×

The practice of econometric theory. An examination of the characteristics of econometric computation. (English) Zbl 1176.62118

Advanced Studies in Theoretical and Applied Econometrics 44. Berlin: Springer (ISBN 978-3-540-75570-8/hbk; 978-3-642-24251-9/pbk; 978-3-540-75571-5/ebook). xvi, 311 p. (2009).
The author starts with the statement that econometric theory is a somewhat disparate collection of findings, rather than a well integrated and organized whole. And furthermore, its nature is more likely to be a set of demonstrated results but rather becoming collectively complete. The consequences thereof are that econometrics inevitably remains a fragmentary body of knowledge on the one hand, and on the other, the practice consists of selecting from and applying the literature, as well as simultaneously evaluating it, so as to test its applicability and range, and support its further advance (preface, p. vii). And consequently, this fragmentary development carries over to the development of econometric software.
Despite of the rapidly growing number of estimators, the author concentrates on the OLS estimator as the most important one in applied econometrics and the supplementary evaluative tests applied in conjunction with OLS. According to this choice, the author presents, classifies, and documents the particular diagnostic tests that are provided in existing econometric software packages. The ultimate goal consists of supplying reliable and useful information to econometric software developers, econometric theorists, and to economists who use this software.
The first two chapters present an introduction to econometric computation and the characteristics of econometric software. The central issue of the book starts with chapter three, dealing with econometric diagnostic tests, followed by the basic statistics in the fourth chapter. A more detailed analysis is given in chapter five which deals with the failure of assumptions such as the presence of heteroscedasticity, properties of disturbances, specification tests related to the functional form, nonlinearities, and simultaneity. Further diagnostic tests are investigated; among those are tests concerned with structural stability of regression relationships, the case of omitted variables as well as linear restrictions on the coefficients.
More recent econometric developments such as cointegration, unit root tests, and nonnested specification tests and their performance in different econometric software packages are discussed in chapter six. The next chapter is devoted to historical considerations and is related to topics such as changes in regression displays and data management issues. The implications of the findings are presented in the final chapter.
The econometric software packages considered by the author amount to a total of 23 packages, including, among others, EViews, GaussX, LIMDEP, PcGive, RATS, SHAZAM, Stata, and TSP. Furthermore, the dataset used throughout the book is given in the appendix (U.S. quarterly macroeconomic data used by Cecchetti and Rich as well as by Greene, as well as the famous Grunfeld investment data).

MSC:

62P20 Applications of statistics to economics
65C60 Computational problems in statistics (MSC2010)
62-04 Software, source code, etc. for problems pertaining to statistics
62-02 Research exposition (monographs, survey articles) pertaining to statistics
62J20 Diagnostics, and linear inference and regression
PDFBibTeX XMLCite
Full Text: DOI