How Data Literacy in Government Is Enhancing Data-Driven Decisions

What Is Data Literacy?

There are many definitions of data literacy, but most center on the ability to interpret and communicate about data. Deloitte defines it as “the ability to read, work with, analyze and use data ethically to solve challenges, drive innovation and create value collaboratively.”

A 2022 Data Foundation report refines the concept to “describe an individual’s ability to read, write, and communicate with data in context.”

The reason context matters so much is that in the modern world, almost anyone can be an end user of data, says Nick Hart, president of the Data Foundation and one of the authors of the report.

“But if you can’t appropriately apply the data analysis to the context that matters, we’re not using it right,” he says. “We want to make sure that context is appropriately matched for decision-making.”

Proper context includes not using state-level data for analysis aimed at counties or ensuring that the intended audience is the right one to receive a particular analysis of data. “The context of the type of data that you have for a particular question really does matter,” Hart says.

LEARN MORE: How public transit agencies analyze ridership data to better serve customers.

Why Is Data Literacy Important to Data Governance?

Data literacy is crucial to state and local governments’ management and use of data for several reasons, experts say. Data literacy is important, Hart says, “because it helps us connect between the decision that is being made and the meaning for what it has done.”

For example, he says, if a state or local government official wants to modify a program, it should make that decision based on the best available information, which may be in different sources and may include scientific or statistical information, descriptions of the problem, data about citizens’ living conditions and more.

“In order for our decision-makers and all of those who support decision-makers to make those decisions using data and evidence, we need to be able to combine the right kinds of information to support those decisions,” Hart says. “How do you do that? You become more data literate.”

Brian Burke, managing director for the Center for Technology in Government at the University at Albany, which has researched data literacy in government, says that data literacy intersects with how agencies manage, store and analyze data. Within agencies, there are some employees who have a range of data analysis skills, from basic research skills to using artificial intelligence and machine learning. Data literacy affects how that data is visualized and presented, usually within teams or to leaders who are making policy decisions.

Derek Werthmuller, director of technology innovation and services at CTG UAlbany, notes that for effective data governance at agencies, everyone involved must have some level of data literacy “that is appropriate for their contribution or use or creation of that data.”

For example, a government worker who is involved in the creation of data might need a deeper level of understanding than an executive who is making a decision based on it. They both might need to know how the data will be used so that the decision-maker can best understand its value, Werthmuller says.

“If someone doesn’t understand how their actions affect others down the pipeline, then it’s going to be difficult to institute change and improvement,” he says.

If data governance is the series of rules that people use to make decisions about how data within an organization is stored and used, then more people knowing  more about data can help to improve those rules, processes and organizational structure, according to CTG UAlbany Director J. Ramon Gil-Garcia.

“They are almost mutually reinforcing,” he says. “If you have better data governance, you can have better data literacy. If people really understand what they are doing with the data and limitations and all of that you can create better rules to govern the data in your agency.”

How Officials Better Measure Data Literacy in Government

Experts agree that state and local governments can do a better job of measuring and improving data literacy — and they say that it’s not as daunting a task as it might seem.

“Based on what we know about the states that don’t have formal data literacy programs, which is a large percentage, I would say that very few are measuring their data proficiency or data literacy today and closing that gap,” NASCIO’s Robinson says.

States must recognize that there is a business need in improving data literacy and that there is currently a systemic gap, Robinson says. State governments also should invest funding in and potentially issue policy directives for improving data literacy for their workforces, Robinson says.

“It’s not going to happen by accident, and it’s not going to happen by default,” he says, noting that it may take several years for states to improve once they decide to invest.

Hart says that government and the data community in general need to move away from the belief that a single data literacy training course is enough, and instead focus on data literacy as a program and practice that is established and sustained over time.

A key first step, Hart says, is measuring and establishing a baseline of current data literacy skills in the workforce. Doing so does not have to be excessively expensive or onerous, he says, noting that such a survey could be qualitative. Such measurement would be based on the organization’s needs.

“We want everyone to have some level of comfort and familiarity with core concepts, but that does not mean that every single person in the organization needs to know how to run a regression model,” Hart says.

Training programs should be tailored to specific users’ jobs and data needs, Gil-Garcia says. For example, there could be one course for data scientists and statisticians that is very advanced and technical and another for a general audience on how to use basic statistical analysis to draw conclusions. There could also be courses for those who are not going to perform data analysis but who need to know how to interpret data. And there could be training solely on data visualization and presenting data.

EXPLORE: How data center optimization helps state and local agencies improve services.

Data Literacy Training Leads to Better Data-Driven Decisions

Experts agree that data literacy is essential for improving data-driven decision-making in government.

Robinson notes that many states don’t have the budget to hire lots of specially trained data scientists. It’s therefore important for a broader number of government employees working directly on programs to be data literate. For example, instead of just having health informatics specialists focused on analyzing data related to the opioid epidemic, it would be beneficial also to have employees working in addiction programs well versed in the data.

“If they improve their ability to interpret and analyze the data and come to conclusions, I think all that is for the good,” he says.

Through a case study involving the New York State Department of Environmental Conservation and the New York State Department of Health, CTG UAlbany researchers found that the introduction of data analytics “helped produce evidence that could be leveraged for broad planning and designing of environmental interventions,” according to the center’s website.

DEC researchers were interested in a wide range of water quality issues in New York, Burke says. “We were able to show how these research scientists improved their processes for using this data for their own decision-making, as well as to inform leadership,” he says.

“Data literacy has a positive feedback loop for evidence-informed policymaking,” Hart says. “We need data literacy in order for evidence-informed policymaking to succeed. And if evidence in making decisions drives better decisions, it also means that we are hopefully seeing improved effectiveness, improved efficiency and ultimately better outcomes at all levels of government.”