How data sharing amplifies benefits programs
When information is shared across agencies, leaders can craft more equitable solutions and unlock the full potential of available resources.
To help state and local government leaders take advantage of untapped funds from social benefits programs, a new resource shows how agencies can leverage data sharing to improve agency efficiency and streamline families’ access to vital assistance.
Data sharing is an underutilized tool in government, and implementation often faces challenges, said Alissa Weiss, the director of government innovation at Benefits Data Trust (BDT), a national nonprofit dedicated to helping government connect residents to benefits systems. For instance, agency benefit systems may operate on different software or tech platforms or house data in incompatible formats or in different locations, making it difficult for leaders to share data and collaborate.
When information remains siloed, millions of residents may lose the chance to enroll in assistive programs, Weiss said. As a result, more than $80 billion in benefits goes unspent, according to BDT. To combat the deficit, BDT recently released “Data Sharing to Build Effective and Efficient Benefits Systems: A Playbook for State and Local Agencies,” a report that highlights key practices and resources for leaders looking to improve their data-sharing efforts.
Laying out the groundwork
Awareness of legal burdens to data usage is a key starting point, according to Weiss and playbook co-authors Jeneé Saffold and Bridget Gibbons Straughan. For example, if an agency wants to send outreach messages via text, leaders should first be sure they are permitted to use cellphone numbers to prevent the upending of plans, Weiss said.
Local government leaders should be aware of the applicable state and federal laws, especially concerning data security and privacy, said Tayyab Walker, an associate commissioner of New York City’s Office of Technology and Innovation. BDT and New York City began partnering in 2014, facilitating nearly 55,000 enrollments for benefits programs since then, Weiss said.
The playbook introduces important data concepts such as “privacy-enhancing technologies” and “privacy-preserving record linkage” so that government leaders can gain a better understanding of data practices. More specifically, the playbook suggests agencies identify the data fields needed to conduct appropriate interventions, analysis and measurements of success. The playbook walks leaders through the process of developing a data sharing agreement that governs what data will be shared and how. An agreement should also outline how information will be protected and should be “written in a way that allows for ongoing operational adjustments,” the playbook states. The playbook provides examples of data agreements for leaders’ reference.
Walker emphasized the importance of maintaining a catalog of data agreements and data documentation for future sharing projects. If an agency has precedent for approved data sharing or information about where available data resides, “you may not have to start from scratch,” he said, which becomes a time-saving measure for staff.
Putting data to use
Demographic and geographic data is vital for agencies advancing equity-focused work, according to the playbook. This data provides insights on residents’ disparities in access to benefits so government can build equitable strategies to provide support to the community, Weiss said. For example, if data shows a community has a large Spanish-speaking population, then leaders can adjust outreach efforts to ensure those individuals receive language-appropriate education about access to services.
For agencies with tech resources to leverage, the playbook suggests that automated data transfer methods can reduce the manpower needed to interact with data, improving efficiency and privacy considerations.
One potential option is application programming interfaces, which enable information sharing without the need to exchange full datasets, according to the playbook. For example, APIs can identify an individual who may be enrolled in the Supplemental Nutrition Assistance Program and is eligible but not participating in the Special Supplemental Nutrition Program for Women, Infants and Children.
Hashed data linkage is another way agencies can share subscriber data while shielding personally identifiable information. For instance, the California Department of Social Services partnered with the California Policy Lab in 2021 to conduct a data linkage project meant to measure how many residents receiving safety-net benefits were at risk for not receiving additional services such as stimulus payments, earned income tax credit or child tax credit. By linking a dataset with safety-net enrollment data and another containing tax filing data, officials were able to identify and focus resources on residents who were not receiving additional benefits because they had not filed state returns.
“The data elements might be siloed, but the individuals behind the data are not,” Weiss said, so breaking down data silos enables the delivery of more comprehensive care to those in need and access to benefits funds that may have gone untouched otherwise.
This story was updated Jan. 20 to clarify that the playbook provides instructions to create a data sharing agreement rather than a recommendation to do so.