Let’s Not Digitize Our Blind Spots and Biases: Hidden Cost of Poor Domestic Violence Data in Child Welfare Systems

By David Mandel, CEO and Founder, Safe & Together Institute 

As child welfare systems across the United States and around the globe modernize their data infrastructure, there's a critical conversation we need to have about the quality of domestic violence information being collected and integrated. Poor data quality in this crucial area undermines the effectiveness of even the most sophisticated child welfare information systems, also known as CCWIS, and can perpetuate harmful practices that put families at risk.

Let's examine how inadequate domestic violence data impacts key aspects of child welfare work.

Telling the Complete Story of Children and Families

When domestic violence data lacks depth and nuance, we miss crucial elements in a family's story. Consider a case where only surface-level incidents are recorded, like arrest history, without capturing the perpetrator's actual pattern of behaviors, including coercive control, financial abuse, or manipulation of children. The system might capture isolated incidents of physical violence or "conflicts" rather than recognizing an ongoing pattern of abuse that fundamentally shapes family dynamics and child development.

Similarly, the absence of comprehensive information about a survivor's protective efforts—from subtle strategies like maintaining routines for children to more visible actions like seeking orders of protection—can lead to mischaracterizing them as "failing to protect" rather than recognizing their ongoing work to shield their children from harm.

Safety, Permanency & Well-Being Decisions

Without high-quality domestic violence data, safety assessments become dangerously incomplete. When systems can't see how substance abuse often emerges as a survival response to trauma, or how mental health challenges may be direct results of abuse, they risk misattributing cause and effect. For example, what may be documented as the failure of an adult survivor to complete a substance abuse program may actually be the result of a perpetrator’s sabotage of her treatment. This can lead to intervention plans that address symptoms while leaving underlying abuse dynamics untouched.

Cultural contexts and community dynamics are particularly prone to being oversimplified or ignored when data systems aren't designed to capture them. A survivor's hesitation to engage with law enforcement might be recorded as "non-cooperation" without capturing how historical oppression, immigration status, or community pressures influence their decisions.

Service Needs Assessment and Resource Allocation from Data

Poor quality domestic violence data creates a cascade of problems in service planning. When systems can't accurately identify the prevalence and nature of domestic violence within their population, they can't properly forecast resource needs. This leads to:

• Underfunding of specialized domestic violence services

• Lack of training for staff in domestic violence dynamics

• Insufficient advocacy services for survivors

• Missing connections to culturally specific support services

The result? Resources are misallocated, and families don't get the specific help they need when they need it.

Financial and Operational Impact of Data

The "garbage in, garbage out" principle has real financial consequences. When domestic violence data is incomplete or misleading, agencies:

• Waste resources on interventions that don't address root causes

• Miss opportunities for early intervention

• Face increased costs from repeated crisis responses

• Struggle to demonstrate program effectiveness to secure funding

Managing by Data: The Amplification Risk

Perhaps most concerningly, as child welfare systems increasingly rely on data for decision-making, poor-quality domestic violence information doesn't just perpetuate existing problems—it risks amplifying them. Machine learning and predictive analytics, built on historical data that reflects biased practices, can systematize and scale harmful approaches to domestic violence cases.

Moving Forward: Breaking the Cycle

To prevent new CCWIS implementations from cementing bad practices, we must ensure they're designed with a sophisticated understanding of domestic violence dynamics. This means:

• Developing data models that capture the complex patterns of coercive control and abuse

• Creating fields to document survivor strengths and protective strategies

• Building in ways to record cultural and community contexts

• Establishing connections between domestic violence, substance abuse, and mental health data

• Implementing quality control measures specific to domestic violence information

The stakes are too high to accept poor-quality domestic violence data in our child welfare systems. As we invest in new technology, we must ensure it helps us see and respond to domestic violence more effectively, not just digitize our blind spots. Only then can we truly serve the safety and well-being of children and families affected by domestic violence.

What steps is your organization taking to ensure domestic violence data quality in child welfare work?