Environment and Planning A: Economy and Space

The Garin–Lowry model is a well-documented spatial interaction model which does, however, leave unanswered questions that relate to the constraint procedure and the supply component. The incorporation of a stock simulation model satisfies both these limitations and if the iterative structure is reformulated the whole is capable of automatic calibration. This paper concentrates on the issue of automatic calibration by presenting a framework within which existing search algorithms can be incorporated, and presents the results of such a calibration and a subsequent run of the model using the automatically derived values for the parameters.

One of the important problems in environmental management is the sharing of responsibility for cleaning the pollution. It may happen that a particular region may suffer pollution caused by the production of inputs and outputs which are consumed outside the region. This is also further complicated by the diffusion of pollutants over space. Within the framework of a balanced regional input–output model, a scheme for subsidy/tax has been presented so that interregional equity is maintained. Pollution control strategies must result in regional unemployment, migration, capital movements over space, changes in interregional competitive position, etc. A dynamic scheme is suggested so that, given some exogenous policy decisions and technological constraints, the activity levels of production in different regions are determined so that the system is in equilibrium.

Two major drawbacks prevent the wide application of Markov processes to land use and other types of planning analysis: the lack of temporal homogeneity of the transition mechanism, and the lack of independence of spatial data. In order to avoid these difficulties Gilbert (1972) proposed the use of the nonhomogeneous, two-state, continuous time Markov process and the semiMarkov process. In this paper we expand upon Gilbert's discussion and illustrate the calibration of the nonhomogeneous (birth and death) model while commenting on the empirical intractability of the semiMarkov process.

Published work on industrial linkage flows has tended to ignore both their relevance in underlying intracity industrial locational patterns and also the extent to which they relate to the present-day evaluation of such locations by firms. A series of hypotheses are set up and tested for Greater London concerning the interrelationships between linkage flows, spatial patterns, and locational evaluation. In general it is found that the volume of freight moving to and from firms in linkage flows is a poor predictor of spatial patterns and locational evaluation, despite the fact that London firms perceive linkages as important, not only in themselves but also in their strong relationship both to the degree to which firms concentrate on local markets and to their rates of decentralisation. Possible reasons for this paradox are suggested and, finally, some practical implications of the findings are indicated.

After a review of different approaches to educational planning, the usefulness of various optimization models for regional educational planning is examined. These models are criticized for not making a distinction between education as an investment and education as a consumption good. Also they do not distinguish between supply of and demand for education, which leads to serious problems. In optimization models for the whole economy the solution depends on a large number of coefficient matrices. Each of these matrices is subject to errors of measurement. These can lead to very large errors in the variables concerning the educational system.

The probability of error is drastically reduced if we limit ourselves to a sectoral model of the educational system, in which the future manpower requirement is an exogenous variable. The optimization model presented in this paper is an attempt along these lines, and takes into account a regional decision structure.

This paper describes an investigation into the gain in the predictive and descriptive abilities of a trip-distribution model which takes into account the differences shown to exist between various categories of trip maker. Four classifications of trip makers are considered—age, sex and marital status, socioeconomic group, occupational classification, and standard industrial classification. A trip-distribution model is calibrated for each of the categories mentioned above. Calibration methods are discussed and a new method whereby the significance of the differences in the values in the trip-distribution model parameters between categories may be examined. It is concluded that it is not worth running separate distribution models for each distinct category of trip maker since the improvement, at least in descriptive ability, is only marginally better than the fit shown by a single trip-distribution model.

This paper presents two kinds of entropy-maximising location model for situations in which there exist two classes of individual. In one class of model, sampling is with replacement; in the second, sampling is without replacement. These problems are solved, and interpreted as describing the distribution of items and the component classes of items by distance, size, or time. The paper emphasises methods of obtaining complete solutions to each model for certain ideal cases. The first class of model has negative exponential solutions, the second is characterised by logistic functions.