No official series measures unmet demand for a skill. The observable proxy is a metro median well above the national median for the same occupation, particularly where local concentration is not high — that combination suggests employers paying up to attract people rather than benefiting from a deep local pool. Anchorage pays petroleum engineers a median of $212,240, the highest of any metro holding them.
Nobody publishes a shortage
Scarcity is not directly published anywhere. What is published is the pay, and unusually high local pay for an occupation is the closest observable signal you can get. Everything else on the subject is somebody’s assertion.
There is no series measuring unmet demand for a skill, and the word shortage covers three different situations. Employers say it when they cannot hire at the price they budgeted, when the local pool genuinely is thin, and when they are asking for a combination of skills nobody actually has. The word covers all three situations without distinguishing them.
All three feel identical from inside a company, and only one of them is a shortage in any economic sense. The budget version resolves itself the moment the budget moves. So the useful question is not who is claiming a shortage but where pay has already responded to one.
The signal that works: a premium without a pool
Start by comparing a metro’s median for your occupation against the national median for the same occupation. A large premium means employers there are paying up for something, though it does not yet tell you what. A premium is a symptom rather than a diagnosis.
Then check the concentration, because that is what separates the two explanations. A premium in a place with a deep pool usually means a high-value local industry paying its people well, which is San Jose for software. Everybody is already there and the pay is still high.
A premium in a place with a thin pool is much closer to genuine scarcity. The employer is paying to get somebody there at all rather than benefiting from everybody already being there. Those two situations look identical in a salary table and mean completely different things.
Anchorage against Houston
Petroleum engineers have a national median of $144,910 across about 18,060 people. Anchorage pays a median of $212,240, which is a premium of 46 percent, to roughly 200 of them in the whole metro. Two hundred people is not a labor market by any measure.
Houston pays $169,870 to 4,770 of them. That is a real premium too, and it comes from a deep market where dozens of employers compete for the same engineers. That competition is what produces the Houston number.
The Houston premium is industry. The Anchorage premium includes the cost of persuading somebody to move to Anchorage, which is a different thing being paid for. Both numbers are correct and only one of them describes a market you could stay in after a job ends.
The second signal: a stretched local distribution
Where employers are competing hard for experienced people, the top of the local range pulls away while entry pay barely moves at all. That produces a distinctive shape in the published percentiles. The gap between the ends widens while the middle holds.
A metro with an unusually wide gap between its tenth and ninetieth percentile for your occupation is telling you that experience specifically is scarce there. The median can look entirely ordinary while that is true underneath it. Medians hide distributions and this is a clear example.
That is also the most actionable version of this signal. It says the premium is available to you specifically rather than to anybody who shows up, which is the difference between a market worth moving for and one that simply pays well. Your own experience is what the premium is buying.
The trap in the other direction
Low local pay does not mean low demand, and reading it that way rules out places that would have hired you. There are at least three innocent explanations for a low figure. None of them says anything about hiring appetite.
It can mean abundant local supply, a low local cost base, or simply an occupation that is inexpensive everywhere in the country. None of those is a statement about whether anybody is hiring. Ruling a place out on price alone is a common mistake.
The comparison that means something is your occupation in that metro against your occupation nationally. Comparing it against other occupations in the same metro mostly measures the metro rather than your prospects. Hold the occupation constant and vary the geography.
What employers mean when they say it
Usually they mean the role has been open a long time at the salary somebody approved months ago. That is a real constraint for them and it is a budgeting problem rather than a market one. Budgets are set in advance and markets move afterwards.
It is worth recognizing because it is also an opening rather than a closed door. A long-vacant role is a role where the budget is considerably more likely to move than the requirements are. Time open is the pressure that eventually reopens a range.
Where a genuine shortage exists, pay has usually already moved and you can see it in the published figures before anybody uses the word. The data leads the language by a comfortable margin. Watch the published figures rather than the announcements.
Using this to pick where to look
Look for three things together rather than any one of them. A metro median meaningfully above national for your occupation, a concentration that is not already very high, and a wide local spread between the percentiles. Any one of the three alone is much weaker evidence.
That combination is where an experienced person has the most leverage available. All three numbers are published for every occupation across 393 metros and none of them takes long to look up. An afternoon covers every metro genuinely worth considering.
It will point you somewhere unexpected more often than not. The obvious destinations are obvious to everybody, and their candidate pools are correspondingly deep with people exactly like you. Competing where nobody is looking is the better position.
A worked example, start to finish
Say you are an experienced petroleum engineer deciding where to look. The national median is $144,910. Houston pays $169,870 and holds 4,770 people, Midland pays $172,070 with 940, and Anchorage pays $212,240 with about 200.
Houston is the deep market: highest employment, a real premium, and dozens of employers within reach. If a job ends there you interview across town rather than across the country. Midland pays marginally more with a fifth of the people, so the premium is concentration in a small place and the market is genuinely specialized.
Anchorage pays 46 percent above national to a couple of hundred people. That is not a deep market paying well; it is a small number of employers paying whatever it takes, with a housing market attached to the same industry. All three are defensible choices, and raw salary ranks them in exactly the wrong order for anybody who values being able to leave.
When the scarcity is you rather than the place
The most valuable version of this is not geographic at all. If your particular combination of skills is genuinely uncommon, the shortage travels with you wherever you go. Geography stops being the constraint at that point.
The question then becomes which employers feel it most acutely, and that is findable rather than a matter of guesswork. Watch the same employer’s listings over a couple of months and note which roles keep reappearing. Repetition is the tell rather than the wording of the posting.
A role posted continuously for a quarter is a role where somebody has become impatient. Impatience is the condition under which budgets move, and it is visible from outside the company for free. Nobody has to tell you and the listings already have.
Common questions
Is there a published measure of skills shortage?
No. Nobody publishes unmet demand, so pay is the closest observable proxy for it.
What does high local pay indicate?
That employers are paying up. Combined with low local concentration it points to genuine scarcity rather than a high-value industry.
What else signals scarcity?
A wide local gap between the tenth and ninetieth percentile, where competition for experienced people pulls the top away.
What do employers usually mean by shortage?
Often that they cannot hire at the price they budgeted, which resolves when the budget moves rather than being a shortage.
Does low local pay mean low demand?
Not necessarily. It can mean abundant supply or a low cost base. Compare against the same occupation nationally, not other occupations locally.
How can I spot a genuine skills shortage?
A metro median well above national for your occupation, combined with LOW local concentration. High pay with a deep pool is usually a rich industry, not scarcity.
What do employers usually mean by shortage?
That a role has stayed open at the salary they approved. That is a budgeting constraint rather than a market one — and it means the budget is more likely to move than the requirements.