The headlines say crisis; the workforce statistics say more GPs than ever; and both are quoting real numbers. The way through is to separate three things headlines routinely blur: unemployment (GPs who cannot find any GP work), underemployment (GPs working less than they want to), and preference mismatch (GPs unable to obtain the sessions, roles or locations they would choose). The data supports different conclusions about each, and the distinction changes what any solution would need to fix.
What the surveys show
The BMA's UK-wide survey of the issue, building on its earlier locum-specific survey, found 15 percent of responding GPs unable to find any GP work at all, and 56 percent wanting to work more NHS hours than they could find, with 60 percent reporting pay-rate declines over the preceding year, 69 percent reporting stress or anxiety attributable to un- or underemployment, 21 percent making definite plans to leave the profession and 47 percent considering alternative careers. Survey caveats apply, self-selection among affected GPs is likely, but the direction is unambiguous and the sessional workforce is clearly bearing the worst of it: the BMA's newly elected sessional committee chair described taking three years to find a suitable role. On these numbers, underemployment is the larger phenomenon, with a smaller but real core of outright unemployment concentrated among locums and the newly qualified.
What the workforce statistics show
England's official series complicates any simple shortage-of-GPs or glut-of-GPs story, because the answer depends on definitions. Counting fully qualified GPs plus trainees by headcount, numbers rose about 18 percent from 2015 to 2024; counting fully qualified GPs by full-time equivalent, the delivered-capacity measure, England had 29,008 FTE in June 2026, slightly below the 2015 baseline, while the registered population grew 13 percent to 63.4 million. A peer-reviewed analysis in the BJGP showed the same decade reads as anywhere from modest per-capita growth to a 15 percent per-capita decline depending on definition, and found the spread between practices widening dramatically, the 95th-percentile practice moving from roughly 4,100 to over 5,500 patients per FTE fully qualified GP. Meanwhile the most recent monthly releases show qualified permanent GP FTE rising year-on-year, up around 2.3 percent, consistent with policy pushing employment toward salaried practice roles. One further honest caveat from the BMA's own analysis: the practice, PCN and ARRS datasets overlap, so no single transparent national count of working GPs currently exists.
Reconciling the two pictures
Put together, the data describes not a labour surplus but a matching failure. Capacity per patient is at best flat against sharply higher demand, which is why access remains poor; yet the employment mechanism, practice finances, the historical shape of ARRS, the collapse in flexible locum spend, geography, has been unable to absorb the GPs seeking work, which is why individual GPs experience a jobs drought inside a service short of GP time. "Unemployment crisis" is accurate for a minority and for the locum market specifically; "underemployment and distribution crisis" is the more precise description of the whole, and it is arguably the more damning one, since the constraint is organisational rather than a lack of either need or doctors.
What the recent policy changes do
Two 2026/27 contract changes bear directly on the matching failure: the practice-level GP employment reimbursement scheme, £292 million repurposed to fund practices, at £4.57 per adjusted patient, to employ additional salaried GPs or sessions, recurrent in the core contract; and the removal of ARRS restrictions so the scheme now reimburses GPs at any career stage at improved rates. Both push money toward the exact point of failure, practice-level employment. The early monthly data showing salaried FTE growth is consistent with an effect; whether it reaches the sessional GPs reporting the worst outcomes, and how pay rates respond, is what the next year of surveys and workforce releases will show. We track the open-vacancy picture quarterly at /blog/where-are-gp-jobs-available-uk.
Frequently asked questions
So is "GP unemployment crisis" fair or hype?
Fair as lived description for a real minority, especially locums and the newly qualified; imprecise as a system diagnosis, where underemployment plus maldistribution fits the data better. Both framings agree on the central absurdity: unused GP capacity alongside unmet patient need.
Why do government and BMA numbers seem to conflict?
Mostly definitions: headcount versus FTE, all GPs versus fully qualified, practice versus PCN datasets that overlap. Nearly every quoted figure is real; the selection is where the argument lives.
Where should a GP affected by this start?
With the practical channel-by-channel approach at /blog/newly-qualified-gp-cant-find-job-practical-guide, and the BMA's support resources for un- and underemployed GPs, which include career, financial and wellbeing help.
What single indicator is worth watching over the next year?
Salaried GP FTE in the monthly workforce release, read alongside sessional-rate reports: if the 2026/27 funding is working, the first shows sustained growth and the second stabilises. Divergence between them would mean the money is landing without reaching the GPs reporting the worst of it.