This article revolves around the search for statutory recognition of the use of algorithmics (including rule-based systems) to make decisions. Thus, the focus is on whether a powerconferring rule is necessary before an administrative authority may issue a valid automated decision. By claiming that such a power-conferring rule is not only desirable but also a legal requirement, one contends that the shifting to automated decision making in the public sector is simply a by-product of reengineering bureaucratic practice through Information and Communication Technology. The issue is essential because there are several cases throughout Europe and beyond in which the latter is precisely the case. Moreover, values such as efficiency and flexibility to adapt to fast technological change could be considered sound justificatory arguments to purport that automated decision-making in the public sector should not be governed through the traditional arsenal of the rule of law. The primary purpose of the article is to offer a first conceptual skeleton to sow the seeds for further research on a normative claim that automated public administration decision-making should, nonetheless, be implemented in light of the rule of law. Secondly, it offers some recommendations for a legal framework to govern such a practice. The background idea is that focusing on the legal power to make an automated decision constitutes a necessary step to guarantee that the ethical and political principles that should undergird the adoption, development and implementation of any form of algorithmic decisionmaking in the public sector are appropriately and adequately taken into account.