Why Data Must Drive Decisions in Government

The Dawn of the Data-Driven Government

The government landscape in 2013 includes a host of challenges and opportunities. The economy is uncertain, politics are highly polarized, and many more people are relying on government services compared to just a decade ago. We’re also living in a time of unprecedented digital openness, convenience, and transparency in everyday life.

It’s this landscape, where eighty-five percent of American adults have internet access and fifty-six percent own a smartphone, which has turned “google” into a verb and made everyone a restaurant critic (Fox, 2013). Citizens expect to find the answers to almost any question online and are demanding more answers from government. Yet, in large part, government-related accessibility has lagged, until now.

By adopting cutting-edge technologies, such as cloud data storage and application programming interfaces (APIs), to unlock and share their treasure troves of data, some government agencies are not only catching up with the private sector in terms of innovation, but also accelerating beyond it. Data has long been recognized as a government asset, but now it can more easily be shared and utilized both inside government and among citizens, entrepreneurs, and researchers, to find solutions to persistent civic problems.

A new era of data-driven government and civic innovation is taking shape, and the results are amazing so far. They provide only a hint of what is possible in the future. It’s not a matter of whether data-driven government can create the best solutions to society’s problems; it’s a matter of how soon different governments will embrace the idea and reap the benefits.

As the former leader of one of the earliest and most successful data-driven programs in government, Maryland StateStat, I want to share my story of how I arrived at that role, what we accomplished during my tenure, and why I have now chosen to move to the private sector to create a platform for data-driven governance.

How I Joined Data-Driven Government

For most people, government is an ancillary part of everyday life; that is certainly how I saw it before I decided to attend law school. I had a general goal to help people with my law degree, but getting into government had never been of interest.

A day that changed so many people’s lives, September 11, 2001, ended up moving me toward a career as a public servant.

At the time, I was fresh out of law school, newly married, and working at a New York City hedge fund. A painful slip in a pothole on September 10, 2001 sent me to the doctor’s office on the morning of September 11, rather than heading to work in the World Trade Center Complex.

The experience prompted a lot of reflection and a shift in priorities. My husband and I decided to move back home to Baltimore. And, I remembered why I went to law school in the first place: to help people.

I changed my career trajectory to pursue a job as a public defender. While studying for the Bar, I worked as a juvenile probation officer to get court experience. The more time I spent at the Maryland Department of Juvenile Services, the more my appreciation grew for the process, as opposed to the practice, of law.

At this time, Mayor Martin O’Malley of Baltimore was undertaking some inspiring work. While leading Baltimore’s government, he began taking cues from New York City’s success with the CompStat program.

For example, the CompStat team in NYC had cross-referenced crime maps with police department resource maps and uncovered a glaring disparity: crime-fighting resources were evenly distributed, but crime wasn’t. By being smarter about dispatching departmental resources, NYC’s leadership was able to clean up the city significantly. When crime disappears, something else usually moves in. In New York City’s case, it was industry, business, tourism, and more vibrant neighborhoods. The data-driven approach proved powerful.

Mayor O’Malley wanted to use this approach to clean up Baltimore, but in a much broader capacity. He was focused on reducing crime, but he also wanted to impact and improve the city as a whole. He wanted the suburbanites who had abandoned their city to return. He wanted a vibrant downtown full of safe and happy visitors. He wanted all of Baltimore’s children to have the best education, and he wanted a cleaner, more efficient city.

As a result, he led a team to create his version of CompStat—called CitiStat. It was a game-changer. The shift to a data-driven approach not only impacted crime, but also provided a data-based decision-making platform for all city agencies. It gave birth to services that benefited citizens, like 311 non-emergency issue reporting and a forty-eight-hour pothole response guarantee.

I observed the success of CitiStat from my role in the juvenile justice system and was naturally drawn to the data-based analysis of problems.

The Birth of Maryland StateStat

When Baltimore’s Mayor O’Malley became Maryland’s Governor O’Malley in 2006, it was clear that CitiStat would be rolled out statewide. After his election, he named me Chief of Staff of the Department of Juvenile Services. Then, based upon my work in the juvenile justice system (and my tendency to disregard the status quo and naysayers), Governor O’Malley invited me to be the Director of StateStat.

In my following five years heading up the program, I was able to play a key role in a fundamental change in the way the Maryland State government operated. The goal from the beginning was to make data-driven decisions and welcome innovation.

We initially focused on the larger agencies within the state government. Governor O’Malley’s vision for Maryland was ambitious and bold. His goals included ending childhood hunger in Maryland by 2015, achieving a twenty percent decrease in crime statewide by 2018, and getting Chesapeake Bay to the “Healthy Bay Tipping Point” by 2025. He knew from his experience with CitiStat that data from multiple agencies would be essential to solving these complicated issues. He purposefully pulled together people from different departments, and their data, to find solutions.

Stories of Success with StateStat

Childhood Hunger

Take, for instance, Governor O’Malley’s goal of ending childhood hunger. It’s a pretty weighty goal. Plus, it’s hard to quantify success in ending hunger. Do you measure the number of free lunches delivered? How else can hunger be measured?

To make things more complicated, the government services that address hunger are disparate and separated by agency walls. Reducing the number of hungry kids is on the radar of schools, non-profits, and social service agencies. Until O’Malley’s StateStat program began in 2006, these stakeholders had never truly focused on pooling their data resources and working together to find solutions, even though they were all trying to accomplish the same end. Add to that the stigma of using food stamps or getting free lunches, and it was obvious that the State did not really have a handle on how many children were “hungry.”

Once we were able to gather all these stakeholders in the same room and compile the data we did have, it became obvious that school meals were the lynchpin. We determined that if we could feed more kids more meals at school, we had the best chance of improving their situation. We worked together to extend free meal services geographically where needed.

Free lunch turned into free breakfast, then dinner, then summer programs with free meals, too. What were the results of our efforts? In addition to no longer being hungry, these well-fed children performed better at school. We saw dramatic improvements in academic results and significant decreases in behavioral issues. While we couldn’t draw a direct correlation between the free meals and the improved academic performance, most leaders in Maryland agree that without the expanded free meals programs, Maryland wouldn’t have been awarded the designation of best schools in the country by Education Week five years in a row: 2008, 2009, 2010, 2011, and 2012 (“Quality Counts 2012,” 2012).

Safety for Foster Children

In another dramatic story of impact, we worked with the Department of Human Resources to find out how foster children in Maryland were faring. Were they safe and well cared for? As part of our analysis, we wanted to overlay a map of foster kids’ addresses with a map of the locations of the state’s most violent criminals and all registered sex offenders.

This location information had never been compared because it was separated in different departments. Also, because of legal barriers, it was not brought together as part of the screening process for foster parents.

Once we gathered together and visualized the data, we found that some of our most vulnerable children were living near some of our most dangerous criminals. We were able to dispatch safety assessments to the most worrisome situations. This effort is still one of my proudest accomplishments as the Director of StateStat.

Financial Stability

In 2007, Governor O’Malley decided to raise taxes. We saw the economic meltdown as an impending reality so we first sought to identify waste and eliminate it. In 2006, the O’Malley administration began the process of eliminating nearly six thousand State positions (O’Malley, 2013). In addition, while most states were cutting taxes like crazy, we made the very unpopular decision to raise taxes—and not insignificantly. Applying data helped us with this process.

Raising taxes, in hindsight, allowed Maryland to weather the recession better than most states. We weren’t immune to widespread, to-the-bone funding cuts, but by making data-driven decisions about revenue increases and service reductions, the state was one of only nine states to retain an AAA bond rating during the recession.

StateStat’s Results

The data-driven approach has made a significant impact in Maryland, from determining the right approach to better air quality to getting unemployed workers back on the job. Compiling information from different agencies has given us new insight and proved the power of data.

All told, StateStat’s success was nothing short of staggering. In its first three years, the state saved $20 million in overtime in our public safety agency alone. We cut costs by consolidating our print shops and state car fleets and reducing duplication in projects.

O’Malley also made good on his commitment to reduce crime. Violent crimes in the state decreased by twenty-five percent from 2007 to 2012. In fact, homicides were down twenty-seven percent in 2011 compared to 2006. The city of Baltimore, in particular, saw a historic reduction in crime.

In addition, the O’Malley administration managed a massive $8.3 billion in spending cuts in its first seven years, while recovering eighty-one percent of jobs lost during the recession—that’s the eighth fastest rate in the nation. Meanwhile, O’Malley’s team helped to expand healthcare coverage to more than 360,000 residents, most of them children.

In 2012, Maryland had the fifteenth lowest foreclosure rate in the nation. Eighty-seven percent of high school seniors graduated from high school, and fifty-six percent more students took Advanced Placement exams in science, technology, engineering, and math-related topics in 2012 than they had in 2006.

The list of achievements goes on and on.

How was such widespread success possible? I believe that it is our commitment to the philosophy of governing by data. It can’t be overemphasized. The data-driven approach wasn’t a side project. The StateStat program oversaw eighty percent of all budgets and personnel in the state. Agencies checked in with StateStat monthly, if not weekly, and StateStat was reviewed quarterly against Governor O’Malley’s fifteen stated goals.

In many ways, the timing of StateStat’s beginning was ideal. Data-driven decision-making is always useful, but in the midst of the Great Recession, we absolutely had to do more with less. In order to drive efficiency and make the right decisions for citizens, we needed to understand which services were the most valuable. The answers were in the data.

The Shift to Open Data

Our biggest challenge with early StateStat was that the data was coming from multiple places. Without secure databases, we didn’t have a central place to gather and store information. Instead, we were manually compiling spreadsheets. We needed a way to give a curated view of data from multiple sources to the public or internal teams.

We looked at business intelligence and database solutions, but in the end, open data won. Our open data portal centralized and standardized the data so any department could see another department’s information—and so could citizens.

We used the portal as the basis for public dashboards. We constructed a framework for setting goals that the public could appreciate and displayed the data related to our goals. By making the StateStat process more public, we helped citizens understand how the government was working for them.

Being open about our efforts helped us engage constituents. When we released public data and contextualized problems, citizens rallied and responded. When they saw the Recovery Act dollars we received, they started to point out projects needed in their communities. We created a website for them to make comments and give feedback on where we planned to spend the money.

Our efforts toward transparency through open data sparked a citizen-led, quality-of-life movement in the state.

Essentials of Data-Driven Government Success

After five years at StateStat, I felt like we had a good method for using data to make better decisions and engage the community. The three basic guidelines I developed were to:

• Curate data and inform people about the government conversation.
• Let developers access the data and allow the ecosystem to flourish.
• Nurture a collaborative environment where data analysts talk with the government and developers, and everyone understands the big picture and feels empowered to take risks and set ambitious goals.

Meanwhile, Governor O’Malley defined the leadership style needed to be successful. He bred this technology-based accountability into every level of Maryland’s government, and then supported it by putting the right people in positions to make data-driven decisions. He had the vision of how StateStat would become a repeatable, scalable machine that, armed with the right data, would be able to tackle practically any challenge the state faced.

This vision from the top and unwavering commitment to make it work is essential to creating the success we had in Maryland. In terms of transferability and scalability, this is the biggest hurdle to seeing a StateStat-style approach reproduced by governments worldwide. A strong leader is the key to ensuring widespread, long-term adoption within an organization.

Why I Moved to the Private Sector

While working on the open data architecture within the StateStat model, I collaborated with our open data platform provider, Socrata, to implement and optimize our open data portal. I recognized that Socrata had the scalable technology necessary to expand open data and data-driven governance to a much larger audience. This idea inspired me.

What if I could take what I learned in Maryland and design a product that would jump-start any government in the world to become data-driven?

Socrata invited me to their team to build that product. We named it GovStat, and its creation has been a career-defining project for me. It takes the best practices that decades of previous “stat” models had experimented with, and that Maryland’s StateStat relied on, and creates a platform that any state, city, or county can use to implement data-driven decision-making. I wanted to make the move to data-driven governance quick and easy and give other agencies a high likelihood of success.

For governments just starting with a data-driven approach, I wanted this tool to guide them through the best practices for setting goals, applying metrics, and tracking performance. For more seasoned organizations that have experience using open data, this platform can be used to elevate their performance tracking and reporting capabilities quickly and efficiently.

For those instances, and in every one in between, built-in citizen communication is key. With GovStat, goals, metrics, timelines, and tracking are reported to citizens through dashboards that give the data context and exposure. This is the completion of the loop, so that citizens can be involved in efforts to improve their quality of life.

People often ask why I left Governor O’Malley’s team to work in the private sector. I think that what we created with StateStat was transformative. I also think that government as a whole—locally, nationally, globally—needs to change. We can’t afford not to adopt a data-driven approach and miss out on the benefits of greater creativity and collaboration in decision-making.

Working with Socrata was the fastest way to move the needle in that direction. In all honesty, once you work for Governor O’Malley, you never really stop. I can now evangelize his message of the importance of this approach and spread the word that the often-difficult task of running government can absolutely work and bring amazing results for all.

The Future of Data-Driven Performance Management in Government

The success we have had in Maryland has drawn attention from other states and cities. During my time as Director of StateStat, many government officials came in to observe and talk to us about transferability. In the six short years since StateStat was born, many states, counties, and cities across the United States have adopted a data-driven approach to leadership. President Obama’s Executive Order on Open Data signed on May 9, 2013, signals a bold new commitment to a data-driven federal government.

Every “stat” program is slightly different, but they all operate based on the four tenets born out of NYC’s CompStat initiative:

1. Accurate and timely intelligence: Know what is happening.

2. Effective tactics: Have a plan.

3. Rapid deployment: Do it quickly.

4. Relentless follow-up and assessment: If it works, do more. If not, do something else. (Godown, 2009)

Data-driven governance is more than the open data it’s fueled by. The traditional view of open data’s greatest impact is that it creates a development engine to get data out into the open where it can stimulate economic activity in the private sector. There are plenty of real-life examples of this in the app development arena alone—from getting restaurant reviews alongside health code violation reports on your mobile device, to receiving a text when the city tows your car. These are the types of tangible quality-of-life improvements open data advocates have prophesied.

I believe, however, that the most transformative products of open data come when government is given efficient, low-cost access to tools that allow them to use data to drive progress toward things that are important to them and their citizens. Then, open data affects not only the people who rely on government services, but also extends to everyone who interacts with government in their day-to-day lives. We all have to interact with government, whether it’s getting a driver’s license or requesting a permit to build a home.

All these interactions influence our happiness index. Each of those interactions is impacted by the way government services its citizens. Open data has the power to be transformative in service delivery and influence the expectations that people have. It has the power to turn the tide on the idea that government can’t work.

One of my favorite sayings is that “the rising tide raises all ships.” There are examples of government working, even during the harshest austerity. The more people we have in government using open data, the better all governments will be. The more opportunities we have to proliferate best practices, figure out ways to build efficiencies into the enterprise of government, propel good behavior, and cultivate entrepreneurial spirit, the more we will see it come to fruition in reality.

This is all possible when governments are data-driven and align their resources to those outcomes that are supported by the tracking of data.

This is my vision. The creation of GovStat is a step toward engaging as many people as possible, as quickly as possible. The faster we can standardize the measurement of common problems, like crime, blight, and poverty, the sooner we’ll be able to replicate and scale what is working in small pockets on a broader scale. If New Orleans has a program that’s wildly successful at combating blight, let’s standardize that and create a platform for sharing that thought leadership. Why should Philadelphia recreate the wheel when much of the problem has already been solved?

We’re at a nexus of open data and performance insight, and the most explosive, ripe, and exciting opportunities lie in a data-driven governance approach. It’s not a matter of whether or not this approach can be adopted on a wide scale, but when. We must use data and be more agile in government in order to solve the problems we are facing locally, nationally, and globally. The tighter we weave together open data and performance, the faster we can realize the results and the richer the data will be.

About the Author

Beth Blauer is director of GovStat for Seattle-based Socrata, an open data platform provider. She served from 2008 to 2012 as director of Maryland’s StateStat and the Governor’s Delivery Unit for Maryland Governor Martin O’Malley.

References

• Education Week. (2012). Quality Counts 2012: The Global Challenge. Education Week, 31 (16). Retrieved from http://www.edweek.org/ew/qc/2012/16src.h31.html
• Godown, J. (2009, August). The CompStat Process: Four Principles for Managing Crime Reduction. The Police Chief, vol. LXXVI (no. 8).
• Fox, S. (2013, July 1). Pew Internet: Health. Retrieved from http://www.pewinternet.org/Commentary/2011/November/Pew-Internet-Health.aspx
• O’Malley, Martin, State of Maryland. (2013, January 30). State of the State: Better Choice; Better Results. Retrieved from http://www.governor.maryland.gov/documents/2013StateoftheState.pdf