Econometrics and the Computer: Love or a Marriage of Convenience?

This essay considers some of the circumstances that have characterized econometric practice during the past fifty to sixty years, including both the invention and use of the electronic computer and the growing accessibility of both micro- and macroeconomic data. However, its particular focus is econometric computation and the degree to which computational algorithms are not determined by econometric theory, thus making it important for historians of econometricians to begin to consider the specific way in which econometrics is practiced as well as how the theory has evolved over the years. At issue are both the way in which particular calculations should be made and the choice of misspecification and other tests that are made available in the existing econometric software packages.