The volume of socio-economic data has risen significantly in recent years. At the same time, its complexity is steadily increasing. A closer look at the data that is compiled for decision-makers reveals that we are far from making full use of the ever-growing mountain of data. A team of researchers drawn from the fields of statistics, machine learning, economics, social sciences and computer science is seeking to develop new methods that will allow the extrapolation of improved conclusions from the data. The project is funded by the Austrian Science Fund FWF.
“We need innovative and powerful tools for the analysis of big data in economics and the social sciences,” Gregor Kastner, coordinator of the project entitled “High-dimensional statistical learning: New methods to advance economic and sustainability policies” states. He joined the Department of Statistics at the University of Klagenfurt in 2020 and is continuing his work on the project here.
The team of researchers hopes to use the new methods to analyse complex data sets involving situations in which either the number of observations, the number of potential time series, and/or the number of variables included is very large. The aim is to provide answers, based on socio-economic data, to questions such as these: How do market and economic uncertainty affect income inequality? What are the relationships between greenhouse gas emissions and macroeconomic indicators? Which role do tweets play in the evolution of the prices of crypto-currencies? Which policy measures are most effective to foster sustainable urban mobility patterns?
Gregor Kastner underscores the need for this research: “Given that policy makers are usually interested in evaluating their policies in quantitative terms, it is crucial to have robust econometric tools at our disposal for the purpose of forecasting and running simulations.” However, considering the increasing complexity of the economy, vast amounts of information need to be used to adequately reconstruct the underlying causal structures and provide a comprehensive picture of the potential conduits through which policy interventions are communicated. The new methods developed within the project, including those based on Bayesian statistics, will ultimately serve to generate better foundations for decision-making.