In brief:
- Nvidia hasn’t had a meaningful layoff since 2008, including through the 2022-2023 wave that cut thousands at Meta and Microsoft.
- Big tech stocks did rise an average of 5.6% in the month after 2023 layoff announcements, but that’s not proof the layoffs caused it.
- A 2024 study of 37 tech firms found operating performance actually fell in the quarter layoffs were announced.
- A 40-year meta-analysis of 114 studies found “little long-term gain” from headcount cuts overall.
The Business Case for Not Cutting Heads: What the Numbers Actually Say
Every layoff announcement comes with a similar line: “we had no choice.” If you look at the numbers, the story is different, at least in tech.
Let’s start with an example that’s the hardest to dismiss. Nvidia hasn’t made a meaningful layoff since 2008. It didn’t touch the headcount during the 2022-2023 wave that hit Meta (8,000 cut) and Microsoft (4,800 cut). It grew instead, while sitting at the centre of the most competitive, fastest-moving segment in the industry.
I know the objection: Nvidia’s rise to a roughly $5.3 trillion valuation by April 2026 (Forbes) is about chip demand, not restraint on layoffs. That’s a fair statement. The AI boom drove the stock, not a no-layoff policy. But that objection actually strengthens the point I want to make. If competitive pressure forced companies to cut, Nvidia was exactly where we’d have expected to see it happen first. Yet it didn’t.
The stock pop isn’t evidence of what caused it
Layoff announcements in 2023 were followed by some real stock gains. Big tech stocks rose an average of 5.6% in the month after the cuts were announced (Money.com). Alphabet climbed 15%, Amazon 19%. And Meta nearly 50%.
It’s really tempting to look at that as the market rewarding discipline. I don’t think that’s what really happened. 2023 was also the year the AI rally lifted every large tech stock at once, and even coverage of the layoffs at the time asked the causal question directly rather than answering it: were the layoffs actually responsible for the turnaround, or would these stocks have risen anyway? Correlation in the same calendar year isn’t automatically equivalent to causation, especially when a sector-wide rally is running underneath everything at the same time.
The operating numbers tell a different story than the stock ticker
If the stock price is the visible signal, the operating numbers are what’s actually underneath it. A 2024 study of 37 large US tech companies covering Q1’20 through Q1’23 found that investment and operating performance actually dropped during the quarter each company announced layoffs (Journal of Economics and Finance, Springer). The uptick that eventually followed was marginal and it came while financing and investment activity were still contracting.
That single-sector observation is aligned with a much larger picture. A meta-analysis conducted in 2024 pooled 905 effect sizes from 114 downsizing studies spanning roughly 40 years. Its conclusion was: “little long-term gain” from headcount cuts (Frontiers in Behavioral Economics). It was not along the lines of: “sometimes it backfires.” Little long-term gain, full stop, across four decades of research.
The pattern is consistent, not coincidental. Tech companies that lay people off tend to see operating performance dips in the short term, headline-friendly stock bumps that aren’t clearly tied to the layoffs themselves, and (if the 40-year meta-analysis is representative) no reliable long-term financial gains to show for it later.
So what are layoffs actually buying?
Based on the numbers above, an honest answer to this question seems to be: a short-term stock reaction that’s more likely explained by broader market conditions than by the cut itself, a real dip in operating performance in the announcement quarter, and no reliable long-term financial upside once the dust settles.
That’s not an argument that companies should never reduce costs. It’s an argument against headcount being reached for as the default lever ahead of alternatives, without solid numbers actually backing up the assumption that it’s a solution that works.
This is also where RETAIN’s second pillar, “Exhaust the Ladder Before the Exit”, stands. It isn’t a values-based objection to layoffs. It’s a sequencing argument: hiring freezes, vendor reduction, furloughs, tiered pay cuts, and voluntary buyouts should all precede involuntary layoffs, precisely in that order. Why? Because each one of them is usually reversible and cheaper to roll back than a severed employment relationship. If the operating performance data above is right, the sequence isn’t just kinder. It’s the financially logical order to test levers in too.
The question worth asking your own company
Before accepting “we had no choice” as the end of the conversation, consider asking two questions related to the numbers above:
- What did the ladder of cheaper, reversible solutions actually look like before headcount was impacted, and were they all genuinely tried first?
- Is the case for the layoff based on operating performance or on a stock reaction that a rising sector would likely have produced anyway?
Next in this series: RETAIN itself, the six-part framework that this data is building toward.
Sources
- Nvidia’s no-layoff history and 2022-2023 headcount growth: Seoul Economic Daily, August 2026
- Nvidia’s valuation: Forbes, April 2026
- 2023 post-layoff stock reactions: Money.com
- 2024 study of 37 large US tech firms: Journal of Economics and Finance (Springer)
- 2024 downsizing meta-analysis (905 effect sizes, 114 studies): Frontiers in Behavioral Economics
