The irony is palpable and unpleasant: unemployed and underemployed GPs are struggling to find NHS work, while patients can’t get a GP appointment when they want one. With public satisfaction with general practice at record lows, this is not a tenable situation for a government committed to improving access to general practice.
For over a decade, England struggled with a lack of GPs. Faced with high and rising workloads, many GPs cut their clinical hours. Others retired, or took jobs outside general practice. Practices struggled to recruit, and GP training places went unfilled. Challenges with high GP workload, burnout and falling job satisfaction remain, but in the past two years the labour market in general practice has changed significantly. From practices clamouring for recruits, GPs are now clamouring for practices to recruit them.
This short paper explores the mix of reasons behind the underemployment of GPs in England.
What’s happened to GP employment?
(i) The number of newly qualified GPs is increasing
After years of broken promises to boost GP numbers, rapid expansion of speciality training places for GPs (posts for resident doctors to train in general practice) is finally yielding results. The number of GPs qualifying per year has risen from 2,751 in 2015 to 4,090 in 2024: a 49% increase.
Although it’s positive that there are rising numbers of GPs qualifying each year, decreasing workforce participation (the amount of time each qualified GP spends seeing NHS patients) means that full-time equivalent (FTE) GP numbers haven’t risen perfectly in line with joiners to the profession.
(ii) The number of available GP jobs has fallen
As the number of GPs qualifying each year has risen, the number of jobs available in practices is falling. In December 2025, NHS England started publishing data on the number of NHS job adverts in general practice. This comes with caveats. These aren’t yet formally ‘official statistics’, and data only includes adverts published on NHS Jobs, which may miss roles advertised by agencies, word of mouth or other platforms.
Nonetheless, the data is instructive. Since the start of data collection in the first quarter of 2022/23, FTE GP job adverts in general practices in England have decreased by 45%, from 919 FTE in the first quarter of 2022/23 to 504 in the second quarter of 2025/26. The summer of 2024/25 saw the lowest number of jobs advertised, at 340 FTE (summer is when most final-year GP trainees complete their training, so usually aligns with the highest numbers of GP jobseekers). It should be noted that this data does not include GP jobs within Primary Care Networks (addressed below).
Although the number of GP jobs being advertised has fallen in practices in all regions of England, some areas are more affected than others. Practice-based GP jobs seem particularly scarce in London, the south east and north west (where FTE job adverts have fallen by 60% and 55.3% and 55.2% respectively since early 2022). The south west of England has seen the lowest fall in advertised roles – a 23% reduction over the same period.
What’s driving the fall in available GP jobs?
The simplest answer would be that we finally have ‘enough’ GPs: that we are training sufficient GPs to offset leavers and fill available posts, such that no more are needed. But this doesn’t entirely add up. GP surgeries continue to struggle with demand, the public continues to remain dissatisfied with access to GPs, and patient list sizes – the number of patients each GP is responsible for – remain stubbornly high. So, what else might be contributing?
(i) There’s been massive growth of other allied health professionals working in general practice
Since 2019, the clinical staffing of general practice in England has changed enormously. Groups of practices working as Primary Care Networks (PCNs) have been funded to recruit an expanded range of staff, now from over 20 roles including paramedics, physios, pharmacists and physician associates. This ‘Additional Roles Reimbursement Scheme’ (ARRS) has allowed practices to recruit new staff effectively without cost to themselves (i.e money does not come from core practice funding).
In part designed to allow the Treasury to put funding into general practice with certainty that it would be used to pay for staff (rather than entering the melting pot of the core GP contract, from which GP partner income is drawn), recruitment under the ARRS has been dramatic. There are now over 42,500 additional FTE direct patient care staff working in general practice (see chart below). This massive expansion in other allied health professionals in general practice wasn’t intended to replace GPs – but the significant numbers recruited, and the absence of cost to practices, means that at least some substitution seems likely.
(ii) Some GPs are working under the Additional Roles Reimbursement Scheme
By the summer of 2024, with few GP jobs available, another large cohort of trainees about to qualify, and headlines about unemployed GPs working as Uber drivers, the new government faced a decision. Rely on practices to recruit new GPs (risking the possibility that they wouldn’t, leaving a cohort of freshly qualified GPs jobless, while patients complained about poor access), or allow GPs to be recruited under the ARRS. The government chose the latter, making £82 million available to employ newly qualified GPs under the ARRS from October 2024.
Knowing precisely how many early career GPs are working under the ARRS is difficult. Data is based on retrospective payment claims and plagued by significant time lags. But by October 2025, NHS England data suggests that 2,614 GPs were working a total of 1,383 FTE jobs in ARRS roles.
The expansion of the ARRS to include newly qualified GPs was explicitly described as an emergency measure to address the immediate issue of GP unemployment. It is not uncomplicated. ARRS roles are fixed term, and don’t offer the stability of permanent posts. They are in PCNs – not in individual GP surgeries – so GPs may be expected to work between multiple practices, limiting their ability to provide continuity of care. And there is also a significant risk that allowing employment of GPs under the ARRS is worsening the GP recruitment crisis: there is little incentive for GP partnerships to employ salaried GPs directly at practice level when they can access ARRS GPs for free.
How does all this relate to practice funding, the GP contract and the partnership model?
There is some merit in the argument that years of constrained practice funding may have limited the ability of GP surgeries to afford new recruits. Although overall government investment in general practice has increased over the past 15 years, it has not kept pace with inflation. The BMA claim that practices have struggled to meet rising wage bills, keep up with utility costs and remain afloat. Employing more GPs – particularly on permanent contracts at a time when annual GP contract negotiations have often been fraught and the outcome of funding deals uncertain – may not be an appealing choice for GP partners.
But two things undermine arguments that underfunding is driving underemployment. First, the 2025/26 GP contract was a major funding boost for general practice, with a £969 million uplift representing a 7.2% total increase to core funds. That direct GP employment by practices doesn’t appear to have risen significantly this year suggests that partners may be relying on ARRS GPs rather than making practice-based posts available.
Second, GP partner income has risen significantly. Earnings data from general practice isn’t perfect – it’s based on anonymised tax data from HMRC, and while major outliers are excluded, the averages presented mask significant variation. But as the way data is collected hasn’t changed much recently, comparison between years is a reasonable indicator of GP pay. On average, GP partner income increased by 13.2% between 2022/23 and 2023/24. Although averaged national data masks practice-level variation, arguments that GP partners cannot find GPs to hire, or cannot afford to employ them in practices, sit at odds with available evidence.
What’s the solution to GP underemployment?
Factors that led to GP unemployment are different from those now driving it. Initially, underemployment of GPs likely did result from over a decade of constrained funding in general practice. Although allowing GP employment under the ARRS solved an immediate crisis in the summer of 2024, it’s now causing problems: evidence suggests that many GP partnerships are relying on staff recruited through the ARRS (now including GPs), rather than recruiting directly at practice level. Limited availability of practice-level GP jobs is likely driving early career GPs to ARRS roles – fuelling a cycle which will ultimately result in a lower proportion of GPs being employed in GP surgeries.
Despite the introduction of health professionals with a range of different skills, it’s clear that public satisfaction with general practice rests heavily on patients having access to GPs. And as expert medical generalists, and leaders of multidisciplinary teams, more GPs are needed to maintain high functioning general practice. But employing additional GPs requires having money to pay for them. Without greater overall funding for general practice, the government has three options: move funding to employ GPs ‘up’ to neighbourhoods, ‘down’ to practices, or keep it at PCN level through the ARRS. Each has different implications for how GPs will work.
The BMA support a traditional model of general practice – arguing that the next GP contract should replace the ARRS with a ‘practice level funding scheme to support practices to hire more GPs’. This may offer a route to permanently employing more GPs at practice level, stabilising the GP workforce and enabling greater continuity of care. Were funding to be prioritised for under-doctored areas, it might also support practices in socioeconomically deprived areas, where GP numbers are lower. But the Treasury’s concerns about how practice-level funding would be spent could only be allayed if funding is ringfenced and transparent – only available for the purpose of GP employment. And policymakers know that once additional funding is put into the core GP contract, they would be unable to claw it back.
‘Neighbourhood health’ and ‘integrated neighbourhood teams’ were key features in the primary care section of the government’s 10 year health plan, but are still evolving conceptually. If policymakers envisage a future where some GPs are employed in wider neighbourhood teams, then redirecting ARRS money to new neighbourhood services would be a sensible bet: funding GPs to work at neighbourhood level would help create new services.
The government knows that it must end the paradox of unemployed GPs, while patients demand more GP appointments. Choosing whether to direct money to employ GPs to practices or to neighbourhood teams represents a major fork in the road – a choice between a ‘traditional’ model of general practice, and a new, less clear ‘neighbourhood’ service. But this is a false dichotomy, created by constrained funding. In creating new neighbourhood services, policymakers must not destroy core general practice. Ultimately, this may require extra funding to pay for additional GPs to work in practices, alongside creating GP jobs in neighbourhood services. If the Secretary of State is serious about shifting more care to communities, this type of increased investment in primary care is exactly what’s needed.
*With thanks to Lucina Rolewicz for her work on the third chart.