EdSystems is releasing a new cost model tool to support states in the development of their longitudinal data systems.
Based on interviews with a variety of SLDS and data organization owners across the country, the tool includes a narrative context for the model as well as a spreadsheet that practitioners can utilize to customize a model budget.
Variation of State Longitudinal Data Systems
What is a state longitudinal data system (SLDS)? There isn’t one answer to this question. SLDSs can be systems for supporting analytics by agency staff or deep research efforts. In some states, the goals of the SLDS are to provide data to the public, while in others, the focus is on legislatures and policy staff. Some SLDSs are focused on K-12 and higher education, some include workforce data, and others include data from health and human services, public safety, and the secretary of state’s office (such as driver’s license data) and also provide insight into early childhood education.
Because SLDSs take on such various goals, the organizations that support them are also quite different. EdSystems has connected with 11 different data organizations to understand their services and the commensurate cost over the last six months. This included nine SLDS organizations, a performance management office, and a university policy lab that works closely with state agencies. Critically, we found that there is not one right approach for an SLDS, but states need to align their organizations with goals and their resources with the capacity required from the organizations. As data needs grow, so should the organizations tasked with managing, organizing, and using the data.
What is Longitudinal Data?
Longitudinal data track information from the same people, or samples, across time, which allows researchers to measure change. Understanding longitudinal data and outcomes is a critical component in determining the efficacy of taxpayer-funded programs and interventions. To support that research, many states have launched longitudinal data systems in recent decades. Learn more about Illinois’ Longitudinal Data System.
Cost Modeling for an SLDS
The cost model aims to avoid a cookie-cutter answer to funding an SLDS. The organizations we spoke to ranged in cost from around $500,000 to $15M annually. Some support data linking for research, while others ensure that various stakeholders can use data. In some states, data management (cleaning data and preparing it for research) is centralized; in others, data management occurs at the agencies that provide data. Some teams have their own legal staff, and some leverage the legal staff of either a parent or sister agency. Similarly, the management of the databases and linking technologies can be done in-house or might be the provenance of a statewide technology department. How states account for these costs can be very different, meaning the budgets also look very different. Based on our findings, we believe states should avoid hidden costs and be cognizant of the roles and responsibilities necessary for their data organization to function well. This means including costs for functions such as legal and communications in their budgets. Critically, any technology costs that are outsourced or provided by a state technology agency should also be included in budgets. Without these elements, a full understanding of the funding necessary for ongoing operations may be underestimated.
Supporting Continuous Improvement of SLDSs
Each data organization should be developed and designed to meet the needs of a particular state, carefully considering questions such as: How involved should the organization be in policy conversations? How do you plan to develop agency and cross-agency data tools? Does your state need a research enclave, a protected online space for research activities, or will you use a third-party environment like the Administrative Data Research Facility? A collective understanding by state agencies of these questions allows SLDSs to go beyond just collecting data and allows them to become centers of excellence that can help support agencies on their data journeys.
As many SLDSs launched in the 2010s, these data organizations are maturing into “adulthood.” The question for states is whether they are prepared to use their data fundamentally differently. While it is critically important that states maintain the privacy and security of the data they are privileged to collect, it is also the case that states owe it to citizens to use their data more effectively. Getting this balance correct is critical, as too many states cannot do the deep research necessary to inform policy. To that end, the intent of the new cost model and brief is to help foster conversation about states’ choices and the resources necessary to support those choices.
Building More Inclusive Manufacturing Pathways: What We’re Learning from Schools
Learn More »