This big data exercise was carried out by a team of experts, Veronika Vilgis, Yu-Chang Ho, Xin Jin, Kangbo Lu, Karla Rascon-Garcia and Matthew Reese, coordinated by Martin Hilbert.
https://repositorio.cepal.org/handle/11362/45484
This report explores the opportunities and challenges of the systematic use of publicly available digital data as a tool for the formulation of public policies for the development of the digital economy in Latin America and the Caribbean (LAC).The objective is to share the lessons learned in order to advance in a research agenda that allows the countries of the region create alternative measuring tools based on the digital footprint.
Its substantial insights stem from the content of its six chapters, which assess the current state-of-art of complementary aspects of the digital economy in LAC, namely the labor market and digital skills; technology prices; micro-, small- and medium-sized enterprises; broadband; cryptocurrency; and social media. We show that the digital footprint left behind by labor market portals, e-commerce platforms, and social media networks allow obtaining unprecedented insights, both in terms of reach and detail. For example, in terms of reach, we are often able to gather naturally harmonized data for more than 15 LAC countries, track 2.5 million small enterprises across the region, or 35 million statements related to the Sustainable Development Goals (SDGs). In terms of detail, we are able to go beyond what official survey statistics contain and track the gender of small business ownership in the Caribbean, or the effective hourly salary for a specific skill like data entry by gender. Beyond what official administrative registries contain, we are able to distinguish active use of 3G and 4G mobile access, or the price for emerging technologies like drones. The report showcases some 30 figures that exemplify the presented opportunities.
The methodological contributions of the report stem from the lessons learned from this exercise of Big Data analytics. In this sense, the report serves a rough guide for practitioners interested in using modern data science for development policies. Naively, when thinking about the ‘big data’ paradigm, people seem to imagine that, with enough computational skills, one would simply go online to collect data, and the entirety of reality would suddenly stand point-blank in front of any observer in all of its details in real time. In reality, the process is more reminiscent of the proverbial inspection of an elephant by a group of blind people, where data scientists take the role of the blind touching very distinct parts of the whole, never being able to grasp it all at once, but trying to piece together irreconcilable pieces of evidence. Making sense of ‘big data’ in a meaningful way includes computational challenges, but goes much further, and touches on the definition of data science as the convergence between computer science, statistics, and its substantive application area. Issues of representativeness, generalization, harmonization, data quality, definition of variables and indices quickly become the main concerns of data science in practice. We report some of the lessons learned throughout this exercise, which took place during January and March 2019, and summarize them in the final chapter.
See also how this report came about:
Prof. Hilbert and a team of five UC Davis students are developing a big data online observatory for the UN Secretariat in Latin America and the Caribbean…. Three students spent three months in Chile, while two others continue to stay there for even longer, and one new student will be joining the team.
Download (, Unknown)
