Giulio Grossi

Post-Doc in Statistics @University of Florence

Hi! I'm a Post Doc  in Statistics at the University of Florence.
During my Ph.D. I have devoted my attention to the study of methods for panel data in causal inference, advised by prof. Alessandra Mattei. 

 During the post-doc period, I have worked with prof. Emilia Rocco on the costruction of socio-economic indicators with neural networks. 

Then, I joined the BayesMeCOS project, funded by European Union - Next GenerationEU, UNIFI Young Independent Researchers Call - BayesMeCOS Grant no.B008-P00634 to study Bayesian Methods for clinical and observational trials. 

Currently, I am PI of the SPARKLING - SPAtial and tempoRal Knowledge for causal learning IN Global climate change awareness, funded  by the UNIFI4FUTURE, University of Florence, grant n. B17G24000250006. 

In this project I am willing to disentangle what are the main drivers for constructing climate change awareness in the citizenship and what are their relations with natural disasters. A relevant portion of the work will be tangled to the development of spatio-temporal methods for causal inference. 

Prior to the Ph.D., I have joined the CINTURS in Faro, PT, and worked as research fellow in IRPET - Istituto Regionale per la Programmazione Economica Toscana, in Florence, IT. 

My main research topics are causal inference under interference, spatial causal inference, Bayesian methods and environmental perceptions. 




Research Statement

I have always considered it crucial for economic theory to describe reality, and I have always looked to real data to explain phenomena.

My research interests span the field of causal inference in longitudinal, spatial, and network contexts. In particular, I am currently working on Synthetic Control Method and Matrix Completion estimators. I am interested in working in contexts of inter-unit interference, treatment diffusion, and optimal policy learning. I am also interested in working in non-traditional policy evaluation contexts, e.g.: without control units, with full interference between units, with clustering observations. I think these kinds of applications, although very challenging, are representative of the real world.

Nonetheless, I am very interested in the intersection of the econometric literature and causal inference, which I believe to be one of the major areas of development in econometrics over the next 10 years, along with applications of Bayesian statistics in econometrics, and applied machine and statistical learning. 

Applications for these methods range from fiscal, health, transportation, energy, and economic policies. I'd like to think of my work as a tool for enhancing the experiences of citizenship, helping to provide tools capable of improving policies.