How Data-Driven Insight Is Transforming Government

The IBM Center recently released Seven Drivers Transforming Government, a series of essays exploring key drivers of change in government. It is based on our research and insights shared by current and former government officials. What follows is an excerpt from that report.

The language around performance-related data in government decision making has evolved over the last quarter century. Today there are frequent references to evidence-based decisions, strategic analytics, and data-driven progress reviews. Policy makers have promoted the use of open data—both within and outside government. At the same time, evolving technologies have reduced the cost of collecting and reporting such data. Yet the original challenge remains: how can government make sense of vast and growing amounts of data to develop new understandings that inform decisions?

Transforming Data into Insight  

The use of data and analytics goes beyond just collecting and reporting evidence of program outcomes. New technologies, such as cognitive computing, are helping decision makers identify meaningful and actionable information that can transform data into insight leading to effective action.

Insight offers the power to gain understanding from data, people, or a situation not immediately evident. It enables leaders to uncover solutions and act based on keen and clear observation. Observation, however, does not lead immediately to insight. Government leaders must make interpretations shaped by context, knowledge, and experience—implicit knowledge, often accrued after years of experience.

Increasingly, there is an opportunity to base decisions on data-derived insight derived from advanced analytics. For example, local police now use multiple historical data sets—past burglaries,  weather, time of day—to predict future crime patterns, leading to strategic redeployments of officers and sharp reductions in crime.

Increasing the Availability of Data

The Government Performance and Results Act of 1993 and subsequent legislation have contributed over the past two decades to the development of a more robust supply of useful data and performance information that serve as the foundation for evidence-based insights and decisions. Examples of this trend include:

• More open data has contributed to a growing supply of useful information. Presidential commitments to open data, the adoption of the Digital Accountability and Transparency Act, and a commitment to making routine administrative data more widely available have supported this trend. For example, the federal one-stop website, data .gov, makes nearly 200,000 data sets available to other agencies, the public, and entrepreneurs.

• Technology innovations have made it possible to collect, organize, share, and interpret data on a much larger scale, with greater immediacy. For example, after the 2008 financial crisis, the public could track spending of the $831 billion Recovery Act by ZIP Code. More government data will be available in the future with the growth of micro-data via the internet of things (IOT), such as real-time monitoring of pollutants in waterways.

Even with this increase in data availability, governments must continue to address data quality and reliability—especially with data from multiple sources, where users lack a deep view of how data were collected or interpreted. In addition, statutory barriers impede the collection and sharing of certain kinds of data, such as limits imposed by the Paperwork Reduction Act.

Pursuing Evidence-Based Policy Initiatives

Governments are now working to improve analytical capacities to develop evidence-based insights that can inform decision making. Building on the foundation of data and evidence developed during the early 2000s and the analytic capacity developed in federal agencies, the federal government undertook a series of initiatives beginning in 2009 to use these data and capabilities to improve government policy and  funding decisions. A host of external foundations and nonprofits championed these efforts, including Results for America, the Pew Charitable Trusts, and the Arnold Foundation.

New policy tools that rely on evidence to inform resource allocation include social impact bonds and tiered evidence grants. Additionally, new laws require some agencies to set aside funding to conduct program evaluations or to earmark a portion of grants for programs that could demonstrate effectiveness. Many of these initiatives have benefited from bipartisan support. For example, the statute creating an Evidence-Based Policymaking Commission—co-sponsored by Speaker Paul Ryan and Senator Patty Murray—identified ways to lower barriers to collecting and sharing data and to better use evidence in government decision making.

Building Analytic Capacity

Federal agencies’ capacity to act on the data they collect has grown steadily in the past decade. Multiple factors have influenced this growth:

• The availability of timely data, based on technology that makes it easier to collect, analyze, and display, and the growth of cognitive tools to help interpret data;

• An evolving organizational infrastructure that supports the use of evidence—several agencies have designated chief data officers and a number of agencies have created positions to drive the use of program evaluation and real-time analytics;

• Congress has provided funding to support analyses, leading to increased experience in the design and conduct of evaluations; and

• Agencies have begun to conduct regular reviews of the progress of their key priorities, which in turn rely on data and evaluations.

Finally, the availability of new technical capabilities—such as sentiment analysis of real-time social media data, tools that speed social network analyses, and the increased application of behavioral science insights—have all contributed to more sophisticated capacities that bolster government decision makers’ confidence in relying on comprehensive and robust evidence when making tradeoffs.

Continuing the Shift to Insight

Even with this progress, significant opportunities remain to move from creating a supply of data to creating a demand for insight that informs decision making. These opportunities include:

• Agency leaders need to ensure the usefulness of program-level data and analytics and make them readily available to frontline employees and managers who are then incentivized to use the data and confirm its quality. For example, the U.S. Office of Personnel Management has granted federal managers access to a website with annual survey data on employee engagement with their work so they can develop strategies to improve.

• According to University of Maryland professor Donald Kettl, analytics tools are developing faster than the strategy (and theory) about how to use them. This is, in part, a cultural divide between “data people” who understand how to collect and analyze data and leaders who are only beginning to figure out why they should rely on data versus their instincts.

• Multidisciplinary approaches, such as the use of strategic foresight, risk management, and creating institutional resilience, can increase the use of data analytic tools. For example, developing strategic foresight frequently relies on analyses of macro-trends in economic, environmental, demographic, and social data.

There will be many opportunities to address issues of turning data into actionable insight over the next few years. Agencies can embed these approaches in their reform plans and their budget requests. The Commission on Evidence-Based Policymaking’s final report, “The Promise of Evidence-based Policymaking,” provides additional direction and impetus. Encouragement from nonprofits, such as Results for America’s nine-point legislative agenda advocating the integration of evidence into decision making also helps create momentum in Congress.