Econometrics: Specification and Evaluation

Author: (British) Granger
Publisher:
Publish Date: 2005-10-01
Features: When constructing economic models, it is essential to consider the integration of the model with economic theory. If a model is statistically significant but economically unreasonable, it has little reference value for economic analysis. The book includes a wide range of cases that are simple and engaging, making it accessible even to non-professionals who can understand the process of constructing economic models and benefit greatly from it. —Zhao Guoqing, Professor and Doctoral Supervisor at Renmin University of China
When comparing mathematical economics and econometrics, we often feel that the former resembles a science, while the latter is more like an art. Indeed, in an empirical study, everything from the selection and specification of the econometric model to estimation, evaluation, prediction, and interpretation is left to the researcher's discretion. As a result, empirical research can appear unreliable: researchers can find "suitable" data and "suitable" models to support their economic theoretical viewpoints. However, Granger rigorously and engagingly demonstrates the entire process of constructing empirical models through some interesting real-world examples in this book. It shows us the scientificity, credibility, and of empirical research. —Chen Yanbin, Economics PhD
In this book, Granger elucidates the process of constructing and evaluating empirical models. The book extensively draws on and references cases and anecdotes from economics, finance, political science, environmental economics, as well as art, literature, and the entertainment industry, blending precision with accumulated intuition to provide us with unique and highly engaging insights on this topic. The first chapter analyzes the problem of model specification, using the example of deforestation in the Amazon basin of Brazil to discuss the process of model specification. The second chapter considers the issue of evaluation, pointing out the neglect of evaluation by economists and emphasizing that models should be evaluated based on the quality of their results. The third chapter employs more complex and technical methods than the first two chapters to further discuss how to evaluate predictions. This book brings us new ideas about how to construct and evaluate models, suggesting that in future modeling, we should give more consideration to the economic significance of models rather than their statistical significance, and more to their practical value rather than their superficial elegance.