Outcomes of the Evaluation Phase
The Multi-Agency Data Integration Project (MADIP) is a partnership between Australian Government agencies. It demonstrates how to maximise the value of existing public sector data for policy analysis, research, and statistical purposes, in a safe and secure way. Now in an evaluation phase, the project has brought important national datasets together to look at patterns and trends to help agencies and analysts address complex policy and service delivery questions facing Australia.
The Australian Bureau of Statistics is the Accredited Integrating Authority for the project and is responsible for linking data, managing the integration process, and ensuring that the project data is strictly confidential and is used exclusively for policy analysis, research, and statistical purposes.
There are six agencies working together on the project:
- Australian Bureau of Statistics;
- Australian Taxation Office;
- Department of Education and Training;
- Department of Health;
- Department of Human Services; and
- Department of Social Services.
The MADIP combines existing data on Medicare benefit claims, government payments, and income tax with the 2011 Census to create a linked dataset that provides a high quality snapshot of Australia in 2011.
The benefits of the project include:
- Providing insight into the effectiveness of government policies, programs, and services and how these interact, to ensure that they are delivering value to the Australian community;
- Combining existing datasets to provide a more comprehensive picture of Australia and its people;
- Making better use of existing public data for policy analysis, research, and statistical purposes;
- Providing safe access to data for approved researchers and analysts;
- Facilitating better targeting of services (such as healthcare and early childhood services) to the people and communities who need them; and
- Enabling people and businesses to make more informed decisions.
The MADIP is for policy analysis, research, and statistical purposes only. It is not used for compliance purposes.
The internal practices, methodologies and technology protecting MADIP information are consistent with the Australian Government information and protective security standards. Data can only be accessed via highly secure systems, by approved researchers and analysts. Practical measures to safeguard the information include strong encryption of data in transit, restricted access to data on a need-to-know basis, and auditing of staff roles and use of MADIP data.
Personal information is supplied and stored separately from other data and is only used to bring the data together. Analysis is only performed on a file that does not contain personal information. The data does not identify individuals, but looks at patterns and trends among groups of people in the community.
Linkage is performed in a secure facility, and within that facility no person will ever be able to access both personal information (e.g. name, address, date of birth) and analytical information (e.g. occupation, income, health services use). This is known as the separation principle.
There are strong legislative protections in place to safeguard people's privacy. The MADIP complies with the Privacy Act 1988 and with the legislative requirements of each party. In particular, MADIP data is protected by the Census and Statistics Act 1905 , which ensures that no data is published or disseminated in a manner that is likely to enable the identification of a particular person or organisation. The MADIP is also conducted in accordance with the High Level Principles for Data Integration Involving Commonwealth Data for Statistical and Research Purposes.
OUTCOMES OF THE EVALUATION PHASE
Key achievements of the ‘learning by doing’ approach to governance and partnership taken in the evaluation phase of the MADIP include:
- Partnering to establish an enduring national integrated data resource;
- Streamlining governance frameworks within existing legislation to extract more value, increase efficiency, and manage risk;
- Building capacity through collaboration; and
- Learning from the experience of others to inform development of the project.
The project's technical achievements in the evaluation phase include the pioneering nature of the project, and the high linkage rates which have been achieved (as high as 96% between the administrative datasets). The evaluation phase has demonstrated that there is a sound basis for extending across time (longitudinally) and expanding to include datasets from across a broad range of sectors and jurisdictions.
- Improve efficiencies in government spending and better targeting resource allocation;
- Refine services and policies to achieve better outcomes;
- Evaluate key programs by investigating what works, how well, for what cost and under what circumstances; and
- Model the impact of potential policy changes on vulnerable populations or government expenditure.
The next steps are to complete the evaluation of the MADIP and build on the lessons from the evaluation phase to put in place arrangements that make better use of existing public data, safely and securely.
The evaluation of the MADIP has noted that additional data, particularly longitudinal data, would be extremely useful for policy analysis, research, and statistical purposes. It has also noted that there would be considerable benefit in broadening access to integrated public data in a manner that maintains the confidentiality of individuals.
1. Overview - a one-page summary of purpose, benefits and future of the project.
Download PDF - MADIP Overview.pdf
2. FAQ's - responses to frequently asked questions about the project.
3. Case studies - four case studies that highlight the policy value of analytical findings from MADIP data.