Tech Layoffs: Panic or Pattern?

May 19, 2026

The headlines have been relentless. Google cuts 12,000 jobs. Meta drops 21,000. Amazon, Microsoft, Salesforce, Spotify — the list of household names shedding staff has been long enough to convince a lot of people that tech is broken. That the golden era is over. That maybe this is a bad time to be trying to get into the industry.

Is it though?

Because when you go past the headlines and look at what twenty years of employment data actually shows, the picture is a lot less dramatic. And a lot more interesting.

It has happened before. Several times.

Tech employment does not move in a straight line. It never has. It moves in cycles — sharp drops followed by surges that consistently exceed the previous peak. We have watched this play out four times in the last twenty-five years, and most people seem to forget the earlier chapters every single time.

Tracked tech layoffs during major downturns:

Period	Tracked layoffs
Dot-com bust 2001	~168,000
Dot-com bust 2002	~110,000
Great Recession 2008–09	~300,000
COVID flash crash 2020	~80,000
Tech winter 2022	~165,000
Tech winter 2023	~265,000
Tech winter 2024	~153,000
Tech winter 2025	~124,000

Sources: Computerworld / TechCrunch dot-com records; Layoffs.fyi 2022–2025; Fortune / CNBC analysis.

The dot-com implosion (2000–2003)

The first and still the most dramatic. Companies hired on pure speculation through the late 1990s. When the Nasdaq lost 78% of its value, the reckoning came fast. In 2001 alone, tracked tech layoff announcements topped 168,000. Peak IT unemployment followed a year later, in 2002 and 2003, after most of the corporate failures had already happened. Tech workers at the time were 4.5 times more likely to face mass layoffs than workers in other sectors.

The financial crisis (2008–2009)

The Great Recession hit everything. Tech fared better than banking or manufacturing — software demand proved more resilient than physical goods — but the sector still shed hundreds of thousands of roles. The recovery came fast. By 2012, tech was growing faster than at any point since the 1990s.

The COVID flash crash (2020)

This one is the most counterintuitive. The pandemic is remembered in tech as a boom. And mostly it was. But the first months of 2020 brought a real shock, with tens of thousands of layoffs across startups and travel-adjacent businesses. Then digital demand exploded and companies pivoted from cutting to hiring at unprecedented speed. The flash crash became the foundation for the biggest hiring bubble the industry had ever seen.

The tech winter (2022–2025)

Which is where we are now. Interest rates rose. Venture capital dried up. The valuations that had justified those massive headcounts collapsed. Companies that nearly doubled their staff between 2019 and 2022 suddenly found themselves badly overstaffed for a slower-growth world. The correction arrived in waves: roughly 165,000 layoffs globally in 2022, around 265,000 in 2023 (the worst single year since the dot-com crash), then easing to approximately 153,000 in 2024 and 124,000 in 2025.

2023 was the second-biggest year of tech layoffs on record. But it followed the biggest hiring binge the sector had ever seen. You cannot understand the correction without understanding the excess that came before it.

Nobody talks about the hiring binge.

Here is the context that almost every layoff headline leaves out.

Between 2019 and 2022, major tech companies grew their workforces at a pace that had no historical parallel. Remote work removed geographical hiring limits. Pandemic-era digital demand looked permanent. Capital was essentially free. Some employers nearly doubled their headcount in three years.

Some large tech firms nearly doubled their headcount between 2019 and 2022
US tech employment peaked at ~3.1 million in early 2023 — the highest ever recorded
Tech employment grew at an average 3.1% per year between 2010 and 2020

The layoff wave of 2022–2024 looks terrifying in isolation. Measured against the binge that preceded it, it starts to look like a painful but not irrational correction. Many of the workers laid off in 2022 and early 2023 had been hired in 2021 and 2022. The sector did not shrink below pre-pandemic levels. It corrected back toward them.

What does the BLS actually say?

This is where it gets interesting.

The Bureau of Labor Statistics Job Openings and Labor Turnover Survey — JOLTS — is the most rigorous longitudinal dataset on US labour market flows. It does not rely on press releases or tracker websites. It measures actual layoff rates across every major industry. And when you look at it, the story it tells about tech is not the story LinkedIn has been telling.

Annual average layoff rates by industry, 2020–2024 (% of employment):

Industry	2020	2021	2022	2023	2024
Information (Tech)	2.0%	1.0%	1.2%	1.1%	1.1%
All industries (avg)	2.4%	1.0%	1.0%	1.1%	1.1%
Construction	3.7%	2.1%	1.9%	2.2%	2.0%
Leisure & Hospitality	6.6%	1.3%	1.3%	1.4%	1.5%
Manufacturing	1.9%	0.8%	0.8%	0.9%	0.9%
Health Care	1.5%	0.6%	0.6%	0.7%	0.7%

Source: U.S. Bureau of Labor Statistics, JOLTS Table 24 — Annual average layoffs and discharges rates by industry, not seasonally adjusted.

The BLS classifies technology inside the "Information" sector. In 2022 — the year tech layoffs dominated every business outlet — the Information sector's annual average layoff rate was 1.2%. The national average was 1.0%. Construction: 1.9%. Leisure and Hospitality: 1.3%. Arts and entertainment: 3.0%. Professional and Business Services: 1.6%.

In 2023, when 265,000 tracked global tech layoffs prompted analysts to call it the worst year since the dot-com crash, the Information sector's BLS layoff rate was 1.1%. That is exactly the national average. Healthcare came in at 0.7%. But nobody was writing breathless articles about healthcare because healthcare workers do not have millions of LinkedIn followers posting about being laid off.

The raw numbers from JOLTS Table 23 say the same thing. In 2023, the Information sector recorded approximately 399,000 layoffs in absolute terms. Retail: 1.88 million. Health Care: 1.78 million. Construction: 2.07 million. Tech generates enormous coverage per layoff. That coverage is not proportional to the economic scale of the disruption.

The Information sector's layoff rate in 2023 was 1.1% — the national average. Construction was at 2.2%. Leisure and Hospitality at 1.4%. What makes tech feel different is not the data. It is the visibility.

So why does it feel so bad?

Three things.

First: concentration. Tech workers are disproportionately on LinkedIn, Twitter and in the media. When a Google engineer loses their job, they write a thread that goes viral. When a retail worker loses theirs, nothing happens online. The signal-to-noise ratio on tech layoffs is extremely high.

Second: brand salience. The companies doing the cutting — Meta, Amazon, Microsoft — are household names. A 5% workforce reduction at a company most people have never heard of is invisible. The same cut at a company whose app is on every phone generates global coverage for a week.

Third: expectation. Tech workers had grown used to near-full employment and exceptional pay. When a sector that had been functionally immune to downturns starts shedding jobs, the psychological shock is out of proportion to the statistical reality. The contrast with recent experience makes things feel worse than they are by any historical measure.

The long view: tech still grows.

Every previous cycle — dot-com bust, financial crisis, COVID shock — was followed by tech employment that exceeded the prior peak. There is no data-based reason to assume the current one ends differently.

BLS occupational projections look through the cycle, not at the moment. Software developer employment is projected to grow 17% from 2023 to 2033 — "much faster than the average for all occupations", in the BLS's own words. Data scientist roles: 34% projected growth to 2034. Information security analysts: 29%. These are among the fastest growth rates across the entire US occupational landscape.

And current unemployment rates confirm the demand is real today, not just projected. Software developers sit at approximately 2.8% unemployment. Systems analysts: 1.8%. Security analysts: 2.3%. The national average is around 4.2%.

Looks pretty good, doesn't it?

What does this mean for you?

If you are considering a career move into tech, the noise of layoff headlines can feel like a stop sign. The data says it is more accurately read as a map — one that shows which roads are closing and which are opening.

The closing roads are fairly clear: mid-level roles that existed mainly because hiring was cheap, content and operations positions that got swept up in the pandemic binge, and anything AI can now do reliably on its own.

The opening roads are also clear. Full-stack development with AI fluency. Data engineering and analytics. Cybersecurity. Applied machine learning. The BLS projects these roles will add hundreds of thousands of positions per year through 2034, and the current unemployment rates in these areas confirm the demand is not theoretical.

Our bootcamps are built around where the hiring is actually happening. The Web/Mobile Development Bootcamp is a 9-week intensive — in Barcelona or online — covering JavaScript, React, React Native, Node.js, MongoDB and now AI-powered app development. The Deep Learning & AI Engineering Bootcamp is a 13-week online part-time programme for those who want to work in AI and ML directly. And the UX/UI Design Bootcamp prepares you for a field the World Economic Forum listed as #8 in the fastest-growing jobs for 2025. None of these are bets on a sector in decline. They are bets on where the sector is going.

The paranoia in tech circles right now is understandable. It came from a real shock to a sector that had convinced itself it was recession-proof. But paranoia is a poor guide to career decisions.

The data, looked at across twenty years, tells a more useful story. Tech employment is cyclical, volatile, and consistently higher after each trough than before it. The question is not whether tech is a viable destination — it demonstrably is. The question is whether you are building for the version of tech that existed three years ago or the one being assembled right now.

Skills do matter. Which ones you choose matters even more.