Modelling human resources policies with Markov models: an illustration with the South African nursing labour market

This article proposes a modelling framework to simulate and assess the immediate and long-term effects of policy interventions to attract and retain nurses in rural areas. Specifically, we use a Markov model to model the dynamics of movements of health care workers in a professional labour market. A model is developed to simulate the movements of South African nurses between different segments of the labour market over time. The model builds upon a series of assumptions that are stated in details, and uses predictions generated by discrete choice experiments. The results demonstrate the ability of Markov models to model the effects of human resources policy interventions in the short and long run. They highlight the effects of time on the effectiveness of some potential policy interventions, whose immediate positive effects can be eroded as different adverse effects appear. Despite its complexity, this innovative method provides a transparent and useful tool to inform the design of policies to address rural staff shortages.