IE 11 Not Supported

For optimal browsing, we recommend Chrome, Firefox or Safari browsers.

Government Must Use Data to Drive Decision-Making

Open data portals were a good first step toward putting the massive amount of information government holds to work. But now, experts say data should drive storytelling and decision-making across the enterprise.

Adobe Stock
When San Francisco launched its first open data portal more than a decade ago, it marked itself as a leader in the space. That portal even came with an executive order promoting open data from then-Mayor Gavin Newsom, helping the city reduce 311 calls and save millions. This was in 2009.

Since then, data work in government has continued transforming across the country. Today, data enables states and localities to improve storytelling so that the layperson can draw useful conclusions with data. It also informs investments and helps government direct resources to jurisdictions with the greatest needs.

Indeed, experts say public-sector data work is maturing as government agencies invest more and their teams gain experience. Stephen Goldsmith, a Government Technology columnist, is a leading voice on public-sector data use. He heads the Civic Analytics Network, a Harvard-based group that is the premier support organization for data work in local government.

Goldsmith explained that as cities got better data, it led to better visualizations. City issues are often connected and using data to map a problem’s prevalence — along with the resources to combat it — allows for collaboration between public, private and nonprofit.

This, however, is just one example of how data in local and state government has evolved. Goldsmith and other experts say data’s evolution in the public sector is ongoing, but one thing is clear — better use of data still has great potential to help governments better serve the people.
A street in Arlington, Texas, at night
Arlington, Texas, has established a Data Governance Innovation Team that comprises 40 members from across city agencies to work on data together.
Adobe Stock


Getting better with data starts with the collection process itself.

One of the challenges at the local level is systemic, as the fragmentation of data sources — often coming from different sectors — can create a limited view of what’s needed to solve community problems, Goldsmith said.

As Michael Lawrence Evans, program director with the Boston Mayor’s Office of New Urban Mechanics (MONUM), put it, cities need to think about not only quantitative data, but also qualitative data. Boston acquires this through interviews with residents.

For example, in a project the city has embarked on to map areaways, city staff are using oral histories to fill data gaps. It’s proving effective. And tech leaders in other cities agree with the approach.

“Collecting data is easy; collecting data with integrity is difficult,” explained Craig Poley, CIO of Arvada, Colo. In his city, the work to improve data collection involves formalizing the approach to data governance.

In Arlington, Texas, the city works with data from the data-as-a-service perspective. Sarah Stubblefield, Arlington’s strategic initiatives manager, emphasized the importance of understanding community context so residents connect to the story being told.

One common issue many local governments face with data collection is that certain sources of data, such as feedback from town hall meetings and 311 call logs, come from a small group of residents who are most active. It is, essentially, not representative of the whole or even the majority of people.

Milda Aksamitauskas, a fellow with the State Chief Data Officers Network at the Beeck Center for Social Impact + Innovation at Georgetown University, says there are several major trends in data usage at the state level, too.

States are increasingly hiring chief data officers, dedicating someone to lead the work. Beyond that, states are also building teams to support this work and add capacity. Aksamitauskas credits this change in part to the COVID-19 pandemic, as public expectations of data availability changed.

As states increase investment in data and data capabilities to expand infrastructure, Aksamitauskas says they are more likely to find the answers to the increasingly complex questions they’re asking.

Indiana Chief Data Officer and Director of the state’s Management Performance Hub Josh Martin explained that there are many challenges in making data collection more equitable, some of which can be attributed to legacy systems, while others trace to historic means of gathering or saving information.

“On the other side of that coin, just because you don’t have it all there doesn’t mean you can’t try to fill in gaps,” Martin said.

Martin suggested states consider adopting data collection standards that evolve continuously with the world of data to help better fill holes.


Data can be a powerful tool in storytelling, a strong way to back up engaging narratives with a bedrock of hard facts, and visualizations can enhance the story for those who are not data experts.

Goldsmith elaborated on this idea, saying that visualization around place, specifically, drives a lot of activity in cities, whether that be looking at a neighborhood’s environmental data to improve public health, or looking at a community’s walkability and the need to invest in transportation access in certain areas.

“It’s the visualization of data that allows not only better use by public employees, but more engagement by the community as well,” said Goldsmith. “And visualization of data exposes inequity and drives action.”

One common visualization tool is dashboards. Martin believes COVID-19 has played a role in the shift to government entities making data more accessible for public consumption through such tools.

In Arlington, Stubblefield says visualizations like web maps make data within the city’s open data portal so much easier to read than a raw data set, which to an untrained eye is essentially just a spreadsheet cluttered by numbers. She also noted that public confusion about the complexity of government operations can be mitigated through visualizations like the city’s development dashboard.

“Storytelling is a primary skill for CIOs, and we have technical realities that we have to convert to business language,” Poley said, describing a data team’s work in three parts: data engineering, data science and visualization.

Goldsmith believes the advancement of data storytelling over the next five years will deliver something akin to intuitive products like smartphones, with better user interfaces and more usable data, configured for specific needs.

Ariel view of Baltimore
Data is the “lifeblood” of decision-making in Baltimore, said Assistant City Administrator Ariel Giles, helping predict violent crime and inform the distribution of federal funding.


But good data use in government isn’t just about getting information to residents. In fact, data is equally valuable — if not more so — for government internally. Right now, for example, it can literally make your local government money.

Currently, a series of sweeping federal funding bills have allocated big chunks of money for states and localities to invest in digital equity, infrastructure and other areas. But making a strong case for that money requires data.

This allocation of funding — which Goldsmith says the feds set up well, requiring the use of data to both apply for and report the use of their funds — is a state issue. The way the system is designed, the money will first go from the federal government down through each governor’s office, and then to communities and cities. What this means is that — once again — good data is vital at all levels of government.

One thing changing with federal funding is that applications are less focused on the narrative and more focused on the ability of an entity to tell a story through numbers. “Not having the data is a huge disadvantage to being able to fill out these applications,” Stubblefield said.

She also noted that having updated data easily accessible simplifies the process of reporting back to the federal government.

Boston CDO Stefanie Costa Leabo said that being intentional about collecting information beyond the minimal federal requirements can help the city track the outcome of federal investments several years into the future.

In Baltimore, which is resource-constrained like most cities, Assistant City Administrator Ariel Giles said that federal funding from the American Rescue Plan Act has led to significant investments in the city’s technology posture through projects like the creation of a Digital Services Team.

To ensure equitable investments are being made throughout the community, the city is building ways to track impact internally in addition to satisfying federal reporting requirements. Smart city technology, like sensors to monitor infrastructure, will help Baltimore gather data to extend the usefulness of those federal dollars and reduce operational costs moving forward.

Laptop with the screen showing Boston's CityScore dashboard
Boston’s CityScore dashboards were originally conceived to help city leadership understand department performance, but they are also public-facing, creating more transparency for residents.


Some government entities are specifically structured in a way that enables collaboration, whereas others have to get more creative. But when it comes to collecting data, collaboration is vital.

For Indiana, the Management Performance Hub acts as a facilitator, working as part of other state agency teams to break down silos, with the unique role of focusing on leveraging data as a strategic asset, Martin said. It expands the state’s capacity to leverage data, empower agencies and convene stakeholders.

In Boston, the Mayor’s Office of New Urban Mechanics works closely with the Department of Innovation and Technology, often leveraging the knowledge of both teams to solve city problems with data, pairing the former’s analysis skills with the latter’s user experience knowledge.

“If we have more people doing data work, more data work will get done,” Costa Leabo said.

Giles stressed that in Baltimore, data is the “lifeblood” of the administration’s decision-making process, but it does not exist in a vacuum. In recent years, the city has worked to collect and disseminate information across the enterprise internally. The administration also works with the Office of Performance and Innovation, a team that focuses on how the city uses data to measure and improve.

In Arlington, bringing this data together is achieved in part through the work of the Data Governance Innovation Team, which includes 40 members representing different departments within the city to make use of the large quantities of siloed data.

This coordinated approach can help cities tackle complex issues whose contributing factors span departments.


While the term “data-driven decision-making” has become a sort of buzzword, the fact remains: There is a lengthy and growing list of examples of how data drives government.

Storytelling is a primary skill for CIOs, and we have technical realities that we have to convert to business language.
One such example is CityScore, Boston’s ranking system that uses data to provide executives with key metrics to understand how departments are performing, said Costa Leabo. And although this was designed to be used by leadership, the city also made dashboards public so that it could double as an outward-facing accountability tool.

This work inspired a similar project in Arvada. Poley says his city is currently in the phase of determining the core metrics that will be used to score different departments within the city.

A similar program exists in Baltimore, with CitiStat, the central performance management mechanism that city government leaders have used to measure service delivery for decades.

While that initiative uses data internally for executives, another data-driven project in Baltimore that has directly impacted residents is the Group Violence Reduction Strategy. As Giles noted, this effort leverages data to identify where violent crime is and which individuals may be at risk of being victims or perpetrators. The city then works to connect those individuals with resources to help reduce violent crime: “And it has been impactful in that it’s seen significant reductions in violent crimes in those districts that have it.”

States are also asking more complex questions, Aksamitauskas said, which require cross-agency collaboration to answer, from workforce development, to safety net benefits, to education.

Looking forward, government entities are exploring how to best adopt tools like predictive analytics and AI that use data. As Poley put it, being able to effectively leverage predictive analytics is “the dream.”

While many entities are still building capacity to effectively get to this point, some are already implementing these tools — for example, Baltimore is using AI to predict and prevent collapsing rooftops.

“Today, the use of data and the definition of efficiency is how we’ll use the data and technologies to make our department run better,” Goldsmith said. “Five years from now, that will be as obsolete as PDFs and open data are today, because five years from now, the issue will be, ‘How are we delivering the data that we have to the individual who wants to use it?’”
Julia Edinger is a staff writer for Government Technology. She has a bachelor's degree in English from the University of Toledo and has since worked in publishing and media. She's currently located in Southern California.