Matthew Harding

Department Affiliation:
Assistant Professor, Stanford Department of Economics

Centers & Institutes:
SIEPR Faculty Fellow
Founding member, Carbon Action Partnership

Current Research:
His work is concerned with the estimation of econometric models with latent variables and unobserved heterogeneity. He uses Random Matrix Theory to uncover meaningful economic structures from large panel data. This work focuses on the statistical properties of the spectral decompositions of generalized dependency matrices such as delay correlations, commutators and adjacency matrices in both linear and nonlinear factor models. His research also explores the use of nonparametric Bayesian methods and quantile regression methods in the estimation of nonlinear econometric models with unobserved heterogeneity such as random coefficients models, panel probit and heterogeneous treatment models. This research has important implications for the estimation of financial risk, estimation of social networks, the measurement of consumer preferences from scanner data, predicting the effect of economic news on trading activity, modeling unemployment durations and evaluating the determinants of R&D activity.

Professor Harding believes that solving the current energy and environmental crisis requires an in-depth understanding of both consumer preferences and political economy. His research on energy is primarily concerned with (a) the political economy of the Middle East and its complex connections to world economic factors (b) understanding the role of consumer expectations and macroeconomic factors as drivers of the oil market (c) the measurement of preferences for energy efficiency and the impact of economic policy on consumer choice.

Website: http://www.stanford.edu/~mch/

Contact:

Matthew Harding's PEEC Funded Projects

Appliance Efficiency and Long-Run Energy Demand

Matthew Harding's Publications

“Dynamic Panel Probit with Flexible Correlated Effects”, with M. Burda, Journal of Business and Economic Statistics, revise and resubmit, 2009.

“A Quantile Regression Approach for Estimating Panel Data Models Using Instrumental Variables” with C. Lamarche, Economics Letters 104(3), 2009.

“A Bayesian Mixed Logit Probit Model for Multinomial Choice”, with M. Burda and J. Hausman, Journal of Econometrics, 147(2), 232-246, 2008.

“Explaining the Single Factor Bias of Arbitrage Pricing Models in Finite Samples”, Economics Letters 99(1), 2008.

“Using a Laplace Approximation to Estimate the Random Coefficients Logit Model by Non-linear Least Squares”, with Jerry Hausman, International Economic Review 48(4), 2007.

“Flexible Parametric Estimation of the Taste Distribution in Random Coefficients Logit Models”, with J. Hausman, presented at the Econometric Society Meeting, Chicago, January 2007.

Matthew Harding's Presentations

“Identification of Communities in High Dimensional Stochastic Networks”, with K. K. Nair.

“Estimating the Number of Factors and Lags in High-Dimensional Dynamic Factor Models”, with K. K. Nair.

 “Finite Sample Bias Corrections for IV Estimation with Weak and Many Instruments”, with J. Hausman.

“ Understanding Choice Intensity: A Poisson Mixture Model with Logit-based Random Utility Selective Mixing”, with M. Burda and J. Hausman.

“Least Squares Estimation of a Panel Data Model with Multifactor Error Structure and Endogenous Covariates”, with C. Lamarche