Data at Scale

The prevalence of data in the social sector means that government agencies and nonprofit providers, in theory, have a wealth of information at their disposal to understand what works and more effectively deploy resources to achieve better outcomes. Administrative data (data collected by governments or nonprofits as part of administering a specific program) in particular provides a cost-effective and accurate way to understand participant outcomes and break the silos between agencies serving similar populations who operate in different policy domains.

The practical reality is that providers and government agencies cannot easily access administrative datasets or do not have the expertise to analyze them. Thus, the promise of administrative data in improving outcomes of social programs often remains out of reach.

Our work with a cohort of government partners under the Administrative Data Pilot aims to catalyze progress towards solving this challenge. In January and February 2017, together with our project partners at the Stanford Center on Poverty and Inequality (CPI), Third Sector hosted an open competition to select three government agencies from across California, Oregon, and Washington state to receive services under our Administrative Data Pilot grant from the Social Innovation Fund, a federal program of the Corporation for National and Community Service. Services under the grant support government agencies in accessing and analyzing administrative data, and subsequently using data-driven insights to enter into outcomes-based contracts for social services.

Three issue areas or themes emerged from the deep and diverse pool of motivated governments who responded to this opportunity, and our final selections under the competition ultimately represented each of these themes. The issue areas that emerged were:

• Homelessness and Care Coordination
• Employment Services for Vulnerable Populations
• Child and Family Success

The emergence of these themes provides a unique framework for us to consider how this work can help catalyze larger-scale change in how governments and nonprofits access and utilize administrative data. Each of the selected projects provides a clear pathway for scaling the insights and unique solutions that will emerge from this work within their area of focus.

County of San Diego (CA) Health and Human Services Agency –  Homelessness and Care Coordination

• In San Diego, the County is participating in California’s Whole Person Care pilot effort, funded under a Medicaid Section 1115 waiver through the state Medi-Cal These pilot projects are targeted to high-utilizers of multiple health care systems who continue to have poor health outcomes. Eighteen counties in the state received grants under the pilot (totaling over $3 billion over the next four years), with most targeting the homeless or near-homeless population, so there is a clear conduit for our work with the county to be applicable to other jurisdictions in the state. The project scope for the service award also includes another local project, Project One for All, which is designed to improve outcomes for essentially the same population via the use of supportive housing and mental health services. The County is interested in understanding the comparative effectiveness of these programs and using the flexibility of the local program to further their use of outcomes-based contracting for providing homelessness services.

Santa Cruz (CA) County Human Services Department – Employment Services for Vulnerable Populations

• In Santa Cruz, the County is focusing on subsidized employment services under California’s Welfare-to-Work program, a component of the state Temporary Aid to Needy Families (TANF) program (CalWORKS). The County is developing a data system that they intend to use to greatly expand their ability to efficiently and comprehensively pay service providers based on attainment of long-term outcomes, rather than based on caseloads or other traditional metrics. CalWORKS represents over $5 billion in annual expenditures for the state of California, and Welfare-to-Work remains a program requirement for all TANF programs nationwide, meaning that project learnings have the potential to inform the future expenditure of significant government resources.

Washington State Department of Early Learning – Child and Family Success

• In Washington, the Department of Early Learning previously completed a Pay for Success feasibility assessment with Third Sector, and has taken tremendous steps to implement the recommendations and act on the insights generated through the feasibility assessment process. Additionally, a state blue ribbon commission recently recommended re-organizing several aspects of the early childhood and youth development services in the state; the current legislation authorizing this re-organization mandates that 100% of contracts and grants for the new agency must be outcomes-based, meaning that there is a clear pathway to implement and deploy administrative data for performance-based contracting as part of this grant.

The learning community operated under the grant provides one conduit for facilitating the cross-agency learning and catalytic change we hope will emerge, particularly within the issue area themes described above. As a mechanism for doing so, our intention is to include additional governments in the learning community, not only those selected for services under the Administrative Data Pilot. We will continue to seek out other opportunities to ensure that this work helps increase the capacity for government agencies to utilize administrative data in service of achieving better outcomes for populations in need in their communities.