“Hands on DATA”: How to use data science to design public policies in Argentina

CAF -development bank of Latin America- organized in its offices in Buenos Aires the working group "Manos en la DATA" together with the System of Information, Evaluation and Monitoring of Social Programs (SIEMPRO) and the Sadosky Foundation, of which also officials from the National Ministries of Production and Labor and Education participated, as well as officials from the Government of the City of Buenos Aires and from the Municipality of Córdoba, among others.

The introduction was in charge of the CAF representative in Argentina, Andrés Rugeles, who assured that the appropriate use of data science techniques is key for the good design and monitoring of public policies. He was accompanied by SIEMPRO's national director, Soledad Cubas, and CAF's vice president of Knowledge, Pablo Sanguinetti, who added that the cost of using these techniques is particularly low for the State that, due to its scale and frequent and administrative relationship with people physical and legal, generates a large volume of information that can be used for statistical purposes to help devise better public policies.

During the opening panel, the director of the Data Science program of the Sadosky Foundation, Leandro Lombardi, spoke of machine learning techniques. "There is a myth that a huge brain in the cloud is predicting things. What we have is a set of techniques that, applied carefully, allow us to extract valuable knowledge from the available data, and thus respond to the demands of the public sector to improve its programs and policies. That's why I think that this workshop that we organize with CAF and SIEMPRO fits perfectly to meet these objectives."

Also, researcher Walter Sosa Escudero (CONICET and Universidad de San Andrés), added: “Big Data is a consequence of the data boom generated spontaneously by things that are interconnected. But be aware: Many people believe that we are proportionately better the more data we have, and that is unfortunately not true. This a completely different scenario, with all the good and bad that it entails.”

Finally, the chief economist of CAF, Lucila Berniell, gave details on the goals of this meeting: "With the workshop 'Hands on DATA' we proposed to bring data science to public policy. In this way, we will have better information to assist governments in the design, monitoring and implementation of their public projects and programs, with which we intend to contribute to reinforce a virtuous cycle of good policies informed by rigorous evidence."

Six workshops were conducted on different issues during the event: Detection of evasion (labor issues), identification of school drop-out risk (education), identification of cases of voluntary interruption and monitoring of pregnancy (health), prescription and description of detours and frequencies of urban buses (transport), metrics and maps (accessibility) and identification, characterization and prediction of deregistration from the universal child benefit program (social development).

To close the event, public officials, data scientists and CAF executives who participated in the working groups delivered their conclusions on the topics discussed, which will be followed-up until implementation of the projects.

With “Hands on DATA,” CAF contributes to the design of more effective public policies that promote development in Argentina and the region.