Data on household consumption expenditure in South Africa are certainly headline grabbing. According to Statistics South Africa, households spend more on clothes than they do on education, and more on alcohol than on out-of-pocket healthcare.
Expenditure patterns in lower income households are particularly curious, especially to those of us with strongly paternalistic middle class tendencies; households that earn less than R3 500 a month spend more on clothes than they do on education and healthcare combined.
The data is fascinating, but more so are the drivers that shape expenditure patterns. Of course we are never really sure why other people spend the way they do, but we can point to some key factors.
The first factor is the state. State delivery and state failure have a material impact on the way households spend money. As a starting point, the government’s extensive grant system provides more than R9.4 billion in cash to roughly 44 percent of households every month. But more than this, the government’s provision of free products and services to poor households is also significant. For instance, the housing subsidy programme provides free houses to beneficiary households that earn less than R3 500 a month. The impact of the programme on housing patterns is astounding. According to census data, over 3.5 million more households lived in proper housing in 2011 compared with 2001, a good proxy for the addition to the housing stock. Over that time Stats SA reports that the private sector built about 660 000 units. The balance was provided mostly by the state.
This is significant, not least because of the impact of housing form on household spending priorities. Put simply, the home you live in drives much of the way you spend your money. Its location determines your transport expenditure, and formal housing typically comes with services such as electricity and running water. Together, transport and housing-related expenditure account for almost 40 percent of household expenditure. Beyond that, the size of your house determines what you can fill it with – again shaping purchasing decisions.
Beyond housing, the provision of free services such as education and healthcare to lower income households clearly affects consumption patterns, as does the provision of free municipal services. Free essential services frees up income for other purchases, not all of them necessities.
But there are problems with “free”. In the first instance, as they say, some free services can be worth every cent you pay for them. Indeed, while we have excelled at increasing access to education and healthcare, we have been less successful at maintaining quality. While delivery of free services means that households do not have to allocate expenditure to these categories, poor quality encourages those who can to buy up, often at significant cost, into the private sector.
The second problem with free is that it is not, in fact, free for all. A range of means tests and income thresholds are in place to determine who qualifies for free services. Those who earn above these thresholds must pay. But the distorting effects of subsidies discourage private sector providers, who fear they cannot compete with free, and who deliver comparatively higher spec products and services to differentiate their offerings from those provided by the state. In the case of housing, the cheapest newly built units developed by the private sector in major metros appear to cost upwards of R350 000. Even households earning more than R12 000 a month would struggle to afford these units financed by a 20-year mortgage. We therefore find the existence of a so-called gap market for housing comprising upwards of 3.5 million households that are not quite poor enough to qualify for a free house provided by the state, but at the same time are too poor to afford to purchase what the market can provide.
The same is true in healthcare. Assuming households can allocate 10 percent of their expenditure to funding medical expenses, entry level medical schemes are simply not affordable for 2.6 million households that earn above the R4 167 income threshold to qualify for free hospital access in the public sector. Likewise in education, recent research published by the CDE highlights that even the most affordable private schools cost about R10 000 a year per child, excluding the cost of transport, school uniforms and books. Even if these households want to spend on housing, healthcare and education, there appears to be nothing affordable for them to buy.
A third problem with free is that there is, in reality, no such thing as a free lunch. Someone has to pay. The burden falls on a relatively small base of taxpayers and rate payers. Municipalities have to increase rates and service charges levied on households that can pay to fund a growing number of households who do not. A recent report published by the SA Cities Network found that households living in properties valued at R1 million in major metros were paying on average 37 percent more in real terms for a package of municipal services in 2012 compared to 2009. Clearly, for these households the need to divert cash to pay for higher-cost necessities has a material impact on discretionary expenditure.
Aside from free, another critical factor shaping expenditure patterns is the availability, or otherwise, of credit. Here too the data is headline grabbing. In 2008, roughly half of all new loans by value were mortgages. Last year this had declined to about a quarter, with other secured loans (principally vehicle finance) accounting for one third of new credit granted. Unsecured loans, which accounted for 10 percent of credit granted in 2008, were neck and neck with mortgages last year. The decline of mortgage lending has occurred as we continue to see the growth in an emerging middle class. This has entirely predictable and unfortunate consequences for the ability of the emerging middle class to accumulate leverage-able housing assets. It is a story that is deserving of far more attention from policymakers and bankers than it appears to receive. For households that have no hope of attaining their dream home funded by a mortgage, the dream might shift to one about a car.
Lower down the income pyramid, credit provided by clothing retailers has a significant impact on the ability of households to spend money on clothing. According to credit bureau data, there are almost 11 million South Africans who have one or more open clothing accounts. The data indicates that an astounding 40 percent of these borrowers are in arrears on at least one clothing account. While the business model based on high merchandise margins, high costs of credit and high default rates is not to everyone’s liking, clothing retailers typically provide an entry point into the credit market for many consumers.
We can try to distil purchasing decisions down to key drivers that we can control, but in reality spending is also driven by unobservable, psychological factors. While these can be hard to quantify, the qualitative research is telling. In lower income markets in particular, respondents refer often to dignity, and highlight the shame of poverty. In post-apartheid South Africa, the word “dignity” has powerful connotations. To quote one respondent, “Even if you are struggling financially, you try by all means to save money to buy that expensive shirt, T-shirt, trousers, shoes because you want to look good. It’s a trend in the township. It covers you so that a stranger can’t tell you are poor.”
The implications are powerful: for the private sector looking to serve new markets, and for policymakers whose decisions – about what to subsidise, who to subsidise and how – shape these markets, often inadvertently. In our society, choices are curtailed, more than just shaped, by factors beyond households’ control.
Illana Melzer – Co-founder of Eighty20