Firm Site Selection and Industrial Land Use

A barber, a factory, and a warehouse can all employ local residents, yet choose very different sites. The barber needs customers to reach the shop. The factory needs workers and supplies to reach the plant. The warehouse needs reliable routes for moving goods. Firm location begins by asking which movements are costly and whose access matters most.

Learning goals

Explain households’ two roles, derive an industrial land-rent bid, distinguish freight access from CBD proximity, calculate residual land values, and evaluate a site using labour, logistics, and externalities.

1. Are households customers or workers?

Imagine two firms selling the same annual value of output. One is a neighbourhood hairdresser; the other makes components sold overseas. They have different reasons to care about nearby households.

Household role	Firm’s location problem	Examples
Client	Make it easy for customers to visit or buy	Barber, local shop, branch-based services
Factor of production	Recruit workers and compensate them sufficiently to reach the workplace	Factory, assembly plant, distribution facility

For local services, customer travel is a friction. A shop can lose sales if customers must travel too far. Population concentrations, passing traffic, visibility, and competing outlets matter.

For an exporting manufacturer, local households may primarily supply labour. An isolated plant might save land costs but need higher wages, staff transport, or other benefits to recruit workers. Those expenses can consume its land saving.

These roles overlap. A supermarket hires workers as well as serving customers. A factory may have a local showroom. The framework asks which flows dominate, rather than assigning every business permanently to one category.

Teach it back: Why could a barber and a factory choose different locations even when both hire the same number of people?

Self-check: identify the costly journey

The barber depends on frequent customer visits, while the exporting factory depends more on workers, materials, and goods reaching the plant. The economically important journeys differ.

2. From central lofts to large industrial sites

The supplied lecture describes a historical shift from central ports and rail terminals, with expensive nearby land and multi-storey industrial buildings, toward road-served plants and warehouses using large parcels and often a single main floor.

The mechanism is straightforward. If goods must pass through one terminal, a nearby site saves expensive local hauling. If trucks can reach many locations through a highway network, firms gain more site choices. A broad production floor can reduce vertical movement and accommodate loading, vehicle circulation, and flexible layouts, but it consumes more land.

Treat this as a stylised historical pattern rather than a universal sequence in which one transport mode replaced all others. Ports, rail, trucks, and airports remain complementary. Some modern locations also use multi-storey logistics buildings when land is scarce. The relevant question is what a specific production and distribution system needs.

Teach it back: Why can cheaper freight movement make a larger suburban parcel preferable to a compact central building?

3. An industrial firm’s maximum rent bid

Imagine a factory selling a standard product. It can pay for production, shipping, and land, while still earning the profit available elsewhere. This remaining amount determines its land bid.

Use the supplied per-unit profit equation:

pi = S - AC - sd - r_f(d)/Q.

Symbol	Meaning	Consistent units
S	Sale price per unit of output	Currency per product unit
AC	Average non-land production cost, including capital	Currency per product unit
s	Shipping cost per output unit per distance	Currency per product unit per mile
d	Distance to the relevant freight hub or market	Miles
Q	Annual output per acre	Product units per acre per year
r_f(d)	Annual land rent	Currency per acre per year
pi	Required profit per output unit	Currency per product unit

Here S means sale price, not the stock of space used in the first lecture. Q means output per land area, not residential lot size q. Keeping units visible prevents these symbols from becoming misleading shortcuts.

Rearranging gives the maximum annual land rent consistent with that profit:

r_f(d) = [S - AC - sd - pi]Q.

Holding S, AC, s, Q, and pi fixed, moving outward by one mile raises shipping cost by s per output unit. The firm therefore needs annual land rent per acre to fall by sQ:

dr_f/dd = -sQ.

A factory example

Let S = 100, AC = 70, required pi = 10, shipping cost s = 0.20 per unit per mile, and annual output Q = 1,000 units per acre.

Distance to hub	Shipping cost per output unit	Maximum annual land rent per acre	Profit after paying that rent
0 miles	0	20,000	10 per output unit
10 miles	2	18,000	10 per output unit
20 miles	4	16,000	10 per output unit

The rent gradient is -0.20 x 1,000 = -200 per acre per year per mile. At 10 miles, profit is 100-70-2-18,000/1,000 = 10. Lower land rent exactly compensates for higher shipping cost.

Equal profit is an equilibrium benchmark for comparable firms and sites under the stated assumptions. It does not mean every observed firm earns identical profit. Differences in productivity, contracts, risk, or amenities can matter. A negative calculated rent bid means this use cannot support a positive land rent at that site under these inputs; it does not imply the market must charge negative rent.

4. Why a modern industrial gradient can be flatter

The gradient’s magnitude is sQ. A reduction in shipping cost per mile s reduces the rent premium for being close to the modelled hub. A fall in output per acre Q also reduces that per-acre premium, holding the other inputs fixed.

For example, change s from 0.20 to 0.05 and Q from 1,000 to 400. The gradient becomes -20 per acre per year per mile, compared with -200 before. The 10-mile rent difference falls from 2,000 to 200.

A single-storey layout does not mechanically prove that Q falls: automation, product mix, inventory turnover, and operating intensity also determine output per acre. Likewise, road transport does not make shipping costs literally zero. Reliability, congestion, tolls, and delivery deadlines can remain important.

Crucially, d is distance to the relevant hub or market, not automatically distance to the CBD. A weak relationship between industrial rent and CBD distance can coexist with a strong premium for direct motorway, airport, or port access.

The supplied notes attribute a weak Chicago industrial CBD-distance effect to a Buttimer and Rutherford study involving about 6,000 properties. Treat the exact sample and finding as a lecture attribution pending the underlying study’s specification; it is not evidence that all industrial access premiums vanish. A flat CBD gradient and an interchange-access premium answer different questions.

Teach it back: Why might two equally suburban sites command different rents if one is five minutes from an interchange and the other requires a long trip through local streets?

5. Competing uses bid for the same land

The previous section concerns annual land rent. Now switch to a development valuation in which P and C are upfront monetary amounts per unit of floor area, and p is a residual land value per unit of land area. Do not compare these values directly with annual r_f without a consistent capitalisation step.

The residual formula is:

p = F(P-C),

where F is Floor Area Ratio, P is property sale price per square foot of floor area, and C is construction cost per square foot of floor area. FAR converts a margin per floor area into residual value per land area.

Illustrative use	FAR F	Price per floor sq ft	Construction cost per floor sq ft	Residual per land sq ft
Industrial	0.40	45	30	6.00
Single-family housing	0.35	250	150	35.00
Townhouse	1.50	200	175	37.50
Mid-rise office	3.00	250	200	150.00

The townhouse value is 1.50 x (200-175) = 37.50, correcting the supplied table’s 37.00. These are illustrative lecture inputs, not current market estimates.

On a 10,000-square-foot site, the industrial residual is 60,000, while the office residual is 1,500,000. Under these assumed prices, costs, and feasible densities, the office use can outbid industry for the site.

This does not prove offices dominate every parcel. Prices vary by location, and a site may lack office demand or permit only industrial use. The simple calculation also omits costs unless they are incorporated in C or deducted separately: professional fees, finance, infrastructure, taxes, risk, and a required developer return can materially reduce the supportable land bid.

The sorting mechanism is conditional: central office or residential demand can support high land bids, while a low-FAR industrial use may find suitable land cheaper on the perimeter. A peripheral location is an outcome of relative bids and site requirements, not an intrinsic rule that industry always belongs far from the centre.

6. Labour access belongs in the cost calculation

The basic freight equation holds AC constant. That abstraction hides an important firm-household link. If labour becomes more expensive at some sites, write AC(d) or model wage and recruitment costs separately:

r_f(d) = [S - AC(d) - sd - pi]Q.

With Q and other inputs fixed, the slope becomes -Q[s + AC’(d)]. If a location reduces labour costs, that saving can offset additional freight costs. If it worsens recruitment, the land discount may be insufficient.

For example, a site may save 100,000 in annual land costs but require 140,000 more in wages, staff transport, and recruitment. Before any freight or productivity change, the move makes the firm 40,000 worse off annually. Compare the whole operating system rather than the property bill alone.

This helps explain why access to suburban population nodes can matter. It also prevents treating “suburban” as a reliable proxy for cheap labour: commuting patterns, skills, shift times, and competing employers differ by place.

7. Access benefits and proximity costs

Access is the ability to reach useful destinations efficiently. Proximity is physical nearness. They are related, but not identical.

Site feature	Potential benefit to the firm	Potential cost or limitation
Highway interchange	Faster distribution and wider labour catchment	Congestion, tolls, turning restrictions
Port or airport connection	Better access to freight networks	Delays, handling costs, route dependence
Large parcel	Yard space, loading, expansion, single-level workflow	More land required and infrastructure costs
Nearby residential population	Worker access	Noise, traffic, pollution, and community impacts
Previously developed industrial land	Existing industrial infrastructure or lower purchase price	Remediation, flood exposure, ground conditions, liability

Industrial uses can benefit from a highway connection while nearby households dislike truck noise. That is a negative proximity externality: a cost borne by neighbours that may not be fully reflected in the firm’s private calculation. Good network access does not erase those costs.

The supplied notes mention recycled, wet, or environmentally constrained land. A low asking price alone does not make such a site economical or suitable. Cleanup, drainage, flood resilience, permitting, and engineering obligations enter the site comparison. Industrial use does not automatically make constrained land usable.

8. A practical site comparison

When comparing two sites, keep the product and annual output consistent. Estimate delivered input and outbound freight costs, labour and commuting implications, facility and land costs, and site-specific operating constraints. Include reliability and disruption where they affect production or service promises.

Separate one-off acquisition and construction costs from annual expenses, then convert them to a common present-value or annual-cost basis. Finally, compare the project’s supportable land bid with the actual site price and alternative uses. A cheaper parcel can be a more expensive production location once access and operating costs are included.

Feynman study lab

Explain it simply

A firm chooses a site by asking which journeys matter. A shop wants customers to arrive easily. A factory wants workers, inputs, and output to move economically. A site that saves shipping costs can command more rent, but its land bid also depends on how much output or floor space it supports. Another use can win the parcel by creating a larger residual value.

Teach it back without looking

Explain the 10-mile factory example aloud in two minutes. Show why a shipping bill of 2 per unit is offset by a land-rent discount of 2,000 per acre per year. Then answer:

If industrial rents barely decline with CBD distance, why can motorway access still be valuable?

Check your explanation after you try

CBD distance may poorly measure the journeys the firm actually makes. Access to an interchange, port, or distribution route can reduce shipping time, cost, and uncertainty even when distance from the old city centre has little explanatory power.

Transfer questions

1. Let S = 120, AC = 80, pi = 10, s = 0.50, Q = 200, and d = 20. What annual land rent per acre can the firm support?
2. What is the rent-gradient slope under those inputs?
3. If s halves while Q doubles, what happens to the slope?
4. At FAR 0.5, a sale price of 100 and construction cost of 70 per floor square foot support what residual per land square foot?
5. Does that residual represent annual land rent?
6. A remote site saves 50,000 in annual rent but adds 30,000 in freight and 40,000 in labour costs. Is it cheaper overall before other differences?

Check the transfer questions

1. Shipping costs 10 per unit, leaving (120-80-10-10) x 200 = 4,000 per acre per year for land.
2. -100 per acre per year per mile.
3. No change: the product sQ is unchanged.
4. 15 per land square foot.
5. No. With upfront sale price and construction cost inputs, it is an upfront residual value. Annual rent requires a consistent income and capitalisation treatment.
6. No. Total annual costs rise by 20,000.

Find the gap and try again

Locate the point where your explanation became vague: customers versus workers, distance versus access, output per acre versus FAR, or annual rent versus asset value. Revisit its example, then close the notes and explain it again using an actual type of business. Change one cost and predict the location incentive before calculating. Tomorrow, rebuild the rent-bid equation from the firm’s revenue and costs.