Md. KAMRUZZAMAN and Nobuyuki OGURA

Transcription

1 J. Archit. Plann., AIJ, Vol. 74 No. 636, , Feb., 2009 Md. KAMRUZZAMAN and Nobuyuki OGURA This paper contributes to the view that the housing construction practices in cities in developing countries are differ in nature and extent from, those of developed countries. The findings of an investigation from a field survey conducted in a typical location in the capital of Bangladesh are presented in order to explore culturally rooted housing construction practices. The survey illustrates that households are intimately involved in the process of creating and controlling modern apartment developments. The owner-builders split the whole construction works into several smaller tasks and then build the apartments in a piecemeal process. They first construct their own dwelling units and then construct additional units for rental use depending on affordability, practicality and expediency. About 75% of all dwelling units in the study area are constructed in a self-help piecemeal construction process. The owner-builders gradually acquire construction skills during the construction period and then continue construction of their apartments over a period of years until completion. Keywords: Apartment, Dhaka, Owner-builder, Piecemeal Construction, Self-Help Housing., 1. INTRODUCTION The indigenous style of housing construction has hitherto received very little attention, and yet it is the most popular way of shelter production for the overwhelming majority of urban populations in Africa, Asia and Latin America (Schwerdtfeger, 1982). In most cities in South-East Asia, people have made efforts to increase the production of modern houses constructed with permanent building materials, following the traditional style, using their own meager financial resources and which benefits the urban majority. When the dwellers control major decisions and are free to make their own contribution to the design, construction and management of their housing, both the processes and the environment that are produced, stimulate individual and social well-being (Turner, 1976). This paper will analyze and discuss how local people develop modern housing in the indigenous construction style of the city of Dhaka, Bangladesh. Housing is produced by private or community efforts in cities and villages in Bangladesh. People are forced to depend, in a large part on private sectors for their housing solutions in most urban areas of Bangladesh (Kamruzzaman and Ogura, 2008). Self-help, piecemeal housing developments have penetrated deep within urban centers, surfacing through spot redevelopment of lots and embedding themselves in the urban fabric. It is an integrated hybrid of contemporary technology, new forms and reinterpreted traditional elements. It can be bland, awkward or whimsical yet is always rational, practical and expedient. Despite similarities of forms and symbols; the array of physical elements, proportions, and motifs create a distinctive regional identity for housing in Bangladesh. The legitimacy of the quest for better living conditions cannot be denied and excluding self-help piecemeal housing from the equation, the lack of housing can loom much larger that it ought to. The contributions of piecemeal developers can not be ignored and the significance of this sector has yet to be addressed. Thus the purpose of this paper, therefore, is to outline techniques or production mechanisms for the supply of housing from indigenous households in urban areas in Bangladesh. 2. OBJECTIVE OF THE STUDY The need for scaling up housing production in developing countries through all possible means has now been accepted and acknowledged by all policy makers and commentators in this field (World Bank, 1993; UNCHS, 1996; Tipple, 1994; Okpala, 1992; Keivani, 2001). The public housing paradigm has failed to meet housing needs in many cities in the developing world. Subsequently, many national housing policies are adopting this enabling strategy for housing developments. It is against this background in Dhaka city, Bangladesh that this study undertakes an investigation on prevalent housing construction techniques in line with housing production strategies as a case study. The urban housing demand in the city of Dhaka has essentially been fulfilled by the private sector including self-help owner-builders, small-scale builders and their associates. This sector provides ready housing units for the vast majority of middle and lower-middle income city dwellers. Thus the Ph. D. Candidate, Department of Civil Engineering & Architecture, University of the Ryukyu s, M. Eng. Prof., Department of Civil Engineering & Architecture, University of the Ryukyu s, Ph. D. 323

2 main aim of this study is phrased as: How do the local people construct modern houses despite the absence of institutional technical knowledge, formal finance, and project management and construction experience? Among all the issues, this paper focuses on construction processes and the specific objectives are: - to delineate self-help housing development in Dhaka city; - to explore the process of building modern apartments by piecemeal construction techniques; - to determine the extend that the self-help piecemeal housing sector is operative in urban housing. In fact, this is an ongoing study on how the indigenous people built their housing to meet their own needs as well as provide shelters to the city s huge tenants. The other parts of the study will cover the financial and project management aspects of self-help piecemeal housing in Bangladesh. 3. METHODOLOGY In this study, four neighborhood areas of the Rupnagar Residential Area (RRA) in Mirpur, Dhaka were selected for investigation in order to identify characteristic features of physical growth in self-help housing 1) developments. Selection of the area to be studied is guided by three general principles. The first principle is to ensure that the area represents a typical type. The second is to consider areas which are newly developed where culturally rooted housing construction styles can be found. Finally, it is important to ensure that current popular housing types have been constructed in the area. A great majority of the urban population in the city of Dhaka lives in settlements where they manage to build and maintain their habitats. The characteristics of the RRA meet the three major criteria and are found to be an appropriate example of housing developments within a time span of 25 years. This enables us to better understand the prevalent housing construction style. The study is based primarily on data from small scale interviews of 50 owner-builders, census materials, and a survey of 1,089 housing plots conducted in The 50 owner-builders were selected randomly from four neighboring areas in RRA. A questionnaire consisting of 135 structured questions was directed at household heads or their representatives, i.e. the owner-builders. Physical observations of all housing plots in the study area were undertaken in order to determine housing types, construction types, building materials, providers, building height, no. of dwelling units and finally, development length. Materials from other secondary data have also been cited. This study intends to undertake the retrieval of prevalent construction techniques that city-dwellers often adapt in their housing projects and remain consistent from the supply end. Existing literature in the field of self-help housing has not widely focused on piecemeal construction techniques and has prompted this study to uncover facts, particularly in the context of multi-storey apartment development in a typical Asian city. To this end, the paper will first present the background of self-help housing situations in Bangladesh. In the following section, the paper will focus on the study area and descriptions of the surveyed housing plots. Different housing types during the construction phases and incremental construction styles from piecemeal developers which are found both in the survey as well as from the interviewee homeowners, will be discussed. In the concluding section, this paper will recapitulate the piecemeal construction process and focus some of its strengths, as well. 4. CONTEXT OF SELF-HELP HOUSING IN BANGLADESH Bangladesh is predominantly a rural country, where self-help building is especially significant in traditional peasant communities. Traditional rural housing in Bangladesh is largely based on the use of locally available resources as prime building materials, usually in a process of self-help building undertaken by the community. This housing is well adapted to a natural environment with widely available resources, and supports people s direct involvement in the construction of their dwellings. Throughout the past, local availability of natural building materials and their employment in self-help construction has sustained traditional architecture. However, during the last few decades, due to the advent of the cash economy, industrialization, scarcity of natural resources and population pressure, traditional self-help building processes have been greatly effected or changed entirely. In Bangladesh, affluent households are shifting to manufactured materials and often use skilled builders. It is no longer possible to generally witness self-help traditional housing in its former idyllic light of bucolic harmony (Ahmed, 1998). The abilities of the middle and lower-middle income households to construct housing are declining due to recent trends of utilizing costly construction materials. Thus self-help, piecemeal construction remains the profound option that the urban households widely practiced in their housing construction activities. 4.1 Construction Materials To understand changes to self-help housing in this context, use of the main building materials are outlined. Only the most predominantly and widely used building materials used in indigenous housing in this region are discussed. Bamboo and thatching materials are two main natural building materials. The results of a national survey in 1982 showed that more than 60% of total households in Bangladesh lived in houses that used bamboo as a main building material, and less than 5% of houses were built of manufactured materials such as brick and cement (BBS, 1988). However, these natural materials are gradually becoming expensive, and are no longer readily available. Brick for wall materials and corrugated iron (CI) sheets as roofing materials is more prevalent in urban areas, and affordable by a relatively larger majority of affluent households. In the recent past, people have become more accustomed to utilizing brick, cement, sand and steel despite the higher cost and the use of traditional building materials have been rapidly declining due to changes in life style, dwelling habits and trends towards western-style living in urban areas in Bangladesh, including Dhaka city. Availability of building materials is crucial to the growth of housing. Increased construction activities, their dependence on high-cost building materials and components and inadequate supply have pushed up the cost of housing. The escalating cost of building materials has had a negative impact on house construction. During the last 10 years, prices of major building materials have gone up by more than twice (Fig. 1). From 1970 to 1977 housing expenses in Dhaka city increased at a relatively gradual rate. From the mid-1980s there was a momentum of increasing housing expenditure and between and there was an abnormal 324

3 and steep increase in the cost of housing compared to food and general trend because of political unrest followed by a huge financial crisis. As housing expenditure has a greater influence on the total living cost, following Fig. 2, affirms the concomitant sharp increase (since 1990) in the general living index. Until today the cost of living index shows an upward trend. The escalating price of building materials and consequent lack of affordability has lead to adaptation of piecemeal development technique arranged within self-help labor by the urban households in Dhaka city. Price Index ( =100) Steel Sand Brick Cement Fig. 1 Price Index of Major Building Materials in Dhaka (Source: BBS, 2006) 4.2 Self-help Housing Index Number ( =100) Housing General Food Fig. 2 Cost of Living Index in Dhaka City (Source: BBS, 2006) Self-help is characteristic of most traditional housing (Oliver, 1987). This does not necessarily imply that a household possesses all the skills to construct a house entirely on his own. Quite often the services of specialist builders are required, traditionally distinguished from professional contractors by the fact that they do not usually receive payment in the form of cash, but provide service in exchange for food or future favors from the household. The specialist might be family friend or neighbor, and his service is an accepted feature of the community. His role is usually minimal; requirements of construction are generally understood by the community as a whole and the specialist helps only in the fabrication of certain items requiring special skills. For example, construction of the roof requires a specialist, while the rest of the house can be built by the household. When necessary, extra labor is also paid for in kind. Supervision of house construction is conducted by households whom are generally not experienced about technical aspects, and hence supervision is generally not adequate (Ahmed, 1998). The most common way of building a house on a self-help basis in Dhaka city can be categorized into three distinct types. Those are: (1) complete self-help (2) partial self-help and (3) contract basis. In the first case, the households recruit the masons, carpenters and several assistants from nearby location according to the size of the work and construct the whole work under their own supervision. In such a case, family members frequently assist with the transportation of building materials and the finishing of interior walls and floors. In the second case, the household assigns major work to a master mason, steel workers and carpenters and the payment is made on a volume-of-work basis. In this case, the households remain in supervision as usual but the small items of works are done through self effort or employ daily labors. In the third case, households assign the whole work to small-scale builders. The homeowner only supplies materials and supervision STUDY AREA It is helpful to first outline a few features of the study area that are pertinent for understanding issues of housing development in these settlements. To ease the housing problem in Dhaka city, the National Housing Authority (NHA) took up the Rupnagar project in 1982 on 35 hectares of land divided into 1,200 plots of different sizes, located at Pallabi in Mirpur, about 6 km from the city center (Fig. 3). Of these, 31 plots were kept designated as commercial in the project profile in order to fund the project without government funds. Rupnagar is a planned residential area; it contains plots earmarked for schools, colleges, park, hospital, lake, shopping complex and other civil amenities. The government developed these areas and allotted them to the people, mostly from the middle class. The whole study area was divided into 4 parts in order to simplify the analysis of the development pattern. Part 1 included the plots between roads 1 and 10; part 2 those between roads 11 and 16; part 3 those between roads 17 and 23, and part 4 those between roads 24 and 33 (Fig. 3). 5.1 Plot Size and Orientation The plot sizes of these projects were examined and found to range between 117 and 335 m 2. In part 1, all plots were 335 m 2 in area. The plots in part 2 were 234 m 2 in area, with the exception of a few corner plots. Plots in part 3 were 167 m 2 in area, and those in part 4 were 117 m 2. The minimum plot size was 117 m 2 (in part 4), while the maximum plot size of 335 m 2 appeared in part 1 (Fig. 3). The typical plot was rectangular in shape and lay along the N S axis, with its house built along the E W axis, thus avoiding the direct sunlight from the east and west and capturing the breeze. Regarding plot sizes, one can guess there is a clear tendency towards reducing plot sizes. This can be taken as evidence that the issue of optimum space requirements has not been yet been fully addressed. This fact might be the outcome of the mechanical re-adoption of old designs without revision in new projects. When the RRA project was launched by NHA in the early 1980s, Dhaka was not densely populated and land was available at reasonable prices. Recently, due to rapid urbanization and its eventual pressure on residential lands for housing in Dhaka, NHA has shifted its strategy to providing apartment units rather than providing land. 325

4 Part 4 Total Plots = 359 Plot Size = 117 m 2 Roads 24 to 33 N Part 3 Total Plots = 326 Plot Size = 167 m 2 Roads 17 to Gridiron Pattern In RRA, a rectilinear gridiron pattern has been adopted as the housing pattern. Plots are arranged along the main road at inclines of about 10 degrees in part 2, 15 degrees in part 3 and 30 degrees in part 4, with respect to the N-axis. This, however, does not seriously affect heat accumulation or natural ventilation. The regular, rectangular forms, devoid of acute angles and redundant spaces, ensure the maximum utilization of the land. The simplicity of the patterns helps residents to form clear mental images of the neighborhood so that residents are seldom lost or confused about their locations at any point within the area. The simplicity also facilitates the design and execution of infrastructure items, especially as a result of the uniform optimum gradients in underground works, which make their provision and maintenance more convenient. 5.3 Densities The gross density of the RRA was about 1,925 persons per hectare. No reliable figures are currently available for housing density in Dhaka. Population density increased in order from part 1 to part 4. 6 PIECEMEAL CONSTRUCTION OF MULTI-STOREY APARTMENTS Piecemeal construction is a construction style wherein the buildings are designed for their Part 2 Total Plots = 201 final constructed size, but within which the construction takes place incrementally over Plot Size = 234 m 2 Roads 11 to 16 time. Homeowners generally move into semi-finished buildings and then gradually continue the construction according to the choice, need and affordability. The straightest answer regarding the reason behind the prevalence of this type of construction is a lack of external finance meeting the construction costs. It is not within the scope of this paper to cover the entire range of causes behind the evolution of the type of construction used in indigenous housing of the region, thus the focus in on construction styles only. Part 1 Total Plots = House Types in the Construction Phases Plot Size = 335 m 2 In order to assess different construction phases, it is necessary to define the different Roads 1 to 10 house types that households often construct before beginning apartment construction. For this purpose, housing forms can be defined as: (a) temporary house (b) semi-permanent house and (c) permanent house (apartment). Mosque The general tendency of most households is to withstand possession of the land first. Health Center High School Thus the households build some form of a temporary structure in order for the tenants to Primary School ensure land possession. Most tenants, in this case are slum dwellers. They pay a lump-sum Public Housing rent or sometimes stay on a free-hold basis. Temporary houses are locally known as kutcha m houses. Kutcha houses generally have an earthen plinth with bamboo (sometimes timber) posts and a wall and roof made of temporary materials. The walls are made of straw, jute Fig. 3 Study Area 2) (1,089 Plots Surveyed, 2007) stick, mud, and un-burnt bricks. Roofs may be made from thatch rice, wheat, maize straw, or catkin grass, along with split bamboo. In the study area, the highest percentages (10%) of temporary houses are found in Part 1 area where the plots are of a large size (Fig. 4). Semi-permanent houses are locally known as semi-pucca house. It has a shallow brick Vacunt Temporary Semi-permanent Apartment foundation with brick soling and ordinary floor finish. Walls are often built with brick or 100% CI sheet. Roofs are with CI sheet with timber framed or steel angle framing in some 90% 80% cases. Generally, the workers have no formal training. Masons are trained by their 70% seniors. The owners are the architect and masons are the engineer for this type of 60% housing. The houses are L-shaped or U-shaped with a central courtyard ensuring 50% 40% maximum utilization of land (Photo 1). A typical semi-pucca house is comprised of 6 to 8 30% rental units. This is a very popular form of housing as there has huge demand for rental 20% units in all four areas of the study. The rental income in a semi-pucca house and an 10% 0% apartment has no significant difference while the investment cost of the latter type is Part 1 Part 2 Part 3 Part 4 quite large. Thus, the households often build this housing type and then accumulate a Fig. 4 Different House Types (%) in Each Part of the RRA House Types (%) in Each Part 326

5 rental income for further investment in building apartments. The presence of this type of housing in large numbers in four areas of study reveals this fact (Fig. 4). Permanent houses are widely termed as pucca house and most often are meant as an apartment. In the study area, apartments may be classified in two different structural forms. The first type is a load-bearing brick structure and the other is a reinforced concrete (RC) frame structure. Traditionally, a 10 inch (25.4 cm) brick wall is used as the load-bearing wall in the first type and a 5 inch partition wall is used in the latter type. Roofs, in both the forms are constructed with RC and floors with mosaic tiles according to the ability, choice and motifs of the households. Apartments made with a load-bearing brick structure are gradually disappearing because of their unsuitability to vertical extension. In the surveyed houses, less than 5% are apartments with a load-bearing structure. Apartments are the most preferred housing option in the study area. As in the Fig. 4; the highest percentages (61.7%) of multi-storey apartments are developed in part 3 area. Average 49.4% of all housing plots in the RRA are developed with multi-storey apartments. 6.2 Apartment Construction The buyers of the plots in RRA belong to the middle class. They own the plots at a subsidized rate through the government site and service scheme. Thus the land cost was not a big burden to the households as compared to the construction costs. The entire financing in order to meet construction costs at different stages of construction is managed by the households with their own savings, informal loan from relatives/friends, and selling assets from rural origin. The long term goal of the households is a monetary return on investments through the addition of rental units. Almost all the interviewee households (90%) respond that they prefer to investments on housing in order to ensure decent housing within the city, income generation and to secure a source of income at the later stages. The households have perfected prototypical layouts that permit a variety of incremental transformations. The key elements ensuring this versatility are: - A building system which can accommodate additions and repartitioning. A reinforced concrete frame with brick or hollow block infill is the most widely used structural system (95%). It is cost-effective in terms of the desired level of safety, durability, flexibility and ease of maintenance. - A staircase to future upper floors conveniently located relative to street access in order to allow internal partitioning of the floor area into independent units. Thus the household choose multi-storey apartments built with RC frame structure which fosters vertical extension and internal partitioning to provide further rental accommodation. Households contract out construction to workers in the building trades. They retain control of the rate of build up, set the priorities, choose from alternatives and decide on the sequencing of works. The workers freely replicate, adapt and blend building techniques and styles. The choice of materials and systems is dictated by economy, practicality and expediency. The choice of motifs and colors is determined by the preference of both the builder and tenants. They share a dislike of oldfashioned styles and a conscious admiration of the new. An illustration of the complete piecemeal construction phases which exist in the surveyed plots is shown in Fig. 5, though this omits some other micro-phases of construction, for reasons of simplicity. Households generally provide a boundary wall (Phase a) soon after purchasing and finishing the land registration process. It involves Full Experience no significant construction works or experience since the process is limited to brick work. No interviewee households were found who have not built a form of temporary Degree of or semi-permanent structure after purchasing the land. The slum-dwellers often build Experience temporary housing that the landowners accept in order to retain the possession of the land. In most cases, households construct semi-permanent structures (Phase b) which No take less capital but generate a handsome rental income (Photo 1). In this phase; the Experience a b c d e f g h i j construction time, investment and households gain in construction skills are little Zero Completion Degree of Completion Full Completion increased (Fig. 5). In the surveyed plots; 42% of housing structures in part 1 area, Construction Phases: 20% in part 2 area, 21% in part 3 area and 51% in part 4 area are semi-permanent Phase a Construction of boundary wall housing (Fig. 5). Semi-permanent structures in the whole study area are, on average, Phase b Construction of temporary/semi-permanent house Phase c Sub-surface construction/pile driving/sand filling 33% of all housing structures. Phase d Construction of girders, columns and brick works Generally the households take some breathing time before starting the foundation Phase e Casing of roof slab (either partial or full) Phase f Finishing the ground floor (Owner s portion) works of apartments. The principle is to accumulate sufficient capital from rental Phase g Finishing the ground floor (for Tenant s) income through semi-permanent housing and then further investment in apartment Phase h Construction of stair case, columns of next floor Phase I Casting the roof slab of next floor (partial or full) construction. Thus semi-permanent houses may be treated as stepping stones towards Phase j Finishing the next floor (partial or full) the building of apartments. The investment recovery time is less and hence further Fig. 5 Typical Model 3) of Piecemeal Construction Process income helps them to build apartments. Almost all the households (99%) plan to build their apartments with 6-storys or with a maximum permissible height as stated within building bylaws. Thus there needs to be a strong foundation which is often in the form of deep (pile) foundations. Cast-in-situ piles are widely used which involve good investment as well as construction time. Thus in phase c, the households acquire quite good construction skills include sub-soil investigations and foundation constructions. After completing the foundation works, the households raise the land up to plinth level (about 0.5m higher than the maximum flood level) by sand filling 4). 327

6 At phase d; the households start to build grade beams, columns and brick works. Generally, people try to cast the roof slabs soon after completion of the brick works. Depending upon affordability, practicality and expediency, people decide to cast whole or partial roof slabs at phase e. Within the study area, both of the following were common practices: (1) casting half of the roof and quickly finishing the interior and (2) casting the whole roof and then finishing the interior over a longer period of time. After casting the roof slabs, the owners first move into their semi-finished houses and gradually complete the remaining work by self labor (phase f). During the next stage, the owners complete the remaining portions of the same floor for their tenants who are then invited to move in (phase g). The households then take more time for further construction. During this downtime, households minimize their financial transactions and prepare for next construction. After a good interval, in phase h, people start by casting staircases, columns and brickwork for the next floor (Photo 2). As usual, households decide whether to cast the whole roof or just half based upon their local situations (phase i). As the interior finishing work require additional financing, people often do this incrementally (phase j) and an incomplete first floor is seen in Photo 3. Again, given favorable situations and time, households continue constructing subsequent floors (Photo 4). Meanwhile, the households attain good construction skills and attain bulk rental income, which they utilize to construct subsequent floors, as seen in Photo 5. In later stages, construction speeds are faster than in the initial stages due to the broader construction skills of the owner-builders and well known track to informal finance. The households continue construction until the final floor, as designed in the master plan. Finance remains the main constraint on construction speed. In Photo 6, two buildings are compared in order to differentiate the outcomes from piecemeal and non-piecemeal construction. Owner-builders of different educational, professional, social backgrounds gradually acquire construction skills during the construction of their buildings and were found to be adequately satisfied and experienced at the end of construction. Thus self-help piecemeal construction remains a popular housing construction style in Bangladesh. 6.3 Timeframe of Piecemeal Development It is a little tricky to outline any time frame or to break entire construction works down into specific incremental steps over time, which can help determine the time required to complete the construction projects. The fuel for construction is money, and hence construction speed depends directly upon the availability of finances Semi-permanent st Story nd Story rd & 4 th Story Fig. 6 Development Timeframe of a Typical Piecemeal Residential Building 328

7 Households manage the finances underpinning construction projects in the face of variable and individual circumstances. Thus construction time varies widely from household to household. An empirical sequence of construction with time is outlined in Fig. 6, taking Mr. Golam Mohammad s house in part 3 as an example from among the 1,089 surveyed housing plots. Mr. Mohammad bought his plot in 1982 and had constructed 4-story apartment by Thus, it took him a total of 23 years and he has no desire to perform additional construction. Examples from other households show no significant disparities with this example and an estimated 20 to 25 years is needed to complete a project. 6.4 Piecemeal Developers Multi-story apartments are being built by the public sector, formal private developers and self-help owner-builders in the study area, and the details of the three providers development figure are presented in Fig. 7. In the study area, only 5 public apartments were found in part 1 area. In the other three parts, there were no public apartments. Good numbers of private formal developers apartments were found in part 1 area rather than part 2 and part 3 area (Fig. 7). This is due to the inability of the owner-builders to build apartments on large plots with their own finances, who thus sold land to the formal developers to build apartments. Self-help owner-builders were the largest housing providers in the study area. About 92.6% of all apartments in RRA were built by owner-builders; and in smaller plots, this figure was above 98% (Fig. 7). These self-help owner-builders included piecemeal developers and non-piecemeal developers and are further analyzed in Fig. 8. As a whole, piecemeal developers contributed 71.1% of the apartments in RRA, while 21.5% belonged to non-piecemeal developers (Fig. 8). Apartment Provider's (%) in Each Part Public Sector Formal Developer Self-help Owner-Builder 100% 75% 50% 25% 0% Part 1 Part 2 Part 3 Part 4 Fig. 7 Apartments Provided by the Different Providers in RRA Piecemeal & Non-Piecemeal Apartments (%) Formal Public & Private Developer Non-Piecemeal Developer Piecemeal Developer 100% 75% 50% 25% 0% Part 1 Part 2 Part 3 Part 4 Fig.8 Apartments Provided by Piecemeal & Non-piecemeal Providers Part 4 Part 4 1-Story 3-Story 2-Story 4-Story Part 3 Part 3 5-Story 6-Story Part 2 Part 1 6-Story 5-Story 4-Story 3-Story 2-Story 1-Story Part 2 Part No. of Apartments Fig. 9 Apartments of Different Heights & Numbers Provided by Piecemeal Developers No. of Apartments Fig. 10 Apartments of Different Heights & Numbers Provided by Non-Piecemeal Developers Fig. 9 is an attempt to evaluate the piecemeal developers overall contribution to housing within the study area. It is evident from Fig. 9 that buildings of moderate height in relatively smaller plots had the maximum piecemeal development rate. For example, in part 3, four-story apartments were most prevalent. In the same context, Fig. 10 demonstrates that the self-help non-piecemeal developers had different growth tendencies. For them, taller buildings on large plots were preferable. Six-story apartments made up about 58% of all non-piecemeal apartments and were common in part 1, part 2 and part 3 as shown in Fig. 10. In the study area, about 80.3% of all housing plots included piecemeal construction techniques including semi-permanent structures, and 75% of the total dwelling units were developed mostly as rental accommodations. 7 CONCLUSION The production of public housing is far too little and too slow to make a dent in the housing shortage afflicting most households in Dhaka. The sharp inflation in construction costs has eroded prospects of increasing affordability and added more stress to their housing solution. In these critical situations, indigenous people solve their own housing problems in their own styles. The main object of this paper has been to provide detailed information on the indigenous construction style for building modern apartments, giving examples from a Bangladeshi city and centering on the physical intensification process in the context of apartment development. Based on the above analysis and discussion, the following conclusions can be drawn. 329

8 - Self-help housing is the main form of housing development in Bangladesh. The study area was a newly developed area, all of whose houses were built within the last 25 years. The relatively smaller plots in the study area were found to have high development rates, whereas the larger plots had slower growth. This was due to the people s inability to support the costs of construction on large plots. About 90% of all apartments were produced through self-help efforts, including piecemeal and non-piecemeal construction processes. - In the self-help, piecemeal construction process, people used to first build semi-permanent housing on their plots. They then allowed a long period of time to elapse before starting apartment construction, in order to save up sufficient money from the rental income derived from the semi-permanent housing. Thus semipermanent housing acted as a ladder to apartment building. The construction project as a whole could be split into several smaller works and the owner-builders build their apartments gradually, simultaneously gaining construction skills, and kept on constructing until the completion of their projects. Owner-builders constructed their own dwelling units first and moved into the semi-finished apartments. Thereafter, they constructed additional units for rental use within the confines of their affordability, practicality and expediency. In later stages, construction speed was faster than in the initial stage due to their having gained construction skills and well known source of informal finance. At the end of construction, owner-builders were found to be adequately satisfied and experienced. Thus, self-help piecemeal construction is the most widely-practiced housing construction style in Bangladesh. - About 75% of all dwelling units are provided by piecemeal developers, mostly as rental accommodations. Within the prevailing difficulties, this sector holds the role of the single largest housing provider in the study area, and has become a popular form of housing construction. - Due to the lack of external financial support, people used to construct their houses according to their own financial abilities. Obviously, the energy for housing supply is slower and thus progresses incrementally. Supply speed can be increased if external finances are fed into the supply chain. - It takes nearly years to complete a house according to its master plan. Obviously, completion largely depends on affordability and other socio-economic characteristics of the owner-builders. There is no immediate panacea for the problems of promoting housing supply, and the sustainable utilization of self-help energy has the potential to produce promising results in cities in Bangladesh. It may offer a possible source of recovery from the decline of culturally-rooted building practices and renew the power of people to be intimately involved with and control the process of creating dwellings for habitation. Notes 1) Apart from squatter settlements, self-help housing is also expressed by the illegal subdivision of land, occupying urban peripheral areas (Gilbert and Ward, 1985). 2) Aerial photograph of the study area can be found at 3) In order to construct a typical model for piecemeal construction process in Dhaka city, we used degree of completion of construction works and degree of construction experience as a comparing tool. This is an empirical estimation and both the parameters are unit less. Models of these kinds can be found in Payne (2000) for further reference. 4) In order to cope with the frequent flood, plinth level of buildings is generally raised well above the highest flood level by filling materials like sand. References 1) Ahmed, I. (1998) Crisis of Natural Building Materials and Institutionalized Self-help Housing: The Case of Grameen Bank in Bangladesh, Habitat International, 22(4), ) BBS (1988) Households and Housing Structure in Bangladesh: Evidences from a Demographic Sample Survey, 1982, Bangladesh Bureau of Statistics (BBS), Dhaka. 3) BBS (2006) Statistical Pocketbook of Bangladesh (Dhaka: Reproduction, Documentation and Publishing Wing, BBS) 4) Gilbert A. and Ward P. (1985) Housing, the State and the Poor: Policy and Practice in Latin American Cities, CUP, Cambridge. 5) Kamruzzaman M. and Ogura N. (2008) Housing for the Middle Income Group in Dhaka, Bangladesh With a Focus on Affordability and Options, Journal of Architecture and Planning (Transactions of AIJ), 73(627), ) Keivani R., and Werna E., (2001) Refocusing the Housing Debate in Developing Countries from a Pluralist Perspective, Habitat International, 25(2), ) Oliver, P. (1987) Dwellings, Phaidon, Oxford, p.11. 8) Okpala, D. C. I. (1992) Housing Production Systems and Technologies in Developing Countries: A Review of the Experiences and Possible Future Trends/Prospects. Habitat International, 16(3), ) Payne, G. (2001) Urban Land Tenure Policy Options: Titles or Rights, Habitat International, 25(3), ) Schwerdtfeger F. W. (1982) Traditional Housing in African Cities, John Wiley & Sons Ltd., Bath, Avon. 11) Tipple, G. A. (1994) A Matter of Interface: the Need for a Shift in Targeting Housing Intervention. Habitat International, 18(4), ) Turner, J. F. C. (1976) Housing by People, Marion Boyars Publishers Ltd, London. 13) UNCHS (Habitat) (1996) The Istanbul Declaration and Habitat II Agenda. Istanbul. 14) World Bank (1993) Housing: Enabling Markets to Work. Washington, DC: World Bank