The Missing Generation

The hiring landscape in the 2026 tech field is one of the most ruthless and damaging job markets since the 2008 employment crisis. This, is a breakdown of the how and why the software engineering field, on every level from junior to senior, needs to be fixed, and soon.

Entry-level software engineering postings in the US dropped 67% between 2022 and 2026. That figure comes from multiple independent analyses tracking job board data, and it's not the part of this story that should actually worry you. The part that should, is the fact that job postings labeled "entry-level software engineer" actually grew by 47% between October 2023 and November 2024, while actual hiring into those roles fell 73% over the same window of time.

Companies are advertising junior openings at a growing rate and filling them w/experienced engineers instead. That's not a hiring freeze, but something closer to a bait-and-switch happening at industry scale, and almost nobody entering the field right now realizes it's happening to them specifically.

The Numbers, Plainly

I'm going to start w/the most credible figure in this entire discussion, because it's the one measuring something real rather than something aspirational. Note that, from here, I'll also be using a lot of bullet points instead of just talking here, but I promise it's for a good reason. I have a lot of statistical points to make clear to your brains, so please don't skip the "boring" stuff.

Side Note: I've been working on this article for a little bit now, and when I tell you this stuff is important, I genuinely mean the fact that we should all be talking about these issues, and working on some actual fixes in our (once-beautiful and respected) tech field.

Stanford's Digital Economy Lab, working directly w/the ADP payroll records covering millions of actual workers, found that: employment for software developers aged 22 to 25 fell nearly 20% from its late-2022 peak — and kept declining into 2026. Mind you, this wasn't some survey like Family Feud asking random people how they feel about the damn job market. It's an actual count of actual paychecks…stopping dead.

A few more figures worth having on hand, because the pattern holds up across every source measuring it independently. Yes, this is the boring, but important part:

• SignalFire's State of Talent Report 2025 — found entry-level hiring at the 15 biggest tech companies dropped 25% year-over-year from 2023 to 2024 alone (that's only 1 fiscal year).

• The junior share of overall tech hiring — the percentage of new hires who are entry-level rather than experienced — dropped from roughly 15% to 7% over 3 years, according to multiple 2026 industry analyses. For every 100 new engineers a company hires today, roughly 7 are junior devs. 3 years ago, it was closer to 15.

• New-grad hiring at "Big Tech" in specific — cratered from about 32% of hires in 2019 to roughly 7% by 2026 (not a joke). This collapse actually predates most of the AI-tool adoption curve (vibe coding drama), which matters for the causal argument coming later on in this article, you'll see what I mean.

• CompSci graduates now face 6.1% unemployment — with computer engineering grads at 7.5% — both notably higher than several non-STEM majors (according to Federal Reserve data). That's the actual, uncomfortable rebuttal to "just get a CS degree and you'll be fine!", or "just go learn to code, everybody can do it!". A piece of advice that was actually once reliable for over a decade and has quietly stopped being true.

• UK entry rolls are hit as well — in the UK, entry-level tech roles fell to 46% in 2024 alone (a wide variety, as a whole, by the way), w/projections putting the decline at 53% by the end of 2026. This isn't just a US phenomenon, and I'm sure if you looked elsewhere, you would find the same issue starting to bleed over across the world.

The Bait-and-Switch Nobody's Talking About (For Some Reason)

Here's the mechanism behind that whole 47% postings-growth vs. 73% actual-hiring-decline gap, because it's the most important and least discussed part of this whole story, at least in my 'old man shouts at cloud' opinion. Companies haven't stopped writing job listings that say "entry-level" or "junior" in the title. They've largely stopped meaning it when they have their HR department write it in the description, along with the myriad of unnecessary requirements for the job itself. A posting gets labeled junior for a mix of reasons that have nothing to do w/an actual willingness to train someone to be in this field:

• broader applicant pools

• lower advertised salary bands

• willingness to train someone

• expected role and department work

• complete misunderstanding of role requirements (truly)

And then the job gets filled quietly, by an experienced engineer willing to accept a junior title in an exceedingly tight market, or simply sits open far longer than a genuinely urgent role would, or never gets filled at that level at all.

Layer onto that a genuine and well-documented shift in what "entry-level" is now expected to mean on arrival. One 2026 industry analysis put it plainly which is just about one of the saddest and stupidest things I've heard in a while:

"The 2026 Junior Candidate is now expected to arrive with system-design understanding that would have counted as mid-level knowledge in 2020."

Combine higher expectations w/a shrinking number of real openings (not 'ghost jobs', that's a whole other topic for now), and you get exactly what's happening on the ground for people actually trying to retain a career in this field, let alone the ones attempting to break in right now:

• Automated resume screening rejects roughly 70% of applications before a human ever sees them.

• A typical process now runs 5 to 7 stages (ew) — resume scree, algorithmic assessment, TCI's (test coding interview), a multi-day take-home project, multiple interview rounds — consuming 15 to 25 hours of unpaid candidate time (across 4 to 8 weeks!) for just a single applicant.

• Candidates are routinely applying to dozens of roles in parallel just to generate a handful of real conversations, because the volume math has gotten that unfavorable. Because why not, right?

None of this reads like a temporary correction that self-resolves once the broader economy actually stabilizes, but like a structural redefining of what the first rung of the engineering 'career ladder' requires to reach at all.

Who's Actually Responsible: AI, Economics, or Both

The dominant narrative in most coverage of this is simple: AI did it. AI coding assistants now handle a meaningful share of exactly the tasks junior engineers used to cut their teeth on. Things like bug fixes, boilerplate code, review and analyze legacy code, scaffolding, routine implementation work. These are just a few of some important things companies stopped needing bodies to do it. It's a clean story, and not entirely wrong. But it's also not the whole picture here, and intellectual honesty means presenting the actual disagreement instead of picking the more viral half of it.

Scarily enough, several of these independent analyses out there attribute a meaningfully smaller share of the collapse directly to AI than the popular framing suggests. One detailed breakdwon put AI's direct causal contribution at roughly 10%, w/the larger share of the decline traced to a broader economic correction: the end of the zero-interest-rate era that had funded years of the growth-at-all-costs hiring sprees most of us know well, tighter capital generally, and a hard pivot across the industry from "hire ahead of need" to "hire only against immediate provable ROI". Well under that kind of framing, AI didn't cause the collapse so much as it gave companies already looking for ways to cut costs a clean (and plausible), forward-looking justification for a decision that budget pressure from the higher-ups was already pushing them toward. And always is, let's be honest.

The blunt version of this economic argument stated openly by way more than just 1 company:

A junior engineer typically needs 6 to 12 months of ramp-up before contributing meaningfully to the company, while an AI coding tool costs roughly $20 to $30 per/month starting on day one.

Framed as a straightforward cost comparison rather than an investment in future capability, that math doesn't favor the junior — and a lot of hiring decisions in 2025 and 2026 got made on exactly that scope, w/o much visible consideration of what the comparison misses entirely, which is the actual subject of this next ranting and raving section. Sorry not sorry in advance, but somebody needs to give these people a damn voice, because Lord knows nobody is actually listening to them.

The Missing Generation Problem

Here's the part of my story that turns it from a rough patch for new grads into an industry-wide structural risk: senior engineers do not spontaneously generate. Every senior engineer working today was (at some point) a junior that somebody took a chance on, ran through a genuinely uneven first year or two, and mentored into competence. If a whole multi-year cohort doesn't get that chance, hired, ramped, and mentored at any realistic scale, the industry doesn't just have a rough couple of years for new grads or the self-education starters wanting to get into this field. It creates a gap in the pipeline that shows up 5 or 10 years later, exactly when that missing cohort would have been reaching the senior and staff level roles the rest of the industry depends on! Downturn-driven hiring freezes have produced this exact delayed effect in past economic cycles. The pain doesn't arrive when the freeze actually happens, it shows up later, when there's a hole in the experience curve right where the industry needs it filled. But by then, it's too late for all players involved.

Not every company is making the same bet, and the disagreement is worth featuring directly rather than flattening into a single doom narrative that we all constantly see being preached.

IBM tripled its junior dev intake in 2026 under Nickle LaMoreaux (she was the CHRO for IBM but now I think she's the Senior VP if I read her IBM page right, or both I may just be stupid.), deliberately restructuring the roles rather than eliminating them (get this person a freaking medal!).

Juniors at IBM now spend less time on stuff like routine implementation and more on customer interaction and reviewing AI-generated output, w/oversight built directly into how the role is fully designed. Her own reasoning is worth stating plainly: the companies that will be most successful 3 to 5 years out are the ones that kept investing in entry-level hiring through exactly this environment, not the ones that cut it (like most of them do anyway).

It's a genuine bet against the prevailing short-term statistics honestly, made explicitly because someone in a position to see the industry-wide pattern decided the missing generation risk was real enough to act against.

Where entry-level hiring is actually holding up, for anyone trying to orient inside this market right now:

• Cloudflare, Shopify, and OpenAI all continue meaningful junior and intern hiring at real volume, even as many of the similar companies cut back…dramatically, mind you.

• Enterprise software vendors and companies w/decade-old codebases still need people willing ot learn genuinely complex internal systems that AI tools have far less training exposure to than they do to greenfield, well-documented public code.

• Non-tech industries adopting real engineering headcount — healthcare, fintech, gov departments — remain comparatively strong, since they're earlier in their own hiring curve for this kind of talent than Big Tech is.

• Meanwhile, Salesforce halted any and all hiring entirely in 2025, while some other big names like Microsoft, Google, and Meta quote, "…scaled back job postings, restructured specific teams to fund artificial intelligence, and scrapped aspirational diversity hiring goals." They're hiring roughly HALF as many new grads as they were in 2021. The split between "doubling down" and "cutting to zero" is real and it's happening company by company, not industry-wide in sync w/each other.

What Does This Actually Mean

Look. There's a lot to all of this, I realize that, you realize that. If you're trying to break into this field right now, the honest read isn't that it's impossible, but its the fact that the on-ramp changed shape, and pretending it's still 2019 will 100% waste your time. Where you aim matters more than it used to. A generic "I learned React! Hire me now please for $138K w/full benefits!" application is competing directly against a rapidly shrinking pool of postings that mean what they say, while companies restructuring the entire junior dev role (mainly solely) around AI-output review and real client exposure (following the IBM model) are quietly signaling what the next version of "entry-level" is actually going to look like once this correction settles, God willing it does.

For everyone else…

The mid to senior engineers (staff engineers are just fine, wherever they are), hiring managers, people who aren't personally affected by this today, well the honest takeaway is less comfortable.

This isn't a story that resolves itself quickly and quietly on its own timeline. It's a bet, being made right now, company by company, mostly on short-term cost math and statistics; about whether there will be enough experienced engineers in 5 years to run the actual systems that everyone's still going to depend on. Some companies are making that bet deliberately, w/a real argument behind it. A lot more seem to be making it by default, one quietly unfilled junior posting at a time, w/o much visible thought for the bill that comes due once the missing generation was supposed to have grown up.

And I can tell you, from deep inside the tech field, if these issues aren't solved, and soon, at least a plan…

We're all going to feel the brunt of this hit, and the reality is, it won't be the companies taking said hit, it'll be us. The devs, designers, and engineers, fighting for a chance to prove our worth out there.

Best of luck, everyone. We all need a bit of luck nowadays. 🍀

Research and Links

Sources referenced throughout this piece, for anyone who wants to go deeper into the topics. These are things discussed in the article, as well as just research for more information.

• Stanford Digital Economy Lab — Canaries Dashboard — the ADP-payroll-based research (Brynjolfsson, Chandar, and Chen) tracking real employment, not job postings, for workers aged 22–25 in AI-exposed occupations. The single most credible dataset in this piece, and the one I'd point a skeptical reader to first.

• SignalFire's State of Talent Report 2026 — hiring data across roughly 80 million companies, the source for the entry-level collapse figures at major tech firms and startups.

• Federal Reserve Bank of New York — The Labor Market for Recent College Graduates — the ongoing, continuously updated tracker behind the CS/CE unemployment figures. Worth bookmarking directly; it updates quarterly.

• Entry-Level Job Crisis 2026 — Grindhotline — the source for the 67% posting-decline figure, the UK data, and the LinkedIn "entry-level in name only" analysis. This piece also cites a Harvard study (Hosseini and Lichtinger) worth a direct look if you want the deepest causal dive available on this topic.

• IBM Tripling Entry-Level Hiring — Yahoo Finance — the on-the-record reporting behind the IBM/Nickle LaMoreaux counter-example.

• Indeed Hiring Lab — The Labor Market Is Tilting Toward Seniority — a good, more granular breakdown of how the entry-to-senior posting ratio has shifted by industry, not just in tech.

• NACE Job Outlook 2026 — the National Association of Colleges and Employers' annual survey of actual employer hiring plans for new graduates, a useful reality check against job-board data alone.

// EOF