EM 4103 Urban Planning II Lecture 9 Overview

EM 4103 Urban Planning II Lecture 9: Overview on Models in Planning and Use in Retail Planning

Shopping Models § The analytical models used by planners in assessing existing provision and future needs for retail are known as gravity models. § These models depend largely upon knowledge of the relationship between activity locations and the travel behaviour of the users who undertake such activities.

Gravity Models are so called because of an analogy with the ideas of Newton whose Law of Universal Gravitation states that “Two bodies in the universe attract each other in proportion to the product of their masses, and inversely to the square of their distance apart”, as follows: Where F = the force of attraction between the two bodies M 1 and M 2 = the respective masses of the two bodies d = the distance between the two bodies G = the gravitational constant

In the analysis of urban systems, the amount of interaction between two areas A and B, is related directly to the size or attraction of the areas and inversely to the distance separating them: Where IAB = the interaction between areas A and B PA and PB = the respective sizes of areas A and B (as measured, e. g in terms of population) d. AB = the distance between A and B = an exponent applied to the distance variable K = a constant which is empirically determined

Reilly's Law of Retail Gravitation 1. By this law the market boundary between two retail towns can be identified. 2. According to this law, shoppers living at any point between two towns A and B will be attracted to the towns in accordance with the relative populations of the two towns, and inversely with the square of the relative

Market Boundary d. A d. B PA A d. AB PB B A Market Boundary between two Centres

Given the choice faced by shoppers living between two cities A and B, the attraction of the shopping centre of City A, with population PA, to individuals living at an intermediate location, distance d. A from A, will be Where GA = the force of attraction of A

Similarly, the attraction of the shopping centre at City B, with population PB to individuals living at an intermediate location, distance d. B from B, will be: Where GB = the force of attraction of B

If an individual shops in the City with the strongest overall attraction, the market boundary or breaking point between these two retail centres is the point to which one city exercises the dominating retail influence, and beyond which the other City dominates. This is the equilibrium point where the attraction of A is equal to the attraction of B: GA = G B Substituting into the earlier equation: Shoppers living at this point will be indifferent between the two competing retail centres

By cross-multiplying, the market boundary can be identified as follows:

Market Boundary PA = 100000 A 20 miles 10 miles PB = 25000 B Reilly’s law as applied to determine market boundary

§ Shoppers will shop only in the town with the stronger overall attraction. § The application of Reilly’s law means discontinuous trade areas. § In practice, however, shoppers are not confronted with the simple bilateral choices that

§A §D §E §B §C Discontinuous Trade Areas for Retail Centres

Calculate attraction measures for each shopping centre Flow Chart for Typical Shopping Model Find the total attraction that shopping centre have on residential zones by summing Calculate the probability of interaction between residential zones and shopping centres Estimate consumer expenditure available in each residential zone Cj Calculate expenditure flows from residential zones to shopping centres Calculate attraction measures for each shopping centre

A Probabilistic Approach to the Gravity Model K is similar to the gravitational constant, And , , and are the other model parameters, of of which can be empirically determined in the caliberation processes • • For a given period, the number of trips (T) made by those living at i to an activity or centre at j will increase with the number of trip makers (P) located at I, and with the number of opportunities (H) available at j to satisfy the demands of these trip makers. The number of trips decreases with the distance (d) or cost of travelling from i to j.

The model may now be re-specified as follows: S = expenditure flow C = the money available for purchases F = total retail space

The expenditure available in each zone is the total expenditure available in the subregion: Assuming that = = 1,

Cross-multiplying: Thus:

Applying the Constrained Gravity Model: The Three Zone Study Area Zone 1 Zone 3 Existing Shopping Centre A Proposed Shopping Centre C Existing Shopping Centre B Zone 2

Applying the Constrained Gravity Model: The Three Zone Study Area Zone 1 Zone 3 10 2 15 2 A 6 C 15 10 5 2 B Zone 2

Calculate attraction measures for each shopping centre Flow Chart for Typical Shopping Model Find the total attraction that shopping centre have on residential zones by summing Calculate the probability of interaction between residential zones and shopping centres Estimate consumer expenditure available in each residential zone Cj Calculate expenditure flows from residential zones to shopping centres Calculate attraction measures for each shopping centre

Limitations of shopping models 1. They only describe the spatial interaction of activities within the urban system but do not explain the behavioural reasons for any given pattern of location, and thus cannot be an effective predictive tool. 2. Secondly, the method necessarily simulates interaction of one activity, which is modelled, while other parts of the urban system are held constant. The technique is thus criticized as being partial in nature.

§ The approach is criticized as being static and have failed to adequately embrace the dynamic elements of the urban economy. § Gravity models consider activities at an aggregate level e. g. population, expenditure and retail activity, and do not look at the smaller components of

There also practical and operational limitations as well. § There are problems of defining variables as well as caliberating or measuring them. § In shopping studies, there is the issue of determining levels of consumer expenditure.

• In general, gravity models can provide broad and useful information for policy makers. • However, it is important to note that these techniques are demand-side models and ignore supply-side considerations. • They should be supplemented by appraisals of shopping facilities to identify commercial and other factors

A composite General Urban model: The Garin-Lowry Model

A General Urban Model: the Garin-Lowry Model § The shopping models described previously are partial models which simulate one part of the urban system. § A general model attempts to describe several parts of the urban system and simulate the allocation and interaction between several land use activities – people, work and homes.

1. The Garin-Lowry model is one general model that combines a residential location model and a retail location model coupled together through the economic base mechanism 2. The model operates by deriving dependent population and service employment 3. It allocates households and retail employment to zones of the study area.

1. From an estimated level and distribution of basic employment, the model first allocates basic employees to residential zones. 2. Then the population associated with these basic employees is found. 3. This population demands services from retail employees, which are allocated in service centres.

§ In turn, these retail employees are allocated to residential areas and the associated population required to be serviced. § Further increments of population and service employment are derived and allocated, until this process converges, that is, until further increments of population and retail employment are negligible enough to be ignored.

A Steel Mill 1000 workers Shops and Schools for 500 workers A Settlement of 1000 Families The relationship between service employment and population is an iterative one, and there will be an eventual total balance greater than the first round figures Addition to the settlement of another 500 families

Basic Employment Population Service Employment Population The model builds on this simple relationship, adding a land use dimension to the essentially economic relationship

Allocate Retail Employees To Service Centres Retail employees To service Allocate Centres Basic Employees to Residential Zones Retail Employees To Residential Zones Derive Allocate Derive Level & Location Of Basic Employment (Given) Dependent Derive Requirement of Population on Service activities Basic employees (Retail Dependent Employment) Requirement of Population on Service activities Basic Derive (retail Employees Employment) Functional Structure of the Garin-Lowry Model

Notts-Derby Study Output from the Model: § Predicted non-basic workers per zone § Predicted population per zone § A complete inter-zonal journey to work matrix § A complete inter-zonal journey to service matrix; this approximates to a ‘journey to shop’ pattern § Internal migration components for population and non-basic jobs

Application Overall: § The model can be used at a macro level to test large scale changes or different alternative plans. § It can also be used for testing the impacts of policies such as green belts and new expressways.

The Learning Aspects of Models: Lowry in “The Journal of the American Institute of Planners”, May 1965: Ø The process of model-building is educational. Ø The participants invariably find their perceptions sharpened, their horizons expanded, their professional skills augmented.

2 1 3 4

1 67 97 33 397 40 4 2