Part 1 Block-recursive systems and the foundations of structural estimation: on the cost of approximate specification in simultaneous equation estimation; a correspondence principle for simultaneous equation models; causation and specification in economic theory and econometrics; on the relations between econometric sectoral and macro models; on the independent use of two or more sets of policy variables; on the analyses of history and the interdependence of the social sciences. Part 2 Some implications - essays in econometric theory: dynamic structure and estimation of economy-wide econometric models - some further thoughts; simultaneous equation estimation - the state of the art; the relative sensitivity to specification error of different k-class estimators; approximate specification and the choice of a k-class estimator; the effect of simple specification error on the coefficients of unaffected variables; tests of equality between sets of coefficients in two linear regressions - an expository note. Part 3 Empirical work: an econometric model of the world copper industry; regional specialization and the supply of wheat in the United States, 1867-1904; the costs of automobile model changes since 1949; accidents, 1962-1967.

Franklin Fisher's work on the underlying structure of econometric models has been fundamental to the development of the subject. Fisher's early discovery of block-recursive systems, together with his results on continuity for small specification errors, provided the foundation for all structural estimation. It did so by showing that such estimation is possible and useful even though models are inevitably approximate and contain exogenous variables which are themselves determined by equations in a larger stochastic system. Fisher's later work on causation dealt with the implications of regarding simultaneous equation models as limiting cases of non-simultaneous ones. The essays in this volume enlarge on those themes in various ways. They not only deal with such matters as the importance of block-recursive systems for the foundation of structural estimation but also with more practical topics. These include the implications of block-recursive systems for the relations between sectoral and macro models and for the choice of instrumental variables. The book also contains Fishers' well known exposition of Chow tests and closes with some examples of empirical work including the famous Fisher-Griliches-Kaysen study of the cost of automobile model changes.