Training capacity has increased, but the number of graduates was no higher than in 2013 according to the latest comparable data cited by DREES. Coverage of nursing care needs has declined despite rising workforce numbers and would decline further under DREES projection assumptions. Part of the problem occurs after graduation: the birth of a first child reduces the average volume of paid work for former hospital nurses by 0.14 FTE after five years and 0.22 FTE after ten years. Under the assumption of constant consumption by age, nursing care needs would increase by 50% between 2021 and 2050 according to DREES.
The essentials
- The increase in training capacity and graduates is not continuous, and the sources consulted do not measure a national deficit in nursing hours (DREES, IGAS-IGESR).
- Between one and ten years into their careers, average paid volume drops from 0.84 to 0.67 FTE; the study particularly isolates an effect of motherhood on work volume (DREES).
- 87% of nurses are women; the shift to part-time work mainly explains the individual decline in work volume associated with motherhood; its net effect on total hospital supply has not been established by this study (DREES).
- Under the assumption of constant consumption by age, needs would increase by 50% between 2021 and 2050 (DREES).
- Motherhood reduces the paid volume observed in the cohort studied, without allowing conclusions to be drawn about capacity tensions by 2040.
A deficit accumulating quietly
The figure of a nursing shortage has been circulating for several years. It is regularly cited, rarely analyzed. The common error is to read it as a training problem. Graduate cohorts have grown. Nursing schools have opened more places.
Yet the volume of available hours is growing slower than needs related to population aging: nursing numbers have increased, but not as fast as these needs.
The reason lies in a simple mechanism that DREES data documents with precision: the stock of available nursing hours in the system depends not only on the number of graduates. It depends on the hourly volume each professional delivers over her entire career. Between one and ten years into their careers, this volume declines by roughly a fifth according to the DREES cohort.
The study describes a progressive decline in work volume over the course of a career. The shift to part-time work is not systematically compensated by additional recruitment at the department level. Her patients are redistributed among her present colleagues, or are no longer covered at all.
Multiplied across thousands of similar trajectories, the effect is massive. This decline is measured in the studied cohort at ten years into the career, without always appearing as recorded departures. They are recorded as contractual adjustments, extended maternity leave, shifts to part-time work “by choice.”
The structure of a profession 87% female
87% of nurses in France are women. This figure is known. Its organizational consequences are less so.
A profession with a very strong female majority is particularly exposed to the effects of employment-family policies: not because women work less, but because hospital work organization systems were designed for full-time continuous careers. Shifted schedules, night shifts, weekend work: these constraints are manageable for a single adult or a couple without children. They become very difficult to sustain for a mother whose partner also works constrained hours, in a country where childcare in the evening and at night remains exceptional.
The shift to part-time work responds to concrete constraints: work organization designed for full-time continuous careers, where shifted schedules, night shifts, and weekend work become difficult to maintain for a mother whose partner also works constrained hours. This reduction in hours operates gradually, under the pressure of cumulative constraints. The patient waits for lack of available bed.
DREES observes that average total volume decreases after the first years of career. The forty-year-old nurse working at 70% has lost that 30% gradually, sometimes without conscious decision, under the pressure of cumulative constraints.
This phenomenon is not unique to France. It appears in all health systems with a high proportion of female caregivers. In France, hospital nurses’ pay is close to the national average salary, unlike the European Union average where it exceeds it by roughly 20%, which reduces their room to negotiate family trade-offs. Giving up a few hours costs less when hourly pay is low. It is a constrained calculation, not a choice.
This income inequality tied to professional structure is coupled with geographic inequality: services under the most strain, rural hospitals or under-resourced maternity wards, are also those where working conditions are most difficult and where childcare solutions are least accessible.
Maternity wards as a revealing indicator
Maternity wards concentrate the problem in its purest form. These are demanding services, with unpredictable schedules, long shifts and high emotional burden. They are also, precisely, the services where nurses and midwives are of child-bearing age. The overlap is brutal.
The Court of Audit reports that some maternity wards temporarily suspended obstetric activity in 2023 due to staffing tensions and difficulties maintaining permanent staff. The National Order of Nurses signals this in its annual reports. Some closures or suspensions can be linked to human resources difficulties, notably a deficit in care hours.
The medical desertification usually measured by general practitioners has its nursing equivalent, less visible because less publicized. A territory can have its full bed quotas and school places filled and still lack care because available staff work at half capacity.
The increase in training capacity was an important lever in the public response, among other measures for attractiveness and working conditions. It is a logical response to a truncated diagnosis. If the problem were a lack of graduates, training more would suffice. The problem is a lack of available hours across entire careers. Training more without acting to maintain hourly volume amounts to filling a leaking tank.
The inevitable constraints of 2040
Population aging in France is not speculative projection. People who will be 80 in 2040 are already 65 today. Their number is known. Their medical needs are modelable. Population aging increases nursing care needs.
With three years of study, places opened from 2027 would result at earliest in graduates by 2030, with hourly volume declining from the first years of motherhood onward.
Two trajectories are taking shape for the 2030-2040 decade, and public policy decisions made between 2026 and 2030 will determine which prevails.
The first option is continuation of the current model: increased graduate cohorts, managing part-time arrangements as they arise, occasional recourse to temporary staff and overtime. This trajectory manages the deficit according to current trends.
The second trajectory requires structural reform of nursing work organization. It starts from the observation that the decline of roughly a fifth in average total volume over this period is partially avoidable, provided causes rather than symptoms are addressed. This requires, according to leads identified by researchers and professional organizations, at least two things.
First, a guaranteed replacement mechanism in case of shift to part-time work. Some Scandinavian systems have such mechanisms and report better retention of available hourly volume.
Second, access to childcare solutions integrated into health establishments could support maintenance of available hourly volume. Several European hospitals have opened in-house childcare facilities, financed by pooling among establishments.
The signals to watch for evaluating the trajectory actually chosen are precise. DREES measures the change in average hourly volume per nurse between the fifth and tenth year of career: if this figure stops declining between 2027 and 2030, reform is taking effect. The rate of bed closures correlated with nursing hour deficits, tracked by establishment, is the other hard indicator. Beds reopening without net recruitment would mean that hourly volume per caregiver has increased, signaling a reversal.
Other countries’ attempts
France is not alone facing this problem. Aging health systems in Western Europe confront the same equation. Their responses diverge, and their results allow testing of hypotheses.
Germany relies heavily on international recruitment, particularly from Southern and Eastern Europe. This strategy raises its own tensions: it displaces the deficit to poorer countries without resolving it, as studies on health professional migration flows show. It also does not resolve the question of hourly volume across entire careers: an internationally recruited nurse faces the same family pressures once she settles and starts a family in Germany.
The Netherlands experimented with flexible schedule systems over the course of entire careers, with mechanisms for returning to full-time after the early childhood period. Results are encouraging on retention but assume a hospital HR organization that few French establishments have the administrative capacity to deploy today.
The United Kingdom attempted significant pay reform after Brexit, to compensate for the loss of European caregivers. The revaluation had a real effect on recruitment but limited effect on retention of mothers of young children, whose main problem was not salary but work organization.
What these experiences show is consistent with the French diagnosis: the deficit in nursing hours over average careers is an organizational problem as much as a pay problem, and it does not resolve through a single lever. Systems that achieved the best results combined pay revaluation, organizational flexibility, and childcare solutions. None found a shortcut.
The cost of waiting
Reform undertaken late would offer less time to train and place staff before care needs linked to aging reach their peak. People who will need nursing care in 2035 will be 80 or 85, with corresponding dependencies. If care hours are insufficient, their families absorb the burden or they wait. The human and economic cost is real and measurable in either case.
The question facing decision-makers between 2026 and 2030 is not technical. It is one of priority: do we address the problem now, when graduates trained today still have ten years ahead of them to work at full capacity, or do we wait until bed closures become politically untenable?
The formal answer will likely come from a national plan. The National Order of Nurses and Institut Montaigne have both formulated convergent recommendations on employment-family flexibility and guaranteed replacement. The analytical tools exist. Foreign experiences provide testable models. What is missing is less information than the will to address a problem whose most visible effects will only appear in ten years, and whose solutions cost now.
Sources
- Institut Montaigne – What are the challenges ahead for the French healthcare system?
- DREES – Directorate for Research, Studies, Evaluation and Statistics, data on health professions and nursing hourly volume
- National Order of Nurses – Annual reports on the state of the nursing profession in France
- ScienceDirect – Studies on employment dynamics in health professions with female majority (2025)
- Court of Audit – Reports on the situation of maternity wards and hospital capacity closures
