The NDIS Digital Twin: Developing a tool to examine how policy changes affect NDIS participants — The Association Specialists

The NDIS Digital Twin: Developing a tool to examine how policy changes affect NDIS participants (23844)

Marissa Shields 1 , Yi Yang 1 , Huixin Chen 1 , Priyanka Nair-Turkich 1 , Simon Mutch 1 , George Disney 1
  1. University of Melbourne, Parkville, VIC, Australia

Background: The National Disability Insurance Scheme (NDIS) is Australia’s largest social policy reform since Medicare and supports more than 700,000 Australians with disability. The scheme is currently facing substantial reforms, including eligibility changes and the introduction of foundational supports. These changes involve multi-billion-dollar decisions, yet the potential consequences of these choices are challenging to evaluate with traditional epidemiological or policy evaluation methods. This project aims to address this gap by developing a robust and transparent agent-based model to examine the potential impacts of policy change on the NDIS and people with disability who interact with the NDIS.

Methods: We describe the development of an agent-based ‘NDIS Digital Twin’ using the Overview, Design concepts, and Details (ODD) protocol. Using NDIS unit record data and mortality and birth datasets, a ‘population’ model was developed to generate a virtual population of NDIS participants, with each agent assigned demographic, disability, and NDIS participation characteristics. Model design concepts were evaluated for suitability through consistency checks and comparison with observed patterns. Two policy modules were then developed, using the virtual NDIS population to simulate impacts of hypothetical policy changes to a) eligibility criteria for people with psychosocial disability and b) employment supports.

Results: To date, the modeling framework has been implemented, key parameters and transition probabilities have been estimated, and calibration is underway to ensure sensible predictions of demographic processes and NDIS-related transitions (e.g., application acceptance). Preliminary outputs include processes for assessing prediction suitability, descriptive characteristics of the simulated NDIS population, and early learnings from testing hypothetical policy scenarios.

Conclusion: Using an agent-based modeling approach to create a ‘policy sandpit’ is a novel way to explore the potential impacts of policy changes prior to real-world implementation. Given the ongoing reform facing the NDIS, the NDIS Digital Twin offers a transparent tool to support policy development.