Supporting indigenous communities with analytics

Public hearings for the National Inquiry into Missing and Murdered Indigenous Women and Girls have finally begun. Though they make up only 4 percent of Canada’s female population, Canadian indigenous women and girls have accounted for an estimated 16 percent of the country’s female murder victims over the past 30 years.  

This grim statistic leaves little doubt of the need for additional resources to help prevent violence, identify risk factors, and increase access to health care for those most vulnerable.

That’s why earlier this year, SAS joined the Vulnerable Persons Project to help the Mohawk Council of Kahnawà:ke (MCK) develop an analytical approach to help identify those at the most risk. “This project is a perfect fit for SAS, a proud participant in the Data for Good movement,” said Cameron Dow, President, SAS Canada. “Using data to improve outcomes for at-risk indigenous women and children is truly a data for good story that SAS is honored to be a part of.”

By integrating data from case management systems with data from a variety of sources – social benefits, education, health care, law enforcement, criminal justice systems, etc. – SAS will work with representatives from Blackberry, Forest Green and other technology partners to securely apply advanced analytics to spot trends, uncover patterns and identify key relationships. The plan will enable community-run data collection and system management, emergency notification and crisis communications. It will also allow families to securely share sensitive records with law enforcement agencies and health care providers. Ultimately, the project will help to identify high-risk situations faster so case workers can intervene sooner, saving more lives and providing proper resources and support before situations turn dire.

The Indigenous Data and Analytics Forum

On June 7, 2017, the Indigenous Controlled Technology Forum in Ottawa brought First Nations Chiefs from across the country together with representatives from Blackberry, SAS, Forest Green and other technology partners to discuss this crisis and how data analytics technologies can be used to prevent violence and improve health outcomes in indigenous communities.

The panel discussed the challenges of working on data collection and analysis. How can indigenous communities improve data collection? How should data structure and data sharing be modified in these communities? How can the quality of data be used to improve outcomes for both communities and their residents? Collectively, the group will work with the indigenous community to:

• Get a comprehensive view of all available data for indigenous women and girls by applying data management techniques to integrate this data across all available sources.
• Apply data matching and link analysis to determine important relationships.
• Quantify the safety risk for indigenous women and girls using advanced analytical methods that will calculate overall risk scores for each woman and child, employing predictive models to address a broad range of vulnerabilities.
• Quickly alert caseworkers of important status changes. Software will monitor all available data sources and automatically alert case workers when changes occur that affect a woman or child’s risk score. Risk scores are continuously recalculated based on the most up-to-date information.

“Effective community leaders need current and accurate data to make decisions, allocate funding, and apply staff,” explains John Paul, Executive Director, Atlantic Policy Congress of First Nations Chiefs Secretariat. “In many cases, indigenous leaders have fewer administrators and IT staff to perform analysis than their municipal, provincial, or federal counterparts. This highlights the importance of using off-the-shelf solutions where other governments and private sector companies have paid for research and development.”

No level of abuse or neglect is acceptable. But it’s even more disturbing to learn that in many instances of abuse or neglect, information was available that could have helped identify high-risk situations before tragic outcomes occurred. Canada is not immune, and the cases of Alexandru Radita and Jeffrey Baldwin are sad reminders that critical missing data pieces can hold the key to a child’s well-being.

Unfortunately, caseworkers can’t readily access the information they need. This prevents them from effectively using all available information to continuously monitor each child’s overall safety risk. Making matters worse, budget constraints have overburdened personnel tasked with ensuring the safety of at-risk children, and large caseloads limit the amount of time they can allocate to each individual case.

SAS analyzed nearly six years of data on children that had some contact with the Florida Department of Children and Families. The SAS analysis considered factors such as prior removals because of sexual abuse or drug abuse, as well as physical or mental disabilities. The resulting five-year Child Fatality Trend Analysis is helping investigators better predict the needs of families in crisis. Some of the key findings include:

• Overall, child deaths within the agency are trending downward.
• Children who received prior service from the agency saw their odds of dying reduced by 90 percent.
• Children who experienced prior removal due to physical abuse were 14 times more likely to die.
• Children who experienced prior removal due to parental drug or alcohol abuse were 15 times more likely to die.

Case study: New Zealand Ministry of Social Development

Social welfare accounts for almost a quarter of New Zealand’s gross domestic product. The country’s largest government agency, the Ministry of Social Development (MSD), is charged with providing child protection and youth, family and employment services to the island nation’s 4.6 million citizens, a task that costs $22 billion a year. Research revealed some startling figures about the country’s social welfare system:

• One million New Zealanders relied on MSD in some way, shape or form for support.
• 13 percent of the working population is on an adult benefit, and many had been for more than a decade.
• Even more shocking, more than 70 percent of MSD’s liability was attributable to those who had entered the welfare system under the age of 20.

MSD’s strategy was to take an “investment approach” to achieve better social and fiscal outcomes by better targeting services to help New Zealanders become less reliant on the welfare system. MSD focused in particular on the most vulnerable — teen parents and young people unable to live with their families — developing a data model to estimate the risk of welfare dependency, matching and analyzing data across government agencies using SAS Data Management, and predicting the likelihood of entering adult welfare programs. This allowed MSD to offer targeted services — personal mentorship, teaching budgeting skills, providing more access to skills and training — to reduce their long-term benefit dependency.

The approach has paid dividends. Those who received the extra services moved on to adult benefits at the lowest rate in five years, and their employment rate rose by 9.3 percent.

“I now hear from single parents every week who are grateful for the support they receive from caseworkers … who are often the first to ask them what they want to do with their lives, and then help them find work,” Paula Bennett, then Minister of Social Development, said in 2014.

Analysis can help shift the approach to child welfare from mitigating tragedy to improving outcomes. The Vulnerable Persons Project hopes to achieve a similar experience for the indigenous women and children of Canada by leveraging data and insight to pinpoint services and support that can lead to the most positive outcomes for those most at risk.