July | August 2017

Governing Data

by Katherine Barrett and Richard Greene
The emergence of digital government and the power and potential of data have expanded dramatically in the last 10 years. Social media has become ubiquitous. Open data, which makes data sets available to all citizens, often through state websites, has become ever more commonplace since 2009. And the daily media attention to “Big Data” started around 2012.
But although state governments are floating in a sea of data, the management and governance of this new kind of asset has tended to be weak, and sometimes close to nonexistent. The National Association of State Chief Information Officers’, or NASCIO’s, survey of CIOs in 2015 noted that technology directors are “wrestling with a host of challenges around data governance, legacy data, data access and sharing, and major new flows of data from new sources.”
In 2015, data management appeared for the first time on the CIOs’ top 10 priority list. Still, fewer than 5 percent said they had “formal data management policy and practices” at an enterprise level. Less than half said the current role of the state CIO was to develop an enterprise data strategy. Only 38 percent believed their current role involved convening stakeholders for data governance decisions.
Legislatures haven’t jumped into the breach with any frequency. “There’s a significant gap in the understanding of the power of analytics and what needs to happen on the legislative side,” said Doug Robinson, executive director of NASCIO. “They’re not having these conversations in legislative bodies.”
Agency apprehension doesn’t help. Part of the reluctance to share data is the sense among agencies that it will be used against them. Said Connecticut’s chief data officer, Tyler Kleykamp, who has held the newly created position for two years, “We’ve tried to communicate to the agencies that their data has value beyond the purpose for which they collect it. And that purpose isn’t to look over their shoulder and tell them they’re doing a bad job. We’re not looking for problems. We’re looking for solutions.”
Said Beth Ashcroft, executive director of Maine’s Office of Program Evaluation and Government Accountability, “There really need to be . . . dedicated resources and focus in each agency to work with the Office of Information Technology in matters of IT significance, including data governance.” Without centralized data governance, resources can be wasted in a constant reinvention of the wheel as individual research projects have to go through lengthy meetings and legal maneuverings to gain permission to work together, and share data, for the common good.
Fortunately, while most states are still very early on in setting up structures for data governance, a growing number of states have made progress. In 2010, Colorado became the first to appoint a chief data officer and followed that move with the creation of the Government Data Advisory Board. Since then, states such as New Jersey, New York, Connecticut and Texas have created data chief or data coordinator positions. Some states, like Virginia, have devoted attention to data governance, without actually creating a data chief role. Others, including Indiana, have inaugurated statewide programs for data sharing through executive orders.
Texas appears to be advancing as quickly as any, which is a particularly difficult challenge, given its deeply decentralized structure. It established both a statewide data coordinator position and an interagency data transparency commission in the 2015 legislative session. Chaired by the governor’s office, the commission is made up of representatives from the House, Senate, legislative budget board, comptroller’s office and the Department of Information Resources, with support from executive leadership from Texas agencies. The creation of new communities of interest has already sparked some unexpected collaborations such as one between the Texas Veterans Commission and food and nutrition programs associated with day care and run by the Department of Agriculture. The goal is to make sure that veterans who have children, and who were eligible for the reduced meals program, were using it, by matching data between the two programs.
Ed Kelly, who had been the chief administrative officer for the Texas Department of Agriculture, took on the statewide data coordinator job in September 2015. Data governance, open data and data analytics/business intelligence are the “three towers” under “data utility” in the state’s strategic plan. “Data is an extremely important strategic asset,” said Kelly. “There’s opportunity, even in a decentralized government environment, to open up and share so agencies can be more effective and efficient in the way they serve their constituents.” So far, Kelly has met with 30 agencies. He says the predominant reaction to his appointment is “overwhelming;” the phrase he hears is: “It’s about time we had a person in this role.”
Working on data governance involves a potent effort to change the culture of a state’s approach to technology. It involves multiple individuals spread out across programs and incorporates both technological, political and business decisions. Who decides whether data is open or not? How are decisions made about what is collected and what it is used for? Where do you need approval and from whom?
These aren’t easy questions. “The data was safe when it was locked in a box,” said David McCurdy, the director of technology in Colorado. But the more you open it up, the more potential risk. “But also we know we can use open or shared data to really change lives.”

About Barrett and Greene

CSG Senior Fellows Katherine Barrett and Richard Greene are experts on state government who work with Governing magazine, the Pew Charitable Trusts, the Volcker Alliance, the National Academy of Public Administration and others. As CSG senior fellows, Barrett and Greene serve as advisers on state government policy and programming and assist in identifying emerging trends affecting states.