The Obama administration likes to say it is the most data-driven presidency ever. The White House launched Data.gov and hired three chief technology officers whose focus has been, in part, on unlocking the data from the bowels of the government. And now, agencies are on the hunt for chief data officers, data scientists and experts to fill similar roles.
DJ Patil, the White House’s chief data scientist and the first person to hold that title and role, references an analogy used by former Federal CTO Todd Park to describe the evolving role of federal chief data officers.
“We have this unique opportunity to truly be that team. I think Todd’s analogy, he calls it the Avengers, you get called out and you got to do this stuff,” Patil said in an exclusive interview with Federal News Radio. “The part we often have to remember, if we are taking this analogy of Avengers, is that there is S.H.I.E.L.D. And that’s really the people who are inside the agencies and they are the ones who own this every single day. The real heroes are the people who are there.”
While it’s easy for agencies to jump on the big data bandwagon—or on S.H.I.E.L.D’s Helicarrier to follow Patil’s analogy—the question that must be asked is, where does the chief information officer fit into this equation?
Insight by Kodak Alaris: Practitioners provide insight into how states and the IT industry are dealing with Real ID in this exclusive executive briefing.
As agencies collapse their technology infrastructures into the cloud, looking to both internal and external shared services, the CIO’s role is to provide the mission areas with the technology they need, and that’s quickly becoming data analytics, data visualization and data mining tools.
So, if a CDO is supposed to manage data and a CIO is supposed to manage information, then what’s the difference between information and data?
In Federal News Radio’s special report, Deconstructing the CDO, we look at how agencies are creating the role of the chief data officer and how these data experts fit into the existing federal IT oversight infrastructure.
The buzz around big data is growing in volume as eight agencies, including the departments of Energy, Commerce and Transportation have named formal chief data officers.
But agencies must find the right balance of roles between these hip CDOs and the tried-and-true chief information officers as federal IT evolves through cloud and the commoditization of technology.
NASA is in the midst of hiring a new data expert, but it isn’t calling that person a CDO quite yet.
“There’s a lot of energy right now about putting that position in the CIO because we are talking about a lot of IT. It’s hard to separate it out when you are trying to modify or upgrade your systems for dealing effectively with data management,” said Larry Sweet, NASA’s CIO. “I think there is a natural belief that the leadership of big data comes from the CIO because it is chief information, and chief data isn’t a far stretch from that. As we focus on where we want to go with a data strategy, it doesn’t necessarily need to reside in a CIO organization and maybe it shouldn’t. It’s something we are trying to explore a little bit further and see where the benefits are if we keep them separate or if we keep them tightly coupled.”
Sweet, like many CIOs, is trying to figure out if the CDO should start in his office, but over time move toward a more independent setup.
Frank Baitman, CIO at the Department of Health and Human Services, is caught in a similar conundrum as Sweet and many other agencies are when it comes to the CDO.
Baitman said there are plenty of good reasons to hire a CDO, especially as the understanding deepens of how data can be used.
“When you begin to look at the HHS enterprise and say, ‘Well, it’s a bundle of 10 operating divisions with different missions.’ Well, that’s all true, but they have in common some support to public health,” he said. “There is information that is created in one part of the enterprise that can and should inform other parts of the enterprise so that we choose to treat and pay for and advance therapeutics that have the greatest impact.”
But Baitman is unsure whether HHS headquarters needs a CDO or if the data expert would be better suited for some or all of those operating divisions.
One of those operating divisions, the Centers for Medicare and Medicaid Services, decided hiring a separate chief data officer was the path it wanted to go down.
Niall Brennan is the first chief data officer at CMS and also leads the Office of Enterprise Data and Analytics. He started in December and works outside of the CMS CIO’s office.
“My overall position as CDO is to maximize CMS data for internal and external users. Under that rubric, our absolute number-one overarching priority is the protection of beneficiary privacy,” Brennan said. “We feel we’ve been able to push the envelope quite a bit in the release of data sets that give us good insight into how providers practice care in America.”
He said up until two or three years ago, CMS was prohibited from releasing data at the individual physician level. But a successful court case let CMS release much more granular data about how doctors practice care and prescribe drugs in the Medicare program.
This is where the CDO role is beginning to take hold—in digging out information for the benefit of citizens.
And, CDOs are filling a void that CIOs and others can’t at this moment.
Jane Snowdon, IBM’s Chief Innovation Officer for Federal, said the CDO is a growing trend.
“The government, as well as private sector organizations, are realizing the strategic value of putting an executive in charge of its data and analytics,” she said. “Leading organizations are relying on chief data officers for innovation and transformation across their organizations.”
Snowdon said the three big fundamental shifts happening in government today—cloud, mobility and citizen engagement—are all dependent on data.
Snowdon and other experts said the CIO has to be focused on developing the infrastructure to handle this technology and services shift at least for the near future.
Snowdon said the role of the CDO is much different and works on the data strategy side of the equation.
“The chief data officer views data as an asset and tries to find the best ways to leverage existing internal data, exploiting new data sources from business partners or from social media. I think the CDO is really a change agent. They are changing the culture. They are trying to find ways to enrich the data, perhaps monetize the data and, in some cases, figure out how to reuse data and create new services. The chief data officer also has to have knowledge of machine learning, statistics and other advanced analytics techniques.”
Patil, the White House’s chief data scientist, said the role of the CDO has grown over the past few years.
“The data scientist and the CDO is Spock on the bridge. The important thing to remember is Spock is on the bridge, not because Spock is supposed to have a chair on the bridge, but because Spock has earned the right to be on the bridge and is a critical component of every conversation,” he said. “That is that operational execution focus and that is what I think we have seen as the CIO continues to evolve, in the same way the CTO role continues to evolve. It will have overlap. Now is the opportunity for that CDO role to bloom.”
Nick Sinai, a former White House Deputy CTO and now a Walter Shorenstein Media and Democracy Fellow at the Harvard Kennedy School of Government, said CIOs are focused on cybersecurity, email consolidation, budget and contracting, and many other strategic and operational issues, and don’t necessarily have time to focus on data innovation challenges.
Sinai said the CIO and CDO roles will likely continue to evolve along with the discussion of how the roles fit together.
“There are times when a CDO reports to a deputy secretary or the front office in a way the CIO doesn’t. There are both advantages and risks associated with that,” he said. “I think this will be a great period of experimentation as probably a dozen agencies develop and have some experience with a CDO role.”
Sinai said in the private sector CDOs are less likely to report to the CIO, and most likely coupled with the vice president of engineering or the CTO.
It’s clear from these experts, many agencies have room and a need for CDOs today. But as agencies move to the cloud and stop worrying about the day-to-day IT challenges, how will the management of data and information change?
HHS CIO Baitman said the CIO and CDO roles are complementary.
“I’m never going to move out of that role of supporting the infrastructure for the HHS enterprise. That is, I think, an appropriate and necessary role to be performed by this thing by the Office of the Secretary. That is where we bring this entire large enterprise together,” he said. “That being said, there is a lot of information and data that travels on those networks. And there is the opportunity to begin to drive standards for how that data is captured and to begin to drive common APIs for sharing that information, so that other people who weren’t the creators of that information can make use of it.”
The White House’s Patil said the CDO must be part of the conversation with the CIO, CTO and mission owners.
“Everybody needs to have their data strategy. That is a fundamental thing that every organization must have,” Patil said. “Is it the CDO or someone else? What people have to have, which is more important than anything, is not a title but a focus on being data driven. What does data driven mean? It means that you instrument, you measure, you acquire the data, you process the data, you turn that data into action, insights, build data products, return it back, open it up and allow investment to continue to build on top of that investment. That is more important to me than any title.”
He said sharing data and information is the quickest way to improved services and decision making. And it’s something the CIO, the CDO and many others have to institutionalize in the government over the short term.
Read part 2 of Jason Miller’s special report, an interview with DJ Patil, the White House’s first-ever chief data scientist