Skip to main content
main-content

Über dieses Buch

Spatial econometrics deals with spatial dependence and spatial heterogeneity, critical aspects of the data used by regional scientists. These characteristics may cause standard econometric techniques to become inappropriate. In this book, I combine several recent research results to construct a comprehensive approach to the incorporation of spatial effects in econometrics. My primary focus is to demonstrate how these spatial effects can be considered as special cases of general frameworks in standard econometrics, and to outline how they necessitate a separate set of methods and techniques, encompassed within the field of spatial econometrics. My viewpoint differs from that taken in the discussion of spatial autocorrelation in spatial statistics - e.g., most recently by Cliff and Ord (1981) and Upton and Fingleton (1985) - in that I am mostly concerned with the relevance of spatial effects on model specification, estimation and other inference, in what I caIl a model-driven approach, as opposed to a data-driven approach in spatial statistics. I attempt to combine a rigorous econometric perspective with a comprehensive treatment of methodological issues in spatial analysis.

Inhaltsverzeichnis

Frontmatter

Introduction

Chapter 1. Introduction

Abstract
The importance of space as the fundamental concept underlying the essence of regional science is unquestioned. Since the early growth of the field in the late 1950s, a large number of spatial theories and operational models have been developed which have gradually disseminated into the practice of urban and regional policy and analysis. However, this theoretical contribution has not been matched by a similar advance in the methodology for the econometric analysis of data observed in space, i.e., for cross—sections of regions at one or more points in time.
Luc Anselin

Foundations for the Econometric Analysis of Spatial Processes

Frontmatter

Chapter 2. The Scope of Spatial Econometrics

Abstract
In this first chapter in Part I, I outline the scope of the book in more detail, and describe the meaning of some important terms in an informal manner.
Luc Anselin

Chapter 3. The Formal Expression of Spatial Effects

Abstract
To facilitate the reading of the later chapters, I present an initial discussion, in fairly general terms, of the notion of connectedness in space, and the main operational tools by which spatial effects are encompassed in econometric work: the spatial weight matrices and spatial lag operators. I also briefly discuss some complicating factors related to the notion of space implied by the various techniques.
Luc Anselin

Chapter 4. A Typology of Spatial Econometric Models

Abstract
A large number of model specifications for spatial processes have been suggested in the literature and empirically implemented. This great variety may seem unwieldy, and give the impression that every particular model necessitates its own methodological framework. Fortunately, some structure can be imposed, guided by the principle that econometric techniques can be applied in essentially the same manner to models grouped in terms of salient characteristics.
Luc Anselin

Chapter 5. Spatial Stochastic Processes: Terminology and General Properties

Abstract
In this chapter I discuss in general terms the formal mathematical and statistical background for the estimation and testing of spatial econometric models. Even though this material is highly abstract, I will keep the treatment rather informal. An in-depth discussion of the various definitions, theorems and proofs would clearly be beyond the scope of this book. Since the relevant theoretical literature is extensive, a main motivation for this chapter is to bring together some important results that have been presented in a variety of sources.
Luc Anselin

Estimation and Hypothesis Testing

Frontmatter

Chapter 6. The Maximum Likelihood Approach to Spatial Process Models

Abstract
In this chapter, I consider the application of the maximum likelihood principle to estimation and hypothesis testing for spatial process models. The models follow the taxonomy for cross-sectional situations presented in Chapter 4. Space-time formulations will be discussed in Chapter 10.
Luc Anselin

Chapter 7. Alternative Approaches to Inference in Spatial Process Models

Abstract
The maximum likelihood approach to estimation and hypothesis testing in spatial process models is by far the better known methodological framework. Moreover, in most of the literature in spatial econometrics it is the only technique considered and implemented. Nevertheless, several alternatives can be suggested to avoid some of the problems associated with ML estimation. Specifically, the numerical complexities of the nonlinear optimization and the restrictive parametric framework are features that the techniques discussed in this chapter attempt to deal with in a more satisfactory manner.
Luc Anselin

Chapter 8. Spatial Dependence in Regression Error Terms

Abstract
The analysis of the effects of spatial dependence in the error terms of the linear regression model was the first specifically spatial econometric issue to be addressed in the regional science literature. Initial problem descriptions and the suggestion of some solutions were formulated in the early 1970’s, e.g., by Fisher (1971), Berry (1971), Cliff and Ord (1972), McCamley (1973), Hordijk (1974), Martin (1974), Bodson and Peeters (1975), and Hordijk and Paelinck (1976). This was followed by many further assessments of the properties of various estimators and test statistics, which continue to be formulated to date. Spatial error autocorrelation is also the only spatial aspect of inference that has been recognized in the standard econometric literature, though only very recently and to a limited extent, e.g., in Johnston (1984) and King (1981, 1987).
Luc Anselin

Chapter 9. Spatial Heterogeneity

Abstract
Many phenomena studied in regional science lead to structural instability over space, in the form of different response functions or systematically varying parameters. In addition, the measurement errors that result from the use of ad hoc spatial units of observation are likely to be non—homogeneous and can be expected to vary with location, area or other characteristics of the spatial units.
Luc Anselin

Chapter 10. Models in Space and Time

Abstract
Up to this point in the book, the empirical context for the various estimators and tests has been limited to a purely cross—sectional situation. In this chapter, I consider models for which observations are available in two dimensions. Typically, one dimension pertains to space and the other to time, although other combinations, such as cross-sections of cross-sections and time series of time series can be encompassed as well. This situation has become increasingly relevant in a wide range of empirical contexts. It is referred to in the literature as panel data, longitudinal data, or pooled cross-section and time series data.
Luc Anselin

Chapter 11. Problem Areas in Estimation and Testing for Spatial Process Models

Abstract
In this chapter I will briefly review some fundamental methodological problems associated with estimation and testing in spatial process models. The attention paid to these issues in the spatial literature varies. Some problems, such as those resulting from pre-testing, have essentially been ignored. Others, such as the boundary value issue, have resulted in a considerable body of literature, but have not yet been resolved in a satisfactory manner. In general, the issues discussed in this chapter can be considered to form important directions for future research, crucial to an effective operational implementation of spatial econometric techniques.
Luc Anselin

Chapter 12. Operational Issues and Empirical Applications

Abstract
To round off the discussion in Part II, in this chapter several of the estimation methods and tests for spatial process models that were previously developed in formal terms will be illustrated empirically. Three aspects are considered in particular. In the first section, I focus on some operational issues related to the implementation of maximum likelihood estimation and the associated nonlinear optimization problem. In the second section, the analysis of cross-sectional data is considered. Using a simple model of determinants of crime for 49 contiguous neighborhoods in Columbus, Ohio, several estimation methods and tests from Chapters 6, 8 and 9 are implemented empirically. In the third section, attention shifts to space-time data sets. A Phillips curve model is estimated at two points in time for 25 contiguous counties in South-Western Ohio, to illustrate the instrumental variable methods from Chapter 7 and various diagnostics for spatial effects from Chapter 10.
Luc Anselin

Model Validation

Frontmatter

Chapter 13. Model Validation and Specification Tests in Spatial Econometric Models

Abstract
In philosophical discussions about the nature of progress in science it is often maintained that the development of theory cannot proceed without some form of reality—based validation. This leads to a continual process of interaction between theory formation and empirical assessment, motivated in part by the observation of phenomena which are not explained by existing theories, and by the failure of theoretical constructs to be reflected in experimental situations. Essential to this process is a set of clear standards or criteria of validity, and a uniform methodology to apply these to the theoretical constructs under scrutiny. Typically, this is based on a formal probabilistic framework.
Luc Anselin

Chapter 14. Model Selection in Spatial Econometric Models

Abstract
In this chapter, the statistical validation of spatial econometric models is approached from a different perspective, based on model selection or model discrimination techniques. In this context, the issue of selecting one model from among a number of alternative candidates is considered as a decision problem. More specifically, attention is focused on the nature of the trade-off between the number of parameters included in the model and the model fit, in the sense of how well the estimated model predicts in-sample observations.
Luc Anselin

Chapter 15. Conclusions

Abstract
For a book such as this one, it is hard to formulate a single conclusion that accurately captures the range of materials covered and concisely reflects the main results. Instead, it may be more useful to point out the distinguishing characteristics of this collection of methods and models, and to summarize the most relevant contributions.
Luc Anselin

Backmatter

Weitere Informationen