Appendix Six
Forecasting methodology
Workforce forecasting methodology
The Analytics and Forecasting group, National People Operations, Health New Zealand - Te Whatu
Ora has developed health workforce forecasting models for professions including almost all
regulated professions and some unregulated professions.
The workforce supply forecasting models are based on data on individual practitioners. Each
practitioner’s new entries, re-entries and exits are tracked based on annual changes in the work
history. Entry and exit rates are calculated for the group of practitioners in each five-year age band:
in the forecasting model they are moved between age bands as they age.
In other words, the forecasts are based on ‘rates tables’ which record the actual numbers entering
and leaving the workforce over recent years. The entry numbers include both NZ-trained and
internationally trained professionals and include those entering for the first time (new entries, with
their first annual practising certificate) and those coming back into the workforce after a break (re-
entries). Entries and exits in each age band are treated separately because those age-specific
patterns vary. For a ten-year forecast we have to allow that current workers will be ten years older
and their likelihood of leaving (or having already left) wil be different from now, and that new
entries and re-entries will be more likely to happen in certain age bands - in other words, the aim is
to forecast not just the total numbers but also the age structures of the future workforces.
Testing the model's forecasts against what has actual y happened in past years shows it to be 98 per
cent accurate for five-year projections for a large occupational group, namely all general
practitioners.
The basic algorithm and specific models have been reported in academic publications as fol ows:
• Jo, Emmanuel, Kimberly Mathis and Justin Goh, Forecasting future medical specialty
workforces supply with age distribution using health workforce annual practising certificate
data, Operations Research Society of New Zealand, 2017, 1-12.
http://orsnz.org.nz/conf51/wp-content/uploads/sites/3/2017/12/ORSNZ17_JoE.pdf
• Seleq, Sam, Emmanuel Jo, Phillippa Poole, Tim Wilkinson, Fiona Hyland, Joy Rudland,
Antonia Verstappen and Warwick Bagg, The employment gap: the relationship between
medical student career choices and the future needs of the New Zealand medical workforce,
New Zealand Medical Journal, 29 November 2019, 52-59.
https://journal.nzma.org.nz/journal-articles/the-employment-gap-the-relationship-
between-medical-student-career-choices-and-the-future-needs-of-the-new-zealand-
medical-workforce
• Dunn, Alex, Shaun Costello, Fiona Imlach, Emmanuel Jo, Jason Gurney, Rose Simpson and
Diana Sarfati. Using national data to model the New Zealand radiation oncology workforce.
Journal of Medical Imaging and Radiation Oncology 2022, 1-9.
https://doi.org/10.1111/1754-9485.13448
New Zealand Government Parliamentary Media Announcement about the model:
• beehive.govt.nz/release/new-modelling-helping-improve-workforce-planning
General assumptions for the forecasting models
The Analytics and Forecasting team’s standard (baseline) workforce forecasts are based on patterns
which have been evident in the last three or five years, and on projecting these patterns into the
future (the next ten years). These forecasts assume that current patterns of work wil continue, that
is:
• no changes in technology or models of care
• no changes in scopes of practice of other professions or additions of new professions which
could affect the practice of the model ed profession
• continuation of age-specific patterns of new entry to each profession or specialty, and re-
entries after periods of absence
• continuation of current age-specific exit rates
• new entrants include those who have completed training in New Zealand and ful y-qualified
overseas-trained professionals registered for the first time in New Zealand, and the model
assumes that the historic patterns of entry of the two groups continue
