arXiv subjects

Jonathan Libgober

Publications and source records attributed to Jonathan Libgober.

12 recordsLinked to original sources

arXiv2026

Organization Design for Complex Worlds

I study the role of \emph{horizontal complexity} -- defined as the variation in actions that similar tasks require -- in organization design. A continuum of workers each choose an action to adapt to a local state that follows a Gaussian process across locations. Headquarters can group workers into \emph{teams}, simplifying the attention problem it faces in coordinating across the organization, at the cost of inhibiting adaptation. Tools from spectral theory identify how horizontal complexity affects team size: roughly speaking, variation left unresolved by headquarters expands teams, while variation concentrated along dimensions that headquarters absorbs through attention shrinks them.

econ.TH↗

arXiv2025

Misspecified learning and evolutionary stability

We extend the indirect evolutionary approach to the selection of (possibly misspecified) models. Agents with different models match in pairs to play a stage game, where models define feasible beliefs about game parameters and about others' strategies. In equilibrium, each agent adopts the feasible belief that best fits their data and plays optimally given their beliefs. We define the stability of the resident model by comparing its equilibrium payoff with that of the entrant model, and provide conditions under which the correctly specified resident model can only be destabilized by misspecified entrant models that contain multiple feasible beliefs (that is, entrant models that permit inference). We also show that entrants may do well in their matches against the residents only when the entrant population is large, due to the endogeneity of misspecified beliefs. Applications include the selection of demand-elasticity misperception in Cournot duopoly and the emergence of analogy-based reasoning in centipede games.

econ.TH↗

arXiv2023

Incentivizing Forecasters to Learn: Summarized vs. Unrestricted Advice

How should forecasters be incentivized to acquire the most information when learning takes place over time? We address this question in the context of a novel dynamic mechanism design problem in which a designer incentivizes an expert to learn by conditioning rewards on an event's outcome and the expert's reports. Eliciting summarized advice at a terminal date maximizes information acquisition if an informative signal either fully reveals the outcome or has predictable content. Otherwise, richer reporting capabilities may be required. Our findings shed light on incentive design for consultation and forecasting by illustrating how learning dynamics shape the qualitative properties of effort-maximizing contracts.

econ.TH↗

arXiv2023

With a Grain of Salt: Uncertain Veracity of External News and Firm Disclosures

We examine how uncertain veracity of external news influences investor beliefs, market prices and corporate disclosures. Despite assuming independence between the news' veracity and the firm's endowment with private information, we find that favorable news is taken ``with a grain of salt'' in equilibrium -- more precisely, perceived as less likely veracious -- which reinforces investor beliefs that nondisclosing managers are hiding disadvantageous information. Hence more favorable external news could paradoxically lead to lower market valuation. That is, amid management silence, stock prices may be non-monotonic in the positivity of external news. In line with mounting empirical evidence, our analysis implies asymmetric price reactions to news and price declines following firm disclosures. We further predict that external news that is more likely veracious may increase or decrease the probability of disclosure and link these effects to empirically observable characteristics.

econ.TH↗

arXiv2022

Familiarity Facilitates Adoption: Evidence from Electric Vehicles

This paper shows that a non-price intervention which increased the prevalence of a new technology facilitated its further adoption. The BlueLA program put Electric Vehicles (EVs) for public use in many heavily trafficked areas, primarily (but not exclusively) aimed at low-to-middle income households. We show, using data on subsidies for these households and a difference-in differences strategy, that BlueLA is associated with a 33\% increase of new EV adoptions, justifying a substantial portion of public investment. While the program provides a substitute to car ownership, our findings are consistent with the hypothesis that increasing familiarity with EVs could facilitate adoption.

econ.GN↗

arXiv2022

Learning Underspecified Models

This paper examines whether one can learn to play an optimal action while only knowing part of true specification of the environment. We choose the optimal pricing problem as our laboratory, where the monopolist is endowed with an underspecified model of the market demand, but can observe market outcomes. In contrast to conventional learning models where the model specification is complete and exogenously fixed, the monopolist has to learn the specification and the parameters of the demand curve from the data. We formulate the learning dynamics as an algorithm that forecast the optimal price based on the data, following the machine learning literature (Shalev-Shwartz and Ben-David (2014)). Inspired by PAC learnability, we develop a new notion of learnability by requiring that the algorithm must produce an accurate forecast with a reasonable amount of data uniformly over the class of models consistent with the part of the true specification. In addition, we assume that the monopolist has a lexicographic preference over the payoff and the complexity cost of the algorithm, seeking an algorithm with a minimum number of parameters subject to PAC-guaranteeing the optimal solution (Rubinstein (1986)). We show that for the set of demand curves with strictly decreasing uniformly Lipschitz continuous marginal revenue curve, the optimal algorithm recursively estimates the slope and the intercept of the linear demand curve, even if the actual demand curve is not linear. The monopolist chooses a misspecified model to save computational cost, while learning the true optimal decision uniformly over the set of underspecified demand curves.

econ.TH↗

arXiv2022

Sequentially Optimal Pricing under Informational Robustness

A seller sells an object over time but is uncertain how the buyer learns their willingness-to-pay. We consider informational robustness under \textit{limited commitment}, where the seller offers a price \textit{each period} to maximize expected continuation profit against worst-case learning. Our formulation considers the worst case \textit{sequentially}. We characterize an essentially unique equilibrium under general conditions. We further show that, under mild conditions on the prior distribution, the equilibrium profit coincides exactly with the profit guaranteed by the equilibrium price path even under arbitrary (unrestricted) learning processes.

econ.TH↗

arXiv2021

Retractions: Updating from Complex Information

We modify a canonical experimental design to identify the effectiveness of retractions. Comparing beliefs after retractions to beliefs (a) without the retracted information and (b) after equivalent new information, we find that retractions result in diminished belief updating in both cases. We propose this reflects updating from retractions being more complex, and our analysis supports this: we find longer response times, lower accuracy, and higher variability. The results -- robust across diverse subject groups and design variations -- enhance our understanding of belief updating and offer insights into addressing misinformation.

econ.GN↗

arXiv2021

Identifying Wisdom (of the Crowd): A Regression Approach

Experts in a population hold (a) beliefs over a state (call these state beliefs), as well as (b) beliefs over the distribution of beliefs in the population (call these hypothetical beliefs). If these are generated via updating a common prior using a fixed information structure, then the information structure can (generically) be derived by regressing hypothetical beliefs on state beliefs, provided there are at least as many signals as states. In addition, the prior solves an eigenvector equation derived from a matrix determined by the state beliefs and the hypothetical beliefs. Thus, the ex-ante informational environment (i.e., how signals are generated) can be determined using ex-post data (i.e., the beliefs in the population). I discuss implications of this finding, as well as what is identified when there are more states than signals.

econ.TH↗

arXiv2021

Iterative Weak Learnability and Multi-Class AdaBoost

We construct an efficient recursive ensemble algorithm for the multi-class classification problem, inspired by SAMME (Zhu, Zou, Rosset, and Hastie (2009)). We strengthen the weak learnability condition in Zhu, Zou, Rosset, and Hastie (2009) by requiring that the weak learnability condition holds for any subset of labels with at least two elements. This condition is simpler to check than many proposed alternatives (e.g., Mukherjee and Schapire (2013)). As SAMME, our algorithm is reduced to the Adaptive Boosting algorithm (Schapire and Freund (2012)) if the number of labels is two, and can be motivated as a functional version of the steepest descending method to find an optimal solution. In contrast to SAMME, our algorithm's final hypothesis converges to the correct label with probability 1. For any number of labels, the probability of misclassification vanishes exponentially as the training period increases. The sum of the training error and an additional term, that depends only on the sample size, bounds the generalization error of our algorithm as the Adaptive Boosting algorithm.

stat.ML↗

arXiv2021

Machine Learning for Strategic Inference

We study interactions between strategic players and markets whose behavior is guided by an algorithm. Algorithms use data from prior interactions and a limited set of decision rules to prescribe actions. While as-if rational play need not emerge if the algorithm is constrained, it is possible to guide behavior across a rich set of possible environments using limited details. Provided a condition known as weak learnability holds, Adaptive Boosting algorithms can be specified to induce behavior that is (approximately) as-if rational. Our analysis provides a statistical perspective on the study of endogenous model misspecification.

econ.TH↗

arXiv2020

Evolutionarily Stable (Mis)specifications: Theory and Applications

Toward explaining the persistence of biased inferences, we propose a framework to evaluate competing (mis)specifications in strategic settings. Agents with heterogeneous (mis)specifications coexist and draw Bayesian inferences about their environment through repeated play. The relative stability of (mis)specifications depends on their adherents' equilibrium payoffs. A key mechanism is the learning channel: the endogeneity of perceived best replies due to inference. We characterize when a rational society is only vulnerable to invasion by some misspecification through the learning channel. The learning channel leads to new stability phenomena, and can confer an evolutionary advantage to otherwise detrimental biases in economically relevant applications.

econ.TH↗