career

The Job Market in 2025: Year-End Reality Check

The official 4.6% unemployment hides a darker truth. When you count contractors, discouraged workers, and the H-1B exodus, the real number approaches 9%. Here's what a full year of data shows.

Chris Groves

21 Dec 2025 — 8 min read

What if the unemployment rate everyone quotes is only half the story?

The official number says 4.6%. The broader measure - counting everyone the standard U-3 rate excludes - is 8.7% (U-6).

2025 didn't just shake the job market. It exposed the gap between headline metrics and lived reality.

As the year closes, we have the full picture: 1.17 million announced job cuts through November - the most since the pandemic. Federal workforce reduced by 317,000 positions. Tech layoffs that, while lower than 2024, still eliminated 122,000+ workers. Contractors cut without appearing in any headline. H-1B workers receiving deportation notices within their 60-day grace period. And 514,000 Americans so discouraged they've stopped looking entirely.

This isn't a recession. It's a restructuring. And the gap between official metrics and reality is significant.

The Numbers They Report vs. The Numbers That Matter

Let me start with what you'll see in headlines, then show you what's actually happening.

Official unemployment (U-3): 4.6% in November 2025 - the highest since September 2021 (excluding pandemic). The labor market added just 64,000 jobs that month. The number of unemployed Americans hit 7.1 million.

Real unemployment (U-6): 8.7% in November - up from 8.0% in September. This includes everyone the U-3 conveniently ignores: part-time workers who want full-time, marginally attached workers, and those employed below their skill level.

The hidden workforce: 514,000 discouraged workers have stopped looking entirely. Another 1.8 million are "marginally attached" to the labor force. 6.4 million want jobs but aren't actively searching.

That's nearly 9 million people the headline number pretends don't exist.

The hiring collapse: Through November, employers announced just 497,151 planned hires - down 35% from 2024. It's the lowest year-to-date total since 2010.

The 1.17 Million: Breaking Down the Cuts

Challenger, Gray & Christmas tracks layoff announcements. Through November 2025: 1,170,821 announced job cuts. That's 54% higher than the 761,358 announced in the same period of 2024.

It's only the sixth time since 1993 that announced cuts through November have surpassed 1.1 million.

Where the cuts came from:

Reason	Job Cuts	Notes
DOGE Impact	293,753	Direct federal workforce + contractors
DOGE Downstream	20,976	Loss of federal funding to private/nonprofit
Store/Unit Closings	178,531	Retail, branches, departments
Restructuring	128,255	Corporate reorganization
AI	54,694	Explicitly AI-attributed
Tariffs	~8,000	Trade policy impact

October alone saw 71,321 cuts - the highest October total in 22 years.

The DOGE Exodus: 317,000 and Counting

The Department of Government Efficiency (DOGE) initiative, led by Elon Musk, drove the largest peacetime federal workforce contraction in decades.

The timeline:

February 14, 2025: The "Fork in the Road" buyout offer. More than 150,000 federal employees accepted deferred resignation.
July 8, 2025: Supreme Court overrides lower court orders that had frozen firings, allowing reductions to continue.
August 26, 2025: Partnership for Public Service reports just under 200,000 federal workers have left.
December 2025: Bloomberg reports 317,000 total departures - about 1 in 8 civilian workers.

The Office of Personnel Management confirms: federal employment fell to 2014 levels. For context, 115,900 workers left in all of 2023.

Courts ruled many firings illegal. The Supreme Court let them continue anyway.

Tech Layoffs: The Third Consecutive Year

Different trackers show different totals, but the picture is consistent:

Layoffs.fyi: 122,549 tech employees across 257 companies
TrueUp.io: 209,838 people impacted across 716 layoff events
Crunchbase News: 2025 layoffs have already surpassed 2024's 95,667

The monthly breakdown tells the story: October hit 18,510 (the highest month), followed by July at 16,327, May at 10,397, and November at 8,932.

Notable December cuts: The Trade Desk, McKinsey (10% of non-client roles), Amazon (84 in Seattle/Bellevue).

The silver lining: tech layoffs are approximately 20% lower than 2024 - the lowest in half a decade. Stabilization may be beginning.

The H-1B Crisis: 60 Days or Less

For H-1B workers, layoffs mean something far worse than job loss. The rules give 60 days to find a new employer willing to sponsor your visa. Miss the deadline? Deportation.

But in 2025, even that grace period isn't guaranteed.

USCIS expanded the circumstances triggering Notices to Appear (NTAs) in a February 2025 policy memo. The result: deportation notices issued within the 60-day window, sometimes within two weeks.

A Blind poll in August showed one in six Indian H-1B workers - or someone they know - had received an NTA within their grace period.

The regulation includes a clause granting DHS discretion to "eliminate or shorten this 60-day period." That discretion is now being exercised aggressively.

Microsoft, Amazon, Meta, Google - they all appear in the top 50 H-1B employers. They're also all on the layoff trackers.

The message is clear: the talent pipeline that built Silicon Valley is being systematically restricted.

The AI Factor: 76,000+ Jobs and Counting

In 2025, 76,440 workers lost their jobs explicitly to AI - approximately 491 per day.

Challenger, Gray & Christmas attributes 54,694 layoffs to AI through their tracking. The difference reflects methodology - some cuts cite AI as a contributing factor rather than the primary cause.

But here's the nuance everyone misses:

The age gap is brutal. Unemployment among 20-30 year-olds in tech-exposed occupations has risen by almost 3 percentage points since early 2025 - notably higher than same-aged workers in other fields.

The paradox: According to the World Economic Forum's 2025 Future of Jobs Report, 41% of employers worldwide intend to reduce their workforce in the next five years due to AI. Yet AI-related roles are among the few still growing.

The corporate reality: 44% of companies say employees will "definitely" or "probably" be laid off due to AI in 2025 - up from 37% in 2023.

The ladder's bottom rungs are being removed while the top expands.

Software Engineering: Bifurcation in Real Time

Software engineering job postings are at just 65% of February 2020 levels - a five-year low.

The Indeed Hiring Lab confirms: tech postings stood 36% below February 2020 as of mid-2025. The hiring freeze has entered its third consecutive year.

The breakdown by role is stark:

Software engineers overall: down 49% from 2020
Android, Java, .NET, iOS developers: down 60%+
Machine learning engineers: up 59% from 2020

Entry-level tech job postings are down 34%. Senior roles are down 19%. At startups, under 6% of hires are new grads.

The market isn't dead. It's splitting in two.

AI-powered coding tools like GitHub Copilot have enabled 30% productivity increases at companies like Salesforce - while maintaining flat headcounts. Approximately 75% of engineers now use AI assistance.

The result: smaller teams doing more work. Fewer entry points for new talent.

Consumer Confidence: The Canary

Gallup's Economic Confidence Index fell to -30 in November - a 17-month low, the most negative since January 2021 during the pandemic.

The outlook is bleak:

57% expect the economy to weaken - the least optimistic since Deloitte started tracking in 1997
63% say it's a "bad time" to find a quality job - the most negative since January 2021
Americans' holiday gift budgets collapsed by $229 since October - the steepest mid-season cut in nearly two decades

Holiday spending tells the story:

Average planned spending: $1,595 - down 10% year-over-year
Gen Z pulling back 34% year-over-year
Millennials: down 13%
Only Gen X increased spending (+3%)

Yet here's the paradox: Consumer spending keeps driving GDP. The National Retail Federation projects record holiday sales past $1 trillion.

The explanation: Nearly all inflation-adjusted spending growth is concentrated in the top 20% of income earners - many supported by AI-related stock gains. Everyone else is trading down to Walmart and T.J. Maxx.

The economy isn't fine. It's bifurcating along the same lines as the job market.

What the Data Actually Shows

Add it together:

Official unemployed (U-3): ~7.1 million
Underemployed/part-time for economic reasons: ~4.3 million
Marginally attached: ~1.8 million
Discouraged workers: ~514,000
Contractors who lost work (uncounted): Unknown millions
H-1B workers in 60-day limbo: Tens of thousands

The 4.6% headline is a convenient summary. The reality is closer to 9% - and that's before counting contractors and gig workers who simply stopped getting calls.

This isn't AI's fault alone. The economy is restructuring. Technology is accelerating it. But the fundamental shift is corporate: maximum output, minimum human cost.

The Mindset Shift Required

The traditional model is breaking:

Value determined by employer, not skills
Security from job title, not capability
Success means climbing someone else's ladder
Worth tied to how cheaply you can be replaced

That paradigm served its purpose. It's dying now.

What replaces it isn't clear. But waiting for corporations or governments to solve this would be foolish. They're the ones doing it.

What You Can Do

If you're worried about your position - good. Worry is appropriate. Panic is useless.

Build capability, not credentials. Another certification won't save you. Demonstrable skills that solve real problems will.

Work with AI, not against it. The bifurcation is real. 75% of engineers now use AI assistance. People who augment their work with AI have opportunities. People who resist become the jobs that get automated.

Build in public. Document projects. Write about what you learn. Help others. This creates reputation, relationships, and opportunities no resume matches.

Focus on what's hard to automate. Strategic thinking. Creative problem-solving. Human judgment in complex situations. Deep expertise in niche domains.

Diversify income sources. The 27.7 million full-time independents figured something out: multiple clients beats one employer who can fire you.

The Honest Assessment

2025 was the worst year for layoffs since 2020. AI adoption is accelerating. The job market for knowledge workers has structurally changed.

The official statistics hide more than they reveal. When you count everyone - discouraged workers, contractors, underemployed, H-1B workers in limbo - the picture is significantly worse than headlines suggest.

I think we're facing genuine economic disruption over the next 2-3 years. Not apocalyptic - but painful for millions more than the comfortable 4.6% number implies.

We're in transition. It's uncomfortable. Many people will experience hardship.

But we're not powerless.

We can build alternative models. Share knowledge instead of hoarding it. Create value outside traditional employment structures. Develop capabilities that matter regardless of employer.

That's what we can build together - systems where humans and AI work together, where knowledge is accessible, where capability matters more than credentials.

Sources

Official Unemployment Data:

Layoff Trackers:

Tech Layoffs:

DOGE/Federal Layoffs:

H-1B Impact:

AI Impact:

Software Engineering Market:

Consumer Confidence:

Discouraged Workers: