Broken Cities. Deborah Potts

as roads, electricity, water, shops, schools and clinics. The vast majority of the city’s low-income residents lived in such settlements and only a small proportion in areas with illegal tenure. Squatting was rare, as the new independent African government maintained a vigilant battle against housing informality, as had its white minority and racist predecessor. These five areas represented very different phases of housing in the city. The first three were built by white minority governments to house African families. Africans were not allowed to own urban land or housing and the general policy was to discourage permanent settlement by Africans in the towns of Southern Rhodesia (as colonial Zimbabwe was called). Nonetheless, a constantly oscillating migrant labour force had its downsides as the urban economies matured since it restricted the availability of skilled labour. Thus, a limited number of family houses – for rent and not ownership – were built in ‘townships’ such as Highfields to allow an element of more skilled and permanent urban labour to emerge. At independence, the ownership of most of these small, one-storey houses was transferred to their main tenants. The history of the way in which this housing was costed to its main residents both before and after independence bore no relation to the forces of supply and demand; as such, it is discussed in Chapter 5 as an example of mass public housing provision. Chitungwiza and Kuwadzana were very different, both being site-and-service settlements where people were allocated land (called plots or sites in African housing parlance) and were then supposed to construct their own dwellings. These could then be connected to the water and electricity supply provided in the settlements; these supplies were metered, and households defaulting on their bills could be disconnected. Chitungwiza was begun by the white state in the late 1970s, just before independence in 1980. Kuwadzana (meaning ‘cooperation’) was begun shortly after independence with support from USAID. All the homeowners in these areas at this point had been significantly subsidised by the government or by aid programmes. Profit-oriented market forces had had little to do with their housing.

      

      However, most interviewees in the survey were lodgers renting rooms from these homeowners. Their rents were set by market forces operating across the city and were completely unregulated, even though rent control legislation was on the books. The rents for one or two rooms were generally very similar, with some variation between settlements according to distance from the main employment areas of downtown Harare and its industrial zone. Many migrants were young, unmarried and lived alone in one room or shared with a friend. Married migrants usually lived with most of their immediate family in town for at least part of the year, in contrast to the situation before independence, but there was a significant proportion whose spouse and/or children were still living in the rural areas. Most of those divided from their families in this way aspired to bring their families to town and live together, but, as they often pointed out, the problem was money. They could not afford to support them in town, largely because their outlay on housing would increase or they would all have to live in one room together. This pattern of an entire household living in one room is common enough in the rental sector in African cities, but it is obviously neither desired nor desirable, being a response to the housing dilemma at the centre of this book.

      When asked about their plans and aspirations for the future, it turned out (rather unexpectedly) that most of the migrants expected to leave the city eventually and return to rural areas, usually their birthplace, in a pattern similar to that previously enforced by racist legislation. There was no social security net in place and few had any expectations of a realistic pension. As they explained, what would they eat and how would they pay their rent if they became unemployed for any length of time or had to stop work due to old age or ill health? Staying in the city permanently, they argued, was possible only ‘for those with a house and a pension’.

      The survey’s main aims had not included the issue of housing per se, but the key components of household budgets did emerge. And housing came up again and again during interviews, particularly in relation to family life and the risks of proletarianised life: unemployment, old age, illness. For this reason, during a subsequent survey in 1988, the residential area chosen was Glen Norah C: a new site-and-service settlement with access to water and electricity. Here, private-sector building societies had been brought in and incentivised via some new tax arrangements to provide loans to so-called ‘beneficiaries’ (homeowners). The shift from generous subsidisation of housing costs had begun, in line with trends across the world as the neoliberal phase of contemporary capitalism took hold. Precisely the same forces were at work in nearby Zambia and Malawi, for example. They were also in full swing in London.

      

      In the hope of maintaining a reasonable rate of repayment from the new borrowers in Glen Norah C, a minimum income was set as a condition of eligibility for the scheme, as well as a maximum. This was in contrast to the situation in Kuwadzana in 1985, at which time only a maximum income was set, above which people were ineligible for the scheme. Kuwadzana was thus much more definitely a pro-poor housing scheme, although Glen Norah C was also promoted as such by the government and the World Bank. Applying a minimum income condition at Glen Norah C showed that it was realised that the poorest Harare households could not afford to service a loan for a basic site-and-service plot, let alone build anything on it. In other words, it was proof of the premise of this book: that it is impossible for the formal private sector to provide housing for the poorest urban groups. It is also proof that the World Bank knew this. The loans available at Glen Norah C presumed that most of Zimbabwe’s urban workers would not be able to pay more than 30% of their monthly income to pay back the building society over 20 years. This guideline of a 30% level of income as a maximum for housing costs for low-income households is quite common in such calculations across the world, although nowadays many renters and homeowners in the largest cities of the GN are paying much more. However, where incomes are very low in real terms, as in most of Africa, it is understood that the opportunity cost of paying another 10% or 20% or more of your income for housing does not mean setting aside meals in restaurants or new clothes, but cutting out some food altogether, taking children out of school, or leaving the purchase of medicines so late that a sick family member will suffer unnecessarily or even die. In other words, when people are really poor, the welfare losses imposed by increasing payments for housing can be very serious and potentially life-threatening.

      

      It is worth noting that, unlike a mortgage for an already constructed house, the sort of housing loan made in Glen Norah (and in thousands of similar projects across the urban GS) did not provide any actual housing in the first instance – just access to a bare plot. Until something is built, the borrower has nowhere to live (or they must rent somewhere else, compounding the affordability issue). For those on the minimum income level applied at Glen Norah C, the maximum loan they could get covered the cost of the plot but not much else. In other words, there was no provision for the cost of building most of the actual house but they would already be paying out 30% of their income to service their loan. This is something of a dilemma, evidently.

      But what housing loan could typical residents living in Glen Norah C actually afford? The 1988 survey collected data on household incomes and it turned out that the average income of the 227 migrants interviewed was almost identical to the average for beneficiaries of the programme. In other words, they were good proxies for the households living in the area. Many of Harare’s low-income households are from other parts of Zimbabwe and, in the 1980s, most worked in the formal sector, as did the rest of the city’s workers. Unemployment levels were low. The survey data were used to construct housing affordability curves (see Figure 2.1), which showed how many could afford the housing payments needed to actually build a house, even if gradually. Not many, it turned out. Based on the household head’s income, only 8% could afford the monthly payment of Z$141.20 that was necessary to repay the size of loan required both to buy and occupy their plot and then to build the minimum type of house required by local regulations. The average loan provided by the building society was thus completely insufficient to build the required house. This was also obviously problematic. Yet half the migrants were even too poor to borrow that amount. A third had incomes below the minimum required for the scheme. The affordability curve based on total household income, including income earned by other household members, was slightly more favourable, with about 19% able to build the minimum house. This would still have left four-fifths unable