The Zimbabwean and South African examples demonstrate the hard limits set by typical urban incomes on housing solutions if those solutions are bound by the norms of free-market pricing and financing. When these limits are recognised, the logical necessity of thinking about what determines incomes as the first building block of understanding housing issues becomes apparent. This is a complex topic but the essential element in relation to housing affordability is that urban people are caught between two sets of markets. Labour markets determine what people are paid and housing markets determine what housing costs. Crudely speaking, across the cities of world, the pay rates for millions of jobs or types of work are simply too low for the workers involved to pay enough for the types of housing that can be provided by private-sector markets that comply with regulations and have legal tenure.
The existence of such hard limits for individual families set by their incomes and the requirement to meet certain basic needs in order to survive are fully understood by poverty analysts working in the GS when they are working on food security. Indeed, this is the starting point. Obviously people have to eat, and they should eat every day. Food costs money. It is fairly straightforward to calculate the cost of a basic food basket needed for adequate nutrition and compare this to incomes. Much of this sort of food security work is done in rural areas in the GS where most people are often poor or very poor. Most national poverty datasets show that the incidence of poverty is much lower in urban than in rural areas in GS countries. However, the way in which such data are analysed is problematic. First, fairly obviously, most rich people tend to live in cities and so do most of the ‘real’11 middle classes. The number of middle-class people in countries such as Brazil, India and China is now very significant. These skew the figures. However, if the typical livelihoods, real disposable incomes after all necessary urban living costs, and lifestyles of most people living in urban low-income settlements are compared with those in rural areas, the seemingly large gap in ‘poverty’ narrows. Indeed, the underestimation of urban poverty levels in the GS is a key data problem for really understanding cities in these regions.12 As anyone who has worked on livelihoods in urban Africa will attest, there are very many households living day to day and hand to mouth, who spend most of their money on food. When things are tough, meals are missed. As already described above, this is a situation in which an extra few per cent of income spent on rent will cause real reductions in household welfare. This is poverty. As will be discussed in Chapter 4, if you cannot significantly increase your income in such a situation, then you may have to house yourself informally because that means your housing expenditure plummets. This ‘magically’ increases your disposable income and, depending on the local politics of urban poverty datum line setting, may make you apparently non-poor.
But suppose that urban poverty datum lines were set differently and in accordance with the frequently asserted view that capitalist property norms, functioning urban land markets and ‘legal tenure’ are essential to the functioning of efficient, productive cities. Let us say that the basic urban household income required for people to be classified as ‘non-poor’ in any city across the world must include an allowance for, say, current market rents for the cheapest types of housing but in a legal, planned dwelling with secure tenure and legal connections to water and electricity. (Let it be remembered that current definitions used by UN-Habitat label any other type of urban housing as ‘slums’.) In essence, this is the situation in most of the GN. The immediate political argument that ensues is over space – how many rooms per person in the household are being factored in. Yet, however that were worked out, the basic principle would reshape our understandings of the incidence of urban poverty, which would shoot up in most of the cities of the GS. The corollary, of course, is that poverty rates in the GN would, apparently but absurdly, decrease were ‘cheap’ illegal slums to reappear at scale in the world’s wealthiest cities, and if their inadequacy were not factored in to poverty measurement.
This approach has been used for years by the Jesuit Centre for Theological Reflection (JCTR), an organisation that, among other things, monitors prices and household budgets across the city of Lusaka in Zambia. At first they published food budgets based on the food needed for a family of six to keep them healthy. The diet is entirely in accordance with local norms, with mealie meal (maize) being by far the largest item. In 2002, the JCTR updated its budgets with the necessary other costs incurred in town, of which rent is by far the most significant. Information on the ‘basic needs’ basket is regularly used by trade unions and other groups when lobbying for wage increases. In the first quarter of 2002, the basic food basket cost 324,510 Zambian kwacha per month. The monthly ‘take-home pay’ of a selection of formal-sector workers could not cover this cost. For example, the lowest-paid secretaries, nurses and police officers at the time earned between K120,000 and K270,000. Even those at the top of their scales earned only K300,000 to K370,000. Primary school teachers could earn between K280,000 and K309,000, so even they could not afford the basic urban food basket. But this was only the start of the problem, because once other basic urban costs, including housing, were added, the costs rose to K823,510. Even secondary school teachers at the top of their scale earned only 60% of this amount. While this was startling enough, these budget figures relate only to those in the urban formal sector. The great majority of the urban population works in the informal sector where incomes are generally even lower, so the situation was actually much worse. The lowest-paid security guards recorded by the JCTR – on K40,000 per month in 2002 – were better surrogates for many in this sector.13 Surveys in markets in 2004, when the basic needs basket costs had increased by a further 25%, found that in Lusaka’s largest market, Soweto, 64% of customers were earning under K100,000 per month. If the small amount of meat, eggs and dry fish were removed from the 2002 budget, it went down to K278,110 – still far beyond the means of most and leaving a diet seriously deficient in protein. These depictions of poverty in the city are completely rejected by the government, however, which produces its own figures that show far less urban poverty. Yet a key reason for this is that the housing element of the government’s urban household budget is based on costs in informal settlements. The JCTR, on the other hand, used the costs of renting modest accommodation in formal, legal, ‘decent’ housing. This was because, while it was true that the government’s approach was using more typical housing costs, these were characterised by conditions that any multifaceted approach to understanding poverty would factor in to the measurement of who is poor and who is not. Such multifaceted approaches are strongly promoted by all poverty experts and agencies across the world. And the Zambian government is not above knocking down the occasional informal settlement because it is ‘illegal’.14 In other words, the JCTR’s approach was that being unable to afford anything except very inadequate and insecure housing is incommensurate with being ‘not poor’.
In cities across the GS, the truth of hard limits on housing expenditure can be seen in other ways. In Harare, the average rent for one room in Kuwadzana, the city’s largest low-income housing project, remained at around 10% of the average income of a formal-sector wage worker in both 1994 and 2001. Rents were seen as burdensome and people complained about them rising but in real terms the share of income spent on rent was not changing much. The proportion spent on food did rise after structural adjustment started in the 1990s. But despite a shortage of housing, rents only kept pace with incomes – there was a cap effect on rents set by what people could afford. A similar situation was found by the urban development specialist Philip Amis in Nairobi in the 1990s,15 and there is evidence of this effect operating in the low-income rental sector in Buenos Aires and in Mexican cities, according to Professor Alan Gilbert of the Geography Department of University College London, who has long championed the importance of the rental sector in the cities of the GS.16 As Gilbert says, ‘however grasping landlords may be, poor people can only be exploited to a certain point’.17
But so what, it might be asked? The examples so far come from the GS and we all know that countries there are poorer than the societies of the GN. So perhaps it is to be expected that many urban residents there will not be able to afford ‘decent’ housing and are too poor to be a viable market opportunity for formal housing finance