Targeting with machine learning: An application to a tax rebate program in Italy

B-Tier
Journal: Journal of Economic Behavior and Organization
Year: 2018
Volume: 156
Issue: C
Pages: 86-102

Authors (5)

Andini, Monica (Banca d'Italia) Ciani, Emanuele (not in RePEc) de Blasio, Guido (not in RePEc) D'Ignazio, Alessio (not in RePEc) Salvestrini, Viola (not in RePEc)

Score contribution per author:

0.402 = (α=2.01 / 5 authors) × 1.0x B-tier

α: calibrated so average coauthorship-adjusted count equals average raw count

Abstract

This paper shows how machine learning (ML) methods can be used to improve the effectiveness of public schemes and inform policy decisions. Focusing on a massive tax rebate scheme introduced in Italy in 2014, it shows that the effectiveness of the program would have significantly increased if the beneficiaries had been selected according to a transparent and easily interpretable ML algorithm. Then, some issues in estimating and using ML for the actual implementation of public policies, such as transparency and accountability, are critically discussed.

Technical Details

RePEc Handle
repec:eee:jeborg:v:156:y:2018:i:c:p:86-102
Journal Field
Theory
Author Count
5
Added to Database
2026-01-24