FORMATION OF PRIVATE RENTAL STOCK IN TURKEY Murat Balamir [Paper first received, April 1999"; in final form, June 1999]
ABSTRACT Withdrawal from socially owned housing provision and the privatization of the stock have stimulated growth of the owner-occupied sector in many European countries. These processes have also led to a more significant role for the private rental sector. Apart from a model of 'social deliberation', there are other models to explain the evolution of this sector in its socio-political and historical context. In the formation of a large private rental sector in Turkey, public policy had no contribution to make. Instead, it was a particular mode of urban development that allowed the proliferation of the sector as a subsidiary process of urbanization. The relative conditions of tenants and the policy requirements in this context are very different than those observed elsewhere.
1
Current concerns in the rental sector
The rental sector has attracted the renewed interest of practitioners in the field of housing studies, who have made it the focus of significant intellectual effort. The state of this sector with respect to the national housing system, and the different forms of tenancy in particular, have been the subject of comparative analysis, both across countries and through time, spanning periods of present and past experience. ~ Some of the issues that have been discussed are the relative disadvantages that tenants face, changes in the size of the sector, increases in rents and housing costs, and the reshuffling and concentration of lower-income groups in specific parts of the stock. Attention has also been devoted to policies like providing ineffective subsidies, selling off the social rented sector, abolishing rent control, etc. which aggravated these problems further. The fundamental issues addressed in the analyses range from building a capacity to judge whether low levels of tenancy are sustainable to determining the direction in which housing systems are moving and how policies could be reshaped. In this context, the rental sector has become an indicator in the evaluation of housing conditions and policy. Comparative studies have been relevant not only to a better understanding of how rental systems function in the different contexts but also to devising more incisive policy tools.
Neth. J. of Housing and the Built Environment, Vol. 14 (1999) No. 4. 385
Drawing attention to the 'under-theorized' nature of housing research, and to the bias prevailing within the English-language housing literature, Kemeny (1994) pointed out that the existence of public and private segments of the rental sector is conventionally related to the dualist nature of the national political systems accommodating the Labour and Conservative fronts. Yet there are countries where the political power structure is not necessarily aligned in this manner and therefore the implications for the housing system deviate, even to the extent that the meaning as well as the social and legal content of tenancy cannot be considered identical with that of the liberal political model. In many other contexts, little mention is made of housing problems in general, let alone the rental sector, in the domain of social policy, and housing is not part of programmes of political deliberation. This deprives researchers and policymakers in this context, of the use of concepts and approaches employed by the dualist model. The 'deep-rooted ethnocentricity' of the dualist model (Kemeny, 1994), brings in the necessity of understanding other systems and developing a more neutral conceptual framework. Reviewing the different contexts, Maclennan (1998) could identify three models of tenure-composition within the more advanced economies, differentiated by the different shares of privately-owned, private rental, and the publicly-owned stocks. In the first model, home ownership exceeds all rental provision, while private rental exceeds socially owned housing. The second model refers to the dominance of non-market housing in rental provision. Where the rate of renting relative to owning is high, as in the third case, private rental provision dominates. Even though this classification constitutes a set of mutually exclusive groups, it probably does not deplete all possible combinations. Policy development and implementation in housing are complex activities, and they become even more cumbersome in the rental sector. As privatization and the transfer of public housing stock to households took hold in many European countries, alarming reports about the worsening conditions of residual tenant households focused attention on the necessity of formulating new policies for this sector. Europe's owneroccupation sector has expanded under the practices of deregulation, while tenancy rates have declined and the production of units for rental purposes has been cut back drastically. In other parts of the world, however, observations are partly in line with the European experience. The number of households living in rental accommodation is still on the rise, even though public and large-scale private investment in rental housing is at its lowest level or in some cases is non-existent (UNCHS, 1993). In this context, where the provision of public rental stock is most unlikely, policy recommendations are often concerned with the promotion of the interests of small-scale private landlords (Watson and McCarthy, 1998; Kumar, 1996). This paper explains how a significantly large private rental sector has grown in Turkey despite the supposedly adverse conditions of a 'no-policy', 'no-regulation' environment at the local and central administration levels. The potential problems identified and appropriate policies recommended in this analysis may also be relevant for other contexts.
386
2
The context and nature of the rental sector in Turkey
The case of the rental sector in Turkey differs in many respects from the European experience and is unlike that in the developing countries. The general context of housing and that of the private rental sector may be described in terms of the following points: - No public policy or provision and no publicly owned stock exists in any form as rentable housing. This condition seems to have had no effect on the everincreasing number of tenants in the country. The share of tenancy in Turkey is no less in proportion than the rates reached at the peak of the housing provision efforts in many European countries. These efforts have been effectively supported by the social housing programmes dominating housing policies for almost a century. Although public intervention in the sector has been negligible, total housing production maintained through private investments in an economy of relatively low GDP per capita has been significantly high by international standards during the past four decades. This performance has largely been dependent upon the informal rearrangement of specific forms of property rights which in turn are based on tendencies toward more home ownership. Urban households in Turkey prefer home ownership, not only for occupation but also for investment and financial security. The prevalent policies and the inflationary economy all in effect promote ownership. Such conditions did not hamper tenants' chances of access to the newly built stock, even though no overt policy was followed by any designated authority to regulate the terms of this access. Contrary to widely held belief, tenants in Turkey do not necessarily bear high costs. Despite a loose mechanism of rent control that conforms with the market conditions, and despite the no-regulation climate, tenants seem:to enjoy relative privileges in their housing conditions and options. Analyses of tenancy in Turkey cannot ignore the question if how a rentable stock is formed and landlordism emerges as a subsidiary process of urbanization. Nor can the analyses ignore the role that housing plays in the dynamics of urban social stratification. This in itself is not the result of explicit policy. It poses problems of a different nature and begs a range of questions not necessarily referred to in the European context. Significant local variations in tenancy point to the need for policy tools in the hands of urban governments rather than reliance on regulation by central government. Tenancy in Turkey has evolved in a specific historical and institutional setting. Subsequent governments seldom pretended to provide welfare measures in housing. Only during the Second World War was any effort made to protect tenants; that was when rent control (freezing rents) was imposed" for a brief period. Apart from a minute segment of the housing stock constructed for government employees, provision of rental housing was never a practical objective or a concern of the public authorities,
387
central or local, even though such functions are explicitly described in their statutes. Almost by tradition, all discourse employed and every step taken in the housing sector purported to encourage home ownership. Governments have subsidized the prospective homeowners and have provided credits for new construction through intermediary financial institutions. Yet a comprehensive and consistent model of social policy and control has been distant in the ownership model as well. Even in the case of occasional projects that aim to upgrade squatters, preserve the historical stock, or reconstruct settlements after natural disasters, tenants have been ignored. It was in mid-1980s that the State Housing Development Administration (HDA), attached to the Prime Ministry, came into action, providing subsidized credits to both the supply and the demand sides for new construction, giving deliberate priorities to large-scale cooperatives. This move contributed to the volume of production, primarily in the form of large-scale housing estates. Yet it did little to serve the needs of tenants, who were left entirely unprotected in the market. In the European tradition, large-scale housing developments in physical terms represent publicly subsidized social housing to support tenants. In contrast, cooperative investments subsidized by HDA in Turkey imply support for ownership alone, ousting tenants. Currently, the only protection available for the tenants is that of the judicial system and court decisions concerning restrictions on evictions and annual rental increases, unless otherwise overridden by the terms of privately drawn up tenancy agreements. Despite the absence of ideological-political and practical protectionism, the number of tenant households has increased four-fold since 1965. The current total number of tenants in Turkey is about 3.5 million households. An overview of tenancy rates (Table 1) also indicates a rise from 16 per cent to 30 per cent between 1965-1990 for the country as a whole. In absolute terms, urban tenant households increased by almost five times within the same period. Rural areas have always had negligible levels of tenancy. Thus, factors which act to inflate the overall tenancy rates in Turkey seem to be related to processes of urbanization. The very same factors also tend to stabilize this rate at some asymptotical level. The underlying nature of this rental system may be elaborated by a number of statements: - Tenancy in Turkey is fundamentally an urban phenomenon in that almost all 96 per cent of the rental transfers made in a year (1985 figures) are made in towns and cities (SIS, 1989: 170). z In absolute terms this represents a substantial volume of payments compared with total investments in housing or with the total pecuniary incomes of tenant households. Tenancy in rural areas is becoming less and less significant, since tenant households are more likely to migrate to the cities and metropolitan areas. - Tenancy in Turkey is fundamentally a market phenomenon. Tenancy agreements are drawn up between tenant households and the private landlord and not with public or semi-public bodies or private companies acting as landlords. It was observed that only about 12 per cent of the tenant households have had some form of acquaintance with their landlords prior to signing rental agreements (Balamir, 1992). At the other extreme, tenants having a first-degree relationship with their landlords as relatives and offspring constitute only seven per cent of the total, 2.5
388
Table 1 Rates of tenancy in Turkey 1965-1990 (% of households) Total Urban areas* Rural areas 1965 15.86 38.66 1970 18.42 37.96 5.22 1975 19.31 36.14 5.62 1985 22.57 36.51 4.46 1990 29.71 41.08 4.03 *Urban areas comprise provincial centres and sub-provincialtowns. Sources:General Population Census of years indicated, SIS. The 1980 Census contains no relevant data.
per cent of these still making formal pecuniary rental payments. Almost all of the rental transfers in the market are transactions carried out between individual private landlords and tenant households. Tenancy in Turkey is fundamentally a local phenomenon. Significant disparities prevail in the distribution of tenancy rates of settlements, which vary between 25 and 62 per cent. This wide range could not be explained by settlement size or age of the stock, nor by some regional feature. Rates of tenancy within the settlement system also reveal significant variations in their pace of change (1955-1997). Towns that have changed more rapidly account for increases that are 13 times faster than the more stagnant ones. Faster-changing settlements are not necessarily the largest cities or metropolitan areas, nor do they represent a consistent pattern of regional concentration. Tenancy in Turkey is fundamentally a contingent phenomenon. Rentable housing is not provided by some purposeful authority or decision-maker responding to rental markets. Instead, it is exogenously procured and delivered, depending on general conditions of development. In this sense, the need and demand for rental units, rental levels, supply surplus, and vacancy rates are not always consistently interrelated in any settlement. In this sense again, the source and origin of the rental stock is in none of the four models, which I describe as 'social deliberacy', 'business management', 'historical residue', and 'landlord professionalism'. Rather, the rental stock originates in another model, which could be defined as the 'by-product' model. The first two simply refer to the provision and management of rental stock by public and commercial enterprises. The third is related to filtering processes, whereby the stock produced originally for owner-occupation is extensively passed on to tenants through successions of generational moves. In the fourth model, individual households acting as petty-landlords are encouraged to invest in second homes as well as in furnished and partial lettings, exploiting the marginal capacities of the system. The processes taking place in Turkey could not be accommodated in these models. Rather, the growth of the rental stock will have to be explained with reference to the processes of urban development and housing investments. In that context, the formation of a rentable stock is not necessarily the primary objective, and even 'reluctant landlordism' may be observed. It is therefore relevant to expand briefly on the nature of urban growth 389
and describe how tenancy rates might be determined within processes of property development.
3
Formation of the rental system as a subsidiary process of urbanization
In contrast to most European post-war housing regimes, governments in Turkey did not subscribe to comprehensive housing programmes, let alone to an explicit policy for the rental sector. Yet overall rates of tenancy in urban housing remained firmly above 30-35 per cent. At no stage in the post-war history of Turkey did central or local authorities administer a housing stock at their own prerogative. How then could almost one-third of the stock be allocated to tenants and how could households be differentiated along tenancy cleavages? The answer to this question is hidden in the intricate evolution of specific forms of property relations in urban areas (Balamir, 1969, 1975, 1996a). The performance of housing production in Turkey has been particularly striking during the past four decades. Despite the low levels of capital accumulation and per capita income, and despite the scarcity of public resources available for the procurement of land for development, housing production exceeded even the volumes reached by many of the high producers of the world. This performance has been dependent upon anonymously devised informal rearrangements of property relations, primarily functioning as a framework to facilitate home ownership. These are of several types, each one leading to a different form of stock. I tend to identify these rearrangements as processes of 'appropriation', 'apportionment', and 'appurtenance' to be differentiated in terms of legal status and parties involved, apart from their physical qualities (Balamir, 1996a, 1996c). The first process usually refers to the formation of squatter settlements, whereby public or private land is illegally appropriated and unauthorized construction is carried out. The second is a process of informal sharing of land that is legally owned, based on joint ownership, but illegally subdivided for purposes of unauthorized development. But it is the third type of rearrangement in property relations that is the dominant and ultimate form of development in Turkey; ultimate in the sense that this form of relations cannot be transformed into another in the free market interactions. The process of 'appurtenance' is a model of cooperation between the entrepreneur, the landowner, and the participating investor households in the construction of residential buildings. The entrepreneur, with the consent of the landowner, takes advantage of a form of access to land with a minimum of costs and obtains the advance payments from third parties. All parties involved in this cooperation are prospective owners of the predetermined (i.e. with reference to the building project at hand) and appertaining parts of the building, each entitled to separate deeds expressed as a ratio of the property in its entirety. Land and common areas in the building are subject to joint ownership pro rata (Balamir, 1969, 1975, 1992). Not only has this method of construction remained undefined in legal terms, but many constraints have also been imposed by the Civil Law itself. 3 Such cooperation therefore had to rely on private agreements. With the extensive adoption and proliferation of the appurtenance
390
procedures throughout the country, immense political pressure had built up, leading to its formal recognition in 1965 with the legislation of 'flat easements' and 'flat ownership'. When operating within this model of property relations in the market, the landowner acquires a substantial share equivalent to the 'land rent' or 'surplus'. This share, in return for his permission to develop the land, is usually defined in terms of property, 25-60 per cent of the prospective block of flats. The former landowner thus experiences a metamorphosis and turns into an owner of several flats. The property owner has the privilege of either renting them out or disposing of them in order to reinvest in yet another cycle of similar development and thus to expand his personal stock of property. A group of households, thus sequentially participating in such processes, succeed in progressively and rapidly increasing their wealth and improving their status. The size of such a group could be determined more or less by the extent and pace of physical development. High-density and concentrated physical developments in central cities generate higher surpluses and thus a smaller group of elite monopolist landlords. Lower densities lead to a more extensive and egalitarian sharing of the development gain. 4 It is the number of units transferred to the landowner within this set of relations as defined in the appurtenance process that represents a surplus and a contribution to the rental stock. Rent relations in urbanization thus simultaneously determine the pace of development and the size of the rentable stock. For this reason, it is not surprising to observe a high correlation between the distribution in cities of tenant households and the stock of fiats in blocks (r:: 0.99). This is much higher than the correlations of tenants with the distributions of the number of 'houses' or 'squatter dwellings' produced as the other possible forms of the housing stock in settlements. Variations in inter-urban distributions of tenure groups are thus dependent on this system of property relations and on the extent of investments made in housing, but not necessarily on some exogenous decision of providing a rentable stock or the rental market demand levels. In this context, it is possible to group cities with 'supply surplus' and 'supply deficit' and to formulate differentiated policy recommendations. The process of appurtenance is also responsible for the intra-urban distribution of the various tenancy forms generated. Tenants, rentier households, and owner-occupiers are all interrelated and accommodated in each block of flats at proportions based on the economics of development and sharing of the property. This leads spatially and functionally to a mosaic of different tenures all taking part in a symbiotic coexistence, rather than to spatial differentiations and enclaves of tenures. Tenants are therefore well distributed in spatial terms and in terms of stock age and type. The implication of such distribution is that tenants are not discriminated or constrained in any manner. Rather, they have a wide range of options in every district of the city, in housing of different ages and type, and have comfortable access specifically to the newly completed parts of the stock. This leads to an extensive market of rentable stock in the city, with a consolidated range and relatively lower rental prices for new units. The only restriction tenants could experience in such extensive lebensraum is probably the case of subsidized large-scale housing cooperative developments. Ironically, cooperatives too are realized under the aegis of 'flat-ownership' relations. Yet
391
considerably smaller numbers of units are available for tenant households, especially during the earlier phases of occupation of cooperative housing, since no surplus is generated in the process. Cooperative developments, however, represent only a very small part of the total housing production. Thus, in a policy climate that favours ownership, about one-third of the new additions to the stock, as if 'naturally!, are allocated as rentable units, almost defying the necessity of formulating and structuring social policies to regulate rental provision.
4
The process of 'appurtenance' as a distinct context for the private rental sector
Coexistence of independent ownership rights (or easements) in buildings is not unique to Turkey. Similar property-sharing ownership arrangements in multi-unit structures have existed since early 20 ~ century in a number of European and other countries in real (de facto) and legal (de jure) terms. As stock transfers and privatization of the public rental stock took place more recently, the introduction of such rights into the body of existing legal systems have been inevitable. The Turkish case is unique, however, for a combination of factors: - The property-sharing arrangements are exploited mainly for the purposes of development and the formation of stock in Turkey rather than for transforming the ownership relations in existing stock, as in cases of privatization, or for the purposes of marketing a previously produced building by some independent entrepreneur, or by planning decisions to turn some valuable stock into collective capital to ensure its maintenance and upkeep. The procedures of arranging such property rights in Turkey, on the other hand, function mainly as a catalyst in any process of development. - The process of appurtenance has evolved within free-market interactions as a historical phenomenon at a particular stage of urbanization in Turkey and has an economic and a legal history. Its proliferation has been based on the excessive demand for housing and buildings on the one hand, and scarce public resources to supply infrastructure and serviced urban land for development, on the other. - As an innovation based on shares related to prospective buildings, this model of property development was adopted throughout the country within two or three decades as the dominant form of housing provision and mode of urban living. This sudden expansion of its exploitation is incomparable to any other example available. It is by means of this method of capital formation that intensive investments and large volumes of housing production have been possible over decades, and the physical stock in urban areas has grown extensively and massively. Today almost 70 per cent of the total urban housing stock is under the regime of flat-ownership in the country. In terms of the size of the population involved, the scale of new investments, and the extent of the distribution of physical development throughout settlements, the appurtenance process exceeded all other forms of provision, which is unique in the world.
392
-
-
-
-
Development by means of the process of appurtenance is unique again in its evolution and structure of relations within the legal context. The Civil Law of Turkey, based on the Roman model, could not tolerate the sharing of buildings (present or prospective) on a permanent basis, Its extensive adoption and practice through private agreements based on economic necessities nevertheless implied insecure legal status, leading to intensive political demands for its formal recognition. Special and separate arrangements had to be tailored to the legitimization of this process. The legal history of procedural developments therefore constitutes another distinct aspect of this new form of property ownership. In most countries, on the other hand, legal provisions either already existed in civil and property law, business law, and law concerning capital markets, prior to any interest and extensive practice of such form of ownership, or in the purposeful making as in the case of 'Commonhold Law' in England under consequences of privatization policies. The emergence and spread of the appurtenance process is a purely market phenomenon in Turkey. In many other countries, depending upon the purposes of encouraging multiple ownership in multi-unit buildings, explicit public policy and manipulative decisions are responsible for its practice. The latter may have the objective of upgrading central areas, stock renewal or rehabilitation, or the privatization of public housing stock. The nature of the capital involved in development through the appurtenance process is again specific to the circumstances in Turkey. In this set-up, investments in construction are not necessarily transferred from another source or sector of the economy. In an economy of scarcities, insignificantly small-scale investment capacities of the individual households are gleaned together and organized into operational capital to carry out moderate amounts of construction work in stages. It is this minimalist exploitation of capital in construction that facilitates the smooth success of development business (Balamir, 1975, 1982). The builder employing the partial and staged investments of prospective owners (of the independent parts) of the intended building puts up almost no capital of his own, making profit on capital advanced by others. No capital is again advanced by the landowner other than his allowance of a collective investment on what is now a jointly owned plot. With respect to scale and staging on the other hand, capital transferred by the prospective flat-owner households, often have no other outlet for an efficient exploitation and therefore represent capital procured with the lowest opportunity cost. In almost all other countries where similar ownership patterns are observed as forms of condominium or flat-ownership, large-scale capital already active in development or another sector is responsible for development in the ftrst place. Sharing of buildings is often a secondary concern or procedure. In general, the landowner, the entrepreneur-developer, and the investing households constitute the three distinct parties involved in the process of appurtenance. Development activities of a similar nature in other countries may reveal a bi-partite or a singular party (roles unified) operating during development. A landowner could have sufficient capital for the production of multi-unit structures.
393
-
-
-
-
394
Alternatively, the entrepreneur could acquire and procure land for development later to produce blocks of flats and dispose of them. The stock produced by means of the process of appurtenance has been the basis upon which urban populations have been restructured into new classes of wealth. The housing stock, as a new material resource, and social values differentiate urban society along cleavages of property ownership in a manner incomparable to differentiations based on conventional forms of tenure as tenants and owneroccupiers (Balamir, 1982, 1992, 1996c). The social and spatial distributions of the population in urban areas are determined directly by this model of unequal sharing of values generated, with dissimilar consequences compared to other and previous cases reported elsewhere in the world. Property ownership of a similar nature elsewhere in the world would help to consolidate existing social stratifications, rather than generate new alignments in society. Restructuring of the urban population along tenure cleavages by means of the process of appurtenance, although dependent on a very real differentiation of wealth and styles of living, has little impact on the spatial distancing of the emerging 'classes' in urban areas. Rather, a symbiotic existence in the stock is the rule which blurs the perception of polarizations and generates a mosaic of tenures in large and homogeneous urban regions. Tenants, owner-occupiers, rentier households, janitor households, all huddle together in the same building as parties interdependently involved in the process of appurtenance. This outcome is very dissimilar again to the much theorized geographic structuring processes of many cities of the Western world. It is not possible here to discuss the spatial models of differentiation, although very real social differentiations do prevail. After four or five decades of the appurtenance process, it is only recently that some of the wealthiest of the households are observed to be moving to urban peripheries to consolidate themselves in exclusive housing neigbourhoods. Social roles and property relations designated by the appurtenance system are not confined to the construction and stock formation stages of development alone. The mode of social organization in the management of multi-unit buildings is also determined by the very same system. The 'fiat-ownership' law constitutes a basis upon which all regulation in environmental management could rest. This includes methods of running property, democratic procedures for electing a manager for the year, meeting the expenses and how they are to be shared, constraints on the use of flats, resolutions to conflicts arising between various forms of shareinterests in buildings, and even the time-sharing arrangements are all formally regulated. Such detailed regulation of management issues by law with reference to ownership, unless specified within private agreements, is not the overall rule elsewhere. Flat-ownership is the ultimate form of property relations in Turkey, in the sense that a reunification of fragmented rights in any building is almost entirely discarded (Balamir, 1975). Although individuals' coalescing interests constituted the basis upon which the initial development took place, a second consensus for redevelopment purposes is highly unlikely to obtain under free market interactions. Market relations that favour the institution of the fragmentation of rights in build-
ings, could not in a similar fashion guide the parties to an exit, and lead to the reunification of fragmented rights in the building. An exit from such relations is possible only in three ways: public acquisition, unanimous decision of all shareholders, or the total demolition of the building. This implies an extensive and absolute urban paralysis in physical terms. In almost all regulations elsewhere, however, action for renewals or major changes is facilitated by the re-imposition of some unified decision-making power, either at the end of some specified period or whenever the need arises. Causes and concomitants of the process of appurtenance are unique to Turkey therefore, specifically in terms of capital formation and investment behaviour, development procedures, sharing of property wealth generated, social stratifications based on this wealth and their distributions in space. As such, the process of appurtenance is also the main generator of the private rental stock in the country. It is true, on the other hand, that social relations in squatter areas also give rise to some rentable stock as well. The numbers and values involved in these rental markets are minute compared to the mainstream relations of tenancy under flat-ownership. Still, the squatter units, in their evolution towards more substantial investments, will inevitably follow the process of appurtenance in physical and social terms, even though they may remain unauthorized forms of development in legal terms. It is relevant, therefore, to focus on the process of appurtenance as the main generator of rental stock and tenancy relations in urban areas.
5
Landlord and tenant profiles in the supply-driven rental system
The supply-driven rental sector is made up of a stock, the by-product of some structural economic process, a class of landlords designated by this process, and 'eventually' a specific population of households using this resource. Since the rental stock is generated largely as a by-product of investments intended to meet home-ownership demand (rather than to meet some existing rental market demand) in Turkey, the characteristics of the tenant households resident in this stock may not necessarily resemble those observed in many of the European and other contexts. The process of appurtenance not only allows a group of households to become homeowners, it also forces a minor group (land-owning individuals) to turn into property-rich landlords and a rentier class. These are private individuals, and in the case of Turkey, which is different from the Western experience, they comprise no companies or public bodies as a rule. This process contributes to the polarization of the owneroccupier and tenant households in a manner unlike the reported European cases (Maclennan, 1998: 398). Reviewing the basic indicators like 'accommodation size', 'household size', 'income and wealth', 'housing expenditures', 'duration of residence', 'previous tenure status' etc. reveals further discrepancies." Tenant households occupy larger dwellings in the aggregate with respect to household size than owner-occupiers (Balamir, 1989, 1996b). Their position in terms of accommodation and stock use is not inferior to
395
Table 2 Distribution of households of different income levels according to tenure (% of households) Income level (20% of households)
O-o
1969 T
O-o
1978 T
O-o
1987 T
O-o
1994 T
Lowest 18.00 13.53 12.41 13.50 1 2 . 0 6 26.65 16.31 6.37 Low 40.34 38.89 35.28 38.31 3 1 . 3 1 36.35 18.81 2.25 Middle 18.54 21.12 28.73 29.01 2 1 . 9 2 1 7 . 1 0 1 9 . 0 7 21.54 High 8.96 12.92 19.21 15.34 1 7 . 4 7 1 2 . 3 0 21.98 16.57 Highest 14.06 13.64 4.37 3.84 16:94 7.60 2 3 . 8 3 13.27 Total 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 Source: Sample surveys of housing income and expenditure SIS (O-o: Owner-occupier, T: Tenant).
those of the owner-occupiers. It is most probable that the lower mobility of the owneroccupiers and the existence of 'latent tenants' do generate this type o f an incongruence, s When the proportion of tenants in any city is smaller, this relative privilege seems to be more intensive. In terms of household size distributions, tenant households are more stable than those of owner-occupiers, with a slightly biased growth in the single-person and four-person size categories of households. Owner-occupier and tenant households are observed to have moved towards the opposite extremes of the income ladder, with tenants concentrating at the lower income quintiles (Table 2). This is partly explained by the existence of the flourishing rentier class. In the distributions of housing and rental expenditures relative to incomes, tenant households stand in no inferior position compared to the owner-occupiers. The lower rates of expenditure o f the higher-income tenants, on the other hand, reveal the existence of a sub-group of voluntary tenants. A significant relative fall in the aggregate volume of rental transfers is observed, at least for the 1975-1985 period. Differentiations by wealth may be more significant in that tenants are generally believed to have less real property. In a survey of households living in the multistorey blocks, only 25 per cent of all tenants had property other than dwelling units, as compared to 34 per cent of the owner-occupiers (Balamir, 1992). This gap is slightly greater in the case of other possessions like cars, which show respective rates of 23 per cent and 37 per cent. The 1990 Census reveals that the ratio of owneroccupier households in posession of dwelling units other than occupied is 15 per cent, while the rate for tenant households is 16 per cent (SIS, 1993: 186). According to a private survey of four provincial centres (1992), 30 per cent of the tenants have a house somewhere else. Another 18 per cent are in the process of saving to own, and for another 13 per cent the prospective dwellings are under construction. With respect to style of life and type of stock occupied, 83 per cent of all tenants occupy blocks of flats, and a great majority of these have had similar units as their previous accommodation. It is only the cooperative type of housing provision that seems to accommodate the lowest tenancy rates. Tenants were tenants again to a great
396
extent (78 per cent) in their previous accommodations. Only 14 per cent of the households currently were previously owner-occupiers. Of the current owner-occupiers, 61 per cent were tenants in their previous accommodation. In the aggregate, 41 per cent of the total households prove to have the capacity to turn into owner-occupiers at any period, whereas 32 per cent either must retain their status as tenants in the system or choose to do so. A more comprehensive analysis of the stratification of households, as determined by the specific reorganization of property relations (the appurtenance process) in urban development, yielded four distinct groups with reference to tenure. A four-part distinction of households by tenure and property ownership, (i.e., tenants as well as owner-occupiers with and without property) revealed a better fit to existing distributions of household features than the conventional two-party (owner-occupiers vs. tenants) classification. In other words, with respect to the social attributes of households, a four-party distinction proved to be a more successful division than the conventional two-party (tenants vs. owner-occupiers) model of tenure. 6 This peculiar condition substantiates the arguments in favour of avoiding 'ethnocentricity' in housing research.
6
Prospects and policy alternatives: is tenancy an ephemeral phenomenon?
The rental sector in urban Turkey is dominated by petty landlords and by a growing class of rentier households. This holds true even in the case of squatters districts, where significant illegal monopolies of land and family holdings have been generated. Even though tenants find relatively favourable housing-market conditions at the moment, whether this is a sustainable condition remains debatable. Most of the advantages of tenancy seem to emerge from the current high rates of investment in development and the dictated supply of a specific form of property sharing in the process. As long as new housing construction maintains its current form and rates, the relative advantage of the tenants is likely to last. The only threat to the tenancy sector at the moment is from the cooperative segment of new construction, which tends to repel the tenant households. It is not clear how the tenancy sector could contract if this symbiotic interdependence between tenure forms comes to an end when the pace of construction slows down, or if cooperative development is stepped up further through policy measures. In the next cycle of events, the surplus stock owned by individuals will probably be dissipated by transfers to inheritors or filtered down to be sold to owner-occupiers. So the stock could once again be taken over by owner-occupiers to a large extent, and the rentable stock and supply will be pushed back to its lowest ebb. Since tenancy is not an ephemeral social phenomenon, on the other hand, scarcity is likely to occur in the markets at some later but not too distant stage. Production of social rental stock is not necessarily the only alternative proposal to be advanced against policies of ownership in Turkey. Under the constraints of the current political trends, there are other policies preferably to be advanced as substitutes for policies of ownership, proving higher-social and political returns at lower risk and cost. Emancipation from a policy-barren environment in the rental sector should be
397
possible through several lines of action to be followed by the authorities at different levels. For the central government, the primary task is to enable the local administrations in several directions. Empowering municipalities to practice rent control in an integrated manner with their existing property taxation and planning powers could generate favourable results. Modest reform in the incomes and property taxation systems is a second area of intervention which could revolutionize market relations between tenants and landlords and improve the tax base for governments at both local and central levels. Concessions in tenants' income taxes could terminate the current symbiotic interests of these two parties which perpetuate lower declaration of rental and transfer prices, causing immense rates of tax evasion. Income tax exemptions to tenants (and purchasers) only, proportional to their payments, could suffice to break up the pervasive 'property tax evasion corporation'. A vital area of institutionalization in a country of the magnitude of Turkey, and at its stage of urban growth and stock formation, is the integrated information banking and surveillance of the housing system. Finally, the central government could in some cases be obliged to selectively subsidize subgroups of tenants like single elders within the social security system. In some cases, such selective subsidies may be related to specific large-scale development projects like squatter-clearance and upgrading, which could also merit international credits. The HDA could pluralize its singular policy of subsidizing home ownership and contribute in more way than one. Even in its allocation of credits for ownership at the settlement level, prioxities could be set upon indicators such as tenancy rates, share of population living in squatted premises, rental transfer levels, vacancy rates and overoccupation ratios. The wide variation in tenancy rates justifies differential interventions in the system and calls for the formulation of discriminatory policies concerning tenancy among settlements. The re-introduction of tenants in the housing cooperatives could be enforced by introducing low-interest credits for additional rental units to be included in production programmes, owned and managed under collective ownership. It is critical that subsidies end up in the hands of the real tenants rather than those of the rentier households. This is the way to maintain the vital balance between total volume of tenant households, rates of production, and supply of new rentable stock. Given the peculiar nature of the housing system in Turkey, it is particularly relevant and effective to structure policies for the tenancy sector at the local level. The local implementations of rent control, the piecemeal and long-term acquisition of a stock of public rental units, and the publicly initiated management of information concerning the private rental sector are just some of the possible forms of approach that could prove particularly potent in the vitalization of local democratic processes. Allowing municipal administrations to practice rent control may lead to the institutionalization of an effective method of allocation of local resources to meet local needs. Accountable directly to the local electorate, in their practice of a rent control mechanism, municipalities will have to develop a political discourse and take into account those parties involved in the housing markets. Coupled with the existing local powers of property taxation and planning control, rent control could be responsive to spatial characteristics as well as to the type and age of the stock, encouraging devel-
398
opment at specific locations and curbing it at other places. This should be an appropriate approach in dealing with the variations in the rental sector among settlements. Not only could housing conditions get priority on the local political agenda, but also a variety of alternative approaches could be formulated in practice through periodic experimentation, as the office changes hands in a plural political environment. For those municipalities still confident and devoted to the idea of directly owning and controlling a stock of rental units, there are probably more opportunities and options in Turkey. No local authority need be involved in large-scale production and management tasks of housing, which demand major investments and public spending, expertise, and commitment. But the option of acquisition of individual fiats at different blocks and locations could be exploited with a minimum of commitment, with modest levels of investment in existing stock as resources permit at any time. The 'right of preemption' could here not only enable municipalities make the most favourable selections in the market, but it could also act as a means of controlling market performance. This allows the building up of a municipal stock distributed throughout the city in different districts, consisting of units of different types, age, and size of stock, constituting a system of price observation posts in the housing markets. A gradual acquisition of a stock in this manner could also help to build experience in the management skills of property resources. Procedures of acquisition could be integrated with the planning powers and be part and parcel of development activities of the municipality. Legal provisions could be arranged so that it is more difficult for the municipal administration to dispose of such rental units than to acquire them (Balamir, 1995). Still within a third approach, municipal administrations could effectively act as intermediaries in the local housing markets and provide information services about the rental stock at cost price. An information base could be started as a registry of the rental stock and of tenant and landlord households. Such information management capacities could be achieved by any municipality with a minimum of resources, allocating a few computers and personnel for running the service. Yet the impact of this service on the rental markets is likely to be extensive, especially in those cities where tenancy rates are high. 7 The model provides incentives and protection both to landlords and tenants. Landlords could receive at-cost maintenance and rehabilitation services for their property, secured lease agreements, at-cost legal services, and minimization of vacancy periods. For tenants, this would mean relatively lower rental payments, at-cost services of moving, secured lease agreements, free legal services and immediate identification of candidate accommodation with free visits. The larger the market to be served, the more sophisticated the technology and software that is to be used, including the likely exploitation of GIS capacities. The analysis of the rental sector in Turkey, which proliferated in a more distinct context than in many European countries, could not only remind us of the ethnocentricities that abound in current housing research, but also provides an environment for the experimentation of alternative policies in which complementary roles of the public and private interests could be articulated. 9
399
Notes
*
2
3
4 5
6
7
400
The article was first submitted as a contribution to the policy and practice column. Contributions to this area of study include works of Kemeny (1994), Harloe (1995), Oxley and Smith (1996), Balchin (1996), and others. Examples of organized work cover the UNCHS (1990, 1993), which has published a collection of studies on the developing world, and the special issues of Housing Studies (October 1997; volume 12, number 4) and Netherlands Journal of Housing and the Built Environment (1998; volume 13, number 3), which focus on the European countries. The latter two journals have complemented each other, concentrating their special issues on social and private rented housing respectively. No consistent information and statistical series or Annual Surveys specific to the housing system are available in Turkey. Data used here are either General Census data compiled by the State Institute of Statistics (SIS) or results of sample surveys carried out by SIS for a set of representative settlements. The annual construction statistics, prepared monthly by the municipalities and compiled by SIS, provide basic information about the stock. In addition to the above, results of two other privately conducted sample surveys have also been referred to here. A sample survey of the housing markets of four cities (Ankara, Bolu, Manisa, Gaziantep) carried out in 1992 and a sample (10 per cent) survey of households with reference to the stock of flats that I conducted in Ankara (1985) are further sources of the empirical information presented in this paper. In the 1985 Census, the three major parts of the housing stock in provincial centres (67 settlements) consist of: squatters (12 per cent), flats (48 per cent), and houses (40 per cent). This distribution is rapidly changing in favour of blocks of fiats. Turkish Civil Law which is based on Roman law, treats multiple ownership in property as an anomaly and could tolerate it only as a temporary condition. Almost all forms of land and property easements are as a rule limited to a period of 10 years, after which time decision-making power is once again concentrated under a single authority. The overall rate of accumulating landlords in the metropolitan areas could be estimated as nine to ten per cent of urban households (Balamir, 1992). The majority of 'latent tenants' are newly married couples living with their parents, in the expectation of moving out or inheriting the unit in the long run. A study carried out with a sample of such couples indicated that about one-third moved to the parents' (usually to groom's parents') home (Oktar, 1991). This investigation used Discriminant Analysis to establish in which subgroup any household would fall, depending upon the combinations of known social attributes (age, size, duration of residence, income, educational status, occupation, etc.) of the household. It was observed that the probabilities of successfully allocating the households into the four-party grouping was greater than their allocation to a twoparty grouping of tenants and owner-occupiers (Balamir, 1992, 1996c). This model implies the recapturing of a privatized service by the public sector, relying on the technological opportunities available, and has ideological connotations. Our surveys of the annual volume of payments for such services, the
mobility rates of tenants, vacancy rates, and the size of the tenancy sector corroborated the feasibility of such a project (Balamir, et al., 1998). This service in Ankara for instance, could generate resources other than those necessary to cover its own costs in less than nine months, should the operations capture four to five per cent of the rental sector of the rental market, even though the services were provided to the lower end of the market. Municipal information services on rental stock is primarily a public-relations project, which could prove to be a highly rewarding political activity, to make it worthwhile to complement and at times to compete with the private sector providing such services. The feasibility of such public-led organizations of the rental sector comprising a multitude of individual landlords could also trigger more courageous and formal organizations and large-scale companies or REITs to channel landlords' assets into 'equity stakes with tradable shares', leaving all management concerns to professionals and leading to an 'efficient, expanding and permanent rental sector' (Maclennan, 1998: 399). Even if social housing may be irreversibly converging on a 'residualist model' in Europe, new forms of socialization in the rental sector may be envisaged, "targeted on sectors of the population who are in the economic and political mainstream", as suggested by Harloe (1995), and depending on emerging interests and on means provided by technology.
References Balamir, M. (1969) "Spread of ownership fragmentation in rapid urbanization", Unpublished dissertation proposal submitted to University College London. Balamir, M. (1975) "Kat M~ilkiyeti ve Kentle~memiz", [The process of ownership fragmentation in the urbanization of Turkey], METU Journal of the Faculty of Architecture (2), no. 1, pp. 295-318. Balamir, M. (1982) "Kentle~me, Kentsel Stiregler ve Kent Yaplsx" [Urbanization, urban processes, and urban structure], in Gtil6ksiiz, Y. (ed.), Tiirkiye Birinci Sehireilik Kongresi (1), pp. 13-54, METU Faculty of Architecture, Ankara. Balamir, M. (1989) "Evolution des structures des families urbaines et du logement en Turquie", in Haumont, N., and M. Segaud (eds.), Families Modes de Vie et Habitat, pp. 15-30, Paris: Edition L'Harmattan. Balamir, M. (1992) "Ttirkiye'de Kentle~me ve Kat Miilkiyeti" [Flat ownership and urbanization in Turkey], Unpublished research report, Yunus Nadi Social Sciences Award, 305 pages. Balamir, M. (1995) "Tiirkiye'de Kiracdar ve Kira Kesiminde Politika Segenekleri" [Tenants in Turkey and policy alternatives in the rental sector], in Konut Araw malarl Sempozyumu [Housing Research Symposium], pp. 75-88, Ankara: HDA. Balamir, M. (1996a) "Making cities of apartment blocks: transformation of the builtenvironment in Turkey by means of reorganizations in property rights", in Sey, Y., (ed.), Housing and Settlement in Anatolia: A Historical Perspective, pp. 335344, Istanbul: The History Foundation.
401
Balamir, M. (1996b) "Small households as another", in Komut, E., (ed.), Housing Question of the Others, pp. 523-536, Pre-Conference of Habitat II, UIA and Chamber of Architects, Ankara. Balamir, M. (1996c) "Housing as a process of differentiation: property relations and housing classes in Turkey", Upublished paper presented at the Helsing6r Conference of the ENHR. Balamir, M., A. 0zta~, and M. Tuna (1998) "Yerel Y6netimin Konuta hi,kin Ytildimliiliilderine Bir Farkh Bakl~: Kirahk Stok Kamu Bilgi I~letmecili~i Modeli" [An approach to the housing responsibilities of the local government: the case of public information management on rental stock], Ege Mimarhk (8), no. 4, pp. 1823. Balchin, P. (1996) "Introduction to private rented housing", in Balchin, P., (ed.), Housing Policy in Europe, pp.25-35, London: Routledge. Harloe, M. (1995) The People's Home? Social Rented Housing in Europe and America, Oxford: Blackwell. Kemeny, J. (1994) Understanding European Rental Systems, SAUS Publications, Working Paper 120, University of Bristol. Kumar, S. (1996) "Landlordism in third world urban low-income settlements: a case for further research", Urban Studies (33), pp. 753-782. Maclennan, D. (1998) "The future for private rental housing: surviving niches or flexible markets?" Netherlands Journal of Housing and the Built Environment (13), no. 3, pp. 387-407. Oktar, 13. (1991) "Housing problems of the newly married", Unpublished MCP thesis submitted to the Department of City and Regional Planning, METU, Ankara. Oxley, M., and J. Smith (1996) Housing Policy and Rented Housing in Europe, London: E and FN Spon. SIS, State Institute of Statistics (1989) General Census of 1985: Social and Economic Properties of the Population (1369), Ankara. SIS, State Institute of Statistics (1993) General Census of 1990: Social and Economic Properties of the Population, Ankara. UNCHS (1990) Rental Housing: Proceedings of an Expert Group Meeting, Nairobi: UNCHS. UNCHS (1993) Support Measures to Promote Rental Housing For Low-Income Groups, Nairobi: UNCHS. Watson, V., and M. McCarthy (1998) "Rental housing policy and the role of the household rental sector", Habitat International (22), pp. 49-56.
About the Author
Murat Balamir has degrees in M. Arch. in Architecture, M. Phil. in Planning, Ph.D. in Political Science and Public Administration. He is affiliated with the City and Regional Planning Department of the Middle East Technical University (METU) in Ankara. He is a founder member of the Housing Research Center and the Disaster Management Center at METU.
402