‘Until the great recession, I don’t think anyone would have called the data that was coming in “big data”,’ said Micheline Casey.
Data consultant Micheline Casey most recently served as the first-ever chief data officer (CDO) for the Federal Reserve Board – the central bank of the United States – and the first CDO for a state government in the US, Colorado.
An internationally recognised expert in using data for critical, data-driven decisions, Casey has been honoured several times, including being named in DC’s Top 50 Women in Technology list, and in the Top 25 Information Managers from Information Management magazine. She was recently featured on Siliconrepublic.com as one of 25 leading women breaking the CIO mould.
‘Those who are in charge of organisations, including the board of directors, need to be quite thoughtful and systematic in their approaches to enabling diversity within their organisations’
– MICHELINE CASEY
What path led you to a career in technology?
My first job out of school was with a software start-up in Atlanta, Georgia, and I fell in love with the tech industry. We were a one-year-old company when I joined it and we ended up doing an IPO, which was exciting. I had an opportunity to wear lots of hats and have a rapid career progression from market research assistant to engineering, to consulting with clients, and I just totally fell in love with this space.
I went to work for a data broker in 2002 called ChoicePoint, a spin-out of Equifax. It was my first experience of big data in 2002 and no one had coined that term, but I was very unaware that there were companies out there aggregating these massive amounts of data, to be able to package up and deliver to companies for a variety of purposes, such as credit management, employment background checks, or helping the government with tracking terrorists and things like that.
Today, there are so many industries built on those things, but back in 2002, it was an eye-opening experience and I fell in love with the power of data and the ability for data to help companies run their organisations much more effectively.
Since that time, I have been in the data space and I have been fortunate to be able to parlay that first role at ChoicePoint into some fairly impressive (on paper) titles.
What was it like breaking new ground as the first chief data officer in the US?
The governor’s office in Colorado created the first big government chief data officer role in the United States and, at the Federal Reserve, I served as its first chief data officer.
In between the two, I ran my own consultancy, where I have worked with companies across a variety of industries in the public and private sectors on data strategy, data management, governance and data monetisation.
Looking back, it feels weird to say, personally, that you’re a groundbreaker, but I guess I was.
I worked for the Ritter administration from 2007 to 2011. At the time in 2007, there just weren’t many chief data officers, regardless of industry sector, so it was a fairly new title.
I don’t think there was a ton of understanding of what that role was and what it meant. Organisations of all types were becoming more aware that data was really powerful, and they should and could be doing more with their data. I was really fortunate to work for a governor and a CIO that had a very strong and distinct vision that we needed to do digital transformation; we needed to be able to leverage data and technology that serves our citizens and the businesses in the state more effectively. We were ready to make the changes that were necessary from an infrastructure perspective, but also from a policy perspective, to enable us to really leverage data and analytics in very different ways from what other states were doing.
As I was going around talking to the department heads of the various agencies and executive branches, and putting together our enterprise data strategy, there were several anchor projects that rose to the top.
They rose because they were funded. People with stakes in the game could actually make it forward and – if we made progress in these few areas – start making those like Legos and have a state umbrella cover.
We did a lot of work in education, longitudinally early-childhood education through to higher education.
There was also set work we did around juvenile justice and looking holistically at kids coming into the correction system, and pulling in education data, helping human services data, behavioural and health data – a lot of highly sensitive areas.
While there was federal and state grant funding coming, the areas we focused on were very cross-functional in nature, and designed to get people from various agencies more aware of the work we were doing, and it was good to have such multifunctional participation.
Around the same time, President Obama passed the Affordable Care Act and the health information exchanges were starting to come online. There was a lot of planning in the state around health information exchange data that tied into other benefits data, and that became another area of importance for us.
The fourth pillar was identity management and being able to uniquely identify a citizen and, in today’s vernacular, understand the business from a 360-degree point of view in terms of the benefits, services and revenue that the citizen was contributing to the state.
How was big data embraced by the Federal Reserve, the central bank of America?
The two primary missions of the Federal Reserve are to maintain optimal employment in the US, and then maximise interest rates and manage what that does to employment and spending.
Most people don’t understand what their central banks do at all and I certainly didn’t. As a citizen of the United States, I didn’t have much insight into the Federal Reserve and how it operated until I got inside that institution.
For me, what I found quite interesting coming out of the great recession was how we all felt the impact of the systemic failure in the financial markets, and the federal reserve was a key force in trying to provide stability – not just in the US but worldwide.
So those datasets were not only important to the Federal Reserve but had global impact, which, from a mission perspective, was just very interesting and fascinating.
Was it a huge leap in responsibility jumping from state level to federal level?
There were major differences. In terms of the amount of data, no, it wasn’t a huge leap.
Some of the primary differences are that in the United States (I’m sure its different in EU countries), federal government has to be fairly far removed from the day-to-day; the likes of people and businesses.
Working in the governor’s office in the state of Colorado, we were directly impacting benefits that people got on a day-to-day basis. The flow of money from the federal and state government to people and local government was dispersed, so you were much closer to your customer at the state and local level than you were at Federal level. As such, there was much more of an awareness at state level – you are dealing with real human beings and you are impacting their lives.
If you don’t approve benefits applications pretty quickly, there are fairly dire consequences; or if you release information inappropriately, it could be a life-or-death situation for a child.
The Federal Reserve is generally a policy shop, dealing with money and making policies around interest rates, which, at the end of the day, also impact people.
The sense of urgency and culture were quite different and so there were a lot of adjustments that I needed to make to staff, and how things were handled. The other big difference was, at the Fed, I had to stand up a whole new business unit.
A new infrastructure was needed for that – people, processes and technology.
The bigger lesson was the culture and change management in the institutions that needed to happen, to accommodate a new department, increase headcount by 50 people and steer changes in the day-to-day management of data.
The amount of the data was actually probably greater in the state of Colorado. At the end of the day, the Federal Research has existed for more than 100 years and they do get lots of data but, in terms of the scale of the data, they are not generally data generators. They pull in data from the federal government, statistical agencies such as the Bureau of Labor, statistics like commodity prices, energy prices and data from a handful of surveys they do.
Again, until the great recession, I don’t think anyone would have called the data that was coming in ‘big data’. But, post-recession and after the passage of the Dodd-Frank Act, certainly the scale and the volume of data changed, although it was actually less than I was dealing with in Colorado.
The Fed manages the budget of 500m and has 2,500 employees. In Colorado, it was a $19bn organisation with close to 30,000 employees, serving 5m citizens and another half a million businesses. The scale in Colorado was definitely bigger.
How do you think governments look at data today to make better decisions?
I think there is still a long way to go from an analytics perspective – things tend to be very much BI [business intelligence] operational reporting versus predictive.
Part of that is just that the infrastructure within the environment can be very old. In Colorado, we were dealing with a 40-year-old mainframe for our Medicaid data. In the Federal Reserve, we had to remodel the data architecture that was pulling in the economics data, and that data architecture (when I got there) was 25 years old, and so we needed to modernise that.
Certainly, there are structural problems in terms of government agencies being able to do predictive analytics that you would hope they would be able to do, for policymaking and programmatic purposes. We still have a long way to go.
Speaking of a long way to go, what are your feelings on diversity in the tech sector from the perspective of being a woman leader?
I was just speaking at a conference two weeks ago and, as usual, on the three panels I sat on, I was the only woman. There is still some way to go.
Over the course of my career, going back to the early days of serving in a software start-up, I was aware that there weren’t that many women and I don’t think that it particularly bothered me until later. Earlier in my career, I focused on good work and adding value to organisations that I was in, and certainly when I accepted the chief data officer role, I didn’t think about being the first woman.
I think we should talk about diversity being much larger than just women.
I think about it as a leader in a role and my responsibility, to my employees and to the organisation, to mentor and provide leadership internally and externally.
I tended to try to make sure I had time in my calendar for mentorship, certainly with women on my team. Certainly, at the Fed, I was very conscious that I wanted to highlight a diverse management team and individual contributor team.
My management team ended up being half women, and it also ended up being half minority. I had Egyptian American, African-American and Nigerian-American people as part of my team, and I had people in their 30s and in their 50s who were later in their career, so it was a nice blend of different perspectives from a number of parameters.
And then my team ended up being minority white, and the ratios were pretty evenly split for males and females, though I didn’t have as many women as I hoped for on the purely technical roles. But we did spend a lot of time in our leadership team meetings talking about that and making sure that we were thinking about it.
Part of the role of leader is to raise awareness and make sure people are thinking about it.
It is very easy to forget that and pull in talent very quickly. From a larger perspective, those who are in charge of organisations, including the board of directors, need to be quite thoughtful and systematic in their approaches to enabling diversity within their organisations.
Want stories like this and more direct to your inbox? Sign up for Tech Trends, Silicon Republic’s weekly digest of need-to-know tech news.