Content area
Full Text
By Joshua D. Angrist and Jorn-Steffen Pischke. 2014. Princeton, NJ: Princeton University Press, Pp. 304. $79.50 hardcover, $24.50 paperback.
Mastering 'Metrics' is unique among the books on econometrics, not only because of its downplaying of statistical theory and standard econometric terminology but also by its whimsical humor. The text is interspersed with citations from Kung Fu and philosophical dialogs between "Masters" (econometricians) and "Grasshoppers" (students).
The content of the book focuses on five econometric methods called Furious Five .
These are randomized trials, regression analysis, instrumental variables, regression discontinuity design (RDD), and difference in-differences (DiD). Regression analysis is the foundation stone of classical econometrics and the concept of IV is also well-known. The randomized trials approach is rare in econometrics because of data shortage but RDD and DiD are new and widely used, but not well represented in most econometric textbooks.
The first chapter of the book deals with randomized trials (RT), a topic virtually nonexistent in standard econometric courses. However, the authors regard RT as the "pride place" in their toolbox. Randomized trials are the standard medical research tool, where the procedure involves a random selection of patients who are then randomly divided into two groups. One group receives the treatment in question and the other group is not treated. After the treatment is finished the average outcomes of the two groups are compared. The fact that the two groups are virtually identical in all respects except for the treatment leads to the elimination of the possible effect of "selection bias." The averages of all the existing but unobserved control variables, (such as age, health, socioeconomic status, and so forth) average out when calculating the differences between them. For the authors this method...