Econometrics
Statistical methods for economic data, causal inference, and forecasting.
econ.EM · 119 papersCalibeating Prediction-Powered Inference
Calibeating Prediction-Powered Inference introduces a method to post-hoc calibrate prediction scores on labeled data, improving semisupervised mean estimation efficiency.
Flexible Bayesian Models for Time-Varying Income Distributions
This paper introduces flexible Bayesian models that dynamically link income distributions across years, improving stability and precision for time-varying inequality and poverty measures.
Participation and Representation in Local Government Speech
Analyzing a decade of California city council meetings, this study reveals demographic biases in public participation and the limited impact of remote access.
On-chain Peak Shaving
This paper studies "on-chain peak shaving," scheduling Ethereum transactions to off-peak hours to reduce gas fees, finding varied firm strategies and cost savings.
Recent Advances in Causal Analysis of the Stochastic Frontier Model
This paper reviews recent advances in integrating causal inference methods with stochastic frontier models to analyze productivity and efficiency.
Probabilistic Forecasting for Day-ahead Electricity Prices, Battery Trading Strategies and the Economic Evaluation of Predictive Accuracy
This paper reveals flaws in using battery trading strategies to evaluate probabilistic electricity price forecasts and proposes a new stochastic programming method.
From Clerks to Agentic-AI: How will Technology Change Labor Market in Finance?
This paper tracks how technological waves (computerization, passive investing, AI) have impacted labor productivity in financial asset management.
Clustered Local Projections for Time-Varying Models
Clustered Local Projections (LP) is a new method for estimating impulse response functions in time-varying models, using k-means for data classification.
Factor-Augmented Panel Regressions and Variance-Weighted Treatment Effects
This paper shows that factor-augmented panel regressions consistently estimate variance-weighted average treatment effects under nonparametric assumptions.
Causal inference for social network formation
This paper introduces a design-based framework for causal inference in social network formation, using repeated observations and random initial ties to identify key drivers.
Subsample-based Estimation under Dynamic Contamination
Subsample-based estimation in dynamic time series fails under contamination due to residual propagation, but a new patch removal operator restores consistency.
Bootstrap consistency for general double/debiased machine learning estimators
This paper establishes the theoretical validity of bootstrap inference for Double/Debiased Machine Learning (DML) estimators, filling a critical gap.
A Model and Estimation of the Bitcoin Transaction Fee
This paper develops and estimates a structural model of Bitcoin transaction fee choice using novel mempool data, treating it as a market for scarce blockspace.
The Virtue of Sparsity in Complexity
This paper shows that in asset pricing, complexity in feature spaces complements factor sparsity, enabling discovery of parsimonious risk structures.
Decision Traces: What Multi-System Data Fusion Reveals About Institutional Knowledge in Enterprise Hiring
This study operationalizes "decision traces" in enterprise hiring, showing how fusing siloed data reveals critical insights about candidate success.
Integrating Diagnostic Checks into Estimation
A new method integrates diagnostic checks into estimation via residualization, improving inference, reducing variance, and minimizing bias.
Path-Explosive Behaviour in Economic Time Series: A Realization-Centred Exploratory Framework
This paper introduces a realization-centred framework to detect and characterize path-explosive behavior in economic time series without distributional assumptions.
The Econometrics of Matching with Transferable Utility: A Progress Report
This paper reviews recent econometric methods for matching markets with transferable utility, finding the separable approach robust to omitted variables.
True and Pseudo-True Parameters
This paper explores pseudo-true parameters in misspecified models, finding their decision relevance fragile but deriving robust confidence intervals.
Tweedie Calculus
Tweedie Calculus introduces a general framework for deriving Tweedie representations in additive-noise models, simplifying posterior mean estimation.
📬 Weekly AI Paper Digest
Get the top 10 AI/ML arXiv papers from the week — summarized, scored, and delivered to your inbox every Monday.