Zambezia (1990). XVII (ii).HOUSING POLICY, PRODUCTION AND CONSUMPTIONIN HARARE: A REVIEWPART IIC. RAKODIDepartment of City and Regional Planning,University of Wales College of CardiffandN. D. MUTIZWA-MANGIZADepartment of Rural and Urban Planning, University of ZimbabweTHIS IS THE second part of a two-part article the main aim of which is toreview published and other secondary material on housing markets inHarare. For organizational purposes we have considered production, con-sumption and exchange separately, although these issues are clearly inter-related. In the previous issue of Zambezia (1990, 1-30), material on theproduction and supply of housing was reviewed, concluding with asummary of the constraints on housing supply. In this part we summarizethe literature on the consumption of housing, include a brief section onthe exchange of housing and, finally, draw general conclusions and outlinedirections for future research.THE CONSUMPTION OF HOUSING IN HARAREEstimates of housing needTwo indicators of housing need have been used in Zimbabwe, the officialhousing waiting lists and estimates of need based on population growthprojections. In the past the housing waiting list for Harare (then Salisbury)was added to on a chronological basis. By 1977 the number of applicationshad reached 20 000 (Smout, 1977,50) and 36 000 by 1984 (Taylor, 1985) or63 000 if Chitungwiza were to be included (Mutizwa-Mangiza, 1985b, 85).Although this list was used as a guide to housing need, it was recognizedto be out of date and that many households did not bother to apply as thechance of their being allocated a house or a plot was so remote. In 1984the list was reformed in two stages. Firstly, all those who wished to beplaced on the list had to reapply and a system of annual updating wasadopted. In 1985 the list had dropped to 23 000 because of the non-renewal of applications but by August 1987 it had increased to 35 000 and111112HOUSING POLICY IN HARARE: PART II360 new applications were being received each month (Financial Gazette, 21Aug. 1987). Secondly, a scoring scheme was introduced to replace thesimple queueing system which had operated previously: this new schemetook into account length of urban residence, household size, income, etc. Inorder to be eligible for registration a household head has to be employed orin acceptable self-employment; be married or have parental responsibilities;not own or rent a house elsewhere; and have an income below Z$450 amonth (this amount was increased in 1985 to Z$600 a month). Certaincategories of household were thus ineligible, especially those with house-hold heads who were not working in wage employment or in registered self-employment. These included many households headed by women, althoughwidows were permitted to stay in houses already allocated to their families.The 1982 census registered 15 per cent of households in Harare as havingwomen heads, but only 7 per cent were beneficiaries of public housing.Women's property rights have, however, been strengthened since 1980 andmany women, although not all, are now eligible for public housing (Lee-Smith, 1987).The second way in which housing needs have been estimated is byprojections of new households, plus an estimate of the backlog or shortageand sometimes also replacement needs (Rhodesia, 1979; Zimbabwe, 1985,33-4). The 1979 estimates were given for particular cities, the 1985 onesfor the national urban population only. Not surprisingly, the estimatesvary widely, given the scope for different assumptions and uncertaintywith respect to future rates of urban growth. A pre-Independence estimatebased on an arbitrary, but deliberately generous, national urban growthrate of 11,5 per cent per annum between 1977 and 1983 which allocatedhousing need between centres led to a prediction that 67 000 dwellingunits would be needed for Harare, excluding Chitungwiza (Rhodesia, 1979).In the long-term plan (Zimbabwe, 1985) housing needs are estimated forall urban areas together, based on an anticipated population growth rateof 6 per cent per year up to the year 2000 and on an average householdsize of 6 people (Harare grew on average at 4,6 per cent per year between1969 and 1982).Alternatively, housing needs are based on an employment projection,in which it is assumed that employment in urban areas will grow at 3 percent per year between 1985 and 2000 and that two people will be workingper household. The figure realized from this projection is then added to anexisting backlog of 244 000 which in turn is added to estimates of annualreplacement, upgrading and new dwelling construction to ease over-crowding (Zimbabwe, 1985,33,36). In a further projection by the Ministryof Public Construction and National Housing, this time using an averagehousehold size of 4,7, it is estimated that 61 000 housing units will beneeded per annum between 1985 and 2000 to eliminate the housing shortageCAROLE RAKODI & N. D. MUTIZWA-MANGIZA 113in urban areas, excluding, it appears, estimates for replacement and up-grading (Zimbabwe, 1986a). The basis for these estimates of housing needis not alwaysxlearly spelt out and the assumptions made are, of necessity,arbitrary. They are not given for individual urban areas.Effective demand for housingTo provide more realistic inputs into policy formulation, estimates of theability of each group in the income distribution hierarchy to consumehousing are needed. For this, accurate information is needed on householdincome, on the proportion of that income that is available for expenditureon housing and utilities, as well as on the ability of households to mobilizeresources for housing from other sources, including the household itself,the extended family and credit. Thus information is required on the avail-ability of conventional and unconventional housing finance, in terms ofboth volume and eligibility criteria. The availability of credit is of relevanceto existing or potential owner-occupiers. In understanding urban housingmarkets, however, rented housing must also be analysed, including theextent and nature of the rented housing sector over time, the characteristicsof tenant households and the nature and outcomes of government inter-vention. Government intervention to change the nature of demand mayinclude the provision of credit, policies intended to influence income orhousehold savings, and subsidies, the distributional impact of which needsto be assessed.Estimates of housing need in Zimbabwe, arbitrary as they are, aregenerally accompanied by some recognition that the ability to pay is ofrelevance. Thus the pre-Independence estimates quoted above noted therelatively low incomes in Bulawayo and Harare (then Salisbury) in relationto public housing costs (Rhodesia, 1979,11-12). As a result, it was recom-mended for 1978-83 that 20 per cent of the total of 128 700 new unitsrequired would be in transit areas with communal ablution facilities, 45per cent would be serviced plots with individual ablution facilities, 30 percent conventional 'low-cost' housing, and 5 per cent medium-cost houseson 2 000 m2 stands (Rhodesia, 1979, 20-1).In the long-term housing plan the estimated number of new house-holds resulting from employment growth plus the backlog of those lodgedor housed in the informal sector is divided into five income categories.These are based loosely on an income distribution obtained from surveysof high-density areas in 1982, and on an assumption that 25 per cent ofincome is available for housing and that each household is liable forpayment of a Z$15 service charge and eligible for a housing loan on whichrepayment is Z$9 per month for each Z$l 000 borrowed. Rather than beingused to illustrate what households can afford to pay for housing and thuswhat type of residential package would be feasible, the figures are used to114HOUSING POLICY IN HARARE: PART IIestimate total capacity required for loans (Zimbabwe, 1985, 28-36). Asomewhat different approach has been adopted by the Ministry of PublicConstruction and National Housing, in which it is assumed that 46 per centof new households will be low income, 44 per cent medium income and 10per cent high income. These proportions are then applied to estimates ofanticipated new urban households, the total cost of the housing programmethen being estimated and allocated to the public and private sectors (thepublic sector is responsible for 68 per cent of the cost of low-income-sector housing, 20 per cent of the cost of medium-income-sector housingand none of the cost of anticipated high-income-sector housing) (Zimbabwe,1986a, 39, and 1986b, 14). Again, the purpose of the projections is not toanalyse what urban households can actually afford to pay but to provide aguide to housing policy. This has resulted in the Ministry being alerted tothe discrepancy between recent investment levels and the resourcesrequired if housing needs defined in this way are to be met.A more systematic attempt to relate incomes to the housing optionsavailable has been made by Mutizwa-Mangiza (1985b) using the incomefigures available from a 1982 survey of households in Harare. Minimum-wage legislation was introduced immediately after Independence. Althoughthe poverty datum line was over Z$100 per month in 1979 (Astrow, 1983)the recommendation of the Riddell Report that a minimum wage of Z$128be phased in over three years was resisted by the business communityand the mimimum wage was set at Z$70 for urban workers in July 1980(the minimum wage for domestic workers was set at Z$30) and increasedto Z$105 in January 1982 and Z$115 in January 1984 (Riddell, 1984, 467).However, the real value of these increases was eroded by inflation andreductions in food subsidies, and by January 1984 an income of Z$125would have been needed to purchase the same quantity of goods that itwas possible to purchase for Z$70 in 1980 (Riddell, 1984).Although urban workers earning less than Z$150 benefit from freehealth care and all workers from free primary-school education and somecategories of worker, especially domestic workers, have benefited fromminimum-wage increases, for many industrial workers the minimum wagewas less than their current wage. In fact, mimimum wages legislationprovided an excuse to hold wage rates down (Astrow, 1983) and thus realstandards of living in urban areas have deteriorated, while since 1983 theconsumer price index for low-income families has increased more thanthat for higher-income urban families (326,6 net of tax, compared to 304,1in December 1989, base 100 in 1980). Although a poverty line does notseem to have been defined recently, in 1982 the ceiling for-the housingwaiting list was about Z$400 and only 7 per cent of households hadincomes above this (Mutizwa-Mangiza, 1985b, 89). Patel (1984) notes thatan income of Z$480 is needed for a building society loan of Z$12 000 for aCAROLE RAKODI & N. D. MUTIZWA-MANGIZA 115private sector house. Using the local rule of thumb (that not more than27,5 per cent of income is available for housing), Mutizwa-Mangiza (1985b,90) notes that while 16 per cent of households were unable to afford any ofthe available strategies (because their incomes were less than Z$80), 37,5per cent could have afforded a serviced plot without wet core, 13 per centa serviced plot with wet core, 17,5 per cent a core house of between oneand three rooms, and only 16 per cent (those with incomes of Z$260 andabove) the official four-roomed core house (see Fig. 2). However, he alsonotes that the 1976/7 Lower Income Expenditure Survey records house-holds as spending an average of 17,6 per cent of their incomes on rent,water, fuel and light. Applying this alternative rule of thumb (see Fig. 3),the proportion of households resident in high-density areas who wereunable to afford any of the available strategies rose to 42 per cent (thosewhose income was below Z$123), while 36 per cent were able to afford aserviced plot with no wet core, 7 per cent a serviced plot with a wet coreand only 7 per cent (those with an income of Z$406 and above) a four-roomed house (Mutizwa-Mangiza, 1985b, 90).Although recent information on income distribution and amountsactually spent on housing and services is not available, the conclusionsare clear. The great majority of low-income urban residents, includingthose below the income ceiling of Z$ 175 used in Kuwadzana, cannot afforda four-roomed house with a fully serviced 300 m- plot. The implications ofthis analysis for housing policies, while acknowledged in general terms inofficial publications (Zimbabwe, 1985), have not been faced and the contra-diction between the aim of providing affordable housing and currentprovision practice remains.Housing financeTo analyse effective demand, information is required on the resourcesavailable to individual households, including the structure of householdincome, the likely availability of savings, the income of household membersother than the head, and inter-household transfers. No recent informationof this sort is available for Harare. The cost of housing in relation tohousehold income makes the availability of credit crucial, and this mustbe backed by an institutional system which can make capital available forlarge-scale public-sector investment as well as for individual house pur-chasers or builders. Information on the housing finance system in Zimbabwerefers solely to the conventional formal-sector system, but the absence ofinformation should not lead us to assume the absence of informalmechanisms. We will examine, firstly, the system of financing the public-sector housing investment programme, secondly, the existing private-sector sources of housing finance and, lastly, recommendations whichhave been made to improve the system and the impact of initial reforms.116HOUSING POLICY IN HARARE: PART IFigure 2: HOUSING AFFORDABILITY: 27,5 PER CENT OF INCOMEDEVOTED TO HOUSINGCAROLE RAKODI & N. D. MUTIZWA-MANGIZA117Figure 3: HOUSING AFFORDABILITY: 17,6 PER CENT OF INCOMEDEVOTED TO HOUSING£600I>5003 40030020010012040four room core ( 7%) "three room core (1 %) -two room core (2%)one room core (5%) ŠS 9I71 0965-60 -57-1447-1138-2521-4760 80 100% Low-income population118HOUSING POLICY IN HARARE: PART IPublic-sector housing investment programmeThis programme has been financed from four sources: the National HousingFund, internal sources, borrowed private-sector funds, and aid programmes(Mutizwa-Mangiza, 1985a; National Council of Savings Institutions (NCSI, 1985).(a) The National Housing Fund, an intermediary organization adminis-tered by the Ministry of Public Construction and National Housingas part of the Public Sector Investment Programme, allocates loansto local authorities at an interest rate of 9,75 per cent over 30 yearsfor annual programmes and particular projects. Although in thepast it has obtained some loans from private-sector sources, since1980 it has been entirely dependent on the central government forfunds, and during the period of the Transitional Development Plan(1982-5) was allocated only 25 per cent of the estimated amountneeded (NCSI, 1985). Local authority repayment arrears havethreatened its self-sufficiency in recent years and additionalgovernment funds have been required to cover the deficit. In addition,the number of units financed in the country as a whole has decreasedeach year since 1980, partly because of declining allocations in realterms and partly because of the increased unit cost (from Z$l 440 in1980/1 to Z$6 940 in 1982/3, Z$7 580 in 1983/4 and an estimated Z$12740 in 1984/5), only a small part of which can be attributed toinflation (NCSI, 1985, 39).(b) Internal sources of funds include rents, service charges, and waterand electricity charges. In addition, 50 per cent of the profits fromthe sale of beer in low-income areas was available for a HousingRevolving Fund which was used for loans for housing and infra-structure development. In addition funds raised from the sale ofbeer were used to provide facilities and amenities in these areas.Since 1980 an 80 per cent tax on profits on the sale of beer hasbenefited central government at the expense of local authorityrevenue. Although in Harare about Z$l million per annum is obtainedfrom the established Fund, inflation exceeds interest and so de-capitalization is occurring. Between 1961 and 1976 employers werealso taxed to subsidize transport and housing for Blacks (the ServicesLevy Fund) (Patel and Adams, 1981).(c) The Harare City Council has in the past been able to borrow private-sector funds directly.(d) Aid funds, notably from USAID, the World Bank and th& Common-wealth Development Corporation, accounted for 58 per cent of thecapital funds available to the Ministry of Public Construction andNational Housing in 1985/6 and of which 59 per cent was channelled toCAROLE RAKODI & N. D. MUTIZWA-MANGIZA 119Harare for the provision of 6 500 plots. However, this clearly cannotbe relied on in the long term.The funds from central government and external agencies appear to rep-resent the only subsidies to the low-income housing programme, which isotherwise, with the exception of some cross subsidies operated by a poolrental system, based on full cost recovery. Administrative costs of Z$3,85per month per loan in Harare are added to funds otherwise re-lent to plotpurchasers at the same interest rate (9,75 per cent) and over the samerepayment period (30 years), as the funds are obtained from the NationalHousing Fund (NCSI, 1985, 51). In new aided self-help housing schemes,allottees qualify for local authority loans related to their incomes. Thus inKuwadzana, following a 5 per cent down payment on the purchase price ofthe plot (Z$410-855) and repayment of the remainder over 30 years,allottees are eligible for a loan of between Z$294 and Z$3 230 (averageZ$2 500) repayable in monthly instalments which do not exceed 27,5 percent of their incomes (Gore et al., 1985; Mafico, 1987). In view of theexclusion of lower-income people from the purchase of the plots currentlybeing offered, the subsidy is not very effectively targeted at those who, itcould be argued, have the greatest need of it.Private-sector sources of housing financeThese include building societies, other private-sector financial institutions,employers and households. Three well-established building societies whichhave operated for over twenty years constitute the main source of fundsfor high-income households' house construction or purchase. Buildingsocieties are dependent on savings and have been adversely affected bypost-Independence developments: Including the increase in the tax rate(to 44 per cent on average) and the broadening of the tax base, which havereduced potential savings; the increase in deposit and lending rates in1981 which increased the cost of interest above returns on outstandingmortgages; the increase in mortgage interest rates in the same year to13,25 per cent, which increased arrears and thus collection and legalcosts; the loss to the Reserve Bank in 1984 of Z$55 million worth ofblocked funds belonging to emigrants; and the abolition of tax on theinterest from Post Office Savings Bank (POSB) accounts which was instit-uted in order to increase POSB savings (they are a source of governmentsecurities). As a result, the tax-free return on a POSB account for a highest-rate taxpayer is potentially 26 per cent per annum. Between 1980 and1984, while POSB deposits increased by 90 per cent, building societydeposits increased by only 7 per cent (NSCI, 1985). Although 24 539 newloans were made between 1980 and 1984 (average amount Z$14 240)mostly from reflows and backed up in part by the Ministry of Public120HOUSING POLICY IN HARARE: PART IIConstruction and National Housing's Housing Guarantee Fund (see below),these loans were confined to the higher-income groups (minimum loanZ$6 000) and may be insufficient to meet future demand.Building societies rely on share capital (the interest rate is now 11,25per cent or marginally above the POSB maximum interest rate of 10 percent) and savings accounts (until 1981 the interest rate was 3,5 per centbut is now 7,75 per cent in comparison with the recently increased rate of8,5 per cent for POSB accounts) (NCSI, 1985). Loans are made availableover 25 years at between 12,5 per cent for residential loans of less thanZ$12 000 and 14,75 per cent for commercial loans. This represents areasonable return over the average cost of money (9,45 per cent), buthides the problem of high administrative costs of a large number of smallsavings accounts (those with a balance of less than Z$300) which, forexample, made up 81 per cent of the Central African Building Society's(CABS) accounts in 1984 (NCSI, 1985, 31), and which accounted for only12,9 per cent of savings but 79 per cent of withdrawals and cost an averageof Z$l,13 per transaction. Although the building societies do not appear towish to cease to handle such small accounts, it is noted that they do notbreak even and that overall profits in recent years have been modest.However, the recently introduced paid-up permanent shares (PUPS) withtax-free interest have significantly increased the flow of money into buildingsocieties.A mortgage guarantee scheme is already operated by the governmentvia the Housing Guarantee Fund (HGF). While building societies arenormally willing to lend up to 70 per cent of the price of the house, the HGF100 per cent scheme enables civil servants to obtain 100 per cent mortgages(ceiling Z$50 000 and payments deducted at source), while members ofthe public who can raise the 10 per cent deposit may qualify for the 90 percent scheme (maximum price Z$17 000, thus allowing a maximum loan ofZ$15 300 which at 22,5 per cent of income places an income ceiling ofZ$475 per month on beneficiaries). These schemes are of roughly equalimportance and just over 2100 guarantees per year were issued on averagebetween 1980 and 1984 (NCSI, 1985, 44-6). Between 1979 and 1984 therewere 66 foreclosures under these schemes, the dwelling units being addedto the HGFs housing stock (value Z$7,2 million in 1984) the rents fromwhich, ev$n when let at fair rents to civil servants, enable it to make asmall annual surplus.Other private sector financial institutions, such as commercial banks,pension funds and insurance companies, are relatively well developed inZimbabwe, the last two controlling 15 per cent and 20 per cent of financialresources, respectively. Commercial banks were not involved in the housingmarket before 1980, but their contribution to mortgage lending increasedduring the early 1980s to compensate for a shortage of building societyCAROLE RAKODI & N. D. MUTIZWA-MANGIZA 121funds. Pension funds and insurance companies have in the past invested aproportion of their funds in the building societies. With the recent require-ment that 60 per cent of their assets be held in government securities, thissource of funds has become less available for housing finance.Employers have in the past been able to adopt one or more housing-related policies. They have constructed housing for rent (at a partly orwholly subsidized rent), made loans to employees for house purchase or adown payment, provided guarantees to enable employees to get access toa building society loan, or given a monthly housing allowance. Althoughno information is available on employers' roles since Independence, andthey may never have been very significant in Harare, they have probablybeen reduced because of the high standard and cost of construction,pressure to freeze or lower rents and the difficulty of evicting an employeeif he leaves his current employment (NCSI, 1985).Recent developmentsConsiderable attention has been given in recent analyses and policystatements to the structure and role of the formal-sector housing financesystem (NCSI, 1985; Zimbabwe, 1985, 1986a, 1986b). These have referredboth to capital supply to the sector and household access to credit. Theshortage of public-sector capital funds to finance housing loans in additionto land acquisition and infrastructure installation, plus the well-developedbut deteriorating position of the building societies, have led to a variety ofrecommendations and actions.Firstly, it has been recommended that the advantage of the POSB overbuilding societies in attracting savings be removed. It has been noted that,while at the end of 1984 there were, nationally, 842 375 non-term and 15160 term accounts in the POSB, most were used as current accounts and77 per cent had a month-end balance of Z$100 or less. These accounts,however, accounted for only 3,2 per cent of total savings, illustrating theattraction of tax-free interest to larger individual and corporate savers. Ithas been suggested that restoring tax on interest on POSB accounts wouldnot affect the majority of savers and would give the building societies animproved chance of competing for the main volume of funds, which isheld in relatively few accounts, while government tax revenue would, ofcourse, increase. Instead, in 1987 building societies were (as has beenpointed out earlier) permitted to issue 9 per cent tax-free paid-up permanentshares (ceiling on investment Z$75 000 for individuals and Z$35 000 forcompanies) (Zimbabwe, 1987b), which is expected to enable buildingsocieties to attract additional investment funds.The suggestion that a secondary mortgage market be encouraged todevelop by the redirection of a proportion of the assets of life insurancecompanies, and pension and provident funds, now invested in government122HOUSING POLICY IN HARARE: PART IIsecurities, into the housing market under the auspices of established discounthouses (NCSI, 1985) still appears to be under consideration (Zimbabwe,1986b). However, it has also been recommended that local authorities onceagain be allowed to borrow from these institutions (Martin et al., 1985),which, even if healthy, may not have the resources to support all these newdemands. Under the catchphrase 'public- and private-sector partnership'(Zimbabwe, 1986a, 1986b), it has been suggested that the aim should be thatfor every Z$l of public money invested, Z$l of mortgage money from theprivate sector and Z$l from household savings should be made available(Zimbabwe, 1985, 51). This recent intention to move towards public- andprivate-sector partnership appears to be a response to emerging policyamong international aid agencies, including the United Nations Centre forHuman Settlements, as reflected in A New Agenda for Human Settlements(United Nations Centre for Human Settlements, 1987). Minor additionalchanges recommended, but not yet implemented, include requirementsthat building society borrowers be savers with the societies and that civilservants be required to pay a deposit.As we have noted, at present building society mortgages are availableonly for houses costing more than Z$6 000 and are thus only available tomiddle- and upper-income households. In an attempt to make housingfinance available to lower-income households, it has been suggested thatfunds could be channelled to local authorities via a National HousingCorporation, which would develop appropriate solutions for every incomegroup, ensure the availability of land, organize the design and constructionof housing schemes, and ensure that finance was available for the purchaseof units (Martin etal, 1985). This report by Martin et al. (1985), sponsoredby USAID and United Nations Centre for Human Settlements, implies thatcentral government funds would still have to be made available for housescosting Z$6 000 or less and that some housing might need to be subsidized.The official attitude to subsidies, however, remains ambivalent. Whileemphasizing its objectives of minimizing subsidies and providing affordableshelter, the Ministry of Public Construction and National Housing alsoappears to acknowledge a need for shelter assistance programmes forworkers earning below the minimum wage Š 'informal-sector employees'and the unemployed (Zimbabwe, 1985, 37, 45).An alternative approach has been adopted by USAID under its policyemphasis on public- and private-sector partnerships in housing and urbandevelopment (Zimbabwe, 1986b) and the World Bank (1987), which aims totap the existing expertise of the building societies to provide constructionloans for low-income housing. In an experimental project in Kwekwe, amedium-sized town, and Gutu, a rural growth-centre, which was partly fundedby USAID and UNDP and implemented by local agencies with United NationsCentre for Human Settlements' assistance, only the Beverley Building SocietyCAROLE RAKODI & N. D. MUTIZWA-MANGIZA 123could be persuaded to participate by making loans for houses which onaverage cost Z$2 655 to construct (sufficient in Kwekwe for a four-roomedcore house). The vast majority of participants in this experiment had incomesof Z$150 or less (Zimbabwe, 1986c). The housing finance component of theproject had a number of teething problems, as mechanisms were devised tosatisfy the building society's administrative and collateral requirements. Asimilar approach has been adopted by the World Bank, in which the threebuilding societies are being coaxed to move into lending finance for low-income housing and serviced plots, with assistance of Z$9,6 million seedmoney from the Commonwealth Development Corporation to add to theirown funds (World Bank, 1986). Since then, all three building societies havelent a proportion of their funds for low-income housing.THE EXCHANGE OF HOUSINGPeople exchange dwellings to meet changing family needs, as a result of achange of job or of altering economic circumstances, as a hedge againstinflation, or simply because of personal preference. Systems of exchangemay be either formal or informal and may be subject to a greater or lesserdegree of government regulation. The various means of exchange availablein a particular city should be analysed in order to assess whether appro-priate services are provided to those who wish to exchange dwellings, tosee who benefits from the systems and to assess the impact of governmentintervention on the various actors involved. Nothing has been written sofar on the process of exchange of land and housing and its effects on thehousing sector in Harare. What is known, however, is that the high-incomeprivate-sector housing component has a well-organized system of estateagents and conveyancers, all of whom are properly registered. Here thepublic sector has virtually no role except in the registration of deeds.With respect to the exchange of properties within public-sector housingschemes, beneficiaries of sites and services projects are prohibited fromselling a plot before any developments have been implemented. However,where a beneficiary whose plot is partially or fully developed wishes tosell his/her rights, interests and obligations in that property at a timewhen he/she has not yet finished paying back the Municipality, a fairlyelaborate procedure (cession procedure) has to be followed. Here, boththe cedant (person selling) and the cessionary (person buying) have toapproach their District Housing Office. The cedent, if married, is requiredto bring his/her spouse, who must consent to the sale of the property Šthe reason being to prevent irresponsible spouses from selling their housesbehind their partners' backs. The District Housing Office will refer bothparties to the Head Office of the Department of Housing and CommunityServices. The Head Office will then look at the reasons why the cedent124HOUSING POLICY IN HARARE: PART IIwishes to sell, whether the cessionary/buyer owns another property inHarare or Chitungwiza, whether the cessionary is a Harare resident andwhether he/she is on the Council's housing waiting list, and, finally, whetherthe proposed price is commensurate with Council's prices rather thanbased on prevailing market forces.The procedure outlined above is meant to prevent the infiltration ofhigher-income people and property speculators into public-sector low-income areas, and also to prevent the development of a fully fledged low-income housing market which would push prices beyond the reach of theintended beneficiaries. However, an increasing number of low-incomeproperties are appearing on the market. For example, in one newlydeveloped housing scheme, Warren Park, 78 cessions were completed in aperiod of 17 months (October 1987 - February 1989). This figure does notinclude properties sold after title deeds had been transferred to beneficiariesupon completion of purchase from the Municipality. A common practiceused to avoid the Municipality's cession procedure is first of all for the sellerand buyer to agree on a price privately (usually well above what theMunicipality would have set). After the buyer has paid this amount, theseller then uses part of it to pay off the outstanding loan in order that the titledeeds may be transferred to him. Thereafter, the property is transferred tothe purchaser through private-sector conveyancers, without any hindrancesfrom the Municipality. In this way, low-income properties are slowly butsteadily creeping into the open market, their use value is turned into exchangevalue, and former owner-occupiers are turned into tenants and lodgers. Theextent of this process has, however, yet to be quantified.CONCLUSIONThe public sector has had a major role to play in the housing market inHarare. The political decision-making process and the administrativestructure for policy-making and implementation are, therefore, importantinfluences on the approaches formulated and their outcomes. Most of theavailable sources which discuss the operation of particular institutions,such as the main government ministry involved, the building societies andthe local authorities, describe and, in some cases, assess their operation,although in fairly superficial terms. The precise political and organizationaldynamics of these organizations have not been discussed, nor has theinterface between them and the national interests which are used inexplanation by other authors.In 1980 the new government had no plans to reorganize urban localgovernment; instead it aimed to de-racialize the representative bodies instages. Although Dewar (1987) has outlined the administrative history ofHarare, the strength of his analysis lies in the description of settlerCAROLE RAKODI & N. D. MUTIZWA-MANGIZA 125assumptions and values: there is no discussion either of the interestsrepresented by the Black membership of Township Boards after 1971 or ofthe City Council after 1980, or of the influence of these councillors ondecision-making and the operation of local government services. Such anexplanation has been attempted for Marondera (then Marandellas) byHodder-Williams (1982). Dewar also describes the organizational structureadopted by ZANU(PF) in its attempts to consolidate and extend its powerbase, but he does not analyse the interests represented by Party leadersor their influence on decision-making in the city, other than to suggestthat Party membership may be an essential credential in obtaining accessto the right to vote, to housing and to jobs in the bureaucracy Š anassertion which itself is highly suspect and is not demonstrated empirically(Dewar, 1987, 47).Similarly, at the national level, the impact of these administrative changes,the effect of inter-ministerial rivalry and the relationship between ministriesand the locations of decision-making power have not been analysed, althoughPatel (1984) describes the splitting up, in 1982, of the Ministry of LocalGovernment and Housing into the Ministry of Local Government and TownPlanning (now the Ministry of Local Government, Rural and UrbanDevelopment) and the Ministry of Housing, the latter being merged in 1984with the Ministry of Construction to form the Ministry of Public Constructionand National Housing.The disjunction between ideology and practice identified by writerson housing such as Schlyter (1985), Teedon and Drakakis-Smith (1986)and Drakakis-Smith (1987) is attributed to the petty-bourgeois class baseof ZANU(PF) leadership (see also Astrow, 1983) and the peasant base of itspopular support. The stated objectives of the government are to provideadequate and affordable housing and related services for all and to eliminatethe urban backlog by the end of the 1980s. That achievements fall far shortof these aims has become only too clear in this analysis, but to attributethe absence of socialist housing policies and the failure to redistributehousing resources to a state which represents the interests only of capital,the bourgeosie and peasants is to advance only a partial explanation, andin particular is to ignore the limited room to manoeuvre open to thepresent government (Libby, 1984; Riddell, 1984).A more realistic assessment of effective demand and of the costs ofadopting high infrastructure and construction standards in new low-incomehousing development from the outset, together with a realization of theadverse effects of measures to constrain supply, would have made asignificant difference to the availability of land and housing in the city.Some tentative explanations of government's failure to adopt such policychanges have been advanced, but too little evidence is available to reachfirm conclusions on why these inappropriate policies have been continued.126 HOUSING POLICY IN HARARE: PART IIA variety of areas for future research have been identified and may besummarized as follows:(a) A survey of the characteristics of the middle- and high-incomehousing stock Š including its price trends and present ownership,the volume and location of sources of initiative, capital and pur-chasers for new construction now taking place, and attitudes ofresidents to service levels and densification Š needs to be made.(b) An investigation of land available for housing, including privatesubdivision and serviced-plot schemes, the potential for locatinghigh-density schemes in or near low-density areas, constraints whichinclude the attitudes of current residents, physical and socialinfrastructure, the location of employment centres and the provisionof public transport needs to be undertaken.(c) The organization of the construction sector, both large- and small-scale, should be reviewed. This review should include the currentactivities of firms, their capital sources and ownership, preferredwork, and constraints on other ways of organizing construction,such as direct labour forces, building brigades and co-operatives.(d) The rented housing market, including the effects of rent regulationin high-and middle-income areas, and the registered and unregisteredrented housing market in high-density areas and up-graded areasshould be studied.(e) A follow-up to Butcher's (1986) study investigating the impact ofupgrading on Epworth should be made.(f) Both formal and informal housing exchange systems should bereviewed, including the mechanisms of exchange, constraints onexchange, government regulations governing exchange and the long-term effect on occupancy.(g) A survey of the building societies, including the results of the 1987measure to increase their share capital, and their reaction to theKwekwe-Gutu pilot scheme and the schemes funded by the WorldBank and the Commonwealth Development Corporation intendedto encourage their participation in the low-income housing market,should be undertaken.(h) Household incomes, especially those of low-income households,should be researched and should include the following topics;sources of income; availability of income for housing; availability ofother resources for housing, including savings; inter-householdtransfers; and the reactions of members of low-income householdsto the loan programmes available. This research should lead to aCAROLE RAKODI & N. D. MUTIZWA-MANGIZA 127realistic assessment of the resources available for housing and forma guide for policy-making.(i) Realistic estimates aimed at matching supply to demand, while notneglecting needs, should be formulated to provide a basis for financialestimates and bids for government and other funds.ReferencesANGEL, S. et al. 1977 'The low income housing delivery system in Asia',in R. P. Pama, S. Angel and J. H. de Goede (eds.), Low Income Housing:Technology and Policy (Oxford, Pergamon), 1177-218.ANGEL, S. et al. (eds.) 1983 Land for Housing the Poor (Singapore, SelectBooks).ASHTON, E. H. 1969 'Economics of African housing', Rhodesian Journalof Economics, III, iv, 29-33.ASTROW, A. 1983 Zimbabwe: A Revolution that Lost Its Way? (London,Zed Press).ATKINS, G. 1983 'Town planning and economic design', ZimbabweEnvironment and Design, V, 5-9.BOLEAT, M. 1987 'Housing finance institutions', in L. Rodwin (ed.), Shelter,Settlement and Development (Boston, Allen and Unwin), 151-78.BRATTON, M. 1981 'Development in Zimbabwe: Strategy and tactics',Journal of Modern African Studies, XIX, 447-75.BURGESS, R. 1978 'Petty commodity housing or dweller control? A critiqueof John Turner's views on housing policy', World Development, VI,1105-133.BURGESS, R. 1985 'The limits of state self-help housing programmes',Development and Change, XVI, 271-312.BURGESS, R. 1987 'A lot of noise and no nuts: A reply', Development andChange, XVIII, 137-46.BUTCHER, C. 1986 'Low Income Housing in Zimbabwe: A Case Study ofthe Epworth Squatter Upgrading Programme' (Harare, Univ. ofZimbabwe, Dept. of Rural and Urban Planning, Occasional Paper 6).CHRISTOPHER, A. J. 1972 'Urban encroachment on a large farm inRhodesia', Land Economics, XLVIII, 287-90.COLQUHOUN, B., O'DONNELL, H. AND PARTNERS 1985 'Overview of Con-struction and Materials Production Sectors for Inclusion in the NationalHousing Corporation Study' (Harare, Report to the Ministry of PublicConstruction and National Housing).DAVIES, D. H. 1987 'Harare, Zimbabwe: Origins, development and post-colonial change', African Urban Quarterly, II, 131-7.DEWAR, N. 1987 'Salisbury to Harare: Citizen participation in publicdecision-making under changing ideological circumstances inZimbabwe', African Urban Quarterly, II, 37-48.128 HOUSING POLICY IN HARARE: PART IIDOEBELE, W. A. 1987 'Land policy', in L. Rodwin (ed.), Shelter, Settlementand Development (Boston, Allen and Unwin), 110-32.DOWALL, D. E. 1989 'Bangkok: A profile of an efficiently performinghousing market', Urban Studies, XVI, 327-39.DRAKAKIS-SMITH D 1986 'Urbanisation in the socialist Third World: Thecase of Zimbabwe', in D. Drakakis-Smith (ed.), Urbanisation in theDeveloping World (London, Croom Helm), 141-57.DRAKAKETSMITH, D. 1987 'Zimbabwe: The slow struggle towards social-ism', Geography, LXXII, 348-51.ENGLAND, R. and ALNWICK, D. 1982 'What can low-income families affordfor housing?', Habitat International, VI, 461-70.GILBERT, A. and VAN DER LINDEN, J. 1987 'The limits of a Marxist theoreticalframework for explaining state self-help housing', Development andChange, XVIII, 129-36.GILBERT, A. and WARD, P. 1985 Housing, the State and the Poor: Policy andPractice in Three Latin-American Cities (Cambridge, Cambridge Univ.Press).GORE, C. H. et al. 1985 'Small Scale Informal Construction Sector atKuwadzana Low-Income Housing Project' (Harare, Report to the Min-istry of Public Construction and National Housing, mimeo.).HARVEY, S. D. 1987 'Black residential mobility in a post-IndependentZimbabwean city', in G. J. Williams and A. P. Wood (eds.), GeographicalPerspectives on Development in Southern Africa: Papers from the RegionalConference of the Commonwealth Gergraphicfal]Bureau, Lusaka (Towns-ville, James Cook Univ. of North Queensland for the CommonwealthGeographical] Bureau), 179-87.HODDER-WILLIAMS, R. 1982 'A Small Town in Crisis: Marandellas, Zimba-bwe' (Birmingham, Paper presented at a conference of the AfricanStudies Association).HOEK-SMIT, M. 1982 'Housing Preferences and Potential Housing Demandof Low Income Urban Households in Zimbabwe: Vol. 1' (Harare, Reportprepared for the Ministry of Housing, mimeo.).JOHNSTONE, M. 1984 'Urban housing and housing policy in peninsularMalaysia', International Journal of Urban and Regional Research, VIII,497-529.LEE, M. 1985 'Myths of affordability', Third World Planning Review, VII,131-42.LEE-SMITH, D. 1987 Overview of Women and Shelter in Africa (Nairobi,Mazingira Institute).LJBBY,R.T. 1984 'Developmental strategies and political divisions withinthe Zimbabwean state', in M. Schatzberg (ed.), The Political Economyof Zimbabwe (New York, Praeger), 144-63.MAFICO, C. J. C. 1987 'An Analysis of Public Sector Urban Low IncomeHousing in Zimbabwe: Towards an Appropriate Housing Policy' (Liver-pool, Univ. of Liverpool, Ph.D. thesis).MALPEZZI, S. and MAYO, S. K. 1987 'The demand for housing in developingCAROLE RAKODI & N. D. MUTIZWA-MANGIZA 129countries: Empirical estimates from household data', EconomicDevelopment and Cultural Change, XXXV, 687-722.MARTIN, R. et al. 1985 'National Housing Corporation: Feasibility Studyfor the Republic of Zimbabwe' (Harare, United States Agency forInternational Development and United Nations Centre for Human Settle-ments).MEHTA, M. and MEHTA, D. 1989 Metropolitan Housing Markets: A Study ofAhmedabad (New Delhi, Sage).MOAVENZADEH, F. 1987 'The construction industry', in L. Rodwin (ed.),Shelter, Settlement and Development (Boston, Allen and Unwin), 73-109.MOLLER, V. 1974 'Some aspects of mobility patterns of urban Africans inSalisbury', Proceedings of the Geographical Association of Rhodesia,VII, 22-32.MUTIZWA-MANGIZA, N. D. 1985a 'An Analysis of Low-income HousingStrategies in Harare' (Bradford, Univ. of Bradford, Project PlanningCentre for Developing Countries, Discussion 111, mimeo.).MUTIZWA-MANGIZA, N. D. 1985b 'Post-Independence urban low-incomeshelter policies in Zimbabwe: A preliminary appraisal of affordability',in S. D. Romaya and G. H. Franklin (eds.), Shelter, Services and theUrban Poor (Cardiff, University of Wales Institute of Science andTechnology), 81-98.MUTIZWA-MANGIZA, N. D. 1986 'Urban centres in Zimbabwe: Inter-censalchanges, 1962-1982', Geography, LXXI, 148-51.MUTIZWA-MANGIZA, N. D. and MARCIANO, M. 1987 'Building Brigades andUrban Low-cost Shelter Provision in Zimbabwe: A Preliminary Evalu-ation' (London, International Centre for Technical Research, Paperpresented at an International Conference on Urban Shelter in Devel-oping Countries, mimeo.).NATIONAL COUNCIL OF SAVINGS INSTITUTIONS 1987 'Housing Finance in Zimba-bwe' (Washington, Report prepared for the United States Agency forInternational Development).PATEL, D. H. 1984 'Housing the urban poor in the socialist transformationof Zimbabwe', in M. Schatzberg (ed.), The Political Economy ofZimbabwe (New York, Praeger), 182-96.PATEL, D. H. and ADAMS, R. J. 1981 Chirambahuyo: A Case Study in Low-income Housing (Gwelo, Mambo Press).POTTS, D. 1987 'Recent Rural-urban Migrants to Harare, Zimbabwe: TheMaintenance of Rural Ties' (Univ. of London, Centre for African Studies,School of Oriental and African Studies, Paper presented at a workshopon Rural-Urban Links, mimeo.).RAKODI, C. 1983 'La politique du logement pour les bas revenus aLusaka, Zambie', trans. A. Durand-Lasserve, in Le Logement, I'etat etles pauvres dans les villes du Tiers-Monde (Talence, Centre d'Etude deGeographie Tropicale et Centre National de la Recherche Scientifique,Practique Urbaines 2), 155-72.130 HOUSING POLICY IN HARARE: PART IIRAKODI, C. 1987 'Land, layouts and infrastructure in squatter upgrading:The case of Lusaka', Cities, IV, 348-70.RENAUD, B. 1984 'Housing and Financial Institutions in DevelopingCountries: An Overview' (Washington, World Bank, Staff Working Paper658).RENAUD, B. 1987 'Financing shelter', in L. Rodwin (ed.), Shelter,Settlement and Development (Boston, Allen and Unwin), 179-204.RHODESIA 1979 Urban Development in the Main Centres (Harare, Ministryof Finance).RlDDELL, R. C. 1984 'Zimbabwe: The economy four years after Inde-pendence', African Affairs, LXXXIII, 463-76.SCHLYTER, A. 1985 'Housing strategies: The case of Zimbabwe', Trialog,VI, 20-9.SEAGER, D. 1977 'The struggle for shelter in an urbanizing world: ARhodesian example', Zambezia, V, 83-90.SHENG, Y. K. 1989 'Some low-income urban delivery systems in Bangkok,Thailand', Environment and Urbanization, I, 27-37.SIMON, D. 1985 'Contemporary decolonization and comparative urbanresearch in Southern Africa', Comparative Urban Research, X, 32-41.SIMON, D. 1986 'Regional inequality, migration and development: Thecase of Zimbabwe', Tijdschrift voor Economische en Sociale Geografie,LXXVII, 7-17.SMOUT, M. A. H. 1977 'Urbanization and development problems inRhodesia', Journal of Tropical Geography, XLV, 43-51.STEVENS, R. F. and EATON, K. J. 1984 'Feasibility of Setting up a BuildingResearch Establishment in Zimbabwe' (Harare, Report to the BritishOverseas Development Administration and to the Ministry of Con-struction and National Housing, Zimbabwe).STRASSMAN, W. P. 1980 'Housing improvement in an opportune setting:Cartagena, Colombia', Land Economics, LVI, 155-68.TAYLOR, J. C. B. 1985 'Zimbabwe: One local authority's response to thechallenge of decent shelter', in T. Blair (ed.), Strengthening UrbanManagement (New York, Plenum), 24-8.TEEDON, P. and DRAKAKIS-SMITH, D. 1986 'Urbanization and socialism inZimbabwe: The case of low-cost urban housing', Geoforum, XVII, 309-24.UNDERWOOD, G. C, 1987 'Zimbabwe's low cost housing areas: A planner'sperspective', African Urban Quarterly, II, 24-36.UNITED NATIONS 1976 World Housing Survey 1974 (New York, UnitedNations).UNITED NATIONS CENTRE FOR HUMAN SETTLEMENTS 1984a Community-basedFinance Institutions (Nairobi, UNCHS).UNITED NATIONS CENTRE FOR HUMAN SETTLEMENTS 1984b Government PolicyGuidelines to Improve the Efficiency of the Construction Industry (Nairobi,UNCHS).CAROLE RAKODI & N. D. MUTIZWA-MANGIZA 131UNITED NATIONS CENTRE FOR HUMAN SETTLEMENTS 1984C The Promotion ofNonconventional Approaches to Housing Finance for Low Income Groups(Nairobi, UNCHS).UNITED NATIONS CENTRE FOR HUMAN SETTLEMENTS 1987 A New Agenda forHuman Settlements (Nairobi, UNCHS).VAN DER LINDEN, J. 1986 The Sites and Services Approach Reviewed (Alder-shot, Gower).WELLS, J. 1986 The Construction Industry in Developing Countries:Alternative Strategies for Development (London, Croom Helm).WORLD BANK 1986 'Financing shelter: The economic dimension', UrbanEdge, X, ix, 1-6.WORLD BANK 1987 'Making housing markets more efficient', Urban Edge,XI, ix, 1-5.ZIMBABWE BANKING CORPORATION 1987 Economic Review (Harare, The Cor-poration).ZIMBABWE 1983 Transitional National Development Plan 1982/83-1984/85(Harare, Ministry of Economic Planning and Development).ZIMBABWE 1985 Long Term Plan 1985-2000 (Harare, Ministry of PublicConstruction and National Housing).ZIMBABWE 1986a 'Public and private partnership in housing', ZimbabweScience News, XX, iii-iv, 31-9.ZIMBABWE 1986b 'Public and Private Sector Partnership in Housing andUrban Development' (Harare, Ministry of Public Construction andNational Housing and United States Agency for International Develop-ment).ZIMBABWE 1986C 'Low-income Housing Pilot Projects: Kwekwe and GutuProject Evaluation Report' (Harare, Ministry of Public Constructionand National Housing and United Nations Development Programmeand United Nations Centre for Human Settlements).ZIMBABWE 1987a Monthly Digest of Statistics: March 1987 (Harare, CentralStatistical Office).ZIMBABWE 1987b 'National Action and National Shelter Strategies up tothe Year 2000' (Nairobi, United Nations Centre for Human Settlements,submission by the Government of Zimbabwe to the Tenth Session ofthe UN Commission for Human Settlements).ZIMBABWE 1988 Quarterly Digest of Statistics: December 1988 (Harare,Central Statistical Office).ZIMBABWE 1989 Quarterly Digest of Statistics: June 1989 (Harare, CentralStatistical Office).ZIMBABWE 1990 Quarterly Prices Statistics Bulletin: March 1990 (Harare,Central Statistical Office).ZIMBABWE AFRICAN NATIONAL UNION (PATRIOTIC FRONT) 1985 Zimbabwe at5 Years of Independence: Achievements, Problems and Prospects (Harare,ZANU(PF)).