The next CLEAN seminar will take place on Friday 27 October in distance at 4.30pm Italian time.
The presenter will be Fabio Martinenghi (UNSW Sydney). The title of the paper is: "Does refusing legal aid cause harm? A design-based double-machine-learning approach".
Legal aid enforces the right to counsel, which is fundamental to a fair trial and was recognised by the United Nations in 2012. Nonetheless, little is known about the impact it has on court outcomes. This is likely due to a mix of data access and identification issues. I ask whether legal aid denial causes harm, focusing on court outcomes. I use double machine learning on a new administrative dataset linking data from legal aid application forms to their court outcomes. Access to all factors used in the decision to provide aid allows me to learn the unknown treatment assignment function and identify ATTs and ATEs. I find that aid denial does not cause harm. Instead, because aid is denied to those who can afford a private lawyer, I find that aid denial leads to better court outcomes, namely to a 10 p.p. reduction in the probability of being incarcerated. I argue that this approach, combining high-quality administrative data with double machine learning, may allow policy-makers to track the performance of public programs that are considered hard to evaluate.
The seminar will take place in room Alberto Alesina Seminar Room 5.e4.sr04, Via Röntgen 1.