this provision first went into effect but had 2 fewer senators (1 fewer if you count the tie-breaking VP). Why didn't they try to change it? Is there some reason a change in the tax code like this can't be modified or repealed once its in place?
Politics are complicated.
Generally, in tax bills they try to keep them "neutral" where any tax cuts or tax breaks are coupled with tax increases elsewhere BUT they tend to report the 10-year affect for whatever reason. This bill provided a ~30% cut in corporate tax on profits, with a delayed increase in tax cost on Software R&D pushed to the next term.
If the next party wants to reverse it, they'd have to find the money with an increase in tax - directly undoing it would be a ~50% increase in corporate tax rate, which (I guess?) would be a tough sell politically. Meanwhile, the tax code on software engineering sounds too niche to expend political capital on.
Either way, its another example of how corporate America is trading long-term growth (R&D, product development) for short term gain (lower taxes today).
They tried. They had Senate spoilers.
As a progressive, it seems like the Democrats always have Senate spoilers...
> As a progressive, it seems like the Democrats always have Senate spoilers...
With Republicans usually being dominant in a number of states, if Democrats have a Senate majority, it is usually both narrow and dependent on a very small number of Democratic and/or Dem-leading moderate independent Senators from Republican-majority states who vote with the party on leadership, but are soft (or firmly opposed to the progressive preference) on a number of issues important to progressives.
If the US were approximately an equal democracy, this might be less of an issue.
>If the US were approximately an equal democracy, this might be less of an issue.
How? Evenly divided voters and representatives are the issue. Each side can barely afford to lose 10% or so during votes
No, the reason the "there is always an in-party Senate spoiler" effect (when they have a Senate majority) seems to be more true of Democrats is because it is more true of Democrats, and the reason is that when the two parties in rough balance by popular support (or even rough balance in Presidential electoral prospects, which has the same directional bias as the Senate but of lesser magnitude), the Republican Party has a systematic edge in dominance of states, which translates into a systematic advantage in the Senate, which means that when the Democrats have a Senate majority, it tends to have a decisive segment in red-state Democratic Senators who are unreliable on key priorities.
The issue being discussed in the Senate is not a symmetric issue resulting from near balance in support between the parties.
It’s also because republicans politically punish dissent, while it is more tolerated in the Democratic Party. The consequences of “disloyalty” are higher in the Republican Party.
This might change. After party leadership got 20% of democratic senators to vote for trump’s procedural blank check, the party’s approval rating dropped to 27%.
If it doesn’t change, I suspect the party will split.
> If the US were approximately an equal democracy, this might be less of an issue
Equal to what?
Equal in voting rights. Gerrymandering has been perfected by Republicans. Through that they manage to dilute votes of the opposition. Other measures discourage voters likely to vote against them, like people who cannot easily take time off to vote in person or who have changed their name. Blocking rank choice and maintaining first past the post also disenfranchise third parties, and reinforces the power of incumbents.
Trump himself admitted it's better for Republicans when fewer people vote.
> Equal in voting rights. Gerrymandering has been perfected by Republicans. Through that they manage to dilute votes of the opposition.
This thread is talking about the Senate. The senate isn't gerrymandered. Both senators are state-wide races.
If you want to view it that way, you can view the senate as "pre-gerrymandered". But the last time that was an option was in 1959, and both of those are just "the entire area the US owned, but wasn't a state yet. To get senate gerrymandering, you have to go back to 1912 and the admission of New Mexico/Arizona.
> If you want to view it that way, you can view the senate as "pre-gerrymandered".
That is quite explicitly the history of the US Senate (and House), FWIW.
The Connecticut Compromise was reached to give low-populations states outsized legislative power in the senate. This is the main reason the senate exists.
Building on that, the 3/5th compromise was reached as part of this to give slave states outsized legislative power in the house.
The state of Maine used to be part of Massachusetts, but it was later set up as an independent state in order to increase the number of anti-slavery states in the senate (the Missouri compromise).
Gerrymandering can affect voter sentiment and trigger polling location changes during redistricting, both of which can affect voter turnout[1][2][3] (though the research doesn't seem conclusive on the effect).
And thinking about it more, though I haven't seen if there are studies on it: there are probably manpower/fundraising effects from gerrymandering.
If you're able to protect your political power in one area that probably better enables you to amass resources to use in the area you can't gerrymander.
But all that said, both parties practice gerrymandering and I don't think there's strong evidence of a significant advantage over a major party from current gerrymandering at the national level.
[1] https://da.lib.kobe-u.ac.jp/da/kernel/90008864/90008864.pdf
[2] https://electionlab.mit.edu/articles/gerrymandering-turnout-...
[3] https://stateline.org/2022/05/20/check-your-polling-place-re...
> On a percentage basis, over three times as many districts were competitive in states where independent commissions drew maps as in states where Republicans drew maps.
https://www.brennancenter.org/our-work/analysis-opinion/how-...
That’s just confusing cause and effect. If your seats are safe, you have no reason to agree to forming an independent commission. The same is true in both heavily blue and heavily red states. Are districts more competitive in states where Democrats draw maps? I don’t think so.
This totally ignores values and motivations, and I would argue that only one group in your comment values winning at any cost.
I don’t even know which group you mean, but “my group has good values and motivations, but the enemy group just values winning at any cost” is exactly what a total partisan who values winning at any cost would say.
The evidence is that independent commissions drawing maps makes for more competitive districts. Which party is most opposed to such commissions? Which party is gleefully dismantling all accountability and oversight positions and departments? Which party is openly inviting corruption and pardoning those they should be prosecuting?
Governors are elected by popular vote.
?? Both sides happily gerrymander. It’s been around since 1812 and both sides are equally guilty at this point.
I didn't say democrats were innocent. I said Republicans perfected the (ab)use of districting.
https://www.brennancenter.org/our-work/analysis-opinion/how-...
Hell, just first past the post would eviscerate the current parties.
Argh. Too late to edit. Something else outside first past the post* like ranked choice voting.
Which is why they’ll never vote for it. Such changes are remarkable rare. :(
And get blamed for it. If every single Republican and two Democrats vote against something guess who people blame?
But this is the type of thing that progressives would like support (tax big corporate America).
No, this is a misunderstanding of the kind of taxation policy progressives tend to favor. Taxation on profit for businesses should be high, and taxation on upper tiers of individual income should be high, but taxation on funds businesses use to reinvest should be exempted or deductable. Basically the taxation we had in place after WW2 and on, with a steep corporate tax rate and more or less a maximum income for individuals. The R&D exemption removed in the 2017 bill, and discussed in the article, is key to that, because it encourages corporations to reinvest their income in building new products and paying workers rather than taking it directly as profit-- after all, at least they could reap the rewards (in growth and revenue) of the R&D later, instead of just giving the money to the government as taxes.
I don’t think most progressives think about it in that detail. Raise taxes on the rich tech companies that are gentrifying san francisco.
At first glance I support ... "social and economic equality" and "reforms to improve human conditions, combat corruption, and reduce inequality". Am I progressive?
If you ask me "should corporations pay more taxes?" I will say, yes. Famously so does Warren Buffet, is he also a progressive?
If you ask me, "hey should we gut tax incentives for R&D spending in the USA?" I will say, uhhh no? probably a bad choice?
Recently the progressives have latched on to culture war agendas against the wealthy, educated, white, male, straight and/or over the age of 35 crowd.
In other words, they have a popular agenda, but are political morons that are going to eventually wonder why they can’t break out of solidly blue districts.
https://runforsomething.net/run/candidate-support-system/
I think that is a misrepresentation of the fundamental progressive position, which is to make progress but never at the cost of the marginalized. Because we historically make most progress at the cost of the marginalized it can feel limiting or even discriminatory when we make sure they don’t beat the brunt of continued progress.
There is nothing against the group you mention except that it might be the group that most fights against progress toward equality.
I‘m not American but the above description of a tax policy is what I hear a lot from progressives in media.
This tax is far more consequential for small companies than for large ones. It probably actually benefits larger companies because it hobbles competition.
This time bomb was created because the bill slashed the corporate tax rate from 35% to 21%. Maintaining the status quo would mean taxing big corporate America more than this bill does.
What makes you think this?
But it isn't tax big corporate America. Did you read the article?
It's a 10% tax cut for big corporate America, with some economic poison for blue states in the future.
Providing spoilers was the explicitly designed purpose of the US Senate. It's not a one-sided problem - Senate spoilers are also why the Affordable Care Act didn't get repealed in 2017.
Explicitly?
US Senator was an office initially designed to be selected by state legislatures rather than by direct popular election like the representatives. To a populist or a party boss, that might count as a spoiler to the will of the people or to the will of those in DC, or to both. But I may misinterpret GP's point.
I assume the person you're replying to is talking about the Filibuster and supermajority requirements not the direct election history. The filibuster is a senate rule not a constitutional design, so it wasn't part of the "design". Maybe they're both different ways of adding veto points to the same effect, but I think spoilers as "explicit design" is probably not how I'd describe it.
https://en.wikipedia.org/wiki/Filibuster_in_the_United_State...
Not parent but the founders were like folks writing smart contract code, thinking about various exploits and vulnerabilities (that might reduce the wealth of their class) so many of the seemingly dysfunctional elements of the system turn out to be designed deliberately to be dysfunctional. Feature not bug.
They were not thinking about various exploits and vulnerabilities but rather making whatever compromises were necessary in order to form the union. It was negotiation, not planning.
A compromise can also be a feature to resolve a bug, from the point of view of the one demanding it.
Both parties tend to when there is a narrow majority, e.g. McCain thumbs downing at the repeal of the ACA.
Why should they? Why did we allow a president to put in tax raise for the future. Replicants were playing politics from the start. Pass a bad bill, and then hope to get about it when the bad parts kick in when the other side woo be in power
If this was passed in 2017 to go into effect during the next presidential term, wouldn't that only work as a time bomb for Biden's presidency if Trump didn't expect to win a second consecutive term?
Given the history of prior presidents winning 2 consecutive terms, it seems like Trump could have reasonably expected a 2022/2023 tax change to be his problem.
if you retain power, you can fix it. the US government currently has the significant problem that one party campaigns on the government being dysfunctional, so they do their best to make it so.
So.. criminal racketeering?
But what would trump have done if he retained the presidency and lost congress? That's also been pretty common over the last few decades if I'm mistaken, a president with one or both sides of Congress is reelected but Congress flips to the opposition party.
He would do nothing because his supporters believe misinformation and worship him.
Prices haven't gone down at all nor will bringing manufacturing to the US do this (likely causing them to go up) but his approval rating is 50%
> He would do nothing because his supporters believe misinformation and worship him.
Interesting, that hasn't been my experience.
I live in a very red part of the country and most people I know are Trump supporters, including some family members have been very MAGA since 2016.
I've been hearing more and more complaints over missed promises: no Epstein files, raising budgets, RFK is starting to water down his promises, no end to the Ukraine or Gaza wars, etc.
He missed effectively every promise from 2016. Why did these people vote for him 2 more times, especially after an attempted coup? Maybe these "complaints" are just an attempt to dodge personal responsibility for having supported a catastrophe.
> But what would trump have done if he retained the presidency and lost congress?
Trump is blaming Biden for the obvious outcome of Trump's tarrif nonsense. What do you think Trump would have done?
I would suppose that the Democrats would remove the policy regardless of who was in charge.
But they didn’t, so that supposition is bunk
When did they have control of the senate without a BD being the lynchpin?
How?
Vote with Republicans in removing it
[flagged]
That’s a cute quip but remind me which party controls the governments in Baltimore, California, Detroit, etc?
Are there any parties running in a track record of functional government?
This civic control correlation can simply have more to do with the most-white-supremacist Democrats switching to the GOP en masse and also simultaneously leaving multiethnic cities and school districts en masse after the 1960s. That self-selection left Republicans not a competitive amount of credibility or voter pool behind to work with. Your implication that policy dysfunction has ensued on that account rather than because of fiscal drain -- that's a separate topic. Individual states and individual cities have too many fiscal policy similarities and differences, overlapping, to responsibly compare in any online discussion.
> That self-selection left Republicans not a competitive amount of credibility or voter pool behind to work with.
So by your logic New York is a better governed state than Florida? Net internal migration would seem to disagree.
> New York is a better governed state than Florida
Yes, New York is significantly more successful than Florida in almost every way: Better education, better healthcare, longer life expectancy, less pollution, lower crime, more productivity, higher wages, more amenities, better transportation infrastructure, less poverty, happier residents, and so on.
Is it your opinion that the only factor relevant for those deciding what state to move to is quality of government?
I'm surprised that things like the job market wouldn't come into play, for example.
I think quality of governance is a major reason, yes. When my parents immigrated to this country, they moved to a deep red state (Virginia) instead of the deep blue state next door (Maryland). Why? A focus on good schools, low crime, and low taxes, instead of a focus on economic redistribution.
The two are related: bad governmental policy can make employers leave a state and make employers that choose to stay less prosperous.
... because nobody moves to Florida for (what they perceive of) the weather, right? Especially not retirees tired of the idea of one more winter in NY.
What makes California's government dysfunctional?
What about Detroit?
A government that runs the richest city in the country (SF trades this spot with NY every few years) and makes it look the way it does is the definition of dysfunction.
And Detroit... well, I guess now that they've bulldozed all the abandoned buildings it looks less like a post apocalyptic hellscape and more just abandoned. An improvement I suppose.
California also has easily solvable housing, education, transportation and mental health crises that are entirely driven by mismanagement by the state government. They haven’t done anything meaningful to address these issues in 25-40 years depending on the issue.
Heck, they ignored the water crisis for twenty years, and what they’re doing now for aquifer replenishment is still less than what makes sense.
I say they are easily addressed because simply reverting to California’s policies from ~ 1975 would greatly improve the current situation.
The biggest root problem with California governmental structure is harmful constitutional features added by public referendum, especially Proposition 13 (1978). I guess you can blame "government" for that, but it doesn't seem like quite the right target.
The water crisis is a difficult problem because water rights are complicated and central valley farmers are an influential political group very focused on short-term preservation of water access and not as concerned with long-term sustainability.
> easily solvable housing, education, transportation and mental health crises
I submit that these are much less "easily solvable" than you claim. (What have you personally done to work on these problems, if they are so "easy"?) Legislators don't get to wave a magic wand, but need support of a wide variety of stakeholders who have contradictory demands and expectations (some of which are fairly unrealistic, but anyway..).
Education for example has competing goals of local funding vs. inter-city equity. Should the wealthiest towns get to spend arbitrarily much local property tax money on their own children's public schools while the poorer town next door is running out of toilet paper, or should the state try to equalize funding between schools to give every child the best opportunity? There's not really a "correct" answer to this, and every possible choice has some serious disadvantages.
You are all over this thread treating HN like reddit or twitter. Please go.
Yes sir! Mam? Xer?
But more seiously: you are saying that there's this poor little party that can't manage to vote for any laws because big bad Republicans keep coming in and aww shucks. Do we really want such incompetence in government?
No. It’s after re-election. Bad news late in your second term isn’t that big of a deal, unless you care about legacy.
Most presidents care about legacy, at least to me it seems like trump holds that as a higher priority than most.
Trump seems to care what people think today, every day. He doesn't seem like someone who puts a lot of thought into the future.
You know, I wonder if that's a sad but valuable trait for a politician.
Public opinion can change daily, and external events can appear with no warning. These things can make a prior path of action vanish, or even make it madness to pursue.
If you try to plan everything long term, I bet you hit a lot of disappoint as a politician. If you only see today, then you're not fighting for things that are now not possible.
I imagine one would be far less stressed as a result. And maybe more popular than otherwise.
Some problems require years or even decades to address. Consider how quickly a COVID vaccine was developed, yet it depended upon many years of quietly studying SARS and R&D around MRNA. Or consider trying to address developing or maintaining infrastructure.
A chaotic politican whose mind is changed by the last person they spoke with won't do well facing serious long term problems.
It gets worse if the only things they consistently stand for is their own power, personal wealth, their sycophants, and their grade-school-level (mis)understanding of complex matters.
>me it seems like trump holds that as a higher priority than most
Why?
He seems to care immensely about being viewed as the "winner" and the "best" at everything.
I also have to assume that anyone interested in slapping their name in big gold letters on as many buildings as possible is interested in the perception of legacy.
This is just wrong. It was passed in 2017 (during Trump’s presidency). It was to go into effect in 2020 (a presidential election year during Trump’s presidency). He hoped to be re-elected.
No. It went into effect in 2022 [0], which means the timeline absolutely does track with OP's theory. That gives a hypothetical Trump term 2 a full year to fix it but also imposes enough of a time crunch that they could plan on sabotaging attempts to fix it by another party.
I'm not saying it's the actual story, but the timeline does track.
[0] Page 60, Sec 1306(e) sets the date: https://www.congress.gov/115/bills/hr1/BILLS-115hr1enr.pdf
I don’t think it really changes the narrative. 2020 was also a congressional election year, even had Trump won (as he appeared to want desperately) he could not have been assured of a Republican congress.
My argument is simple: Occam’s Razor
The Republicans in congress put the provision in solely as a gimmick to get past the CBO.
Frankly I don’t think legislators in either party are competent enough to have foreseen the consequences and even if they had been they wouldn’t have put a bomb like this in that would be more likely than not to backfire and affect them.
I just think that too often people interpret incompetence as malice, especially nowadays when things are so polarized that it’s fashionable to hate people who differ with one’s political opinions.
I think you’re wrong that legislators are incompetent. They’re human, and they’ve spent much of their lives learning how to get votes, but they’re not incompetent. A lot of them aren’t malicious, but there seems to be a small group of people outside of the legislative branch who are hell bent on taking control at all costs. And if you want to keep the votes rolling in you have to work with those people or get primaried. The dysfunction follows from fear more than incompetence these days.
I have no doubt it's bad, but I can't believe that it's both fueling mass layoffs and also almost nobody has heard of it. Those are unlikely to both be true.
I'm not sure why TFA makes it sound like almost no one has heard of it, but it was extensively discussed on HN in early 2023 as being a primary cause of layoffs, before it was cool to blame AI.
Software firms across US facing tax bills that threaten survival (924 points, 981 comments) April 18, 2023 https://news.ycombinator.com/item?id=35614313
Ask HN: How are you handling Section 174 changes for bootstrapped companies? (298 points, 187 comments) Feb 2, 2023 https://news.ycombinator.com/item?id=34627712
Why the big tech firms that suddenly laid off a bunch of people the instant they started looking at their 2022 tax bill didn't tell everyone explicitly that that's what was happening I can't say, but it's not like this has been happening in secret.
Obviously interest rates also play a role, and probably a larger one. But this is objectively a very very bad contributing factor, far worse than the impact of coding LLMs.
By what logic?
At least one person at every company impacted knows why they're firing people, right? Likely several people who are in a decision making capacity. At every company.
Those companies have R&D for a reason. A company _wants_ to make things, right? If this is impacting their ability to make things, wouldn't it be in a company's best interest to advocate openly against the tax code, rather than be silent about the reason, fire their staff, and just not make things?
It doesn't make sense to me how so many people are aware of this to the point that many many companies are all doing the same thing for the same reason, but seemingly nobody was talking about it before this post. That doesn't make a lot of sense to me.
It's not like this.
It's like this: Company wants to do R&D, they have a budget, they do math that says they can afford to pay X number of R&D workers with Y budget.
Government changes tax laws in unexpected way, that changes the math so that Y budget only can support X-A R&D workers because the "A" goes to taxes now.
Also important to note, the tech R&D space is a very small part of the overall economy. We exist in a little thought bubble here on HN.
Assuming this change is causing the layoffs, then these companies know about this.
If these companies don't know about this change, then why would we believe this change is causing it.
They didn't know the change was coming. Then it happened, now they have to revisit their budgets. Some thought it was going to be changed back before it went into effect, it wasnt.
I thought wages were deductible anyway. Say you pay a developer $250,000 a year. The employee pays the tax on their own wages.
No, not in this case. 250k is an expense for the company. Company had to amortize this expense over 5 or 15 years. (15 for software engineers outside of the US)
> (15 for software engineers outside of the US)
Yikes. Does that apply to outsourcing?
W2 salary is different from fully loaded labor cost.
It might be "quite short" but it's full of click bait style text. This tax law will change everything, but we won't say what it is for 4 or 5 paragraphs, nor what changed for another 3
Edit : sorry I just realised you meant the tax law is short. The article itself is very annoyingly written
Agreed. I think the article's text was responsible for the confusion in the comments prior to me posting this. They could have been much more clear and straightforward.
Who sponsored this text in the bill?
Here's the cosponsors of the bill:
https://www.congress.gov/bill/115th-congress/house-bill/1/co...
I think the purpose of the change was to "increase revenue":
> Requiring that certain research or experimental expenditures be amortized over a five-year period or longer, starting in 2023, would increase revenues by $109 billion over the period from 2023 to 2027.
https://www.congress.gov/congressional-report/115th-congress...
> I think the purpose of the change was to "increase revenue"
Yes, but in a specific way: they were trying to offset the tax cuts they wanted so they could pass it via the reconciliation process and avoid the Senate filibuster. They didn't actually care about this revenue and the assumption from most people was that the specific carve-out would disappear in some future bill.
And now with their attempts to keep the tax cuts around, they've just decided to ignore the rule entirely and pretend that extending a temporary tax cut counts as not costing anything. Of course, there's nothing that would stop them from getting rid of the filibuster entirely either, but that honestly just makes it weirder to pretend that this somehow fulfills the requirements rather than just is taking advantage of the rules being only self-enforced.
They want to call it anything other than a tax.
It's a specific tax, on a particular class of better educated workers in specific jobs.
That typically don’t vote for that party and are unsympathetic.
Unsympathetic because?
Generally R&D software developers aren't considered to be poor or disadvantaged.
What if they lose their job?
Then they are still arrogant brats.
Also, that administration was pissed off at tech.
Idk but it was under trump. And the meta issue was balancing the budget after all his tax cuts so he needed to find more tax revenues. Which this accomplishes pretty handily
I think partially dismissing the question due to the bill happening "under trump" doesn't help the conversation here. If the bill was sponsored by particular reps/senators, then it's worth identifying those, so their voters can factor this bill in to their decision to vote for/against in the future, etc.
This was also a gift to big tech, boxing out competitors.
> All software expenses must be treated as research and experimental expenses
From what I've read, not for software fixes to ongoing products, but for new products and I can't remember for new feature work. Also if you contract for someone else I heard you can still write off expenses without amortization.
I would guess that because in these cases software development costs would be classified as “cost of goods sold” instead of “research and development”
I worked for a UK company that amortised it’s development costs… it led to the false belief that the company was profitable when it really wasn’t
OK, but you've changed the topic from tax accounting to financial accounting/reporting.
In the US, it remains the case that programmers salaries must be treated as an expense (i.e., cannot be amortized) when calculating the company's income statement, balance sheet, etc. Not following that rule will get the accounting firm signing off on those financial reports in trouble (with the SEC, the Public Company Accounting Oversight Board, and maybe even the Justice Department if the purpose of the violation was to defraud investors).
"Profitable" in an accounting sense has nothing to do with your cash position. This is something that people in tech don't really seem to understand.
i think people in tech are usually focused on free cash flow, aka playing around money.
Tech likes free cash flow because share based compensation isn't included in it.
Exactly. And if you’re more profitable on paper, you have to pay more tax, making you even less profitable in reality.
Yes, that is tremendously important aspect here - the US tech would look better on paper - higher paper profits due lower paper expenses - while getting increased cash flow stress due to decreased deductability of the salaries which are among the main expenses in software dev business.
>Yes, that is tremendously important aspect here - the US tech would look better on paper
It's completely unimportant. Nobody is getting fooled "on paper" by amortized salaries.
People is getting fooled by "adjusted" earnings that reduce salaries "on paper" by hiding the "non cash" component.
Except for example the millions of stock market casual participants.
Want a bet?
I’ve seen it used in UK listed companies to massage the profit numbers and make divisions of the company seem more profitable than they are
I don’t get the big hoopla. Here in Germany I’m opting to turn development costs into assets (I simplify a bit). I need assets on the balance sheet, otherwise we’re over-indebted. As long as the development costs are much higher than income (I.e. as long as you’re not profitable), then it shouldn’t matter. And once you are profitable, you pay some more corporate taxes, but aren’t they kind of not too high in the us anyway?
the big hoopla is this: you're a newish startup. You have $300k/yr revenue and $$270k/yr expenses of which $250k is paying your programmers.
prior to this rule change, what you pay your programmers is just a deductible expense, so you owe taxes (in this very simplified example with no other expenses etc. etc.) on just $50k.
after the rule change, you can deduct only $50k of the labor cost (in this year), so now you owe tax on $250k.
there is a very good chance you do not have the cash available to make this payment.
of course, after 5 years, things all balance out and are effectively "back to normal". but you have to get through those 5 years first.
How does this work, in your first year, to have 300K revenue on 270K development cost. It sounds like some very specific boot strapping case, and even then after 70K deductions its 230K profit, so like 50K in taxes on 30K profit. Sure it’s annoying, but it doesnt sound like total doom. Its more likely youll have 250K programmer expense in the first year, 100K in revenue, need 200K in outside investment, and pay very little taxes. After a couple of years it evens out anyway.
Consider some sort of logistics company. They might pay 250K for their hardware (e.g. vehicles), 50K in expenses, get 300K in revenue. Theyll be taxed the same as the startup building up their IT assets.
Excellent summary.
What i like about US is that compare to other countries (like for example Russia where i'm originally from) there is almost no lying and cheating here. Instead there is a respect of the law and an army of talented creative accountants and lawyers. Remember that stale "multi-used" sandwich served with the drink which by virtue of its existence converted drinking establishment into a food serving restaurant? Not being an accountant, i'd just speculate, out of sheer fantasy, that some hardware chip/gadget added to your software may similarly convert your software development into hardware/gadget one.
I'm not sure I buy into this. Sure, compared with Russia it's probably a lot less, at least in terms of being something everyday people engage in. But in terms of comparing with countries like Germany or Sweden I don't know.
Here's some food for thought:
* Global financial crises: Banks were paying (bribing) ratings agencies to rate junk bonds AAA.
* Savings & loan crisis: widespread fraud & insider abuse.
* Bernie Madoff: Ran the largest Ponzi scheme ever, with an estimated fraud total of $65B raking in $17.5B in invested cash.
* Enron: straight up accounting fraud sprinkled with intentionally causing brownouts in California to pad their pockets with a side bonus of making Gov Davis unpopular & get him recalled (Enron was closely aligned with the Bush administration).
* Nixon straight up using psy-ops against Democrats & finally trying to burgal the DNC offices.
In terms of stats, the FBI does a few hundred bribery and corruption cases annually. Are they good at catching white collar crime? Well such crimes regularly take more than 5 years to investigate.
And hell, some things that are basically lying and cheating are straight up legal. Usury is legal with minimal to no regulation of payday loans. Pyramid schemes are legal as long as you call it multi-level marketing.
The list goes on and on.
If the list doesn't go on and on that is a clear sign that corruption is being hid.
Could be a sign that the commenter was not willing to spend all his day writing, especially once his point was made.
I can’t tell if they’re trying to refute the point (ie the list being long means that it’s definitionally not being hid) or support (ie the list I made is finite and therefore the corruption is hid).
But I think of it like trying to estimate the size of an iceberg by observing the tip. Just because you know a little bit doesn’t mean you actually know about the scale. And there’s every reason to believe it’s quite extensive given how easily money flows from corrupt countries through USD and US persons and companies (eg the major bank that’s constantly getting fined for laundering terrorist and drug cartel money - either they’re the only ones and they’re making a killing providing this service anyway or they’re the only ones anyone is bothering to investigate, but that business is clearly lucrative to continue to engage in).
> What i like about US is that compare to other countries (like for example Russia where i'm originally from) there is almost no lying and cheating here. Instead there is a respect of the law and an army of talented creative accountants and lawyers
I thought you were being sarcastic here at first because, good lord, there is plenty of corruption here in the US (though those doing it used to care more about hiding it). The US, especially in its current state, is certainly not a place I'd describe with "almost no lying or cheating". I do understand that Russia is on another level, though, given the open assassinations and doing things like what was done to Navalny.
> I thought you were being sarcastic here at first because,
You've never been in Russia. There is no clear law abiding business there. That opens a lot of opportunities for those with some power. Corruption is one of them, selective punishment is another. I'm sure in most 3d world situation is not better, but they at least don't have laws to cheat and bribing isn't a crime.
You have no idea about corruption if you haven't lived it in a BRIC country.
The funny thing, is that people not from America say that there IS corruption, but at least it happens in the open. I think OP is saying the same.
>say that there IS corruption, but at least it happens in the open
How is that corruption?
Breaking the law is still breaking the law even if you don't hide it. If anything, not getting in trouble for breaking the law noticeably often means that there's also corruption from the ones who should be holding the corrupt accountable.
Breaking the law isn't "corruption"
A good definition from AI "Is when government officials misuse their power for personal gain or to benefit their friends or associates"
Right, which is illegal. Thus breaking the law. And remains both corruption and breaking the law even when it happens in the open.
Where in that definition does it say that it can’t be done in the open?
See for example Trump’s shenanigans, which are done in plain sight for all to see, but with few if any repercussions (a very brief selection: having foreign dignitaries stay at his hotel in DC while he’s in office; having the Secret Service stay at his resorts when he goes golfing; scamming the public with his family’s meme coins; etc)
Because this is SO much better..../s The only difference in style is that the American billionaire will corrupt everything and still say it is for your own good.
You forgot putting a dead guy on trial.
> almost no lying and cheating here
Are you living in an alternate world?
No, he's saying that people respect the law, which they do. It's all about finding loopholes, and sticking to the letter of the law while working around the law to do whatever the law prohibits but doing it in a way that remains legal. This entire situation came up because of loopholes. A great way to offshore money was to spend it on software developed by overseas subsidiaries.
If you’ve never lived outside the US you have zero idea how bad it gets. It literally is an alternate world.
The amount of daily activities in the US that just work 99.999999% of the time that would have a corruption aspect in some other countries is mind boggling.
The closest analogy I can come up with is imagine if every money transaction involved cash tipping the parties involved. And that’s just the beginning.
Like tipping the licensing agent every time you had to renew your driver’s license or get plates. Or tipping the judge for a favorable judgement.
Or are cynical Americans living in an alternate world, blind to how much better the rule of law is here than most other countries? The commenter's comparison was to Russia. When was the last time Putin lost an election?
I'd say we're slightly behind western Europe as far as rule of law goes, not really sure about the advanced east (Japan, Korea), and miles ahead of just about everywhere else (eastern Europe, Russia, Africa, China, etc). Yes, even with Trump in office, though he really makes me worry.
I mean, the sitting President was shilling cars on the White House lawn and runs an active meme coin bribery slush fund.
This is not slightly behind Western Europe. This is miles behind any developed country. China may be corrupt, but Xi Jinping hasn’t yet sold beans or cars via press conference.
The rule of law gets down to nitty-gritty levels, too, not just a reality show at the highest altitudes: trust the police don't extort you, the ability to gain relief in court (small claims or civil), trust things you build won't be looted overnight, trust in your neighborhood to walk at night or leave something unlocked, trust in your bank to wire things, trust in your title companies, trust in your package deliveries, etc.
It's not perfect, but you could do so, so much worse.
"Could be much worse" is a platitude not an argument. "It is improving" might be, if it were true.
I often couch my arguments in soft language like a conversation would be in order to have a discussion. The idea that the US is miles behind developed nations is nonsense.
Why are these alternatives? I believe it is true that the situation in the US is better than many other countries (not most), and also that "almost no corruption" is false.
Being better than others really isn't the only thing that matters.
If you pursue excellence, you compare yourself against the best, not the worst.
Then the russian fellow shouldn't have dared to compare USA to his country?
salary is already counted as an expense right? does that mean we were double billed as two expenses? salary and R&D?
more like triple. do not forget the sales tax when you're spending it.
But where an established company invests steadily in software, whether it is amoritized or deducted year to year is a wash. Rather than harm tech, this would seem to protect established US companies at the expense of startups. Thats probably great for shareholders in publicly traded companies. It seems just another querk of taxation meant to maintain the established order
Only bootstrapped startups, funded startups will get the amortization by the time they need to deduct from earnings.
This is what I was thinking too
It's only a wash if they've been amortizing all along. There's been no advantage to doing so, so established ones have all been deducting, and will have the same five year window of increased taxable income that startups will.
Startups have to face that five year window every time they start up.
Each and every startup will have a year 0 where they're spending more than they earn, and under the new Section 174 they will only get to deduct 10% of their employee's salaries that year. In year two they get to deduct 20% of year 1's salaries and 10% of year 2's salaries, which is still 30% of what the established players will be able to deduct. By year 4, if they make it that far (which most startups don't) they'll finally be at 90% of a full deduction.
Add to that the fact that startups also by definition have a much higher rate of growth than established companies and you'll find that a startup almost definitionally will be paying substantially more in taxes as long as it remains a startup, because they only get to deduct an average of the last 5 years of expenses from this year's revenue in order to calculate this year's profit. That's fine when your last five years are more or less similar to this one, but it's terrible when you've been growing.
The net effect of this change can only be to disincentive startups and cement big, slow established players.
The net effect of this change can only be more tax income which benefits the society. Tax the rich
Increasing taxes does not always increase tax revenue. It’s easy to do enough damage to the economy that total tax revenues fall. Past that point is easy for the economy and government revenue to fall into a dwindling spiral.
Not if it limits growth to a commensurate extent (or more)
A big part of why America is as rich as it is in 2025 is Big Tech. If laws and regulations had prevented that industry from taking off by stifling the now-giants back when they were starting up, you may have been more equal today (you’d have fewer billionaires), but there would also have been a lot less wealth to go around, even for the working class
The OBBBA (“Big Beautiful Bill”) suspends amortization requirements for domestic R&D expenditure, and explicitly allows domestic software development as an R&D expenditure eligible for immediate expensing.
The new rules would apply from 2025 to Dec 31, 2029:
https://www.crowell.com/en/insights/client-alerts/house-comm...
Repealing SB174 has bipartisan support. The house already passed its repeal but it died in Senate because a separate took (that also repealed it) took its place but that separate bill stalled out.
174 is so small it can't go through both chambers on its own so it needs to get attached a larger bill like OBBA.
It's unfortunate because it appears both sides want this repealed to allow immediate amortization of domestic R&D expenses.
https://abgi-usa.com/section174/latest-and-greatest
It's so depressing to hear that congress can't even do small things that everyone agrees upon.
If they could be required to craft single issue bills, this wouldn't be as big an issue. Instead we get the clusters of good and bad that inevitably die or sometimes worse, pass.
If everyone agreed on it, Congress would have no problem doing it (Congress itself, after all, is a subset of "everyone".)
That's still not true. As long as a group within "everyone" (or multiple groups) decide that their support is required to pass the bill, they can suddenly demand concessions and the bill now gets complicated with good and bad.
> As long as a group within "everyone" (or multiple groups) decide that their support is required to pass the bill, they can suddenly demand concessions
Well, yes, but then everyone doesn't really want it, do they? Someone wants something else, and wants that something else enough that it is worth jeopardizing the supposedly universal goal for it.
Yes they do want it.
If you've ever negotiated, I bet you've done the same thing of jeopardizing something you want in order to get something else you want. If you never do that, you'll make a lot of deals where you're riding the edge of just barely acceptable and the other person is taking advantage of you. But in this case, with a standalone law, doing it gets pretty rude and we'd be better off if nobody did it.
It’s possible for everyone to want something, while simultaneously being incapable of getting it done.
A highly disproportional subset of everyone, maybe. Though uncapping the house wouldn't fix the Senate (maybe adding some more states would)
> 174 is so small it can't go through both chambers on its own so it needs to get attached a larger bill like OBBA.
There's a minimum size for laws?
Theoretically no, but congress won't vote anything that has no collateral advantages to everybody involved.
Which is one of the many ways in which the system is broken.
It's almost funny that small code reviews are preferred in software engineering (https://google.github.io/eng-practices/review/developer/smal...), but in Congress we have these stupid "big beautiful bill(s)" that are sometimes thousand pages long, and sometimes only released hours before a formal vote. Almost like these bills are intended to fool constituents, cripple opponents' plans, and created just in the hope that they get passed and signed into law without anyone looking carefully.
Code reviews for 10 lines of code: 100 suggestions
Code reviews for 1000 lines of code: LGTM
Omnibus bill packages get the same LGTM treatment where legislators don’t even read the whole thing.
The short timeline actually makes this an excellent opportunity for LLM analysis if you have a model which could digest the entire bill without reaching token limits.
Why would a senator from Kentucky vote for a bill that doesn't benefit his state to a meaningful amount?
Because they are a patriot. If not why are they in politics..?! Whom am I kidding...
It does benefit his state, they will get higher quality software.
But they can play games, drag their feet, get more concessions, and then pass it anyway with no repercussions.
Seriously?
Sadly, that is entirely serious.
Arguably, that's the whole of politics: why should I give you something if you don't give me something?
The people involved are, generally, not deep thinkers, aren't aren't thinking much beyond their direct short-term advantage. The system selects against that.
No, the system selects for people who are brilliant for using it to maximize their own individual benefit. Never take the bait that these people are in anyway stupid.
Anything that's not a budget reconciliation bill can just get filibustered in the senate by the minority party. That's why they're attaching everything to the OBBBA.
I think there is a limit on the number of bills that can make it through the procedures so it’s too low profile to get scheduled.
It's perhaps noteworthy that OBBBA is not the first bill to attempt to revert this tax law. It's simply the latest. There have been other attempts to revert section 174.
Other attempts that come to mind: 1. Tax Relief for American Families and Workers Act of 2024 (H.R. 7024) 2. American Innovation and R&D Competitiveness Act of 2025 (H.R. 1990)
This article is informative: https://www.cebn.org/media_resources/section-174-sign-on-let...
That "suspends" should be understood as "continues to hold-hostage" / "renews as a time-bomb to screw over some other party".
That isn't the reason. They sunset in the bill so it has a lower CBO score (which calculates costs out to 10 years). If you sunset in the bill after 5 years, even if you know it will get renewed, the apparent cost goes down. Get it?
The CBO score is a perfect example of the metric becoming the measure and teaching to the test.
Removing it would make Congress less powerful, and we can't have that now can we.
If anything it has been the opposite problem, with modern congresses having been more than happy to delegate away their powers. You might have heard the recent tariff news for example.
Modern being what? The main tariff laws in question happened 50+ years ago. The republicans in congress have been going out of their way to not interfere, but that's significantly different from creating new delegations.
Presidencies last 4-8 years, congressional careers last decades.
Avg for house is 12.5 and 8.5 in senate
Mitch McConnell has been serving in the senate since 1985. He's been a senator longer than most Americans have been alive.
I think the outliers are far more consequential than the averages in this case.
How could Congress get less powerful than now?
That's how they like, more time to campaign and less work that may be impopular.
Thinking about it, the US government is going exactly the same way of the Roman republic.
> US government is going exactly the same way of the Roman republic.
We're gonna conquer and annex Egypt? What an awesome time to be alive!
That was the first imperial conquest. The republic had been gobbling up states long before that.
Egypt? No. But how about Greenland, Canada, and Panama?
Kayfabe
That would be the one positive I have heard regarding OBBB. This should be put into its own bill.
That isn’t how legislation is passed. If anything, it needs a section about acceptable tar shingle application standards for roofs within 6 nautical miles of any heliport operated in a subarctic area on the west cost. Then it’s looking like a bill.
Just last year, Congress snapped to attention and wrote and quickly passed a bill to ban the eminent national security threat of a video-sharing app. That bill doesn't do anything else.
Just a reminder that Congress, even now, can rapidly act on a laser focus when it is sufficiently motivated.
Perhaps not the best example to choose given that the president managed to fully ignore that law. Tik-Tok remains unbanned to this day despite there being no sale.
It was a perfectly fine example. Nothing in your comment has any relevance to what it's an example of.
The President is ignoring a lot of laws of various vintages. That's not generally under the purview of Congress.
Because almost no voters would be against the bill and it harms almost none of the supporters of representatives
If you think no voters would be against the bill, I would suggest your model of our media landscape could use a refresher.
Is there a good summary of that episode somewhere? I've tried to read up on it as I don't really understand how it was an eminent (imminent?) security threat.
Is the TikTok debacle not a way higher profile case?
Did it have a larger financial impact than the tax code issue being discussed here? Absolutely not.
It was higher profile because Congress decided it should be higher profile.
I'm pretty sure that is precisely what GP referenced.
Yeah that’s my point
And it gets so bizzare that even legislators have to laugh when they read it out loud, like in this case here in Switzerland:
https://www.youtube.com/watch?v=S9hztUCq15o
There's a little of this, but more so, you only get one reconciliation bill per year. And anything that's not a reconciliation bill has to be bipartisan.
You forgot renewing the Patriot Act :)
> domestic software development
So now it seems its like a pseudo tariff against any other freelancers and producers for software outside of US.
This is a highlight in an otherwise shitty bill.
I saw let Trump’s ugly bill die and then a small fix up to the tax code could be this. Should be able to pass.
This bill is goated for upper middle class and tech and defense sector
And I’m tired of pretending like we aren’t going to be beneficiaries
Every Congress increases the debt, we can acknowledge that the cuts they picked are going to wreck the lower class especially with the medicaid, we can acknowledge that it won’t meet its goals of cuts
but are you guys just scared to acknowledge its going to super charge things that you are a beneficiary of too? so busy saying it just benefits billionaires as if we’re trying to avoid guillotines. not gonna happen and many people here are going to try to take advantage of new programs
You don't want to live in a society where an increasingly large percentage of the population have nothing to lose.
Regardless of whether it benefits our industry or socioeconomic status, it'd be incredibly shortsighted to just do all of that at the expense of the lower classes.
Also someday you might find yourself needing things like medicaid. I once met a former SE who was on disability because he lost his sight. Protecting those things for others also protects them for yourself.
I don’t have to be in support of something just because it benefits me
In fact condemning something despite the fact that it would benefit you personally ought to be considered noble.
I am very much in support of Section 174A no matter who does it, what riders goes alongside it, or what it was a rider to
Congress can always pass anything else at any speed. This slow motion filibuster thing is a choice, and the powerlessness of doing anything about that choice just means everyone else should have a single they care about too to correct the laws and riders that shouldn’t have passed.
> no matter who does it, what riders goes alongside it, or what it was a rider to
Really? You can’t think of anything you wouldn’t support in order to get this thing that benefits you?
I agree but I think the huge cuts across public research & development basically hurt everybody in the long term.
And China will pick up the slack.
I have to say that you sound very biased in the way you're talking about the bill. If a primary goal of this administration is to cut and make government more efficient, this bill is about as big of a failure as one could conceive. Any reasonable person would have to admit that.
I'm fine admitting that I would benefit greatly from this bill. I also hope to heaven it doesn't pass because an additional trillion dollars to suit me sounds asinine. I don't need help.
>Every Congress increases the debt,
And yet only one party campaigns on fiscal responsibility, debt concerns, and reduced spending
And funnily enough it is the other party that actually deliver improvements, while the party that campaigns on it just cut taxes when in power.
[flagged]
Amortization is bad policy when it comes software. Software is inherently high risk. Every piece of software is unique and does not guarantee steady income over 5 years. Most startups won't survive 5 years to fully realize the deductions. This is the end of US software dominance.
Amortization makes sense for things that have some inherent value. Like a microscope or computer.
A bankrupt company can still sell their computers. Selling you code, lol -- code is more of a liability really :)
The software that companies make is sold off in bankruptcy all the time.
I have a few friends who specialize in it with 2 ongoing contracts for splitting off pieces of software.
The value is far less than the amount being amortized for its development.
> Amortization makes sense for things that have some inherent value. Like a microscope or computer.
I am nitpicking but since a microscope or a computer is a tangible asset, the correct term is depreciation. Amortization applies to intangible assets.
> Selling you code, lol -- code is more of a liability really :)
It's important to consider that lawmakers (who are not well informed or downright stupid) might think code has intrinsic value because of media married with a lack of real-world experience.
Lawmaker is a misleading word. The people who actually make the law and lobby for it probably know quite a bit. The representatives are law voters not makers. They don't design the laws literally. They vote because they are told to.
Continuing your observation, this presumes they read and think deeply about the bills they vote on. They do not.
I remember the day I mentioned this in my high school^ honors sociology class and the eventual valedictorian exclaimed that I was stupid to think that. The system has been broken for longer than I have been alive, but the indoctrination has been working to make up for it.
This was a Blue Ribbon School 1992-1993 yup. https://www.ed.gov/sites/ed/files/programs/nclbbrs/list-2003...
Amortization is bad policy, period. If cost is actually incurred, it should be fully deductible immediately. No matter if it's a piece of equipment or software.
i'd disagree heavily with that... let's say you have an expense of an insurance policy that covers you for the next 10 years. You're paying for 10 years of service, that should be amortized over 10 years.
Yeah but if the insurance policy requires me to pay upfront, I'm out the entire ten years' worth of insurance premium. Amortization forces it to be divorced from actual cash flow.
> end of US software dominance.
What is that? Software sold by companies that have HQ in the US? Or software created by someone in the US? Because if it is only the first, good riddance.
Does a machine guarantee steady income during the period of its depreciation?
Is US software dominance because of our startups? Or because of the giant trillion dollar monopolies we have?
Didn't AAPL, GOOG and FB all create products _before_ they had any taxable income? Would this change have had any actual impact on their foundings?
> Is US software dominance because of our startups? Or because of the giant trillion dollar monopolies we have?
Most likely neither: It is its massive trade deficit, the one it strangely wants to get rid of now, that has allowed US consumers to consume more than they produce (i.e. you can take something with no real expectation of having to give anything back in return). Which, as it relates to tech, has enabled offering services for what is effectively free to dominate the market. Nobody else in the world can compete with that.
> Didn't AAPL, GOOG and FB all create products _before_ they had any taxable income?
Wouldn't you say they had no taxable income because of it? If Facebook brought in $100,000, and paid $100,000 to developers, then there would be no taxable income under normal regimes. But if the developers were not tax deductible, then that $100,000 in revenue would be taxable, even though the bank account is empty. This isn't nearly so simple, but it has changed the calculus in a similar way. The business models of old no longer work because of it.
Well, presumably the claim would be that a factor in their not having taxable income was the fact that they didn't have to amortize their development cost.
Yeah; start-ups will start paying tax much sooner since salaries are the main expense in software development, and only a fraction can be deducted per year. The tax change must make things marginally more difficult for young companies that have some revenue, aren't cash-flow positive, and have a short horizon.
It's not marginal. It significantly impacts sub-$10MM companies.
Hi, I'm just really curious about something. Why write AAPL, GOOG, and FB, and not Apple, Google/Alphabet, and Facebook/Meta?
Stock tickers are common and more readable than FAANG
It started with small and nimble innovators. Then it was shifted to Big Tech with the squeeze of patent trolling in the 2000's applied. It was capturing massive created value into the hands of few, connected, corrupt shitbags.
I’m not familiar enough with the very early days of Apple which started out as a hardware company to rebut you; but perhaps you mean the current Apple that has re-invented itself?
This impacts deductable expenses, not profits directly. The labor you pay for internally owned IP related to software must be amortorized. This screwed up an enormous number of business plans because software has more risk than many other endeavors. For small businesses, you basically can't do your own software.
It applies to things like configuring your internal tools too. Good luck at audit time.
I don’t know how it works in the U.S., but we had HMRC in the U.K. write us a cheque every year, as if you have a greater R&D claim than your tax bill, you get a rebate.
HN has taken a sad turn over the last few years where we see genuine curiosity - such as your reply - met with downvotes instead of replies.
I don't have an answer for you. But I support your intrigue.
It's worth noting that FB was quite possibly being secretly funded with taxpayer money by national intelligence interests at inception, which would have substantially reduced or eliminated commercial pressure early on.
DARPA was working on Project LifeLog starting in 2003, was to be "an ontology-based (sub)system that captures, stores, and makes accessible the flow of one person's experience in and interactions with the world in order to support a broad spectrum of associates/assistants and other system capabilities". The objective of the LifeLog concept was "to be able to trace the 'threads' of an individual's life in terms of events, states, and relationships", and it has the ability to "take in all of a subject's experience, from phone numbers dialed and e-mail messages viewed to every breath taken, step made and place gone".
The program, at least officially and publicly, was cancelled on February 4th, 2004, the exact same day that Facebook was founded.
https://en.m.wikipedia.org/wiki/DARPA_LifeLog
https://en.m.wikipedia.org/wiki/Facebook
You can call it a coincidence if you want, I just tend to be very skeptical of "coincidences" where massive, powerful, unaccountable, immoral, unethical institutions like the US intelligence community get exactly what they want at the expense of our civil liberties.
I often wonder if national intelligence interests are behind or have taken control of major corporate players like Microsoft, Google and Apple. There was an article [0] back in 2015 that brought forth the proposition that google was created by the CIA. It would explain the current enshitification of these companies and the lengths they are going to take away choice.
[0]: https://medium.com/insurge-intelligence/how-the-cia-made-goo...
Google Is Not What It Seems
https://wikileaks.org/google-is-not-what-it-seems/
Ever wonder why Microsoft bought Skype?
This is insane, how does it make sense? Employee salary expenses are no different from other expenses to run your business. Imagine they did this for raw material instead, a restaurant could only expense 20% of the food that they sell. If they purchased $100 worth of food, but could only sell $50 worth of it, they have to pay tax on that even when making a net loss overall. It just does not make any sense. There would've been a huge uproar if this was done for cost of goods. Why are employee salary expenses any different?
It makes sense when you consider that there is no minimum tax rate on businesses.
Given the choice, Amazon would rather spend 100% of its profits on itself than allow any of its profits to be paid out in taxes. Section 174 was implemented without a minimum tax on corporate profits before voluntary deductions such as research. Therefore, it’s exploitable and all companies ought to hire and fire staff to ensure their profits show as 0%.
This tax code defect is now closed by accident, but could have been done much more intelligently than it was. Oh well.
(EDIT: My first sentence is potentially confusing when I reread it later. To restate: section 174 was defective as implemented due to the uncapped 100% deduction, but the concept of a significant research exemption is still excellent. Just need to close the effective 0% corporate tax rate loophole.)
The company already pays payroll taxes on those salaries, and the employees pay income taxes. And the people hurt by this aren't the shareholders or top executives, it's the rank and file workers getting laid off, losing benefits, and being asked to work more for the same pay.
What this change effectively did was make software developers significantly more expensive, without increasing the amount those developers get paid.
The company does not pay payroll taxes. Individuals pay those taxes.
It’s split but the company pays more. Both pay SS and Medicare. Company also pays unemployment.
It doesn't actually matter that much who actually writes the check to the government (although in the US, both parties pay taxes).
Either way, the total cost of employment is higher for the employer than the after-tax income of the employee.
Software developers are already too expensive in US, so this applies some downward pressure on those salaries. Frankly the economy will be much better off when tech salaries equalize across geos, thus avoiding the deep whole US manufacturing is in (for example, manufacturing wages in Vietname are one tenth of US manufacturing wages, and thus it is better to open new plants there).
If you want equalized poverty, feel free to move to the EU. Say goodbye to owning a nice house, or building any kind of wealth - that's reserved for the old money class.
In the US, software is one of the few remaining ways to achieve the American dream. I came to this country to work hard and earn money.
Weird, I live in the USA make $180k yet still can’t own a house and building wealth is extremely hard.
EU has better societal benefits than the US (access to healthcare, education, mandated vacation time (often starting at 3-4 weeks).
The vast majority of people care about living a life without suffering. In the US this is only reserved for the rich it seems.
Yeah, make everybody equally poor. That'll solve things.
If you look at happiness and indexes versus taxation rates - yes, making everybody poorer does tend to solve things. Not too soon in the growth curve - but certainly not never.
Those two scenarios are only comparable if you isolate happiness and taxation and completely ignore things like social services and inequality.
I think you're referring to Nordic countries which consistently rank as the happiest countries and also have relatively high tax rates (4 of 5 Nordic countries rank in the top 11 tax rates globally. Norway has oil.) The high taxes that "make everybody poorer" also fund extensive social services that contribute to happiness.
However, this conversation is about making (a class of) workers poorer by using tax policy that puts downward pressure on their salaries. Tax revenues will stay the same, so social services will not be increased. Economic inequality increases because the workers became poorer, the C-Suite and Board Members don't.
[flagged]
Don’t forget the other stakeholder - the general public.
Yes it sucks for developers, but does it make any difference for any other employee? Why does Joe’s plumbing have to pay those taxes, but Jane’s AdTech company doesn’t?
Sure, there are benefits to investing in R&D in general, and tech has fueled a lot of growth, so incentivizing it has likely paid off for the whole economy. But will that forever be true? Maybe?
If Joe's plumbing hires an assistant plumber, they get to fully deduct the assistant's salary.
Why do I, the hardworking tax payer, have to subsidize Joe Plumber, who already has a big house with a pool?
In some parts of the world we have a sales tax which is a form of minimum tax on business outputs. The consumers of plumbing and software pay 10% regardless on a businesses profitability.
Yeah, VAT would help tremendously in alternative here, but for gestures at United States sociopolitics reasons the existing U.S. taxation methods can’t keep up and won’t be repaired any time soon. I could boil the ocean on this down to bedrock (citizens should be taxed on [redacted] in excess of threshold, services and goods should be VATed) but I stand by “section 174 with a sub-100% cap” as what at minimum would have balanced research and taxation.
In many parts of the US there are sales taxes, but they are state or local taxes, not federal taxes.
Joe's plumbing doesn't have to pay those taxes. Operational costs, including paying employees for normal operations, is deductable.
But with the change, the cost of R&D employees is now only partially deductible (right now, you can eventually deduct the full amount over the course of several years), and software development has to be considered R&D.
after 5 years then every year is deducting a whole year's worth of R&D - as long as that investment is not too lumpy from year to year you are back where you started
Which is fine for steady companies, but perpetually drags down any rapidly growing company
In this theory you should tax revenue and not profit. Welcome to VAT.
> Given the choice, Amazon would rather spend 100% of its profits on itself
And why is this bad, exactly? Money will be spent and will go back into the economy. Amazon will have to use the funds to build new offices, datacenters, do research, whatever.
And even if execs give themselves $10^11 USD in bonuses, they will be taxed as personal income, at even higher rates than corporate income.
It is complex - is it better for the money to go back into the economy by paying high salaries to a specific group of highly-educated people? Or is it better for the money to go back into the economy through taxes, then disbursing the benefits to lower-income benefit programs?
I’m not sure what the answer is. The former is likely to drive some innovation, which I’m sure varies by company. Where the latter could also unlock innovation by giving the bottom-quartile of earners a chance to improve their situation.
Those salaries are also taxed, and at the highest tax brackets. The government may end up getting more revenue that way.
The answer is simple: it's the biggest growth generator in USA.
Growth has its own problems of course (I don't want to estimate the health impact of Coca Cola), but it's a prerequisite of a country not falling behind others.
It can do both, by eliminating corporate taxes.
At that point, do we need to fundamentally rethink political donations by companies (outright ban them) and SuperPACs? No representation without taxation.
> It is complex - is it better for the money to go back into the economy by paying high salaries to a specific group of highly-educated people?
Yes. Also, the salary will not go _only_ to highly-educated people. For example, if Amazon decides to build a new distribution center, it will employ blue-collar workers to build it, not software engineers.
> Or is it better for the money to go back into the economy through taxes, then disbursing the benefits to lower-income benefit programs?
No.
> I’m not sure what the answer is.
The answer is pretty clear: invest money into the private sector, rather than divert it into the Federal budget. Private actors are more efficient at allocating funds than the government.
I'm not against social spending, it's a necessary evil for any real state. Pure libertarianism leads to dystopian outcomes. But it should be understood that it's a very real artificial inefficiency that is imposed on the economy.
There are also situations where additional social spending is necessary, but they are VERY easy to detect: when your interest rate is near zero.
Jesus man, how can you look at the economic history of the past 30 years and still think neoliberalism is the way to go?
Same as the communists, in that it hasn't been truly implemented anywhere?
The brand of neoliberalism where the fox sets up shop in the henhouse does not work.
State spending is not a panacea.
Real neoliberalism (with land value tax and pigouvian tax) has never been tried.
https://en.wikipedia.org/wiki/No_true_Scotsman
Then you misunderstand, the markets and economies of the past 5 decades have been two children playing Candyland. Saying it's not is a No True Scotsman fallacy, because clearly since I labeled it as Candyland economy it must be so.
because a high percentage people on HN fall into the group that benefits more from neoliberal economics than the larger group of people within those economies who don't benefit.
I used to think like you, until I saw what the lack of neoliberalism does to countries. And before I witnessed the magic of market economy that adapts to changes far, far, far better than anything else.
If you want a static economy that supports gradual decline (preferably with a mineral-based income stream), then a lot of state spending is fine.
Mostly, Amazon will do stock buybacks, so that its investor can invest into other top stocks.
Funds for the stock buybacks are not R&D, they'll be taxed.
It’s the same for movies, other intangible assets that are valuable and produce income over several years. And it’s done for many tangible goods, like servers in a datacenter, the kitchen equipment in a restaurant.
I think you may misunderstand. For most of those, you get the choice to amortize if you prefer. In this instance, you must amortize, which is a big problem for startups.
Employee salary cost isn't always 100% an expense.
Imagine you are BigCarCo, you make cars. The salary for your factory workers that build cars to be sold is an expense, incurred in that year, to be matched against the revenues earned by selling those cars. But the cost to build the factory needs to be amortized over the lifetime of the factory - and that's true whether you buy a factory from BigFactoryCo or hire a bunch of people to build it.
Now, I'd argue that a) most software dev work is closer to the factory worker than the factory builder and b) the lifetime for most software is less than 5 years, but the idea that some cost of developing software should be amortizable is pretty reasonable.
Actually, if the company isn't selling the software they build, what their software devs do is closer to building a factory rather than working in it.
Mostly developing software is about automating things that are expensive and slow to do manually. So, to stick with the factory analogy, it makes the factory a bit better and more efficient. If you stop doing that because it is too expensive, you fall behind with your factory.
Of course the whole issue in the US is that it outsourced much of what happens in factories to China and software has become one of the main things the country runs on.
There are other expenses that are also amortized.
Now imagine that a restaurant buys 100 tables, 500 chairs, kitchen equipment, cutlery for 800 people, signage, a security system, and does a remodeling before opening. (Or an airline buys an airplane. Or a hotel chain builds a hotel.)
Should they be able to expense all of those items that provide value for multiple years in a single year?
Does software development provide value exclusively in the year it's done? Or over multiple years?
The reason that we require you to deduct an expense over years for some things is because they have a resale value that needs to be accounted for. It's not a pure expense because you have an asset with real value that came out of the purchase. Employee time has no resale value. Once used it's gone, so employee salaries are expenses, not investments.
The only possible justification for the Section 174 R&D changes is that employees working in R&D theoretically are producing something which does have a resale value, so there's a small tax dodge enabled by direct-expensing your R&D costs but then ending up with an infinitely-copyable asset that came out of it.
If that's what you're saying, then I'd reply to that argument by saying that paying humans to design new things has historically been a business strategy that the government has wanted to incentivize in a way that buying and holding physical assets has not been. I've seen no justification for the government deciding that from 2022 on we should actively discourage R&D, it just seems to be a mistake.
Software is like Art, it doesn't have value until sold or can be used. If they sell services based on the software, they are generating revenue and then taxation on that revenue can occur.
Same as if they sell the software, either as a copy or ownership.
But not being able to take salary as a business expense seems like as thing that would happen if software in and of itself has value, which is largely does not.
> But not being able to take salary as a business expense seems like as thing that would happen if software in and of itself has value, which is largely does not.
To me it seems like a thing that just wouldn't happen. Forget software.
Say you own a McDonald's, and as part of your operations you have some people on staff to take orders, prepare food, and clean the bathrooms. Why are their wages not a deductible business expense?
If the answer is "they are, don't be stupid", then... what exactly was the R&D tax break?
The software itself has no value, it's the licence to use the software.
> I've seen no justification for the government deciding that from 2022 on we should actively discourage R&D, it just seems to be a mistake.
Removing a specific tax exemption to create a level playing field isn’t discouraging R&D.
That’s the thing, every year such exemptions exist the US taxpayers are handing out money. Just because we subsidize say EV’s or Corn doesn’t mean that’s the baseline forever more.
> Removing a specific tax exemption to create a level playing field isn’t discouraging R&D.
If the end result of removing this exemption is that there is less R&D done in the US, then yes, empirically, removing the exemption discourages R&D. Assuming the mass layoffs were indeed fueled by the removal of this exemption (I don't know if the article is correct or not), then it is reasonable to assert that it is true that removing the exemption has reduced the amount of R&D done.
Or, you could also say that the "default state" is some low level of R&D, and the tax exemption encouraged and incentivized more of it.
Either way you slice it, though, the status quo prior to 2022 was some level of encouraged/incentivized R&D. That status quo changed to encourage/incentivize less R&D, and companies have followed these lack of incentives and have fired a lot of their R&D staff. Is that a good thing for the US? I can't see how it could be.
> empirically, removing the exemption discourages R&D.
Not clearing a road means fewer people use it, but you not going out with a shovel to clear a public roads isn’t you discouraging their use nor is you canceling your plans to clear said roads.
Having zero subsidies is the default situation.
It didn’t create a level playing field, it just discouraged a very specific type of R&D while ignoring all others. All other types of employee salaries follow certain rules and some can optionally follow R&D rules. Software is now the only one required to follow 5 year R&D amortization so the deck is now stacked against software.
Software is an asset. If you pay people to build a building you don’t get to deduct their salaries as an operating expense.
The default situation is whatever was yesterday. I’d be astonished to learn that even a single significant civilization functioned without subsidies or patronage of priorities held by a society’s leaders.
> The default situation is whatever was yesterday.
If Amazon delivered you a TV yesterday that doesn’t suddenly become the default where you can expect another one today and every day after that.
The US government does a new budget every year, making every year a new ballgame.
No but if a TV was in my house yesterday I’d bet that it’ll be there today.
And my point about there being no natural state of subsidies is more important.
Those subsidies lasted a long time, but just as with a TV they didn’t last forever.
So if your argument is some subsidy will probably happen next year sure, but individual subsidies change over time. No specific subsidy is the default.
This doesn’t seem connected at all to your previous claim. You said that the default is an absence of subsidies?
There’s no contradiction between saying:
For any specific situation the default is no subsidy.
With millions of situations some of them are not going to be at the default.
In 500 years will some specific things be subsidized? Vs in 500 years will something be subsidized?
Level playing field for whom? Who does incentivizing R&D disadvantage?
Restaurants weren't competing with R&D-heavy corporations in any way. R&D-heavy corporations competed with each other, on a level playing field where all of them can build new stuff without having to pay taxes on negative income in their early years.
The only change this has made is un-level the playing field in favor of old, established corporations that already have the revenue streams in place to fund their new R&D projects.
> Who does incentivizing R&D disadvantage?
Taxpayers who end up with the bill and every company is competing for workers, office space, etc. Incentives across decades shift what people study, what business get created, etc. R&D sounds great abstractly, but it’s not some panacea where unlimited funding results in pure gains.
The economy is generally more efficient without central planning, and dumping money into anything that can be classified as R&D is simply inefficient.
> every company is competing for workers, office space, etc
My company is all-remote and none of us would work for a company that isn't doing R&D. Most of an entire profession now has to be amortized over 5 years.
> The economy is generally more efficient without central planning
The old tax code isn't "central planning", it just had the very reasonable property that the government wouldn't force you to pay taxes on a loss.
This scenario [0] is now possible. It wasn't before. That is a catastrophic level of stupidity, and you can't justify it with invisible-hand nonsense.
https://news.ycombinator.com/item?id=44204353
> none of us would work for a company that isn't doing R&D
So you’d just be unemployed for the rest of your lives? That’s a possible edge case not worth adjusting the tax code for, but it seems unlikely.
> wouldn't force you to pay taxes on a loss.
R&D is an investment, you only pay taxes if the rest of the company is profitable.
If your company is spending 1M / year on R&D and not adding 800k in long term value then in theory you’d be correct. But at that point you either aren’t doing R&D, or are doing such a poor job of it that the government shouldn’t be encouraging that activity.
The problem here is that all software development (excepting that done for hire) is classified as R&D. The software developer working on your Wordpress or Magento site (and arguably the accountant building a spreadsheet, to take the statute at face value) isn't an operational expense, they're now an R&D expense that has to be amortized and can't be taken as an expense against revenue. Previously, this was an optional choice (and many large and mature companies were amortizing anyway), but under the current tax treatment it's required, which essentially turns early-stage startups into cash bonfires, given how many small companies don't make it to year five.
> Early-tags startups into cash bond fires
As a practical measure it’s really not. The transition is difficult for existing companies, but a future startup is going to be minimally impacted.
Year 0 you’re unlikely to have any profits, future years you have multiple years of R&D to offset with.
But let’s assume the worst case. Taxes are 21% of profits and at minimum deduction 20% of R&D so the theoretical maximum distribution is 0.8 * 0.21 = 16.8% increase in R&D expenses if profits = R&D year 0. But that maximum case is only year 0, you’d be able to fund R&D with those same profits and easily be profitable after that.
If profits where say 40% of R&D in year 0 you’d have to pay 16.8% of 40% so an increase is only 6.72% hardly likely to tank the business if it’s already generating that kind of income year 0, and again after that point you’ll deduct for multiple years.
More realistic numbers are going to be really low multiples here, more importantly they represent significant investments not operating expenses.
> Year 0 you’re unlikely to have any profits, future years you have multiple years of R&D to offset with.
You're only unlikely to have no profits if you have no revenue. And you only get to break even 5 years in, which most startups will never reach.
In practice what is likely going to happen is that we'll see more and more startups deliberately avoid revenue in the early days. More and more free tiers followed by rug pulls when revenue actually becomes an asset rather than a liability.
There is no unplanned economy, only different outcomes from better or worse plans. And I'm having a hard time imagining a worse plan than one that intentionally disincentivizes businesses from adopting a sustainable business model early in their lifetime.
> unlikely to have no profits if you have no revenue.
It’s much easier to have revenue than profits, set the price lower and suddenly zero profit. Some company avoiding profits because of the 21% tax on profit like that would be mathematically dumb.
> There is no unplanned economy, only different outcomes from better or worse plans. And I'm having a hard time imagining a worse plan than one that intentionally disincentivizes businesses from adopting a sustainable business model early in their lifetime.
There’s zero advantage to avoiding revenue or profit here. You’re tilting at windmills.
You simply need less investor money for R&D when other parts of the company are profitable. As to central panning, the mistake you just made is mitigated when many people are all independently making plans. Governments always need to get it right, the market is fine if some people get it right and therefore can reinvest in their success.
It sounds like you’re talking about government funding of research? This is about private companies funding the costs of making product ideas into actual sellable products.
Money is fungible there’s zero difference between a tax break for 100$ and handing out 100$ directly.
Are you asserting that software and other labor-heavy startups should raise additional private capital so that they can pay taxes before they’ve established themselves in the marketplace? I’m not sure what you mean to say exactly.
I’m saying investers should pay the full cost of R&D without assistance from taxpayers.
When the non R&D portion of the business is profitable they should start paying taxes. Assuming a company isn’t miss classifying operations as R&D it shouldn’t be a major issue.
Thanks for clarifying.
This will of course discourage “riskier” startups and dampen innovation and give more power to profitable incumbents who will have less incentive to innovate. (Perhaps the result of this looks like Europe?)
Risky startups with multiple years of R&D before revenue would be the least impacted.
You’re only paying taxes if the business is profitable ignoring investments like R&D spending.
You seem extremely confused.
Section 174 specifically made those R&D costs “ignorable” from a tax standpoint. When it ended R&D costs could no longer be used to offset income.
What specifically do you disagree with? That R&D is an investment? I mean outside of the tax code that’s what it means to do R&D.
As to my other point, the highest risk category of startup has zero customers for years they also have zero revenue, zero profit, and zero taxes to pay here. On the 5th year they can deduct R&D from each of those years making the net effect on them minimal vs a startup with profits on year 0.
> The economy is generally more efficient without central planning
Big fat "citation needed" there. I know you chose the term "central planning" to try to invoke the communism boogeyman, but overall, free markets do not exist, and have never existed. Governments constantly use various levers (taxation being one of them) to encourage or discourage certain kinds of business activity. This is nothing new, and I find it laughable to suggest that this kind of thing should be done away with entirely.
There’s a lot of evidence for this outside of communism. Housing markets for example are a clear example of economic inefficiency created by subsides. But you also see problems with farm subsidies, flood insurance, and a host of other related issues.
Markets operate on revealed preferences, which is just a massive advantage in terms of giving people what they want. There’s definitely a role for governments in economies around information asymmetry, safety, etc, but allocation of resources specifically doesn’t work well.
What about construction worker and other labor time to build a factory? That’s the analogy being made here by the tax code: Software whose development is a capital expense with value returned over time.
From a quick search it appears to me like construction labor is deductible as an expense in the year it is incurred. Do you have evidence that says otherwise?
My reading of § 1.263A-1 is that construction labor must be capitalized.
§ 1.263A-1.a.3.A indicates that it's in scope: Real property and tangible personal property produced by the taxpayer
§ 1.263A-1.e.2 specifies that Direct Costs are subject to capitalization: Producers. Producers must capitalize direct material costs and direct labor costs.
(I'm just a taxpayer, not a tax lawyer or even an EA or CPA.)
What tax code references or treasury regulations did you find to support your belief that construction labor can be expensed in the year performed?
> Dear ChatGPT, is construction labor deductible as an expense in the year it is incurred according to GAAP? Please answer in a few lines.
Under GAAP, construction labor is not immediately deductible as an expense in the year it is incurred if it relates to the construction of a long-term asset (like a building). Instead, it is capitalized as part of the asset's cost and then expensed over time through depreciation. Only labor costs not tied to asset creation (e.g., routine maintenance) are expensed as incurred.
Though your answer is correct for the tax code as well as GAAP, Generally Accepted Accounting Principles are not necessarily followed by the tax code.
Unfortunately my understanding of the R&D expensing rule is that it is lifted directly from GAAP, which means private companies have to adhere to those (heavyweight) rules to comply.
Fair point. I changed the question to "according to the tax code" and it told me that
Construction labor is generally not deductible as an expense in the year incurred if it is related to the construction or improvement of a capital asset (like a building). Instead, under the U.S. tax code (IRC §263A), these costs must usually be capitalized and recovered through depreciation over time. Exceptions may apply for certain small taxpayers or repairs.
But the employee time that had a one time use was turned into software. That software is the thing that has value longer than "right now"
And the value of that software will be taxed if and when it starts to draw cash.
They should have the choice.
If software lasted longer than 18 months or was otherwise tangible, this could also make sense.
I have seen a lot of software development where what's been done has been changed beyond recognition over the course of less than a year.
It's only shifting what year the government gets its revenue. The government should simply let the company choose how to do it, but if they choose anything other than year 1 interest will be payable at government bond rates.
It's also massively shifting the companies' cash flows. The company paid $X for R&D this year, but for tax purposes 80% of that $X expense is moved to next four years. So for this year's tax purposes, the company R&D expenses are much lower than what the company paid.
Ironically I think they would want to claim that over multiple years unless they have other profitable operations under the same company. E.g. other restaurants.
The appropriate analogy is:
Imagine a restaurant spends money on employees to build 100 tables, 500 chairs, etc. Those tangible goods would be capital assets, so the labor costs of building them would also be capitalized.
This change to the tax code is just bringing the tax treatment of software development in line with how every other industry is treated. IOW, it was closing a loophole. A very valuable loophole, whose beneficiaries used it to get filthy rich, and bragged about how their industry was so much more valuable than everything else, even though a lot of that value was due to the exception software was getting in the tax code.
Notably, in the current version of the budget as of 6/6, the loophole is temporarily coming back, though given the Musk-Trump feud, it's very possible it will get pulled again to try to mollify the hardline deficit caucus.
If the restaurant buys e.g. a fancy oven or a delivery truck, it can't expense 100% of that cost in year 1, it has to spread that cost over the lifetime of the oven or truck.
Labor that operates the business day-to-day would be an expense, labor that creates a capital asset is more complicated.
I happen to think most employee time in software dev is more on the day-to-day operation side, and should be expensed, but I can see an argument that some should (or could) be amortized.
> If the restaurant buys e.g. a fancy oven or a delivery truck, it can't expense 100% of that cost in year 1, it has to spread that cost over the lifetime of the oven or truck.
The difference of course is that you'll have a truck or oven that can be sold. If you could count the full value in the first year then you could sell and buy one each year to reduce your taxes without actually changing anything.
Thus if we want to go that route for software the salary of the R&D employees should be counted against the value of the software they created (As in, the value were it to be sold wholesale to another company). The time spent by the employees is not an asset, once you pay the employees for their time it's gone even if they generated nothing of value. The actual value is that of the software, but that's obviously not easily assigned a value.
It was literally just a shot at California and New York, that’s all it was. “Own the libs” ya know
“if we aren’t rich then no one else will be”
Worth noting: the version of the Big Beautiful Bill passed by the House ends this particular change, starting in tax year 2025. We'll have to see if this provision makes it through the Senate, and in what form.
That's crazy. We're 3 years into a 5 year depreciation cycle, and now they "change their minds". Sure convenient when you know you are in power to supercharge growth and leave a time bomb for the next admin.
The destructive power of the Section 174 change cannot be overstated. It has been reported on a lot, but its harms are generally poorly articulated.
I do not like many things in BBB, but I am glad to know there is at least something in there that I can be glad for.
Lol only for it to kick back in in 2029 during the next administration. Your employment has now become a bargaining chip in the GOP's handbook.
I don't follow. Why is the GOP to blame here?
If the other party allows these cuts to expire, why wouldn't you blame that party?
Why sent tech companies and tech workers kick up a fuss when this bill passed in 2017. I remember being mad about it
Yeah I think we did kick up a fuss?
The better question is why the tech industry seemed to forget that the first Trump administration was terrible...
> ends this particular change
Temporarily, for 5 years.
If remember correctly, this was put in by Trump first round, set to activate when Biden was in office.
Now Trump second round fixes it, but expires in next (presumably) Democrat administration.
That is correct. Some historical context is much appreciated in this thread.
> tl;dr on Section 174, Research & Experimentation costs went from being fully deductible in the year incurred to being deductible over a 5 year period.
Larger tax bills and a tightening on what roles/activities are deductible as R&E are likely what OP is pointing at with his comment.
To the best of my non-inside baseball research, Section 174 changes were simply one part of a package of revenue generating measures to offset the large tax cuts from the broader tax act they were a part of.
The changes came from The Tax Cuts & Jobs Act of 2017 that was introduced to the House of Representatives by Congressman Kevin Brady (R) Texas. The bill passed both houses of Congress along party lines. Then President Trump signed the bill into law. Section 174 changes did not take effect until 2021.
https://news.ycombinator.com/threads?id=heymijo&next=4332098...
The title of this article implies that it is a major or even the only cause for mass tech layoffs, which I strongly doubt.
For example, rising interest rates I'm sure also independently contributed. I would be interested to if anyone has gotten a sense of exactly how much this has contributed.
agreed, the interest rates and the overestimation on the stickiness of the pandemic’s increase in internet usage post-pandemic are the primary other contributing factors that, imo, represent the lion’s share; even allowing that the tax changes are tertiary is a stretch much less as the primary/secondary reason
I only recently learned about the Section 174 change, and honestly didn’t expect it to have such a big impact.
I used to work at a small startup, and most of our spending went toward engineers’ salaries. If we had to amortize that over several years back then, I don’t think we would’ve made it.
It’s surprising how a single line in the tax code can quietly make it harder for small teams to hire. Makes me wonder how many other policies are silently shaping things behind the scenes.
As a non-American, it seems strange to me that the cost of regular software development, i.e. that is neither “research” nor “experimental” in a conventional sense, would be deductible in the first place (amortized or not). Isn’t that subsidizing a whole business sector? Maybe I’m misunderstanding something.
We're not talking about a tax deduction in the sense of a special privilege, we're talking about simple calculations of profit.
Before this change, tax for software development was calculated against:
* Profit = Revenue - Expenses
And software developer salaries fell neatly into Expenses unless you were looking for an R&D tax credit.
After this change, tax for software development is calculated against this new equation:
* Profit = Revenue - (1/5 * YearlyExpenses[-1]) - (1/5 * YearlyExpenses[-2]) - (1/5 * YearlyExpenses[-3]) - (1/5 * YearlyExpenses[-4]) - (1/5 * YearlyExpenses[-5])
Which means that if you are in Year 1 of operation, your values for YearlyExpenses[-2:-5] are all 0 and you only get to deduct 1/5 of your actual operating costs for the year from your "profit". So you can be in the hole but still owe taxes on your "profit" for the year because what you spent money on was classified as R&D.
Wasn't there something when this went into effect about the mid-year being the start so it is 10% in years 1 and 6?
Yeah, I just read that. So it's actually 10-20-20-20-20-10, which is both weirder and also slightly worse than my formula above.
That part is not so weird, you didn't pay all salaries on January 1st. But amortizing salaries in general is ridiculous.
It is a subsidy!
Why should money spent on software _development_ not have to be deprecated over time like other money spent on _development_?
I get that it sucks from a cash flow standpoint but the same is going to be true of other R&D expenses. It's just that we're more exposed to this specific R&D expenditure and not others.
The root of this subthread makes it clear why the current provisions to force software expenses to be amortized are different than other kinds of R&D.
The article makes it clear it's a subsidy.
> Originally enacted in 1954, Sec. 174 has historically allowed taxpayers to deduct SRE expenditures in the year incurred. Its original aim was to level the playing field for small businesses, those without dedicated research teams, that may be unable to deduct product development expenses under Sec. 162 because the costs were not ordinary and necessary expenses paid or incurred in carrying on a trade or business
Straight-up, any deviation in the tax code for a special group is always a subsidy.
I mean, yes, it will be true for other R&D types. But that's also new and also broken for the same reason: it means new R&D companies are at a massive disadvantage in their first few years compared to the established players who have lots of expenses queued up to deduct. It's wealth redistribution from young startups to established players who have 5 years of past expenses to use in their favor, and that is going to be a very bad thing for the health and vibrancy of our economy.
And, as a sibling points out (and as I pointed out in a comment at the top level), software is in this regime singled out from all other possible R&D expenses, making it particularly vulnerable. A skilled accountant/lawyer can probably turn big chunks of other R&D expenses into something that doesn't fall under 174. No amount of skill can do that for software, because we're singled out.
Because you slinging a React component or Vibecoding a security pile requires no Technology Readiness Level assessment nor does it have development liability. Rather, what we call Software Development is more appropriately labeled Software Engineering.
Salaries in general (not just of software developers) are tax deductible in many countries. This is desirable because we do not want companies to be paying taxes on revenue.
I dunno, I pay taxes on revenue.
Do you take a standard deduction?
It's clearly not enough to cover all of the expenses that are required to generate your "revenue", but it's a gesture in that direction.
Are you a company? If not, then you probably don't have revenue -- you have income.
In the US, unless you are a C Corp then you probably also pay taxes on net income of some form. C Corps have some different accounting, where dividends are double taxed unfortunately.
Small business owners are very impacted by the R&D schedule.
If only I can pay tax just for income - expenses (rent, food, etc.)
Businesses are taxed on profits, not revenue. Paying people to write code is an expense, so you'd normally deduct that expense (plus all your other expenses) from your revenue to arrive at an amount that should be taxed.
That's the rub. Is it an operational expense, like rent or a capital expense, like buying machinery?
It is sort of between the two in my view and is highly dependant on what the software engineer does each day.
Are they fixing a bug, helping a customer, refactoring? I think that is operational.
Are they building out a new feature? That is capital. But it is not quite like buying equipment because it adds no value to the books. So depreciation seems off.
But the same issue applies to other roles. Is a sales persons day trying to land a sale, or trying to develop the business.
It all comes down to "intangible assets" and whether you are making them.
I think it is easier to just say if you are paying someone to work then you can deduct. There must be better ways to claw it back.
The whole reason for most business to exist is to use operations (operational costs) as a lever to increase the growth and intangible value of the business.
The answer is that it's an operational experience when it's a salaried employee and a capital expense when it's a contractor. Like not in a theoretical sense, this is how it's classified right now.
Source?
Consider a contractor in a software maintainer role; accounting for this as capex makes zero sense.
It's how they're classified at $dayjob. It doesn't matter what they do, it matters that their contract is a fixed expense rather than an ongoing one.
That's probably something they are doing wrong then.
> Isn’t that subsidizing a whole business sector?
It is if the only thing your company does is create software. No operations, no sales, no physical assets to purchase sell or manage.
It stems from the difference in treatment of capital gains and income. Either way it’s deductible, the difference being when it is deductible and how much tax is saved. Capital deductions are typically done later since they require a taxable event.
It’s a fudge to make projections look better to allow congress to pass a budget neutral reconciliation bill with the intent that congress would remove the fudge before the consequences triggered.
Governments in general are pushing for capital gains tax normalization where instead of requiring a taxation event the capital gains tax would be levied yearly. In such a scenario the only difference remaining would stem from the difference taxation rates.
> Governments in general are pushing for capital gains tax normalization where instead of requiring a taxation event the capital gains tax would be levied yearly.
You’re alluding to wealth taxes, right?
Because taxing unrealised gains are wealth taxes.
Or maybe I’ve misunderstood?
> Because taxing unrealised gains are wealth taxes.
No, wealth taxes are a tax on retained wealth (a stock). Taxing unrealized gains is a tax on income (a flow), it just changes the point at which taxation attaches from a realization event to the actual gain.
But you haven't gained... you could be taxed over and over again, and if the stick drops or hits zero then what? It's all on paper and not "real".
> But you haven't gained...
Yes, you have. You have an asset of greater value which you can leverage in a number of ways without liquidating it and "realizing" the gains. That's a real gain, with real value.
> you could be taxed over and over again
Only if you make new unrealized gains.
> and if the stick drops or hits zero then what?
Then you have a negative unrealized gain, or, equivalently, an unrealized loss. If you are taxing unrealized gains instead of taxing gains when realized, then the natural assumption would be, just as is done with taxing gains at realization, that negative unrealized gains are either offset against current income or against future unrealized gains, and so effectively create (considered on their own) negative (current or future) taxes. The simplest form of this is to offset only against future gains, by the simple mechanism that when gains are recognized for tax purposes, they adjust the basis value of the asset, and when unrealized losses occur, they don't effect the basis value at all, so you don't have a taxable unrealized gain again until the market value exceeds the basis value established at the prior peak.
More complex versions would allow you to offset some or all of the unrealized loss from the prior basis value against current income of other forms, but the amount of that offset would reduce the basis value of the asset.
The unrealized value is notional, not actual. This is a very important distinction. The notional value is often not remotely realizable. In many cases, the realizable value can be a tiny fraction of the notional value.
Most laypeople grossly conflate notional and real value. Taxing notional value massively inflates the adverse impact of tax incidence on expected returns relative to people’s casual intuition based on the relative tax rates for realized and unrealized gains.
A tax on unrealized gains is in effect a way of laundering a steep tax rate so that it looks “small” and therefore reasonable to the unsophisticated.
> The unrealized value is notional, not actual.
No, its an actual thing, measurable by some mechanism. Otherwise, this would be a non-discussion, as taxing it would be impossible, not a possible thing that we can argue about the merits of.
> The notional value is often not remotely realizable.
Whether it is or is not immediately realizable is immaterial to the desirability of taxing it; it may be material to designing the forms of taxation that should be acceptable. E.g., if the difficulty of realizing the value is, across the tax base, likely to making collecting the tax in cash or equivalents difficult, it would argue for permitting a fallback option for the tax to be collected in-kind, e.g., by the taxing jurisdiction acquiring a proportional interest in the asset equal to the share of the value of the asset represented by the taxes not paid by other means.
> A tax on unrealized gains is in effect a way of laundering a steep tax rate so that it looks “small” and therefore reasonable to the unsophisticated.
If you allow carry forwarded losses, even just by the simple method of adjusting basis values, and include taxes on realized gains (and carry forward, offsetting against current income with perhaps a negative net, etc., for realized losses), then taxing unrealized gains is identical to taxing realized gains if the gains are eventually realized, but simply avoids the ability to find maneuvers to benefit from leveraging the value of the asset without paying taxes by avoiding realization. It doesn't make a "steep" tax rate look small, it makes the tax rate look like exactly what it actually is, unlike taxing only realized gains, which makes an effectively non-existent tax on capital gains look like something more, when people can benefit from assets without realizing the gains.
For many assets, like real estate, there are liquid markets with market prices. There are a number of US states that already tax based on real estate value, you can dispute the assessed value but that impacts other things like insured value.
Being difficult to assess value is a problem they’ll make you pay an accountant for and punish you if you get it wrong, it’s not going to stop them.
In the US, most recent studies of asset portfolios suggest that 60-70% of notional asset value has no liquid market. We already generate fictitious valuations for compliance purposes in many cases (e.g. 409A) that no one confuses with being representative of actual value. Tax policy based on overt fiction is bad policy.
Even in the case of real estate, a large amount of value is locked up in extremely non-liquid markets. You might get a vaguely representative market-clearing transaction once per decade, with high price volatility that makes it nearly impossible to predict what the next market clearing transaction will look like. I’ve owned assets in these types of non-liquid markets; differences in subjective valuations can vary by an order of magnitude and there is no evidence from the market to support any of those values.
If you only include extremely liquid markets for tax purposes in order to make valuations vaguely plausible, assets will be made non-liquid such that they are excluded from consideration. Ultimately this is why taxes on unrealized gains have been a challenging proposition in practice. We have no way to accurately model realizable value for the majority of assets and current simple approaches produce extremely wrong estimates a substantial percentage of the time.
They’ll make up a number and make you spend money proving otherwise. The government won’t care about your inconvenience when they need the money.
Of course this is a prediction of something that hasn’t happened before but looking at the chess prices move this does appear to be an intended destination.
If your asset valuations were swinging by an order of magnitude (10x), that wasn’t real estate, it was gambling in the fog.
That isn’t the asset valuation, that is the range of assessments interested parties make with respect to asset value. Because market clearing transactions are rare in many asset markets, those extremely high variance estimates of asset value are all you have to work with. It is only marginally better than no information at all. Too make matters worse, the rare transactions in these markets frequently have a lot of complicated structure such that the nominal price is not reflective of the underlying value.
tl;dr: Many assets have no meaningfully assessable fair market value. These are investments with extremely long and indefinite time horizons before the asset value can be assessed in a reasonable way. You can look at it as a peculiar type of risk capital portfolio with an extraordinarily long time horizon.
imo, it's in the best interest of the market for people to have to realize their gains otherwise the price of an item is pretty imaginary if it's never realized.
Gains are frequently not realizable as a matter of law and/or contract, for good reason. Additionally, there are many assets with notional value conditional on not liquidating them, which makes them de facto not realizable. And of course, the majority of assets have no liquidity, so realizability is a practical fiction.
The unrealized values are a fiction. There is significant value in treating values as unknowable when they are, in fact, unknowable. Forcing people to make up a fake valuation creates a lot of adverse incentives.
> The unrealized values are a fiction.
Then instead of taxing the gains, you'd accept the government nationalizing the assets by eminent domain and paying fair compensation that was significantly less than the "fictional" unrealized value?
Or if someone unlawfully deprived you of the asset, you'd accept as restitution or seek as civil damages for the loss something significantly less than the "fictional" value?
Or, when it was no longer an excuse to avoid fair taxation, would that "fiction" suddenly be a lot more real to you?
It would be much better to tax the benefits of the unrealized gain that a person realizes.
It’s much easier to do because there is no disputing the assessment since the person implicitly agrees to the valuation. And it allows people to forgo realizing any benefit from the unrealized value at all to avoid taxation.
Say take x% of the top of the money lent to someone who uses their unrealized gain to secure a loan. Make the money paid count against any tax they owe if they sell the asset later.
"uses their unrealized gain to secure a loan" sounds impossible to define with enough precision that you could base taxes off of it.
Have you taken out a secured loan in the last year over x amount?
Take the value of the asset assessed by the bank and the price paid for the asset to find the total value that is unrealized gain.
Divide that by the total value to get percent of collateral that is unrealized gain. Multiply that by the loan value. Then multiply that by the tax percentage.
All you need is for banks to report secured loans to the IRS and it’s easy.
Secured loans of that form are easy, sure.
But if those skyrocket in price from tax, they'll be more subtle about convincing banks they're good for the money and pay a slightly higher rate for unsecured loans.
Or maybe they'll just treat the asset securing the loan as having the pre-gains price. Get the bank to agree it's worth at least what you paid, with no further analysis.
If you try to plug those loopholes you lose the "much easier to do because there is no disputing the assessment since the person implicitly agrees to the valuation" factor.
>treat the asset securing the loan as having the pre-gains price
That’s no different than if the asset had no unrealized gain at all.
>they’ll be more subtle
It only takes a small rise in interest rates before it’s cheaper to pay the tax—assuming the tax isn’t outrageous.
Unsecured are much riskier because of the way unsecured creditors are treated in bankruptcy, so they already have higher interest rates.
It would be very easy to tweak bankruptcy laws to make unsecured loans over a certain amount a bit riskier to increase the delta even more.
We also already have regulations governing how banks assess creditworthiness, and the percent of their capital they can lend unsecured based on risk. As well as the amount of unsecured loans they can make to signal individual. If necessary tweak those values.
Another easy way is to add a surcharge to large unsecured loans where the loan amount exceeds the taxpayer’s assets based on acquisition price by some large margin.
None of those impact implicitly agreeing to the valuation and they are all pretty easy to do.
> That’s no different than if the asset had no unrealized gain at all.
It lets you get loans based on 100% of your pre-gain money with zero taxes paid, which I think is too generous. You wouldn't do that if it was actually all your money. It mostly fits the idea of only taxing the "used" money, but not entirely, and I don't really favor that idea in the first place.
> It only takes a small rise in interest rates before it’s cheaper to pay the tax—assuming the tax isn’t outrageous.
15% tax is pretty big. And if we put capital gains back in line with income tax it would be double that.
> Unsecured are much riskier because of the way unsecured creditors are treated in bankruptcy, so they already have higher interest rates.
If the unsecured loans only go up to 90% of the post-gain asset value, that's not much riskier, is it?
> We also already have regulations governing how banks assess creditworthiness, and the percent of their capital they can lend unsecured based on risk. As well as the amount of unsecured loans they can make to signal individual. If necessary tweak those values.
Yeah okay we could stop unsecured loans from happening. That seems awkward though.
> Another easy way is to add a surcharge to large unsecured loans where the loan amount exceeds the taxpayer’s assets based on acquisition price by some large margin.
I don't like this one at all.
You are basically making my case for me. It is widely recognized that replacement value differs materially from notional value. In such cases you may be required to pay much more than notional value because you have to pay for liquidity costs that only exist when transactions are forced to occur. The act of transacting can intrinsically change the asset value, sometimes by a large factor.
Are you oblivious to the extensive litigation that occurs in cases like eminent domain because there are substantial differences of opinion on even the notional value, never mind the realizable value?
Notional valuations are fiction, everywhere and at all times. Treating them as some kind of objective reality is just enabling a lot of abuse and motivated reasoning.
If pegged to inflation then they are not, but I think they generally will not be pegged. People who might think this is great should understand that the government makes more money increasing wealth inequality aligning the interest of the government and the ultra rich.
Most businesses let you deduct inputs and capital expenses from your revenue so that tax only applies to profit.
Since this is done on annual buckets it's very common to try to move items in both columns between years to minimize tax.
So if company A pays company B to develop some software, that revenue for company B (or rather, its profit) is still taxable? Then it makes sense I guess.
The revenue minus expenses is taxable, yes. And if the business itself makes no money, that means all of it was taxed through payroll.
Yes and part of the reason America is so rich
No, you have it right. Software was getting a special exception from the normal rule that salaries spent on creating capitalized assets are capitalized (which is the general rule for most industries, as well as for software development in most of the EU).
If your payroll ends up being about the same, after 5 years it all evens out in the sense that you will be expensing 100% of your payroll each year (but the expensing will be 20% from each of the prior 5 years).
If your payroll is quickly growing You experience the problem on all payroll growth.
If your payroll is decreasing, you get a tax benefit. Your outgoing cash is less, but you are getting deductions from prior year expenses.
Sure if you big enough to ride out 5 years but if your a hungry struggling bootstrapped startup, this can be game over.
No, it's always strictly worse because you could have bought bonds or deployed the capital in some other way with that money.
Your not taking into account the time value of money. You always want to expense sooner.
Additionally, having to wait 4 additional years to deduct that 80% is a huge drain on capital.
Combine this with higher interest rates and the effect is essentially pouring sand into the gears of the tech industry.
There is definitely a lot of misunderstanding here.
This provision can and does lead companies to owe significantly more in taxes than they make.
The only reason it hasn't been bigger news, is because most companies are pretending it doesn't exist and just sweeping it under the rug, hoping it will get fixed before enforcement gets serious.
I think the real reason it isn't bigger news is because the second you talk about tax code people start to tune out. It's easier to wind people up over AI taking jobs than it is to try and explain what amortization means.
> The only reason it hasn't been bigger news, is because most companies are pretending it doesn't exist and just sweeping it under the rug, hoping it will get fixed before enforcement gets serious.
Why pretend that it doesn’t exist? Why not vocally lobby for a change in the tax code?
There is bipartisan support to repeal the change. Meanwhile, further changes to the tax code are being prepared by the administration, very probably containing further such time-delayed footguns that will be the problem of the next administration to clean up, making them look like they raise taxes.
This change was added in 2017, triggered in 2021/2022. It's been the policy for years now.
There is very little pressure on elected officials because big cos can afford it and it bankrupts their tiny future disruptors.
Why would you let it be fixed?
Because then the people we watching might notice you and dig in.
My reaction to learning about this is that it is good news: this explains the weakening of demand for programmers, but unlike the AI explanation, this explanation does not come with a large risk of the demand becoming much weaker than it is now.
Also, finally programmers with the right to live and work in the US catch a break: salaries for US-based programmers can be amortized over only 5 years as opposed to the 15 years of non-US programmers.
I mean, the link not many people have made is that executives want to replace programmers with AI because of Section 174.
It has effectively become a lot more expensive and difficult to employ a programmer. Once this change went into effect we started to see hundreds of thousands of layoffs.
Then tech executives started aggressively talking up how you could use AI to write code instead of having humans write it.
Now of course reducing headcount and the associated expenses and replacing them with a bot sounds tempting to executives no matter what. But it sounds REALLY tempting when you've been on a hiring freeze since 2022 due to the fact that you can no longer deduct employee salaries in the year you pay them out.
Bear in mind that both Republicans and Democrats say they want to fix this and haven't done so due simply to gridlock and government incompetence.
I think most software businesses are taking a wait and see approach. Don't hire until this thing gets fixed. In the meantime, double down as hard as you can on automating those programmer jobs out of existence, in case the law never gets fixed.
Section 174 is the root cause.
> For cash-strapped companies, especially those not yet profitable, the result was a painful tax bill just as venture funding dried up and interest rates soared
Can someone explain this? What taxes do unprofitable US businesses owe that this would be deducted against?
Here's a toy example that hopefully makes this clear:
In 2024, your business has $1m in revenue and has $2m in expenses. 100% of these expenses are R&D salaries (engineers you hire.)
Your company loses $1m/year. (You brought in $1m and spent $2m.)
Under the old rules, you'd owe no tax because you were unprofitable.
After Sec 174, what the IRS now says is:
You had revenues of $1m. But you only had $400k in expenses (because you now have to spread that $2m in R&D expense over 5 years).
So actually you had a profit of $600k! And you owe tax on that $600k profit (~$120k)
So you now have an additional $120k tax expense, making your business even more cash-flow negative.
.
Amusingly, if you're pre-revenue, none of this matters (you have no income at all, so it doesn't matter what your expenses are.) You get hardest hit by this change when you have some revenue and when you do a fair bit of R&D.
> Amusingly, if you're pre-revenue, none of this matters (you have no income at all, so it doesn't matter what your expenses are.)
https://youtu.be/BzAdXyPYKQo
There is so much gold in that show. Just rewatching it now. Fucking billionaires...
Wait - they are saying that employee salaries are not expenses?
That is surely wrong? Just because those salaries are for R&D?
I could understand if there was some additional tax break for R&D which was being removed. I can't see how basic operating costs cease to be expenses.
They're still expenses, they just now need to be amortized.
Buying a truck is an expense, as is buying gas for the truck. But the former you have to amortize over x years, the latter you can expense immediately.
The law used to be "employee salaries for software are like buying gas" and now it's "employee salaries for software are like buying a truck".
The critical difference is that the business owns the truck but not the employee. The amortization assumes that the asset can be sold for value. An employee can quit at any time for any reason. You don’t retain the right to their labor for five years.
If they're producing a capital asset, you do retain the right to the fruits of their labor, even if they quit.
The rationale behind amortization isn't exactly the idea that the asset can be sold, it's that the asset is producing revenue over multiple years. For software, the asset is the codebase.
Let's say you hire a single software dev, for one year, and they write Excel++, which you can sell for the next ten years. It would be entirely appropriate to amortize the cost of creating that software over those ten years, based on the matching principle (a fundamental idea of accounting, matching expenses with revenue).
The issue in the real world is that's not how the software industry actually works, 99% of the time.
> The issue in the real world is that's not how the software industry actually works, 99% of the time.
What would be a more appropriate model from accounting perspective?
As anyone that has ever sold software IP knows, most of the value is vested in the person that wrote the code, not the code itself. The code is not a factory, it is the output of the factory.
> ...most of the value is vested in the person that wrote the code, not the code itself.
You must have misphrased what you intended to say. If what you wrote was true, a software company's most valuable asset would be the specific programmers in its employ. If true, average tenure of a programmer would be way longer than 1.5->2 years as companies worked really, really hard to keep their most valuable assets from walking out the door into the doors of another company just to get reasonable pay increases.
Perhaps your opinion is influenced by doing post-collapse-sales of a whole bunch of software houses that built just plain bad software? I can't see why else you'd be selling "software IP" independently of the rest of the business.
Anyway. Given that information, how should you have phrased what you wanted to say?
Most programmers do approximately zero work that is R&D. The most you lose if they walk out the door is institutional knowledge.
On the other hand, I've worked almost exclusively on software R&D for decades and seen the loss of a single person effectively end a project even when the software was essentially finished. Software R&D is about developing abstract knowledge, concrete implementation code is just a useful byproduct of that since R&D is typically motivated by a specific novel requirement.
If the software IP that results from R&D is not core to your business or a competitive risk, there is money to be made by licensing it. I've licensed this type of IP to big tech companies a number of times. If you are not actually doing software R&D, you are unlikely to be in a position where this is a possibility.
In almost every IP sale and licensing deal for software R&D I've seen, the value of any code is almost entirely conditional on retaining the services of person(s) that designed and wrote it. The entire "acquihire" phenomenon is an explicit admission of this. This is true even when the code is in a mostly finished form. Companies are buying capability, not revenue, so the code can't be a black box to their engineers. Companies usually spend more to acquire people with the code knowledge than the actual code.
The practical reality is that it is difficult to reverse engineer abstract knowledge from a concrete implementation. No one wants your code per se, they want to adapt your code to a different application that requires having a deep understanding of the domain the code represents -- they don't know what they don't know.
If you are just grinding out software that could be vibe coded then there is minimal asset value being created in the software artifacts. Anyone else would be better off reimplementing it themselves.
So yes, almost all of the value of code produced by software R&D vests in the people that wrote it. This is evident across many software IP transactions.
Based on my exchange with wdaher, who seems to understand this well, it's a bit more subtle than that:
The salaries are of course expenses, but they are exactly offset by the value of the IP created by the R&D activities.
It's a bit as if you spent money on buying some materials. As long as the material doesn't degrade, the cash is gone but the value is the same and therefore won't reduce your taxes.
If that IP is amortized over a single year, it does not contribute to taxation, but it does if it is amortized over a longer period.
They are expenses, but amortized over 5 years. So if you spent $2m on employee salaries, you would then deduct $400k from your revenue every year for 5 years.
If your employee expenses remained constant, then by year 5 you would be deducting $2m from your revenue since you'd be accumulating the deductions from the previous four years.
So in steady state it wouldn't necessarily be a big problem. But for a startup which is hiring many new employees and whose revenue is growing it's a huge problem.
This was my first reaction when I heard about it before it passed. I was horrified.
>Wait - they are saying that employee salaries are not expenses?
>That is surely wrong? Just because those salaries are for R&D?
The same would be true if you hired a bunch of scientists/engineers and got them to do R&D.
Would it also be true if you hired a bunch of construction workers and got them to build a stadium?
Buildings have to be depreciated, so probably? If you have to depreciate a building if you buy it, why should you get a free pass just because you built it yourself?
What other cost do you think goes into software development? Companies are not spending that much money on IDE licenses. The vast, vast majority of software/R&D costs are labor
So? They are all costs, whatever their source.
In the UK, business gets taxed on profit, which is what is left from revenue after subtracting costs.
Is this true even if you don't capitalize the immaterial IP asset generated by the R&D salaries on the balance sheet? Is that required in the US?
Otherwise I'm quite amazed that salaries can be carried forward as future expenses.
This is what the Sec 174 change said: it says that you do have to capitalize it.
Elsewhere in the world (under IFRS accounting rules) capitalization of R&D costs has been a firm requirement for a while. The US has been somewhat unique in allowing them to be expensed instead, until recently.
Taxes are calculated according to tax accounting rules, not IFRS, though?
I know of at least two Western European countries where you don't have to do that. Don't worry, we pay enough taxes either way ;)
Yeah, seems I was wrong about that. Apparently most IFRS countries allow expensing R&D for tax purposes, regardless of accounting. Many even have an R&D superdeduction nowadays.
Sorry for the noise :(
Some countries have uniform accounting where the tax accounting rules closely follow IFRS.
I was confused and has to double check. In Australia you can deduct them https://www.ato.gov.au/businesses-and-organisations/income-d...
Came here to ask about the Aussie RDTI. So, if I spend $10M on R&D and make $5M, what's the difference between US and Aussie net?
so you are okay, if you start getting revenue when you're five years in?
You can deduct 100% of salaries paid 5 years ago, but only 20% of salaries last year (etc.), and since companies tend to hire more people over time, most of your expenses will have been in the last few years that are still amortizing. You might have enough losses to carry forward in your first year of revenue, but 6 years in that could run out. It depends on the exact circumstances.
But nobody’s forcing you to classify software developers as R&D.
No, that's literally the Section 174 change. You now must count them as R&D.
The relevant paragraph from Section 174:
> (3) Software development
> For purposes of this section, any amount paid or incurred in connection with the development of any software shall be treated as a research or experimental expenditure.
https://www.law.cornell.edu/uscode/text/26/174
So that would include everything? - cloud/hosting expenses - system administrators/devops engineers and their laptops, workstations - project management software, office software, support, etc - project managers, designers, technical writers, qa engineers - software licenses, domain names, certificates, etc - internet bandwidth, data-centers, HVAC, backups
What "in connection with" means is vague. I think a reasonably competent tax attorney could probably argue that the costs of running your production cloud serving existing customers don't count, but IANAL.
what if you don't call it "software development"?
how about "business process mechanization"?
At that point you’re so into tax fraud that you light as well call them “postage and shipping”
Then you risk going to jail.
I mean sometimes fraud works, sometimes it doesnt.
What if some executive tweaks a "no code" tool? Technically, the name says that there's no coding involved.
Presumably that still counts as "developing software"- the regulation doesn't mention "coding" at all.
A fair point.
Or is it "using software"?
A person typing an essay with a word processor in doing more work than many of the users tweaking no code software.
How would that help? R&D developers helped saving taxes, now they don’t.
Classifying them as non R&D doesn’t help saving taxes again.
If the business has some revenue, but is not yet profitable after deducting development costs, it can become profitable on paper (and owe tax) if R&D is capitalized instead.
That is kind of strangely worded, but I think I see what they're getting at.
Say you would have been exactly not-profitable ($0) if you could expense all of your R&D as in the old system, therefore avoiding tax. Now with the new rules you may be on-paper profitable because you can only deduct 20% of the R&D as an expense this year. The remaining 80% of that expense tips you over, becomes profit, and that's taxable.
Right. With concrete numbers, say your main expense is $1 million in developer salaries and you have $500k in revenue. Going by the previous rules, you have a loss of $500k and don't owe income tax. With the new rules, you can only deduct $200k of expenses which gives you a "profit" of $300k, on which you'll owe $62k in taxes.
I thought you could carry forward losses or something. i.e. Once profitable you can use your previous losses as 'tax credits'.
So you're essentially giving the government a 0% interest 5 year loan, in the amount of the pre-paid taxes
If you make it that far.
Yes, but businesses operate on cash, not tax credits.
What taxes do unprofitable US businesses owe that this would be deducted against?
An unprofitable business doesn't pay income taxes. Businesses are taxed on their net income (i.e., profit).
People are railing against this as the cause of tech's recent underperformance, but it was a non-factor for the vast majority of tech companies, because most tech companies aren't profitable and wouldn't have paid taxes anyway.
From what I understand, this does not actually affect Google. They were already amortizing their R and D expenses.
Over long time scales (and big company revenue streams), this is sort of a wash. I think this hurts startups a bit more due to the long timescales involved which eats up much needed cash in the short term.
For me the worst things is that they treat all software as R&D. I understand in maybe some situation it could be abused but imagine established company having non innovative software that keeps engineers only for bug fixing and security patching and basic maintenance. In true spirit this is not research for sure. It's equivalent of someone having a hotel and suddenly telling that their cleaners, security, gardeners, receptionist qualify as R&D which would be nuts.
AFAIK it was also affecting more freelancers outside of US since amortisation is 15 years. For EU citizen IMHO this is equivalent of US putting tariffs on outside world. I wish EU at least try to fight back and revenge on US Tech by increasing taxes or also making all US tech bought by EU companies to be 15 years amortised so they have taste of their medicine.
It seems that there is quite a bit of confusion about this. What this does is that it reduce your deductible cost in the tax year.
First you have to make a profit (tax is on profits). Secondly, what this does is to limit your software development expenses for tax purposes in the current year because the development cost is seen as a capital cost that will be amortized over five years opposed to operating expenditure in the same year.
If you are a startup and not make profits, then the loss will be less in the current year, but either way, your tax liability is the same: $ 0.
So software development is moved from opex to capex.
Profit is determined by expenses though.
A simple example to illustrate:
Say you had 100k revenue and 1 software developer you pay 100k per year.
Under the new law, you can only deduct 20k of the developer’s salary, so your profit is 80k, which you have to pay taxes on.
However, you have $0 in the bank because you earned 100k and paid out 100k in salary.
See how that is problematic?
That's a horrible revenue to expense ratio for ditch digging, nevermind software development.
I can see why it would affect startups not making a profit but why would it dramatically affect FAANG (e.g. some of the most profitable companies in the world that have been running for decades)? The article contributes all these large layoffs in FAANG, in part, to this tax rule.
Because they are profitable. So the cost is deductible over 5 years, instead of one year.
A very simple example:
Revenue: $ 1 000 All other cost except software: $ 500 Software cost: $ 100
Net profit (if software is allowed as opex): $400
Tax on $400 (@30%): $120
Net profit after tax: $280
However, if it is capex(amortized over 5 years):
Revenue: $ 1 000 Other cost (except software): $500 Software cost: $ 100
Net profit before tax: $ 400
Important: But now for tax purposes you can only deduct $20 this year as a cost ($100 amortized over 5 years)
So now you have to add back $80 to net profit for tax purposes: $480
Tax (@30%): $ 144
Net profit after tax: $400 - $144 = $256
So the difference is $280 - $256 = $24
Just a few notes:
1. I assume tax rate at 30%, it can be something else, principle stay the same
2. That all other expenses are tax deductible
There's a difference of $24 but I have $1200 in cash reserves. And I make up the difference later. Oh no! Guess I have to lay off 10% of my employees now.
This is about way more than software. It's all R&D
It's effectively 6 years too. You only get to depreciate 10% in 1st year. This might have killed my company if it was around during first years.
See my comments on the previous discussion (Nov 2023) here: https://news.ycombinator.com/item?id=38145630
I made one of the original posts on HN about this years ago after hearing about it from my CPA. Both then and now these changes make zero sense to me as a matter of good policy. I am also still surprised at the number of people in tech who either haven’t heard about this or are willfully ignoring it and likely filing their taxes incorrectly.
It's not only that, ZIRP[1] contributed.
Also, the 10+ years before the layoffs started tech companies were on a hiring binge. Much of big tech was hiring to keep people off the market and off their competitors payrolls (this is from friends of friends in FANG HR departments). These were high paying jobs, too.
[1] https://news.ycombinator.com/item?id=44141650
We are at a bad inflection point of Zirp, tax changes, and AI.
All of which make hiring engineers unattractive
In knowledge-based white-collar work, there has been a significant increase in productivity in recent times. Tasks that once took days or even weeks—such as research, content creation, and visual generation—can now be completed within minutes. At the same time, both the speed of production and the quality of output are continuously improving. The outcomes are unavoidable.
It should be illegal to post graphs that start in 2020 when talking about tech hiring trends. The relevant comparison is probably the 2014-2019 era, not the peak pandemic craziness.
If you didn't know about Section 174 until 2025 you have no business being in a leadership position anywhere, period.
This has been a slow moving disaster for years now and people have repeatedly tried to raise the alarm.
Just crickets and layoffs.
The most fascinating question is not "How did a single line in the tax code help trigger a tsunami of mass layoffs?" but how did a single line in the US tax code help trigger a tsunami of mass layoffs in other countries?
There's a 15-year amortization period if a US company hires an offshore developer. It's only a 5-year period if the developer is American.
Because people make stupid correlations for clicks.
Is there a flaw in saying, "salaries should always be considered a business expense and cannot be amortized over many years." ?
Yes, I think there are two similar but subtly different flaws in that always.
1) Accounting rules are to match revenue with the expenses responsible for them, which I think is a good principle. If your workers make something now that provides revenue for 5 years, it makes sense to spread that expense over 5 years too. In many cases, you would want to do that as a business, makes it more clear how your business is profitable vs not.
2) Decisions whether to "build vs buy" a capital asset should not have massive tax implications. If I buy CoolSoftwareProduct from someone and resell it for the next 5 years, I'd have to amortize that. Should be similar if I hire a coder to write CoolSoftwareProduct instead.
(This doesn't mean that "salaries should always be amortized" is the right answer, of course, I think it's a very silly law)
Thank you.
I would say, yes there is a flaw there, because salaries are often a huge chunk of R&D expenses, and for the sake of long term growth, we want to disproportionately incentivize R&D spending
I think you inverted your understanding. My phrasing is inline with incentivizing R&D, or not dis-incentivizing it.
How about the salaries of employees being paid to create a new line of business? Say, the business runs restaurants and you decide to break into the tax software business. Can you avoid taxes on restaurant profits right now in order to build your new unrelated venture. ISFICR, tax law allowed outright deducting of costs of the current business, but not costs for starting a new one. Not deducting the new costs against the old profits.
After that, we can nitpick: should the development costs of new software be encouraged the same as maintenance costs of existing software. If you want to encourage startups, then yes they should. If you want to discourage startups or very temporarily increase tax collection, then no.
> restaurant profits
lol
Discouraging starting new businesses would be unconstitutional. All freedom in the US is derived from being able to participate in controlling capital.
They may want to encourage certain kinds of business over others.
Considering they are probably the largest component of R&D expenses... yes, _if_ you think R&D should be tax-subsidized in some way.
I don't understand how that is a subsidy, are the people paying the employer?
The employer makes less profit due to salaries, but they won't "lose less" or make more money due to salaries.
Under that argument, the government would have a direct incentive to dictate how businesses do business to maximize taxable revenue.
Tax subsidies are when the government taxes you less, thereby reducing your tax burden. You don't receive funds, you just owe less.
That's not what the law says. You'll have to take it up with Congress.
I think they're asking whether there would be a flaw with making that change.
>The delayed change to Section 174 — from immediate expensing of R&D to mandatory amortization, meaning that companies must spread the deduction out in smaller chunks over five or even 15-year periods.
Doesn't this just amortize out to be roughly the same amount of deduction over the long term?
All the big companies mentioned should be relatively unaffected over an N>5 year time period. Also this was something that's been in the works for years so their accountants should have been planning for it so it wasn't a financial shock (and company financials seem to indicate no such shock).
If you look at the time value of money[1], a $1,00,000 deduction this year is worth more than $200,000 deductions over the next 5 years.
But more importantly, the article claims it was used as a tax shield to grow.
"Basically, as long as spending counted as R&D, companies could report losses to investors while owing almost nothing to the IRS."
"Once those same expenses had to be spread out, or amortized, over multiple years, the tax shield vanished. Companies that were still burning cash suddenly looked profitable on paper, triggering real tax bills on imaginary gains."
1: https://www.investopedia.com/terms/t/timevalueofmoney.asp
Sure, but that doesn't account for the allegedly apocalyptic layoffs from companies that don't fit into the "real taxes on imaginary gains" mold.
I get that this is bad for the VC monopoly bucks scene, but they were already down for the most part. If the changes are as the article alleges than all these big tech companies that are posting huge layoffs should mostly be fine because it's not a serious change from status quo for them.
My company was affected. The amount of money paid in taxes more than quadrupled from one year to the next.
It hurt small businesses that were slightly profitable. No one else. VC shops aren’t profitable anyway, so no taxes to pay. Microsoft took a 4 or 5 billion dollar write off, but they can literally write a 5 billion dollar check.
The issue is that the IRS wants you to pay them today on profits and cash that literally don’t exist. You make $1M in revenue and pay 5 developers 200k/year? You have no money left at the end of the year, but you pay taxes as if you profited about 900k.
Yes, but you are still talking about a loan. That 1M in write offs are guaranteed to come over a five year period.
From the horror stories I read from other founders, the bank wouldn’t loan on those numbers.
When we asked our accountants what they were seeing from other companies, the answer was “mortgage their house.” That assumes they had enough equity to mortgage.
Interest rates are bigger motivator of the layoffs than these changes. When interest rates are high that means investors far more heavily prioritize profits today over profits tomorrow.
This tax change just made it worse.
> Doesn't this just amortize out to be roughly the same amount of deduction over the long term?
With steady enough employment numbers, sure. Google has a weird one-time cost where they get hit with extra taxes at 80%, 60%, 40% and 20% of their employee's salaries for five-years and then it's all balanced. You can turn the money Google needs to borrow (or not invest) at some interest rate into a known number.
Any startup that is cash poor and especially one that is growing struggles. In year 3 you get to write off 20% of year 1's salaries, 20% of year 2's salaries and 20% of year 3's salaries.
> Doesn't this just amortize out to be roughly the same amount of deduction over the long term?
Yes, but if your business is not yet profitable, having to pay tax on money you don't actually have in the bank (because expenses exceeded revenue during the year) will cut into your runway, perhaps to the point that your company might not exist in five years... or even two or three.
Google, Facebook, Microsoft and many other of those old big companies are profitable though and they dont go anywhere in next 5 years (even if first 2 bleed out users)
Yes, if you are a profitable company operating at a steady state and your investors have a time horizon of (in other words, are locked in for) a decade or more.
Most companies in question don't fit these criteria. They are either large public companies subject to the reactions of the market to quarterly earnings, or small private startups that have limited cash (a runway of far less than 5 years) and are facing a perfect storm of a historic rise in the cost of capital coinciding with this change.
In either case, their cost of labor just went up by a lot and will continue to cause layoffs, labor market shrinkage, and diminished ability to develop new products.
Yes, big tech companies are lesser affected and they were already amortizing their expenses as “cap labor”.
It’s a pretty bad article general and to blame the law change on this is all kinds of disingenuous: “It’s no coincidence that Meta announced its ‘Year of Efficiency’ immediately after.”
I wonder if this was an unintended consequence, or if the politicians backed by big business really wanted to disrupt the software infrastructure.
If this article is accurate it doesn't sound like it. The change was a political tactic to make the tax bill it was part of comply with Senate budget rules on paper. Apparently this is a common tactic with tax bills, with the expectation that the changes will be repealed or altered in a later bill. There is a movement to repeal this change, but the effects have already been felt.
> Apparently this is a common tactic with tax bills, with the expectation that the changes will be repealed or altered in a later bill.
None of this adds up. You're saying, the legislators were trying to cheat and because it's a "common tactic" that kind of cheating is somehow good, but it's bad when the cheating doesn't go through?
On the other hand, being a common tactic implies that the possibility of it remaining in the books was well understood, and the declared "expectations" carry zero weight as evidence, even less than zero when coming from politicians.
Legislation like that has far reaching consequences and pretend "surprise" just confirms the intent behind it. It's only prudent to assume that we have a common tactic case of throwing sheet at the wall to see for how long it'll stick. If there's no backlash the "tactic" will remain there forever.
As another example of the same common tactic, consider the fact that all popular browsers have been used as Trojan horses into the users' local networks for like forever. At some point back in 2015 somebody objected so the browser makers started talking about fixing the problem but then stopped talking without fixing it because public opinion moved on to other areas affected by abundant sticky materials... Thus, that particular sheet remained on the wall for another 10 years and counting, and the story may repeat itself again.
When using bill reconciliation in order to avoid Senate filibusters to pass a budget, certain conditions must be met otherwise regular Senate rules and the need for 60 Senators to be onboard to avoid a filibuster come into play.
It's not cheating, it's playing by different rules to get most of what you want/need done and then sometimes those that played and gambled were intending to, or hoped to, make the changes later that require rules. Their hope is that 60+ Senators would be onboard for those changes because they (those that gambled and pushed the budget bill thru) managed to get what they wanted at the expense of #$%#ing something up that most others would then be willing to fix/address.
Agreed. If the members of the majority party can compromise within their single-party system, and play by certain rules, everyone else is powerless to amend or block the legislation.
In tax policy, every single change looks reasonable to one interest group, and like a cheat to a different interest group. That is just the nature of tax policy. Any change hurts some people, harms others.
Changes to Section 174 happen rarely and are not a “common tactic.” Changes to tax policy in general are common, especially in the reconciliation process. They can have unforeseen side effects. As well as side effects that are foreseen but considered more acceptable than other side effects.
> Any change hurts some people, harms others.
Not quite the sentiment you intended.
Ha! Call it a Freudian slip…
> You're saying, the legislators were trying to cheat and because it's a "common tactic" that kind of cheating is somehow good, but it's bad when the cheating doesn't go through?
I don't think GP made any kind of value judgment either way; they were just stating how things seem to usually work.
Around ~2010 I still had a lot of coworkers who claimed tech was basically incapable of defending its interests against other sectors. Maybe a bit different than today. I don't doubt that they thought they would get this repealed, but I would suspect the risk of the live grenade went to the sector with the least lobbying competence per revenue for the tax equations.
It was a very explicit change with its own very specific paragraph. Some stuff can be unintentional. This could not.
Bloomstink has a short article on R&D expenses/tax credits as does Reuters on some of the back and current history.
But just as an accounting note: R&D expense has nothing to do with the company having revenues for an existing product, which already is allowed to deduct cost of goods sold, selling and admin expense. It is a cost related to future business and in that regard, it is not crazy to say it should be amortized. That in the past this did not happen, or that accelerated depreciation for other assets is in the IRS code is a function of the government wanting to effectively subsidize business investment.
https://pro.bloombergtax.com/insights/federal-tax/rd-tax-cre...
https://tax.thomsonreuters.com/news/the-future-of-rd-expensi...
But most employee salaries are deductible right? If you hire a chef at your restaurant, you aren't depreciating their salary.
Doesn't that make software engineers one of the few employees with much worse tax treatment?
I think that is the simplest and best analogy I have read so far in the comments.
The chef doesn't create a meal once that you can sell for the next 10 years though. You pay him for time X, he makes a meal, you sell that meal.
That's fundamentally different from regular software development outside of agencies where there is no direct relationship. Software development is closer to an investment than an expense.
Amortization sucks in general, yes, because the money is gone and it doesn't affect your taxes to the same amount, but that's not different for any company doing manufacturing or anyone needing specialized tools or vehicles that cost significant amounts.
> The chef doesn't create a meal once that you can sell for the next 10 years though. You pay him for time X, he makes a meal, you sell that meal.
What if the chef invents a new signature dish that makes your restaurant famous for the next 10 years?
If El Bulli was in the US that would be an interesting question, but for "normal" cuisine it doesn't match, I think.
For chains like McDonalds where they actually research and develop ways to make pink slime look like a burger, maybe? But do you call them chefs?
When someone pays for labor to build an apartment building they profit off for decades, do they amortize that labor?
I think there is an argument that both could be valid. I wonder why we cant let the company decide to pick one over another and not be so fixated on one tax code to rule them all.
A comment I read in the thread here says the answer is "Yes". To which I have to say, that sucks.
That is production, not R&D. Basic research into, for instance, semiconductors may one day lead to the production of a new product. The R&D costs would be amortized, the eventual production costs would not.
What is "Bloomstink"? Neither of your links references it, there are no references to the thing that makes sense when I do a web search.
Bloomberg
Deferring depreciation and deductions decreases their value as inflation happens. So it is a double whammy-not only is your profit reduced this year (and the impact that has on your stock price or return to your private equity investors) butthe value of that deduction decreases over time. So it is not a 'wash'.
I was part of a small R&D company that had a promising product (can't say more, NDA) and we had to shut down because of this. Thankfully the founders were able to get us acqui-hired or I'd be in a much worse position. But that IP is just lost to history AFAIK, in spite of significant investment of US research $.
Honest question, is there a community / grassroots effort I can participate in so that this this section 174 change can be reverted to its pre-2022 state?
I'm wondering, if such a movement doesn't doesn't exist already, do I need to start it myself?
'mjwhansen founded the Small Software Business Alliance specifically to work on this issue: https://ssballiance.org/
- Gather up about 10 million dollars (more will help)
- Bribe the right people
I hate to provide such a cynical and lazy response but we've got until midterms (maybe) before you really have a shot at 'democratically' influencing the system. For the time being you'll have to work with the mafia that's currently running things and outbid whoever wanted this to happen in the first place.
Ironically the debate/discourse here is healthier than anything we see from US Congress. Presidents have a limited number of terms they can serve, but there is no limit on the House and Senate and change is hard if not next to impossible because of this. Term limits would be a good start to introducing positive changes, but good luck finding the necessary majority to vote against their power and very cushy and comfortable lifestyles.
Question, if you have to amortise it over 5 years, and you can survive the initial 4, does it break even in year 5 (assuming stable employment)? Ie you amortise the previous 5 years (20% each) which works out to 100% anyway?
If you give me 1000$ today and I give you back 1000$ in 5 years do you break even?
This US American idea of paying taxes = giving money away is weird. Especially coming from tech people, as what is preventing other nations from taxing the hell of big/US tech is the US government and its threats.
Related. Others?
Big Beautiful Bill R&D Tax: Will tech go on a hiring spree again? - https://news.ycombinator.com/item?id=44028106 - May 2025 (19 comments)
The Consequences of Limiting the Tax Deductibility of R&D - https://news.ycombinator.com/item?id=43639202 - April 2025 (64 comments)
House restores immediate R&D deduction in new tax bill - https://news.ycombinator.com/item?id=39212650 - Feb 2024 (8 comments)
Ask HN: Best country to run a boostrapped startup from? (After Section 174) - https://news.ycombinator.com/item?id=39098371 - Jan 2024 (31 comments)
US tech innovation dreams soured by changed R&D tax laws - https://news.ycombinator.com/item?id=38988129 - Jan 2024 (3 comments)
Ask HN: IRS section 174 – cause of layoffs? - https://news.ycombinator.com/item?id=38957651 - Jan 2024 (21 comments)
Will US companies hire fewer engineers due to Section 174? - https://news.ycombinator.com/item?id=38931860 - Jan 2024 (37 comments)
Will US companies hire fewer engineers due to Section 174? - https://news.ycombinator.com/item?id=38870429 - Jan 2024 (20 comments)
IRS tax code change in Section 174: R&D is an expense - https://news.ycombinator.com/item?id=38642461 - Dec 2023 (23 comments)
New tax rules on R&D expenses may lead to layoffs for devs - https://news.ycombinator.com/item?id=38636866 - Dec 2023 (7 comments)
Tell HN: People laid off in my company due to IRS Section 174 changes - https://news.ycombinator.com/item?id=38633668 - Dec 2023 (6 comments)
Tell HN: Submit comments to IRS re tax treatment of software dev expenses - https://news.ycombinator.com/item?id=38120388 - Nov 2023 (225 comments)
Software firms across US facing tax bills that threaten survival - https://news.ycombinator.com/item?id=35614313 - April 2023 (981 comments)
Ask HN: How are you handling Section 174 changes for bootstrapped companies? - https://news.ycombinator.com/item?id=34627712 - Feb 2023 (187 comments)
https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que...
Not a big thread, but a layoff one from a year ago: https://news.ycombinator.com/item?id=38633668
Added. Thanks!
> “I work on these tax write-offs and still hadn’t heard about this,” a chief operating officer at a private-equity-backed tech company told Quartz. “It’s just been so weirdly silent.”
Hasn't Ben Thompson of Stratechery spoken about this a number of times? I'm aware of this 'feature' and I'm not even in the USA, let alone a COO at a private-equity-backed yada yada.
Earlier discussion: https://news.ycombinator.com/item?id=44028106
I think people are misinformed about how to deal with 174. That said, yes, a repeal would be a good idea. Hopefully that happens through BBB in the next month or two.
You do not have to amortize 100% of your engineering costs. Not even close.
Here's the key:
How does this work?You are going to design a new robot arm.
In January, you spend $100K to "remove uncertainty". In rough strokes, this means discovering all the things you don't know and need to know for this robot arm to become a product. This amount will be amortized over five years under 174.
Now, with uncertainty removed, you spend an additional $1.1MM from January until December for engineering implementation. No uncertainty being removed. Just building a product. This is 100% deductible that tax year.
Analogy: You want to build a new brick wall with specific properties. You spend $100K to develop a new type of brick and $1.1MM to build the wall using that brick. The $100K is amortized, the $1.1MM is deductible in one shot.
BTW, at year 6 the amortization schedule reaches steady-state and you are amortizing the full $100K every year. In other words, the impact of 174, if treated intelligently, is the time value of money until steady state is reached for the engineering costs incurred to remove uncertainty.
That said, I hope the BBB repeals this.
https://www.law.cornell.edu/cfr/text/26/1.174-2
That’s not what “ghost in the machine” means
This seems a clear disadvantage to corps with employees but it also works to the advantage of LLC solopreneur types who aren't paying themselves.
If anyone cares about combatting government propaganda in the US, do this:
Query any search engine for "are US income taxes direct or indirect taxes"
Every one will tell you that they are direct taxes. This is false. The supreme court has exclusively held that income taxes have always been indirect taxes (excises specifically, read about what an excise is in any authoritative source on tax law) in a constitutional sense. (See Brushaber v Union Pacific RR Co. 1916, Moore v U.S. 2024)
The sixteenth amendment did not give congress the power to directly tax citizens (or domestic corporations) and the complexity of the tax code is an attempt to obfuscate this fact, but the code is not inscrutable, it has rules.
Unsure of why this matters? Look up the difference between direct and indirect taxes in US law. None of these deductions matter unless you are a foreign corporation. I have tried commenting about this in other income tax related threads (this is my alt account), but people here don't like the idea that there is government propaganda in the US, or that most people are wrong and blindly accept the socialization about taxation without verifying what the law says.
I realize this is a disturbing truth to accept, not least because it involves accepting that most people who have been prosecuted for income tax crimes are only guilty of ignorance of the true legal purpose of the forms they signed. You can easily verify that most accountants and tax attorneys do not know what they are talking about by asking them this simple question about direct vs indirect taxation.
This is not legal advice, it is a wakeup call.
> This is not legal advice, it is a wakeup call.
The masses are in a persistent state of slumber, so they will never wake up. Depressing but true.
And there I was wondering why R&D was getting moved to Canada.
Note that Trump's Big Beautiful Bill as it passed the House of Reps would bring back 100% expensing of R&D expenses including software development costs/salaries.
If the article is to be believed, though, the damage is already done. Companies have already laid off large portions of their R&D staff, and have canceled lots of forward-looking technical work. Re-hiring those people and restarting those projects can take years, and that's if companies feel confident enough that the exemption will stick around, and not get removed again in a few years.
It's always young companies that hire everyone, so it still helps.
Still, I'll take it
"I'll give you a candy bar if you let me stab you multiple times"
Didn't realize this was due to Trump's first term.
Why aren't the All In podcast bros ragging on Sacks about this!?
This article was so clearly written with AI it hurts.
Manufacturing is affected also it’s not just software. Best way around it is deficit spending for growth
That’s not what “ghost in the machine” means.
Meta: god articles like this drive me crazy. What ever happened to the inverted pyramid hierarchy of information? I have to read 3 paragraphs of unmitigated filler before they actually tell me what's changed. It's not just this article, it seems like every article on newer media sites is like this. I understand why, but fuck them very much and the incentives that drive this behavior.
Could this lead to a new financial product that lends money to companies to pay this tax secured on the future deductions?
This would be a no go for startups though.
Maybe Shareholders are demanding companies to cut overhired and improve profits
Why is this the first we’re hearing about this, three years in? The article says these companies blamed other factors for the layoffs - why?
There were a ton of stories around this in 2022 as a bunch of startups scrambled to make ends meet due to this bill taking effect.
It's not the first time many have heard about it.
“We are heavily subsidised by taxpayers” is not great optics.
To head off the likely questions, I downvoted this comment because it is a gross misrepresentation of section 174 and the changes made to it.
What have I grossly misrepresented? Or are you arguing that a tax rebate is not a subsidy? I haven’t even mentioned the changes made.
To start, rules for deductions aren’t tax rebates. Rebates also aren’t necessarily subsidies unless they’re targeted.
Deciding whether labor is a capital or operating expense and deciding how to depreciate it if capital is also not a subsidy.
Would you like a razor with which to better split hairs? Also, I don’t appreciate your blatant racism.
I thought this was changed in this new spending bill they are trying to pass now?
So, we want incredibly profitable companies like Google, Microsoft, and Apple to take their software development costs and subtract that from their tax bill? These are the same companies that file patents so nobody else can use the ideas that they developed at the expense of public services. How about making it a tax break only for small and medium sized companies?
Fix the patent problem. Leave the r&d right off alone.
Oh interesting. In 2022, the company I worked for folded because a primary investor spontaneously pulled out. We were nearly 100% R&D and the sudden change in relationship was surprising.
Let me guess - the keyword here is "Section 174", just from the title alone :)
Dealing with Section 174 amortization in those first one to three years is a real headache (and your tax bill ends up higher than if it didn’t apply). Once your startup survives that the first few years of doing Section 174, things do get easier... but, sadly, most don't make it that far.
> For almost 70 years, American companies could deduct 100% of qualified research and development spending in the year they incurred the costs
This is an artificial subsidy. That’s not how the tax code treats other types of investments that generate recurring income.
Its so funny to me that people freak out about amortization when I spent several years at a public company having to document my work as being R&D to amortize it to make our EBITDA look better.
How this impacted our business is that when you are doing next year planning, and the goal is to grow the business, it made ads and other marketing investments more appealing versus tech hiring to expand product capabilities.
bit of a self-own it seems. Start-ups and early stage companies might simply decide to start in a more friendly tax jurisdiction. E.g.: Switzerland offers a 135% deduction on R&D-related salaries in the year they are incurred, making it an attractive location for tech development
EU provides a large pool of experienced developers seeking new opportunities on salaries well below SV. Why pay 500K for a burnt out "rockstar" who spends more time on twitter than doing actual work when you can hire highly skilled people in Eastern-EU (or even in Berlin).
Section 174 seems unlikely to progress unless attached to broader legislation.
> "More promising is the Tax Relief for American Families and Workers Act of 2024 (H.R. 7024), which proposes restoring immediate expensing for U.S.-based R&D investments through the end of 2025. " -- https://www.pwc.com/us/en/services/tax/library/tax-committee...
That's a good policy but Switzerland is awful for startups: expensive, strict labour laws, few funding opportunities, risk-averse customers, fragmented European market.
If I could start anywhere in the World, Switzerland would be above all the war-torn and crime-ridden places, but business-wise it's no good for a tech startup.
The top of the article blames Trump for some reason, when every time this is brought up, everyone sites Biden Admin for messing this up.
Investors and stock holders should be extremely outraged that all of these businesses are knee capping their future profitability. Can't make all those future pension payments if all your investments can't stay relevant in the market.
Some people will point out that AI will fix this, no it won't:
1) The real cost is higher than anything you'd pay for a person an there is not likely any real change there.
2) AI will be lies like Actual Indians that won't scale
3) Here's the kicker: If AI does succeed, now these multi-billion dollar firms will have to compete with multi-billion dollar single person businesses, that eat their lunch
Its a race to the bottom right? That means you need to invest in the business and all these layoffs are exactly not that, and will leave companies unprepared for the next 10 years.
Remind me in 2035.
I worked as a network engineer for a software company and had to report how many work hours was R&D. It was very silly.
Indeed. This is only a problem because software companies try to classify most of their expenses as "R&D" only to look more profitable than they really are, but at the same time they don't want to pay taxes on those supposed profits. For honest companies which don't capitalise their wage expenses, nothing substantial changes.
Love articles that are 39 pages long with one paragraph and 3 ads per page. mmmm.. good journalism.
Where is the comparison with the "control group" : non-US software companies ?
“Some spoke on condition of anonymity to discuss sensitive political matters.” - yep this is fine.
I reject this framing.
What really changed things was the end of ZIRP [1] and even then it was opportunistic. Labor costs are a massive cost for tech companies. They have continually tried to suppress wages. In the 2000s, it was the anti-poaching agreement between Steve Jobs, Eric Schmidt and others. In the 2010s, high growth ahnd zero interest meant labor costs continued to balloon.
But then Covid came along and was a massive opportunity. A few companies may have needed to do layoffs but that created the opportunity for everyone else. Big Tech just went full Corporate America with a page straight out of Jack Welch: fire the bottom 5-10% every year. Call it "layoffs". It's a direct pay decrease for those who remain (who get assigned the work). Those are still there won't be asking for raises because they're now afraid of their jobs.
Very little of this was ever necessary. None of the big tech companies ever came close to making a loss. They've remaining insanely profitable, in total and on a per-worker basis. At different times Google's per-worker profit has approached or exceeded $1 million.
The other factor is these companies eventually reached their size limits where antitrust stopped them making any more significant acquisitions.
Consider the timing: this change came in 2017. Where were the mass layoffs in 2018? 2019?
Also, the 2017 tax cuts contained a massive tax holiday for the repatriation of foreign profits.
Mass layoffs are simply wage suppression. It's the end state for any company that can't keep growing the way the market demands: eventually it comes down to cutting costs to make those quarterly profit targets. And in that, they sow the seeds of their own demise.
[1]: https://en.wikipedia.org/wiki/Zero_interest-rate_policy
> Consider the timing: this change came in 2017. Where were the mass layoffs in 2018? 2019?
The bill passed in 2017, but the changes to R&D didn't kick in until 2022.
Things can have more than one cause. Even the article only claims this change "has contributed to the loss".
Yes. And I reject that claim.
Big tech companies are both doing mass layoffs AND hiring. How does this fit the narrative that the tax change is at least in part responsible? The new hires still have the same deduction issue, right? So what impact does this really have?
Think of it this way: if this passes, will the layoffs end? Or reduce? Absolutely not. All this does is give line the pockets of shareholders. That's it.
I'm a big fan of tying certain benefits to NOT doing layoffs. This can include:
1. You get this deduction only if you've fired fewer than 1% of your workforce in the last calendar year;
2. You don't get to sponsor for an H1B if you've conducted ANY layoffs in the last calendar year; and
3. The tax deduction only applies to unionized workers.
And while we're at it, let's roll back this ridiculous tax structure where IP can be "sold" to a subsidiary in Ireland and then royalties paid.
Things wouldn’t be called layoffs then, people would just be aggressively PIPed out
That's a solvable problem and probably already solved. Fire more than a certain threshold of your employees over a certain period for any reason and it's a layoff in effect, say 3% over 12 months.
> Big Tech just went full Corporate America with a page straight out of Jack Welch: fire the bottom 5-10% every year
Plenty of "big tech" already did it. Microsoft could not be more famous for stack ranking dating back to the 90s. Amazon have long had that kind of culture too.
This doesn't quite fit into the article and is probably too inside baseball for a general business audience, but as I see it, there’s a real and serious argument to be made here about how Section 174 changes restructured the cost architecture of tech employment (yes, even for big, cash-rich companies). When salaries could be fully expensed, the effective marginal cost of headcount was lower. Amortization means the same engineer now triggers a significantly bigger near-term tax bill. At scale, that’s a serious shift in how labor costs flow through the P&L… functionally, op-ex becomes capex, and cash flow implications for big players run into the billions. But maybe it’s me!
Isn't this literally the content of the article? What you just wrote down is basically this paragraph from TFA:
> And so, on schedule in 2022, the change to Section 174 went into effect. Companies filed their 2022 tax returns under the new rules in early 2023. And suddenly, R&D wasn’t a full, immediate write-off anymore. The tax benefits of salaries for engineers, product and project managers, data scientists, and even some user experience and marketing staff — all of which had previously reduced taxable income in year one — now had to be spread out over five- or 15-year periods.
[0] https://news.ycombinator.com/item?id=34627712
Agreed that's literally what the article is about
> op-ex becomes capex
i.e. some humans get the same tax treatment as humanoid robots, while LLMs ("AI") are always deductible as op-ex, regardless of function.
Draft 2025 spending bill in Congress would revert Section 174 changes for 2026-2029.
Only external LLM use is ‘always deductable as op-ex’. If you build your own server farm and/or developer your own LLM, those are capital expenses which must be depreciated.
What percentage of LLM use is based on internal purchase of $30K H100 GPUs? OpenAI projected 2025 revenue is $12B.
Contractors licking their lips at the prospect of being a clients op-ex. I think you’re right and hence the slow down in hiring top talent for top dollar.
so will this incentivize a return to revenue/profit-driven business models? will we start to see a reduction in venture capital burning money on revenue-negative startups?
I have to say I do have some sympathy for what's happening in the US at the moment.
Trump is just getting started. By the time he is finished, your economy will be shot to pieces. The US dollar will no longer be the reserve currency for global trade.
[dead]
[dead]
[dead]
[dead]
[dead]
[dead]
at least all the future expenses will be AI instead of people /s
[flagged]
This doesn't explain the mass tech layoffs. According to the article, the rule applies to R&D. The vast majority of tech workers laid off in the last two years didn't work in research and development. They wrote regular software for sale, like games, for example.
The games industry, while hugely profitable and bigger than TV, movies, and music combined, laid off tens of thousands of people. It's unmitigated greed is all it is.
> For almost 70 years, American companies could deduct 100% of qualified research and development spending in the year they incurred the costs. Salaries, software, contractor payments — if it contributed to creating or improving a product, it came off the top of a firm’s taxable income.
According to the article, as long as the tech workers contribute to improving or creating a product (be it games or apps), they count as R&D cost.
I worked in games 2 years before the studio shutdown. It wasn't because of "R&D" tax breaks. None of the recent layoffs or studio closures are explained by that. Nor are the Microsoft, Dell, or Intel layoffs which aren't game-related.
To qualify for R&D tax breaks, IIRC having identified qualifying work for a segment of my firm, there must be elements of hypothesis, experimentation, results, etc that I would consider more science-y 'Research' than just turn the crank software 'Development.' It has to be both. And that has to be documented. And offshore research+development doesn't get you a tax break. The irony is that the R+D tax actually discourages onshore pure development as a 'trade' and encourages a split of onshore R+D and offshore D.
This sort of thing appears to be self-reported; I don't know if it ever gets audited. I don't know if big tech lies or creatively interprets what counts and that has contributed to the issue. But this article sort of over-represents what qualifies as R&D for US tax purposes.
Under the new rules, all software development, excluding bug fixes, must be expensed in this manner. "Turn the crank" development is included.
https://larsco.com/blog/section-174-updates-navigating-the-i...
Which makes sense. Software is functionally a capital asset, so really it should be depreciated across the length of the copyright term (unless the company wants to release it to the public domain to fully depreciate it early).
Maybe software should be a capital asset, but these depreciation rules don't fix that issue.
The rule says if you pay someone $200k to develop software: then you now have a $200k asset that then devalues to value of $0 over 5 years (starting midyear). That's just plain weird.
For our example a depreciation table might look like:
The final effect of the 174 rule change is that you still finally end up with a software asset worth $0. However you now have taxable income of $200k in year one and expenses equalling $200k spread over 5 years. The taxes paid could be a lot: although the taxation money is really just being lent to the government for a few years at 0%. The actual financial costs are fucking complicated.Understanding accounting and taxes are two absolutely essential skills if you ever wish to be a founder (and useful anyways).
Finding a solution to dealing with the valuation of assets is difficult. The historical solution of depreciation is broken for software, intellectual property and goodwill. In theory, taxes on dividends and capital gains taxation already deal with the issue (company taxation at x% kinda ends up at $0 because the shareholder pays y% and claims back the x% through imputation).
And remember that salaries are properly taxed.
Right, that weirdness is why it should be depreciated over the length of the copyright term. You spend $200k this year, and now you have a useful asset for the next 95 years (or 120 years if you never publish it).
If it turns out it's not useful, we could then allow companies publish the source and release it into the public domain to immediately "destroy" the asset (the copyright) and claim their deduction. So failed r&d projects would be deductible right away as long as the public gets them, and ones that result in a useful asset get depreciated based on how long they actually last, which is currently potentially multiple lifetimes.
I don't think copyright term is a good rubric/measure here. For SaaS, a company can keep the software locked up indefinitely, regardless of copyright term. Employees can be contractually obligated not to publish source code, even if the copyright has expired.
Amortizing development cost over the useful life of the software is maybe a reasonable thing to do (I don't think it is, but let's for a minute say I agree), but determining "useful life" is not simple.
I get your thinking here but copyright isn’t the only relevant intellectual property constraint.
Software built by a business is a trade secret independent of its copyrightability. Even after the expiry of copyright a business can continue to exploit it as a proprietary asset.
At the last couple of companies we worked at, they just sent out surveys on what time went to different activities. We couldn't possibly fill that out honestly, as that wasn't tracked.
Which, I think is an overlooked part of this. They must constantly have gotten feedback that people were lying to them.
> The vast majority of tech workers laid off in the last two years didn't work in research and development.
I bet they were classified as R&D for accounting purposes. Product development largely falls into R&D - it doesn't matter what the product being developed is for.
Every job I had at a megacorp was classified as R&D, and I know because I had to track hours against such.
> I bet they were classified as R&D for accounting purposes.
It's not just that. Section 174 now explicitly calls out Software as always being an R&D expense:
> (3) Software development
> For purposes of this section, any amount paid or incurred in connection with the development of any software shall be treated as a research or experimental expenditure.
https://www.law.cornell.edu/uscode/text/26/174
> They wrote regular software for sale, like games, for example.
Even though it sounds unintuitive, that activity is considered R&D for tax purposes.
Greed is too easy as a target.. industry space has shifted because of slower innovation and less growth, so cost cutting being more a focus would be a reasonable strategy
If your company is already profitable to the tune of billions annually, "cost cutting" isn't necessary. You're just cutting people out of jobs and out of economic participation in society -- which affects a far larger group than just themselves when those folks can't spend their salaries in other businesses.
There is no justification for "cost cutting" when it hurts the larger economy. If the company were losing money, that would be different, but these mass layoffs are all from firms that make obscene, enviable levels of profit. It's greed.
If a company is making $10 billion in annual profits, and they discover that they're spending $100 million on a useless project, are they morally obligated to continue that spending indefinitely?
There is no justification for "cost cutting" when it hurts the larger economy.
It is not good for the economy to have people doing work that doesn't produce value.
> It is not good for the economy to have people doing work that doesn't produce value.
This is a political statement, not an economic one. What is or isn't good for the economy is up to the goals of that economy. In Japan, for example, they've more or less adopted the opposite principle as an important plank in their political system. Or perhaps it would be better to say that they've adopted the idea that people having jobs is more important than those jobs having a direct connection to some measure of productivity.
You can call any self-interested decision "greed" if you need to just turn off your brain and emote.
But they were making high profits for decades, and being greedy for decades. Then there were a lot of layoffs. What changed ?
You are correct saying it's not the R&D deductions.
But it's not "greed": it's the end of zero interest rate policies.
https://newsletter.pragmaticengineer.com/p/zirp
> They wrote regular software for sale, like games, for example.
Wrote software, like, you know, "developed" it?
I’m not sure I fully understand the problem here
1. I start “Facebook for dogs” It’s gonna be massive. For the first year me and five guys code away in the garage and I use my savings / credit card / family trust fund to pay them 100k each. Expenses are 500k, revenue is, amazingly, 1.5M and taxes owed is 500k.
At this point turning round and saying the development was R&D, and claiming 500k of tax breaks is just (to me) ripping off the American Taxpayer.
And I’m not even an American Taxpayer.
If the revenue was zero would anyone suggest that the taxpayer give me 500k to help ? (Ok I would because I like free money but most people won’t)
Or am I missing something?
Everything you just said but imagine revenue is $500K, and you spent $500K on salaries for the team.
You can only expense $100K of the salary costs this year, so even though you're break-even, you pay taxes on $400K in income.
Or, even worse, imagine revenue is $250K, and you spent $500K on salaries for the team.
You can only expense $100K of the salary costs this year, you're already -$250K on the year, and now you're paying taxes on $400K in income. You're destroyed.
VC-backed startups aren't designed to get profitable quickly, and I don't see that as a problem for the American taxpayer, and nobody is saying the taxpayer is giving money or helping. A business losing money should not have to pay taxes on income, as if it's not losing money.
the idea is that normal business expense are deductible.
in this case, your taxable income is $1.0M, and cash flow is $500k ($1.5M - $500k salaries - $500k taxes).
now you have to amoritize it over 5 years. so your taxable income is $1.4M ($1.5-500k*20%), taxes are 700k, and cash flow is $300k.
Uncle Sam just reduced your cash flows by 40% by a simple tax change. You eventually make up the difference, but for fast growing tech companies, that's a large shift in current flows and significantly changes their investment strategy.
See my example here for where it ends up biting you: https://news.ycombinator.com/item?id=44180533#44204246