Employment Decentralization and Edge Cities

If suburban land is cheaper and workers already live nearby, why would a firm pay to stay downtown? And if leaving downtown saves money, why do firms build suburban clusters rather than spread evenly across the map? Employment decentralization is the balance between the benefits of being together and the costs of reaching one another.

Learning goals

Measure decentralization, explain compensating wage differences, describe local rent peaks and commuting catchments, and compare agglomeration benefits with congestion costs. Use these ideas to reason about housing diversity, transport capacity, and commercial zoning.

1. Decentralization is a distribution, not just a photograph

A CBD, or Central Business District, is the main downtown business centre. Decentralization means a larger share of metropolitan activity is located away from that centre. A polycentric region has several employment centres. Here, an edge city means a substantial suburban concentration of jobs and services, not simply any housing estate on the urban fringe.

Suppose a region has the following cumulative shares of jobs within each distance from the same CBD:

Distance from CBD	Earlier distribution	Later distribution
Within 5 km	50%	30%
Within 15 km	80%	60%
Within 30 km	100%	100%

The later distribution puts relatively more jobs farther out. The cumulative curve rises less quickly near the centre. Saying it becomes “flatter” is useful shorthand, but it need not have a smaller slope at every distance: a suburban cluster can create a steep local rise farther out. Cumulative curves can also cross, so inspect the whole distribution.

Shares are not job counts. If total employment doubles, the CBD can gain jobs while losing its regional share. Likewise, office stock is not the same as employment: vacancy, floor space per worker, sector mix, and remote work affect the relationship.

The supplied notes describe historical US office and employment decentralization, including Washington, New York, Boston, and Los Angeles. Treat these as historical lecture context, not a statement about each city’s current office market. Sector differences matter: functions that depend heavily on specialised interactions may remain clustered while more separable operations disperse. Neither “finance stays central” nor “technology goes suburban” is a universal rule.

Teach it back: Can downtown gain 10,000 jobs while its share of regional employment falls?

Self-check: distinguish counts from shares

Yes. If the rest of the region grows faster, downtown’s absolute employment can increase while its percentage share declines. A suburban share gain alone does not establish downtown decline in absolute terms.

2. A wage premium can compensate for a costly commute

Consider a worker comparing otherwise identical jobs. A simple measure of the job’s value is annual wage minus annual commuting cost. Hold housing, job amenities, skill requirements, and other conditions fixed.

If both jobs must provide the same net reward, then:

w_CBD - t_CBD = w_suburb - t_suburb, hence

w_CBD - w_suburb = t_CBD - t_suburb.

If the suburban job pays 50,000 and commuting costs 2,000 annually, its net reward is 48,000. If the CBD commute costs 6,000, a CBD wage of 54,000 provides the same net reward. The 4,000 nominal wage premium compensates for the extra commute; it is not an extra 4,000 of spendable benefit after travel.

This mechanism can give a suburban employer a labour-cost advantage. The supplied notes mention CBD labour costs up to 15% higher in an empirical setting. Without a specified sample, occupation mix, and date, retain that as a lecture observation rather than a universal wage rule. The example above is invented and gives an 8% premium relative to 50,000.

Actual wages also reflect productivity, worker sorting, bargaining, amenities, and employer characteristics. Some workers live centrally and have shorter CBD commutes. A suburban job far from their home may require a premium instead. There is no universal wage gradient determined by CBD distance alone.

Teach it back: Why can a higher-paying downtown job leave a worker no better off?

3. The firm’s comparison includes productivity

A firm compares revenue or productivity with its full site cost. Lower suburban wages and office costs favour dispersal, but losing access to clients, specialised workers, or collaborators can reduce output.

For example, a 100-worker firm saves 4,000 per worker annually by using a nearby suburban labour pool. That is 400,000. If it also saves 200,000 on premises, the gross annual saving is 600,000. If leaving the central cluster reduces annual contribution before these site costs by 700,000, the move makes the firm 100,000 worse off. If that productivity loss is only 300,000, the move improves the result by 300,000.

Compare premises expenses on the same basis and avoid counting a land rent and an all-inclusive office lease as separate versions of the same cost. The wage saving also depends on being able to recruit suitable workers; a nearby population is not automatically the required labour pool.

4. Several centres create several access premiums

In a one-centre model, residential land rent can decline smoothly with distance from the CBD. Add a productive suburban employment node and nearby homes can gain access value. The resulting rent surface can have a local peak around that node as well as around downtown.

A commute shed is the area from which a workplace or centre draws workers. It is a travel catchment, not necessarily a circle. Roads, public transport, congestion, and work schedules shape it. Workers may travel suburb-to-suburb, outward from the CBD, or across centres instead of always travelling inward.

Local peaks are conditional predictions, not guarantees that the parcel beside an office tower has the highest residential rent. Noise, traffic, schools, housing quality, and planning rules also matter. Job centres can overlap, and workers do not always choose the nearest employer because jobs require different skills and offer different wages.

Teach it back: Why might distance to the nearest suitable workplace explain a home’s accessibility better than distance to the old CBD?

5. Why firms form clusters instead of dispersing uniformly

Being near other firms can create agglomeration economies: benefits from concentrated economic activity. Specialised workers and employers find better matches; suppliers can serve several clients; face-to-face contact can help exchange information; and a concentration of workers supports restaurants and other services.

Some benefits are market transactions, such as buying from a nearby supplier. Others are spillovers that a firm does not fully pay for or capture. Not every cluster benefit is an externality, and firms can also suffer from nearby competitors bidding up wages.

Subcentres offer a compromise: enough concentration to share useful connections, with less of the central city’s land and commuting burden. Growth eventually brings its own costs. A larger catchment can mean longer commutes, higher compensating wages, scarcer premises, and congestion.

A marginal-benefit example

Consider successive increments of employment in a hypothetical subcentre. The figures below are illustrative net changes per increment, measured in thousands of currency units per year. Agglomeration benefits exclude the wage, premises, and congestion costs listed separately.

Next employment increment	Additional agglomeration benefit	Additional labour, premises, and congestion cost	Net gain
First	120	60	60
Second	100	80	20
Third	80	100	-20

The first two increments improve the result; the third does not. In a smooth interior model, the best scale occurs where marginal benefit equals marginal cost, provided the net-benefit function has the appropriate maximum. Do not compare total benefits with marginal costs.

This is a benchmark, not a claim that real centres automatically attain a unique optimum. Coordination, uncertain growth, indivisible infrastructure, and planning can produce several possible outcomes. A private firm’s preferred expansion can also differ from the socially best scale if it imposes congestion on others or creates uncaptured spillovers.

6. What limits decentralization?

Constraint	Why cheap suburban land may not be enough
Housing diversity	A suburb with limited housing options may not accommodate all worker incomes or household needs
Skills and matching	Nearby residents may not have the required skills; recruiting farther afield can raise costs
Regional access	A well-connected CBD may reach a broader and more varied labour pool
Network layout	Radial routes favour inward trips; orbital connections can support cross-suburban access
Transport capacity	Congestion can consume the initial advantage of a growing centre

The supplied examples of Boston’s Route 128/I-495 and Washington’s Beltway illustrate the role of orbital routes. They do not imply that highways alone determine the location of jobs: transit, planning, business networks, and land supply also contribute.

If zoning limits the types or amount of housing near a centre, some workers must commute farther. Firms may need to pay more or accept a narrower recruitment pool. The suburban wage discount can then erode. Conversely, mixed housing near jobs can improve access without guaranteeing that every household can work locally.

Greater transport capacity can support a larger centre before congestion becomes severe. In a simplified equilibrium with comparable competing centres, expansion can continue until their relevant generalised marginal costs are balanced. Literal equality of perimeter congestion is a special case, not a rule for centres with different productivity, amenities, or transport modes.

7. Commercial zoning can transfer the location surplus

Suppose many firms want an accessible subcentre but planning caps office supply. Firms bid against each other for scarce permitted space. Office rents may rise enough to absorb part of the wage or access advantage that initially attracted them.

For example, a location gives a firm a 100,000 annual operating advantage before premises costs. If competition for limited offices raises its annual premises bill by 80,000, only 20,000 remains with the firm. Part of the location advantage has been transferred to property owners.

Compare an annual office rent with an annualised replacement-cost benchmark, or compare capital value with replacement cost. A rent in currency per year cannot be directly compared with a one-off construction bill. A persistent excess over the relevant benchmark can capitalise into a scarcity or entitlement premium, but risk, quality, and other costs must also be considered.

The phrase “net wages remain low” needs care. A suburban nominal wage can be lower while workers’ wages after commuting costs are equalised with other jobs. Zoning does not mechanically fix either wage: workers, employers, housing markets, and commuting conditions can all adjust.

8. A congested CBD and a changing land-use mix

Higher wages needed to attract commuters can reduce what an office occupier can pay for space, holding revenue and productivity fixed. That can reduce the office development residual. If residential use supports a higher feasible land bid, conversion or redevelopment may become attractive.

But an office building need not immediately deteriorate or turn into housing. Existing leases, physical layouts, conversion costs, planning permission, and the option to wait can delay adjustment or make conversion unviable. Compare feasible net values rather than assuming that every congested CBD must shrink.

When travel becomes more expensive, a polycentric region has several potential responses: firms move nearer workers, households move nearer jobs, centres add housing, transport improves, or some work becomes remote. Durable buildings and job matching slow the process. Polycentricity can create opportunities to shorten journeys, but it does not guarantee low total commuting when homes and suitable jobs remain mismatched.

Teach it back: Why can adding housing near a job centre help accessibility without eliminating commuting?

Feynman study lab

Explain it simply

Downtown offers connections, but those connections can be expensive to reach. A suburban workplace may save workers travel and let firms pay less for labour and premises. Firms still cluster because they benefit from shared workers, suppliers, and ideas. Each new cluster eventually develops its own scarcity and congestion. Housing and transport determine how far its advantages can extend.

Teach it back without looking

Explain the 50,000 suburban wage and 54,000 CBD wage example to a beginner. Then explain why firms might gather at one suburban node rather than scatter across every cheap parcel. Finally answer:

If a subcentre saves firms money on wages, why might office rents there rise sharply?

Check your explanation after you try

The savings make the location attractive. If office supply cannot expand, competing firms bid up rents for the available space. Property owners can capture part of the operating advantage. Workers’ nominal wages and rewards after commuting costs must still be distinguished.

Transfer questions

1. The CBD has 40,000 of a region’s 100,000 jobs. Later it has 50,000 of 200,000. What happened to its count and share?
2. A suburban wage is 60,000 with commuting cost 3,000. What CBD wage matches the net reward if commuting there costs 8,000?
3. A firm saves 300,000 in wages and 100,000 in premises expenses but loses 450,000 of annual productivity contribution. Is relocation beneficial under those assumptions?
4. Another increment of subcentre employment creates 90 of benefit and 110 of cost. Should it be added under the simple net-benefit rule?
5. Why can an office-supply cap increase property values without creating an equivalent increase in productive efficiency?

Check the transfer questions

1. Its count rises by 10,000; its share falls from 40% to 25%. Both absolute growth and relative decentralization occur.
2. 65,000: both jobs provide 57,000 after commuting costs.
3. No. The net annual change is -50,000.
4. No. Its net change is -20 in the stated units.
5. Scarcity can transfer an existing location advantage into higher rents and asset values. A higher entitlement value is not itself evidence of additional output or a social welfare gain.

Find the gap and try again

Record where your explanation became vague: counts versus shares, nominal wage versus net reward, access versus distance, or marginal versus total benefits. Rework that example and explain it again without formulas. Change one assumption, such as housing availability or office supply, and predict who gains and who pays. Tomorrow, reconstruct the wage-compensation argument and explain one reason decentralization might fail to shorten commutes.