The phrase talent shortage does a great deal of quiet work. It describes a condition external to the company, like a drought, about which the reasonable response is to compete harder for what exists and to lobby someone else to produce more.
For a large share of the roles it is applied to, this is a description of a decision employers made and have declined to revisit.
There was a period in which firms hired for aptitude and produced skill themselves. Apprenticeships, rotational programs, formal training with real duration, and the ordinary practice of hiring someone who could not yet do the job and expecting them to learn it. That system was expensive and it worked, and it was dismantled for a defensible reason: a trained employee can leave, and the firm that trained them absorbs a cost their next employer does not.
The rational decision that became a collective failure
Every individual firm was right. If skills can be acquired by hiring rather than by developing, hiring is cheaper and faster and carries no risk of training someone for a competitor. The trouble is that the strategy only works while somebody else is still training, and once enough firms reach the same conclusion the supply they were all drawing on stops being replenished.
The evidence that this is a training problem rather than a scarcity problem is in the job postings themselves. Roles described as entry-level requiring several years of experience. Requirement lists no single candidate plausibly satisfies. Searches that run for months while internal candidates who could do most of the job, and learn the rest in a quarter, are passed over because they do not match on paper.
That last one is the tell, and it connects directly to the internal mobility gap most companies have. An organization that will not consider a partially qualified internal candidate but complains that no qualified external one exists has revealed that the constraint is its own willingness to develop anyone.
The employers moving away from this are, not coincidentally, the ones that had to. Skills-based hiring began as a response to shortages in roles where the credentialed supply simply was not available, and it has held up because assessing what someone can do turns out to predict performance better than what they studied. Once a firm can assess capability directly, developing it internally becomes tractable, because progress can be measured.
Public programs have moved in the same direction, with federal workforce funding restructured around employer partnership rather than standalone training that produced credentials nobody was hiring for. That is an improvement and it does not resolve the underlying free-rider problem, which is that no public program can substitute at scale for firms declining to develop their own workforce.
The aggregate numbers also do not close by training alone. In many regional labor markets the working-age population has grown for one reason only, and no employer development programme substitutes for that arithmetic.
The honest version of the complaint would be: we would like to hire people who are already trained, we are not willing to train them ourselves, and we would prefer someone else pay for it. Stated that way it is still a defensible position. It is simply not a shortage.
An earlier argument in this space, The Case for Hiring Slowly in a Fast Market, made the case from the other direction.
