Location Theories and Analytical Models

Introduction

Facility location is a long-term operations decision concerning the geographic position of a plant, warehouse, service outlet, distribution centre, hospital, data centre, or other operating facility. Location determines access to markets and suppliers, transportation expenditure, labour availability, regulatory exposure, resilience, environmental impact, and the future configuration of the supply network. Because buildings and infrastructure are relatively immobile and costly to change, a location decision has strategic consequences that often persist for decades.

Location analysis combines economic geography, operations research, transportation analysis, spatial analytics, and managerial judgment. No single model is universally superior. Weberian theory explains why firms may minimize transportation and labour costs; central-place theory examines the spatial distribution of service facilities; agglomeration theory explains benefits and costs of geographic concentration. Quantitative models then convert operational assumptions into comparable alternatives.

A sound decision process therefore has two stages. First, the organization defines the strategic and operational requirements of the facility. Second, it evaluates feasible alternatives using multiple methods, tests assumptions, and subjects the recommendation to governance and stakeholder review.

Learning Objectives

By the end of this chapter, the reader should be able to:

1. Explain Weberian, central-place, and agglomeration theories of location.
2. Apply factor-rating and break-even methods to location alternatives.
3. Use load–distance and centre-of-gravity models appropriately.
4. Formulate and interpret transportation-model results.
5. Explain how GIS supports spatial location analysis.
6. Design a multi-criteria decision process that integrates quantitative and qualitative evidence.
7. Critically evaluate assumptions, uncertainty, equity, environmental effects, and implementation risks.

1. The Strategic Logic of Location

A location decision should begin with the facility’s role in the operating system. A manufacturing plant may prioritize supplier access, utilities, labour skills, export routes, and production scale. A retail outlet may prioritize customer accessibility, visibility, population density, and competitive positioning. A hospital may emphasize travel time, emergency coverage, demographic need, and regulatory standards. A warehouse may prioritize transport connectivity, service-time requirements, inventory pooling, and land cost.

Location decisions occur at three levels:

The decision horizon matters. A strategic location model may use projected demand, wages, taxes, energy prices, and transport costs over 10–25 years. A tactical relocation decision may use a shorter horizon but still require sensitivity analysis. The relevant cost is not simply the purchase price. It includes recurring operating cost, one-time transition cost, risk-adjusted disruption cost, and the opportunity value of future expansion.

2. Classical Location Theories

2.1 Weberian industrial location theory

Alfred Weber’s industrial location theory seeks a location that minimizes total production-related cost, especially transportation and labour cost, while recognizing forces that may attract firms toward or away from concentrations of activity. The classical framework uses three central ideas:

• Material index: the weight of localized input materials divided by the weight of the finished product. A high index suggests that the process is material-oriented because substantial input weight is lost or transformed during production.
• Locational triangle: the geometric relationship among two material sources and one market. The least-transport-cost point depends on the quantities and distances involved.
• Isodapanes: contours joining locations with equal transport cost. A firm may move away from the minimum-transport point if labour savings or agglomeration benefits exceed the additional transport cost.

In simplified form, total transport cost is:

where TC is transport cost per planning period; w_i is the quantity or load moved on route i; d_i is distance; and c_i is cost per load-distance unit. The equation assumes that cost is proportional to load and distance and that route-specific rates are known. It is not adequate where rates contain quantity discounts, congestion effects, backhauls, or mode changes unless those features are modelled separately.

Weber’s theory remains useful as a first explanation of trade-offs, but its assumptions are restrictive. It tends to treat transport costs as uniform, markets as points, technology as fixed, and labour as spatially variable but otherwise homogeneous. Contemporary decisions must add taxes, infrastructure reliability, carbon cost, exchange-rate exposure, supplier ecosystems, and institutional risk.

2.2 Central-place theory

Central-place theory, associated with Walter Christaller and later developments, explains the spatial organization of settlements and service facilities. A central place supplies goods or services to a surrounding market area. Two concepts are especially important:

• Threshold: the minimum demand or population required to support a facility.
• Range: the maximum distance or travel effort that customers are willing to accept for a good or service.

Routine, low-value services generally have a small range and low threshold, so they occur frequently and close to customers. Specialized services have a larger range and threshold and are concentrated in fewer, larger centres. The theory’s idealized assumptions—uniform terrain, evenly distributed consumers, equal purchasing power, and frictionless competition—rarely hold exactly. Nevertheless, it helps managers distinguish convenience facilities from destination facilities and understand why a network may need several small outlets rather than one large centre.

In public and healthcare planning, central-place ideas can be combined with coverage standards. For example, an emergency facility may be evaluated by the proportion of the population within a specified travel time rather than by average distance alone.

2.3 Agglomeration economies and diseconomies

Agglomeration refers to the geographic concentration of related or complementary organizations. Localization economies arise from concentration within an industry: specialized suppliers, trained labour, technical services, and shared infrastructure. Urbanization economies arise from the broader scale and diversity of a city: universities, financial services, airports, cultural resources, and large labour markets.

Agglomeration can lower procurement and recruitment costs, accelerate knowledge spillovers, and improve access to logistics services. It can also create diseconomies: congestion, higher rents, wage inflation, pollution, competition for scarce labour, cyber or infrastructure interdependence, and exposure to a common regional shock. A location team should therefore assess both the direct benefit of clustering and the correlated risk it creates.

A useful strategic question is not simply “Where are competitors located?” but “Which capabilities are shared, and which dependencies become concentrated?” A technology cluster may improve innovation while increasing exposure to a common skills shortage. A port cluster may reduce transport cost while increasing vulnerability to climate hazards or labour disruption.

3. Factor-Rating and Break-Even Analysis

3.1 Weighted factor rating

Factor rating is a transparent method for combining quantitative and qualitative criteria. The analyst identifies factors, assigns weights, scores each site, and calculates a weighted total:

where S_j is the total score of site j; w_i is the weight of factor i, with ∑w_i = 1; and r_ij is the rating of site j on factor i. Ratings must use a common scale, such as 1–10, where a higher value consistently means better performance. The method assumes that factors can be represented meaningfully on the selected scale and that weighted additivity is acceptable.

Example: three sites are rated on transport access, labour availability, utilities, and regulatory fit.

The totals are:

• Site A: 0.35(8)+0.25(7)+0.20(8)+0.20(6)=7.35
• Site B: 0.35(6)+0.25(9)+0.20(7)+0.20(8)=7.45
• Site C: 0.35(9)+0.25(6)+0.20(8)+0.20(7)=7.65

Site C ranks first, but the result is not a proof of optimality. The team should test whether the ranking changes when weights or ratings change. A criterion that is legally mandatory should not merely receive a low score; it should be treated as a screening constraint.

3.2 Break-even analysis

Break-even analysis compares fixed and variable costs for alternatives. For site j, total cost at output Q is:

where F_j is fixed cost per period, v_j is variable cost per unit, and Q is volume in units per period. The model assumes a linear cost relationship over the relevant range, constant prices and productivity, and comparable output quality.

The crossover volume between sites A and B is obtained by setting their costs equal:

provided v_A ≠ v_B. If Site A has lower fixed cost but higher variable cost, A is preferred below the crossover volume and B above it. For example, Site A has fixed cost $400,000 and variable cost $18 per unit; Site B has fixed cost $700,000 and variable cost $12 per unit. Thus:

Q*=(700,000−400,000)÷(18−12)=50,000 units.

At 40,000 units, A costs $1,120,000 and B costs $1,180,000. At 60,000 units, A costs $1,480,000 and B costs $1,420,000. The preferred site changes at 50,000 units. Break-even analysis should include relevant transport, labour, tax, and service costs rather than only building cost.

4. Load–Distance and Centre-of-Gravity Models

4.1 Load–distance method

The load–distance method evaluates candidate sites by the weighted distance between the site and points of demand or supply:

where LD_j is the load–distance score for candidate j; l_i is the load associated with point i; and d_ij is the distance from candidate j to point i. Lower is better. If different routes have different cost rates, use cost-weighted load–distance:

Distances may be Euclidean, rectilinear, road-network, travel-time, or generalized impedance. Straight-line distance is appropriate only when terrain and network structure do not materially distort travel.

4.2 Centre of gravity

For a continuous coordinate approximation, the centre of gravity is:

where (x_i,y_i) are coordinates of demand or supply points and w_i are loads, shipments, or demand quantities. Coordinates must use a consistent projected system, not uncorrected latitude and longitude for large areas. The model assumes proportional distance cost and does not account for road barriers, fixed facility costs, capacity, congestion, or legal restrictions.

Suppose three customers have coordinates and annual loads: A (2,8; 100), B (8,6; 150), and C (6,2; 50). Total load is 300. The centre is:

x*=(100×2+150×8+50×6)÷300=5.67

y*=(100×8+150×6+50×2)÷300=5.67

This point is a screening location, not necessarily a feasible site. The team must inspect nearby parcels, roads, zoning, flood risk, and actual travel time.

5. Transportation Models

A transportation model allocates shipments from sources to destinations at minimum transportation cost. Let x_ij be the quantity shipped from source i to destination j, and c_ij the cost per unit. The balanced model is:

subject to:

Here s_i is source supply, d_j is destination demand, and balance requires ∑s_i=∑d_j. If supply exceeds demand, a dummy destination can represent unused capacity; if demand exceeds supply, a dummy source can represent external procurement. Real applications may add lane capacities, minimum shipments, fixed-charge decisions, multiple products, service-time constraints, and carbon costs.

For facility location, the transportation model is commonly embedded in a mixed-integer formulation. Let y_k equal 1 if candidate facility k is opened and 0 otherwise. A simplified capacitated model is:

subject to demand satisfaction, facility capacity, and x_ik ≤ U_k y_k, where F_k is fixed opening cost and U_k is capacity. The binary variable links shipment to opening: no flow can use a closed facility. This model is more realistic than a centre-of-gravity calculation but requires reliable data and optimization software.

6. GIS-Based Location Analysis

Geographic information systems integrate spatial data, maps, network algorithms, and attribute databases. A GIS location study may include:

1. Data preparation: geocode demand points, standardize projections, check missing values, and define the analysis period.
2. Constraint screening: exclude protected land, flood zones, incompatible zoning, unsafe areas, or parcels below the required size.
3. Accessibility analysis: calculate road-network travel time, service areas, delivery reach, and congestion-adjusted distances.
4. Spatial demand analysis: estimate population, income, industrial activity, customer density, or future growth within catchments.
5. Overlay and scenario analysis: combine suitability layers with weights and test alternative assumptions.
6. Field validation: inspect candidate sites and verify that mapped conditions are current.

A raster suitability model may calculate a standardized score:

where p is a map cell or location, r_i(p) is its standardized rating on criterion i, and weights sum to one. Standardization is essential because distance, cost, risk, and environmental indicators use different units. A benefit criterion may be normalized as:

For a cost criterion, the direction should be reversed, for example r_i(p)=(max x_i−x_i(p))÷(max x_i−min x_i). GIS results can create false precision if data are outdated, spatial resolution is coarse, or weights conceal value judgments. Maps should therefore display uncertainty and data vintage where possible.

7. Multi-Criteria Decision Methods

Location decisions commonly involve conflicting objectives: cost, resilience, employment, emissions, customer access, community acceptance, and strategic flexibility. Multi-criteria decision analysis (MCDA) makes these trade-offs explicit.

A basic weighted-sum model is:

where z_ij is a normalized performance value. Benefit and cost criteria must be oriented consistently before aggregation. The method is attractive because it is understandable, auditable, and easy to perform in a spreadsheet. Its weakness is compensability: an excellent score on one criterion may offset a severe weakness on another. This is inappropriate when a criterion is a non-negotiable safety or legal requirement.

A robust MCDA process includes:

• Eligibility screening: remove alternatives that fail legal, safety, capacity, or environmental requirements.
• Criteria definition: state operational meaning, measurement method, time horizon, and data source.
• Weight elicitation: use management workshops, pairwise comparisons, budget allocation, or stakeholder surveys; document disagreements.
• Normalization: convert indicators to a comparable scale without hiding the original units.
• Scenario and sensitivity analysis: vary demand, weights, costs, lead times, and risk assumptions.
• Dominance and robustness checks: identify alternatives that perform well across plausible scenarios rather than only under one base case.
• Decision governance: record assumptions, conflicts of interest, approvals, and reasons for the final choice.

The Analytic Hierarchy Process can support pairwise comparison of criteria and alternatives, but its judgments may be affected by scale choices and rank reversal. Outranking methods are useful when decision makers do not accept full compensation between criteria. No method removes managerial judgment; the purpose is to make judgment visible and challengeable.

8. Integrated Illustrative Case: Regional E-Commerce Fulfilment Centre

Background

The following is a fictional illustrative case. Meridian Home, an e-commerce retailer, must locate a regional fulfilment centre to serve three metropolitan markets. The centre must support next-day delivery, accommodate 120,000 orders per month at launch, and expand to 200,000 orders per month within five years.

Problem

Three candidate sites are available. Site A is near the largest customer market and has a low travel distance but expensive land. Site B is in an established logistics cluster with strong labour availability but moderate congestion. Site C has inexpensive land and renewable electricity but is farther from customers and requires road improvements.

Analysis

The team first applies constraints. All sites satisfy zoning and minimum parcel size. Site C is rejected for the base case if the required road upgrade is not funded by the public authority; it remains a scenario alternative.

For the two feasible base-case sites, annualized fixed and variable costs are estimated as follows:

The crossover volume is:

Q*=(8.0−6.2) million ÷ (2.75−2.40)=5.14 million orders per year.

At the launch volume of 1.44 million orders per year, Site B has lower annual cost: approximately $10.16 million versus $11.456 million for A. At the projected 2.4 million orders per year, B remains lower: $12.8 million versus $13.76 million. However, cost alone omits service risk. Network analysis shows that A can reach 92% of customers within the next-day promise, while B reaches 86%. B has a larger labour pool and lower disruption risk because it has two independent highway approaches.

The team uses a weighted evaluation after treating legal compliance as a screen. Weights are: total operating cost 0.35, next-day coverage 0.25, labour availability 0.15, resilience 0.15, and renewable-energy potential 0.10. Scores are normalized on a 0–10 scale. A scores 8, 10, 7, 6, and 5; B scores 9, 8, 9, 8, and 7. The totals are:

• A: 0.35(8)+0.25(10)+0.15(7)+0.15(6)+0.10(5)=8.00
• B: 0.35(9)+0.25(8)+0.15(9)+0.15(8)+0.10(7)=8.45

Approach and solution

The recommended decision is Site B, subject to a service-contingency plan. Meridian should reserve peak-season carrier capacity, establish a small forward-stock arrangement near the largest market, and negotiate expansion rights at B. Sensitivity analysis is essential: if next-day coverage receives a weight above approximately 0.40, A may overtake B; if demand exceeds the forecast range, B’s labour and expansion advantages become more valuable.

Outcome and learning points

The case illustrates why no single model is sufficient. Break-even analysis clarifies volume economics; network analysis measures service; factor rating integrates strategic concerns; and scenario planning addresses uncertainty. The final recommendation is not simply the highest score. It is a conditional decision with explicit monitoring triggers, such as coverage falling below 85%, labour vacancy exceeding a defined threshold, or demand exceeding capacity.

9. Limitations, Ethics, and Contemporary Practice

Uncertainty and model risk

Forecast error, uncertain fuel prices, changing wages, new roads, climate events, and competitor actions can invalidate a base-case model. Analysts should use ranges, probability distributions, stress tests, and real-options thinking. A site with slightly higher expected cost may be preferable if it preserves expansion flexibility or reduces catastrophic downside.

Equity and community effects

Location decisions redistribute employment, traffic, pollution, tax revenue, and access to services. A facility may appear efficient while imposing noise or air-quality burdens on a low-income community. Ethical appraisal should assess distributional effects, meaningful consultation, accessibility, labour standards, and environmental justice. Community acceptance is not merely a public-relations variable; opposition can create legal delay, reputational damage, and project failure.

Environmental and regulatory governance

Relevant issues may include land conversion, water use, emissions, hazardous materials, biodiversity, building codes, labour law, data protection, and climate adaptation. Carbon accounting should define system boundaries and avoid double counting. Regulatory compliance is a minimum condition, not evidence that the location is socially optimal.

Technology and AI

Modern systems can combine GIS, real-time traffic data, digital twins, demand forecasts, and optimization. Machine-learning models may improve demand and travel-time prediction, but they inherit historical bias and may be difficult to explain. Governance should require documented data lineage, validation against out-of-sample periods, human review, cybersecurity controls, and monitoring for drift. AI should support—not replace—strategic judgment and stakeholder accountability.

Best-Practice Decision Protocol

1. Define the facility mission, service promise, planning horizon, and capacity range.
2. Separate mandatory constraints from preference criteria.
3. Assemble reliable, current, and geographically consistent data.
4. Screen regions before conducting detailed site analysis.
5. Use at least one economic model, one spatial model, and one multi-criteria model.
6. Distinguish fixed, variable, transition, risk, and external costs.
7. Test assumptions through sensitivity, scenario, and stress analysis.
8. Validate model outputs through site visits and stakeholder consultation.
9. Document weights, assumptions, data sources, and conflicts of interest.
10. Establish post-decision metrics and trigger points for corrective action.

Chapter Summary

Location theory provides complementary explanations of facility geography. Weberian theory emphasizes transport, labour, and agglomeration trade-offs; central-place theory explains thresholds, ranges, and service networks; and agglomeration theory addresses the benefits and risks of concentration. Factor-rating models integrate qualitative and quantitative criteria, while break-even analysis compares fixed and variable cost structures across volumes. Load–distance and centre-of-gravity methods provide useful spatial screening tools, but they require careful distance definitions and do not guarantee feasibility. Transportation and mixed-integer models represent flows, capacities, opening decisions, and network costs more rigorously. GIS adds spatial constraints, network travel times, and visual scenario analysis. MCDA makes conflicting objectives and value judgments explicit. Graduate-level practice requires sensitivity analysis, governance, ethical appraisal, and recognition that a model is a decision aid rather than a substitute for professional judgment.

Key Terms

Weberian location theory: A theory emphasizing transport and labour-cost trade-offs in industrial location.

Material index: The ratio of localized input-material weight to finished-product weight.

Locational triangle: A geometric framework linking material sources and a market.

Isodapane: A contour joining locations with equal transport cost.

Central place: A settlement or facility that supplies goods or services to a surrounding area.

Threshold: Minimum demand needed to support a facility.

Range: Maximum distance customers will travel for a good or service.

Agglomeration: Geographic concentration of related or complementary activities.

Localization economy: Benefit from concentration within a particular industry.

Urbanization economy: Benefit from the scale and diversity of a city.

Factor rating: Weighted evaluation of location alternatives against selected criteria.

Break-even volume: Output level at which two alternatives have equal total cost.

Load–distance method: Evaluation based on weighted distances between a site and demand or supply points.

Centre of gravity: Coordinate point calculated as the load-weighted average of locations.

Transportation model: Optimization model allocating supply to demand at minimum cost.

GIS: Geographic information system used to store, analyze, and visualize spatial data.

MCDA: Multi-criteria decision analysis for comparing alternatives across conflicting criteria.

Sensitivity analysis: Examination of how results change when assumptions vary.

Scenario analysis: Evaluation of decisions under alternative future conditions.

Review Questions

Short Answer Questions

1. What are the threshold and range concepts in central-place theory?
2. Why can agglomeration create both economies and diseconomies?
3. State the assumptions of a basic break-even location model.
4. What does a centre-of-gravity calculation provide, and why may its result be infeasible?
5. Why should mandatory legal criteria be treated as screening constraints in MCDA?

Long Answer Questions

1. Explain Weberian location theory and evaluate its relevance to modern supply-chain location decisions.
2. Compare factor rating, load–distance, and transportation models in terms of purpose, data, assumptions, and limitations.
3. Explain how GIS can be used from regional screening through final site validation.
4. Discuss the role of central-place and agglomeration theories in planning service and logistics networks.
5. Develop a governance framework for an AI-supported facility-location decision.

Analytical / Application Questions

1. Two sites have costs TC_A=500,000+20Q and TC_B=800,000+14Q. Calculate the crossover volume and identify the lower-cost site at 40,000 units.
2. Using the centre-of-gravity method, calculate the coordinates for locations (1,5), (7,5), and (4,2) with loads 100, 200, and 100 respectively. Explain two reasons why the calculated point may not be selected.
3. Design an MCDA for a hospital location decision with at least five criteria. Identify one criterion that should be a hard constraint, propose weights for the remaining criteria, and describe a sensitivity test.

Multiple Choice Questions with Answers

Question 1

Which concept identifies the minimum demand required to support a facility?

A. Range

B. Threshold

C. Isodapane

D. Material index

Correct Answer: B. Threshold

Question 2

In a weighted factor-rating model, what must the criterion weights normally sum to?

A. 0

B. 10

C. 1

D. 1000

Correct Answer: C. 1

Question 3

A site has fixed cost $400,000 and variable cost $18 per unit. A second site has fixed cost $700,000 and variable cost $12 per unit. What is the crossover volume?

A. 20,000 units

B. 30,000 units

C. 50,000 units

D. 60,000 units

Correct Answer: C. 50,000 units

Question 4

In the basic load–distance method, a lower score generally indicates:

A. Greater weighted spatial burden

B. Lower weighted distance

C. Higher fixed cost

D. Higher facility capacity

Correct Answer: B. Lower weighted distance

Question 5

What is the principal purpose of a centre-of-gravity calculation?

A. To estimate a load-weighted coordinate

B. To determine legal ownership of land

C. To calculate labour productivity

D. To forecast exchange rates

Correct Answer: A. To estimate a load-weighted coordinate

Question 6

Which variable in a facility-location mixed-integer model commonly indicates whether a facility is opened?

A. A continuous demand variable

B. A binary variable

C. A distance coefficient

D. A unit transport rate

Correct Answer: B. A binary variable

Question 7

Why must GIS distance measures be selected carefully?

A. All locations have identical road access

B. Straight-line distance always equals travel time

C. Network structure, terrain, and congestion can alter actual accessibility

D. GIS cannot store attribute data

Correct Answer: C. Network structure, terrain, and congestion can alter actual accessibility

Question 8

Which is the best treatment for a legally mandatory safety requirement in MCDA?

A. Give it a very small weight

B. Ignore it if total score is high

C. Use it as an eligibility constraint

D. Replace it with a marketing criterion

Correct Answer: C. Use it as an eligibility constraint

Question 9

At volumes below the crossover point, a site with lower fixed cost and higher variable cost is generally:

A. Preferred under the linear cost assumptions

B. Always infeasible

C. More expensive because fixed cost dominates

D. Identical in cost to the other site

Correct Answer: A. Preferred under the linear cost assumptions

Question 10

Which issue is a potential diseconomy of agglomeration?

A. Shared specialized suppliers

B. Knowledge spillovers

C. Congestion and wage inflation

D. Improved infrastructure access

Correct Answer: C. Congestion and wage inflation