4 WHAT CONSTITUTES A TEST OF AN ECONOMIC THEORV

. 7

Economic theory

Econometric model

Estimation

Specification testing and diagnostic checking

Is the model adequate ]

J L .

Tests of any hypotheses

Using the model for predictions and policy

Figure 1.2. Revised schematic description of the steps involved in an econometric analysis of economic models.

report that the signs of the estimated coefficients in an econometric model are correct. This approach may be termed the approach of confirming economic theories. The problem with this approach is that, as Mark Blaug points out:

In many areas of economics, different econometric studies reach conflicting conclusions and, given the available data, there are frequently no effective methods for deciding which conclusion is correct. In consequence, contradictory hypotheses continue to co-exist sometimes for decades or more.

A more valid test of an economic theory is if it can give predictions that are better than those of alternative theories suggested earlier. Thus one needs to compare a given model with earlier models. This approach of evaluating

M Blaug, Metiiodology of Economics (Cambridge: Cambridge University Press, 1980), P 261.

Summary and an Outline of the Book

The preceding discussion suggests some changes that have taken place during the last decade in the development of econometric methods. The eariier .emphasis (during the 1950s and 1960s) was on efficient estimation of a given model. The emphasis has now shifted to specification testing, diagnostic checking, model comparison, and an adequate formulation of the expectational variables given the pervasive role of expectations in almost all economic theories.

A large part of the book, however, will be devoted to developments that took place in the 1950s and 1960s because these developments form the basis for recent work. However, Chapter 12 is devoted to recent developments in the areas of specification testing, diagnostic checking, and model selection. In addition, the other chapters also describe recent developments. For instance, rational expectations models are discussed in Chapter 10, recent developments on tests for serial correlation are reviewed in Chapter 6, and vector autoregres-sions, unit roots, and cointegration are discussed in Chapter 14.

The plan of the book is as follows;

Chapter 2 contains a review of some basic results in statistics. Most students will have covered this material in introductory courses in statistics.

Chapter 3 begins with the simple (two-variable) regression model. Some of the more complicated sections, such as those on inverse prediction, stochastic regressors, and regression fallacy, are denoted as optional and are marked with an asterisk. Beginning students may want to skip these.

In Chapter 4 we discuss the multiple regression model in great detail. It is the longest chapter in the book because it covers several issues concerning the interpretation of multiple regression coefficients. A detailed discussion of the model with two explanatory variables is given to fix the ideas.

Chapters 5 and 6 deal, respectively, with the problems of unequal error variances and serially correlated errors in the multiple regression model. Chapter 6, in particular, deals extensively with the usual approach of using the Durbin-Watson test and discusses several situations where its use would be undesirable.

Chapter 7 is devoted to the multicollinearity problem. A separate chapter on this problem is perhaps not necessary, but it was found that inclusion of this material in any other chapter would have made that chapter far too long.

Chapter 8 deals with dummy variables that are used as explanatory variables in several contexts. We also discuss special problems created when the dependent variable is a dummy variable or a truncated variable. There is a discussion of logit, probit and tobit models.

alternative theories has received greater attention in recent years. The problems associated with model comparison and model selection are dealt with in Chapter 12.

SUMMARY AND AN OUTLINE OF THE BOOK 9

In Chapter 9 we give an elementary introduction to simultaneous-equations models and to the identification problem in econometrics.

In Chapter 10 we consider distributed lag models and models of expectations. We also give an elementary introduction to rational expectations models. Models with lags in behavior and expectations play an important role in most econometric models.

In Chapter 11 we consider errors in variables and proxy variables.

In Chapter 12 we present an introduction to some recent work on diagnostic checking, specification testing, and model selection.

In Chapter 13 we present an introduction to time-series analysis.

In Chapter 14 we discuss recent developments in time-series analysis.

Throughout, an attempt has been made to explain complicated material in simple terms. To avoid cumbersome algebra, many results have been stated without proofs although at times algebraic detail is unavoidable. Given readily available computer packages, even this algebraic detail may be unnecessary, but it is good for students to understand at least some minimal steps involved in the construction of computer packages. This will help in understanding the deficiencies and limitations of the various procedures.