Job Market Paper
Double Machine Learning for Static Panel Models with Instrumental variables: Method and Applications (with A. Baiardi, P. S. Clarke and A. Naghi)
Short abstract. This paper proposes a Double Machine Learning (DML) estimator for static panel data models with one endogenous treatment variable and unobserved heterogeneity. We derive a panel IV DML estimator by combining first-difference transformations with Neyman-orthogonal score functions that account for both unob served heterogeneity and treatment endogeneity. The method improves estimation accuracy and delivers reliable inference, even under weak identification. Empirical applications on the effects of immigration on natives’ voting behaviour, together with Monte Carlo simulations, demonstrate improved performance of the proposed estimator relative to conventional estimators, such as two-stage least squares, due to previously undetected nonlinearities
Publications
2. Clarke, P. S. and Polselli, A. (2025). Double Machine Learning for Static Panel Models with Fixed Effects. Econometrics Journal. DOI: 10.1093/ectj/utaf011.
[EctJ article] [xtdml (R package)] [MiSoC explainer]
1. Leoncini, R., Macaluso, M., and Polselli, A. (2024). Gender segregation: analysis across sectoral dominance in the UK labour market. Empirical Economics. https://doi.org/10.1007/s00181-024-02611-1 [EE article]
Policy and Technical Reports
Buono, I. and Polselli, A. (2022). An International Map of Gender Gaps (with Ines Buono), Questioni di Economia e Finanza (Occasional Paper), no. 714, Bank of Italy. Available at SSRN: https://ssrn.com/abstract=4462822 (Technical report for Bank of Italy) [article] [media coverage: SUERF; Il Sole 24 Ore, p.42 under Governance]
Working Papers
xtdml: Double Machine Learning Estimation to Static Panel Data Models with Fixed Effects in R, (2025). https://arxiv.org/abs/2512.15965. (under review)
Influence Analysis with Panel Data. [arXiv] [explainer] [slides] [ado files]
Robust Inference in Panel Data Models: Some Effects of Heteroskedasticity and Leveraged Data in Small Samples. [arXiv]
Work in Progress
Double Machine Learning for Static Panel Data Models with Interactive Fixed Effects (with B. Chen, and P. S. Clarke)
Double Machine Learning for Panel Data with Heterogeneous Effects (British Academy Fellowhip)
Double Machine Learning for Dynamic Panel Models (with P. Cizek, and P. S. Clarke)
Mental Health
Brandizzi, M., Polselli, A., et. al. (2022). Psychiatric emergencies during, after, and before the COVID-19 health crisis; what happened to our patients? A naturalistic observational study. Annals of General Psychiatry, 21.1: 29. (Statistical consulting and data analysis) [article]