In Print: Volume 85: Number 6
By Danielle Keats Citron
85 Wash. U. L. Rev. 1249 (2008)
Distinct and complementary procedures for adjudication and rulemaking lie at the heart of twentieth-century administrative law. Due process requires agencies to provide individuals notice and an opportunity to be heard. Through public rulemaking, agencies can foreclose policy issues that individuals might otherwise raise in adjudication. One system allows for focused advocacy; the other features broad participation. Each procedural regime compensates for the normative limits of the other. Both depend on clear statements of reason.
The dichotomy between these procedural regimes is rapidly becoming outmoded. This century’s automated decision making systems combine individual adjudications with rulemaking while adhering to the procedural safeguards of neither. Automated systems jeopardize due process norms. Hearings are devalued by the lack of meaningful notice and by the hearing officer’s tendency to presume a computer system’s infallibility. The Mathews v. Eldridge cost-benefit analysis is ill-equipped to compare the high fixed cost of deciphering a computer system’s logic with the accumulating variable benefit of correcting myriad inaccurate decisions made based on this logic. Automation also defeats participatory rulemaking. Code, not rules, determines the outcomes of adjudications. Programmers inevitably alter established rules when embedding them into code in ways the public, elected officials, and the courts cannot review. Last century’s procedures cannot repair these accountability deficits.
A new concept of technological due process is essential to vindicate the norms underlying last century’s procedural protections. This Article will demonstrate how a carefully structured inquisitorial model of quality control can partially replace aspects of adversarial justice that automation renders ineffectual. It also provides a framework of mechanisms capable of enhancing the transparency, accountability, and accuracy of rules embedded in automated decision-making systems.