What Creates Demand for Work?

One job for a hundred carpenters

One job brought around one hundred carpenters to Phillip Street, Sydney, on 17 June 1935. A State Library of New South Wales photograph shows the applicants waiting in coats and hats beneath the stone face of a city building. Some read newspapers; others look towards the doorway. The catalogue supplies the harder context. New South Wales still recorded 85,000 unemployed people, with another 58,000 classed as council relief workers.

They had a trade, and Sydney still had buildings worth repairing. What the city could offer that morning was one paying job. Ninety-nine of the carpenters would leave without the work they had come to find.

That queue exposes a weakness in every promise to prepare people for the jobs of the future. Capability and training can help someone do the work. They cannot make a customer buy, persuade an employer to hire or cause a government to fund a service.

Demand for paid work arises when someone directs resources towards an output and a producer responds by using human labour. A household may buy the output. A business may purchase inputs or invest in the hope of a future market. Governments and other institutions can fund services, commission work and build infrastructure. Trade allows buyers elsewhere to support the production.

The route is never automatic. A pressing need may come with no purchasing power. Spending can go to an import, or a firm can meet it by raising prices, drawing down inventory or automating a task. Revenue can rise while employment hours stand still; hiring can grow without wages following. Meanwhile, households perform necessary work every day without a sale, an employer or a payslip.

Getting from need to a wage therefore takes a chain of decisions. Each link asks its own question.

Why employers rarely want labour for its own sake

Alfred Marshall explained the basic mechanism in 1890 by starting with a house. Building one creates what he called a joint demand for bricks, stone, wood, builders' organisation and the labour of several trades. The buyer does not want plasterers' hours in isolation. The demand for them is derived from demand for the finished house.

The same logic applies to a school lesson, an evening meal, a legal opinion, a freight delivery or a software service. An employer pays for hours because those hours, combined with knowledge, tools, materials and organisation, are expected to produce something that can be sold, funded or used.

The customer's purchase can travel through a network of work they never see. A chair connects the buyer with forestry, design, machinery maintenance, transport, warehousing, retail and waste recovery. A bridge contract can reach surveyors, engineers, concrete plants, equipment hire and traffic management. Funding home care supports the care worker, but also scheduling, training, transport and administration.

This also explains why work can disappear from one place even though the underlying want survives. The household still buys a chair, but it is made elsewhere. The client still needs an account reconciled, but software now performs some of the tasks. A government recognises the need for care, yet sets a budget too small to pay for all the required hours.

Marshall's house gives us a durable first answer: demand for work is usually borrowed from demand for an output. Questions about whose demand counts, where the output is produced and who shares in the proceeds still remain.

A need is only the first link

The word demand makes a desire sound more powerful than it may be.

Adam Smith drew this distinction in The Wealth of Nations. His “effectual demanders” were buyers willing to pay the price required to bring a commodity to market. Because they could complete the purchase, their wants could organise production. Someone else might want the same commodity just as keenly yet have no effect on the quantity offered.

John Maynard Keynes used effective demand to examine a different, economy-wide problem. Writing during the Great Depression, he rejected the expectation that an economy would reliably return itself to full employment. Businesses make production and hiring decisions against the proceeds they expect. When total spending is too weak, the economy can settle with people and equipment idle.

Smith and Keynes were working at different levels. Both give us a reason to keep want, expenditure and employment separate.

The table is an analytical sequence, not a universal timetable. A public hospital may receive its budget before delivering services. A self-employed worker may commit labour before knowing whether a customer will pay, while a household growing food for its own use has no revenue stage at all. Keeping the stages separate prevents a claim at one level from quietly carrying a conclusion about the next.

Labour demand · Route to paid work

How does a need become paid work?

Need is only the beginning. Funding, production choices, capacity, location, technology and bargaining decide whether it becomes local employment and whose income grows.

Need or desire→Effective demand→Output→Revenue or funding→Employment→Wages and income

Case 1Household buys a chairShow detailsHide details

Resource committed: Household purchasing power

Possible route: A producer receives an order and decides where, how and by whom the chair will be made.

Diversion: The order may be imported, filled from inventory or produced with less labour.

Case 2Firm invests in equipmentShow detailsHide details

Resource committed: Business investment

Possible route: Expected future sales support equipment production, installation and complementary work.

Diversion: Weak expectations, financing limits or automation may change the tasks and headcount.

Case 3Government funds home careShow detailsHide details

Resource committed: Public budget and service mandate

Possible route: Collective funding turns care need into commissioned services, roles and paid hours.

Diversion: Non-provision, capacity limits, low wages or contracting choices can break the route.

Case 4Community meets a need directlyShow detailsHide details

Resource committed: Time, reciprocity and unpaid effort

Possible route: Care or community work occurs without a market sale or wage relationship.

Diversion: The work remains socially necessary even though it never becomes paid employment.

Where the route can change

No purchasing powerPublic non-provisionImportsInsufficient capacityHigher pricesInventoryAutomationNew tasksUnpaid workUnequal bargaining power

A six-stage chain moves from need or desire through effective demand, output, revenue or funding, employment, and wages or income. Four disclosure panels trace a household buying a chair, a firm investing in equipment, a government funding home care and a community meeting a need through unpaid work. Diversion tags show no purchasing power, public non-provision, imports, insufficient capacity, prices, inventory, automation, new tasks, unpaid work and unequal bargaining power. The figure produces no multiplier, job forecast or wage estimate.

Who gets to turn need into spending?

National accounts provide a disciplined map of where money moves. The United Nations' 2025 System of National Accounts traces how production generates income and how households, businesses, governments and foreign buyers direct that income towards consumption, saving and investment. It can measure a transaction precisely. Deciding which needs deserve priority lies outside its reach.

For households, purchasing power sets the limit

A household pays for food, rent, transport, entertainment and care from its income, savings, credit and transfers. That expenditure becomes someone else's income. How purchasing power is distributed helps determine which kinds of paid work receive support.

Two people can need the same service and have very different power to create market demand for it. A high-income household might buy private tutoring or home care. A household with less income may rely on family, do the work unpaid, receive a publicly funded service or simply go without. The market records the purchase and says nothing about the unmet need beside it.

Wages feed back into the same process because employment income becomes purchasing power. Cutting a wage may lower one firm's costs while leaving the worker with less to spend elsewhere. The distribution of income belongs inside the demand story.

Businesses invest in the future they expect

Firms generate demand for other firms when they buy energy, components, software, freight, professional services and equipment. National accounts avoid counting every intermediate transaction as new value, but the production network itself is real. A final order may pass through many workplaces before it reaches the customer.

Investment works in two periods at once. Building a factory, installing a solar farm or developing a software system requires labour now. Once finished, the asset can expand future productive capacity, alter people's tasks and change the amount of labour needed for each unit of output.

Technical possibility alone does not produce an investment. A firm can know that something is feasible and still hesitate because of uncertain customers, regulation, finance, supporting infrastructure or reliability. “This can be built” is a much weaker statement than “someone will commit the resources to build and use it”.

Public budgets can put idle capacity to use

Schools, courts, roads, emergency services and public health need not depend on an individual's ability to pay at the point of use. Governments can tax, borrow, regulate, employ, procure and transfer resources. These choices turn some shared needs into budgets, and those budgets into paid work.

Australia confronted the scale of this power under brutal conditions. The available historical series shows unemployment reaching 29.0 per cent in 1932. The ABS cautions that these pre-1960 estimates, drawn largely from trade-union reports, cannot be compared directly with the modern Labour Force Survey. Whatever the precise measure, capable people and productive equipment stood idle while families lacked essentials.

Wartime mobilisation then drove unemployment to an extraordinary low. It did so in the service of war, with all the destruction that entailed, but policymakers carried an economic lesson into the peace. On 30 May 1945, the Commonwealth tabled Full Employment in Australia. The White Paper argued that full employment depended on total expenditure providing a market for what Australians could produce with the equipment and materials available. Public expenditure would need to help close the gap when private expenditure fell short. The Reserve Bank's history of Keynesian ideas in Australia describes the paper as the high point of a debate transformed by the Depression and wartime administration.

Public spending still has limits and consequences. Money directed into an economy already at capacity can raise prices. A programme may purchase imported equipment and generate little domestic production. A badly chosen project can employ people while wasting resources or causing harm. Debt, taxation, foreign exchange and political consent all matter. The 1945 claim was narrower. Useful work alone leaves idle labour idle; public institutions can alter the expenditure that employers see.

Trade changes where the work occurs

Exports bring demand into domestic industries from households, firms and governments elsewhere. Imports direct domestic expenditure towards production abroad. Supply chains divide both flows among several countries.

This boundary matters whenever a headline leaps from sales to “jobs created”. A surge in domestic purchases may support local retail and logistics while most manufacturing takes place abroad. An export contract, by contrast, can sustain specialised work in a region whose local customers could never support it. Trade changes the location of derived demand, but tells us little by itself about the security or quality of the work.

The same spending can produce very different employment

Money enters a production system with its own limits and choices.

The ILO's 2022 review of demand-side employment policy makes productive capacity explicit. Additional expenditure can support more output and employment when equipment, infrastructure, intermediate inputs and suitable labour are available. If one of those inputs is scarce, the result may instead be a higher price, a longer queue, an import or a delay. Sustained expansion therefore needs investment that increases capacity as well as spending that uses it.

Productivity changes how many hours are attached to a given quantity of output. The ABS defines labour productivity as output per unit of labour input and warns that it reflects capital, technology, materials and organisation as well as hours. A bakery that doubles output with the same total hours can grow its revenue and production faster than its employment. Yet if lower costs or prices bring enough extra demand, the bakery's total hours may still rise. The balance depends on the size of both changes.

Employers can also adjust working time before adding permanent jobs. A temporary order might be met through overtime, extra shifts, casual workers or a contract with another firm. The word “jobs” hides these choices. For a self-employed person, stronger demand may appear as longer hours or more reliable bookings rather than a new headcount.

Technology introduces forces that can pull employment in opposite directions. Daron Acemoglu and Pascual Restrepo describe automation's displacement effect, in which capital takes over tasks previously allocated to labour. They also identify a reinstatement effect when new tasks emerge where labour has a comparative advantage. Productivity and scale may expand output even as displacement reduces the labour required for existing tasks. A technical demonstration alone cannot tell us which force will dominate in a particular occupation, place or period.

Finally, revenue has many claims upon it: suppliers, taxes, interest, rent, labour and returns to owners. Wage-setting institutions influence the division. The ILO's review of collective bargaining documents how agreements can establish floors, pay structures and ways of sharing productivity gains. Employer concentration, minimum wages, bargaining coverage and the alternatives open to workers shape the result as well. Strong demand may improve a worker's bargaining position, but no fixed wage arrives inside it.

Some needs create work without creating employment

The market chain leaves a great deal of work outside its frame.

The International Labour Organization defines employment as work for pay or profit. Its wider framework also recognises own-use production, unpaid trainee work, volunteer work and other activities as work. A family growing food for its own consumption is working. The same is true of someone cooking, cleaning or caring for a dependant in their household, even though no transaction appears.

Care shows how much the institution matters. An unpaid family member, a directly employed domestic worker, a private provider, a charity or a publicly funded service might all meet the same need. The activity can look similar while its work relationship, income, accountability, access and statistical treatment change.

A purely market account therefore begins too late. Without purchasing power or public provision, the need remains. It may be met through unpaid work, postponed until later, or left to produce exhaustion and deprivation. The ILO's care-work synthesis places paid and unpaid care in one system because the supply of labour to the rest of the economy depends on what households and care institutions do.

Subsistence production offers a similar correction. A household may farm, fish, build, repair and preserve goods for its own use. Access to a market can bring income and greater specialisation, but it can also expose the household to price risk or displace production that sustained it outside cash exchange. Moving into employment changes the work relationship and the source of income. It does not mark the beginning of productive activity.

Demand redraws the map of work

Over time, shifts in demand move labour among industries. Household purchases change with income; governments expand or contract services; investment adds new capacity. Trade connects local producers with distant buyers, while technology changes prices, products and tasks.

Australia's long historical series shows the scale of this redirection. Primary industries and manufacturing together accounted for 52 per cent of employment in 1910–11 and 19 per cent in 1999, while wholesale and retail trade, finance, education, health and other services grew. The figures span many changes in classification, so they cannot tell us how much any single cause contributed. They do show that “how much work exists?” and “which work is growing?” are separate questions.

A new capability opens a possibility. Adoption and demand determine whether it becomes a direction of work. Mobile networks generated work in devices, towers, software, repair, content and services because households, businesses and states committed resources to using them. Earlier tasks shrank, some production took place far from the users, and some of the value accumulated in firms with relatively few employees.

Care follows another route. Ageing, disability, childhood and illness create need, and demographic change can enlarge it. How much becomes paid employment, and under what conditions, depends on public budgets, household incomes, migration rules, service design and wages. A forecast based on population alone skips every institution in between.

Demand runs through the library's six threads. It can deepen specialisation, attach status to expanding occupations, limit the work people can choose and change who has access to a livelihood. Its direction is social as well as technical: purchasing power, budgets and institutional rules decide which possibilities command labour.

Demand is not a vote on worth

Demand can look like a record of what society values. People spend, firms respond, and labour appears to flow towards useful activity.

Markets do transmit information. Prices and orders coordinate production across distances no planner could inspect transaction by transaction. Public budgets can give collective priorities material force, and investment commits resources to an expected future. These are powerful ways of directing work.

Yet the signals are unequal and incomplete. A dollar has the same numerical value whoever spends it, but people possess vastly different numbers of dollars. A luxury renovation can create clear market demand while essential care remains unpaid. Advertising shapes desire, monopolies can restrict supply, and public institutions may neglect a need or fund harmful work. Coercive demand is still demand.

Demand answers a bounded question: who can currently command resources for what? It is a poor ranking of human importance, effort or justice.

This boundary changes how we speak about careers. An interest or aptitude result tells us something about a person. A vacancy tells us something about an employer or institution. Growth in an occupation depends on expected demand, production and workforce organisation over time. A career-fit question needs evidence at all three levels, along with a route the person can actually enter and sustain.

What the chain lets us say

That human need begins the story, but cannot guarantee paid work.

That labour demand usually comes from an output someone can buy, fund or require.

That the path from spending to employment passes through productive capacity, location, technology, tasks and working arrangements.

That revenue and wages are different outcomes, connected by institutions and bargaining rather than arithmetic.

That unpaid and subsistence work continue to meet needs beyond the market's field of view.

The carpenters in Phillip Street had the willingness and the trade. What ninety-nine of them lacked that morning was a buyer for their labour. Ten years later, Australia's Full Employment White Paper treated that absence as a public problem, not a private defect.

The next question is what happens inside a job when new technology changes the tasks from which demand for labour is made.

Notes on the evidence

The economic chain in this article is Guidebeam's synthesis of primary economic texts, modern national accounting and labour-market research. Marshall's derived-demand example concerns the factors used to produce an output; Keynes's effective-demand argument concerns employment across an economy. The article connects them without treating them as the same theory. Its 1930s figures come from the historical series and caveats published by the ABS. Because the ILO demand-side review focuses on low- and middle-income countries, its discussion of capacity, imports and distribution is used to explain mechanisms, not to supply universal numerical estimates. Technology effects are presented through a task framework rather than an occupational forecast.

Sources and further reading

• State Library of New South Wales. “Unemployed workers in Phillip Street”. 17 June 1935.
• Australian Bureau of Statistics. “A century of change in the Australian labour market”. 2001.
• Smith, Adam. An Inquiry into the Nature and Causes of the Wealth of Nations. 1776.
• Marshall, Alfred. Principles of Economics, Book V, ch. VI. 1890.
• Keynes, John Maynard. The General Theory of Employment, Interest and Money, ch. 3. 1936.
• Commonwealth of Australia. Full Employment in Australia. 1945.
• Reserve Bank of Australia. “Keynes and Australia”. 2000.
• Aboobaker, Adam and Jo Michell. Demand-side policies for employment promotion in low- and middle-income countries. International Labour Organization, 2022.
• United Nations Statistics Division. System of National Accounts 2025. 2025.
• International Labour Organization. Resolution concerning statistics of work, employment and labour underutilization. 2013.
• International Labour Organization. Care Work and Care Jobs for the Future of Decent Work. 2018.
• Australian Bureau of Statistics. “Labour productivity”. 2026.
• Acemoglu, Daron and Pascual Restrepo. “Automation and New Tasks: How Technology Displaces and Reinstates Labor”. 2019.
• International Labour Organization. A Review of Wage Setting through Collective Bargaining. 2023.
• International Labour Organization and OECD. “Policy measures to address inequalities and increase the labour income share”. 2025.