ISSN:
1432-0592
Source:
Springer Online Journal Archives 1860-2000
Topics:
Architecture, Civil Engineering, Surveying
,
Geography
,
Economics
Notes:
Abstract In regional science, there is a class of linear, multisectoral models whose solutions were first found as the summation of a power expansion series. Regional input-output models, Garin-Lowry models and Markov chain models represent familiar examples. Each begins with a partitioning of system state space into discrete sectors. Each generates sector-specific multipliers. Accurate estimation of multipliers depends, in part, upon the degree of homogeneity within system state space sectors. This paper's first objective is to explore the effects intrasectoral heterogeneity on multipliers used in a Markov model of residential vacancy chains. Its second objective is to develop a method for defining internally homogeneous sectors. Our data is a linked 1975-'80-'85 file on housing and population in Gävle, Sweden. It is used to calibrate vacancy chain models with crude and refined sectoral definitions. Matrix algebra is used to find the effects of sectoral definitions on vacancy chain lengths. Long-linear analysis is used to measure the contribution of sectoral refinements to overall data variation. A two-stage, nested clustering strategy is used to define homogeneous sectors for two time periods. Finally questions are advanced about the stability of sectoral definitions, multisectoral relationships, and the dynamics of sectoral formation.
Type of Medium:
Electronic Resource
URL:
http://dx.doi.org/10.1007/BF01583573
Permalink