Charts for a technical jungle

This collection of classic papers in econometrics from the late 19th century to 1950 is an invaluable source for anyone interested in not just the history of econometrics but also the history of modern economic thought.

Recent years have seen a significant growth of interest in the history of econometrics. The interdisciplinary nature of econometrics (an amalgam of economic ideas, economic data and mathematical statistics) together with its current youthfully mature status has made its history a stimulating new territory not only to economists, historians of economics, econometricians, students of economics and econometrics, but also to historians of statistics, historians of science, as well as philosophers of science and social science.

Although a number of books and monographs on the subject have been published in the past decade, this is the first time that we have been presented with a collection of the most important econometric writings from the early days. Needless to say the book provides an easy access to these otherwise widely dispersed or deeply buried writings. But the real question is how important or necessary, except for professional historians, it is that the original should be read, especially by the largely forward-looking econometrics profession.

I would certainly recommend the book highly to applied economists and students of economics. Econometrics has grown into such a technical (mathematical) subject that it is quite difficult for students as well as applied economists to learn and understand the essence of the subject merely from econometrics textbooks. Books and papers on the history of econometrics can be thought of as a series of signposts through the technical jungle of econometrics to its origins, whereas this accessible collection represents the ground plan of the subject.

As for the professional econometricians, it may be tempting for them to assume that these early writings are too rudimentary to be interesting. Indeed, the mathematical and statistical methods employed in the current econometrics are so much more sophisticated than before that the subject has undergone a considerable division of labour. Nowadays there are labels like applied econometrician versus theoretical econometrician, microeconometrician versus macroeconometrician, time-series and panel-data specialists etc. It seems almost impossible for anyone in the profession to obtain a full knowledge of the entire enterprise. The lack of a full vision can, however, make researchers susceptible to futile or even erroneous routes. This simple truth is evident not only in the history of various sciences, but also in the history of econometrics. This book provides econometricians working on dissimilar branches a clear view of the trunk and roots of the econometric tree.

On opening the book, one cannot but notice that the contents are strikingly different from most econometrics textbooks and monographs. Most noticeably, there is a highly focused attention to applied economic issues, ie on how to retrieve economically meaningful information from data by using appropriate statistical tools on the basis of given applied issues of interest. In contrast, the part devoted to statistical tools per se is extraordinarily small. Clearly, there have been significant "structural shifts" between current and past research, and current research which operates without any reference to the past is in danger of losing information.

Another remarkable feature of the book is the unusually long and detailed editors' introduction. A mini-sketch is always convenient for fully occupied econometricians, although not comparable with the original. What is unique in this introduction is the replications of some of the results of the early works. Scanning through these replications, I cannot help wondering how much our insight into the subject matter has progressed from these ingenious results from nearly a century ago. For example, Lehfeldt's wheat price model, or Henry Schultz's sugar price and quantity model, are considerably superior to many applied examples in a number of widely used econometrics textbooks. Often such examples are infected with "spurious regression", or ambiguous underlying theory, or other problems recognised already by the pioneers. Thankfully, the editors have placed the relevant data sets in the Internet for free anonymous "file transfer protocol" download.

Duo Qin is a lecturer in economics, Queen Mary and Westfield, University of London.