Title: Safeguarding Statistical Independence in the Age of Political Pressure and Automated Decision-Making: A Case of the the United Republic of Tanzania Experience. IPS 1393 Official statistics play an important role in the effective functioning of the modern state, guiding governance, structural transformation, evidence-based policy, and resource mobilisation. However, the integrity of National Statistical Systems (NSS) is increasingly contested under the dual pressures of executive political control and the rapid proliferation of Automated Decision-Making (ADM) systems driven by advancements in artificial intelligence (AI) and machine learning. This paper examines the safeguarding of statistical independence through the empirical case of Tanzania a jurisdiction whose recent legislative trajectory provides greater latitude for data production within the national statistical system without compromising professional ethics, alongside ongoing developments in public digital infrastructure and critical lessons for statistical governance across emerging economies. The Tanzanian experience sheds light on the evolution of the Statistics Act of 2015 and its pivotal 2019 amendments, detailing the advantages gained and the lessons learned in engagement with domestic stakeholders and international development partners. Following these reforms, the Government of Tanzania established a well-defined National Statistical System mandated to compile official statistics adhering to a commonly agreed National Quality Assurance Framework. Concurrently, Tanzania’s rapid expansion into digital public administration has introduced a new frontier of statistical vulnerability. The integration of the National Identification Authority (NIDA) biometric database with mobile financial services, electronic tax systems, and the Tanzania Social Action Fund (TASAF/PSSN) automated cash transfer programmes illustrates how statistical and administrative data directly govern algorithmic resource allocation. When foundational statistical baselines—such as regional poverty lines, consumer price indices, or census registers—are subject to executive sanitisation, automated targeting mechanisms risk institutionalising systemic exclusion in public service delivery. In communities characterised by high levels of informality and limited biometric coverage, unrepresentative or politicised training data systematically disenfranchises vulnerable populations under the guise of algorithmic objectivity. While the enactment of the Personal Data Protection Act of 2022 marks a positive milestone towards comprehensive data governance, frameworks regulating algorithmic transparency, automated profiling, and model validation remain underdeveloped. To address these compounding risks, this paper proposes an institutional and legal reform model grounded in the Tanzanian experience. It advocates for the constitutional entrenchment of the National Bureau of Statistics (NBS), guaranteeing merit-based, non-revocable leadership tenure and ring-fenced domestic funding mechanisms. Crucially, the paper calls for a statutory mandate requiring algorithmic impact assessments and public registers for all automated systems deployed across public administration, thereby ensuring administrative contestability and independent methodological oversight. Preserving statistical independence in Tanzania is thus positioned not merely as a technical standard of scientific precision, but as a vital constitutional safeguard for democratic accountability, civil liberties, and equitable governance in an increasingly algorithmic state. Keywords: Statistical Independence, Tanzania, National Bureau of Statistics (NBS), Automated Decision-Making, NIDA, Data Governance, Algorithmic Accountability, Statistics Act.
