Industrial clusters: explain proximity and persistence
Several factories close together form a spatial pattern. Explaining that pattern requires more than saying that the factories benefit from being close. What service do they share? How does the service become available? Can each firm actually reach it? Dated records also help explain why an established cluster may remain after its original raw-material source changes.
The district, businesses, demand units, timings and cost units below are entirely invented. They describe a geographical model, not real firms, measured industrial performance or advice about investing in a location. The aim is to connect an observed arrangement to a specific mechanism and a clear limitation.
Distinguish shared benefits from growth within one firm
An external economy of agglomeration is a benefit related to the surrounding concentration of activity. Several separate firms may jointly provide enough demand for a specialised supplier. A pool of workers with relevant skills may make matching workers and vacancies easier. Exchange of ideas may support learning, where an actual channel for exchange exists.
An internal economy of scale has a different boundary: it arises within an individual firm as its scale of operation changes. If Firm A alone spreads the cost of its own machine over more output, that is not the same mechanism as three independent firms supporting a nearby repair business. Identify who gains, what changes and whose activity makes the benefit possible.
Text alternative: A, B and C are west of the river and connect by roads to one repair supplier. D appears nearby on the east bank, but there is no direct crossing. Its route continues to a bridge outside the diagram. Supplied journey records give A, B and C five to eight minutes to the supplier; D takes 45 minutes. The dotted gap is not a usable route.
| Firm or site | Repair demand | Journey to shared supplier | Record status |
|---|---|---|---|
| A | 2 service units | 5 minutes | Established firm |
| B | 2 service units | 6 minutes | Established firm |
| C | 2 service units | 8 minutes | Established firm |
| D | Not supplied | 45 minutes | Candidate site across river |
The model supplier operates only if it has at least six service units of local weekly demand. A service unit is a standardised repair job in this exercise, not money or an hour. The model also assumes the supplier has enough capacity for six units. A threshold for demand is different from a guarantee of spare capacity at any future level.
Worked case 1: pooled demand supports a specialist
A, B and C supply 2 + 2 + 2 = 6 weekly units. That meets the supplier’s stated minimum. A alone supplies only 2, which is 4 short. The cluster therefore supports a service that one model firm would not support on its own. The causal chain is shared demand, a viable local service under the given rule, and short recorded access for each firm.
The benefit is not simply “three factories mean cheaper production.” No repair price or production cost saving is given. We can establish availability and access within the model, but cannot calculate a saving. If C’s demand fell to zero, the remaining total would be 4, below the threshold; the conclusion depends on continued demand rather than permanent geographical magic.
Worked case 2: the input source moves but the cluster stays
| Evidence | 2022 | 2026 |
|---|---|---|
| Main input source | Nearby source supplies cluster | New source farther away supplies cluster |
| Input delivery cost per year | 10 cost units | 16 cost units |
| Shared repair supplier | Available | Available |
| Specialised labour | Relevant skills available locally | Relevant skills still available locally |
| Factory buildings and connections | Existing and usable | Existing and usable |
The delivered-input cost rises by 16 − 10 = 6 cost units per year. That weakens the original location advantage, but does not remove the repair service, skilled labour or usable installed facilities. Continued operation after the source changes is consistent with industrial inertia: persistence after an earlier locational influence has weakened or disappeared.
An additional scenario says relocating would require 30 one-time cost units and would save 6 recurring units each year, with all other future costs unchanged. Simple undiscounted recovery would take 30 ÷ 6 = 5 years. This limited calculation illustrates a switching barrier. It is not a valuation: timing, discounting, risks, moving disruption and alternative uses of assets are omitted.
Do not argue that money spent in the past must be recovered by staying. Irrecoverable past spending is sunk. The relevant continuing facts are that buildings remain usable, replacement or moving may require new expenditure, and shared services still exist. Firms can eventually relocate if the future comparison changes; inertia does not mean immobility forever.
Worked case 3: apparent proximity without useful connection
D is drawn close to the repair supplier, but the river has no direct crossing there. Its supplied 45-minute route takes 37 minutes longer than C’s eight-minute route. The map supports a physical separation, while the journey record establishes the access difference. Measuring a short straight line between the boxes would answer a different question.
D’s record supplies no repair demand, skill availability or operating cost. It cannot be added to the six-unit demand total, and its closeness does not prove that a fourth firm would benefit. A connection-based explanation is stronger than a proximity-only explanation because it checks whether the proposed shared service is practically accessible in the given model.
Three solved explanation tasks
1. Name the mechanism
Two additional fictional firms each provide three service units to a supplier needing six. Explain one benefit without inventing a price.
Answer: Combined demand is 3 + 3 = 6, meeting the minimum. Their joint demand can support a specialised local service. That is an external agglomeration mechanism; no numerical cost reduction is established.
2. Distinguish assets from recurring benefits
Classify an already usable factory building and continuing access to a shared skills pool in the persistence explanation.
Answer: The building is an existing fixed asset. Its future usability may avoid replacement needs, but its historical purchase cost is not an annual saving. A skills pool provides a potentially recurring matching benefit. Neither should be counted twice as both a purchase cost and yearly income.
3. Test a nearby-location claim
“Every firm within one centimetre of the supplier on this drawing receives the same benefit.” Is that supported?
Answer: No. The schematic has no distance scale, D lacks a direct crossing, and demand and access differ. Compare route times, service requirements and actual connections before asserting equal benefits.
The World Bank’s Reshaping Economic Geography of East Africa discusses specialised suppliers, labour pooling and knowledge exchange as agglomeration mechanisms. The numerical district here is original. Congestion, competition for land or service overload could offset benefits, but their size cannot be invented from this map.
Connect mineral deposits and energy supply to industrial location
The cluster lesson above explains shared services and persistence. Its input source was deliberately unspecified. This separate fictional district supplies mining and energy conditions, showing why a factory need not sit directly on a deposit. Coordinates, material quantities, deliveries and costs are invented. They describe no mineral reserves, real firms, rights to land or current power availability.
A mine is constrained by where a workable deposit occurs, but finding mineral-bearing rock is not enough to show economic extraction. USGS guidance on mineral extraction identifies deposit location and shape, rock strength, grade, costs and commodity price as relevant factors in selecting a mining method. A later processing plant also needs connections to inputs, energy, water, skills and markets. Different stages can therefore occupy different places.
| Node | Easting | Northing | Supplied function |
|---|---|---|---|
| M | 0 | 0 | Only proved deposit in this model; mine and concentrator |
| J | 20 | 0 | Junction, grid connection and shared repair service |
| K | 20 | 30 | Market and alternative factory site |
| H | 0 | 10 | Potential hydro site on a supplied river, not an operating station |
North is up. The only transport links are M–J (20 km) and J–K (30 km), both straight and bidirectional in this model. H’s river is not a road and no M–K direct transport connection is supplied. Route M–J–K is 50 km; its straight-line M–K separation is √(20² + 30²) ≈ 36.1 km, which cannot replace the route distance for hauling. The table is a complete native map alternative; do not measure kilometres from cell widths.
Worked resource-to-industry link: reduce the material transported
A complete model process starts with 10 t of run-of-mine material at M, containing 2 t of the tracked valuable component and 8 t of other material. The stipulated concentrator recovers all 2 t of that component plus 1 t of other material into a 3 t concentrate, leaving 7 t of residue at M. This perfect component recovery is a teaching assumption, not a real plant efficiency.
Transporting all 10 t to J uses 10 × 20 = 200 tonne-km; transporting the 3 t concentrate uses 3 × 20 = 60 tonne-km. The model reduction is 140 tonne-km per batch. This supports an initial mass-reducing stage near the mine when other requirements permit. It does not prove a financial saving because processing cost, residue management and haulage charges are missing. “Tonne-km” measures mass multiplied by distance, not a mass of product or a currency.
Worked energy-location contrast
Hydropower depends on water flow and usable fall, often called head; EIA’s hydropower explanation identifies both. For an explicitly ideal comparison, H has flow 4 m³/s and head 30 m, while a second candidate H2 has flow 6 m³/s and head 10 m. If water density, gravity and efficiency are held equal, potential power is proportional to flow × head: 120 versus 60, so H has twice the model potential. These products are not megawatts.
A gorge alone does not establish reliable year-round generation. Seasonal flow, machinery, environmental flow needs, reservoir effects where relevant, connection capacity and permissions remain unassessed. Similarly, EIA describes geothermal reservoirs as underground settings where heat and water occur together; an electricity project needs an adequate accessible resource and connections. A volcano outline or sunny photograph is not a generation record. This case compares the stated hydro inputs only; it invents no present Kenyan energy supply.
| Recurring category | Factory at J | Factory at K |
|---|---|---|
| Delivered concentrate | 12 | 20 |
| Electricity | 18 | 14 |
| Product delivery to market | 10 | 0 |
| Labour and shared services | 8 | 15 |
J totals 48 and K totals 49 cost units/year. K’s cheaper supplied power does not make its listed total lower. The 1-unit difference is conditional on this incomplete model, which omits capital costs, environmental impacts, reliability and capacity. The earlier cluster cases explain how a repair service can depend on pooled demand; preserve that reasoning rather than assuming that every nearby factory automatically receives a benefit.
Independent linked-location task
In a new batch the 10 t input makes 4 t of concentrate; all other route lengths are unchanged. Calculate concentrate transport work from M to K via J and compare it with hauling the full 10 t. A separate revised cost scenario raises J’s electricity cost from 18 to 22, with all other listed costs unchanged. Which site has the lower listed recurring total now, and what remains missing before a real siting decision? Suggested classroom allocation: 6 marks.
Explained answer
The route is 20 + 30 = 50 km (1). Concentrate transport work is 4 × 50 = 200 tonne-km, compared with 10 × 50 = 500 tonne-km, a 300 tonne-km reduction (2). J’s revised total is 12 + 22 + 10 + 8 = 52; K stays 49, so K is lower by 3 cost units/year (2). Actual supply reliability, capacities, capital and residue-management costs, environmental/social effects or permissions remain necessary before a real decision (1). Keep this changed scenario separate from the earlier 48-versus-49 record.
Frequently asked questions
Does a cluster automatically lower costs?
No. A concrete mechanism and supporting records are needed. Shared services can help, while congestion or competition can create costs.
What makes this an external economy?
The benefit depends on surrounding independent firms jointly supporting a service, rather than only one firm increasing its own production.
Is industrial inertia permanent?
No. It describes persistence despite changed initial influences. Future changes can still make relocation attractive or necessary.
Does past expenditure justify staying?
Irrecoverable past spending alone does not. Remaining asset usefulness, replacement needs and future costs matter to the decision.
Why keep one-time and annual costs separate?
They have different time bases. The simple recovery calculation compares a one-time outlay with a stated annual saving; adding them without a period is misleading.
Can I infer knowledge exchange from nearby buildings?
No. Proximity may facilitate contact, but actual exchanges or a credible connecting mechanism need evidence.
Does the map choose a real factory site?
No. It omits many environmental, economic and access considerations and makes no real siting recommendation.
WhatsApp Us Now