Initiative overview
This work is based on the analysis of a large administrative dataset obtained from the DIAN (Colombian Tax and Customs Authority), comprising 686,215 records. The data originate from official tax reports, including: the Annual Simple Alternative Minimum Tax Return (IMAS) for self-employed individuals, the Annual Alternative Minimum Tax Return for employees (IMAS), the Declaration of Income and Complementary Taxes for resident natural and assimilated persons and resident liquid successions, as well as the Declaration of Income, Supplemental Income or Wealth for legal and assimilated persons, non-resident natural and assimilated persons, and non-resident liquid successions.
The analytical procedures applied to the DIAN database and to the Chambers of Commerce datasets were conducted separately during 2018, ensuring methodological independence and robustness. The evidence obtained and the proposed methodological framework are highly relevant for the design and implementation of algorithm-based public policies, particularly those aimed at improving the targeting, prioritisation, and effectiveness of tax auditing and fiscal control processes.
Results, outcomes and impacts
As a result of this initiative, 1,000 companies that evaded or failed to pay departmental taxes were detected.



























