We're excited about administrative data-sharing! Here's why

Behind the scenes at Powered by Data, we’ve been quietly working on a new project focusing on administrative data-sharing that we’re very eager to finally share with you!  

Okay. We know what some of you may be thinking: why are the folks at Powered by Data so enthusiastic about something that sounds so...dry? Well, 1) because we’re nerds, and 2) because we think that administrative data-sharing presents a very cool new avenue for tackling a range of pressing social issues. In this blog post, we’ll highlight how this kind of data-sharing can offer innovative solutions to problems that nonprofits are currently working hard—and sometimes struggling—to address.

Unique problems, shared solution

A non-profit working with recent offenders wants to better understand the impact its interventions are having on reducing the reincarceration rate of their program participants. Over the years, nonprofits often lose direct contact with participants, and so it is extremely difficult to track their long-term outcomes.

Survivors of interpersonal violence have complex service needs—often requiring access to a combination of housing, mental health, and social assistance services. They need to navigate a fragmented system of service providers, which can be a confusing process.

An organization focusing on First Nations child welfare wants to better understand how First Nations youth in care navigate the system; they need evidence to build a case for change. Currently, they have to file dozens of Access to Information requests, accessing data through First Nations child and family service agencies on a case-by-case basis.  It is slow, inefficient, and deeply frustrating.

These scenarios are markedly different, but administrative data-sharing could help address each of them in a unique way.  

But first: administrative data—what is it? Not survey data, not open data

Administrative data refers to records that government and social services keep on the people they serve—information collected for operational purposes. Examples could include: physician visit records, high school completion records, or tax returns.  Because this data is not collected for research purposes, administrative data is different from survey data (so if you’re thinking of census data—nope, that isn’t administrative data!) And since this data often consists of person-level records that contain private and sensitive information, to protect confidentiality, it cannot be made openly available. Administrative data isn’t open data, either.

Why are we interested in these kinds of government and social service records if they can’t be made open?  Although administrative data cannot be released openly, these datasets can be “linked” between government ministries. This means that previously discrete personal records can be joined up, resulting in a richer dataset. Data-linking could look like matching an individual’s health records with their education records; and then doing that for a whole set of people.

Administrative data can also be shared in an anonymized, aggregated format between government, nonprofits, or academic researchers. For example, imagine a youth education nonprofit receiving statistics on high school graduation rates of a cohort of interest—calculated from data accessed through education records. You may already be able to see how access to this kind of knowledge could be game-changing for organizations wanting to better understand the long-term outcomes of their service-users.

Opportunities of administrative data-sharing for the social sector

If we revisit our three opening scenarios, administrative data-sharing presents a unique solution to each issue.  We’re excited about the ways this could promote new opportunities for measuring program outcomes, providing more collaborative care in the social service system, and supporting evidence-based policy.

Outcomes Evaluation

In our first scenario, what if the nonprofit working with past offenders could know how many of their program participants were re-incarcerated five, or ten years down the line? This is already happening outside of Canada. The Justice Data Lab—a service run by the Ministry of Justice in the UK—helps access and analyze data about the reincarceration rate of these organizations’ service users, by accessing administrative data through the Police National Computer database. These findings allow organizations to develop a richer understanding of the program’s potential impact on reducing imprisonment.

The UK's Justice Data Lab helps nonprofits working with those recently incarcerated access and analyze data about the reoffending rate of these organizations’ service users.

Integrated Service Delivery

In our current siloed system of social service delivery, it can be exhausting for individuals with complex needs to access services in a comprehensive way.  Development of a shared administrative data infrastructure across different services could allow for more collaborative care, more streamlined referral processes, and increased service consistency.  This is already happening to some extent in Toronto, where Reconnect Community Health Services facilitates the integration of over 100,000 client records across community mental health, community addictions, and community support services in the city.

Evidence-Based Policy

Canada lacks a coordinated national system for collecting administrative data on children in care, making it difficult to get a full picture of First Nations youth navigating the various services in this system. The First Nations Child and Family Caring Society (led by Dr. Cindy Blackstock) have expressed a need for coordinated administrative data sharing in the child welfare system to support their policy education work. Administrative data-sharing could begin to address the siloed nature of child welfare services, which currently acts as a barrier to understanding and addressing the overrepresentation of Indigenous youth in care.

Is this kind of data-sharing happening in Canada yet?

Administrative data-sharing is already happening to a limited extent in Canada—Alberta’s Child and Youth Data Lab, or the Research Data Centres by Statistics Canada—are some examples. However, despite the wide range of opportunities offered by this kind of data-sharing, in general, there are policies in place that prohibit the linking of administrative data across government databases here. This gets even more complicated when we consider that services in Canada are divided across provincial and federal levels. And of course, any effort to change these policies would need to account for potential risks around privacy, digital security, and data-driven inequities.   

So far, there has been no coordinated policy agenda for increasing social impact through administrative data use in Canada—but we’re excited to play a role in potentially changing that.  In the coming weeks, we’ll be sharing our plan for researching and co-creating a policy agenda around administrative data sharing with a diverse coalition of key stakeholders in Ontario’s social sector. We’ll also be writing about some of the ethical questions and risks associated with this kind of data-sharing—and how we hope to address them through a coalition-building process anchored in equity and inclusion. Stay tuned!

Interested in learning more, in the meantime? Download a PDF of our full issue brief on administrative data.

Powered by Data is grateful to the Ontario Trillium Foundation for their ongoing support and partnership in developing this initiative.