Complete and Incomplete Econometric Models

Complete and Incomplete Econometric Models

John Geweke
Copyright Date: 2010
Pages: 176
https://www.jstor.org/stable/j.ctt7t5jp
  • Cite this Item
  • Book Info
    Complete and Incomplete Econometric Models
    Book Description:

    Econometric models are widely used in the creation and evaluation of economic policy in the public and private sectors. But these models are useful only if they adequately account for the phenomena in question, and they can be quite misleading if they do not. In response, econometricians have developed tests and other checks for model adequacy. All of these methods, however, take as given the specification of the model to be tested. In this book, John Geweke addresses the critical earlier stage of model development, the point at which potential models are inherently incomplete.

    Summarizing and extending recent advances in Bayesian econometrics, Geweke shows how simple modern simulation methods can complement the creative process of model formulation. These methods, which are accessible to economics PhD students as well as to practicing applied econometricians, streamline the processes of model development and specification checking. Complete with illustrations from a wide variety of applications, this is an important contribution to econometrics that will interest economists and PhD students alike.

    eISBN: 978-1-4008-3524-9
    Subjects: Economics, Business, Finance

Table of Contents

  1. Front Matter
    (pp. i-iv)
  2. Table of Contents
    (pp. v-vi)
  3. Series Editors’ Introduction
    (pp. vii-viii)
    Philip Hans Franses and Herman K. van Dijk
  4. Preface
    (pp. ix-x)
  5. 1 Introduction
    (pp. 1-6)

    Models are the venue for expressing, comparing, and evaluating alternative ways of addressing important questions in economics. Applied econometricians are called upon to engage in these exercises using data and, often, formal methods whose properties are understood in decision-making contexts. This is true of work in other sciences as well.

    There is a large literature on alternative formal approaches to these tasks, including both Bayesian and non-Bayesian methods. Formal approaches tend to take models as given, and the more formal the approach the more likely this is to be true. Whether the topic is inference, estimation, hypothesis testing, or specification...

  6. 2 The Bayesian Paradigm
    (pp. 7-33)

    The Bayesian paradigm provides a powerful and practical structure for managing the risk inherent in decision making. This chapter discusses the elements of this structure, which is standard in the subjective Bayesian approach to inference and decision making. A number of recent texts provide more detailed consideration of this approach in econometrics, including Poirier (1995), Koop (2003), Lancaster (2004), Geweke (2005), Rossi et al. (2005), and Greenberg (2007). This chapter also reviews the Bayesian literature on model evaluation: the effort to assess whether the structure under consideration corresponds to reality. Model evaluation is an inherently difficult question from a Bayesian...

  7. 3 Prior Predictive Analysis and Model Evaluation
    (pp. 34-85)

    Prior predictive analysis, described in section 2.4.1, is a versatile tool that provides insight into the characteristics of a model and the means to evaluate a model’s adequacy for given data. This chapter illustrates prior predictive analysis and introduces two new techniques: one for studying model characteristics and the other for model evaluation. The emphasis is on the serious application of subjective Bayesian methods, and therefore all of the methods used are consistent with the likelihood principle.

    The illustration involves a well-used data set: the monthly Standard & Poor’s (S&P) 500 return series. The models considered here are comparatively simple...

  8. 4 Incomplete Structural Models
    (pp. 86-121)

    The dynamic stochastic general equilibrium (DSGE) model has become a central analytical tool in studying aspects of economic behavior in which aggregate uncertainty is important. Models in this family abstract sufficiently from measured economic behavior that clarification of the dimensions of reality they are intended to mimic is important. This clarification is essential if these models are to deepen our understanding of real economies. If the relation between DSGE models and measured economic behavior can be made formal, explicit, and simple, then the analytic power of this approach and the understanding of economic behavior will be enhanced. This chapter explores...

  9. 5 An Incomplete Model Space
    (pp. 122-160)

    The formal solutions of most decision problems in economics, in the private and public sectors as well as in academic contexts, require probability distributions for magnitudes that are as yet unknown. Point forecasts are rarely sufficient. For econometric investigators whose work may be used by clients in different situations the rationale for producing predictive distributions is clear.

    Increasing awareness of these requirements, combined with advances in modeling and computing, is leading to a sustained emphasis on these distributions in econometric research (Diebold et al. 1998; Christoffersen 1998; Corradi and Swanson 2006a,b; Gneiting et al. 2007). In many situations several models...

  10. References
    (pp. 161-165)