The Double-Edged Sword of Data Governance: Integrity at Risk
In the June 28 post Planning for a Modern U.S. Data Infrastructure Must Proceed I recommended we plan now to improve how government data programs are governed.
That article supported strengthening collaboration and coordination across programs that have traditionally operated independently. While the immediate inspiration for that article came from a May 29, 2025 Science article, Fixing the US statistical infrastructure, my own experience in project management and consulting has always reinforced the need to improve how data are shared across multiple systems within any organization—even when those systems operate on different platforms in separate departments. There’s a reason why activities related to standardization, data lakes, data warehouses, and maintaining system interoperability consume so much time, money, and technology resources.
Unfortunately, there may also be a downside to more coordinated or centralized data governance: More centralized data governance can also make data manipulation, misrepresentation, and outright falsification easier.
The classic fictional example of this is found in Orwell’s 1984, where the government falsifies chocolate ration data to make it appear that rations are increasing, even though they are actually decreasing. We are seeing similar concerns today concerning the trustworthiness of critically important government data, especially in light of the current U.S. president’s firing of the head of the Bureau of Labor Statistics.
This type of situation presents a fundamental dilemma when justifying improvements to how government-sponsored data programs are managed.
On one hand, more coordinated data governance can improve data sharing, interoperability, and quality.
On the other, more coordinated data governance can also enable lying, manipulation, and censorship.
Over the course of my 50-year career in and around both government and non-government organizations that collect, manage, analyze, and distribute data, I’ve always championed the importance of data sharing, standardization, and accessibility. Batter and more organized data governance has always been a key element of this.
But now I find myself in a quandary. The very values I’ve long promoted—those that argue for strengthening, not weakening, data governance—can also be twisted to support misinformation and outright lies.
How do we solve this problem? I suspect the solution lies less in new or more clever technology than in the world of politics. Nevertheless, we need to be planning now for concrete steps to repair the damage being done to our national data infrastructure.
Copyright © 2025 by Dennis D. McDonald
