Matthew Harding is an Economist and Data Scientist who uses Big Data to answer crucial economic policy questions in Health/Nutrition and Energy/Environment. Since joining Duke he was promoted to Assocate Professor (with tenure) of Public Policy and Economics at Duke University. He is the PI and Director of the Duke-UNC USDA Center for Behavioral Economics and Healthy Food Choice Research (BECR Center) and the Co-Director of the Duke Energy Data Analytics Lab. He is also an Associate Director at the Information Initiative at Duke (iiD) and a Faculty Fellow in the Duke Energy Initiative.

He aims to understand how individuals make consumption choices in a data rich environment, and quantify the individual and social welfare impact of their choices. Building on a rigorous foundation in econometric methods, he explores the potential of
Big Data to estimate better models and predict the choices made by individuals, while taking into account both traditional economic models and recent developments in behavioral economics. He is interested in the potential for Big Data to solve “wicked problems”, complex interdisciplinary problems in public policy by the use of three fundamental levers: prices, behavioral nudges and technology. In particular he is interested in identifying “triple-win” solutions. These are solutions to the most challenging problems of our time that benefit individual consumers, are profitable for firms, and have a large positive impact on society at large. Examples include policies that promote healthy food choices or energy efficiency.

He designs and implements
large scale field experiments in collaboration with industry leaders to understand the unintended social consequences of individual choices and the extent to which prices, nudges, and technology can be used as cost-effective means of improving individual and social welfare. His research relies on terabyte sized data sets on household food purchases, health and drug prescriptions, and energy consumption, to build a comprehensive framework for understanding economic choices and develop new strategies for achieving triple-win solutions.

As a Data Scientist he also develops cutting edge econometric methods for the practical analysis of Big Data econometric models using Bayesian and Quantile techniques, by focusing in particular on the role of unobserved heterogeneity in complex massive data. His econometric work is concerned with the estimation of large panel data models involving
latent variables and unobserved heterogeneity. His research also explores the use of nonparametric Bayesian methods to estimate choice models with random coefficients, duration models and heterogeneous treatment effects. Additionally, he uses quantile regression methods to develop new estimators for forecasting in Big Data.

His research is supported by the
U.S. Department of Agriculture, The Robert Wood Johnson Foundation, and the NRDC.

Recent Media Coverage:



BECR Header Online