The case for cross-agency data sharing: Unlocking government's hidden potential

Agency leaders are stewards of increasingly valuable citizen data. As such, they face a growing obligation to use this data to drive maximum value for citizens — increasing program efficacy, fighting fraud, and driving efficiency. 

Yet many agencies operate in isolation, unable to access the insights of neighboring departments. This fragmentation creates missed opportunities that extend far beyond administrative efficiencies.

But cross-agency data-sharing isn't a pipe dream. Across the U.S., governments are already leveraging shared data to boost wages through targeted workforce development, prevent disease outbreaks through coordinated public health responses, and partner with community-based organizations to move people out of poverty.

Proven Success Stories: Learning From Implementation Leaders 

Real-world examples show how governments are creating value through strategic data sharing. 

Pivot, Indiana's workforce recommendation engine, showcases the potential for longitudinal data sharing in workforce development. Pivot integrates education, workforce and economic data to provide personalized guidance that transforms abstract statistics into meaningful career recommendations for the unemployed workforce. 

Resulting wage data highlights its success: users who accepted Pivot's top recommended job saw wage increases of $3.98 per hour compared to just $1.42 per hour for nonparticipants (average wages are $18.29/hour for UI claimants). 

That translates to more than $8,000 in annual wage increases. This wage increase generates approximately $15,000 in total economic activity per worker annually through increased consumption, reduced public assistance costs, and productivity improvements. 

The fiscal impact is equally significant: for every 100,000 workers experiencing similar wage gains, states can expect roughly $280-300 million in new tax receipts from income, payroll, sales, and corporate taxes. Beyond economic benefits, higher wages correlate with measurable social outcomes: a 10% wage increase can reduce crime by between 2 and 6% in urban settings, while also improving health outcomes and workforce stability.

The system's success demonstrates how longitudinal data sharing can create direct value for citizens — putting them in the best possible position to optimize their future by unlocking insight on labor market dynamics and helping them successfully navigate a complex landscape of programs and services. 

Colorado's Social Health Information Exchange represents a sophisticated approach to improving patient outcomes through connecting health and human service providers, community-based organizations, and public health agencies. 

This system enables healthcare providers to deliver better care by providing crucial social context about their patients — information that directly supports the strong incentives providers face to improve patient outcomes. The exchange also simplifies navigation of complex behavioral referral systems, particularly valuable given high care coordinator turnover. 

It helps coordinators understand the entire network of potential providers and their proven efficacy in similar circumstances, enabling them to make the right referrals to meet each patient's specific needs. By providing more comprehensive support for individuals with complex health and social needs, the system demonstrates how healthcare data can be responsibly shared across organizational boundaries while maintaining strict privacy protections and regulatory compliance. 

United Way of Central Indiana, through the community-based organizations it supports, acts as a vital component to Indiana’s social services landscape. It grapples with the same challenge that governments often do: it is virtually blind to the positive impact it creates for the families it serves — making it exceedingly challenging to optimize programs and services for future families in need.

Through a novel privacy-protecting data sharing agreement, it was able to effectively leverage key education, workforce, and outcomes data to measure impact. The data-sharing partnership with Indiana's Management Performance Hub has allowed the UWCI to track real outcomes at the individual level while maintaining strict data privacy standards. The data allowed UWCI to track income changes among program participants, a critical component of their goal to move 10,000 families out of poverty. 

In recent funding rounds, UWCI reallocated nearly $10M in grants to reward organizations with proven results. Further, longitudinal analysis revealed that participants staying in programs 4+ years saw significantly higher income gains, prompting UWCI to redesign their approach to emphasize long-term support for the families it serves today.

AI: Amplifying What's Possible 

Artificial intelligence is lowering the bar for success as the stakes for effective data sharing rise. Modern machine learning and AI techniques can dramatically improve the value and usability of imperfect data. Unstructured data like narrative text, images, and video were largely unusable for analysis at scale until recently, making them highly valuable data assets. 

Beyond improving data usability, AI systems excel at pattern recognition in large datasets. They can identify subtle relationships across multiple data sources that would be nearly impossible for human analysts to detect, potentially revealing new insights about program effectiveness, citizen needs and intervention opportunities.  

The scale advantages are particularly compelling for resource-constrained agencies. AI can process and analyze years of cross-agency data in hours rather than months, enabling real-time program adjustments that were previously impossible. This means social workers can receive immediate alerts when multiple risk factors align for a family, or workforce agencies can identify emerging labor market trends as they happen rather than discovering them in annual reports.  

Perhaps most importantly, AI is enabling proactive decision-making. Instead of waiting for problems to manifest, agencies can identify and address risk factors before they escalate. This shift from crisis response to prevention not only improves citizen outcomes, but also reduces long-term costs. When agencies can predict which job training programs will be most effective for specific populations, or which neighborhoods need additional health resources, they can allocate limited resources with precision. 

Emerging AI technologies also offer new possibilities for privacy-preserving analysis, including federated learning approaches that enable agencies to benefit from shared models without sharing raw data, and synthetic data generation techniques that can support research and development while protecting individual privacy. 

The Path Forward 

Governments face a fundamental obligation to effectively steward the data they collect in service of citizens. This obligation extends beyond compliance with regulatory requirements. Citizens deserve to have their data used to deliver measurably better services. The future of effective government depends on breaking down the barriers that prevent agencies from leveraging their collective data resources. 

The agencies that embrace this challenge will define the next generation of data-driven public service delivery. The question isn't whether cross-agency data sharing is possible—the success stories prove it is. It’s how to navigate the legal, technical, and political challenges that stand between vision and implementation.

John Roach is the president of Resultant, a leading consulting firm shaping how local, state and federal government agencies use data.