No longer just nice to have: Data literacy essential to digital government

At least a basic level of data literacy is becoming table stakes for digital transformation at public-sector organizations. To ensure workers have that, many agencies are developing their own data literacy programs.

While it’s imperative that staff get up to speed, a goal of training every employee in every possible skill set is unreasonable, said Adita Karkera, chief data officer at Deloitte Government and Public Services and an author of “Data Literacy for the Public Sector,” published in March 2022. Instead, agencies can take a big step forward by just getting started.

“If we can start off with a phased, well-thought-out approach on how we will bring change about in a phased manner, that is still a victory for that organization,” Karkera said. 

Consider the city of Syracuse, New York. As recently as five years ago, its Department of Public Works had no data to help prioritize which roads needed repaving. Now, it uses a data-based model not only to make those predictions, but also address equity.

A gradual increase in data literacy enabled that digital transformation. First, the Syracuse Metropolitan Transit Transportation Council, a state-designated metropolitan planning commission, began scoring the quality of the city’s roads each year on a scale of 1 to 10. 

When the council shared those scores with the city, its Office of Analytics, Performance and Innovation developed a road pavement prioritization model by combining the ratings with additional data, such as information from field operations, spatial data and SYRCityline, an app through which residents make requests for non-emergency services, including road repair. 

After using the model for three years, the innovation team began looking at ways to ensure that neighborhoods with historically underserved populations weren’t overlooked for paving projects. That meant finding and adding more data to the model.

The innovation office partnered with the the Department of Public Works to look at previous road reconstruction projects and match that information with demographic data from the census. They checked how many residents living along road construction projects were people of color and living below the poverty line. They also factored in how many were over 65, had disabilities, were rent-burdened and had low educational attainment to see whether those factors affected or predicted which roads were resurfaced, said Jason Thomas, a data analyst with the innovation office. “Last year, we included what we call an equity score in the model,” he said.

“It’s this iterative process where five years ago, we didn’t have any data on it, then four years ago, we created this data-driven model and a year ago, we created this model that now includes equity,” he added.

The result is real change: In April 2023, the city announced a plan to reconstruct almost 23 miles of roads in every quadrant of the city and seal and resurface another 40.

Although many public-sector leaders prioritize use of data—data management and analytics ranks sixth on the National Association of Chief Information Officers’ State CIOs Top 10 Priorities for 2024, for instance—they are at varying levels of data literacy promotion. 

For example, Syracuse’s innovation office serves primarily as a consultant for other departments. “We’re focused on specific projects, not so much data literacy for the sake of it,” said Jason Scharf, data program manager there.

By contrast, data literacy is part of orientation for all new employees of Tempe, Arizona, government offices. 

“Everyone should have basic data literacy,” said Stephanie Deitrick, Tempe’s chief data and analytics officer. “I think it’s just important for everybody to have that foundational understanding, whether they’re decision-makers or not.”

Still, not everyone needs the same type of data skills, which is why she offers a variety of  educational opportunities. Information technology training includes classes on analyzing geographic information system data, for example.

The city’s first responders are among the most robust data users, Deitrick said. Some of them recently completed voluntary multiday GIS training designed to help them think about using data in planning for special events, such as concerts. 

“I’ve had a couple of them already requesting access to be able to do some new things,” Deitrick said. “The nice part about things like that is even if they aren’t the ones that end up doing the work, if you understand what’s possible, it lets you think of new things that you … might not have thought about otherwise because you don’t know what you don’t know.”

In Texas, Monica Smoot, data literacy program administrator at the Department of Information Resources, is building a three-tiered program to help data management officers educate business leaders and stakeholders about data. 

Although still a work in progress, the program launched in October 2022 with the “Introduction to Data Literacy” course. So far, Smoot’s team has completed seven of nine first-level courses. And as of December 2023, 37 state agencies and public institutions of higher education were participating in department-hosted courses. Sixty-seven people have enrolled in them so far.

“Originally, we planned to offer the data management officers our courses that we develop on the learning management platform hosted by the Texas Department of Information Resources,” Smoot said. “However, the demand for making those courses available to a wider audience from those data management officers and other data users prompted us to explore additional options.”

As a result, the courses are available on the department’s YouTube channel, and they’re sharing course files with other agencies, so that they can host them on their own learning management platforms.

Despite the recognition that a lack of data literacy is an obstacle to transformation, progress is slow. One reason is that reskilling efforts are often focused on final decision-makers and don’t trickle down to people using data for operational decisions, Tempe’s Deitrick said.

“There’s this assumption that if you tell people to use data to make decisions, somehow magically everybody knows what that means,” she said. “And if they don’t, I think people are afraid to ask because nobody wants to feel stupid or like they don’t know what they're doing,” she said.

“It’s exciting to see people really digging into the basics,” Deitrick said. People are helping lift each other up “so that everybody is starting from a common understanding.”