On this page

Summary

Just as in 2023/24, our workforce data is not high quality. District health boards had different ways of collecting and reporting workforce data. We are still living with these disparate systems and will continue to for some years yet.

This means there will be some degree of inaccuracy in our data – but it is the best data we have available and demonstrates the pressures we will face if we do not work differently. Our shortage estimates have assumptions built into them:

  • Most importantly – these estimates are not targets. We will not be able to achieve the growth they identify in full; both because the workforce is not available in New Zealand or globally, and because it would be unaffordable. But they demonstrate that we will need to work differently to address the strain our workforce faces.
  • In most areas we only have direct visibility of the health workforce that work for Health NZ. We can access some information on the wider workforce from regulators, but the information gathered by each one is different. In some areas partners in the primary and community sector have helped us build more accurate estimates, which is hugely helpful.

In some cases, including primary and community data has resulted in significant changes in our estimates of shortage since 2023 – notably, in pharmacy. This is not necessarily because our pharmacy workforce is in a worse position, but because we have improved our data.

  • We assume that vacancy rates for the community sector are the same as those in our hospitals, where we don’t have better data. This is not always the case. This methodology also means that where labour flows disproportionately from the primary and community sector into Health NZ, our data may inaccurately signal an improvement in supply.
  • Our data and modelling assumes our health system will look the same tomorrow as it does today. We know this will not be true.
  • We do not account for “frictional” vacancies – which arise as people move about the health system between jobs, and because of gaps between people retiring or leaving and new people starting. Frictional vacancies will probably always account for an average 3-4% vacancy rate even in a fully staffed system. This factor drives overestimation of true vacancy rates.

Over the next three years, we will work to improve our data and build more accurate models of current and future need.

There are some areas where we’ve improved our estimates and data since 2023/24:

  • We have drawn on expertise from our community pharmacy workforce, and the mental health and addictions sector, to improve our modelling of primary and community shortages in these sectors.
  • We have layered intelligence from frontline staff and leaders on top of our data to better understand where our data might over- or under-state gaps, and to better understand the relationship between workforce shortages and service pressures.
  • We have in places adopted different internal models to better estimate workforce need.

These factors mean that figures from 2023/24 and this Plan are not always directly comparable. We expect this to stabilise in future years; though there will continue to be pockets of incomparability as we improve data for various workforces (e.g. by incorporating better estimates of primary and community shortage).

Data sources

We have used three key data sources to develop these estimates:

  • Data on our Health NZ district employed health workforce – including information about how many roles are currently vacant (vacancy rates). 
  • Data gathered by the regulators of our health professions and from professional bodies (including Medical Colleges), whose professions are required to have annual practicing certificates. 
  • Data from partners in the primary and community sector.

We have extrapolated from this data to identify estimates of current workforce need. Assumptions underpin each analysis, including:

Current shortage

  • All data is presented as full-time equivalents (FTE).
  • Our data assumes that vacancy rates are a proxy for unmet workforce demand – which we know is an imperfect assumption.
  • Where we don’t have specific primary and community data, we assume the workforce profile is similar to Health NZ’s – and extrapolate using Health NZ vacancy rates and regulatory or professional body data on total workforce size (latest available).

2033 shortage 

  • Our data assumes that there will not be any unanticipated changes in population demand, technology, infrastructure or models of care (how we care for people, including the mix of professionals who provide that care) – which we know will not be the case.
  • The impacts of the actions in this Plan, including addressing current pressures, are not modelled in.

We are working to improve this data for future use. 

Health Workforce Information Programme

The Health Workforce Information Programme (HWIP) is the source of this Plan’s data on Health NZ’s workforce, collected from across Health NZ districts. HWIP collects data from districts as at the end of each quarter; this Plan uses data as at 31 March 2024, for seasonal consistency with the Health Workforce Plan 2023/24.

Data excludes casuals, contractors, and people on parental leave or leave without pay. All FTE values are contracted, funded FTE (where 1 FTE equals 2,086 hours per year). This means that overtime, for example, is not included in FTE calculations.

Headcount, FTE and vacancy data is reported as an aggregate of district payroll data. This excludes Health NZ national offices; for the purposes of the Plan, we do not think this results in a material exclusion of workforces covered by the Plan.

Canterbury and West Coast districts do not currently report vacancy data. In lieu of such reporting, we have established estimated vacancy rates for these districts based on the national average, to avoid underreporting. This means that in some cases, the vacancy rates reported in the Plan may exceed those reported in other Health NZ sources which do not include an extrapolated vacancy factor.

Vacancy rates are calculated as the vacant FTE divided by the sum of the vacant and employed (contracted) FTE. The definition of a vacancy for the purposes of the HWIP collection is as follows:

  1. An unfilled position that has funding allocated and will be actively recruited for within the next six months.
  2. It is a permanent position that is part of the FTE allocation (if applicable).
  3. Where a vacancy exists, it remains a vacancy when temporarily filled.

HWIP data extracted for this Plan is primarily segmented by Australian and New Zealand Standard Classification of Occupations (ANZSCO) code, where ANZSCO codes align to Health NZ workforce segmentation; but where required, job title searches are used to differentiate workforces.

2033 forecasts

Our 2033 forecasts use Health Workforce Analytics and Intelligence forecasts of workforce demand, one decade from the point of calculation. These forecasts use an algorithm to forecast workforce supply based on recent age-specific rates of entry, re-entry and exit for occupational groups – designed to model changes in patterns of work as our workforce ages.

These supply forecasts use data collected from responsible authorities, mandated under the Health Practitioners Competence Assurance Act 2003. Data is also collected from select other sources, including professional bodies for key professions where those bodies shape training volumes (such as medical colleges, and clinical psychologists), and from medical schools on medical student volumes. Data is generally collected annually.

The demand forecasts in this model use estimated changes in population demand for health services, measured as the change in the size of a relevant population group for each service; for example, the change in the size of the population aged four or younger for paediatric services. The relevant population group identified for most services is the 60+ population group, to account for the impact of aging in our population as best as possible using this methodology.

Because of this methodology, this data does not:

  • Cover all workforces; workforces which do not have responsible authorities, or with only recently established authorities (such as paramedicine) are generally unable to be analysed using this model.
  • Cover all medical sub-specialties; in some areas (such as internal medicine) we are only able to disaggregate our medical workforce projections at these more general levels.
  • Account for potential changes in workforce supply other than due to changes in the age composition of the workforce – for example, it makes no bespoke assumptions about cultural changes in work patterns.
  • Account for potential changes in workforce demand other than due to population growth – for example, it makes no bespoke assumptions about changes to models of care, technology or specific growth in care volumes for specific conditions.

Primary and community data

Pharmacy

Estimates of vacancy rates for community pharmacists are sourced from survey data, courtesy of collection by the Pharmacy Guild as of March/April 2024.

This analysis makes use of survey data gathered from ~10% of community pharmacies, extrapolated as a representative sample of the wider community pharmacy sector. We are unable to validate that the survey is representative, but we are confident this represents a better estimate than is available using Health NZ data alone.

Mental health and addictions

Estimates of the size and vacancy rates for mental health and addictions workforces for the primary and community sector are sourced from a blend of Te Pou More Than Numbers workforce data, available online here: More than numbers workforce data - Te Pou for the primary and community mental health workforce, and Health NZ data for significant hospital and specialist workforces employed by Health NZ. This data was published in aggregate in October 2023, but makes use of data collected earlier in some instances.

Specific caveats and notes

Nursing workforce estimates

Our FTE estimates of current nursing gaps are not based on Health NZ data like our other major workforce estimates. This is because we do not consider that this figure, extrapolated to the whole sector from Health NZ vacancy rates, are an accurate representation of the number of nurses we are currently short as a nation.

This is because of how our method for estimating vacancies interacts with the current state of our nursing workforce.

To determine an estimate of health system-wide shortage, we extrapolate Health NZ’s vacancy rate to the total number of health professionals working in New Zealand, based on data from responsible authorities and / or professional bodies. This is a valid approach where we have unmet health workforce demand; but because we have filled generalist nursing vacancies and likely have more generalist nurses than we need in New Zealand today, these greater numbers of nurses artificially amplify estimated shortages using our methodology.

While for most workforces this does not invalidate our methodology as a best estimate, it does so for nursing in 2024.

This produces a clearly inaccurate result: an estimate of ~4,500FTE more nurses required today. Given that we have added around 3,000FTE of nurses just within Health NZ, this is not a realistic estimate of current nursing need. We have therefore adopted an alternative estimate.