Nikos Askitas

Opinion copulas

In a paper entitled “Explaining opinion polarization with opinion copulas,” Nikos Askitas shows how “opinion copulas” might help explain why in a world without communication obstacles polarization arises despite our best models (i.e. stochastic bounded confidence models) being unable to explain why. In the parlance of causal inference, opinion copulas, well supported by findings in social psychology, contain unobserved confounders. The core technical observation of the paper is that joint distributions (e.g. ordinal sums of copulas) of uniform marginals might be multimodal.

Mr. Askitas is a mathematician, data scientist and economist who explores new avenues of research and policy in labor economics and in social science more generally. His mathematical research has been in topology and more specifically in four-dimensional topology and classical knot theory. His more recent empirical work is centered around using the internet as a data source for social science and more specifically for “nowcasting” in a wide range of topics which span unemployment forecasting, health, housing, business cycle “nowcasting” or even traffic jams among others. He is also interested in evolutionary game theory and in non-linear opinions dynamics. His work has been published in leading journals.

Nikos Askitas’s innovations include JoSuA, a web application for remote processing which allows roaming researchers to compute against data which is not allowed to leave the premises of its data custodian. He is also the inventor of EDDI, the annual European DDI user conference which is a roaming conference jointly organized by the IDSC and GESIS under the auspices of the DDI alliance.

Mr. Askitas is responsible for the Toll Index, a monthly leading indicator of the German business cycle. The Toll Index, an example of “economic telemetry,” is based on MAUT data and exploits the idea that just-in-time delivery turns the highway into a warehouse and hence production and transportation are coincident. Since transportation measurements are faster due to the underlying technology of the MAUT, they provide a sneak preview of production.

Nikos Askitas joined IZA in 2000 and is the Institute’s Coordinator of Data and Technology. As head of its Research Data Center (IDSC) and Head of its ICT unit he is responsible for all data and technology issues and is hence a member of both the research and the service units of IZA. Nikos Askitas holds a PhD and an MA in Mathematics from Indiana University, Bloomington. His BA in Mathematics and Computer Science is from Aristoteleion University, Thessaloniki.