This is a paid article which your subscription is allowing you to read.

The Common Sense has spent the past few weeks trying to understand the household finances of South African domestic workers. At first, the exercise seemed straightforward. Work out what a domestic worker earns, estimate what they spend each month, and see whether there is anything left. It turned out to be much harder than that.

After interviewing a handful of domestic workers and the employers of domestic workers, it became apparent there is no particularly useful “typical” domestic worker from which to infer the living standards of the broader workforce. Their circumstances vary considerably, and several of those differences can materially affect whether a worker is able to cover their monthly expenses and what standard of living their income can support.

One domestic worker may work one day a week and another five. Some work for a single household while others have several employers. One may live with an employer and face almost no housing or commuting costs, while another rents a room and crosses Johannesburg by taxi several times a week. Some support only themselves, while others support children or relatives. Even the pay varies considerably. The legal minimum is R30.23 an hour, but there are a lot of variances around that.

The Common Sense modelled a broad range of plausible financial circumstances, centered on a domestic worker acting as the sole provider for themselves and their children. The model asked three questions. How much does a domestic worker earn each month, how much do they spend, and is there anything left at the end of the month?

The analysis grouped plausible income and expenditure assumptions to produce possible financial scenarios. The income and expenditure figures that follow are estimates rather than claims that every domestic worker earns or spends exactly these amounts, and readers are welcome to challenge any of the assumptions. As the results later show, however, even pushing those assumptions towards their extremes does not materially change the overall conclusion.

On the income side, we varied hourly pay and the number of days worked. Hourly rates ranged from R28 to R50, combined with between one and five eight-hour working days a week. The table below produces 25 monthly income scenarios for a domestic worker ranging from the extreme low of R970 to the extreme high of R8 665.

The next step was to test those income scenarios against a range of expenditure scenarios. The purpose was to identify the point at which domestic work shifts from being financially viable to no longer covering even the most basic costs of living.

To do this, three expenditure assumptions were developed across the main costs likely to face a domestic worker.

The first was transport to and from work. At the low end, the model assumes a live-in domestic worker who has accommodation on site and therefore spends R0 per month travelling to work. The middle scenario assumes a daily commute costing R50, while the high scenario assumes R100 a day.

The second was accommodation. This ranges from R0 for a live-in domestic worker who is provided accommodation without charge, to around R1 000 a month for a shack, and R2 000 to rent a room in a formal house.

The third was food for the domestic worker and any children. This ranges from R800 a month at the low end, to R1 600 in the middle, and R2 400 at the high end.

Running these assumptions against the income scenarios makes it possible to identify the combinations of pay, working days, and living costs under which domestic work can cover basic expenditure.

Importantly, this is before grant income is considered and before spending on school supplies, medical treatment, clothing, electricity, airtime, and other household costs. It is therefore a deliberately brutal, bare-bones test of what it takes simply to survive.

At the absolute lowest end, monthly expenditure comes to just R800. This assumes free accommodation, no work-related transport costs, and only R800 for food. It is therefore best understood as a baseline rather than a realistic household budget.

Under these exceptionally low assumptions, all 25 income scenarios are viable. The table below shows this, with the income scenarios that meet the R800 monthly expenditure highlighted in green.

The next step is to move to the more realistic middle assumptions. Here, accommodation costs R1 000 a month, food costs R1 600, and transport adds another R1 000, assuming five days of commuting a week at R50 per day. Together, these basic expenses produce a monthly bill of R3 600. The table below shows which income scenarios break even under the middle case assumptions, highlighted in green, and which fall short, highlighted in red.

At this level of expenditure, 13 of the 25 income scenarios remain viable, meaning that the domestic worker earns enough to cover the three basic costs included in the model.

The final step is to move to the high-end expenditure assumptions. Here, accommodation costs R2 000 a month, food costs R2 400, and transport adds another R2 000, assuming five days of commuting a week at R100 per day. Together, these expenses produce a monthly bill of R6 400. The table below shows just how narrow the range of viable outcomes becomes.

Note that only three of the 25 income scenarios remain viable. All three involve full-time employment at wages well above the statutory minimum. Under these expenditure assumptions, not one of the minimum wage scenarios is viable. That is perhaps the most important finding in the analysis: a domestic worker spending just R100 a day getting to and from work, R2 400 a month on food, and R2 000 to rent a room cannot make the numbers work while earning the state-mandated minimum wage, even when working five days a week.

The picture changes only when accommodation is provided free of charge and the need to commute disappears. Yet that comes with an unsettling consequence. A worker whose ability to survive depends on accommodation, transport, and employment all being provided by their employer becomes, in some respects, a vassal to the employer.

That is not an argument critical of the employer. Many ordinary middle-class households are themselves under severe financial pressure from bond repayments, medical insurance, school fees, and other rising costs. Many also spend considerable amounts of their own money, beyond the salary and wages they pay, supporting their domestic worker and broader family.

Nor can the state simply regulate the problem away by continually raising the minimum wage. Push wages beyond what households can afford, and the likely result is fewer jobs and illicit employment that does not have the protections that accompany formal work.

If you want to lay blame, lay it with a state that, in pursuit of its populist ideologies, has engineered an investment rate half that of South Africa’s peers and compounded that failure by presiding over one of the worst education systems in the world. That, not the employers of domestic workers, is what has kept the roughly 860 000 people who work in South Africa’s homes trapped in a degree of vassaldom for the past 30 years.