Nursing Wiki

Where the Nursing Numbers Come From

Almost every statement made about American nursing traces to one of six data sources. They measure different things, on different schedules, with different weaknesses, and most public arguments about nursing are really arguments between two of them.

This page describes each one and what it can and cannot support. The pages across this wiki cite these sources directly, and this is where their caveats live.

The federal wage survey

The Occupational Employment and Wage Statistics programme surveys employers twice a year and publishes annual estimates of employment and wages by occupation, state and metropolitan area.

It is the authority for what nurses are paid. It reported an average annual wage of $101,420 and a median of $97,550 for registered nurses in May 2025.

Three limits. It describes a reference month roughly a year before publication, so it is a benchmark rather than a live quote. It counts employees of surveyed establishments, so self-employed nurses and some contract arrangements sit outside it. And the agency itself warns against reading year-to-year changes as pure wage growth, because sample and job coding shift between releases. See registered nurse pay.

One underused feature of it is the industry breakdown, which is the only official measure of agency-employed nurses: registered nurses working for staffing firms rather than facilities went from 57,270 in 2019 to 130,890 in 2023. See where the assignments are.

The occupational projections

The Occupational Outlook Handbook publishes ten-year employment projections, currently 2025 to 2035, including expected growth and average annual openings.

It is the authority for whether an occupation is growing. It projects 6 percent growth for registered nurses and about 180,800 openings a year, most of which come from nurses leaving the occupation rather than from new posts.

The limit is that projections describe a national labour market under assumptions about demand, technology and retirement. They cannot tell you whether a given hospital can fill a given post, which is what a vacancy rate measures. See the nursing shortage.

The federal workforce projections

The health workforce agency runs a separate microsimulation model of nurse supply and demand, currently covering 2023 to 2038, which is where shortage figures with a precise number attached usually originate. It projects a registered nurse shortage of about 9 percent by 2036, roughly 337,970 full-time equivalents, and a larger shortfall for practical nurses, with rural areas considerably worse than metropolitan ones.

The limit is inherent to modelling: outputs move with assumptions about how many nurses enter, how long they stay and how much care the population needs. Two credible models can disagree by hundreds of thousands of nurses, which is why a projected shortage figure should always carry its source and its year.

The regulators' workforce survey

Every two years, the boards of nursing and the state workforce centres survey a stratified sample drawn from the licensure databases themselves. The 2024 edition is the best available description of who American nurses are: median age 50, 7.9 percent under 30, and 39.9 percent saying they expect to retire or leave nursing within five years.

Its strength is the sampling frame. It draws from licence records rather than from a membership list or a panel, so it reaches nurses who belong to nothing.

Its limits are stated in the report. The headline count of licences is not a count of nurses, because one nurse can hold several. Responses are weighted, and white, female and older nurses may be slightly overrepresented. And answer options change between editions, which is why the jump in intent to leave is not a clean comparison with 2022.

The hospital staffing report

An annual survey published by a nurse recruitment firm collects hospital-reported turnover, vacancy, time to fill and cost figures. The 2026 edition covers 527 hospitals and 262,405 registered nurses, and is the source for turnover of 17.6 percent, a vacancy rate of 8.6 percent, 78 days to fill a post, and a replacement cost of $60,090 per nurse.

Two limits belong with every use of it. Hospitals self-report and choose whether to take part, so the sample is not random. And the publisher sells nurse recruitment, which is an interest worth naming even though the report is the closest thing the field has to a scoreboard. Method changes matter too: the vacancy calculation changed in the 2026 edition, so part of the improvement is definitional. See nurse turnover and retention.

Mandatory facility reporting

Nursing homes must report their staffing hours to Medicare from payroll records, including hours worked by contract staff. This is administrative data rather than a survey, published quarterly by facility, and it is the only official measure of agency use that breaks down by state.

It is the source for contract staff being 7.5 percent of nursing home nurse hours nationally in early 2024, and for a state range running from 25.1 percent in Vermont to 1.0 percent in Alabama in mid-2025.

Its limit is scope. It covers nursing homes only, so nothing equivalent exists for hospitals, whose contract labour is visible only in Medicare cost reports and in aggregate industry surveys.

The nursing school survey

The association of university nursing schools surveys its members annually on enrolment, graduations, faculty vacancies and applications turned away, and runs a separate survey of vacant faculty posts. It is the source for enrolment growth, for the roughly 90,000 qualified applications refused each year, for a faculty vacancy rate of 7.2 percent, and for the salary gap between teaching and practice.

Its limits are membership and definition. It covers baccalaureate and graduate programmes at member schools, so associate degree and practical nursing education is thinner in the data, and applications turned away counts applications rather than individual people, since one applicant may apply to several schools. See the nurse faculty shortage.

Licensure examination results

The regulators publish pass rates by candidate group and quarter. These are administrative counts rather than survey estimates, so they are exact, and they are the earliest signal of how many new nurses will actually reach the workforce. See the NCLEX.

Job listings and candidate profiles

Job boards hold a different kind of record: what employers advertised and what nurses said they wanted, rather than what either did. Several pages on this wiki draw on NurseRecruiter's own archive, which holds 4.68 million travel postings dated between 2009 and 2026 from 739 different agencies and employers, and 375,448 registered nurse profiles.

What it is good for is granularity nothing else offers. It can say which states advertise the most travel assignments per employed nurse, that 82 percent of postings stating a length say thirteen weeks, that medical-surgical and telemetry account for a third of travel demand between them, and that 38.9 percent of registered nurses say they will consider travel work against 45.7 percent for per diem shifts.

Its limits are as specific as its strengths. A posting is not a filled assignment, and agencies repost. Volume over time tracks which firms were feeding listings more than it tracks the market, so shares and per-nurse ratios are usable while raw counts and trends are not. The audience is self-selected. And some fields are unreliable at the level below state: in 38 percent of postings the recorded city belongs to a different state than the job does, which is why no city-level figures are published here.

Every use of it on this wiki is labelled and its method described in where the assignments are.

How to read a claim about nursing

Three questions catch most errors.

Which source is it, and what does that source actually count? Licences, jobs, full-time equivalents and people are four different units, and shortage arguments routinely mix them.

What year does it describe, rather than when was it published? A wage figure published this year usually describes last year.

Who paid for it? Several of the most-used sources are produced by organisations with a position on the answer, including a recruitment firm, an association of nursing schools, a job board, and the professional bodies. That does not make them wrong. It makes the primary document worth opening.

Sources

Find your next nursing job here

100% free for nurses. New jobs every day. Travel, per-diem, and permanent positions nationwide.

See something out of date? Sign in to suggest an edit.