With Zimbabwe's transition to majority rule in 1980, years of restrictions on rural migration to cities ended. Urban population, already high at 23 percent of the total population, was expected to reach 30 percent by 1990, putting serious pressure on the country's major cities, taxing urban services, and creating a severe shortage of housing for the poor.
The Bank's $43 million Urban Development Loan to Zimbabwe (approved in 1984 and closed in 1993) aimed at easing the pressures of urbanization as the country made its transition to self-rule. The project's objective was to increase the supply of affordable housing to low-income urban dwellers in four cities: Harare, Bulawayo, Mutare, and Masvingo, which together represented 70 percent of the country's urban population. The components included (1) strengthening the financial and technical capacities of local governments to provide essential urban services and infrastructure; (2) involving the country's main private financing companies in providing mortgage loans to low-income urban households; (3) strengthening urban transport; and (4) establishing a project management unit.
Project design and implementation
Financing plan and mortgage terms. The project's success hinged on involving Zimbabwe's three main building societies. The largest of them, the Central Africa Building Society (CABS), already had a pilot project to lend to the poor. The project was implemented at a time of rapid devaluation of the Zimbabwean dollar and severe budgetary constraints, which limited the government's ability to commit to its share of financing. With the Bank's help, project designers were able to arrange cofinancing with the Commonwealth Development Corporation (CDC), which reduced the government's required contribution to 22 percent of project costs. CDC agreed to provide about Z$11 million (about US$9 million) on the same terms as the Bank's on condition that the private mortgage lenders (building societies) contributed matching funds of no less than 3 to 1. The building societies responded by contributing about Z$30 million.
The building societies were to provide loans of up to Z$6,000 to qualified buyers, who were required to provide 10 percent down payment on their purchase. The building societies were also to keep account of mortgage repayment and eventually provide all resources needed to finance the mortgage loans. In return, the societies asked to be exempt from the central bank's 23 percent reserve requirement as an assurance against the perceived risk of lending to low-income groups. Although the government initially agreed with their request, it reneged when the country's monetary condition worsened, forcing the building societies to scale down their lending to the poor.
Shortly after loan approval, it became clear that many potential borrowers were unable to raise the required down payment. With the Bank's agreement, the building societies quickly waived the requirement, and they extended the grace period for repayment from three to six months, adding an inexpensive group life insurance to all loans as added protection for beneficiaries.
The Bank also showed flexibility when it agreed to the proposed variable mortgage interest rate of 12.5 percent when inflation was running at 16 percent. But because the cost of loans for the building societies was about 9.5 percent, they were able to make a profit while serving a social need.
Targeting beneficiaries. Given the target population, affordability of plot and housing was a key issue. The project established two income ceilings for eligibility: Z$200 a month and below, for which 70 percent of applicants had to be found, and Z$400 per month, the maximum eligible income level. Project designers believed that the two-tier approach would reach urban households down to the thirteenth percentile of the income distribution. In practice, however, only 45 percent of applicants met the first tier ceiling, compared with the 70 percent targeted, even after the ceiling was raised to Z$300.
The issue of plot and house sizes also related to affordability. The Bank advocated smaller houses of 20 to 29 square meters to maximize affordability, while the Zimbabweans considered houses of 50 square meters on 200 square meter lots to be the socially acceptable size. In the end, the project settled on the bigger houses, but later surveys of beneficiaries showed that they considered the smaller plots to be adequate as well.
The project helped develop more than 20,000 serviced plots, with the eventual construction of as many houses. This was twice the original target, predominantly because the project team used savings accrued from an appreciating US dollar to develop additional plots. Although the project missed the poorest people, it achieved its overall goal of reaching low-income dwellers.
A post completion survey shows that almost 40 percent of the houses cost less than Z$500 (about a month's income). And by renting rooms, about a third of beneficiaries increased their income on average by half. But in 1994, overall shelter costs (including utilities) accounted for more than 40 percent of income for 37 percent of the beneficiaries, partly because of the sharp increase in the variable mortgage rate. Nevertheless, cost recovery was good, largely because the building societies relied predominantly on automatic payroll deductions.
The public sector invested heavily in infrastructure and successfully serviced the plots, although delays in registering plot ownership dogged the project. And changes in the government's monetary policy strained the building societies' ability to maintain their lending levels, with the result that some plots could not be sold as soon as they were serviced.
The project's institution building efforts were less successful. Inadequate staffing made the project unit inoperative for most of the project's duration, and the government's reluctance to use foreign funds for technical assistance hindered efforts to provide needed training.
An integrated approach based on private and public sector partnership can be a viable option in countries with experienced mortgage companies and municipalities dependent on public sector services. The Zimbabwe project succeeded primarily because it involved from the beginning the country's most active mortgage companies in close cooperation with municipal governments, with a clear division of tasks. Moreover, the experience showed that once again private companies are more adept than the public sector at enforcing mortgage repayment.
Flexibility in project design is important. All parties took a pragmatic approach to project design (in waiving the down payment requirement, for example).
Participation of beneficiaries can help establish clearer selection criteria for better project design. The debate between the Bank and the borrower on plot and house sizes and changes in eligibility criteria might have been resolved sooner if beneficiaries had been consulted through surveys early in the project design.