What's next
What we're working on next
Scarcity counts how many people finish training for 24 health and education jobs and compares that with projected need. This page says what we plan to add, how we decide, and what we won't do.
Last updated November 17, 2026. What has already changed is on What changed.
How many people are starting training
Graduate counts look back. Someone who finishes a program this year started two to four years ago. To see what's coming, we want to count new students and applicants too.
What's on the pages now
Most profession overviews now have a Demand section. It shows national totals of applicants and new students from each job's application service, accreditor, or professional group, with credit and a link. They are context only. They don't change any label.
What's still missing
Dental hygiene, nurse practitioners, and teacher preparation have no student counts yet. One is a paid report, one group's terms ask for permission first, and teacher preparation isn't checked yet. We've written to those groups. Where a group's report says it may not be reproduced, we show national totals only until it says yes.
Each Demand section also shows counts of people who may need the service: people 65 and older now and projected, Medicare and Medicaid enrollment, public school enrollment, and survey rates of mental illness and substance use. These explain need. They don't say how many workers are needed. Next we add them to state pages.
What these counts won't be used for
- An applicant who applies to several programs is one person. Applications are not applicants.
- New students are not future graduates. Some leave before they finish.
- Enrollment won't change a label until we write a rule for it and publish the rule first.
Spotting change early
Some changes show up in rules, grants, and contracts years before they show up in programs or graduates. We write short, dated notes on them on Emerging demand, and each profession's Demand section lists the notes that name it.
What a note says
- What happened, when, and the original document.
- How far along it is: proposed, announced, funded, contracted, operating, or staffed. A grant notice isn't an award, and a company's plan isn't a sure thing.
- What work it could add or change, and who would be affected.
- What's uncertain, and the next thing to watch.
A note doesn't claim a shortage or a new job. We keep every note as it was published and add what happened after, including the ones that turned out wrong or stalled.
An example of the kind of event: Medicare began paying marriage and family therapists and mental health counselors on January 1, 2024. That could change how many people seek those services. On its own, it doesn't show that more graduates are needed.
Where we look: the Federal Register and USAspending now. Grants.gov and company filings with the SEC are next. All are public.
Job ads, only if the notes earn it
Later, we may track job ads from a fixed set of employers, the same way over time. We'll only do it if the notes show it's worth the cost. If we use job-ad data we lease, every place it shows will say so.
Steps from a first sign to a label
These are starting rules we'll test, not proven thresholds.
| Step | What starts it | What it allows |
|---|---|---|
| Detected | The same new tasks at 3 unrelated employers, or one documented event | We keep it on an internal list |
| Under review | About 20 job ads from 5 employers in 90 days, with duplicates removed | We look into what the work is and where it came from |
| Research note | The same tasks over 2 quarters, at 10 employers, in 3 labor markets, plus a second kind of source | A published note, with its limits |
| Job page | Clear boundaries, the ways people get in, and evidence, reviewed before it is published | A full page |
| Shortage label | Separate evidence on need and on qualified supply | A label from the written rules |
More jobs
Every job, listed
We plan to list all 867 jobs in the federal job classification at a basic level, each with an ID that never changes. A job we haven't reviewed will say "Not yet reviewed." Full pages come one at a time.
How we pick the next job
Every new page has to pass four questions. We write down the answers, so the reason stays on record.
- Is it useful? Will it help someone pick a path, plan a program, or understand a staffing problem?
- Is the evidence there? Can we say who's counted and get usable data by state?
- Is something changing? Is there a real development, sources that disagree, or a change in the work?
- Can we keep it up? Does it reuse data we already pull, or is new data worth the upkeep?
The next seven jobs
Five reuse the data we already pull. Two each test a new way of counting people who enter a job. They're listed in the order we plan to add them.
| Job | Why it's in the first group | New work it needs |
|---|---|---|
| Diagnostic medical sonographers | A health job close to the ones we cover, where several credentials overlap | Little: IPEDS plus credential counts |
| Occupational therapy assistants | Pairs with the occupational therapist page | Little: IPEDS plus exam counts |
| Clinical laboratory technicians | Tests keeping technicians and technologists apart | Little: IPEDS |
| Epidemiologists | Public health, where one degree leads to many jobs | Little: IPEDS, with care in matching programs to jobs |
| Information security analysts | The first computing job, with many ways in | More: matching job tasks to a federal framework |
| Electricians | Tests counting apprenticeships | New: apprenticeship data |
| Career and technical education teachers | Tests counting teachers by state assignment and license | New: state assignment and license data |
Ways into a job
College is one way into a job, not the only one. Each page will count the ways people actually get in, and say which ones we measured and which are still gaps.
| Way in | What we count | Mistake to avoid | Where |
|---|---|---|---|
| College | Awards that qualify someone to start, one per person | Counting a degree for people already in the job as a new worker | Most pages now |
| License or exam | First-time licenses and first exam passes | Counting renewals or licenses in a second state as new people | Nursing and other health pages now |
| Apprenticeship | Registrations, completions, and journey-level workers | Counting someone twice who also went to college | Electricians, first |
| Teacher assignment | State assignments, endorsements, and shortage areas | Treating a subject degree as a teaching license | Shortage areas on the special education and ESL pages now; career and technical education next |
| Residency | Residency and fellowship entrants and completions | Counting a fellowship as a new doctor | Later |
| Employer training | Entry, staying in the job, pay, and working conditions | Calling a staffing problem a graduate shortage | Later |
Rules for all of it
These hold for every new page, note, and label. When a page breaks one, the page changes, not the rule.
- Keep each judgment separate. Whether a job is new, whether hiring is rising, whether supply falls short, and whether a program would fill are different questions. Each gets its own answer. We don't combine them into one score.
- Not yet reviewed is not "no shortage." A job we haven't checked says so. Missing evidence and evidence against a shortage are different things.
- No shortage from openings minus graduates. Projected openings are not unfilled jobs, and a program's code doesn't say where its graduates go. A shortage label needs separate evidence on need and on supply.
- Never add a job to the jobs inside it. A specialty, its parent job, and the family they belong to overlap. We show their counts side by side and never add them up.
- Sort jobs by the work, not the title. Two jobs with the same title can be different work, and a new title can be old work.
- Our review order is a choice, not a ranking. Which job we cover next is an editorial call. It says nothing about which jobs have the worst shortage.
- No one pays to be covered. No school, company, or group can pay to have a job added, moved up, or featured.
The current rules for labels are on How we measure.
Suggest a job or a source
If you know a job we should cover, or a public source we should use, email corrections@scarcity.co. Tell us what it counts and where it's published. We read every message. Suggestions go through the same four questions, and no one can pay to move a job up.