Most states don’t have data quality programs, despite interest from top IT leaders

Though most top technology officials in state governments recognize the importance of maintaining data quality for the adoption of artificial intelligence, only 22% of states operate data quality programs, according to survey results published Tuesday.

Through a recent survey conducted by the professional services firm Ernst & Young and the National Association of State Chief Information Officers, state CIOs and chief data officers from 46 states shared the gaps between their organizations’ desires and capabilities.

According to the report, states are “grappling” with the task of establishing enterprisewide data quality programs that will allow agencies to use the latest artificial intelligence software.

“For state and local government executives, the imperative of maintaining high-quality data cannot be overstated,” the report proclaims. “Accurate data is the cornerstone of insightful analytics, strategic decision-making and the effective training of artificial intelligence (AI) models.”

Among those surveyed, 89% said data quality was important, very important or critically important, yet less than a quarter of states reported operating data quality programs. The survey results and interviews indicated that states are still in the early stages of preparing their datasets for AI — 57% reported their states are “reactive” when it comes to data quality efforts. Many reported that data cleaning efforts are treated as “one-off” projects.

Report authors recommended states pursue official means to further their data quality governance.

“For data management efforts to be truly effective and enduring, they must be institutionalized through executive orders or, more reliably, through legislation,” the report reads.

Indiana CIO Tracy Barnes is named in the report as recommending that officials frame such efforts so as to align them with the priorities of lawmakers, who are more concerned with project outcomes than abstract notions of data cleanliness.

The report concludes with a handful of recommendations, including that states instantiate data governance and supporting operating models, and that they set data quality standards “at the point of data creation.”

“Implementing data quality standards at the point of data creation ensures accuracy and consistency from the start, reducing the need for costly corrections later,” the report reads. “This proactive approach enhances decision-making and compliance and is a critical component of a data-driven organization’s governance strategy.”