The Autonomy Gap: Response to Citro et al. and the statistical community

Abstract

AbstractWhile the threat of biased AI has received considerable attention, another invisible threat to data democracy exists that has not received scientific or media attention. This threat is the lack of autonomy for the 13 principal United States federal statistical agencies. These agencies collect data that informs the United States federal government’s critical decisions, such as allocating resources and providing essential services. The lack of agency-specific statutory autonomy protections leaves the agencies vulnerable to political influence, which could have lasting ramifications without the public’s knowledge. Citro et al. evaluate the professional autonomy of the 13 federal statistical agencies and found that they lacked sufficient autonomy due to the absence of statutory protections (among other things). They provided three recommendations to enhance the strength of the federal statistical agency’s leadership and its autonomy to address each measure of autonomy for all 13 principal federal statistical agencies. Implementing these recommendations is an initial and crucial step toward preventing future erosion of the federal statistical system. Further, statisticians must take an active role in initiating and engaging in open dialogues with various scientific fields to protect and promote the vital work of federal statistical agencies.

Keywords

• Applications and case studies
• Education
• Federal statistical agency
• Government statistics
• Official statistics
• Other