Preventing the Next Financial Disaster with Data

Preventing the Next Financial Disaster with Data

We spotlight the change agents and what it takes to drive transformation in government.
Micheline Casey, former chief data officer at the Federal Reserve Board of Governors, head of data digital services, Ford Smart Mobility

Driving change in government is notoriously hard — but not impossible. GovernmentCIO Media sat down with those who made things happen to hear their stories on how they successfully drove change and transformation in a bureaucracy. One important takeaway: You don’t have to be in the C suite to make change happen; in fact, sometimes a senior position will be more of a straight jacket, stifling movement in right direction.

Micheline Casey joined the Federal Reserve Board of Governors in May 2013, after serving as the first-ever state-level chief data officer in the Colorado governor’s office. When she first came in as CDO, she took as her mission an imaginary headline she hoped she might one day read in The Washington Post: “Federal Reserve uses all of its data to predict, avert next financial crisis.”

Despite her government experience, Casey says navigating the federal sector was quite the challenge.

“I found the state government experience to be a lot different than the federal government experience,” she says. “The state legislature and governor really wanted to enact transformation in a meaningful way. People at the Reserve were coming out of a financial crisis and dealing with the aftermath. The sense of ‘we need to do things differently’ just didn’t bear as much there as in state government.”

Micheline Casey, Ford
Micheline Casey, Ford Smart Mobility

The federal government was much more bureaucratic than its state counterpart, Casey says, and the organizational structure at the Federal Reserve — including 12 independent regional Reserve banks — added complexity and red tape, too.

“Having more of that agile, lean startup mindset . . . was more of a struggle than I had anticipated.”

Casey’s goal was to build out a new business unit to improve data analytics within the organization. Essentially, create a startup within a 100-year-old institution, doing everything from strategy, vision and branding to meeting with various stakeholders to understand their needs and wants.

She quickly realized her biggest challenge was to get her own team to embrace agile, leveraging pilots and modular development rather than the waterfall method used previously.

“Having more of that agile, lean startup mindset — doing things in pilots and incremental steps and come out with a minimum viable product and move quickly, and continue to build on top of things — and just thinking differently was more of a struggle than I had anticipated,” she says.

Although surrounded by talented people, Casey says the organization didn't fully understand agile and wasn’t used to an environment that encouraged or embraced that methodology.

In one instance, Casey and her team were trying to implement one particular high-priority pilot. The team was supposed to sweep in, spend two to three weeks getting a snapshot of how each business unit was using data and then bring back the newly acquired metadata to enhance the product being built.

“My team just couldn’t get out of this mentality of, ‘we have to do this perfectly from start to finish — we can’t just pick one piece and iterate on that,’” Casey says. “They were really about ‘here’s the big picture, and it’s going to take us a couple of years to get there — it can’t be less than perfect because no one will accept anything less.’”

Casey suggested the team members come up with a minimum viable product every few weeks and continue working on improvements as they went on.  

"It was a real struggle for my team, as well as our internal customers, to be thinking that way,” she says. So the first sprint, intended to take two to three weeks, dragged on for two months.

The second sprint, albeit faster, still took a month to complete.

But then: a breakthrough.

“The light bulbs were finally going off after about four to five months of doing this,” Casey says. “‘Oh, now we get it — we’re learning and our stakeholders are learning and we’re able to develop and redevelop the product more specifically and enhance it more quickly.’”

What Casey says ultimately helped the team embrace agile was “tons of communication, education and outreach around agile.” She bought her team members copies of “Lean Startup,” Eric Ries’ book about how to shorten product development time by taking an iterative approach, and spent “literally hours and hours and hours” discussing how other organizations were using agile.

“But me talking about [agile] wasn’t nearly as helpful or impactful as them experiencing it themselves,” Casey says. “I think they needed to actually go through the exercise of doing this repeatedly before the light bulb actually came on.”

Despite these obstacles, Casey says anyone presented with the opportunity to work in the federal government should seriously consider a stint.

“I really enjoyed building a team in a startup environment in the federal government,” she says. “If you can build something new in the federal government, certainly coming to private sector and doing the same is that much easier.”