How States Use Data to Inform Decisions

Overview

Every day, state governments make decisions that affect the lives of their citizens. Legislators and governors determine which policies to enact and what public problems to address. State agencies establish how programs should be run and where budget dollars are best spent, as well as who qualifies for government assistance.

To effectively serve the public, state officials at every level of government are tasked with ensuring that these daily decisions are prudent and well-informed. Consequently, states are increasingly turning to administrative data, or information—such as vital records, college enrollment data, and Medicaid utilization statistics, collected and maintained primarily for the routine management of programs and services—to make strategic data-informed decisions. This information can include any data that are necessary to implement and oversee a program, such as demographics, outcomes, and enrollment details.

While researchers have explored the use of administrative data in various areas of state government (for example, identifying frequent users of emergency services), little has been published on this trend more broadly. As state leaders seek to harness data in innovative ways, what common themes, noteworthy successes, and notable challenges have the 50 states experienced across a broad cross section of issue areas? To address these questions, The Pew Charitable Trusts interviewed state leaders across the U.S. in 2016 and reviewed relevant laws, documents, and policies in all 50 states. This report is the culmination of that research, and the first comprehensive overview of how data is being utilized in all 50 states.

States traditionally use administrative data to prepare annual reports showing how funds were spent and the impact of a particular program, to demonstrate transparency in describing what a state agency does, and to comply with performance measures set by the federal government, state legislature, governor, or an agency.

More recently, states have begun harnessing existing information through data analytics—procedures that review data to identify meaningful information and correlations. Such efforts open up critical new opportunities for governments to make effective decisions. Analysts can uncover important insights by employing techniques such as integrating and cross-referencing data sets, undertaking calculations to show trends, finding correlations between various factors, running statistical experiments, mapping geographical data to show areas of high activity, and visualizing data in charts and graphs. Additionally, data analytics can reveal the root cause of a persistent issue, diagnose breakdowns in a system, highlight obstacles, and predict future phenomena, allowing state leaders to be better informed in their approach to a problem and make more strategic decisions.

Using data collected from interviews with more than 350 state officials, this study highlights ways in which some government leaders have employed sophisticated data analytics, beyond traditional uses of administrative data, to accomplish the following:

• Craft policy responses to complex problems. In Massachusetts, policymakers sought strategies to reduce deaths from drug overdoses. The Department of Public Health led the effort to integrate 10 data sets from five agencies. Findings from this analysis showed that illegally obtained drugs caused more deaths than prescribed opioid medications and that individuals released from prison were 56 times more likely to die from an overdose than are members of the public. As a result of these and additional findings, Massachusetts passed Chapter 52 in 2016 to address the opioid crisis’ contributing factors through treatment, education, and prevention.
• Improve service delivery. Missouri health officials believed that analyzing Medicaid claims data could improve patient outcomes. To that end, they added claims information into an algorithm that factored in whether an individual frequently used emergency services and had a chronic health condition. Because these often high-risk Medicaid patients could benefit from more intensive, patient-centered health care, officials enrolled them into “health homes”—a type of patient-centered care delivery in which high-cost patients are assigned caseworkers who help coordinate the providers caring for them. The result was improved clinical outcomes: For example, within a year 25 percent of participating diabetic patients with high blood sugar experienced normal levels.
• Manage existing resources. In Delaware, state leaders explored ways to use the state’s vehicle fleet more efficiently. After installing GPS devices, they received real-time data, such as unauthorized vehicle use and excessive idle time. Between 2008-12, Delaware’s analysis of the GPS data allowed managers to better allocate vehicles across the state, saving $874,000 by reducing the miles driven and fuel used.
• Examine policy and program effectiveness. The District of Columbia performed a randomized controlled trial using administrative data to assess how to most effectively boost participation in its Summer Youth Employment Program. The trial revealed the effect of various strategies on program attendance and provided administrators with the necessary information to choose the most effective course of action.

Such innovative uses of administrative data remain relatively rare, and making them more prevalent can require states to clear a number of hurdles. Budget pressures often leave state agencies struggling to maintain funding for research and analysis, and result in salaries that make it difficult to retain staff skilled in data matching and complex analyses. Data quality necessary to support detailed analyses is usually uneven at best. Information sharing—drawing on data from multiple agencies—requires agreements and compliance with privacy protections. Legal thickets such as these are difficult and time-consuming for states to navigate. And above all, a government’s day-to-day struggles to absorb pressure to cut their budgets, respond to the latest crisis in the news, and accommodate requests from lawmakers and the governor’s office leave little bandwidth for the level of complex analyses contemplated here. Through this research, the authors identified five key actions state leaders could take to work through these challenges and maximize the value of administrative data at their disposal:

• Plan ahead by setting up guiding goals and structures. Implement well thought-out, coordinated approaches to using data by writing formal data strategies to guide future efforts; develop governance structures to inform data use and access while prioritizing privacy; and take stock of systems and perform an inventory of data sets.
• Build the capacity of stakeholders to effectively use data. Train existing employees to increase data literacy and analytics or hire skilled analysts; leverage partnerships with universities, vendors, and other organizations that have these skills and the capacity to do the work; and dedicate funding or secured grants to support data- driven projects.
• Ensure that quality data can be accessed and used by stakeholders. Work to improve data quality and accessibility among state government and approved stakeholders—such as research and nonprofit organizations—often by establishing data-sharing agreements, memorandums of understanding, and protocols among offices and agencies, or developing an enterprise, or statewide, view of data assets.
• Analyze data to create meaningful information. Utilize analytical techniques to extract information from data; visualize and disseminate data in the form of charts, dashboards, and reports; use findings to inform, guide, or alter decisions.
• Sustain support for continued data efforts. Encourage leaders’ commitments to data-driven initiatives, enact legislation and policies supportive of data use, and create a culture that prioritizes data as a strategic asset to guide decision-making.

The authors found states that had implemented a combination, or even all five, of the above actions in different policy areas. But no state has managed to apply these actions to a broad range of government agencies and achieve across-the-board improvements in how it develops policy, delivers services, manages its resources, and evaluates existing programs. The next frontier for state governments will be moving from the narrow, targeted use of data analytics to its comprehensive application across policy areas.