Table of Links
3. Theoretical Lenses
4. Applying the Theoretical Lenses and 4.1 Handoff Triggers: New tech, new threats, new hype
4.2. Handoff Components: Shifting experts, techniques, and data
4.3. Handoff Modes: Abstraction and constrained expertise
5. Uncovering the Stakes of the Handoff
5.1. Confidentiality is the tip of the iceberg
6.2 Lesson 2: Beware objects without experts
6.3 Lesson 3: Transparency and participation should center values and policy
8. Research Ethics and Social Impact
Acknowledgments and References
8.3 Adverse impact statement
The primary adverse impact that this work could have would be playing into the hands of those who would weaponize the census for political gain. Given the heavily politicized nature of the census in general, and of the DP debate in particular, we cannot anticipate how or whether this work could be used to undermine the legitimacy of the census. Further, given the importance of the census for essential societal processes such as redistricting and resource allocation (which we address in our paper), we cannot dismiss the potential for such weaponization as inconsequential.
Unfortunately, there is indeed precedent for such adverse impact. During the DAS development process, the Bureau faced direct political threats to its data products, the most serious of which arose in March 2021 when the state of Alabama sued the Department of Commerce and the Census Bureau in federal district court, alleging that by adopting DP the Bureau had “manipulated” and “intentionally skewed” the redistricting data that they provided to states [1]. Furthermore, the coincidence of the decennial count with the 2020 presidential election, as well as the uncertainty around the Trump administration’s proposal to include a citizenship question on the census, drew political attention to the count. In a time when political actors were searching for any chinks in governmental armor, a Federal agency which was public about internal sources of error became an easy target. Indeed, the Bureau has faced bipartisan scrutiny for the troubles made evident by the implementation of DP – including allegations that DP was a Trump administration tactic attempting to ‘game’ federal funding allocations, and directly contradictory allegations that DP was a Democratic tactic to destabilize the Trump administration [18, p. 32]. Of course, such critiques undermine the ultimate role of the Bureau - to produce representative population counts - and further muddy the already-cloudy waters when it comes to identifying an appropriate implementation of DP.
Ultimately, we believe that the benefits that publishing our analysis might provide - hopefully, insights for both more effective algorithmic governance and more critical algorithmic scholarship - outweigh any potential risks for further weaponization.
Authors:
(1) AMINA A. ABDU, University of Michigan, USA;
(2) LAUREN M. CHAMBERS, University of California, Berkeley, USA;
(3) DEIRDRE K. MULLIGAN, University of California, Berkeley, USA;
(4) ABIGAIL Z. JACOBS, University of Michigan, USA.
This paper is
