What the Universal Registry of Income Can Learn from Other Countries

The new universal income registry, known as the Registro Universal de Ingresos (RUI), began operations on August 1st. At its core lies a critical question that has long preoccupied social policy: can one reliably estimate someone’s income when they have left no formal trace? The answer directly determines who receives subsidies and who does not. To address this, the system employs a statistical tool called proxy means testing, which calculates a likely income based on correlated factors such as housing, assets, and household composition rather than relying solely on self-reported data or existing records. While Colombia is not pioneering this approach, many Latin American countries have used similar methods for two decades, the international evidence suggests these models often fail to accurately identify the poorest segments of the population. A review by Development Pathways found that when such models are used to target the top 10% of the poorest individuals, the rate of design exclusion, where poor people are missed by the model, ranges around 60%. Even at lower coverage levels, the error remains close to 50%. These figures stem from statistical limitations, not poor implementation. The models typically explain less than half of the true variation in household consumption, meaning they inherently miss large portions of the population they aim to serve. This issue is not theoretical, it has already played out in other nations. In Chile, during the pandemic, the government quickly developed an emergency income support program using a different metric than the standard household registration system. The result was inconsistent classifications: a family could appear in the top 20% under one measure and the bottom 40% under another. Those who lost jobs between updates simply disappeared from the system until the next review, with no fault attributed to them. The problem lay in the model itself, not in its execution. Closer to home, Brazil's Cadastro Único, which supports the Bolsa Família program, offers a more recent example. This system regularly cross-references information with tax records, pension databases, and formal employment registries. When discrepancies arise, benefits are automatically suspended without prior notice to the affected households. This approach highlights both the potential and pitfalls of integrating multiple data sources. Recent academic studies have also shown that household surveys used to calibrate and audit such systems systematically underestimate the true scale of programs. These surveys often fail to adequately capture populations living in indigenous territories, informal settlements, or without fixed housing, groups that should be prioritized in targeted social policies. As a result, the integration of data sources may not only misclassify households but also consistently exclude the very groups that social programs aim to assist. For Colombia, the challenge with the RUI is twofold. First, the methodology of proxy means testing must remain consistent and transparent, avoiding abrupt changes in criteria, as happened in Chile. Second, the system must ensure that the integration of diverse data sources does not inadvertently exclude vulnerable populations. This requires ongoing monitoring and adjustments to prevent systematic errors. The success of the RUI will depend not just on its initial design but on how effectively it adapts to the realities of the communities it seeks to support.