There's No Skills Shortage. There's a Wages-Too-Low Shortage.

When an employer says they can't find anyone with your skills, run the wage test before you believe them. In a genuine shortage, prices rise. Pay climbs fast to pull people in. When the role they "can't fill" hasn't seen real wage growth in years, that's not a supply problem. It's a price problem. They can find you. They won't pay you.

Here's the part nobody says out loud: the skills-gap story has been running for thirty years, and it has never once been fixed. Real shortages get fixed. Wages rise, more people train up, the gap closes. A story that runs forever and never resolves isn't a shortage. It's a narrative that's useful to keep alive.

What does a real shortage actually look like?

Supply and demand isn't complicated. When something is genuinely scarce and people want it, the price goes up until supply catches up or demand backs off. That's true for apartments, for oil, and for labour.

So the test for any "we can't find talent" claim is one question: are wages in that role rising faster than the cost of everything else? If a role is truly short-handed, employers compete for the few qualified people by paying more. Visibly more. Year over year.

If they're not paying more, they don't believe their own shortage. They've found a price they like and they're waiting for someone to accept it. That's not scarcity. That's a standoff, and they've decided you'll blink first.

Economist Katharine Abraham at the University of Maryland went looking for the structural skills gap across multiple recessions and found no evidence of one, despite policymakers treating it as a given since the mid-1960s. Paul Krugman called the skills-gap claim a zombie idea that should have been killed by evidence but refuses to die.

If wages tracked output, where would your pay be?

Start with the macro picture, because it sets the table for everything else. Between 1979 and 2025, net productivity grew 90.2% while typical worker pay grew just 33.0%. Workers got far more done. Most of the gains went somewhere other than their paycheques.

A fair flag: this number is contested. Critics argue the Economic Policy Institute compares net productivity against gross wages, which widens the gap, and a narrower measure shows a divergence closer to 23 points rather than 57. Fine. Take the most conservative version anyone defends. It still shows pay falling well behind output for decades.

That's the backdrop. When employers say there's a shortage of willing workers, they're saying it inside an economy that has spent forty years paying workers less than the value they produce. The "shortage" is what happens at the bottom of that long slide, when the offered wage finally drops below what people will accept.

Where do the shortage claims fall apart on contact?

Look at four roles everyone "knows" are short-staffed.

Nursing. The United States has 4.7 million licensed registered nurses, more than at any point in history. Yet hospitals declare a shortage while 100,000 nurses left hospital work in two years. Penn researcher Karen Lasater put it plainly: this is a retention issue, not a pipeline issue. The nurses exist. They trained. They got licensed. Then conditions pushed them out, and the hospitals called it a shortage instead of looking in the mirror.

Software and IT. In 1997 a staffing lobby claimed 190,000 unfilled IT jobs. By 2000 the figure had ballooned to 843,000, built on an NSF shortage study that was never published and was later proved false. Meanwhile real computer-programmer wages stayed flat from 2000 to 2023, and STEM hourly pay grew by all of five cents an hour in real terms between 2008 and 2023. Three decades of "talent shortage" headlines, near-zero real wage growth. A real shortage would have bid those wages up. It didn't.

Degree inflation. During the Great Recession, when one in six workers was idle, employers bolted college-degree requirements onto jobs that had never needed them. Modestino, Shoag and Ballance found that labour slack alone explained 18 to 25% of the rise in skill requirements, with the cyclical channel raising the share of postings demanding a four-year degree by roughly 2.2 percentage points and the share demanding two-plus years of experience by about 3.5 points. That experience floor is its own piece of theatre, because the "need experience to get experience" trap is a choice employers make, not a law of nature. As hiring recovered, those requirements quietly dropped again. The "skills gap" expanded and shrank with unemployment, not with any real change in what the jobs required.

The degree premium that buys nothing. Harvard Business School's study of 26 million postings found employers pay 11 to 30% more to degree holders doing identical middle-skills work, while rating them as nearly or equally productive as non-degree workers. The degree holders also quit more often. Employers are paying a premium for a credential that delivers no extra output, and in the same breath claiming they can't find anyone. Once you see how much of the degree requirement is theatre rather than need, the shortage story gets harder to take at face value.

So who cut the training that supposedly causes the gap?

The same employers who say workers aren't ready stopped paying to make them ready.

In the US, corporate training spending fell from more than 0.5% of GDP in 2000 to about one-third of a percent by 2013. In the UK, employer investment in skills dropped 19% per employee in real terms between 2011 and 2022, with the largest firms cutting 35%.

There's a fair counterpoint here. Training is a leaky investment. If you train someone and a competitor poaches them, you paid and they benefited. That's a real market failure, and it's an argument for apprenticeships and public co-funding, not just finger-pointing.

But it doesn't rescue the shortage story. It deepens it. Employers stopped building the skills they say are missing, then blamed schools and workers for the gap they created. The honest version is: we cut training, we want it cheap, and "skills gap" sounds better than "we'd rather not pay for it."

Where does the immigration angle fit?

It's the same pattern with a passport. Around 80% of H-1B workers land in the two lowest wage tiers, earning 20 to 40% below comparable American workers. A program sold to fill a "shortage of talent" is built largely to access cheaper labour for the same work.

Be precise about the limit of this point. At the economy level, skilled immigration produces real innovation and productivity gains. The wage-suppression effect is a firm-level, role-level reality, and it doesn't cancel the macro benefits. The argument isn't "close the borders." It's narrower and harder to dodge: if a role needs imported labour specifically because the imported version is cheaper, the domestic "shortage" was always a price you didn't want to pay.

Are any shortages actually real?

Yes, and being honest about which ones keeps the whole argument credible.

Some shortages pass the wage test. Genuinely new specialisations, where the trained population is legitimately tiny and pay is climbing hard, are real scarcity. Demographic crunches in some skilled trades, where decades of tacit knowledge walk out the door with retirees, are real. And there's a third kind worth naming: the lag shortage. When AI or automation shifts what skills employers need faster than people can retrain, you get a temporary mismatch that looks structural but resolves on a five-to-ten-year delay.

The tell is always the same. In a real shortage, pay is rising fast right now. In a manufactured one, the headlines shout scarcity while the wage line stays flat.

	Manufactured shortage	Real shortage
Wages over the last 3-5 years	Flat or barely tracking inflation	Rising clearly faster than inflation
Employer's first move	Add credential requirements, lobby for cheaper labour	Raise pay, sweeten conditions, train people up
Training investment	Cut or outsourced to workers	Funded, because the role is worth it
How long the "shortage" lasts	Decades, never resolves	Closes as pay pulls people in

How do you run the wage test on your own offer?

This is where it stops being economics and starts being your salary. Here's the move.

Weak version: "Companies just need to pay more." True, useless, and it changes nothing in your negotiation.

Strong version: Pull the real median for your target role from a source the employer can't wave away. BLS Occupational Employment Statistics in the US, Levels.fyi for tech comp, or local salary data for your market. If the offer sits below median and they're citing a skills shortage to explain slow, stingy hiring, you've diagnosed a price problem wearing a supply-problem costume. A genuine shortage would mean they're already paying above median just to get anyone in the door. They're not. So the gap isn't your credentials. It's their number. Keep in mind there is no single market rate either, because pay for any role is a 2x range and where you land is politics, so aim for the top of the band the data supports.

Then the posture shifts. You stop apologising for what you might lack and start naming what the role is worth.

"I've looked at the market for this role. The median is X. Your offer is below that. If the concern is finding qualified people, the data says the constraint is the package, not the candidate pool. Here's where I'd need the number to land."

That's a market-informed stance, not a needy one. You're not asking them to like you more. You're telling them what the role costs.

The trade-off, named

This reframing has a cost, and you should know it before you use it. Walking in with the wage test changes the temperature of the room. Some employers will respect the clarity and move on price. Others will read it as friction and quietly favour the candidate who didn't push. You're trading a little likeability for a lot of leverage.

That trade is usually worth it, because the alternative is accepting a below-market number and carrying that discount forward. An early role priced low doesn't just cost you this year. It anchors every raise and every next offer that gets benchmarked against it. The discomfort of one negotiation is cheaper than years of compounding from a low base. But it is a real trade, and you make it with your eyes open.

What to do now

1. Find your number. Pull the real median for your exact role and market from BLS, Levels.fyi, or local salary data. Write it down before any conversation.
2. Run the wage test on the claim. Has pay in your target role risen faster than inflation over the last three to five years? If not, treat "we can't find talent" as "we won't pay market."
3. Reframe the gap. When you hear shortage talk, ask yourself whether they're hiring above median to attract anyone. If they're not, the constraint is the package, not you.
4. Negotiate from value, not apology. Name the median, name the gap, name your number. Let the data carry the weight.

The skills-gap story was built to keep you focused on what you might lack. The more useful question is whether you're being paid what your skills are already worth. Most of the time, the answer is no, and that's a fixable problem the moment you stop accepting the framing.

Want to know if your offer is a real market rate or a "shortage" discount in disguise? Send me the role and the number on WhatsApp and I'll pull the market median with you, then help you write the counter. That's what I'm here for.