Deparment of Economics
Fordham University
E-mail: pschenone@gmail.com
I am an Associate Professor of Economics at Fordham University. My research is in microeconomic theory, in particular, decision theory, game theory, and networks. Previously, I was an Assistant Professor of Economics at the W.P. Carey School of Business at Arizona State University.
Final topologies for preference spaces (Accepted at The B.E. Journal of Theoretical Economics)
Consistent conjectures in dynamic matching markets, with Laura Doval (Mathematical Social Sciences, Volume 132, December 2024)
Networks, frictions and price dispersion, with Javier Donna and Gregory Veramendi (Games and Economic Behavior, Volume 124, November 2020, pages 406-4)
Revealed Preference Implications of Backward Induction and Subgame Perfection (American Economic Journal: Microeconomics, Volume 12, May 2020, pages 230-56)
Frictions in Internet Auctions with Many Traders: a Counterexample, with Javier Donna and Gregory Veramendi (Economic Letters, Volume 138, January 2016, pages 81-84)
Identifying subjective beliefs in subjective state space models (Games and Economic Behavior, Vol. 95, January 2016, Pages 59–72)
We propose a decision-theoretic model akin to Savage (1972) that is useful for defining causal effects. Within this framework, we define what it means for a decision maker (DM) to act as if the relation between the two variables is causal. Next, we provide axioms on preferences and show that these axioms are equivalent to the existence of a (unique) Directed Acyclic Graph (DAG) that represents the DM's preference. The notion of representation has two components: the graph factorizes the conditional independence properties of the DM's subjective beliefs, and arrows point from cause to effect. Finally, we explore the connection between our representation and models used in the statistical causality literature (for example, Pearl (1995)).
Instructors typically map noisy exam scores on [0,100] into coarse letter grades A, ..., E. This coarsening acts as a noise filter for transcript readers, but it inevitably risks misclassification. We treat grade assignment as a statistical classification problem: mapping observed scores to latent student types. We introduce a novel grading system, denoted \emph{Bayesian Adaptive Grading} (BAG). BAG holds priors over student type distributions and exam difficulty, and propagates parameter uncertainty into final grade assignments via Bayes-optimal decisions under standard loss functions.
We derive optimal cutoffs for Fixed-Scale Grading (FSG), a commonly used benchmark in which instructors pre-commit to numerical cutoffs, and we compare BAG against both FSG and curve-based grading in Monte Carlo simulations. Across a wide range of misspecification regimes, BAG substantially outperforms both alternatives: over a variety of robustness checks, BAG reduces mean misclassification by roughly 55 - 72\% relative to FSG and 45 - 68\% relative to curving, delivers median misclassification between 0\% and about 3\%, and achieves median error reductions of roughly 66 - 100\% relative to both benchmarks. Curving performs particularly poorly because it fails to model the mixture structure of student types.
Finally, we show how BAG performs on real classroom data, and how informative priors on grade distributions can be used to encode institutional targets while allowing data-justified deviations. BAG thus improves classification accuracy, mitigates grade inflation, and protects students against instructor mistakes in a way that is transparent and statistically rigorous.
This paper studies repeated games of incomplete information where each player knows his own payoffs and where the unknown state of the world can be identified by the combined private information of all players. We obtain a condition that is both necessary and sufficient for a Perfect Bayesian Equilibrium (PBE) folk theorem to hold. This contrasts with the existing literature where, due to the difficulty in keeping track of beliefs as play evolves, the analysis has focused on either Nash equilibrium for one-sided incomplete information or has dealt with various ex-post solution concepts. Finally, we also show the condition obtained is also necessary and sufficient to obtain Ex-Post folk theorems.
Maybe that's the real punk rock