Nonrecursive Models : Endogeneity, Reciprocal Relationships, and Feedback Loops

The title will fill the need for a textbook that is directly relevant to researchers wishing to estimate nonrecursive models. Presently, there is no single monograph that provides such explicit guidance to researchers on the estimation and assessment of nonrecursive simultaneous equation models in a clear, condensed and precise form.The first contribution of this title is to provide the applied researcher a reference that guides them through the specification and identification of simultaneous equation models, how to assess the quality of the estimates, and how to correctly interpret results. Given the ease of statistical estimation using newer software programs, we argue that a need for a clear and rigorous discussion of nonrecursive models, indeed, almost a 'œback to basics approach'� has developed. That is, social scientists can now estimate nonrecursive models without fully understanding their unique attributes. Many software programs are easy to use but do not provide equation by equation assessment of the quality of the model proposed for the data.The second contribution is to fill the need for greater attention to the 'œbasics'� of such models by focusing on the assessment of the equations in a nonrecursive model. There is a wealth of information contained in the reduced form equations of such models, and that this information is often opaque to researchers adopting a full information estimation approach.A third contribution of the title is its attention to the relatively new developments in estimating nonrecursive models when one or more outcome measures are categorical or when one has longitudinal data.A final contribution is that by blending the approaches from econometrics and SEM we emphasize the generality of issues surrounding nonrecursive models to a range of audiences.