|Support Measures to Promote Rental Housing for Low-Income Groups (HABITAT, 1993, 132 p.)|
|II. THE EFFECTS OF LEGISLATIVE AND POLICY INTERVENTION|
The forms of rent control introduced around the globe are highly diverse. Some countries freeze rents at a particular date and only allow cost increases to be passed on to tenants. Others completely index rents. Some have multi-tiered systems in which some units are frozen, some increase at an express annual rate, and some only increase on a change of tenants. In the three case-study countries, the main features of rent control have been as described below.
In India, rent controls have been introduced by most state governments. The different rent-control laws are very similar and apply to all residential and non-residential property, with the exception of governmental premises, high-income accommodation and newly constructed property. The laws are extremely rigid and cover rent levels, the responsibilities of tenants and landlords, and eviction. In some states, the landlord does not even have the right to choose the tenant; responsibility for this rests with a government agency, landlords having to notify the agency of any vacancy. Rents are usually fixed at a standard rate established at some date in the past, and rises are permitted only under certain circumstances. The base for establishing the standard rent is normally the cost of construction plus the value of the land. The rent is based on a percentage of this value, ranging from 6 to 15 per cent according to the state. Periodic rises in rents are permitted in some states but usually only every three to five years. Sometimes, too, rents can be increased if major improvements have been made to the property or if there has been an increase in local taxes. Maintenance is the responsibility of the landlord but if repairs are not made the tenant is permitted to pay for them out of the rent. Eviction is possible on the grounds of non-payment of rent, misuse of the premises or the need of the landlord to use the premises for his or her own family needs. The tenancy rights are extended in some states to the tenants heirs.
In Nigeria, rent controls were first introduced in 1941 and control over eviction in 1948. The two laws were incorporated into the Rent Control and Recovery of Premises Law in 1954. The precise regulations vary from one state to another but all are based on federal guidelines. In most states rents are fixed according to the quality of the accommodation with the general approach being to limit rents to less than 20 per cent of household income. Tenants can be evicted only by a court order which is normally granted on one or other of the following grounds: rent arrears of more than one month, the need for substantial repairs, the premises are required by the landlord, the tenant is misusing the property or being a nuisance, the accommodation is required for some public purpose.
In Egypt, rent control and legislation began in 1920 and controls over rents gradually increased during the next five years (Serageldin, 1993:3). In 1941, rent rises above certain levels were prohibited and tenancies were automatically renewed after their termination. This was consolidated in the Rental Housing Law of 1947. More controls were introduced after the Socialist Revolution of 1952 and by 1962 rents on most kinds of housing were controlled. The law was substantially modified in 1981 when the annual rate of return on the combined value of the land and the cost of house construction was established at 7 per cent. Only a limited number of luxury and furnished apartments were excluded from these regulations. The 1981 law also changed the system for raising rents and permitted the sale of housing to the tenants. The Government also established a central agency to provide subsidies to cover part of the rents of the poorest families. The law also allowed tenants and owners to share the costs of maintenance.
It is clear that few landlords or tenants fully understand the rent-control regulations even in developed countries (Malpezzi and Ball, 1991). In most developing countries, whole swathes of rental housing tend to be excluded from the regulations, notably those in the self-help housing areas. Since many of these areas exist outside the formal embrace of the law, insofar as they lack title deeds or planning permission, it is not surprising that they should be ignorant of much of the rent-control legislation. Nonetheless, not knowing the full details of the regulations is not the same thing as total ignorance. In many cities, rent controls work partially. Some landlords know the regulations whereas others do not. Some tenants are aware of the rules but others not. Some landlords and tenants comply with the regulations, others evade them.
Generally, rent controls seem to be implemented effectively only in countries with socialist regimes. In China and Cuba, controls cover every kind of property and seem to be applied fully. Elsewhere controls seem to have worked in certain areas of Bangalore, Buenos Aires, Harare, Kumasi, Mexico City, Mombasa, Rabat and SPaulo (Ghai and MacAuslan, 1970; Keles and Kano, 1987; Kowarick and Bonduki, 1977; Malpezzi and Ball, 1991; Rakodi, 1989; Rivas, 1977; Tipple, 1988). In most third-world cities, including those just mentioned, rent controls do not seem to have worked in the informal housing areas. The lack of legal rental contracts in most self-help rental housing is a clear sign of this. Recent research has shown that few Latin American tenants are issued with contracts (Jaramillo, 1985; Gilbert and Varley, 1991; Gilbert, 1993).
Even where the law is applied with some effectiveness, there are many ways in which landlords can escape the rent-control legislation. One is through key money whereby landlords compensate for low monthly rents by demanding a high initial payment from the tenant before occupation. Key money is a well-known feature of rental housing markets in Egypt and now increasingly in Kumasi (Abt Associates, 1982; Feiler, 1990; Serageldin, 1993; Tipple, 1988). Sophisticated advances on the system of key money have developed in Bolivia and in the Republic of Korea where it effectively substitutes for the payment of rent. The Bolivian anticrco and Korean chonsei contracts both consist of a lump-sum payment which the landlord returns at the end of the contract period (Beijaard, 1992; Green, 1988; Kim, 1992; Renaud, 1989). The owner invests the deposit. The tenant pays no monthly rent and the deposit is fully refunded at the end of the contract period, usually a minimum of one year. The system became popular in both countries as a way to circumvent rent-control measures and restrictions on the eviction of renters.
In the three case-study countries, there is plenty of evidence of landlords avoiding the effects of the rent-control legislation although the degree of evasion varies considerably between the three.
In Nigeria, the rent-control legislation has been almost totally ignored since the early 1980s and did not work very well even when it was supposedly being applied. In fact, Okpala (1985) claims that rents were higher than decreed levels two years after the introduction of the rent controls. Ozo (1993: 50) goes even further and argues that landlords raised rents above what they would otherwise have been, to impress on government the futility of promulgating a decree which it is not capable of implementing. Whether or not that claim is justified it is clear that tenants were in no position to protest about non-compliance. As Ozo (1993: 12) puts it:
most tenants under the condition of acute housing shortage, for fear of being evicted, colluded with landlords in black market transactions about controlled rents. As a result only a tiny minority of cases were ever reported.
Even when the law was working its implementation was haphazard and very limited in coverage. For the last 10 years the act has been more or less moribund. Today, most landlords act outside the provisions of the law.
The record is little better in terms of implementation of the controls over eviction. Less than 5 per cent of landlords and many fewer tenants have used the courts to evict tenants (Ozo, 1993: 11) and most tenants simply leave when asked to. The essence of the problem for the landlord seems to be that there are too few courts and, therefore, cases take a long time to resolve. As a result, landlords do not have the patience to seek possession under the Recovery of Premises Act (Ozo, 1993: 12). In the case of the tenants, there is another reason; their weak position vis-is their landlords. They are afraid of losing their accommodation and of not getting other accommodation when other landlords hear that they have complained to the authorities (Ozo, 1993: 55).
In India, the regulations work a little more effectively although most landlords still fail to comply with the Rent Control Act. In fact, this has been a longstanding problem. NIUA (1989a: 15) reports on a study of 1968 which found that only 10-12 per cent of tenants in Delhi were affected by the Rent Control Act. Today, the situation is little better. As NIUA (1989a: 153) note: a large part of Delhis rental housing market lies outside the purview of the Rent Control Act. Another significant part of this market does not conform to all the provisions of the Act. This is demonstrated by Wadhva (1993: 31) who notes that most of the landlords and tenants are unaware of the rent control act and even if they are aware are not fully conversant with its provisions. In Delhi, for example, Wadhvas (1993: 31) survey reveals that few landlords use the rent-fixing machinery laid down by the Rent Control Act and 85 per cent of rents had been raised after a couple of years despite the Act. In the informal housing sector, the Act is even less effective: the rent control act has minimal impact on rents and new construction (Wadhva, 1993: 73). Of the 390 premises surveyed in Delhi, the rent was fixed by the Controller only in 38 cases (p. 82). Landlords protect themselves from the risk of the tenant approaching the Court to establish a standard rent, either by charging key money or by demanding a deposit on a returnable or non-returnable basis. The survey found that deposits are charged in 47 per cent of cases. Not surprisingly, few receipts are given for this payment. In the 25 per cent of cases where receipts are given for the monthly payment of rent, the correct amount is written in only 45 per cent of the cases. In 5 per cent of cases it is for a higher amount (so that the tenant can charge his or her employer) and in the remaining half, lower amounts. As a farther sign of non-compliance with the Act, 77 per cent of landlords raise the rent every year, 6 per cent twice a year (p. 83).
The Act does not seem to work much more effectively in terms of controlling evictions. Landlords do not use the Act to evict because of the strictness with which the courts interpret the rules protecting the tenant. There is also the problem of time; the lengthy court procedures discourage most landlord from going to court. As a result,
forced evictions for non-payment of rent or for any other reason are the rule rather than the exception. In many cases tenants vacate on their own rather than protest or seek redressal under law (Wadhva, 1993: 34).
The expense of the legal system is too great for the poor to use it.
Only in Egypt, among the three case-study countries, does the rent-control legislation seem to work at all widely. Certainly, Serageldins (1993) survey of 100 Cairo landlords, found that 95 per cent issued contracts, 70 per cent had never raised the rent and 99 per cent had not evicted a tenant in the last year. However, as Malpezzi and Ball (1991: 73) note
key money is a well embedded feature of the market. Side payments account for almost as much imputed landlord revenue as rents there; and for recent tenants, they account for the majority of the revenue.
Formal-sector landlords also protect themselves by renting to relatives or friends; 63 per cent of the landlords surveyed by Serageldin do this. In the informal housing areas, however, rent-control does not seem to work at all. Landlords do not issue contracts, they levy key money and they charge the free-market rent. The chief reason for the lack of implementation in the self-help areas is the Governments reluctance to admit that such areas exist.1 It has to recognize these areas before it can apply rent controls on the landlords.
1 It has tried for many years to protect agricultural land from the incursions of self-help housing.
In principle, rent controls are intended as a means of protecting weaker groups against the ravages of inflation and/or exploitation by landlords. Rent-control legislation is normally premised on the assumption that landlords are affluent and tenants are poor. For this reason many countries have excluded expensive property from the controls and limited it to lower-value property. In India, state rent-control laws always include low-income property but sometimes exempt high-priced accommodation.
A basic problem with rent controls is that the benefits are never transparent and are rarely wholly predictable. It is certainly not always clear whether the poor benefit. Among the tenants, some poor families benefit but others lose out; some affluent tenant families also gain. Among the landlords, some rich ones suffer but so also do some poor ones. As a result, there is no clear link between those who benefit and those who most need help. In Bangalore, for example, Malpezzi and Tewari (1990) found that Bangalore, for example, Malpezzi and Tewari (1990) found that there was no simple relationship between income and the benefits derived from rent controls.
Rent controls also tend to favour sitting tenants over new tenants. As Malpezzi and Ball (1991: 71) put it: often, long term tenants tend to benefit from controls, and recent movers often pay large key money, advance rent, or other side payments. In Mexico City, long-established controls in the central areas of the city used to mean that tenant families paid extremely low rents; by contrast, the majority of the population were not covered by the rent controls (Aaron, 1966). Similarly, in Cairo, Malpezzi (1986) found that while new entrants paid large sums in key money, existing tenants were paying very little of their income in rent. In India, where the Rent Control Act also favours sitting tenants, new entrants into the housing market lose out (Wadhva, 1993).
Poor tenants may also suffer in the longer term if landlords respond to rent controls by failing to maintain the property. In Guadalajara and Puebla, many tenants have given up complaining about the lack of repairs because landlords simply do nothing. Four-fifths of the tenants who had complained about the need for repairs reported that the landlord had taken no action (Gilbert and Varley, 1991: 172-173). In the central areas of these cities, landlords protest that the cost of repairs is far too high to contemplate given the low rents paid by the tenants. In Cairo, however, it appears that maintenance of formal-sector accommodation may not have suffered unduly despite the oppressive rent controls. As Abt Associates (1982: 208) point out: residents of rent controlled properties often assume the financial burden of maintenance themselves, at least for their own units, spending on average a sizeable fraction of income for this purpose.
A further long-term effect of rent controls is that the rental supply is reduced. As Wadhva (1993: 71) puts the argument in the case of India:
the impact of low rents on rental housing construction is evident from the rental housing market. This provision has driven away even the public agencies. Even though rents are not controlled in houses owned by the public agencies, the application of the Rent Control Act in the market has created a psychology of fixedness of rent which has permeated the tenants of public agencies who resist any rent increases vociferously and violently.
However, again Egypt provides some counter evidence. In Cairo, rent control does not appear to have had a major effect on rates of new construction... rental units are still being produced (Abt Associates, 1982: 208).
Perhaps the greatest problem with rent-control legislation, however, is that it is often not applied in just those areas where most of the poor live. As Section II.B has shown, most legislation simply does not work, or works only partially, in the self-help areas of most Latin American cities where the majority of poor tenants are now concentrated. In these areas, neither landlords nor tenants really know much about the law.
Even if poor landlords and tenants were aware of the law, the high cost of using the courts would deter them from taking legal action. They simply could not afford legal advice and without such help they would be unable to cope with the complicated procedures. As Ozo (1993: 16) explains for the case of Nigeria,
the technical/legalistic language and cumbersome procedure of the law, the monetary and time cost of court procedures under the law, the burden placed on tenants to report infringement to courts before erring landlords can be identified, etc. make the law particularly unsuitable to low-income groups who are also quite often illiterate.
In his survey in Benin City, he found that the only tenants who have used the courts have been literate and from the middle class. As a result, In spite of the intention of the law, the low-income groups have had little, or no, protection from it (p. 17). Still worse is the fact that the law is sometimes improperly used against them; it has been known for wealthier landlords to coerce tenants into leaving the premises by getting lawyers to send them letters.
If rent controls have had little effect on the living standards of the poor, so too have other kinds of government intervention. Generally, and perhaps fortunately, zoning laws are ignored in the self-help housing areas. Public housing construction has provided homes for very few poor families and in recent years, none in rental accommodation. Attempts at encouraging the development of a housing-finance system have brought benefits entirely to higher income groups.
Planning and zoning regulations, including rules covering building standards, infrastructure norms and land use, are frequently criticized for the way in which they restrict the production of housing. The World Bank (1993) suggests that they often make housing unaffordable for low-income groups, restrict the residential land supply, and create bureaucratic bottlenecks that cause delay. Such problems, of course, affect the production of every kind of housing, not just the production of housing for rent. Arguably, therefore, they affect rental-housing production no more forcibly than the production of housing for ownership.
Such regulations are often felt to be particularly inappropriate in less developed countries. They were introduced either under colonial regimes or by indigenous planners trained in the planning schools of the former imperial country. The impact of green belts, zoning regulations, land-use controls and building standards, have introduced a paraphernalia of unnecessary controls into third-world cities. Not only has this increased the bureaucratic hassle for anyone involved in trying to penetrate the maze of regulations, but it has also created opportunities for delay and corruption (de Soto, 1989). A typical view is presented by Baken et al. (1991: 29):
building regulations and codes - being based on European models - have been observed to constrain the development of low income settlements. Apart from the restriction of lot sizes for economic activities, two other factors hamper the economic development of planned settlements: the limited proportion of plots for commercial use, and the regulatory framework which often prohibits mixed residential-commercial land-use.
Of course, the main effect of such regulations is on the so-called formal sector. For, as is well-known, large areas of most third-world cities have developed with little regard for the planning regulations. If houses are being built without services, then it is difficult to believe that their builders are being any more scrupulous in following the regulations that might affect their accommodating tenants. Of course, illegal developers do not ignore all of the rules. In Santafe Bogotfor example, illegal subdividers often adopt some of the regulations in an effort to appear as legal as possible (Carroll, 1980; Gilbert and Ward, 1985). The closer they can approximate to the rules, while maintaining their cost advantage, the less likely they are to be harassed by the authorities. But, in general, market forces in determining the price of land and the general attitude of the government with respect to upholding the law are more important influences on the informal housing supply than the nature of the regulations per se. If the government does not apply any planning regulations in the self-help areas, it matters little what the rules actually say.
Clearly, planning regulations affect formal-sector builders directly but, insofar as few build for rent, there will be little impact on the rental-housing sector. By far the most effect will be felt in the field of owner-occupation. The rental-housing sector will only be influenced when specific planning regulations make the construction of housing more or less difficult. In the case of Mexico City, Connolly (1982: 149) argues that this was precisely the impact of the introduction of more stringent building regulations in the 1940s. The implementation of building regulations modelled on North American and European standards,... clearly made it more difficult to build substandard housing on a legal basis. This reduced the profitability of such investment and therefore the amount of rental housing construction.
The enforcement of safety regulations can also cut the amount of rental accommodation especially when it is linked to slum clearance. Since much decayed housing in central locations is occupied by tenants, slum demolition can lead to vast areas of rental accommodation being destroyed. Unfortunately, although they are less common today, examples of slum demolition on a massive scale have been all too common in third-world cities. In the 1960s and 1970s, inner-city renovation programmes led to the loss of large numbers of homes in Lagos, Rio de Janeiro and Tunis (Marris, 1979; Hardoy and Satterthwaite, 1981; Peil and Sada, 1984; Valladares, 1978). In Delhi, some 700,000 squatters were moved in 1976 from the central areas to the outskirts (Risbud, 1991; World Bank, 1993: 30). In the centre of Mexico City, attempts to resuscitate commercial activity in the central area have destroyed vast numbers of rental homes in the central city. Coulomb and Shez (1991: 40) claim that slum demolition and new road programmes during the 1970s led to the destruction of 50,000 rented rooms in three neighbourhoods alone.
Sometimes, however, attempts to beautify the city and to apply planning controls more rigorously have quite the opposite effect on rental housing. In several Latin American cities, where local governments are trying to conserve the colonial architecture in the central areas, the demolition of old houses has been prohibited. The result is that owners cannot sell the homes and are forced to rent them. Admittedly, this sometimes leads to gentrification and much higher rents but it also leads to the retention of some unimproved rental housing for the poor. The contradictions involved are demonstrated by experience in Quito. As Godard (1988) points out: historic preservation of the colonial city has encouraged landlords to renovate some buildings, the rents for which poor people can no longer afford. However, the overall result of colonial preservation in Quito has been inadvertently to retain more central low-income rental housing than in most Latin American cities... (Klak and Holtzclaw, 1993: 263). Similarly, in Mexico City, the introduction of more stringent planning controls in the central areas in the 1940s made it much more difficult to change land uses. As such, many tenants were not evicted because their landlords could not obtain permission to develop new office or residential blocks. While these landlords may have ceased to maintain their property, something which eventually led to the decay or demolition of many rental housing units, the short-term effect was to sustain rental housing in the central area.
Restrictive planning regulations can also stimulate rental housing production insofar as they limit urban expansion. In cities, such as Seoul, where rural-to-urban land conversion is severely restricted because of rigidly enforced greenbelt regulations and master plan provisions that limit residential development to only 25 per cent of the total land area, the effects can be serious (World Bank, 1993: 83). In Seoul, the restrictions have led to explosive increases in the price of land and housing, severely decreased housing affordability, and persistent housing shortages (ibid., 83). Under such circumstances, it is perhaps not surprising that Seoul has a high percentage of households that rent or share accommodation. Similarly, in Caracas, urban expansion has been limited by a decree to keep development below the 1000 metre contour line. This rigorously applied regulation has encouraged urban development to spread farther along the valley floor into more distant locations. The effect has been to encourage the development of rental housing in the informal sector. The only alternative to renting a room is obtaining a shack in the distant periphery.
In Delhi, the planning regulations have severely discouraged all kinds of housing construction. They have also depressed rental-housing production in a number of ways. First, master plans up to 1990 limited plot sizes to a maximum that could accommodate only two families; even now only three families are permitted to live on a single site (Wadhva, 1993: 29). This prohibition clearly limits the scope for rental housing. Secondly, although rental housing can be developed under the Group Housing Scheme, the densities specified under that scheme are too low to make investment commercially viable (p. 27). Thirdly, planning regulations also make it difficult for owners to adapt their homes to create space for tenants. Renting part of an owners flat is virtually impossible because legally only one kitchen can be built (p. 28). Although all of the above regulations are often flouted they discourage renting because of the bribes that have to be paid to the inspectors. Fortunately, some modifications are being made to the regulations and, in 1990, the third-floor barsati terrace, which had been built and illegally rented out for years, was finally legalized (Wadhva, 1993: 28).
In Benin City, the zoning regulations ostensibly limit the profitable development of rental accommodation by stipulating that houses can only have eight living rooms on a single storey, 12 on two floors and 15 on three, with a maximum occupancy rate of two people per room. However, since few of the planning regulations are actually applied in Benin, the restrictions are likely to have had little effect on rental housing production. The most likely impact of planning rules on formal sector construction comes in the form of the prohibition on any single owner using more than 0.5 hectares of government land on leasehold title. Ozo (1993) claims that this prohibition militates against large-scale development of rental housing (Ozo, 1993).
In Cairo, the urban zoning laws limit the number of houses per plot but Serageldin (1993) believes that this has had little effect on the development of rental housing. Potentially more significant is the legislation introduced in 1940 and 1975 to control urban expansion in agricultural areas; a policy linked to the strategy of building new towns in non-agricultural areas (Feiler, 1990). In practice, planning officials admit that the law has never been effective (Abt Associates, 1982: 36).
The main difficulty in commenting on the impact of planning regulations on rental housing lies in knowing precisely to what extent the rules have been implemented. Since many regulations have been applied fully in one part of a city and not at all in another, it is difficult to be categorical about the overall effects of their application. In general, however, it is probably true that quality standards have encouraged the idea of slum demolition, leading to considerable losses in the low-income rental housing stock. It is also likely that small plot sizes in sites-and-services schemes have discouraged rental housing development (Chant and Ward, 1987:14). In the case of new rental property, complicated rules and regulations have probably helped discourage potential landlords in the formal sector from building for rent. If profits from rental housing can only be generated by building cheaply, then planning regulations which push up physical standards can only discourage investment. Whether this is a good or bad outcome is a moot point.
Usually the rental legislation shows few differences between cities in the same country. In Mexico, while the state authorities are responsible for most regulations covering rental housing, they tend to adopt similar guidelines and usually follow the example of the Federal District. Of course, the local authorities sometimes take an exceptional line, usually for political reasons. During the 1980s, for example, the states of Puebla and Michoacboth introduced regulations which limited rent rises (Gilbert and Varley, 1991: 57-58).
In Egypt, the rent-control legislation incorporated certain variations between areas, with controls not applying to rural areas or to a number of tourist centres (Serageldin, 1993: 14). In Nigeria, state governments are responsible for housing but all the legislation follows central government guidelines (Ozo, 1993: 6). For example, the rent-control ordinances that were introduced in every state after the Civil War, adopted the pattern laid down by the Federal Government. In India, however, the controls over rents in different states do vary a great deal (Wadhva, 1993: 17-19). The main variations relate to the flexibility given to the landlord to choose tenants, the degree to which rents can be raised, the kinds of project that are exempt from control, maintenance standards, eviction procedures and the rules under which tenancies can be inherited. To what extent these differences actually influence the amount of renting in each state is uncertain.
On the whole, local legislation has much less influence on the functioning and development of rental housing than a number of other factors. The nature of the land market, particularly the cost of urban land, the way in which the authorities permit or prohibit land invasions, the nature of the transport system, and cultural attitudes to renting and ownership all influence the tenure structure of a city much more than local regulations. This is clear when rural tenure patterns are compared with urban tenure patterns in the same country. The typical ownership pattern of the countryside and the greater importance of renting in most cities cannot be explained in terms of the planning regulations. Similarly the major variations between Indian cities in the amount of renting cannot be explained in terms of the legislation. The difference between Calcutta with 76 per cent of its population renting or sharing and the 30 per cent in Lucknow is simply too large to be explained in terms of planning regulations, regulations which in any case are only partially applied (NIUA, 1989).