Calendrier du 23 janvier 2023
GSIELM (Graduate Students International Economics and Labor Market) Lunch Seminar
Du 23/01/2023 de 13:00 à 14:00
Maison des Sciences Economiques, Salle 116
MOREAU-KASTLER Ninon (ENS Paris-Saclay)
Due Diligence and Legal Havens: Conflict Minerals Exports in Africa’s Great Lake Region
This paper studies how « conflict minerals » trade flows react to due diligence policies imposing sourcing guidelines for importers, in the presence of legal havens. Dodd-Frank Act Conflict Mineral Rule (2010) imposed sourcing disclosure and name and shame for U.S. importers using 3T (tantalum, tin and tungsten) from conflict areas in Democratic Republic of the Congo and adjoining countries. Comparing targeted bilateral trade flows to non-targeted products exporters within the structural gravity framework, I find that this policy decreased targeted countries 3T exports value, creating a new trade barrier for targeted countries. However, 3T flows are diverted to legal havens. Legal havens are countries adopting laws allowing economic agents to « hide illicit activity and to exempt themselves from legal obligations linked to their economic activities ». After the Dodd-Frank Act, D.R.C. and neighboring countries export share of 3T to legal havens increases. 3T trade flows also increase from legal havens to the United States after 2010, which could point to regulation avoidance.
Texte intégral
Job Market Seminar
Du 23/01/2023 de 12:00 à 13:15
R2-21
LIEBER Jonas (University of Chicago)
Estimating Concentration Parameters for Bandit Algorithms
Bandit models are widely used to capture learning in contexts where agents repeatedly choose
actions with uncertain rewards. Examples include firms maximizing profits by experimenting
with prices or advertisement, randomized control trials maximizing outcomes by evaluating
alternative treatments, and consumers maximizing utility by trying experience goods. A popular bandit algorithm is the upper confidence bound (UCB) algorithm. The UCB algorithm
requires sub-Gaussian concentration parameters as inputs. In practice, these parameters are
unknown so that the UCB algorithm is not fully data-driven. I propose a method to estimate
these parameters and use the estimated parameters to conduct inference with Hoeffding’s
inequality. I show that asymptotic inference with estimated parameters is valid under mild
and optimal under stronger conditions. In finite samples, I establish validity of inference under an anti-concentration condition. Equipped with the proposed estimator for sub-Gaussian
concentration parameters, I adapt the UCB algorithm to settings where these parameters
are unknown. In an empirical application, I study price experimentation after the liberalization of the spirits market in Washington State in 2012 and find that the adapted UCB
algorithm leads to considerably higher profits. My theoretical results can also be applied to
non-standard inference problems that arise in partial identification and machine learning.