HOUSING11023

Sector Policy Paper May 1975

Se Chx, oi Po4II Cy Papert P{ P I t'I' 4t t ' d I 2 22 I tf vt t v2'

Documrento de pohiitk ~e,ti'g e u OPoii Pfa

sector floorQKttlr Pi-w PD)52mt9t od 0sn§t4 *"ei5f S.3&9 $%-9 22)93 2 3 33t/'

• Ii IPrqjen t o' ic bIvst~6, I o'r tCU

~(AW% Flavoi Lv 4 U*,.S M,

inks. *2t 9 Dv29 9 o s3eB2 S 2C *V 9' 5 * ,S,4 ,'

Secah '2jS'$

Pohh -.e ,S29 IB3's5< 'sitByw25>tat2+S c39yt2;2S >-22.sR6>ss85S sSs.St , 3 ,; ,39 5Y* t'3ink s ut^t s Mayt,Ss n ,,4 2324,S5S2t eS; e' 9 ; '52i>,e 33 925 2 jN62 ^,

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

HOUSINGCONTENTS

Page

Summary .................................................... 3

Chapter 1: Bank Group Interest in Housing and Urban Development . .. 9Chapter 2: The Urban Housing Situation in Developing Countries ... 12

Income Levels ............................................... 12

City Characteristics . .......................................... 13Housing Policy . ............................................ 15

Facing the Situation ................ ......................... 15

Chapter 3: The Economics of Housing ........ .................... 17Housing in the Urban Economy ......... ....................... 18Housing Costs and Incomes .......... ......................... 19Housing Markets ............................................. 24

Chapter 4: Housing Policy Options for Developing Countries ..... .... 26The Policy Framework ............. .......................... 26Zoning Regulations . .......................................... 27Building Standard Controls .......... ......................... 27Pricing Policies . ............................................ 28

Land Tenure ........................... ............. 29Land Purchase and Development ............. ................. 29Taxes on Land ......................................... 29Building Materials ....................................... 30Improving Financial Institutions ............... ................ 30

Public Housing ........................................ 31Chapter 5: The Lessons of International Assistance ...... ............ 32

The Evolution of International Assistance ......... ............... 32

The Experience of the Bank Group .... ........................ 35

Sites and Services Projects ........... ........................ 36Upgrading of Squatter Housing ................. ............... 36

Self-help and Mutual Help ................................... . 38Direct Construction ....................................... 38

The Target Groups ....................................... 39Financial Arrangements for Projects ........... ................. 40Financial Arrangements for Occupants .......... ................ 41Housing Finance ........................................ 41The Institutional Framework ................................... 42Monitoring and Evaluation .................................. .. 42

Impact of Other Bank Lending . ................................ 43

Page

Chapter 6: Directions for Bank Lending for Urban Housing .... ....... 44The Lending Framework .......... ............................ 44Squatter Upgrading and Sites and Services Projects ..... .......... 45Selective Housing Construction ........ ........................ 46Housing Finance . ............................................ 48Coordination of Bank Group Lending for

Housing and Urban Development ....... ..................... 49

Annexes1. International Assistance for Housing ...... .................. 532. Population Growth and Housing Indicators for Selected Cities . .. 623. Cost of Residential Construction for Low- and Medium-density

Housing in Selected Developing Countries ..... .............. 644. Cost of Basic Construction, Land Servicing and Raw Land as Per-

centage of Total Housing Cost for Low- and Moderate-incomeHousing in Selected Cities ........ ......................... 65

5. Estimates of Monthly Household Income Required to Purchasethe Existing Cheapest Complete Housing Unit and Percentage ofHouseholds Unable to Afford It in Selected Cities ..... ........ 66

6. Reduction of Housing Cost and Percentage of Households ThatCan be Served in Selected Cities ....... ..................... 66

7. Estimates of Cost of Housing Units of Various Standards and Loca-tions and Percentage of Households Unable to Afford Them inSelected Cities ............. ............................. 67

8. Percentage of Households Unable to Afford Cheapest Dwelling atVarious Assumed Space Standards in Selected Cities .... ....... 68

9. Percentage of Households Unable to Afford Cheapest Dwelling atAssumed Low Space Standard in Selected Cities ..... .......... 69

10. Effect of Cost Changes on Percentage of Households Unable toAfford Cheapest Dwelling at Assumed Space Standards in SelectedCities ........... ....................................... 69

11. Projects Assisted by World Bank/IDA ...... ................. 7012. Percentage Distribution of Development Costs for Sites and Serv-

ices Plots in Eight Projects Assisted by World Bank Group ..... 7213. Urban Areas by Population Size for Groups of Developing Coun-

tries, 1970 . .............................................. 74

2

SUMMARY

The Economics of HousingBy 1980, nearly a quarter of the people in developing countries-

some 550 million-will live in cities. This number is expected to in-crease to nearly 1,200 million by the year 2000, when about a third ofthe population of the developing world will be urban. The task ofaccommodating this unprecedented increase in the number of urbandwellers and improving their living standards poses a major challengenot only for urban development and housing policies, but for nationaldevelopment in general.

Housing is important to development in both economic and wel-fare terms. It typically constitutes 15% to 20% of household expendi-tures. For all but the wealthy, it is usually the major goal of familysaving efforts. Investment in housing represents up to 20% or 30% offixed capital formation in countries with vigorous housing programs,and it is increasingly recognized as a profitable investment item, yield-ing a flow of income. For some of the self-employed, housing is alsothe place of work. In countries with substantial underutilized labor,material and financial resources, housing can make such resourcesproductive at low cost. The import content of housing construction isusually relatively low, so that multiplier linkages tend to be substantial.

Housing has substantial social benefits, including the welfare effectsof shelter from the elements, sanitation facilities and access to healthand education services. Improved health and education and betteraccess to income-earning opportunities can lead to higher productiv-ity and earnings for low-income families. It is thus for sound eco-nomic reasons that, after food, housing is typically the largest item ofhousehold expenditure for poor families; and that they are willing togo to great lengths to obtain housing at locations with access to em-ployment, even if this means incurring the risks of illegal "squatting."

A positive housing policy can thus make a substantial contributionto economic development and social welfare. The relationship ofhousing to employment, social services, recreation and other aspectsof urban life is spatially important. Access to employment for severalmembers of a family is particularly important for the poor, and so isaccess to other urban facilities, particularly education. High transportcosts, or a long time taken in walking or otherwise traveling to andfrom work, can negate the advantages of cheap housing on the out-skirts of a city. Housing policy is, therefore, a major tool for influenc-ing the efficiency and equity of urban areas.

3

The importance of housing stands in contrast to current housingconditions in most cities of developing countries. A handful of coun-tries have responded imaginatively to housing problems, but in themajority of countries housing conditions are much worse than theyneed be.

The reasons for bad housing conditions are complex, and varywidely among countries. However, at the risk of oversimplification,they may be characterized as stemming from a lack of understandingof the underlying problems, and resulting misguided policy formula-tion. Many countries have attempted to provide unrealistically high-standard dwellings rather than meet the effective demand forhousing. This has led to regulations discouraging the production ofappropriate standard housing, the expenditure of limited publicresources on high-priced housing for a small proportion of the popu-lation, and the condoning or active encouragement of the removalof squatter and similar housing.

Such policies, combined with the limited resources available tomeet the investment needs of rapid urban population growth, havegiven rise to substantial gaps between housing supply and demand inmost cities of the developing world. In the six cities studied for thispaper (Ahmedabad, Bogota, Hong Kong, Madras, Mexico City andNairobi), from one-third to two-thirds of the population are unableto afford the lowest-cost housing presently being produced by thegovernment.

Solutions to housing problems differ greatly among cities and coun-tries. However, in most countries, the adherence to unrealistically highstandards for new housing construction and the refusal to accept exist-ing low-qualitydwellings, even as an interim solution, are the principalobstacles to better housing. The result is a cycle of construction anddemolition. The poor, unable to construct or purchase dwellings ofacceptable standard, are pushed to ever more distant peripheries.

Although inappropriate housing standards are most important,other imperfections also affect the operation of housing markets. Thesupply of urban land and urban services, such as water, sewerage andelectricity, tends to be constrained and skewed toward upper-incomegroups. Housing finance institutions are typically undeveloped. Thislimits the supply of housing for middle-income groups, and oftenmeans that housing provided for the poor is "raided" by higher-income families.

Income levels, modified by the structure of income distribution,are the most important determinant of housing options. They set thelimit on what the poor can afford. But because of spatial considera-tions, city size is also very important. Housing must be considered

4

together with income earning opportunities, and hence in relation totransport costs. The growth rate of cities affects the rate at which landcan be serviced. As the largest cities also tend to be the fastest grow-ing, growth often exacerbates spatial problems. However, even inrelatively small cities, high growth rates pose a difficult challenge tohousing policy. Finally, general economic policies, as well as thoseparticularly concerned with housing, play a critical role in determin-ing the efficiency of housing markets.

This paper approaches the housing situation from the standpointof effective housing demand. The costs of dwelling units of varyingsizes, standards and locations are compared with the ability of low-income groups to pay. It is suggested that housing (including sites andservices and other self-help methods of construction) could be pro-vided without significant subsidy for at least 80% of the populationof most cities. It appears that in relatively higher-income countriesmost of the lowest 20% could be reached with small subsidies. In verylarge, lower-income cities it is more difficult for the poorer segmentsof the population to be reached. In these instances, squatter areaupgrading appears to hold the most promise.

Though incomes may be increased, and to some degree redistrib-uted over time, the only way to bring housing within reach of sub-stantially increased numbers within a relatively short period is to bringthe supply cost down. This can be achieved, in the first instance andmost rapidly, by reducing standards: permitting and encouraging theuse of lower-cost (frequently indigenous) building materials and alower quality of finish; providing communal rather than privateplumbing and sanitary facilities; encouraging or providing higherdensity construction, with less land per dwelling unit; and providingless living space per dwelling unit. Without being carried to extremes,these steps-singly or in combination-can bring the cost of housingdown so that a much higher proportion of the population than cur-rently served can afford housing with access to employment. Landcosts, and the efficiency of supply of the various components of hous-ing, particularly of construction, also influence the housing the poorcan afford. Self-help is an obvious way of bringing down financialcosts. In some situations, some complementary construction is neededto utilize self-help appropriately. Medium- or high-rise buildings maysometimes be the most efficient way of reaching the poor in circum-stances in which land costs are high, construction costs are low, suchhousing is culturally acceptable, and the location provides the poorwith cheap and ready access to employment.

These policies, it should be stressed, are only effective when co-ordinated in the context of urban and national development. Ex-

5

change, foreign trade, commercial, labor, monetary, credit and fiscalpolicies influence a country's overall competitiveness, the flow ofprivate and public funds to the housing sector and the coordinationof the housing "package." Such policies, along with zoning and landdevelopment policies more directly affecting housing, also havespatial dimensions influencing the range of city sizes within eachcountry and the locations of residences and places of employmentwithin cities. An improvement in these policies, complementary tomore direct measures, can substantially improve the housing situation.

International ExperienceMany of the early efforts to provide international assistance for

housing suffered from the same misconceptions about the nature andcomplexity of housing problems that prevented most governmentsfrom dealing with them effectively. These efforts did not attempt toprovide large resource transfers. They did not expect to have a sub-stantial direct effect, but they expected to be seminal, and particularlyto have an impact on the provision of low-income housing. However,there was a tendency to underestimate the complexity of the problem,and to see it in terms of a shortage of housing finance, without payingadequate attention to the supply of housing of an appropriate stand-ard at appropriate locations. The financial assistance provided thusoften went to lower middle- and middle-income groups, rather thanto lower-income groups; the bulk of the population in urban areaswas not affected.

The World Bank Group1 entered this field, in the 1970s, as a late-comer. It has, as a result, benefited from the experience of others-both governments and international agencies. It has from the outsettreated housing problems within the broader framework of urban andnational development problems, and it has proceeded cautiously andpragmatically, according to a policy of learning by doing.

Bank Group operations in housing have so far taken the form oflending for serviced sites, on which occupants contribute to the con-struction of their own homes, and, increasingly, lending for the up-grading of squatter settlements. In the sites and services approach,land plots are leveled and furnished with access roads, drainage,water, sewerage, electricity and a variety of other individual as well ascommunity services. Particular attention has been paid to locatingprojects within reach of major markets and employment opportuni-ties, and to bringing activities and jobs to the project area.'The 1World Bank Group includes the International Bank for Reconstruction and Develop-ment, popularlv known as the World Bank, the International Development Association(IDA), and the International Finance Corporation (IFC). The fiscal year of the three institu-tions runs from July 1 to June 30.

6

The essence of squatter housing upgrading is to conserve the exist-ing low-income housing stock, particularly in serviceable, relativelycentral locations; and to improve this housing by including additionsand improvements to services and facilities within the entire neigh-borhood. This approach, a complement to sites and services programs,is seen as an increasingly important means of providing benefits tolow-income people and of maintaining their access in relatively cen-tral locations.

So far, the World Bank Group has approved assistance for 10 urbandevelopment projects involving housing. The projects are in Senegal,Nicaragua, India, Botswana, Jamaica, Tanzania, Zambia, Indonesia,El Salvador and the Republic of Korea. These projects (or their hous-ing components) have been relatively small, entailing loans or creditsof less than $25 million, with some loans being as small as $3 million.

Although self-help and mutual-help methods have been appliedas far as possible in the construction of dwellings and some commu-nity facilities, experience has demonstrated that some direct construc-tion is essential and complementary in both approaches. Self-help issometimes technically impossible or economically inefficient. In someinstances, core units may be required to provide shelter from theelements-rain or cold-while the construction process is completed.Families whose homes must be removed in the process of upgradingsquatter areas may require that new, relatively high-density housingbe constructed for them in the area so that they will not suffer costlyremoval to distant sites. Responding to the housing problems of thepoor in a variety of circumstances-to bring standards down and toobtain accessible locations-requires a flexible approach and a fullrange of lending instruments. It is in this light that future directionsfor Bank Group lending are considered.

RecommendationsThe growing recognition of the importance of urban development,

and particularly of the need for equitable development, is leading toincreased demands for Bank Group assistance for housing in the con-text of urban development. The approach of "learning by doing" en-sures that the Bank Group will be able to respond adequately to theurban development needs of member countries in the future. BankGroup assistance for housing is limited to countries and cities broadlycommitted to alleviating the housing conditions of the poor as part ofefficient and equitable urban development. In this context, it is recom-mended that:

1. Squatter upgrading and sites and services projects should con-tinue to be the prime instruments for improving the housing con-

7

ditions of the urban poor. Within this framework, an open attitudetoward housing construction is necessary to ensure efficiency and, inappropriate circumstances, to facilitate more economic use of landand improve access to income earning opportunities.

2. In some cities, the appropriate response to housing for the poorinvolves conventional construction. This may be in circumstances inwhich land cost, construction cost and access to employment makemedium- to high-rise housing the most efficient and equitable solu-tion for low-income earners, provided such solutions are sociallyacceptable and part of a city's overall approach to housing the poor.Lending for conventional housing would form a small proportion oftotal lending for shelter. It is not envisaged that the Bank Group wouldbecome involved in lending for mass housing construction in theforeseeable future.

3. The Bank Group has a particular role to play in responding torequests for assistance in building housing finance institutions. Lend-ing for low-income housing project beneficiaries is one avenue, butit is relatively limited. The Bank Group should continue to provide"seed capital" for appropriate housing finance institutions with theobjective of promoting better housing solutions for all income groupsover time. In addition, it is recommended that "seed capital" shouldbe provided for mortgage insurance institutions to facilitate lendingfor housing and enable lower-income groups to be reached.

Monitoring, evaluation and research should continue to be an inte-gral component of the urban development program. Increasing atten-tion should be given to the coordination of the Bank Group's lendingfor urban development to maximize the benefits of such lending, andto prepare the way for a larger urban development effort in the future.

8

Chapter 1: BANK GROUP INTEREST IN HOUSINGAND URBAN DEVELOPMENT

By 1980, nearly a quarter of the people in developing countries-some 550 million-will live in cities. This number is expected toincrease to nearly 1,200 million by the year 2000, when about a thirdof the population of the developing world is likely to be urban1 (seeTable 1). Urban development thus poses a major challenge for thedecades ahead.

The Urbanization: Sector Working Paper published in 1972 dis-cussed the World Bank Group's role in assisting member countries

Table 1

Estimated Urban Population of Developing Regions1960, 1980, and 2000

(in millions)

Average annualgrowth rate

1960-1980 1980-2000Region 19i0 1980 2000 % %

Asia(') 134 309 647 4.3 3.8

Latin America 69 163 342 4.4 3.8

Africa 31 77 190 4.7 4.6

Total 234 549 1,179 4.4 3.9

Note: Urban areasaredefined aslocalitiesof 20,000or moreinhabitants. Theurban populationfiguresgivenherethussometimes exclude municipal unitsof less than 20,000 inhabitants within urban areas. Growth rates are regionalaverages; those of principal cities are typically higher than these averages (see Annex 2).

") Excludes the People's Republic of China.Source: United Nations, Growth of the World's lUrban and flural Po,onlation, 1920-2000 (New York, 1969).

to meet this challenge efficiently and equitably. A large part of BankGroup lending has always been urban-oriented,2 and much of it hasbeen concentrated in major cities. In recent years, however, theemphasis has shifted from merely expanding a sector or sectors withinurban areas, to "improving the efficiency of the urban centers bothfor production and for living."3 Projects have provided urban re-sources such as transport and water supply in a framework of urbandevelopment, with special attention to the needs of the poor. A sub-stantial increase in support for urban development has been proposed.

'Urban areas are defined as localities with 20,000 or more inhabitants.2World Bank, Urbanization: Sector Working Paper (Washington, D.C., 1972), p. 53.31bid., p. 55.

9

Housing is important to development in both economic and wel-fare terms. Investment in housing is a significant component of totalinvestment. For lower-income groups, it is the largest item of house-hold expenditure after food and, for all but the wealthy, it is usuallythe major goal of family saving efforts. In urban areas, the housingpackage of land, shelter and utilities-in combination with access towork and educational and social opportunities-markedly affects theproductivity of the poor, as well as their welfare. Its location can alsoalter the spatial layout of cities. Jobs can be made more or less acces-sible, traffic problems can be aggravated or relieved.

The Bank Group has been marginally involved in the provision ofhousing for many years in port, mining and land settlement projects.In such projects, however, housing has not been a central feature, andit has formed a small proportion of total project expenditures. Theapproval of the first sites and services project in June 1972 markedthe Bank Group's entry into assistance for urban housing as the primeobjective. This approach has reflected the Urbanization:SectorWork-ing Paper's emphasis on "learning through doing." Substantial expe-rience has been accumulated in the design of these projects, and inthe understanding of housing in urban development through research.

This paper evaluates the experience of the Bank Group's initial ef-forts in housing in the light of the evolving housing situation in de-veloping countries. It is confined to urban housing, because villagehousing problems are more properly discussed in the context of ruraldevelopment, and it places particular emphasis on responding tothe housing problems of the urban poor. Like the Urban Transport:Sector Policy Paper, this paper is part of an intensified investigationwhich is required for a deeper understanding of urban developmentas a whole. It represents a continuation of efforts to assess the con-tribution made by the Bank Group and others to improving housingconditions in developing countries.

The central purpose of the paper is to examine: (1) whether in thecontext of improving the housing situation of the poor, a broadeningof the instruments of lending for housing-in particular to includelending for construction-is advisable; and (2) what other measurescan be taken to improve the Bank Group's involvement in housing.The circumstances in which assistance for housing is likely to be par-ticularly appropriate are also considered.

To answer these questions, it was necessary to assess the currenturban housing situation in developing countries; analyze the rela-tion between urban housing and social and economic development;evaluate the functioning of urban housing markets with particularconcern for the effects on low-income families; and consider current

10

policy responses to housing problems. A special study of housing con-ditions in six cities (Ahmedabad, Bogota, Hong Kong, Madras, MexicoCity and Nairobi) was undertaken to provide information about thehousing that poor people can afford. An analysis of the housing situa-tion is presented in the next three chapters of this paper. The past roleof international assistance in the housing field, with particular atten-tion to the Bank Group's experience, is then reviewed. This paperconcludes with recommendations for directions for Bank Group lend-ing for housing.

11

Chapter 2: THE URBAN HOUSING SITUATIONIN DEVELOPING COUNTRIES

The houses of the Mediterranean, the flat-roofed dwellings of theMiddle East, and the pitched roof, stilt-supported houses of SoutheastAsia are typical of the variety of styles that enrich the human heritageand form an essential part of the "quality of life." Behind this diversity,however, common uses of housing reflect similarities in living ar-rangements based on family needs for sleeping, eating, child rearing,leisure and earning a living. The vast majority of city dwellers findsome form of housing, but in most cities it ranges from spacious dwell-ings for the rich to the miserable hovels of the very poor. Housing isas complex as the societies it serves, but several key factors clearlydominate the housing situation in all cities.

Income LevelsIncome is the most important determinant of housing. At the

household level, it determines the housing a family can afford. Innational terms, it reflects a country's capacity for housing its popula-tion at standards which will not distort other investment allocations.The distribution of income among households determines the num-ber of poor families, and the extent of their poverty. The spatial dis-tribution of income affects the housing characteristics of a city; thereare relatively rich cities in low-income countries, and relatively poorcities in high-income countries.

Relatively high-income countries (those with annual per capitaincomes of more than $450) generally have a vigorous private housingmarket, which usually provides serviced housing with security oftenure for well over half the population. In addition, some of thecountries in this category have made a substantial contribution tohousing lower-income groups through public housing. Hong Kongand Singapore are notable in this respect, with some 40% of theirpopulations living in public housing. Many of these countries have,however, failed to stimulate private housing for middle- and upper-income groups and have almost totally neglected housing for thelower-income groups.

Middle-income developing countries (those with annual per capitaincomes ranging between $150 and $450) account for almost half theurban population of developing countries. They include many largecities, some high-income cities, and half the population in cities overtwo million. Their housing problems are extremely diverse. Some

12

cities, for example Sao Paulo, have successfully stimulated the privatemarket so that about 80% of the population is housed in serviceddwellings. Other, less affluent cities have also shown that the privatemarket generally makes a substantial contribution to housing, butunfortunately many of the countries in this category have not pursuedpolicies conducive to stimulating private housing. The upgrading ofsquatter housing and sites and services approaches can provide hous-ing for lower-income groups, but many countries have rejected theseapproaches on the ground that they would create or perpetuate theexistence of "slums."

The poorest developing countries (those with annual per capitaincomes of less than $150) may be conveniently divided into twogroups. The first consists of four large Asian countries-Bangladesh,India, Indonesia and Pakistan. They have relatively low rates of growthof urban population. However, the size and number of their citiesare very large, and so many people need to be housed that the prob-lems seem overwhelming. Squatters and slum dwellers in Calcutta(1,720,000), Jakarta (1,125,000) and Karachi (811,500), for example,outnumber the total populations of some countries. The second groupconsists primarily of the sparsely populated, less urbanized Africanstates. Even their principal cities are quite small, with populationsgenerally below 100,000, but their rates of urban population growthare very high. This growth has rarely been able to stimulate an ade-quate government response to housing problems. As a result, theproportion of the total population living in slums or uncontrolledsettlements is very high.' However, only very limited new resourcescan be devoted to improving housing conditions in these countries.

City CharacteristicsCity size is a second major factor in housing. The largest cities,

notably those with populations above two million, confront thegravest problems, because it is here that spatial imbalances in thelocation of people and jobs are most severe. The scarcity of land withgood access to employment opportunities is the central feature of

'Traditional housing "deficit" estimates, which use data on slum and squatter housing asindicators of inappropriate housing (see Annex 2), tend to overstate the seriousness of thehousing problem. Although such housing may be illegal, built from traditional materials,or both, it is not necessarily of an unacceptably low standard. Much of this housing, inmost cities, provides both adequate shelter and good access to employment. Some isquite substantial. The study of six cities conducted for this paper suggests (see Table2) that, in the lowest-income countries, the proportion of urban households currentlyunable to afford housing may approach, or even exceed, two-thirds. This measure, too,overstates the problem. Being based on the cheapest housing currently being produced bythe public sector, it implies that the private sector does not produce lower-cost housing,and that the gap between incomes and housing costs has persisted for some time. Re-search and observation in the same six cities and elsewhere suggest that in practice thehousing situation is not always as bad as such figures imply.

13

their housing problem. This scarcity is often worsened by large areasof such land being either kept vacant or devoted to uses whichdo not require a central location. Of course, land is more readilyavailable on the outskirts of these cities at much lower cost. However,such areas are generally not well integrated with the city as a whole.

Table 2

Percentage of Households Unable to Afford the CheapestDwellings Presently Available in Selected Cities

(percentage of population not servedh

Mexico City 55

Hong Kong 35

Nairobi 68

Bogota 47

Ahmedabad 64

Madras 63

Source: Annex 5.

There are seldom substantial employment opportunities nearby, in-frastructure provision is inadequate and transport is usually verycostly. In contrast, in smaller cities, even centrally located land maynot be expensive, and land on the outskirts is usually very cheap. Thisdoes not mean that large cities are uneconomic. A large range of citysizes is in practice functional.

The rate of city growth is a third important factor in the housingsituation. Many cities in developing countries have grown more rap-idly in the last 25 years than at any previous time in human history.Large cities, moreover, are among the fastest growing. Most havebeen deficient in housing the poor at standards offering protectionfrom the elements and sanitation at a cost they can afford. A minimumof 630 million persons is likely to be added to the urban populationsof developing countries by the year 2000 (see Table 1). Associatedchanges in the age structure of populations ensure even more rapidrates of household formation and increases in the effective demandfor housing, making solutions which meet this demand all the moreurgently needed. These estimates are consistent with a rapid accelera-tion in rural productivity and living standards which is more thanmatched by productivity gains in urban areas, maintaining a relativelyhigher standard of living in cities and resulting in continuing high rates

of migration.Topography, climate, social structure and other factors also affect

housing, but in a modifying rather than determining manner. Hilly

14

and marshy land, a site surrounded by mountains or the sea, and sitessusceptible to earthquakes are all relatively difficult to develop.

Housing PolicyThe fourth critical variable in housing conditions is housing policy.

Variations in housing conditions among cities similar in rates ofgrowth, size and income reflect differences in housing policies andtheir administration. Most attempts to date to deal with housing havetaken the form of calculations of the investment required to meet thepresent and prospective housing "deficit." This approach is basicallyunsound. Apart from the conceptual difficulties of estimating "appro-priate" housing for a variety of socioeconomic conditions, it suggeststhat the housing problem is a "bottomless pit." By ignoring the poten-tial for incremental improvement through a better use of existingresources, it has thus encouraged the adoption of counterproduc-tive policies.

The most common such policy is a cycle of demolishing low-in-come urban neighborhoods and using the land for other purposes.Refusal to accept existing low-quality housing as at least an inter-mediate solution to the urban housing problem is common. The con-sequence is a continuous process of construction of temporary dwell-ings which last until the government clears the land for other purposesor until they are washed away by tropical downpours. Thus, as thevalue of land close to the city center increases, squatters are forcedto move toward the expanding periphery away from employment andother opportunities, never along the way having had the opportunityto build up equity in a house.

Many governments have insisted on maintaining high standardswhich raise the cost of housing and prohibit self-help constructionby low-income households. Prohibitive building codes, costly landacquisition procedures and other barriers prevent the poor frombuilding permanent legal houses where they can earn a living. Thesentiment "construct big, beautiful and forever," is not unusual. Thepoor, who are frequently described as "marginal" by those who resentslum and squatter areas, are thus "marginalized" by policy failure.

Facing the SituationInteraction of the principal determining factors-incomes, city size,

the rate of city growth and housing policies-results in a wide rangeof housing conditions. However, income factors are dominant, forthey determine how much households, cities and countries can affordto spend on housing. At present, inappropriate policies often limit

15

the housing available, forcing even middle-income, but particularlylower-income, families to live under worse housing conditions thanresources could permit. Given particularly the rapid urban populationgrowth forecast for the next decades, a more realistic approach tohousing problems is needed if developing countries are to improvetheir housing situation.

16

Chapter 3: THE ECONOMICSOF HOUSING

Housing has multifaceted economic and social characteristics. Itrepresents some 15% to 20% of household expenditures, and it isthe major focus of saving for all but high-income families. It is in-creasingly recognized as a profitable investment item, yielding a flowof income. For some households, it is also a place of business. More-over, it is spatially fixed, making its location relative to jobs, transportnetworks and public services as much a part of its essence as thematerials used in its construction.

Investment in housing can have an important impact on incomeand employment through the utilization of unemployed or under-employed workers. Moreover, the import content of housing con-struction and of the manufacture of construction materials is usuallyrelatively low or potentially low; and the supply of domestic rawmaterials for housing is generally elastic. Multiplier linkages throughthe economy can be substantial. In the Republic of Korea, for exam-ple, approximately 14 additional jobs are created for every $10,000spent on the construction of dwelling units. Housing constructioncan account for up to 20% or 30% of total fixed capital formation inrapidly growing countries with vigorous housing policies.

Housing is thus a tool for macroeconomic development, whichshould be viewed as complementary to other sectors. The emphasiswhich should be given to housing development in a national or anurban economy will depend upon the strength of demand and theelasticity of supply-that is, the degree to which the particular hous-ing produced can draw on underutilized resources.

Housing also has social benefits. An inadequate dwelling, absenceof water and sewerage facilities, or lack of access to income earningopportunities, contribute to low family incomes, poor health' anda low ability to absorb education. Well-planned housing, on the otherhand, can increase national productivity, economize on urban spaceand minimize the cost of urban infrastructure. Improved location ofdwellings in relation to jobs leads to reductions in traffic congestionand increased household take-home pay by reducing commutingexpenses.

Housing as a package of shelter and services must thus also beviewed in terms of substitution and complementarity. The housing-

'World Bank, Health: Sector Policy Paper (Washington, D.C., March 1975).

17

transport relationship is an important case in point. In some instances,either instrument may pose a different, but equally acceptable, solu-tion to a particular problem. In others, they are best applied together.The relationship between housing and health is more clearly com-plementary. In a squatter area, for example, improvement of thehouse, provision of water supply and sewage disposal and the intro-duction of health clinics may seem to be competing for resources,but the realization of benefits available from a health investment maydepend upon some improvements in housing.

Given its costliness and durability, however, there is a limit to theamount of housing construction a country can afford during anyperiod of time. The emphasis must therefore be on (1) mobilizingunderutilized resources; (2) reducing unit costs by increasing theefficiency of construction or reducing standards; and (3) improvingthe interaction among housing, land use and transport in integratedurban development.

Housing in the Urban EconomyHousing has several components. The usefulness of land depends

upon its location and area; the services, such as water, sewerage,electricity and roads, supplied to the plot; and transport available atthat location. Land and the materials, capital and labor required forconstruction all have uses other than for housing, and their coordina-tion on the site is a substantial problem which few societies have dealtwith satisfactorily. The components together determine the total costof the dwelling, and the mix of components may be adjusted to alterthe cost per unit. The result is a wide range of housing types, whichemphasize the substitutability of one component for another.

Location, and hence land, is clearly of critical importance to hous-ing. A dwelling located far from employment opportunities and so-cial services is more expensive than an identically priced unit withbetter access, because travel costs are part of the cost of living at thatlocation. Poor families are well aware of the value of locations whichgive access to jobs (including sources of income to secondary familyearners and supplemental incomes to primary earners), and accessto services such as health and education which affect productivity aswell as welfare. It is thus for sound economic reasons that familiesare willing to go to great lengths to obtain housing at desirable loca-tions, even if it means accepting the risks of illegal "squatting."

The reasons for social investment in housing are equally soundlybased. Housing can modify urban form and community developmentto improve equity and efficiency. It can take a variety of forms. Accessto income earning opportunities is sometimes best improved by in-

18

creasing the low-income housing supply near employment concen-trations; sometimes it is better to locate industrial and commercialactivities near low-income housing concentrations; sometimes linkingexisting residential and employment areas by improved transport isthe best approach.

Housing Costs and IncomesTo be effective, housing policies and programs have to be tailored

to a country's income level and households' capacity to pay. Manyhousing programs have met with little success because they have at-tempted to meet housing "needs" rather than the effective demandfor housing. In contrast to estimations of need according to arbitrarystandards, effective demand is derived from a household's ability andwillingness to pay for housing.

At present, there are wide gaps in most developing countries be-tween the cost of currently produced housing and the ability oflow-income urban families to pay for it. In the six cities studied,the cheapest housing units currently produced by the public sectorranged from $570 to $3,005 (see Annex 5). From 35% to 68% of urbanresidents were unable to afford the full costs of such housing (seeTable 2).2

A 20% reduction in costs would enable the market to reach anadditional 18% of households in Madras and roughly 10% in theother cities, but only an extra 5% in Nairobi. To reach the bottom20% of households, costs would have to be reduced by nearly 70%in Nairobi, but only 20% in Hong Kong and approximately 50% inthe other cities. (See Annex 6 for the foregoing and for reductionsthat would be necessary to reach the bottom 40%, 20% and 10% ofhouseholds in all the cities.)

Rising per capita incomes and improvements in their distributionmay in time make this picture less gloomy. Time, however, is limited.For the current generation of urban dwellers, measurable progressrequires a pragmatic approach to attaining lower production costs.Too often, a government's legitimate desire for good-quality housingfor its citizens prevents seeking ways of building decent dwellingsat lower cost; however, cost reduction can be achieved, in the firstinstance, and most simply, by reducing standards. Indigenous mate-rials and traditional building methods can be substituted for higher-cost materials. A lower quality of finish can be accepted. Interior

'These figures assume that poor families will spend 15% of their household budget onaccommodation at a real interest rate of 10%. They thus also imply that "affordability"means ability to purchase over time on these terms; this is not a normative judgment,i.e., that ownership is in all cases preferred to rental, but simply reflects the most straight-forward approximation of what can be afforded.

19

living space can be reduced. Communal kitchens and sanitary facili-ties can reduce the cost to each household.

Substantial savings can also result from reducing the land areadevoted to each dwelling. This can be achieved by reducing theplot size for single family detached and semidetached houses in sitesand services development; by designing houses, such as two-storeyrow houses, which share walls on both sides; and/or by constructingfour- to six-storey walk-up apartment buildings. Land use is thus animportant consideration in the design of housing and residentialareas. It raises the question of trade-offs between single family andmultifamily structures and between the use of self-help methods toutilize labor with low opportunity costs and multistorey formal con-struction to utilize land more efficiently and reduce its weight in totaldwelling costs. Families usually have at least the notional choice be-tween relatively roomy and good-quality shelter at locations withpoor and costly access to income earning opportunities, and verycramped living conditions in locations with good access.

The nature of the trade-offs involved and the provision of alterna-tive housing options for low-income families in urban areas of widelyvarying characteristics have to be taken into account in housing deci-sions. Data for such decisions are usually not readily available, andall the options therefore cannot be detailed in this paper. A limitedapproach, based on information collected for the six cities studied,can nevertheless indicate the effects of changes in standards whichcan realistically be achieved in housing of varying location and formalconstruction content. This is accomplished by comparing cost andincome data on the assumptions that (1) livable space per household,(2) land area per household, (3) public service levels and (4) locationcan be brought into a realistic relationship with the income levelsof poor households. The comparisons of dwelling costs with house-hold incomes in Annexes 7 through 9 illustrate the range of housingoptions potentially available.

These comparisons take as their point of departure the actual costsof the cheapest single family and multifamily public housing unitscurrently built in the cities studied (see Annex 5). Hypothetical re-ductions in the costs of these units were calculated on the basis ofthe following assumptions: (1) that interior livable space would bereduced to 20 square meters per dwelling unit; (2) that gross densitieswould be increased by associated changes in land use standards (seeAnnex 8); (3) that reductions in the cost of services and constructionwould follow three different service level standards -individual,shared and basic (see Annex 7); and (4) that all costs would be cal-culated on the basis of land prices presently prevailing at each loca-

20

tion. The effects of changes in location were examined using landprices at the city center, at intermediate locations (defined as one-third the distance from center to periphery), and at the periphery (seeAnnex 7). The resulting changes in the income groups that might bereached by these approaches are recorded in Table 3.

Administrative and maintenance costs are included for the HongKong examples, which are based on actual, full-cost experience.Particularly where ownership is not an option, as in multistorey hous-ing, the management costs of public housing may be important.Therefore, a cost increase of one-third, to account for these costs, aswell as interest during construction and possible delinquencies ordefaults in rental payments, was calculated as a sensitivity measure(see Table 4). Vigorous housing policies may be presumed to stim-ulate the efficiency of supply, so that similar cost reductions areincluded in the sensitivity analysis.

These examples illustrate the importance of city per capita incomelevels and their distribution, of city size (and hence of land values),and of the efficiency of supply of the various housing components indetermining the income groups that can be reached by housing pol-icies. The priority attached to housing low-income groups-especiallythe extent to which governments are willing to reduce standards-is

Table 3

Households Unable to Afford Cheapest Housing PresentlyAvailable and with Assumed Cost Reductions in Selected Cities(')

(percentage of population not served)

Cheapest dwelling Assumed cost reductions

presently available Space standard Space standard(from Table 2) A B

Mexico City 55 4 3Hlong Kong 35 25 18Nairobi 68 47 46Bogota 47 11 8Ahmedabad 64 41 36Madras 63 38 35

(1) Dwelling unit cost derived from cost of cheapest currently built public housing, assuming peripheral location, basicservices and space standards given below.

Space standard A: Singlefamily-20 sq. m. livable space per family, 75 sq. m. land. (This is the low-cost solution inBogota and Ahmedabad.) Multifamily-20 sq. m. livable space per family, 15 sq. m. share of land. (This is the low-cost solution in the other cities.)

Space standard B: Single family-20 sq. m. livable space per family, 40 sq. m. land. (This is the low-cost solution inBogota and Ahmedabad.) Multifamily-20 sq. m. livable space per family, tO sq.m.share of land.(This is the low-costsolution in Mexico City, Hong Kong and Madras. Single family and multifamily are equally low-cost in Nairobi.)

Source: Annexes 5 and 8.

21

Table 4

Households Unable to Afford Housing under Assumed SpaceStandards and Alternative Cost Assumptions

(percentage of population not served)

Space standard B Cost reduction Cost increase(from Table 3) of one-third' of one-third(

2)

Mexico City 3 1 7

Nairobi 46 29 61

Bogota 8 4 13

Ahmedabad 36 13 53

Madras 35 11 68

(1) This is also equivalent to 20% of household income devoted to housing instead of 15%.

(2) Allows for administrative and maintenance costs, interest during construction and 20% rate of delinquencies anddefault on rental payments.

Source: Annexes 8 and 10.

also an important factor. If relatively high space standards are main-tained, the option permitting the greatest proportion of the popula-tion to afford dwellings will always be at the periphery. The examplesillustrate that this may be achieved sometimes with single family andsometimes with multifamily housing, depending upon the mix offactors mentioned above. The essential task for housing policy is,however, to develop the trade-offs between other options for furthercost reduction; and this too depends upon all of these factors.

Income level is the first important consideration. In relatively high-income cities, a fairly broad range of housing options is, in most in-stances, technically feasible; and the bulk of the population may bereached with reductions in one or more standards which are sub-stantial but not so great as to reduce the housing produced to unac-ceptable levels (see Annex 8). Furthermore, the proportion of thepopulation that can be reached in relatively high-income cities doesnot appear to be very sensitive to changes in costs or the percentageof family income devoted to accommodation.3 Thus, for example, inMexico City and Bogota, the percentage of the population whichcould be accommodated does not change markedly with either a costincrease or a cost decrease of one-third (see Table 4).

In the low-income Indian cities, and in Nairobi, the percentage ofhouseholds which could not afford similar housing solutions appearsto be much larger, and also more sensitive to changes in costs. In theIndian cities, this result is attributable chiefly to low incomes. Thus,despite the lowest construction costs in the world (see Annex 3), morethan one-third of families would be unable to pay the full costs of3 Except in special circumstances as in Hong Kong, where low-density housing develop-ment at the extremes of the urban-rural fringe is not permitted.

22

what are defined as low-standard dwellings, with basic services at theperiphery. In Nairobi, despite higher incomes, the figure is nearly one-half, a result which appears to stem primarily from the high local costsof construction and construction materials.

The importance of city size and layout, operating through theireffects on land values at central and intermediate locations, comesout most clearly in the case of Mexico City. Thus, unless they areamong the relatively few short-term beneficiaries of rent control, thepoorest half of the population could afford housing at the standardsfor which costs were estimated only at the periphery. In Hong Kongand Bogota, by contrast, despite lower incomes and relatively largepopulations, substantial numbers in the lower-income groups appearable to afford intermediate locations. In India, these groups clearlyhave a wider range of options in Madras with its extended layout thanin Ahmedabad with its highly concentrated city center.

The figures suggest that the problem in the higher-income cities isin principle manageable; that is, a reasonable reduction in standardswould enable the bulk of the population to be housed at peripherallocations. However, particularly in very large cities such as Bogota,Hong Kong and Mexico City, this may not be the best solution. Someportion of the lower-income groups could afford intermediate loca-tions in row housing or multistorey apartments. However, it would bedesirable to provide a broader range of options-including differentstandards, locations and secure tenure arrangements-under whichlarger numbers might be able to afford housing with better access toincome earning opportunities for the entire family. Several possibil-ities emerge: (1) programs which encourage or enable families todevote a larger percentage of their expenditures to housing; (2) in-creasing densities and/or reducing interior living space still further(see Annex 9); or (3) introducing conscious spatial policies, for ex-ample encouraging industrial and commercial activities to locate atthe periphery or, alternatively, making higher-value land accessiblefor low-income housing through subsidies, whether explicit (as in thecase of the Hong Kong and Singapore public housing programs) orimplicit (as in squatter area upgrading programs). Finally, there arenumerous market imperfections which can be reduced or removed,thereby improving the efficiency of supply of the various housingcomponents and of the materials entering into their construction.

In the lower-income cities, the problem remains one of providinghousing solutions for the poorest segment of the population. In theseinstances, many of the same measures may be considered, thoughwith different degrees of emphasis. For example, explicit subsidies ofany size are beyond the fiscal capacities of governments. Furthermore,

23

in these cities, very low-standard serviced sites and squatter housingupgrading may provide the only means of improving housing oppor-tunities for large numbers of the urban poor.

The preceding illustrations should not be interpreted as beingindicative of where and how the poor should live within cities. Nor dothey establish an "optimal" residential density or suggest circum-stances in which higher densities are required. For one thing, presentinformation about the social costs imposed by lack of access to em-ployment opportunities is too limited to permit a more careful weigh-ing of the trade-offs between peripheral and more central locations.Even if full costs were known, the results would be no more thanindicative of what could be accomplished in a variety of political andsociological contexts. Actual solutions will vary with income levels,traditions, personal preferences, political structures, urban develop-ment settings and relative costs of each city. The perceived privatecosts of the housing solutions proposed will have to be weighedagainst those of squatting and other options open to low-incomefamilies if the proposed solutions are to be socially successful.

The data collected for this study are designed to show what mightbe accomplished by reducing standards, improving markets in otherways, changing locations and densities and utilizing self-help meth-ods. In all cities studied, the examples suggest that housing (includingsites and services and other self-help methods of construction) couldbe provided without significant subsidy for at least 80% of the urbanpopulation (see Annexes 8 and 9). In the wealthier cities, it appearsthat many of the lowest 20% could be reached within acceptablestandards and with a low subsidy element. In Hong Kong and Sing-apore, they are being reached now, with some measure of subsidizedaccess for the lowest third of the population. In the lower-incomecities, it is more difficult for the poorest segment of populations to bereached by formal housing construction solutions. Market improve-ments should be stressed, because they can have great importanceover time; however, squatter area upgrading programs and sites andservices approaches, stressing low space standards and high propor-tions of self-help construction, appear to hold the most promise atpresent.

Housing MarketsHousing markets have special characteristics and they are very im-

perfect, with wide divergences between the private and social costsof production and limits on the flow of resources to the entire housingsector, and to low-income households in particular. The scarcity ofwell-located land, the high capital costs of servicing land and the

24

durability of housing make for a relatively inelastic housing supply;increases in urban population, family formation and income levelslead to large shifts in demand which frequently outstrip the supplyresponse. This results in a market rationing process in which thelimited stock of housing is allocated to the highest bidder. House-holds in all but the highest-income categories are forced to accepthousing which does not meet their expectations, and increasedcrowding results. Middle-income families occupy housing designedfor low-income families, and the latter may not be able to afford"formal" housing at all.

Externalities, particularly the quality of the neighborhood environ-ment, and the fact that housing is spatially fixed affect the decisions ofhouseholds to construct or improve housing or to change their resi-dence. Superimposed on these attributes are imperfections which canbe classified as institutional or economic in nature. The latter includerestrictions on the availability of infrastructure needed to servicehousing land. Complications in dealing with indivisible assets yieldingsubstantial economies of scale are frequently compounded by pricingand management rules of utilities which may restrict the access of thepoor to public services at reasonable cost. Imperfections in privateland markets lead to very high land prices, frequently caused in partby speculative withholding. These high prices combine with hightransaction costs to make land assembly difficult for the poor toundertake.

Institutional constraints generally arise from legal barriers and ad-ministrative controls erected to protect specific groups or interests.Zoning often prevents parcels from being allocated to their highestand best use, either by imposing limited categories of use or by re-stricting unit density. Building codes, ostensibly designed to protectthe consumer who has limited ability to judge quality of structures,often protect suppliers of outmoded materials or favored build-ing trades. Minimum wage legislation and unionization have beenblamed for limiting the flow of labor into the housing sector, raisingunit costs. Rent controls discourage the expansion and maintenanceof the housing stock. Institutional arrangements governing housingfinance, such as interest rate ceilings and collateral requirements,may inhibit the mobilization of savings. Countless additional barriersrestrict the access to vital information and make compliance withrequired procedures more difficult, particularly for the poor. Over-coming these constraints requires a sustained commitment to policieswhich provide housing at a cost within the reach of poor families.

25

Chapter 4: HOUSING POLICY OPTIONSFOR DEVELOPING COUNTRIES

Housing, and more generally urban development, is difficult toorganize efficiently and equitably even in industrialized countrieswhich do not have to deal with explosive city growth. However, sev-eral developed and some developing countries have been able toachieve housing standards which combine efficiency with welfare.Developing countries thus have a range of experience on which todraw. Some of the more important lessons from this experience merita brief discussion at this point.

The Policy FrameworkA country's general policy framework has important implications

for both supply and demand factors influencing housing. Exchange,foreign trade and other commercial policies affect the housing marketthrough their impact on a country's overall competitiveness. Monetarypolicies influence the availability of housing finance; lack of develop-ment of grass roots financial institutions, for example, means a lack ofmortgage funds. Fiscal policies are critical to a country's ability tomobilize resources for growth and the distribution of income. Equallyimportant, all these policies have spatial dimensions which influencecity sizes within each country and the location of residences and em-ployment within a city. Countries with a poorly conceived overallpolicy framework cannot expect to deal effectively with their housingproblems.

The economic characteristics and welfare implications of housingmake it a sector of relatively high public involvement in most coun-tries. Important components of the housing package-access roads,utilities and to some degree transport-are "natural" monopolies thatrequire public ownership or regulation if supply is not to be restrictedso that excessive monopoly profits may be reaped. External costs andbenefits also make public involvement necessary. The substantial ex-ternal benefits of urbanization and housing are reflected in increasingland values in the course of city growth. Unless a special effort is madeto appropriate some of these benefits for social use, the increases invalue will accrue to those individuals who are fortunate enough to belandowners. This situation encourages land speculation and "clear-ance" of squatter areas, and may ultimately lead to social conflict.

The welfare implications of housing are another reason for govern-ment assistance. Such assistance may be designed to offset marketimperfections which are too difficult to attack directly, but which

26

discriminate particularly against the poor. Thus, housing subsidies maybe justified on grounds of increasing productivity, as a means of redis-tributing income to the very poor, or both.

An effective housing program requires effective administration.The experience of most cities in developing countries suggests thatexisting approaches give insufficient weight to fostering institutionswith a specialized knowledge of local conditions. Emphasis needs tobe given to fostering institutions capable of formulating urban devel-opment plans and, within them, housing programs. In contrast toprevious planning methods which have relied chiefly on zoning mapsto indicate which development is to take place at which locations,strategic urban development planning implies the analysis of socio-economic influences on development and their coordination in aframework specifying planning objectives and the resources at handto meet the objectives.

However, without a commitment to carry out these plans, the bestconceived institutions and management cannot be fully effective.Government resistance to programs which allegedly "perpetuateslums" can for a variety of reasons be very strong. In other countries,where resistance may not be a problem, there may still be insufficientknowledge about the operation of housing markets and lags in deal-ing with their imperfections. The first step in improving housingopportunities for the poor is often a forthright recognition of theseconstraints and their effects on the formulation and execution ofhousing policy.

The range of instruments available for carrying out housing strat-egies and policies is considerable. Although their application mayfrequently lead to unexpected effects, sufficient experience has bynow been accumulated to enable some general observations to bemade about the most appropriate use of particular instruments.

Zoning RegulationsZoning is a potentially valuable instrument of urban and housing

policy. It is important in protecting an aesthetically pleasing residen-tial environment. It helps in grouping certain (e.g., commercial) activi-ties in an efficient manner, and segregating unwanted activities. Yetthe inappropriate application of zoning regulations borrowed fromdeveloped countries may exclude the poor from residence near in-dustrial, commercial and high-income residential zones which pro-vide their income earning opportunities.

Building Standard ControlsAppropriate building standards can also do much to create a safe

and pleasant environment, for example by ensuring adequate circula-

27

tion of air, preventing structural failure or limiting the likelihood offires, floods and similar hazards. However, if standards are set too highfor existing income levels, either officially or by the architects andengineers responsible for a particular project, their primary effect willbe to push down the living standards of the poor by reducing theamount of housing available at prices they can afford. The poor willthen be driven to bribery to retain or construct "substandard"housing.

Pricing PoliciesThe attempt to prevent the capture of socially caused gains by

private individuals and to offset the imperfections of housing marketsby introducing rent controls has a long and unsuccessful history. Rentcontrols have only been effective for limited periods of strong socialcohesion, as during wartime, and when applied in combination withstrict controls on other incomes and prices. Applied over longerperiods, they have eventually given way in most countries to illegalrationing systems by which some form of "key money" becomes asubstitute for proper rents, and the poor are driven out of much ofthe controlled housing. But this neither undoes nor fully stops thedamage. Because they limit returns on housing to less than marketlevels, rent controls inhibit additions to housing stock and lead todeterioration through lack of maintenance. Attempts to control theprices of public utilities or urban transport have had similar financialrepercussions. In addition, they distort the location decisions ofhouseholds and firms. They also lead to rationing situations in whichthe poor are seriously disadvantaged as capital stock and the supplyof services deteriorate.

Measures to price these services at their full economic value, how-ever, may be difficult in the short run. They frequently run into resist-ance. But even if they do not, their introduction in the absence ofother measures may be inequitable. Obliging poor families, for ex-ample, to pay the full cost of housing services may place them at adisadvantage relative to the nonpoor as well as other poor families,if the practice is not citywide. Yet it is precisely in cities where theabsence of full-cost pricing for housing and some services is wide-spread that the magnitude of the change, and the numbers likely to beaffected by measures to introduce greater efficiency and equity, cangive rise to the greatest resistance. Such changes must thus be intro-duced thoughtfully. One means which appears to be gaining accept-ance is to finance part of the cost of services to the poor from profitsfrom operations involving richer families; these cross-subsidizationschemes have been increasingly applied in a number of countries.

28

Land TenureLandownership in urban areas is at least as complex and as badly

distributed as in rural areas. While an overall improvement in landtenure can usually be only a long-term target, squatter area upgradingis a prime policy instrument for improving the situation for at leastpart of the poor population quickly. Introducing an element of equityinto urban landownership by giving squatters security of tenure is anurgent policy issue in most cities in developing countries, and onewith high returns in terms of retained and improved housing stock,access to earning opportunities and consequent welfare. However,the payment for an appropriate portion of the actual land valueshould be required in exchange for title to avoid excessive subsidiesto only part of the poor.

Land Purchase and DevelopmentMost cities have found it necessary to purchase land that can be

used for urban purposes in order that they may deal with housing andassociated urban development problems equitably. Some cities indeveloping countries have publicly owned land which can be usedto start such a process; others are interested but have no initial landbank, and a few are legally barred from such action. Land within exist-ing city limits is usually very costly, but peripheral land can often bepurchased ahead of use at low cost. As the present owners are gener-ally farmers, a variety of deferred payment schemes can assure even-tual possession at current prices; in the meantime, the land continuesto be productive in agricultural use. Public land servicing has suc-cessfully stimulated private housing in many countries, and it is par-ticularly suited to assisting low-income households at peripherallocations through the sites and services approach. The preparation ofserviced sites ahead of the growing demand for housing can be usedto affect urban form as well as to provide poor households withsecurity of tenure and a sound basis for building a house. With goodmanagement, the increasing value of serviced land can be appropri-ated for social use on a much broader scale. The development of landfor manufacturing and commercial uses, together with housing andtransport, can have a strong development impact. In these circum-stances, city growth can in time become not only more efficient andequitable, but also more financially self-reliant.

Taxes on LandThe capture for social use of some of the gains, or "betterment,"

accruing to land and housing owners from public improvements isanother important issue. By capturing betterment, society does not

29

penalize a landowner for putting his land to productive use, butsimply reduces any incentive to hold land for other than productivereasons. Capital gains and other taxes on land, as well as user charges,have often been used for these purposes. In practice, valuation prob-lems and erosion of the tax base through exemptions have limited theeffectiveness of the taxes. They are, nevertheless, increasingly calledupon to provide revenues, underwrite the cost of specific public im-provements and achieve a more equitable distribution of urban realresources. Raising tax receipts from land can alleviate some of theburden of taxes on capital, stimulating the flow of investment and thedevelopment of financial institutions. Moreover, a program of publicimprovements financed by betterment levies or equipment chargestends to reduce uncertainty about the pattern of future urban growth,thereby encouraging private investors.

Building MaterialsThe greater use of traditional materials and the adaptation of new

techniques to traditional building methods can lower costs, particu-larly in labor-surplus economies. India has scored notable successesin this regard. Traditional brick and tile making methods, for example,may not be able to utilize more than a small fraction of local clays, buta refractory technique may do so. Although secondary timbers intropical countries may not be sufficiently durable for building in theirnatural state, they will withstand climate and predators if appropri-ately treated. Lime, which is in plentiful supply near many cities, canbe used instead of cement to effect cost savings.

Improving Financial InstitutionsThe lack of mortgage funds frequently constrains housing. How-

ever, this is not so much a housing issue as a question of developingthe financial sector. In most developing countries, to build a range ofsuitable intermediaries is the first task. But financial intermediationgenerally requires reforms in financial markets. The freeing of interestrate restrictions, measures to encourage competition among banks,the ending of inappropriate banking practices, the promotion of lifeinsurance companies and pension funds-all increase the availabilityof long-term credit, and hence, potentially, of finance for an efficienthousing sector. If the several specific hindrances to housing financewere removed, housing could compete more effectively with otherclaimants for long-term institutional finance, and that would enhancesocial progress by facilitating home ownership. Mortgage insurancecan be used to facilitate increasing lending for housing and to bringsuch lending within the reach of lower-income groups. Lending in

30

small amounts to low-income borrowers requires special institutionalapproaches.

Public HousingExperience with large-scale public housing construction which

substantially increases housing density is varied. There have beenmany failures; however, countries active and successful in publichousing view it as an instrument for shaping cities and urban pro-ductivity, and as a contribution to the welfare of lower-incomehouseholds. Public housing may improve the spatial distribution ofresidences and employment by locating families where all earnershave access to job opportunities. Enough knowledge has been ac-cumulated, as much from the failures as from the successes, to guidecountries wishing to embark on this course. Standards have to beconsistent with prevailing income levels. If buildings are designedflexibly, initially low standards can be adapted as incomes rise. Con-struction has to be technically sound and economically efficient.Public housing can provide the steady demand that assists a localcontracting industry to develop, but to avoid an inflationary impactlarge new increments to demand should be accompanied by seriousattention to potential supply bottlenecks.

Secondly, public housing has to be well-located in relation to jobopportunities. In Singapore and Hong Kong, public housing is well-located in relation to industry. In addition, "flatted factories" forsmall-scale industry have been built in the middle of housing develop-ments, and large-scale, labor-intensive industries have been attractedby factory space for leasing in the midst of housing estates to employsecondary family workers. A housing area has to be large enough tocontain-or have good access to-functional community facilities,such as schools, health clinics, markets, shops and recreation. It issomewhat easier to manage owned than rental units, and ownershipis often also preferable for social reasons. In any case, the manage-ment input has to be substantial and efficient. This includes not onlyfinance and maintenance, but also staff to collect rental or time pay-ments according to a rigorous schedule and social workers to assistfamilies to adjust to high-density housing and other new aspects oftheir environment. Some delinquencies are to be expected, for avariety of political and social reasons. But in appropriately designedand managed public housing where standards are geared to incomelevels, delinquencies because of inability to meet rental paymentsshould not be substantial. To be effective, public housing has to havemost of these components, in forms which meet the needs and cul-tural patterns of the occupants.

31

Chapter 5: THE LESSONS OFINTERNATIONAL ASSISTANCE

The Evolution of International AssistanceThe economic and welfare gains to society from the provision of

housing have drawn the United Nations, several multilateral andbilateral aid agencies, and private aid agencies into providing assist-ance for housing since shortly after World War II.

The United Nations has been actively concerned with housingsince its inception. The U.N. Centre for Housing, Building and Plan-ning (UNCHBP) is responsible for the U.N.'s activities in these fields,covering research and development as well as technical cooperationwith member states. Other U.N. agencies with programs relating tohuman settlements include the Food and Agriculture Organization(FAO), the International Labour Organisation (ILO), the United Na-tions Educational, Scientific and Cultural Organization (Unesco), theUnited Nations Industrial Development Organization (UNIDO), theWorld Health Organization (WHO) and the World Food Programme(WFP). The International Habitat and Human Settlements Foundation,founded on January 1, 1975, is expected to play an active role. Theregional Economic and Social Commissions of the United Nationsalso have certain responsibilities in this field.

The UNCHBP has progressively moved toward treating housingwithin the comprehensive framework of development planning,either in the form of specific action-oriented programs or by pre-investment studies. Its activities in housing have covered both urbanand rural aspects of the problem. Many years of urban physical plan-ning efforts have been supported in several cities. This strategicplanning is designed to influence growth in ways which improve onwhat a simple continuation of past trends would produce. It mayalso suggest ways in which disenfranchised groups may share moreequitably in the benefits of urban growth. Some cities, notably Singa-pore, have been able to provide a counterpart planning input, backedby administrative measures and financing, to bring these plans tofruition in a relatively short period.

Although the U.N. assists member states in any aspects of housingfor which assistance is requested, the UNCHBP has increasingly fo-cused on the problems of low-income groups. In 1964, it initiated aprogram to deal with slums and squatter settlements in rapidly grow-ing cities of the developing world. The U.N. Economic and Social

32

Council launched a pilot program aimed at assisting developingcountries in encouraging, and whenever possible supporting, com-prehensive programs for the improvement of slums and squattersettlements in urban and rural areas. These programs emphasize areliance on self-help, self-improvement and gradual improvement,rather than the customary method of clearance and wholesale re-newal. Following several exploratory missions, countries were iden-tified for initiating pilot demonstration projects. Although the initialresponse from governments was slow, the policies advocated by theU.N. have been receiving increasing attention.

In earlier periods, the focus was on limited objectives, such as theconstruction of a specified number of housing units. More recently,however, the emphasis has shifted toward adopting a two-prongedattack on problems of physical development: (1) programming oflong-range investments in housing within the overall investment pro-gram for urban or rural development; and (2) programming of im-mediate investments and implementation of a series of short-rangeactions aimed at treating the most urgent housing problems. Sitesand services projects have formed a notable part of this approach.However, even though a measure of success has been achieved insome areas, the integration of housing within a comprehensive frame-work of development has so far been a difficult objective to achieve.

Most other public assistance efforts have been more direct, pri-marily involving funds for housing finance institutions, although tech-nical assistance has also been provided on a limited scale. Theprincipal public aid agencies with housing commitments have beenthe U.S. Agency for International Development (USAID), the CaisseCentrale de Cooperation Economique (France), the Inter-AmericanDevelopment Bank (IDB) and the Commonwealth Development Cor-poration (CDC). The USAID program has been the largest, with some$205 million in capital assistance and nearly $14 million in technicalassistance devoted to housing from 1949 to 1971. This was supple-mented by commitments of $654 million in the Housing InvestmentGuaranty Program through the end of 1974. This assistance has beenconcentrated mainly in Latin America, where USAID funds had con-tributed to the construction of 111,000 units by 1971.

The Caisse Centrale de Cooperation Economique is the secondlargest source of bilateral funds for housing in developing countries.It has concentrated its lending activities in French-speaking countries,primarily in sub-Saharan Africa. Approximately two-thirds of its loanshave been made to housing corporations, and the rest to housingauthorities (Office des Habitations a Loyer Mod&e) seeking to assistmiddle- and low-income groups. The IDB has committed $416 mil-

33

lion, or 7% of its total portfolio from 1960 through 1973, to urbandevelopment projects, including housing. Under IDB projects, theconstruction of 340,000 dwelling units had been completed by Jan-uary 1, 1974. CDC involvement in housing, also substantial, has fo-cused primarily on "seed capital" and technical assistance for mort-gage finance companies and housing estate developments in Africa,South Asia and the Caribbean. Commitments totaled $148 millionfrom 1960 through 1973;' these funds were supplemented by $246million from other sources. CDC has, in addition, extended loanstotaling $25 million to governments and housing authorities forlow-cost housing.

After more than a decade, it has become clear that although theseefforts have helped to increase the total stock of housing, they havehad little direct impact on housing for the poor. Most of the effortshave not dealt effectively with existing market imperfections andhave not succeeded in reaching low-income groups (Annex 1, Tables1:2 and 1:4). The unit costs of housing they have financed haveranged from $1,100 to $7,000.2 Moreover, "seed capital" for housinginstitutions has not in practice altered banking policies which havesubsidized middle-income groups and excluded the poor from hous-ing loans.

These programs might have reached more poor families had theybeen part of a package of housing inputs, conceived in an urbanplanning framework. They might have been more conducive to re-ducing subsidies to middle-income groups, had their implementationbeen made conditional on the charging of appropriate interest rateswithin an appropriate package of repayment terms.

USAID and IDB are aware of the difficulty in meeting the objectiveof helping to provide "minimum shelter" to the urban poor. A 1971report prepared for USAID concluded that the bulk of the housingfinanced over the previous decade had been too costly for the poor.The report recommended "that USAID redress this record by con-centrating upon the housing needs of low-income people and dis-courage . . . loans, grants and technical assistance programs whichdo not attempt to meet their needs." 3 In 1973, USAID adopted a"shelter sector policy" which identified low-income housing as aprincipal target of its shelter program. In October 1974, USAID furtherrefined its policy and published its "shelter policy determination"which places more emphasis on USAID's intention of formulating orsupporting a variety of housing programs designed directly to benefit'Annex 1, Bilateral Institutions. Sterling figures have been converted at $2.60 to the pound.'See Annex 1, Table 1:2.'Foundation for Cooperative Housing, Cooperative Housing and the Minimum ShelterApproach in Latin America, Vol. I/ (Wlashington, D C.. 1971).

34

the poor. The focus is on developing national housing policy andhousing institutions, as the principal means of achieving long-termsolutions to housing problems in developing countries.

Private bilateral aid agencies, for example the Foundation for Co-operative Housing (FCH) and the Pan American Development Foun-dation (PADF), appear to have been somewhat more successful inreaching lower-income groups, although on a very small scale. TheFCH has helped establish cooperatives in several Latin Americancountries, achieving unit housing costs of less than $1,000. PADFprojects organize the poor so that they may take advantage of low-cost credit. The first national foundation in the Dominican Republic,created with PADF assistance in 1966, has since made available about$1.7 million in loans for housing and community development in-volving 150,000 people, for an average loan of $11.50 per person.

Private development agencies have thus been generally more real-istic in defining the housing demands of low-income groups andmore cautious in their approach; they have concentrated limitedfinancial and human resources on a small number of cases. Perhapsfor these reasons, they appear to have made a more innovative impacton housing than public agencies. The public aid agencies, viewing therange of experience-successful and unsuccessful-with low-incomehousing, are reassessing their own programs, and this has contributedto the renewal of interest in "human settlements" in the internationalforeign assistance community.

The Experience of the Bank GroupThe Urbanization: Sector Working Paper of June 1972 set out an

initial strategy for the World Bank Group's entry in the housing field.Until that time, the Group's lending and related institution buildingefforts in urban areas had been directed toward national and sectoralobjectives, rather than affecting the growth pattern of cities and thewelfare of the poor. Taking a comprehensive view of urban develop-ment, the paper indicated a need for projects which would assistcountries to accommodate rapid urban growth more efficiently andimprove the distribution of income in cities. Sites and services proj-ects, by providing access to serviced land and security of tenure, couldbe a key instrument for influencing urban development. Improve-ment of existing low-income housing was also considered important,although specific approaches were not worked out at that time. Theprincipal objective in both these types of projects was to provide low-income families with the land and public utilities components of thehousing package, and a variety of technical and financial assistance toenable them to use self-help to build and progressively improve their

35

dwellings. Direct lending for housing was not immediately contem-plated, but the possibility was not excluded that, at a later stage, theremight be direct financing of permanent housing if the circumstanceswarranted it. Indirect Bank Group assistance to housing through stim-ulation of housing finance institutions was contemplated, providedthere were no subsidies for medium- and high-income groups.4

Sites and Services ProjectsAssistance for the first sites and services project was approved by

the Bank in June 1972. So far, 10 urban development pr