Ten Thousand Residents, Not Ten Thousand Jobs

Part 1 of 10 in the 10,000 People Economy series.

A community of 10,000 residents does not give us 10,000 workers. It gives us a society.

It has children, older people, full-time workers, part-time workers, unpaid carers, students, people with disabilities, people looking for work and people who have stopped looking because the labour market has given them little reason to continue.

This is the starting point of the 10,000 People Economy. It is a model for a resident population, with all the work, care, consumption and institutional capacity that population requires. It is not a promise of 10,000 wage jobs.

The distinction sounds obvious. Job-creation programmes ignore it surprisingly often. A large number is announced. Occupations are listed. Training places are counted. The total becomes the employment claim before anyone has shown who will buy the output, how the wages will be funded, whether the enterprise has working capital, or how many people will perform real work for a meaningful period.

The problem is not ambition. South Africa needs employment ambition. The problem is that the wrong unit of account turns ambition into weak evidence.

Start With the Society

I find it more useful to think of an economy as a system than as a collection of jobs. People produce, consume, learn, care, invest and trade within institutions that have to be maintained. The jobs emerge from that system; they are not a substitute for it.

Starting with a resident population forces the full system into view. The model must account for food, housing, care, education, utilities, safety, maintenance, transport, finance, culture and the productive base that pays for them. It must also account for the people who need those services without assuming that every one of them should hold a standard full-time post.

The current model begins with 10,000 residents and applies a provisional productive-placement ratio of 64.6 per cent. This produces a planning pool of 6,460 people.

That number has a specific meaning. It is the scale of the productive-inclusion problem under the present assumption. It is not a report that 6,460 people have jobs.

The ratio must eventually be replaced by the real age structure, care burden, participation pattern, disability profile and migration behaviour of a selected community. Until that work is done, 6,460 is a design coefficient. It is useful because it makes the obligation concrete. It is dangerous if it is presented as an outcome.

A Person Is Not a Position

Employment reporting becomes unreliable when different quantities are merged.

A person may hold two part-time positions. Two people may share one position. A learner may spend part of the week in productive work and part in structured learning. A funded post may be vacant. A person may be enrolled in a programme but never perform work. Ten short activities may be reported as ten jobs even when their combined hours equal one full-time role.

A serious labour account must therefore report at least five things separately:

• unique people placed;
• funded positions;
• actual hours worked;
• full-time equivalents; and
• people whose work is supported by verified demand and finance.

These measures answer different questions. People placed tells us about reach. Positions tell us about organisational capacity. Hours and full-time equivalents tell us how much labour was used. Demand-backed employment tells us whether the work has an economic or authorised public-purpose foundation.

The public should not have to guess which one sits behind the headline.

A Job List Is Only an Allocation

The 10,000 People Economy currently allocates its 6,460 planning places across eleven sectors. The portfolio includes food, construction and maintenance, care, education, trade and logistics, manufacturing and repair, professional services, utilities, governance and finance, culture and tourism, and research and export development.

This is useful. It forces the model to show what people might do. It also exposes imbalances. A community cannot place everyone in retail and call itself productive. It needs the people and systems that maintain water, buildings, health, skills, safety, equipment and external trade.

But an occupation list is not an economy.

Naming 760 food and agro-processing placements does not show what will be grown or processed, who will buy it, what they will pay, how often they will pay, which inputs must be imported, or how much cash is needed before the first collection. The same is true for every other sector.

The model now forces this problem into the open. At full planning scale, its current demand account requires about R2.011 billion in annual sales, R927.4 million in loaded payroll and R385.2 million in working capital per community. Under a combined customer-volume and collection shock, the supported placement number falls from 6,460 to about 5,791.

These are not forecasts. They are tests of what the employment promise requires.

This is the practical weakness in an occupation-led model: the job count changes when customers buy less or pay later, while the list of occupations remains exactly the same.

Work Must Have a Purpose and a Funding Route

Not every productive role needs a conventional private customer.

Care, maintenance, safety, early childhood development and environmental work can produce public value that markets do not price well. That does not make the work imaginary. It means the customer or obligation must be stated differently.

A public-purpose role still needs an authorised service, a budget or contract, tools, supervision, standards and evidence that the service was delivered. Calling work socially valuable is not a funding model. Calling it non-commercial is not a reason to exclude it from the economy.

The same discipline applies to apprenticeships. A course enrolment is not employment. A credible apprenticeship requires paid productive work, a real workplace rotation, a competent mentor, protected learning time, an assessment standard and a route into more capable work.

The model should be open to different institutional forms and strict about evidence. Private companies, cooperatives, public providers and nonprofit organisations may all create qualifying work. None should be treated as viable because of its label.

The Employment Claim Has a Ceiling

The most important change in the model is a simple constraint.

The number of jobs that can be claimed is the lowest of four numbers:

• the planning allocation;
• the number supported by accepted customer or public-purpose demand;
• the number supported by revenue; and
• the number supported by working capital.

If the plan contains 6,460 places but verified demand supports 2,000, the employment claim is not 6,460. If customers exist but collections arrive too late to fund payroll, the claim must fall to the cash-supported level. If revenue supports the role but the work is unsafe or has no supervision, it fails a separate quality test.

The current independently accepted demand-backed employment number in the model is zero. No site-specific customer, price, volume, collection, payroll and working-capital schedule has yet passed the evidence gate.

That zero is not embarrassing. It is honest.

It tells us that the project has a model and a research programme, but does not yet have an implemented community whose employment outcomes can be reported. Many development proposals would be stronger if they were willing to say the same thing.

Measure the Conversion Chain

The operating question is not how many jobs were announced. It is how many people moved through a complete chain:

resident -> eligible participant -> placement -> verified output -> customer or public purpose -> collected revenue or authorised funding -> sustained work -> progression

Every arrow can fail.

The resident register can be wrong. Eligibility rules can exclude carers. A placement can exist without tools. Output can fail the required standard. A customer can order and not pay. Funding can arrive after payroll is due. A person can remain in low-quality work without gaining capability or income.

This chain gives managers something useful to manage. It also gives communities, funders and researchers a way to distinguish a broken promise from a specific broken conversion.

The Better Promise

The 10,000 People Economy should make a precise public claim:

We are designing an economy for 10,000 residents. Under the current assumptions, 6,460 people require a productive placement. We will claim only the number of qualifying roles supported by accepted demand, revenue, working capital and public-purpose authority.

That sentence does less political work than "10,000 jobs". It does far more economic work.

I am not arguing for a smaller employment ambition. South Africa needs ambition, but it also needs employment accounts strong enough to survive contact with customers, cash flow and actual work. Measurement should make the promise buildable, testable and ultimately worth believing.

Sources

• 10,000 People Economy Project. (2026). 10,000 People Economy Model Specification v1.
• Kim, K. (2011). Ex-ante evaluation of a targeted job program: Hypothetical integration in a social accounting matrix of South Africa. Economic Modelling, 28(6), 2683-2690. https://doi.org/10.1016/j.econmod.2011.08.008
• National Planning Commission. (2012). National Development Plan 2030: Our Future – Make It Work.
• Statistics South Africa. (2026). Quarterly Labour Force Survey, Quarter 1: 2026.
• Tcherneva, P. R. (2018). The Job Guarantee: Design, Jobs, and Implementation. Levy Economics Institute Working Paper No. 902.

---

Next: A Job List Is Not an Economy