Please use one from below list to access this site
The NDIS Demand Map is a new data analytics tool that provides information to help providers to grow in the NDIS.
Use the NDIS Demand Map to help understand the services and locations NDIS participants may choose by 2023.
The NDIS Demand Map provides an up to date forecast of the NDIS demand by postcode across Australia.
Use the map to find out:
The forecasts use de-identified NDIS and other data (census, Department of Social Services data) to predict what the NDIS will look like by 2023, when it is fully operating and the growth of participant numbers starts to ease.
The forecast will be updated quarterly based on new NDIS data as it becomes available.
It is important to note that this model is based on participant spending patterns to date. This pattern may change in the future if their preferences change.
The Demand Map provides a forecast range for future participant numbers, spending and workforce. This range reflects the inevitable uncertainty in making forecasts, particularly in areas where the NDIS has not been operating for long or areas with low populations.
We expect 90% of postcodes to fall within the reported ranges; but not necessarily the midpoint of the range.
Website users should not sum categories or sum totals across postcodes, as this will produce inaccurate results.
Users should make their own assessments when using this data.
We welcome any feedback you have on how we can improve the Demand Map. Please send your views through our feedback form.
In interpreting forecast ranges, Demand Map users should note:
The Demand Map uses a ‘Random Forest Regression’ statistical technique to forecast NDIS participation, spend and required workforce to deliver these services. The model draws on census data, NDIS administrative data and Department of Social Services data from over 2,500 postcodes in Australia to project the following dependent variables: number of participants in each postcode, and value of spend in each postcode (job estimates are inferred from these spending figures). Only summary statistics at a postcode level are used in modelling, such that individual data observations are not linked across sources.
To estimate the values of dependent variables, the model observes NDIS data from postcodes where the NDIS is currently rolling out and links these figures to underlying postcode demographic and aggregated Department of Social Services administrative data.
The model estimates relationships between these data sources and NDIS outcomes, by generating multiple decision tree models to assign dependent variable estimates to each postcode after considering demographic and other variables in those postcodes. The ranges produced for the model are based on the variation in predictions using these multiple decision trees.
To ensure robustness, modelling is evaluated out of sample through tenfold cross-validation. This procedure splits the available data where the NDIS is currently rolling out into ten subsamples, and then trains a model on nine subsamples, and evaluates performance against the remaining subsample. The procedure is repeated ten times, such that each subsample is tested against, and the average performance of the model is compared to a naïve mean prediction model across two criteria:
These forecasts have been produced by AlphaBeta Advisors Pty Ltd. Every effort has been made to provide the most current, correct and clearly expressed information possible on this site. Nonetheless, inadvertent errors can occur and applicable laws, rules and regulations may change. The information contained on this site is general and is not intended to serve as professional advice. No warranty is given in relation to the accuracy or reliability of any information. Users should not act or fail to act on the basis of information contained herein. Users should not rely on the information for any business, commercial or other purpose, and are strongly encouraged to seek professional advice concerning the information provided on this site before making any decision. Users should not rely on sum categories or sum totals across postcodes, as this will produce inaccurate results. Users should not use this tool for the purpose of re-identification. All contributors to this site disclaim all and any liability to any person or organisation in respect of anything, or in consequence of anything, done or omitted to be done by any person, organisation or other user in reliance, whether in whole or in part, upon any information contained herein.
No part of the text or graphics on this site may be reproduced or transmitted in any form or by any means, electronic or mechanical, including by photocopying, facsimile transmission, recording, re-keying or using any information, storage and retrieval system. Certain links on the site lead to resources located on servers maintained by third parties. As such, no representations are made as to the accuracy, currency or any other aspect of the information contained on such servers or the timely, accurate or complete transmittal of such information.