Applied Econometrics Overview

Principles

Definition

Applied econometrics is a field within economics that utilizes statistical and mathematical methods, collectively known as econometric techniques, to analyze empirical data for the purpose of quantifying economic relationships and testing economic theories. It is primarily concerned with applying these techniques to real-world economic problems to provide empirical evidence, facilitate measurement, and inform decision-making and policy. This methodological approach involves model specification, estimation, inference, and forecasting based on observed economic data.

Ontological type

Identification Strategies

Core Methods

Model Specification

Key developments

Key Figures

Specification and Diagnostics era

A. Stewart Fotheringham[1] was affiliated with McMaster University[3] and Indiana University Bloomington[4] during the Specification and Diagnostics era. His key contributions in this era include co-developing Geographically Weighted Regression (GWR) to explore spatial nonstationarity, as detailed in the 1996 paper Geographically Weighted Regression: A Method for Exploring Spatial Nonstationarity[7], which made possible localized inference and challenged globally parameterized models, thereby improving specification diagnostics in spatial econometrics. Martin Charlton[2] was affiliated with University of Manchester[5] and Newcastle University[6] during the Specification and Diagnostics era. His key contributions include co-authoring the Geographically Weighted Regression paper[7], advancing spatial econometric specification by enabling analysis of spatially varying relationships and providing a diagnostic framework that underpins subsequent robustness checks and specification testing in this era.

Heterogeneity and Nonlinearity era

David Roodman [1], active across the University of Minnesota [2] and the University of Pittsburgh [3] during this era, contributed to applied econometrics focusing on heterogeneity and nonlinear response. Roodman [1]'s notable contribution in this era is the 2009 paper A Note on the Theme of Too Many Instruments [4], which analyzes the problems of instrument proliferation in instrumental-variable contexts and explains why careful instrument selection matters for credible causal claims under heterogeneity and nonlinearities. Roodman [1] highlighted the fragility of IV-based identification in heterogeneous and nonlinear settings, informing methodological debates and guiding subsequent econometric practices. By anchoring instrument-related concerns within a broader heterogeneity-aware framework, his 2009 note helped shape later approaches to causal inference with limited dependent variables and nonlinear responses, aligning with the era's emphasis on credible inference under unit-level heterogeneity [1].

Data-Driven Robust Causality era

Hal R. Varian [1] is a leading economist whose work during this era connected to the University of California, Berkeley [2] and the University of California System [3]. In this era, Varian [1]'s contributions are exemplified by the paper Predicting the Present with Google Trends [4], which pioneered using real-time online data to forecast current economic indicators. The paper Predicting the Present with Google Trends [4] demonstrated how online search data could augment traditional economic indicators, illustrating the era's emphasis on scalable, data-driven inference. Varian [1] helped shape the adoption of transparent, reproducible empirical methods in the data-rich econometrics era.