Using Data To Improve Policy Decisions

Every day, human services professionals in the public sector make complex policy decisions that affect citizens, such as how to improve service delivery, allocate budget dollars, and respond to crises. Increasingly, they use insights gleaned from massive amounts of data—originally collected by governments for reporting purposes—to make strategic decisions.

Researchers at The Pew Charitable Trusts¹ published a report in February 2018 examining how state governments are taking advantage of data analytics to improve outcomes for their citizens. Based on interviews with more than 350 officials in all 50 states, the report, “How States Use Data to Inform Decisions,”² also describes the biggest challenges agencies face in using data effectively and offers strategies policymakers have used to successfully overcome those challenges.

Governments Are Using Data More Effectively

Governments already collect a large amount of administrative data, mostly for routine reporting and compliance purposes. And states have recently begun finding more advanced and useful ways of leveraging these data to better allocate resources, measure success, and improve the efficiency of government-administered programs.

For example, during Hurricane Harvey last September, Texas saved millions of dollars—and, potentially, thousands of lives—because the state had built sophisticated new data capabilities long before the storm hit. These systems helped to identify safe evacuation routes, helicopter landing zones, and places where people needed rescuing during the storm.

This effort began in 2015, when the Texas Department of Information Resources and the Texas Natural Resources Information System, working with Google, started a project that uses geographic information systems to combine data from across state agencies and map it in a way that is easy to understand. In the midst of the hurricane, the Google Imagery project³ shared hourly map updates to keep first responders abreast of the situation; in the aftermath, it was used to predict when school areas could be expected to reopen. This strategic use of data allowed rescuers and government agencies to make far more intelligent, real-time decisions than would have been possible previously.

State agencies across the country have also been leveraging their data for a wide variety of purposes. For example, New Jersey analyzed its data on a prescription drug program to facilitate targeted outreach to senior citizens eligible for the Supplemental Nutrition Assistance Program.

Virginia created a system to share data between its human services agencies and nonprofits, making it possible to functionally end homelessness among veterans. Colorado analyzed its Medicaid prescription drug claims data and discovered evidence of fraud, waste, and abuse in opioid prescriptions. Across the nation, states are increasingly overcoming challenges that range from lack of funding to unskilled staff and are using data strategically to achieve better outcomes.

Challenges States Face

In seeking to make better use of data to drive policy, states have come up against a handful of significant, but not insurmountable, challenges, which must be solved for the use of data to be maximized (see Figure 1). The top four challenges are:

• Staffing. To use data effectively, state governments need staff members with an understanding of public policy, the technical skills to manage and analyze data, and the communication skills to present findings to a wide range of audiences. These workers are in short cannot usually offer salaries competitive with the private sector.
• Data Accessibility. Outdated tech-nology can make it difficult for states to extract their data in a usable format. And when computer systems are controlled by contractors, states may not even own their data or they may lack the technical expertise to access it.
• Data Quality. Data are only useful when they are accurate and reliable. When data points are missing, inaccurate, or poorly defined, the information is less useful.
• Data Sharing. Data are most valuable when they can be shared across programs or agencies and combined with other data sets to get new insights. However, many state agencies face uncertainty about how to comply with privacy laws or share data while keeping it secure—thus limiting the insights that can be gleaned from the data.

Five Key Actions for Strategically Using Data To Make Decisions

Regardless of a state’s current data capabilities, Pew’s researchers found that state agencies can take five concrete steps to maximize the value of administrative data at their disposal and move toward the goal of using data to make better decisions.

1. Plan ahead by setting up guiding goals and structures. By planning ahead, states can create a more supply, especially since governments coordinated approach to managing and using their data, which can help better allocate resources as well as prevent data breaches or privacy violations. Three strategies that help a state effectively plan ahead include writing a formal data strategy, developing data governance structures, and taking stock of systems and performing an inventory of current data sets. The Data Governance Program Office in Oklahoma helps agencies put data governance structures in place, including setting up mechanisms to improve the quality and management of information. The office set up the Deliver Interoperable Solutions Components Utilizing Shared Services program, a cooperative governance board for health and human services agencies to standardize and longitudinally connect data across those agencies and identify shared technology and assets.
2. Build the jurisdiction’s capacity to effectively use data. Adequate staff who are data-literate and have skills in data analysis, appropriate funding and time to support data-driven projects, and external partnerships with universities and other stakeholders can strengthen a state’s success in effectively harnessing the power of its data. To address the need for skilled data analysts, the Manage by Data Fellows Program of the New Jersey Department of Children and Family Services trains staff on using data as part of their everyday work through a nine-month project-based curriculum. With more than 200 alumni, graduates have passed along many benefits to the state, including improving the timeliness of the agency’s case management. Similarly, to increase the state’s data capacity, Delaware’s Medicaid program entered into a formal agreement with the University of Delaware to establish a dedicated research arm. The partnership allows the university to research Medicaid-related programs, present findings to state policymakers and administrators, and help decisionmakers have a better understanding of the needs of Medicaid clients and the best way to administer programs. Before this partnership, Delaware did not have the capacity to conduct extensive data analytics to inform decision-making.
3. Ensure that quality data can be accessed and used by stakeholders. Before staff can effectively use data, it must be able to access it. This requires agencies to improve data quality and accessibility, often by establishing data-sharing agreements and protocols among offices within a department, across other offices within an agency, or externally with other government agencies or stakeholders. Illinois created a uniform data-sharing agreement for all state human services agencies—which has encouraged crossagency collaboration and access to data that could help improve client services while protecting client privacy. This has helped state agencies develop a deeper understanding of the clients they serve, for example, by allowing the Division of Alcoholism and Substance Abuse to understand how its clients interact with the Medicaid program.
4. Analyze data to create meaningful information. If done well, extracting information from data can transform the way state government makes decisions. The Research and Data Analysis Division of the Washington Department of Social and Health Services analyzed several years’ worth of Medicaid and Medicare claims data, as well as information from the state’s long-term care program, to build a tool that can predict which clients are likely to have high health care costs in the future. The tool allows the state to identify those patients and enroll them in a program that provides more coordinated health services, which not only helps patients stay healthier but also saves the state an estimated $248 per patient per month, according to one study.
5. Sustain support for continued data efforts. Once a state invests resources into building data-driven systems, efforts to gain the commitment of leaders and enact policies and frameworks to make sure the initiative continues are equally important. In 2015, the Massachusetts Legislature passed Chapter 55, which encouraged five agencies to share 10 administrative data sets related to opioid use and to work with stakeholders to perform comprehensive analyses to determine the root causes and risk factors of opioid-related overdose deaths.

Subsequent research 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 were people who had not been in prison.

These useful insights came about only because the legislature demonstrated its commitment to creating a culture of using data as a strategic asset.

Moving Forward

This study found that no state is consistently applying the five recommended actions on strategic use of data to a broad range of government agencies or achieving widespread improvements in policy development, service delivery, resource management, and current program evaluation.

However, states recognize the need to invest resources toward planning, building capacity, sharing data, analyzing data to create meaningful information, and sustaining data efforts. As agencies begin to more consistently share and integrate data and deploy advanced analytical methods, harnessing the insights held within administrative data will only become easier. As a result, they will benefit greatly from realizing the potential of data analytics in making government more efficient, effective, and responsive to major public problems.

Reference Notes

1. See http://www.pewtrusts.org/en/research-and-analysis/articles/2018/05/18/prison-health-carequality-monitoring-systems-vary-by-state
2. See http://www.pewtrusts.org/en/research-and-analysis/reports/2018/02/how-states-use-data-to-inform-decisions
3. See https://tnris.org/texas-imagery-service

This article was originally published in The American Public Human Services Association's Policy & Practice Magazine.