Abstract
Gravity type spatial interaction models have been popular among spatial scientists for many decades. Their more traditional specifications suffer from specification error attributable to, among other factors, spatial autocorrelation. This spatial autocorrelation has two components, a more conventional one relating to the geographic distributions of origin and destination phenomena, and a second one relating to the network of flows. Recent literature, to which this chapter contributes, reveals that spatial interaction model descriptions substantially improve with an accounting of network autocorrelation in Poisson regression estimation of parameters. Implementation of this specification requires eigenvector spatial filter construction, a principal topic of this chapter, which also furnishes parameter estimate comparisons for infill and increasing domain spatial sampling designs.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
These commuting data were compiled based on administrative sources. Commuters are salaried employees (staff), whose place of employment is outside the administrative borders of their municipality of residence. These data include only flows of more than nine people. See http://stat.gov.pl/spisy-powszechne/nsp-2011/nsp-2011-wyniki/dojazdy-do-pracy-w-polsce-wyniki-nsp-2011,9,1.html.
- 2.
Spatial scientists rarely encounter negative SA. For these origin weights, MC = −0.22 and GR = 1.27.
- 3.
For these destination weights, MC = 0.12 and GR = 0.93.
- 4.
For the Polish province surface partitioning, the maximum positive SA possible is MC = 0.76 and GR = 0.27.
- 5.
Because the Polish census defines a commuter to be a salaried employee (staff member) whose place of employment is outside the administrative borders of his/her municipality of residence, three counties, which actually are cities—Skierniewice, Łódź, and Piotrków Trybunalski—have artificial zero flows. To preserve comparability, these three flows were differentiated from the remaining flows by an indicator variable.
- 6.
As with the employment data (see Chap. 1), the journey-to-work data for Wałbrzych are zeroes and are wrong. Because a time series of journey-to-work data does not exist, these flows data were not corrected with imputations . Rather, this footnote acknowledges this data error; the flows data remained linked to Wałbrzyski for this spatial interaction analysis.
References
Curry, L. (1972). Spatial analysis of gravity flows. Regional Studies, 6, 131–147.
Griffith, D. (2007). Spatial structure and spatial interaction: 25 years later. The Review of Regional Studies, 37(1), 28–38.
Griffith, D. (2009a). Modeling spatial autocorrelation in spatial interaction data: Empirical evidence from 2002 Germany journey-to-work flows. Journal of Geographical Systems, 11, 117–140.
Griffith, D. (2009b). Spatial autocorrelation in spatial interaction: Complexity-to-simplicity in journey-to-work flows. In P. Nijkamp & A. Reggiani (Eds.), Complexity and spatial networks: In search of simplicity (pp. 221–237). Berlin: Springer.
Griffith, D. (2011). Visualizing analytical spatial autocorrelation components latent in spatial interaction data: An eigenvector spatial filter approach. Computers, Environment and Urban Systems, 35, 140–149.
Griffith, D., & Chun, Y. (2015). Spatial autocorrelation in spatial interactions models: Geographic scale and resolution implications for network resilience and vulnerability. Networks and Spatial Economics, 15, 337–365.
Griffith, D., & Fischer, M. (2013). Constrained variants of the gravity model and spatial dependence: Model specification and estimation issues. Journal of Geographical Systems, 15, 291–317.
Griffith, D., & Jones, K. (1980). Explorations into the relationship between spatial structure and spatial interaction. Environment and Planning A, 12, 187–201.
LeSage, J., & Pace, R. (2008). Spatial econometric modelling of origin-destination flows. Journal of Regional Science, 48, 941–967.
Tiefelsdorf, M., & Boots, B. (1995). The exact distribution of Moran’s I. Environment and Planning A, 27, 985–999.
Author information
Authors and Affiliations
Rights and permissions
Copyright information
© 2018 Springer International Publishing AG
About this chapter
Cite this chapter
Griffith, D.A., Paelinck, J.H.P. (2018). Spatial Autocorrelation and Spatial Interaction Gravity Models. In: Morphisms for Quantitative Spatial Analysis. Advanced Studies in Theoretical and Applied Econometrics, vol 51. Springer, Cham. https://doi.org/10.1007/978-3-319-72553-6_9
Download citation
DOI: https://doi.org/10.1007/978-3-319-72553-6_9
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-72552-9
Online ISBN: 978-3-319-72553-6
eBook Packages: Economics and FinanceEconomics and Finance (R0)