The Tech Job Meltdown

Half a million layoffs? It's all accounting.

He wrote me a prescription; he said “You are depressed
I'm glad you came to see me to get this off your chest
Come back and see me later, next patient please
Send in another victim of industrial disease”
Industrial Disease, Dire Straits

The Google campus doesn’t look as friendly as it used to. (this ia actually from Bartertown in Mad Max 3 - you can see Thunderdome in the middle)

Since the start of 2023, more than half-a-million tech workers have been laid off. This isn’t the impact of COVID, this isn’t a sudden realization that tech workers are under-performing, this isn’t (much) a wave of AI making tech workers more efficient, and the other usual shibboleths like “it was overhiring during the pandemic” or “it’s a wave of H1B workers” or “all knowledge worker jobs are being replaced by LLMs” are only vaguely correct. You could make a pretty reasonable case that it was the end of Zero Interest Rate Policy (ZIRP) and the corresponding impact of cost of capital - that the cost of borrowing went up, thus venture capital became a less attractive investment class than other areas, so less money went to building new companies and it was harder for existing firms to borrow, as investors went elsewhere for better returns. That is correct - and it would have had its impact - though that impact would have basically been the slowdown of new venture backed firms, not layoffs at the Big Tech Giants - and that did in fact happen. (There is definitely a knock-on effect at the Big Tech Giants where a lack of tech startups does bad things to the large parts of the ecosystem - but that effect is not as immediate as it was back in in the 2000 dot com crash.) But there’s a much more immediate bottom line reason.

Section 174 of the Internal Revenue Code governs the tax treatment of research and development (R&D) expenditures. For roughly 70 years, American companies could deduct 100% of “qualified research and development spending” in the year they incurred the costs, and this was generally interpreted pretty liberally. Salaries, software, contractor payments… if it contributed to creating or improving a product, it could be deducted “off the top” of a firm’s taxable income. The deduction was originally codified by Section 174 of the IRS Code of 1954, and under the provision, R&D flourished in the U.S. It gave us the dominance of Bell Labs, Microsoft, Apple, Google, Facebook - pretty much all the US technology booms you’ve lived through unless you’re quire venerable.

So the way these regs are written: These expenditures must be for activities intended to discover information that eliminates uncertainty about the development or improvement of a product. (Kind of open-ended.) Prior to 2022, taxpayers could immediately deduct R&D expenditures in the year they were incurred, providing a significant tax benefit for businesses investing in innovation. Alternatively, taxpayers could capitalize these costs and amortize them over a period (e.g., at least 60 months) if they chose to defer the deduction. But it was pretty rare to do this, because you could directly manage your R&D payroll costs versus income to mitigate the tax hit. And societally, we accepted that - we were investing in growing the American economy.

But, the Tax Cuts and Jobs Act (TCJA) of 2017 amended Section 174, effective for tax years beginning after December 31, 2021. Starting in 2022, R&D expenditures must be capitalized and amortized over 5 years for domestic research (and 15 years for foreign research… which is pretty untenable.) This change eliminated the option to immediately deduct R&D costs, increasing tax liability for companies with significant research budgets in the short term. Even more annoying, amortization begins at the midpoint of the taxable year in which the expenses are incurred, using a straight-line method.

The short version is: this rule change has increased taxable income for businesses in the short term, as they can no longer deduct R&D costs immediately. Now, there is actually a category of tax law which you can still use (IRS Section 41 Research and Development Tax Credits) which are different - if you think I’m about to advocate tax reform, yes, but I’m always doing that and I feel like I’m talking to a brick wall. But it’s not as broadly useful, and it’s not a simple recategorization, or we’d all have done it.

Anyway, let me see if I can summarize how it works today. A U.S. company incurs $1 million in domestic R&D (Section 174) expenses in 2025 and let’s assume they can’t reasonably reclassify any of it under Section 41. Under Section 174, it must capitalize these costs and amortize them over 5 years. Amortization begins mid-year, so in 2025, the company can deduct $100,000 (1/10th of $1 million, since you only get to count half the first year. The remaining $900,000 is deducted evenly over the next 4.5 years ($200,000 per year). These are basically “tax credits”. So in some respects, they may be long term beneficial - but they are a short term drag, which is why you see layoffs at the moment. Also they create a lot more compliance paperwork (and potentially you’ll see companies change hands just for their accumulated tax credits, which is a little… unhelpful.)

And of course, this hit the “real world” of companies focused on building tomorrow’s products in some pretty obvious ways.

The inability to immediately deduct R&D costs reduced cash flow, particularly for cash-strapped startups and small tech firms reliant on R&D. Companies had to either take out high-interest loans - since of course, interest rates have recently gone up, cut costs, or face bankruptcy. Many chose layoffs to free up cash to cover these tax liabilities. For instance, a small company might lay off a software engineer earning $200,000 to cover a $189,000 tax bill.

Now, you might think the 15-year amortization period for foreign R&E expenditures would make hiring non-U.S. engineers less tax-advantageous and help bring jobs back to the US, but this mostly did not actually work out this way. You see, larger companies responded by offshoring R&D to countries with more favorable tax regimes, leading to U.S. job losses. For example, Google reportedly shifted some work to Germany, and Microsoft moved a bunch of research work to China - both because pay rates were better and because the local subsidiary company in that jurisdiction operated under the national laws for that nation, which … were not the US tax laws. They were much more like the previous US tax laws, because the rest of the world had realized “hey, we also want to encourage people to invest in R&D and grow the next trillion dollar company here!” And this new tax ruling doesn’t precisely say “we don’t want to do that” but it does say “we don’t want you to be quick about it” - which everyone who believes in the Amazing Growth Story thinks is anathema to their strategy.

Anyway, the impact of this tax strategy turned out to be: layoffs of U.S.-based engineers while companies restructured operations abroad.

This is NOT working out how it was planned.

Now back in 2017, when Congress passed the Tax Cuts and Jobs Act (TCJA), the signature legislative achievement of President Donald Trump’s first term, it slashed the corporate tax rate from 35% to 21%, which looked like a massive revenue loss “on paper” for the federal government. So in order to make the 2017 bill comply with Senate budget rules, lawmakers had to offset the cost… therefore, they put in a future tax hike (well, several) that wouldn’t kick in right away, wouldn’t provoke immediate backlash from businesses, and could, in theory, be quietly repealed later.

But they didn’t repeal it, so when this went into effect, companies panicked as the tax steamroller ran over them.

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… that’s one of them, as I’m sure you’ve figured out. The delayyed start meant that it would not begin affecting the budget until 2022, but it helped the TCJA appear “deficit neutral” over the 10-year window used for legislative “scoring”.

The delay wasn’t legally required, mind you, it was a way to game the system. These kind of political tactics are commononplace in tax legislation. Phase-ins and or delayed-start provisions let lawmakers game how the Congressional Budget Office (CBO) - Congress’ nonpartisan analyst of how bills impact budgets and deficits - “grades” legislation, because it kicks the costs down the road, outside the “official” forecasting windows. Those of you who play table-top games or RPGs would call this a way to “cheese the system” or the traditional euphemism used to be something along the lines of “robbing Peter to pay Paul”.

Many businesses expected Congress to repeal or delay the Section 174 changes before they took effect in 2022, as there definitely was bipartisan support for immediate expensing. However, inaction led to a “shock” when 2022 tax bills arrived in 2023, forcing rapid cost-cutting, including layoffs. Small software firms, in particular, faced “extinction-level” tax bills, with some reporting taxable income tripling overnight, prompting layoffs or salary cuts. And bigger firms - Amazon, Meta/Facebook, Alphabet/Google, etc, Microsoft, Salesforce, etc - have had widespread layoffs in the US and have moved jobs overseas. Twilio cut 22% of its domestic workforce in 2023. Shopify cut 30% (they’re based in Canada, but much of their R&D was in the US - guess what, it isn’t anymore). Coinbase cut 36% of their team and there are still a heck of a lot of crypto bros, so I think they are probably not in a doomsday situation.

Now, I don’t want to say this was the only thing in 2023 that did this. There was a lot of economic turmoil on the horizon then: rising interest rates, reduced venture capital funding, supply chain problems, and post-pandemic over-hiring corrections, all amplifying financial pressures at the time. Companies like Meta announced layoffs during their “Year of Efficiency” in 2023, partly due to these tax changes and corresponding changes in ad spending. While not the sole cause, the Section 174 change drove (or accelerated) layoffs that otherwise would probably have been unnecessary.

Congress has made noise previously about a bipartisan reversal of this tax code change, and I rather hope they do - it would be a welcome boost to many sectors of the American economy, from manufacturing to pharmaceutical to technology to education to electrnics to scientific research and consulting.

Don’t think of this as just a problem for the tech space. I mean, the tech sector is ridiculously dominant for the S&P 500, the “Magnificent Seven” are a third of the value of the S&P 500. (That’s Alphabet, Amazon, Apple, Nvidia, Microsoft, Meta, and Tesla.) But if this does get repealed - it will be very good for those seven stocks - this is not an official stock tip, shush, you SEC guys.

Throughout the 2010s, a broad swath of startups, direct-to-consumer brands, and internet-first firms… heck, basically every company that you would recognize from Instagram or Facebook ads… built their growth models around a kind of synthetic carefully engineered break-even. Lot of online firms, a lot of handheld or wearable tech firms, personal entertainment devices, ride-hailing firms, anything self-driving, all the recent buzzy things.

The tax code allowed them to spend aggressively on product and engineering, then write it all off as R&D, keeping their taxable income close to zero by design. It worked because taxable income and actual cash flow were often not quite the same thing under what’s known as GAAP accounting practices. Basically, as long as spending counted as R&D, companies could report losses to investors while owing almost nothing to the IRS. In short, it costs a lot to invent - and market - the future. Building a better tomorrow can be expensive! Investors generally bought into this, gave them another round of venture capital, and let them defer a public offering. (This is actually another problem with this model - companies have stayed private far too long - but I’ll address that at some other point.)

But the Section 174 tax change absolutely cratered that model. Once those same expenses had to be spread out, or amortized, over multiple years, suddenly you couldn’t write these off - basically, the tax shield vanished or the accounting rules changed, depending on how politely you want to phrase it. But the mechanics of it are the same: companies that were still burning cash suddenly looked profitable on paper, triggering real tax bills on imaginary gains. (If this reminds you of some of the past economic crises - it should - this is one of the things that burned people back in the dotcom crash of 2000 amongst other mark-to-market problems in 2008, though the 2008 crisis mostly wasn’t this.)

The logic that once fueled a generation of digital-first research-focused growth ran straight into a IRS-shaped brick wall and put a mighty dent in the work force. If you were already public and profitable - well, your management team wasn’t going let you suddenly become unprofitable just because of a tax law change, so the answer was “cut expenses” and that mostly became “slow CapEx on data centers - servers are expensive - and lay off employees” so as to preserve profit margins and keep the stock price high. After all, reasoned management, if we had to put up with these crazy new rules, we basically just had to bank up R&D credits for a few years and then we were back to par. It was a couple of years of drag on the economy - and a particularly bad time to fall behind on technology leadership or a chance to reshore manufacturing or stabilize the American economy. One suspects that the current administration, if they had noticed that, might have kicked out a repeal or other clever plan as part of their new budget package (what’s currently being bundled together under the slightly goofy name of the Big Beautiful Bill) to juice economic recovery.

So it wasn’t just tech experiencing effects. From 1954 until 2022, the U.S. tax code had encouraged businesses of all stripes to behave at least a little bit like we think of tech companies behaving, by which I mean “investing in R&D” and more generally “investing in software” for the latter half of that. From retail to logistics, healthcare to media, if firms built internal tools, customized a software stack, or invested in business intelligence and data-driven product development, they could expense those costs - and the IRS generally agreed, which is slightly miraculous. The write-off incentivized in-house builds and fast growth well outside what people generally call the “tech sector”. For comparison, check the OECD research showing that immediate deductions foster innovation more than spread-out ones.

And American companies loved that logic and invested according to it. According to government data, U.S. businesses reported about $500 billion in R&D expenditures in 2019 alone, and almost half of that came from industries outside traditional tech. The Bureau of Economic Analysis estimates that this sector, the broader digital economy, accounts for another 10% of GDP. Probably 20% if you count Big Tech. And there’s a secondary market - all the people who support and are downstream from those workers and those industries (see below); we’re actually introducing relatively-avoidable friction into about a quarter of the American economy here with this particular tax change.

The result? A tax policy aimed at raising short-term revenue basically defined the growth engine for a huge chunk of American companies. But when that rug got pulled out, it also yanked out the incentive for hiring American engineers or investing in American-made tech and digital products. It made building tech companies in America suddenly not economically viable, and the last time we did this with a stupid policy change (Fed policy tightening “to defeat the Wealth Effect” and cool down speculative fervor … yup, it sure did! Lots of wealth went away, was that a good idea?) caused the big Dot Com Crash in 2000. It would probably be a good idea to put this tax credit back in place (or some facsimile thereof) before it continues to strip jobs out of the American economy - let alone all the folks (realtors, contractors, restaurants, nannies, tutors, personal trainers, et al) that are one step downstream from the tech sector jobs.