Accessibility and Spatial Interaction, edited by AnaCondeço‐Melhorado, AuraReggiani, and JavierGutiérrez. 2014. NECTAR Series on Transportation and Communications Networks Research. Cheltenham, U.K. and Northampton, Massachusetts: Edward Elgar. 256. ISBN: 9781782540724. $120.00.
本书汇集欧洲学者研究，探讨如何通过空间互动模型改进可达性测量，并分析可达性对经济活动和人口分布的影响，适合区域科学、空间经济学和交通工程领域的研究者与研究生。
Accessibility and spatial interaction are two core concepts in multiple social science fields, including regional science, spatial economics, transportation, geography, and city and regional planning. Although it is well accepted that these two concepts are interconnected, open questions persist on how to use their connections to provide an accurate measurement of one family of indicators (e.g., accessibility) through the other one (e.g., spatial interaction) and to better understand the economic and social benefits of accessibility improvement. The book edited by Condeço-Melhorado, Reggiani, and Gutiérrez attempts to concentrate on these questions by assembling chapters written by leading researchers mainly from Europe. This book includes three sections. The first section focuses on advanced data and analytical methodologies to enhance the measurement of accessibility and spatial interaction. Östh, Reggiani, and Galiazzo (chapter 2) articulate the importance of distance-decay functions in computing potential accessibility and spatial interaction measured by the magnitude of flows between locations. When accessibility is not measurable, it can be estimated by spatial interaction models using flow data. They further examine which specification of the distance-decay function is least sensitive to the modifiable areal unit problem (MAUP) using Swedish data. These aggregate, social-physics-oriented estimations of accessibility are straightforward but somewhat arbitrary. McArthur, Thorsen, and Ubøe's research (chapter 3) is economics-oriented, developing a spatial general equilibrium model (SGEM) to endogenize the relationship between transportation network, accessibility, and relocations of residents and jobs. This study is particularly useful to understand how a change in transportation network (e.g., removing topographical barriers) affects accessibility, how a change in accessibility leads to relocations of population and employment, and, in turn, how these relocations alter accessibility. Other advantages of SGEM studies are that they can track the change of accessibility and calculate an economically efficient or socially optimal level of accessibility. Nicolai and Nagel (chapter 4) present a logsum indicator to compare the spatial distribution of accessibility under different spatial resolutions of locations on the flow origin side in Zurich. It is somewhat disappointing that the authors coarsely treat the spatial resolution of opportunities and overlook an examination of whether the resolutions of opportunities affects the accessibility computation. Given the location of an origin, the measure of accessibility could largely depend on the resolutions of travel opportunities, rather than the origin. Schintler, Kulkarni, Haynes, and Stough (chapter 5) offer an insightful introduction to how to use geosocial data to calculate accessibility and develop several social network analysis (SNA) metrics (i.e., degree, betweenness centrality, and entropy) to identify the positions (or importance) of locations within the U.S. network of counties. They also apply a community detection algorithm to find the “mobility sheds” (p. 107) and identify territories with close spatial interaction. While the SNA approach generates an innovative understanding of accessibility and spatial interaction, the authors mistakenly regard betweenness centrality as accessibility (p. 97). There are particular indicators in SNA that measure accessibility or reachability, e.g., closeness centrality. Also, we would expect a more thorough comparison of the accessibility computation based on the SNA and traditional approach (e.g., potential accessibility indicators). The chapters in the second section, “The Social and Spatial Dimension of Accessibility,” investigate the spatial and social factors of the accessibility of locations or population groups. De Montis, Caschili, and Trogu (chapter 6) offer a spatial autocorrelation analysis to explore the influence of the accessibility attributes of adjacent counties and regional demographics on the accessibility of U.S. counties. This spatial analysis demonstrates the importance of spatial dependence in accessibility measures but cannot clarify the causality between accessibility and demographics (e.g., whether high population densities attract more infrastructure investment, leading to a high level of accessibility, or high accessibility facilitates population clustering). Salas-Olmedo, Condeço-Melhorado, and Gutiérrez (chapter 7) measure the border effects on intranational and international interaction through manufacturing trade among European nations. They also evaluate how these border effects vary with the distance metrics employed and affect the computation of accessibility (or market potential). Ribeiro, Remoaldo, and Gutiérrez (chapter 8) connect accessibility indexes with transport-related social exclusion and look at the roles of street slope and walking speed in the accessibility computation of elderly people to public transit in the municipality of Braga, Portugal. By its title, the third section of the book appears set to proclaim “accessibility as a driver of spatial interaction,” while all three chapters tackle the influence of accessibility on economic activities on the production side only. Arbués, Mayor, and Baños (chapter 9) examine the output effects of the accessibility of road transportation infrastructure, simply measured by the distance-decay sum of the population in all other provinces in Spain. Holl (chapter 10) uses a similar measure of accessibility at the municipality level in Spain also and tests its impact on firm-level total factor productivity. Both chapters rely on panel data models and control for the effects of production externalities, measured by spillovers to neighboring regions (chapter 9) or agglomeration benefits from the denser population (chapter 10). They are solid empirical studies. However, both fail to discuss the endogeneity between accessibility and production externalities: that is, how accessibility could affect spillovers and agglomeration and indirectly affect productivity. As later noted by Gråsjö and Karlsson in the last chapter, the measure of accessibility “incorporates ‘global’ spillovers and does not only account for the impact from neighbors or locations within a certain distance band” (p. 229). The latter is often done by spatial dependence models (e.g., the spatial Durbin model in chapter 9). Thus, accessibility could play a more important role in spatial economics than spatial dependence. By reviewing various empirical studies in Sweden, Gråsjö and Karlsson further encourage a more extensive application of the accessibility approach in spatial economics studies, including the analysis of knowledge spillovers, spatial variations of productivity and economic growth, new firm formation and locational dynamics, exports, and labor mobility. This is a valuable reminder at the end of the book that there is a complex interplay between accessibility and spatial economics and that the accessibility approach remains overlooked or undervalued in related empirical studies. Although the editors believe that what sets this volume apart from other contributions to the scholarship of accessibility is its focus on “the link between spatial interaction and accessibility” (p. 3), only four chapters directly address this important link in theory or empirics (chapters 1–3 and 11). These chapters represent the state of the art but meanwhile highlight the complexity of situating accessibility within the framework of spatial interaction theory. Much more work is needed to investigate how accessibility affects the spatial distribution of opportunities and spatial interaction and, in turn, how they change accessibility. For empirical studies, a reduced-form estimation probably cannot accurately reflect the connection between spatial interaction and accessibility and structural models are needed. Many chapters in this book are useful contributions to help develop accurate accessibility computations, for example, through the articulation of the influence of the functional specification of distance decay, spatial resolutions, topographical and border constraints, and spatial dependence. Is it rather disappointing, however, that the measures of accessibility devised in the spatial economics contributions presented in Section III are oversimplified. The coarseness of these measures of accessibility could lower the credibility of the estimated relationship between accessibility and space economy. Also, this volume would have benefitted from a review (or synthesis) of the state of the art in accessibility measurement across different empirical applications either at the beginning or at the end of the volume. Overall, the book is a welcome addition to the study of accessibility and spatial interaction. It will appeal to regional scientists, spatial economists, and transportation engineers who are interested in the measurement and properties of spatial accessibility and the complex connection between accessibility and spatial economics. The book would be a useful text to support graduate courses in spatial and transport analysis.