Programmatic Ideas and Targeting Metrics: Changes in the Definition of the Target Group of the Human Development Bonus in Ecuador

This article addresses a central problem in contemporary social policy: how targeting technologies shape the definition of the poor and, in doing so, reconfigure the boundaries of social citizenship. It asks how programmatic ideas and targeting technologies interact to define and redefine the target group of conditional cash transfers within processes of poverty governance. The study examines Ecuador's Human Development Bonus (BDH) and its Social Registry between 1998 and 2023 through a historical institutionalist framework, combining documentary analysis, administrative data, targeting error estimation, and microsimulations of poverty reduction using Foster–Greer–Thorbecke indicators. The findings demonstrate a co‐evolution between efficiency‐oriented programmatic ideas and increasingly sophisticated targeting technologies—from self‐targeting to proxy means tests and machine learning models. Social registries operate not as neutral tools but as political instruments that institutionalize specific conceptions of poverty and deservingness. Targeting errors, rather than anomalies, generate iterative technological reforms that dynamically redefine eligibility and stratify benefits. While the BDH reduces poverty incidence, gap, and severity, its effectiveness hinges on the state's capacity to classify households accurately. The article contributes to debates on targeting versus universalism and the politics of algorithmic governance, highlighting how technocratic designs shape distributive legitimacy in middle‐income welfare states.