{i.l.t.ltun..lil|‘1¢ 1. 1 ‘ OJ. .0»: vvl. 71511.9"! .- III-til!!!- 50. st. 0. :Il ‘v ((-15 .1 2f ID‘OADVIDIDVA . V (a to .151: II‘IIQQSth‘ art: ,5 {Polv'e'irl 50.731. EMU: lhl bv‘ztnx?frntlilh«wflv‘(lli I . 1a.: 12. v. 9 i .2 0? 1V tickuchr? at! 9 (rein! ollll S i. 9‘; ;:::I! I!!! Silo ll Pf IE: 1.3).}... 1|! 4.1!: l li“1|o' . 0n”. II....:¢¢ I It") rt vtlll It!!!“ 1". 21"?le .l ,Alv. I'III‘II ".l.’ "IVI‘. .Ivuio .‘tbtlo. '12:! :0- ...o‘ 123,}! ilvll I... TE UNIVERSITY LIB ARIES W llll‘lllllllllm lllll I 3 1293 00791 4652 LiBRARY Michigan State 1 University This is to certify that the thesis entitled URBAN RANK—SIZE RELATIONSHIPS IN BANGLADESH: 1901-1981 presented by Jashinta D'Costa has been accepted towards fulfillment of the requirements for M.A. Geography degree in Date November 19, 1992 0-7639 MS U is an Affirmative Action/Equal Opportunity Institution PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. DATE DUE DATE DUE DATE DUE MSU Is An Aflirmetive Action/Equal Opportunity Institution cm“.- URBAN RANK-8123 RBLAIIONBHIPB IN BANGLADESH: 1901-1981 BY Jashinta D’Costa A Thesis Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Geography 1992 ABSTRACT URBAN RANK-SIZE RELATIONSHIPS IN BANGLADESH: 1901-1981 BY Jashinta D'Costa This study analyzed the trend in the city-size distributions for Bangladesh from 1901 to 1981. The urban structure is analyzed using rank-size plots, a rank-size regression model, and a rank-mobility index. The study has suggested that dual primacy in Bangladesh has disappeared and the urban system is tending towards deconcentration with primacy still in existence. A concentration which was in a process of emergence during the British period (1901-1947) was manifested in the Pakistani period (1947-1971) and continued through the early period of Bangladesh. Dual primacy was strong during the British regime, diminished during the Pakistani regime and disappeared during the Bangladeshi period. A process of deconcentration has been established. However, the cities in the western part showed a declining trend. The top few cities continued to maintain their higher status and the lower ranking cities their lower status. DEDICATION To my father (late), mother, brothers, sister and their families whose love and support have always been with me. iii ACKNOWLEDGEMENTS I am deeply indebted to the people who have enabled me to successfully complete this thesis. First, I would like to extend my thanks to Dr. Assefa Mehretu for his guidance and support throughout my graduate studies at the Michigan State University. His constructive criticism and guidance have helped me to understand the rank-size rule in a real world perspective. My thanks are also due to Dr. Richard Groop, a member of my committee, for his cooperation and helpful suggestions throughout my academic period. Also, I wish to thank Dr. John M. Hunter for joining my evaluation committee and making helpful suggestions. I am also thankful to Michael Lipsey for providing me with computer assistance, and Ellen R .White for permitting me to use the cartography laboratory. I am indebted to Dr. Gollapalli Prakasa Rao for his prayerful help and suggestions. Also, I wish to thank the members of the Little Flock Christian Fellowship for their spiritual support. I am thankful to my mother and brothers for their encouragement and support throughout my studies. Finally, special thanks to my nice, Margaret, for her love which always kept me cheerful. iv List Of Tables 0 O O O O O O O O O O O O O O O 0 List of Figures . . . . . . . . . . . . . . . . . CHAPTERS I. INTRODUCTION . . . . . . . . . . . . Statement of the Problem . . . . . . . . Purpose of the Study . . . . . . . . . . Organization of the Study . . . . . . . II. BACKGROUND ON URBAN DEVELOPMENT IN BANGLADESH Bangladesh: An Overview . . . . . . . . Urbanization in Bangladesh: An Historical Perspective . . . . . . . . . . . Hierarchy of Bangladeshi Urban Centers . Problems of Urban Development in Bangladesh Primacy . . . . . . . . Site and Situation of Dhaka. Rural-to-Urban Migration . . Functional Concentration . . Internal Characteristics . . Growth of Large Cities: A Theoretical Perspective . . . . . . . . . . . Strategies to Address Large City Problem Summary III. THEORIES AND METHODOLOGY APPLIED IN THE RESEARCH IV. Introduction. . . . . . . . . Theory and Methodology of Analysis The Rank-Size Model. . The Rank-size Plot . . Rank-Mobility Index. . Source of Data. . . . . . RESEARCH RESULTS . Introduction . Rank-Size Plot Dual Primacy . Discontinuity in the Growth of Dhaka Relative Absence of Mid-Size Cities Rank-Interchange. . . . . . . . Increase in the Number of Urban Centers V vi vii hIAF‘H mm 15 15 18 19 23 23 24 3O 32 34 34 34 34 39 40 42 43 43 44 46 48 SO 51 52 Regression Results of Rank-Size Plots Overall Pattern . . Early Pattern . . . Mid-Period Pattern . Late Pattern . . . Urban Patterns Analyzed Using the Rank-Size Mobility Index . . . . . Classification of Cities by Rank-Mobility Index 0 O O O O O O O O O 0 Cities Showing Unique Patterns of Performance . . . . . . . . . . . . . . . V. SUMMARY AND CONCLUSION . . . . . . . . . . . Historical Development of Rank-Size Relationships of Urban Centers . . . Structural Change in the Urban System Patterns of Changes in Rank-Order . . Policy Implications . . . . . . . . . APPENDIX Population and Ranks of Bangladesh's urban centers, 1901-1981 . . . . . . . . . . . . . LI ST OF REFERENCES 0 O O O O O O O O O O O O O O O O 0 vi 53 54 57 57 58 59 63 65 7O 71 74 75 77 80 9O 2.3. 2.4. 2.5. Percent of urban population in selected Asian countries, 1982 . . . . . . . . . . . . . . Hierarchy of urban centers, 1981. . . . . . . Growth Of Dhaka, 1901-1981 . . . . . . . . . Bangladeshi primacy index, 1901-1981. . . . . Component of population growth in six major cities of Bangladesh, 1961-1974 . . . . . . . Growth rates of Dhaka and Chittagong, 1901-1981 Growth of population of Dhaka, 1901-1981 . . Results of regression . . . . . . . . . . . . Rank mobility of Bangladesh's urban centers, 1901-1981 e o o o o o o o o o o o o o o o o 0 Classification of urban centers according to Rank mobility indices . . . . . . . . . . . . vii 11 16 16 19 46 49 54 61 64 LI§I_QE_EI§QBE§ Administrative districts in Bangladesh . . . . Net migration for Bangladeshi cities, 1961-1974 The core-periphery model . . . . . . . . . . . Railway services in the Indian sub-continent . Polarization of development during the Pakistani per 10d O O O O O O O O O O O O O O O O The rank-size rule . . . . . . . . . . . . . . (a) Tendency toward perfect evenness (b) Tendency toward perfect concentration . . Patterns of city size distributions . . . . . Rank-size curves, 1901-1981 . . . . . . . . . Growth of Dhaka and Chittagong, 1901-1981 . . Slope coefficients (b) for the rank-size relatiODShips O O O O O O O O O O O O O O O O Variations of slopes for urban centers, 1901-1981 Rank-size changes, 1901-1981 . . . . . . . . . viii 21 26 27 29 36 38 40 45 47 55 62 CHAPTER I INTRODUCTION A well articulated urban system ensures an equitable distribution of development benefits throughout a country. In such a system, the relationship between a city and its hinterland is generative rather than parasitic and deleterious core-periphery relationships are absent or less pronounced. Unfortunately, most of the less developed countries (LDCs), against a background of a long history of colonization, a rapid population growth, and little industrial development, have a characteristic trend toward polarization in urban development. Theories of urban system growth that postulate initial primacy (polarization) in urban patterns suggest that in most LDC's the rank-size relations (convergence) work poorly (Berry, 1971). t n o Bangladesh, in South Asia, was under the colonial rule of Britain and Pakistan for more than 200 years. Earlier studies have shown that two phenomena, the British use of Bangladesh as a hinterland of Calcutta and the Pakistani concentration of industries in a few large cities, appear to have impeded a systematic and spontaneous evolution of the Bangladeshi urban system. This may have resulted in a parasitic relationship between the primate city and small towns. The resulting functional concentration has caused 2 population concentration in Dhaka, thus making it a dominant city. Furthermore, functional concentration has also caused rural to urban migration, disparities between city and countryside and between large cities and small cities in terms of development. Like other developing countries, Bangladesh is characterized by a strong concentration of urban population growth in the primate city (United Nations, 1983). The United Nations demographic survey report has predicted that by 1991, the population of this city will be three times larger than that of the previous decade (Rondinelli, 1983). This is an indication of what might happen if the observed trend continued. The government of Bangladesh initiated a decentralization policy in the Second Five Year Plan (1980-85). However, it is not well known if the second five year plan made a positive impact on the urban system development and in reducing primacy. As the government's policies aimed at achieving national development through the urban network, the impact of the policy can only be evaluated by examining the changes in the structures and development patterns of the urban system. Studies of the urban system in Bangladesh are disappointingly sparse and thoroughly inadequate. There has been only one study by Moudood Elahi (1972) on city-size distribution in Bangladesh. A critical examination of the structure and development patterns of urban system has yet 3 to be done. Elahi (1972) also emphasized the importance of further study on this vital subject as it would explain the impact of the liberation war of 1971 and antecedent development of urban functions on the urban system in Bangladesh. Later, Carroll (1982) has suggested that a thorough study is necessary to test the observed temporal difference between urban population size and rank, using slope parameters which was not done in Elahi’s study. The study of city-size distributions is one of the several major branches in the field of urban system analysis. Implicit in this line of research is the importance placed on the ’size' of cities as a measure of population concentration (Fan, 1988). The relationship between population size and rank provides an important framework for understanding city-size distributional patterns. Very often, researchers have used rank-size plots for visual comparisons at different points in time. However, since there is a lack of consistency in the number of cities included in each rank-size plot and since a spatial component is not inherent in rank-size plots, interpretation of patterns has not been easy. ose t 0 ud The purpose of this study is to examine changes in the urban structure of Bangladesh from 1901 to 1981 and to evaluate current trends in the urban system. More specifically, the study will: 4 a. Examine the historical development of rank-size relationships of urban centers with a special reference to political changes from 1901-1981. b. Examine the structural change of the urban system. c. Identify and analyze the factors that have influenced the growth and decline of the various hierarchies in the system. This study will address the following major research questions: 1. What changes have taken place in the distributional patterns of urban places over the political history of Bangladesh? 2. Has the urban system moved toward deconcentration (rank-size) or concentration (primacy) over time? 3. What is the spatial trend in the current urban system of Bangladesh? The study will use visual comparisons of rank-size plots at different points in time and perform a rank-size regression analysis for examining the overall pattern of the rank-size relationships. Since the location of each city and its context within the total system are also important for gaining a more realistic understanding of the urban system, this study will also use the rank-mobility index as an additional tool of analysis. 0 at The study is presented in five chapters. Chapter I outlines the background and the purpose of the study. Chapter II reviews the literature relating to the background 5 on the urban development in Bangladesh. Chapter III discusses the methodology used in the study. Chapter IV presents the results of analysis. Finally, Chapter V contains a comprehensive assessment and conclusions based on the results obtained from the study. CHAPTER II BACKGROUND ON URBAN DEVELOPMENT IN BANGLADESH The purpose of this chapter is to review the literature that has particular relevance to the subject matter of this study. The first section gives an introduction to Bangladesh. The second section provides an historical account of urban development in Bangladesh and discusses the contribution and impacts of the various political regimes the country experienced on the urban structure. The third section discusses the urban system in Bangladesh beginning with a classification of towns into different size categories, and analyzing the function of each category of urban centers. The fourth section discusses urbanization problems. The fifth section examines government strategies to deal with problems associated with large cities in Bangladesh. cs ° w Bangladesh, situated in South Asia, lies between the eastern margin of the Indian subcontinent and the western fringe of vast Southeast Asia (Figure 2.1). The country stretches between latitude 20°30' and 26°45' north and longitude 880 and 92°56’ east. It shares common borders with the Indian states of West Bengal, Meghalaya, Assam, and the Union territory of Tripura in the West, North, and East; in the South-east corner with Mayanmar (former Burma); and I l l l ”'2 WE ”'2 we I’d-o- -2e-u Kur‘vun Dim W IN D IA m it w t— m zsm $1M MI W W ' ' Ban! mm. “fini “mi "*"wfl Wflad Gw- F“ M- n,“ INDIA w Fm Conl- INDIA “v“ 8m: Ml NIH Joann pm . Font w Noam-l m- - O m S“. I N <1. Q A F 93 m Baum P p... ~22'N I p 0% 9 O Luna Bay of Bengal MYANMAR LILLaL-u- un- Figure 2.1. Administrative districts in Bangladesh, 1991. 8 in the South with Bay of Bengal which is over 445 miles. Bangladesh is one of the most crowded rural areas in the world with 106.7 million people within 144,000 square kilometers. The average density is 741 persons per square kilometer (The World Bank, 1992). And this is in a nation where more than 80 percent people are agriculturalists. The man-land ratio is 0.36 acres per person which is very low in comparison to other Less Developed Countries (LDCs). According to the Bangladesh Population Census 1981, about 16 percent of the total population live in 491 urban centers. This is very low in comparison with many of the Asian countries (Table 2.1). The only exception is Nepal where the percentage is 6. Table 2.1. - Percent of urban population in selected Asian countries, 1982. Countries Percentage of urban population Bangladesh 16 Burma 28 India 24 Indonesia 22 Iran 52 Malaysia 30 Nepal 6 Pakistan 29 Philippines 38 Sri Lanka 24 Thailand 17 Source: Bangladesh Population Census, 1981: Report on Urban Area, Table 2, p.12. 9 a do : o ca e a c vo The urbanization pattern in Bangladesh has evolved through several distinct phases which are classified as the Mughal, British, Pakistani, and Bangladeshi periods. Urban development in Bangladesh began with the planned cities in the 3rd century B.C. But very little is known about the impact of these cities on the contemporary structure (CUS, 1976). During medieval times (1300 A.D.), there were several highly populated royal cities, viz., Mahastan, Sonargaon, Vikrampur, and Chatgram (present day Chittagong) which served as centers of administrative and commercial activities and religious festivities. Even though these towns had considerable population, their impact on the overall urbanization of the country was very insignificant. The Mughal and the British periods (1300-1947) constitute the first important stage of urbanization in Bangladesh. During the Mughal period, cottage and craft .industries flourished and several urban centers developed around such industrial concentrations. Throughout this period, Dhaka city with its massive population, dominated the region. During British rule, urban development was rapid and cities attained increasing importance of function as collecting centers for raw materials needed for export. Older towns also became more important as administrative centers of different hierarchies. However, functional diversity of urban centers did not occur because of the absence of industrialization in colonial Bangladesh. The 10 urbanization process was also hampered by the massive industrial-commercial agglomeration around Calcutta, which virtually turned the whole of Bangladesh into its hinterland. Urbanization in Bangladesh received an impetus after 1947 when the region became independent got a separate entity from its surrounding Indian territories. This phase may further be divided into: 1) the Pakistani period (1947- 1971) and 2) the Bangladeshi period (1971-present). After 1947, considerable socio-economic as well as political and administrative changes had taken place giving momentum to the urbanization process. But more important is the liberation of Bangladesh in 1971. This abrupt and dynamic change followed by violent political changeover had its impact on urbanization. The liberation of the country not only accelerated the rate of urban population growth, but also multiplied the nature and problems of urbanization. On the other hand, the Bangladesh region and its urban pattern ceased to be affected by external phenomena giving rise to an opportunity for independent settlement planning (CUS, 1976). a c t n ad The nature of relationships between urban centers and its rural hinterland depends on the hierarchy of the urban centers in the system. The hierarchy of the Bangladeshi urban system is closely related with the administrative 11 hierarchy. Higher order urban centers represent higher order administrative headquarters. The administrative hierarchical order in Bangladesh begins with the State at the top followed by Divisions, Districts, and Sub-districts (Upazillas). Following the classification of Islam and Hossain (1975), the hierarchy of the Bangladeshi urban system is presented in Table 2.2. Table 2.2.- Hierarchy of urban centers, 1981. Order Nature of Center Size of No. of Urban Center Centers Lowest 1st Small Market place (Hat, Bazar) 168 2nd Semi-town Thana HQ. 129 3rd Commercial-Industrial 114 4th Subdivisional town 45 5th District town 23 6th Divisional HQ.(Industrial-Commercial- 12 Educational town) 7th Metropolis (National Capital) 1 Highest The first order central places are basically rural markets. Rural market places are called ‘hats' or bazars according to their functions. The ‘hats' are periodic market places with a few permanent shops and a large number of temporary stalls. A bazar is a permanent market. Bazars do not have standard features like paved streets or brick buildings. But they may have post offices, tubewells, playgrounds, and lowest level of government functions. Each 12 has a population below 5,000. These rural markets serve as collection and exchange centers for farm products and basic consumer goods. The hinterland of the rural market is naturally the rural village which has a radius of influence of about 2 to 3 miles, with a great degree of overlapping in service areas. The larger trading centers, or specialized rural markets may have significantly large hinterlands. With the increasing demand of fertilizer and other farm inputs, the importance of rural markets is also increasing. These centers have become very important to diffuse innovations that emanate from higher order centers such as non-formal education, family planning, health, and political ideologies. The second order central places are known as Thana (police) Headquarters. A Thana Headquarter has a minimum population of 5,000 with an upper limit of 10,000. These urban centers are semi-urban settlements with a few urban characters interwoven with many rural features. As administrative headquarters, these settlement has all the basic Thana level functions with at least one commercial bank, an agricultural bank, a post and telegraph office. Also, a few professional medical practitioners, a number of permanent retail and wholesale establishments, secondary schools and such other facilities can be found. Usually Thanas do not have colleges or permanent entertainment like movie houses. The direct zone of influence of these centers 13 other than administrative functions extend up to about 10 miles. The third order central places are the industrial- commercial centers or subdivisional headquarters. In this category, the range of population is 10,000 to 25,000. This order of urban center enjoys most of the basic urban services as well higher order functions such as colleges, hospitals and movie houses. A few may even have weekly newspapers. There are 114 urban centers in this group. These urban centers have a significant role to play in the diffusion of urban services to the rural hinterland. Their influence extends to about 15 miles in any single direction, and covers approximately 50 miles of territory (Islam and Hossain, 1975). The fourth order central places are the sub-divisional centers having population 25,000-50,000. These towns serve as sub-divisional headquarters and district headquarters. They offer all the services of the previous order centers, and many of them have taken on a higher order of trading, educational, and cultural services. The zone of influence of subdivisional towns covers about 100 square miles. The direct zone of influence in one direction is approximately 20 miles. There are 45 urban centers in this category. The fifth order central places are district towns. They have 50,000-100,000 people in them and there are 23 of these centers. Districts towns have all the functional characteristics of the preceding orders and include new 14 functions with growing industrial and commercial activities. Higher order technical education and higher order administrative services differentiate these from lesser order towns. There is no urban mass public transport such as bus services, but bicycle rickshaws and motor-rickshaws are common public services which provide both intra-urban and urban to rural services. Local buses, trains, motor launches and boats, make the major center-hinterland transport links. The hinterland limits extend to about 40 miles in any one direction and cover about 200 square miles. The sixth order central places are divisional headquarters. These have 100,000 to 1 million people. There are 12 of these centers out of which three are divisional headquarters (Chittagong, Khulna, and Rajshahi), and seven are district headquarters. Khulna and Chittagong, are two port towns which have vast hinterlands spreading over nearly half of Bangladesh. Their direct zone of urban influence cover areas of approximately 3,000 miles. These also include the university towns of Rajshahi and Mymensingh which have comparatively smaller hinterlands about 500 square miles. The seventh order central place is the metropolis of Dhaka, the national capital. Dhaka is the administrative center and national metropolis containing the highest order of services in the nation. The urban centers with a population of less than 50,000 hardly command a significant urban influence beyond a 15 distance of 10 miles. There are 36 urban centers with population of 50,000 or more. These do have a more significant influence especially along transport arteries. The largest urban centers which are 10 in total, command multi-functional direct influence over large rural hinterlands. Thus, only one-third of the total territory of Bangladesh come within the zones of direct urban influence. Two-thirds of the rural area require various direct urban services for development in economic and socio-cultural spheres. Primacy Dhaka has experienced a high growth rate since 1951. In 1901, It had 18.35 percent of the total urban population. In 1981, Dhaka's share of the urban population rose to 25.41 percent. The United Nations (1983) has predicted that Dhaka will experience a dramatic growth in the coming decades. Table 2.3 shows the growth of Dhaka city from 1901 to 1981. 16 Table 2.3. - Growth of Dhaka, 1901-1981. Census Year Total Urban Dhaka's urban % of urban - population 1901 702,035 128,857 18.35 1911 807,024 153,609 19.03 1921 878,480 168,510 19.18 1931 1,073,489 196,111 18.27 1941 1,537,244 295,735 19.24 1951 1,819,773 335,928 18.46 1961 2,640,726 556,712 21.08 1974 6,273,602 1,679,572 26.77 1981 13,535,963 3,440,147 25.41 Data Source: Population Census of Bangladesh 1981. According to 1981 Census, about 52 percent of the urban population live in the urban centers having population of 100,000, Dhaka’s population being 25 percent. Table 2.4. Bangladeshi primacy index*, 1901-1981. Year 2-City Index 1901 1.21 1911 1.23 1921 1.34 1931 1.13 1941 1.32 1951 1.16 1961 1.53 1974 1.89 1981 2.47 *The primacy index is the ratio of the population of the first city to that of the second, third and so on, depending A 2-city index denotes P = P1 / P2, where on the criteria. P: index of primacy; P1, and P , represents population of first and second cities respect1vely. = 2.00 or more. Primacy occurs when P 17 Urban primacy has been growing in Bangladesh. The Table 2.4 shows clearly that from 1901 to 1974, 2-city index exhibited a ‘dual primacy' or ‘binary pattern' in the Bangladeshi urban system. The Two-city indices show that the overall primacy ratio steadily increased between 1901 and 1931 and again between 1951 and 1981. The only decline was in 1931 and 1941 and this must have been due to the increased urban functions as it became an official sea port in 1928. However, the overall 2-city primacy indices show the increases from 1901 to 1981. The United Nations (1983) has predicted that by the year 2000 Dhaka will be the nineteenth largest city in the world with 11.2 million people. Experts predicted further that it will be the fifth most populous city by the year 2025. Islam (1989) puts Dhaka's present population at six millions and the other three cities Chittagong, Khulna, and Rajshahi at 2.5 millions, one million, and 0.5 million respectively. Ahmad (1967) investigated city-size distribution of Pakistan, East Pakistan, and West Pakistan of 1951 and 1961. The purpose was to determine if distribution had become more normal over time. Using rank-size and log-normal probability plots, he found that the primacy ratio for Bangladesh, the then East Pakistan, was 42.9 and 45.9 for 1951 and 1961 respectively. The increase of the primacy ratio was the result of a phenomenal growth in the population of the primate city Dhaka. He found that the 18 distribution of cities in each year was more primate than lognormal; but it is best classified as intermediate. Elahi (1972) studied the Bangladeshi urban system between 1901-1961. The purpose was to asses primacy in Bangladesh. He used rank-size and log-normal probability plots and found that in Bangladesh, primacy was diminishing over time and development. Berry (1961) also studied the city-size distribution patterns of Bangladesh from 1901 to 1961. His study clearly shows that the Bangladeshi urban system is dominated by one primate city. None of these studies tested the slope of the curve statistically. These studies were based on a visual inspection of graphs (Carroll, 1982). Moreover, they show different kinds of city size distributions for Bangladesh for the same periods. While both Elahi and Ahmad (1967) found a dual primacy in Bangladesh, Berry's study showed a clear primacy in the Bangladeshi urban system. It appears that the source and definition of data may have contributed to the inconsistency in their findings. site and situation of Dhaka The growth and form of Dhaka city is restricted by physical conditions, i.e., river system and the height of land in relation to flood level. The southern part of the city is bordered by the Buriganga river. The east and west sides are bounded by the low lying flood plains of the Balu and Turag rivers respectively. A greater percentage of land 19 free from floods is located in the north of the city. Thus, the city is only expanding in one direction resulting in higher distances and transport costs to the city center. Rural-to-Urban Higration Urban growth in Bangladesh has been occurring mainly through rural-to-urban migration. In a study on rural-urban migration, Khan (1983) found that 50 percent of national urban growth during 1961-1974 period was due to rural-to- urban migration. The migration pattern varied among urban centers. He found that due to in-migration population Table 2.5. Component of population growth in six major cities of Bangladesh, 1961-1974. City M’ n W 1961 1974 Natural In-Migra- Annexa- Increase (%) tion (%) tion (%) Dhaka 521,034 1,679,572 18 74 8 Chittagong 364,205 889,760 28 43 29 Khulna 175,023 437,304 27 43 a Narayanganj 162,054 270,680 41 17 42 Mymensingh 53,256 182,153 17 25 58 Rajshahi 56,885 132,909 30 36 34 Source: Khan, 1982. Note: Khulna annexed two urban areas during the period, and the combined population was used as the base population. increased by 74 percent in Dhaka, 43 percent in Chittagong, and 43 percent in Khulna. Natural increases for these three cities were 18 percent, 28 percent, and 27 percent respectively (Table 2.5). On the other hand, the in- 20 migration was less important in the small cities of Narayanganj, Mymensingh, and Rajshahi. These migration patterns are presented in Figure 2.2. During the same period small urban centers having population below 15,000 experienced more than 80 percent increase from rural to urban migration, such as Rangamati (81 percent), Chowmuhani (89 percent), and Chuadanga (81 percent). Barisal, the fifth ranking city in that time lost population through out- migration. In other centers, urban growth mainly occurred by natural increase. 21 Io-llguou on S 0‘ ”OI-74 Pop-lo“. lam-Io Figure 2.2. Net migration for Bangladeshi cities, 1961-1974. Source: Khan 1983. A survey on urban squatters in 1974 shows that the main cause of in-migration to large urban centers is primarily to search for employment. other reasons, listed in order of importance are - amenities of urban living, medical care, educational opportunities, social justice, and improved social status. Also, existing severe disparities between 22 rural and urban sectors in levels of living and wage rates and the perceived potentials for employment in high paying jobs have induced rural out-migration. Khan (1983) also remarked that the reason is that rural economics are unable to absorb the rapidly increasing labor force because of the prevalence of the traditional agricultural system and the absence of significant opportunities in non-agricultural activities. Migrants from rural areas are mainly young adults. This has caused abnormal age pyramids both in rural and urban areas of Bangladesh. Urban areas are marked with concentration of younger people, while rural areas have more older people (Khan, 1982). Because of the relatively youthful urban population, natural increase is also an important component in the future growth of urban population (Khan, 1982). Choguill (1987) noted that migrants often originate in the surrounding areas. In big cities, especially Dhaka, Chittagong, and Khulna, migrants are more likely to come from more distant areas in the country. Dhaka receives migrants from Comilla, Noakhali, Mymensingh, Faridpur, and Bakerganj districts. Chittagong drew people from Noakhali, Comilla, and Dhaka districts. Khulna's migrant population mainly comes from Bakerganj, Jessore, Faridpur, and Noakhali districts. 23 Functional Characteristics Dhaka has a virtual monopoly on main urban functions with higher order secondary, tertiary, and quaternary services located in the city. It is also a center of government, international affairs, high level educational and cultural institutions, transport, and industry. About 70 percent of all the manufacturing industries and 69 percent of all the employment in Bangladesh are concentrated in Dhaka city. Internal Characteristics Development disparity exists not only between principal urban centers and their rural peripheries. Polarization in level of living is also found within large cities. A survey conducted by UNDP revealed that 50 percent of the total urban population live in squatter settlements or bustees. Population in the bustees are overwhelmingly poor. The average total household income is about 832 taka (about 42 US dollars at 1 US$ = 40 taka) or less. Among them, 70 percent are below poverty level and their monthly income is less than 600 taka (about US$15)(Islam and Khan 1988). Most of the squatter settlements in Dhaka city are overcrowded with densities over 2,000 persons per acre (Islam, 1988). Dwellings are constructed with bamboo, sacks, mat, plastics, etc. These huts, erected with disposable materials, disfigure the cities and create health and fire hazards. Shanties cause environmental deterioration and unhygienic 24 living conditions. They also adversely affect the moral and social health of the inhabitants (Chowdhury and Asad-Uz- Zaman, 1976). The infant mortality rate in slums is 150 per thousand live births. It is higher than the national average of 116 per thousand (Robson, 1991). A slum house occupies about 10 to 15 square feet with 10 or 15 people sleep in one room. There are 1125 squatter settlements of this type in Dhaka city. In contrast, one upper class residential house occupies close to one acre. o c ° t a v Developing countries are plagued with ‘large city' problems. The ’large city' or primacy problem can be analyzed in terms of a core-periphery relationship described by Friedmann (1966). Friedmann’s core-periphery model is essentially a colonial model. In the core periphery relationship, a polarized spatial structure is inevitable with a strong center or core which dominates production and consumption and lives off peripheral hinterlands. Figure 2.4 illustrates this model. Wubneh (1982) using Friedmann's 25 model in the'case of Ethiopia explains it as follows: ".... the territory of a nation-state has a spatial structure, as depicted in Figure 2.3A, with a national metropolitan center linked to small subcenters by some pattern of relationships that determines the sphere of influence as well as the flow of goods, resources, people, and so on. The national metropole or core has a stronger base, economic and political, because of initial advantages which may be historical, geographic, or resource endowment, for example. The process of ”cumulative causation” reinforces the comparative advantage or the established dominating linkages that the core enjoys over the periphery. With its strong economic and political power, the core organizes the periphery for its own purposes to supply raw materials, to be a market for its products, and so on. Thus, in the model, the subcenters are used as hinterlands to the national metropole and the flow of resources tends to be asymmetrical in favor of the latter. On the other hand, the national metropole's relationship with the colonial metropole (Figure 2.3B) makes the former the hinterland of the colonial capital. Thus, the national metropole is the intermediary - the exploiter and exploited - in the international setting" (pp. 5-6). The core-periphery model explained by Wubneh can be applied in the Bangladeshi situation. But, British colonial rule has complicated core-periphery links by assessing a strong international dimension to the Bangladeshi situation than was the case in Wubneh's Ethiopia. In the Bangladeshi case, two types of spatial structures can be found during the British colonial period: (1) regional, and (2) national. The British colonial spatial organization kept Dhaka as a regional capital while Calcutta’s role was raised to a national center. Dhaka served as a collecting center . of raw materials for shipment to Calcutta. Dhaka remained in the shadow of Calcutta which was used as a colonial 26 metropole, in which major colonial processing including manufacturing took place. Thus, there was no industrial development in Bangladesh. This structural effect of British colonial rule in India and Bangladesh is illustrated in Figure 2.3. I 'O". --T- ~ 5 ‘ s “ I ' I 0900 000‘ 0003 A NATIONAL SPATIAL 8 GOLD I STRUCTURE N AL UN" — National Boundary —. Experts from Hinterland to Care ---D Imports by Hinterland O Colonial Metropole a National Capital 9 Regional Center (Awicultulturd and/or Administrative Center) 0 Hinterland Figure 2.3. The core-periphery model. Source: wubneh, 1982. Figure 2.3 shows that development in Bangladesh lagged well behind that off India. One manifestation of disparity in economic development is the construction of a railroad. 27 AIIAO OLIVSO lam. IDS, It , Q was - .u'. II76- £085 tllc-ICQS Mm I095 Figure 2.4. Railway services in the Indian sub-continent. Source: Berry, 1971. Ahmad (1976) noted that construction of a railroad started in the Bangladesh region in 1852, 10 years after starting in India. But the progress was slow until 1862. This railway system was constructed only to integrate the metropolis and its hinterlands to collect raw materials. Fisher (1989) commented that the railway system constructed by the British in India was one manifestation of the exploitation of the economy of the country. He stated that "British rule influenced the spatial layout of India’s production and circulation systems. Several focal cities grew rapidly, and a railway system was constructed to integrate them with each other and with their regional hinterlands, which could then be developed and exploited 28 easily." Rizvi (1969) also mentioned that the opening of the East Bengal Railway Company was "more political than commercial". Structuring control of the spatial layout is also evident in other countries which were also under the British rule. For example, Dickensen and Hodgekiss (1983) reported that the railroad systems constructed in East Africa were designed for the export of minerals. These rail systems expanded from port cities and did not link cities laterally generating unbalanced urbanization. This observation agrees with Berry (1971) as he observed that the Bangladesh region was the hinterland of Calcutta during the British period. Under Pakistan, Dhaka became the administrative capital of Bangladesh and this created some dynamics for the city. Subsequently, large scale investments in the physical and social infrastructures accompanied by industrial development gave Dhaka a boost in its primacy. The Second Five Year Plan (1965-1970) invested nearly 43 percent of the state's total plan funds in the development of industry, social, and physical infrastructures in Dhaka, Chittagong and Khulna. This rose to 46 percent in the Third Plan. Similarly, a 29 \ . . D O U 20 3040 90 ,;\,¢. bhbhehl j \Js \. ‘ mu at me ( ‘* I '| \,,. .\\ . . 7 , uterine on oevuome up} , ) ////. uernopouuu cams \v."~0’.~.‘.‘.fl-" 5% it '7' cmrneodc <;.a-""\./"' h OAY OF BENGAL I Figure 2.5. Polarization of development during the Pakistani period. Source: Sibli, 1980. major proportion of the private investments in industry, housing, and transport were invested in these three cities. Thus, nearly 60 percent of the total fixed assets in large scale industries are located in these three cities (Figure 2.5). The urban-biased investment policies resulted in an increased concentration of economic activity and fostered further concentrations of population. Such conditions favorable to concentration breeds further growth in the urban sector (Wellisz, 1971). 30 e t s c t P ob In the light of the growing problem of high concentration of population in large cities especially in Dhaka, a strategy has been implemented by the government of Bangladesh to alleviate the current situation. The main aspect of this strategy is to decentralize development by pushing growth in new urban centers. Besides, a number of policies are suggested by the World Bank (1979), Richardson (1987), and regional planners. The main aspects of these suggestions are: (1) development of the existing secondary and tertiary urban centers that show potential for expansion, (2) increase of the per capita urban absorption costs. In order to slow down the growth of large cities and distribute development benefits to the rural areas, the Bangladeshi government has implemented a decentralization policy in 1982 using the Second Five Year Plan (1980-1985). The aim was to create a harmonious urban structure that would serve local communities in raising their standard of living. The government’s goal also included the devolution of socio-economic and development activities throughout the country from center to periphery. But in practice, it was only further administrative control rather than decentralization to share the development resources of the country. Even if the government meant to decentralize development, resource constraints were a major impediment to effect real decentralization. Shakur (1987) noted that it 31 is unrealistic that decentralization of industries to other areas will necessarily prevent the growth of the country's central region, particularly its biggest metropolis, Dhaka. It may still be assumed, however, that if "decentralization" occurs in a significant manner, it might slow down the rapid growth of a few urban centers and thus reduce the attractiveness of major cities and towns from potential migrants. Demographers and planners suggest that what needs to be done is to develop local natural resources and increase industrial activity in peripheral areas. This will keep in urbanizing the hinterland and thereby decentralizing development. The main problem encountered in implementing the Second Five-Year Plan was resource constraints (The Third Five Year Plan, 1980-1985). The literature in urban development reveals that initiative of decentralization to solve problems associated with the creation of new towns, shows that the venture has not fulfilled the intended purpose. Chief reasons for the failure are cost of construction, difficulties in acquiring sufficient land, and the complex mix of social, economic, and political factors that govern where people and business can locate. Reviewing the success and failure of new towns, Williams et a1. (1983) found that the majority of new towns, whether in the West, in the Soviet Union, or in LDCs, tend to fall quite short of intended goals. It is indicated that it is extremely difficult to achieve the goals of new towns without having total control over land use, construction, 32 and the lives of citizenry as is often the case in communist states. Realizing this problem of creating new urban centers, the 1979 World Bank Development Report suggested that in order to counter-balance the growth of the primate cities, secondary cities that show potential should be developed. On the other hand, Richardson (1987) remarked that large city problems in Bangladesh are mainly due to the underinvestment in urban infrastructural development. He believes that Bangladesh needs_to invest money in urban development not only to solve urban problems but also to eliminate the possibilities of withdrawal of foreign aid. Richardson (1987) commented that as the rural poverty is the most important problem in Bangladesh, it is hard to conceive that the government will invest more in the large urban centers. It would be difficult to raise such resources on behalf of urban development. QEEEAEI Based on the review and critique of the literature dealing with development pattern of Bangladesh's urban centers, it is strongly suggested that the colonial rulers, British and Pakistan, always favored large city/cities to invest resources and control the economic condition of the country. Chapter III presents the methodology to capture the impact of different policies of urban system and to 33 identify the present trend of city-size distribution in Bangladesh. CHAPTER III THEORIES AND NETHODOLOGY APPLIED IN THE RESEARCH Introduction In studying the developmental patterns of Bangladesh's urban centers, three approaches have been used: (a) the historical development of rank-size relationships of urban centers, (b) the structural change the of urban system, and (c) the growth pattern of selected urban centers. Three methods are used to study the historical structure of the Bangladeshi urban system: (a) rank-size plots, (b) rank- size regression analysis, and (c) rank-mobility index. T o and e odo o o s The Rank-size Model The rank-size regression model developed by Zipf (1949) is used for observing the urban data in Bangladesh. In his work zipf used communities that contained at least 2,500 inhabitants in the 1930 United States Censuses. He used the rank-size relationship to examine a wide variety of issues. One of these is to examine the rank-size function which is most widely used to investigate whether the relationship between city size and rank conforms to an expected pattern in accordance with the rank-size rule. According to this rule, when urban populations in an area are ranked in a decending order of population size, the population of the second city is half the population of the largest city, the 34 35 third city a third of the population of the largest city, and so on (zipf, 1949). Symbolically, the rank-size rule is expressed as follows: Pr = Pl/r 00(1) where, r = the rank of a city; Pr = the population of the city of rank r; and P1 = the population of the largest city (rank = 1). When plotted on double Iogarithmic graph paper, this relationship produces a straight, downward-sloping line with gradient of -45°. The hypothetical rank-size distribution describes an urban system containing a few large metropolis, a larger number of medium-size cities, and a still larger number of smaller towns. This basic formula is often modified by a constant (b) to allow variations from the strict rank-size rule. This produces the following modification: Pr=P1 . r'b ..(2) where b is the slope of the line joining various sizes of cities arranged in rank order on double logarithmic paper. zipf designated the distribution as "rank-size" if the exponent b equaled -1 (Figure 3.1). He characterized the rank-size distribution as an ideal city-size distribution. 36 Hmmmbnduflbounb) Rank (log scale) Figure 3.1. The rank-size rule. The linear relationship (equation-2) described by the typical rank-size graph can be expressed in the following equation: log P} = log a - b log r ..(3) the population of the city of rank r the rank of the city the intercept; also, the logarithm of the largest city the slope coefficient where P O' OIHH IIII This rank-size relationship expressed in the equation (3) is used to confirm the visual observations made on rank-size plots. For this purpose, the parameter b is 37 estimated from the empirical data by standard least square linear regression. The linear regression equation is as follows: logPr=a+blogr ..(4) Where, P r dependent variable (log population of the r ranking city) r = independent variable (log rank) a = intercept; also, the logarithm of the estimated population of the largest city b = the slope of the straight line The dependent and independent variables, population of urban centers and rank of cities were ranked on the basis of the population size in a decending order and were converted into logarithms for use in the regression equation. The computation was done on a computer, using SYSTAT statistical package, version 5.01. In practice, because the rank-size line necessarily slopes downward to the right, the value of the slope parameter, b is invariably negative. The b-coefficient evaluates the rate of change in population size associated with the rate of change in rank. The b-coefficient can be used to define rank-size distributions between two extreme limits of rank-size relations: perfect evenness and perfect concentration (Fan, 1988). In a situation of perfect evenness, all cities are of the same size and the b-coefficient approaches zero (Figure 3.2a). On the other hand, the b-coefficient can Umpquwhn 38 approach negative infinity if all the urban population is concentrated in one city (Figure 3.2b). elope approaches zero "‘€? elope approaches negative infinity /' r/ 7‘ Unpqwhmm unnu. unnut (a) (b) Figures 3.2. (a) Tendency toward perfect evenness. (b) Tendency toward perfect concentration. MDdified from Fan 1988. Based on this reasoning, Danta (1985) suggested that the b-coefficients associated with rank-size distribution (b - -1) provides a useful benchmark for estimating the distributional characteristics of urban population within a relatively closed urban system. When it is more negative than -1, it indicates a larger percentage of the population living in higher ranking cities than would be expected by the rank-size rule. On the other hand, D values more positive than -1 indicate that lower ranking centers are relatively more populated than would be expected by the urban rank size rule. It means that the parameter b of the 39 rank-size function is indicative of the relative distribution of population contained within an urban system. The Rank-size Plot This method involves plotting of city populations against the related ranks on a double log arithmetic paper and is used for comparison of growth patterns of the urban system within a country. Most commonly, such plots reveal four types of city-size distributional patterns (Sheppard, 1982; Haggett, 1983): 1. Rank-size distribution: In rank-size distribution pattern, the curve is downward straight line on log-log paper with -45°. According to rank-size rule, the population of any city of rank n is equal to the population of the largest city in the system divided by n. BILEEEE distribution: This pattern exists where the top ranking city or two cities dominate the distribution. When two cities dominate with equal magnitude, the pattern is known as ‘dual primacy'. In the plot, steeper curves represent ‘primacy’ and flatter curves suggest that the growth of the primate city is less than that of other cities. Conger distribnrions: This pattern exists where there are a number of equally large cities that dominate the urban system. Mineralization: The s-shaped distribution pattern may be seen where primacy and convex patterns co-exist. 40 These patterns are clarified in Figure 3.3 I.” Pam 500.. L a — “up... 4% a Pattern . 06° .9 35 ‘W:: ‘5 It ”' w— }, -HhmeHMMI w— m— 20— 10 1 11111111 1111 1 2 3 45 10 100 RIMMIOOIGIIO) Figure 3.3. Patterns of city size distributions. Rank-Mobility Index The study of an entire urban system does not tell anything about the behavior of growth of individual cities. In order to understand the growth pattern of individual cities, the rank-mobility index developed by Marshall (1989) 41 is used. The rank-mobility index is given by the equation stated below: Where, H s rank mobility R0 8 rank at the beginning of a given time period R1 = rank at the end of a given time period In other words, the index is the difference between the two rank positions, divided by their sum. The result is positive if the city’s rank rises and negative if it falls. The theoretical limits of the index are -1.0 and +1.0, with a value of zero signifying no change in rank. It is considered that a city is doing well if its mobility index for a particular period is +0.300 or higher, and doing poorly if its index is -0.300 or lower (Marshall, 1989). The computations were done using SYSTAT, version 5.01. All the cities are classified into four groups in terms of their rank mobility as indicated below: £129.92 e v - 1. No net change Zero (.000) 2. Minor change Between -.300 and +.300 3. High degree of gain Above +.300 4. High degree of decline Between -.300 and -.500 5. Very high degree of decline Below -.500 42 £23££2_21_DA§A Urban population data for the decades 1901 to 1981 are used in this study. The cities included in this study are those which according to the Bangladeshi government have fulfilled the criteria stated below and have populations above 5,000. (a) A Municipality, a Town Committee, or a Cantonment.Board. (b) An urban area which has a population of at least 5,000 persons living in a continuous built-up area. (c) Centers which have public utilities such as, roads, supply of electricity with street lights, water supply, sanitary arrangements, etc. (d) Centers having trade and commerce facilities and non-agricultural population. Bangladesh Censuses are conducted once in 10 years. As such, data for nine decades, from 1901 to 1981 are available and are used for analysis. The 1971 Census was interrupted by the liberation war and was later conducted in 1974. The data sources for this study are: (1) Census data of Pakistan 1961 for urban population 1901-1961, (2) Population Census of 1974, and (3) Population Census of Bangladesh, special report on Urban Areas, 1987. CHAPTER IV RESEARCH RESULTS Introduction The purpose of this study is to examine the structural effects of colonization on city-size distribution and how the policies of independent Bangladesh impacted subsequent patterns and trends of urbanization in the country. In relation to that purpose, the following questions were raised: (a) what changes have taken place in the distributional patterns of urban places over the political history of Bangladesh? (b) has the urban system moved toward deconcentration (rank-size) or concentration (primacy) over time? (c) what is the spatial trend in the current urban system of Bangladesh? The results are presented in three sections. First, rank-size relationship for the years 1901 to 1981 are presented along with rank-size plots. The plots are further examined by a second section which describes the results of regression for drawing inferences on the nature of the rank- size relation. The third section presents the rank-mobility indices and attempts to draw some inferences using the analysis. 43 44 T a -8 2e ot The rank-size plot is the first method used in the study to examine the development pattern of Bangladesh’s urban system. The rank-size relationships were plotted keeping the rank of city on x-axis and the population on y- axis. Figure 4.1 shows the rank-size plots which exhibit a process of evolution of urban centers from 1901 to 1981. The plots represent rank-size distributions for nine decades (1901-1981). It is clear from the plot (Figure 4.1) that rank-size relationships are not linear on a log-log scale. The observations made from Figure 4.1 are noted under five sections: (1) dual primacy, (2) discontinuity in the growth of Dhaka, (3) relative absence of mid-size cities, (4) rank-interchange, and (5) increase in the number of the urban centers. Population Size 45 10,000,000 ' 3.000.000‘ 1.000.000 500.000 ‘ 100.000 80.000 60.000 50.000 40.000 30.000 20.000 10.000 l l l l l I 1 2 3 4 5 6 78 10 20 30 4O 60 80100 200 300400 1.000 Rank Figure 4.1. Rank-size curves, 1901-1981. 46 1. Dual Primacy From the Figure 4.1 it is observed that dual primacy existed during the decades 1901-1951 as indicated by flat upper sections of rank-size curves. These two cities were Dhaka and Chittagong. Strong duality can be seen in 1911, 1931, and 1941. This dual primacy diminished after 1951 and the primacy of Dhaka city had increased (Figure 4.1). These variations in the degree of dual primacy can be observed from the variations in the rates of growth of Dhaka and Chittagong as presented in Table 4.1. Table 4.1.- Growth rates of Dhaka and Chittagong, 1901-1981. Year Dhaka thIIAQQEQ WW 1901 128857 - 106848 - 1911 153609 19.20 125226 17.2 1921 168510 9.70 125968 15.27 1931 196111 16.38 173577 37.79 1941 295735 50.80 224732 29.47 1951 335928 13.59 289981 29.03 1961 555712 65.42 364205 25.60 1974 1679572 202.24 889760 144.3 1981 3440147 104.82 1390684 56.3 The variation of the growth rate of these two cities are graphically represented in Figure 4.2 for a visual examination. 47 0 I I 1 l t I I l TNT-11 1911-21 1821-31 193141194151 195161 1”1-74 1874HS1 ImwummPuhh Figure 4.2. Growth of Dhaka and Chittagong, 1901-1981. Historically, Dhaka and Chittagong have always been the main cities of Bangladesh. Dhaka due to its advantageous central geographical location in the country and also due to its proximity to natural resources such as the tea gardens. In addition, the fine cotton fabric industry located in the city has also made it prominent at all times. Chittagong, though has always been a port city even before 1901, was officially declared a port city only in 1930. Chittagong has long been the only sea-port in the country and its functions were concomitant to a process of greater internal interdependence of cities, and its economic importance was dependent on the export of jute and other raw materials and some important functions. The British also had developed Chittagong in order to facilitate shipping of jute, rice, 48 and tea to foreign countries. As such, from 1930 Chittagong caused an increase in the duality in Bangladesh. The importance of Chittagong diminished in 1950, boosting Dhaka's primacy. This conforms with Elahi’s study that Chittagong lost its functional monopoly after the establishment of another sea-port near Khulna and expansion of river-ports such as Narayanganj and also Chandpur. In 1981, Dhaka had a relatively higher rate of growth compared to Chittagong further diminishing that dual primacy and the tendency toward primate city dominance. The existence of dual primacy is common in South Asian countries; for example, in India, Calcutta and Bombay, in Pakistan, Karachi and Lahore (Dutt, 1983). Harris (1970) noted that the existence of two or several large cities is suggestive of lack of full spatial integration. 2. Diacontinuity in the Growth of Dhaka The first ranked city which is Dhaka, the capital city of Bangladesh, shows a continuous growth from 1901 to 1981. It can be noted from the plot (Figure 4.1) that the growth of population of Dhaka was very slow between 1911 and 1921 and very rapid between 1961 and 1974. The population growth rates for Dhaka are listed in the Table 4.1 and compared with the observations made from the Figure 4.2. 49 Table 4.2. - Growth of population of Dhaka, 1901-1981. Year Population % Increase 1901 128857 - 1911 153609 19.20 1921 168510 9.70 1931 196111 16.38 1941 295735 50.80 1951 335928 13.59 1961 555712 65.42 1974 1679572 202.24 1981 3440147 104.82 The growth of Dhaka which is indicated by the increment on y-axis shows slow growth during the British period (1901- 1951), rapid growth during the Pakistani era (1951-1974), and very rapid growth in the early period of independent Bangladesh (1974-1981). It is clear that in the earlier decades from 1901 to 1941, the urban population growth was slow, especially between 1911 and 1921. It is also seen from Table 4.1 that between 1911 and 1921, the increase in population of Dhaka was 9.7 percent compared to the previous decade's 19.2 percent (Table 4.2). This is also shown by the very small increment on the y-axis for the 1921 curve (Figure 4.1). In 1912, the East Bengal State (the present Bangladesh) was merged with the West Bengal State which resulted in Dhaka losing its prominence. After the establishment of Dhaka university in 1921 and development of infrastructures by the government, growth incentive came back for Dhaka and it continued to grow until 50 1974 when it reached the peak increase of 203 percent. This rapid growth of Dhaka fell to 105 percent in 1981. This trend may be explained by the observation of Mamun (1982) that in 1971, after the liberation war the law and order in rural areas and small towns had deteriorated causing migration of rural population to large cities which were considered safer. A large number of people who had collaborated with the Pakistanis fled to the central or regional cities. Also, because of the rapid shift in socioeconomic conditions after the violent political changeover, many individuals chose to start anew in an urban setting; members of a rural or small-town elite were conspicuous in this group. The rapid growth which occurred in 1971 as an immediate effect of the liberation war decreased in 1981 when migration slowed. These may be the main causes for the slow-down in Dhaka’s growth. A slow down in the growth of Dhaka had also occurred in 1951 which may be attributed to the fact that a large suburban section of Dhaka city was separated and given the status of a city under the name Narayanganj. 3. Relative Absence of Hid-Size Cities The rank-size plot (Figure 4.1) also exhibits clearly the lack of medium size cities in the urban system in the beginning of the century as indicated by the steep slope or sharp decline in the rank-size curves. In 1921, urban 51 centers having population of 25,000-49,999 and from 1941, centers with population of 50,000-99,999 started to grow up. The absence of mid-size cities prior to 1941 shows an uneven urbanization process during the British period. As Fisher (1989) commented, lack of spatial integration reflected by the absence of mid-size cities may be the result of the spatial layout of India's production and circulation system made by the British. This urban system during the early decades of this century may have contributed to the subsequent patterns of urban development dominated by Dhaka. 4. Rank-Interchange From Figure 4.1 it is seen that the curves of smaller urban centers for 1901 to 1931 overlap suggesting rearrangement in the ranking of the towns. Between 1961 and 1981, small urban centers exhibited a balanced and higher growth which is reflected by the parallel and widely spaced curves for small urban centers. One interesting feature for the rank-size lines of 1974 and 1981 is the crossing of 4th and 5th ranking cities, Narayanganj and Rajshahi. It is also noted from Figure 4.1 that the curves of 1971 and 1981 meet at ranks four and five. The cities which are represented by these meeting points are Narayanganj, Rajshahi, and Mymensingh. This situation may be explained by the fact that Narayanganj which ranked four in 1974 was merged with Greater Metropolitan Dhaka in 1981 leaving its 52 rank to Rajshahi. The city Mymensingh which ranked five in 1974 also ranked five in 1981 with very little population growth. 5. Increase in the number of Urban Centers The horizontal expansion in the rank-size plot indicates an increase in the number of centers (Figure 4.1). This horizontal increase is more prominent for the census years 1974 to 1981 suggesting a growth of a large number of urban centers during this period. In the year 1974, there were only 104 urban centers having a population of 5,000 and above. In 1981, this number increased to 322. This increase in the number of urban centers is the result of the Bangladeshi governments national decentralization policy. From the above discussion it may be concluded that the rank-size relationships among urban centers in Bangladesh are not linear on a log-log scale. In the early decades, the urban population growth was rather slow specially during 1911-1921 which was a discontinuity in the urban growth. But, the growth was rapid from 1951 which may be attributed to the development of Dhaka's infrastructures. In the Bangladeshi urban system, dual primacy had existed during 1901-1951 and thereafter diminished. The rapid growth of Dhaka and Chittagong becoming relatively less important due to the emergence of Khulna as a port city after 1951 may explain the decline in the duality. During the British period the urban system clearly lacked in medium size cities 53 which may have contributed to the subsequent pattern of urban system dominated by Dhaka. It is seen that the rank-size relationships have not been stable through the decades which may be attributed to the political changes this region has experienced. The bottom-up development policy of the Bangladeshi government has clearly resulted in a substantial increase in the number of the urban centers contributing to an integrated spatial system. ass as - t The rank-size relationships as observed in the previous section highlight the effects of change in the relative distribution of the urban system. Wubneh (1982) suggested that in order to understand and appreciate the magnitude of shifts in ranks, the relative change in rank-size relationships over time needs to be ascertained. This can be achieved by examining the change in the parameters between rank and size of cities in a regression framework (Wubneh, 1982). This approach is strongly suggested by Carroll (1982) also. As such, the rank-size plots are further examined by a regression analysis for drawing inferences on the nature of the rank-size relation. A regression analysis was performed using the equation (4): log Pr = a + b log r for the nine census years, 1901 to 1981. The results of this analysis are presented in Table 4.3. 54 Table 4.3. - Results of regression. Year a Change b Change R2 Sign. (intercept) in a (Slope) in b Level 1901 11.539 - -0.783 - 0.899 0.000 1911 11.689 0.150 -0.792 0.009 0.897 0.000 1921 11.795 0.106 -0.807 0.015 0.910 0.000 1931 12.059 0.264 -0.861 0.054 0.934 0.000 1941 12.616 0.557 -0.949 0.088 0.954 0.000 1951 12.879 0.263 -0.999 0.050 0.952 0.000 1961 13.263 0.384 -1.031 0.032 0.971 0.000 1974 13.978 0.715 -1.044 0.013 0.974 0.000 1981 14.307 0.329 -0.968 0.076 0.985 0.000 Table 4.3 displays estimates and shows significant changes in the temporal progression of the coefficients both for intercept (a - size of the largest city) and the slope (b - relative change in rank). The patterns manifested in this temporal progression of coefficients are identified and discussed below under four sections: (1) overall pattern, (2) early pattern, (3) mid-period pattern, and (4) late pattern. Overall Pattern The regression results in Table 4.3 show that the negativeness of b-coefficients has increased over time except for 1981. This indicates a tendency toward concentration until 1974 and thereafter deconcentration till 1981. The steady relative rapid increase in the size intercept (Table 4.3) indicates a continuous growth of 55 primacy in the Bangladeshi urban system. As seen in the rank-size plot (Figure 4.1), Dhaka has always maintained its first place in the urban system. A graph is plotted between b-coefficients and the corresponding years for a closer look at the variations in b-values over time (Figure 4.3). 417 Math-HIM / I l I I l 1 l I I 1901 1911 1921 1831 1941 1951 1961 1974 1981 Caunvuu Figure 4.3. Slope coefficients (b) for the rank-size relationships. The rank-size relationship curve in Figure 4.3, progressively approaches the rank-size position of b = -1.00. Then from 1951 to 1974 the curve continues to fall below -1.00 approaching a primate pattern. But, for 1974 to 1981 the curve moves upwards crossing the -1.00 value. The 56 variations in the slopes of the urban system is plotted in Figure 4.4. Log Popustion o_o 0.5 1.0 1.5 2.0 2.5 10 unfiuk Figure 4.4. Variations of slopes for urban centers, 1901- 1981. From Figure 4.4 it is seen that the slopes have steadily increased from 1901 to 1974. The increase is more prominent from 1901 to 1951. From 1974 to 1981, the slope has declined which can be seen from the dispersion of the slope on the x-axis. From the above observations it is evident that the development of the urban system in Bangladesh has not been 57 diachronically monotonic. It is marked with variations. There existed a favorable city-size distribution in 1951 (b = -1.00) and then a tendency towards concentration with primacy until 1974 and a reversal of that trend after 1974. Early Pattern The period 1901 to 1951 was associated with the British rule over this region. It is noted from Figure 4.2 that the absolute values of the b-coefficients which are quite below 1.00 for 1901 and 1911 have progressively approached 1.00 in 1951. It indicates the existence of several large urban centers in the system until 1951. This is also clear from Figure 4.1. Dhaka and Chittagong were dominating the urban system during this period of time, which is noted as dual primacy. It is further noted from Table 4.2 that from 1901 to 1941, the decennial increment in the intercepts is very small indicating a small rate of increase in the population of the largest city Dhaka. Mid-Period Pattern The results of regression analysis have shown that for the period 1951 to 1974, the years associated with the Pakistani regime, the b-coefficient has become more negative. This indicates a tendency of primate pattern in the city-size distribution. But for 1951, the early part of the Pakistani rule, the b-value is .999. This shows a close 58 rank-size distribution pattern in 1951. But the absolute b- value increased to -1.031 suggesting a drift towards a more primate pattern. This according to the model indicates concentration of population in one large city. From Table 4.2 it is also noted that there is a large increase in the intercept for 1951 to 1974 which indicates primacy condition in the urban system. This may be seen as a result of the Pakistani government investing heavily for development of industries including Chittagong and Khulna. Late Pattern The Bangladeshi period started in the year 1971 after the Pakistani regime. From Table 4.2 and Figure 4.3 it is noted that the b-coefficient is decreased between 1971 and 1981. It indicates a deconcentration trend in the urban system. Deconcentration occurred after 1971 as a result of creation of a large number of regional growth centers under the decentralization policy of the Bangladeshi government. This observation is in agreement with the observation made from Figure 4.1 which shows a large horizontal expansion of the curve for 1981. It is also seen that the intercept has increased from 1971 to 1981. This indicates an increase in the population of the primate city Dhaka and at the same time growth of smaller centers in the urban system of Bangladesh. The study has indicated that during this period the relative growth of population in the second city Chittagong 59 decreased compared to Dhaka. This is reflected by the relatively steeper slope of 1971 at the upper portion of the rank-size line (Figure 4.1). This trend continued till 1981. The R2 values for 1974 and 1981 (.974 and .985 respectively) are higher than the corresponding values for the preceding years suggesting a higher linearity in the rank-size relationship. This also indicates higher spatial integration (Harris, 1970). Tha_Bank:§isa.hebilitz_ladas The relative performance of growth of individual cities have been examined by looking at changes in rank. Table 4.4 presents an analysis of the changes in rank experienced by Bangladeshi towns during British, Pakistani, and Bangladeshi regimes in terms of rank-mobility indices. The Table 4.4 lists the cities, ranks, and their indices from 1901 to 1981 for the nine census years. All towns that had 5,000 or more inhabitants in 1901 are included and the table lists these thirty-eight places in rank order according to their sizes in 1901. The ranking of cities for decades 1901-1961 are based on the data from the Census Commission of the Pakistan government. The ranks of 1971 and 1981 are according to the data published by the Bangladesh government. The ranks of these towns are recorded at intervals of ten years. As discussed before, Marshall's (1989) rank-mobility index is positive if the city’s rank rises and negative if 60 it falls (see the methods section for details). To understand the dynamics of rank-order change, the ranks of these cities over the period 1901 to 1981 are plotted in Figure 4.5. In the Figure 4.5, the black lines indicate increase in rank and gray lines show decrease in rank. From Figure 4.5, it may be seen that the rank-order lines for smaller cities cries-cross much more than for larger cities suggesting a high degree of instability in the lower ranks of central places. 61 Table 4.4. Rank mobility of Bangladesh's urban centers, 1901-1981. City Rant: liability 01 11 21 31 41 51 61 74 61 01-11 11-21 21-31 31-41 41-51 51-61 61-74 74-61 01-61 Dhaka 1 1 1 1 1 1 1 1 1 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 Chittagong 2 2 2 2 2 2 2 2 2 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 Serajganj 3 3 5 4 6 10 10 11 13 0.000 -0.250 0.111 -0.200 -0.250 0.000 -0.001 -0.001 -0.625 Eli-nan: 4 4 6 7 7 6 7 6 4 0.000 -0.333 0.067 0.000 -0.067 0.001 0.001 0.001 0.000 Brahmanbaria 5 6 9 9 9 9 12 14 17 -0.231 -0.059 0.000 0.000 0.000 -0.002 -0.001 -0.002 -0.545 Camilla 6 6 4 5 5 5 6 9 6 0.000 0.200 -0.111 0.000 0.000 -0.002 -0.001 0.002 0.000 Barisal 7 7 3 3 3 3 5 7 7 0.000 0.400 0.000 0.000 0.000 -0.001 -0.001 0.000 0.000 Pabna 6 11 12 11 11 14 13 15 12 -0.156 -0.043 0.043 0.000 -0.120 0.001 -0.001 0.002 -0.200 Jamalpur 9 9 10 10 13 17 15 17 15 0.000 -0.053 0.000 -0.130 -0.133 0.001 -0.001 0.001 -0.250 Hadaripur 10 12 6 6 16 24 24 36 30 -0.091 0.333 -0.143 —0.333 -0.200 0.000 -0.006 0.003 -0.500 Nauabqanj 11 5 23 20 19 21 21 23 16 0.375 -0.643 0.070 0.026 -0.050 0.000 -0.001 0.003 -0.165 Tanqaii 12 15 20 19 20 22 26 19 19 -0.111 «0.143 0.026 -0.026 -0.046 -0.003 0.004 0.000 -0.226 Kiehoreqanj 13 13 11 22 22 26 27 30 34 0.000 0.063 -0.333 0.000 -0.063 -0.001 -0.001 -0.002 -0.447 Ranqpur 14 14 13 13 10 16 14 12 9 0.000 0.037 0.000 0.130 -0.231 0.001 0.001 0.002 0.217 Mymensingh 15 10 7 6 4 6 9 5 5 0.200 0.176 0.077 0.200 -0.200 -0.002 0.002 0.000 0.500 Pirojpur 16 23 21 29 30 29 37 50 75 -0.179 0.045 -0.160 -0.017 0.017 -0.004 -0.006 -0.012 -0.646 Syihet 17 16 16 12 15 12 16 16 6 -0.029 0.059 0.143 -0.111 0.111 ~0.002 -0.001 0.005 0.360 Dinajpur 16 16 14 15 14 11 17 16 14 0.059 0.067 -0.034 0.034 0.120 -0.003 0.001 0.001 0.125 Sharpe: 19 17 15 14 16 25 26 31 37 0.056 0.063 0.034 -0.125 -0.163 -0.001 -0.003 -0.003 -0.321 Faridpur 20 20 19 21 17 16 22 22 27 0.000 0.026 -0.050 0.105 -0.029 -0.002 0.000 -0.002 -0.149 Netrokona 21 19 26 26 26 32 34 43 50 0.050 -0.191 0.000 0.000 -0.067 -0.001 -0.004 -0.003 -0.406 Khulna 22 21 17 16 12 7 4 3 3 0.023 0.105 0.030 0.143 0.263 0.002 0.001 0.000 0.760 Iajitpur 23 25 25 25 27 31 42 74 121 -0.042 0.000 0.000 -0.036 -0.069 -0.006 -0.016 -0.023 -0.661 Chandpur 24 22 16 17 6 15 16 20 16 0.043 0.100 0.029 0.360 -0.304 -0.002 —0.001 0.001 0.143 Notchandpur 25 30 32 43 49 51 53 66 102 -0.091 -0.032 -0.147 -0.065 -0.020 -0.001 -0.007 -0.017 -0.606 Netore 26 29 29 35 36 37 39 53 63 —0.055 0.000 -0.094 -0.014 -0.014 -0.001 -0.007 -0.005 -0.416 Satkhira 27 24 26 27 26 26 31 25 35 0.059 -0.040 .0.019 0.019 -0.037 -0.002 0.003 -0.005 -0.129 Jesaore 26 27 27 26 25 20 11 10 10 0.016 0.000 0.019 0.020 0.111 0.005 0.001 0.000 0.474 seats 29 26 24 23 21 19 19 21 26 0.055 0.040 0.021 0.045 0.050 0.000 -0.001 -0.002 0.055 Noakhali 30 31 31 24 24 27 32 35 31 -0.016 0.000 0.127 0.000 —0.059 -0.003 -0.001 0.002 -0.016 Huktaqacha 31 32 35 42 41 49 55 64 132 -0.016 -0.045 —0.091 0.012 -0.069 -0.003 -0.014 -0.023 -0.620 Saidpur 32 26 22 16 23 4 6 6 11 0.067 0.120 0.100 -0.122 0.704 -0.001 -0.001 -0.002 0.466 Heherpur 33 37 40 41 43 46 60 63 66 -0.057 -0.039 -0.012 -0.024 -0.034 -0.007 -0.001 -0.012 -0.455 Debhata 34 36 39 46 51 55 74 - 461 -0.056 -0.013 ~0.103 -0.030 -0.038 -0.009 0.036 -0.233 -0.663 Kushtia 35 35 30 31 29 23 25 29 21 0.000 0.077 -0.016 0.033 0.115 -0.001 -0.002 0.004 0.250 Hobiganj 36 33 36 36 33 36 42 62 66 0.043 -0.070 0.027 0.043 -0.043 -0.003 -0.010 -0.012 -0.410 Jhalokati 37 36 37 36 36 39 47 52 61 0.014 -0.014 -0.013 0.000 -0.013 -0.004 -0.002 -0.004 -0.245 Patuakhaii 36 34 34 39 34 36 40 41 36 0.056 0.000 -0.066 0.0;; -0.056 -0.001 -0.000 0.001 0.000 62 1911 1921 1931 1941 1951 1961 1974 1981 19” mmmw w Ralza‘ssalagalal ‘l “121.4567! 1.“ Mm. mm ‘- Decrease (1901 101981) Increase or Stability (1901 to 1981) ---~------- 'Cifles which are not Incbded In the analysis Rank-size changes, 1901-1981. Figure 4.5. 63 Classification of cities by Rank-nobility (u) Index All the 38 cities included in this study are classified into five groups in terms of there rank mobility using the following criteria: 9:222 V6 ‘ 1. No net change Zero (.000) 2. Minor change Between -.300 and +.3OO 3. High degree of gain Above +.3OO 4. High degree of decline Between -.300 and -.500 5. Very high degree of decline Below -.500 The classification of the cities under the above groups is presented in Table 4.5. It is noted that out of the 38 cities with population above 5,000 in 1901, 14 have indices below -.300, which suggests that these are doing poorly in terms of rank stability. Further, out of the 13 downward moving cities, 8 cities have indices below -.500 suggesting severe decline of the cities. It is noted that 5 of these severely affected cities are situated west of Dhaka 64 .558... «and. 88.6.. :38... .88 53...... 5.3. m8...“ .838: 5:82 «RS 538.. 8.50 3.8 55.5.. 558m 5... 6:38 5552 3...... 56.. .880 «as. .8338: 5.88 «2...... 8.15. .580 8...: 59...... 5.3.. 5.8 .558: .650 22... .126... .5558 am..." 585.38. 55.8 an 5.08: .5552 80.3. is... 556m Sufi she... 5:82 an: 22;. :38... 5.3. 55. 5.82 5...... 52.3.2 5.....8 :3. 5.8.3: .880 3.... .888: 5...... .88. 6...... .580 2.... 5.1.5. .5552 5.8. .358 .5652 5.: 556 .850 :3... 5585.2 5..82 2...: 8.8x. .56... «8...... 5.5.5.6.. .580 5.8 5.5.5. 55.3 5.5. 03.3 885.0 .558» 8...": 833 3...... 5:82 noes. 5.... .3. 5.80 .3. 5.80 .8. 5.80 .8. 5.50 8.83883 8...: 8.53 8.38.... 8...: 888.. 8.38.... 8a.... 8.56.. 8.3.2.6.. 8...: 8.59. 8.3.88 5.5 89:0 :85 88:993.- ".85. A8? "50.5 . . 88$ 0. 8m: 0.8.5 2480 30—: >mm> 24qu ECHO Emu—m Z75 mun—Own EOE moz<=U £052 50...... 3.85-0.5. 6. .5888 5....0 8.... .o 88.0.5.0 .3. 6.8a 65 between Calcutta and Dhaka city. Then, 5 have indices above +.300, which indicates that they are doing well. On the basis of the above observations, it is also noted that stability varied from city to city. The shift in rank-order for small cities is considerably higher compared to that for larger cities. It is also noted that the top ranking cities continued to maintain their higher status. In Figure 4.5, the top two horizontal lines represent the two cities of Dhaka and Chittagong continue to dominate the urban system or they never experienced negative mobility during the study period. cities Showing Unique Pattern of Performance Four of the cities included in this study have shown unique patterns of performance in rank maintenance. Dhaka continued to dominate the urban system by maintaining its first position; Khulna showed highest upward mobility; Debhata showed highest downward mobility; and Barisal which had experienced both upward and downward mobility during the study period finally attained the rank it had in 1901. These four cities are further discussed below: Dhaka: The index of Dhaka being 0 (zero) indicates a steady growth and rank stability overtime. The growth pattern analysis of this ever-growing city has shown that its growth has remained stable overtime. Dhaka, as the capital and 66 premier city, is centrally located in the country on the left bank of the Buriganga river, a tributary of the Meghna- Ganges system. It is an old city and is established since the seventh century A.D. At that time it was known as Bengalla, a small marketing town with its center near the present Bangla Bazar (Dutt, 1983). Dhaka showed a continuous growth since 1947 when British was partitioned and Pakistan was created. Since 1947, Dhaka grew very fast in population and areal extent as the capital of the eastern region of Pakistan and more as the capital of an independent country since 1971. Before 1947, Dhaka city was a district headquarter and a university town. There was one glass factory, two match works, one cotton textile mill, one general engineering works, and one pharmaceutical manufacturer. There was no industrial area. After 1947, Dhaka became the provincial capital of the then East Pakistan and was the chief administrative and business center of the province. The main capital was at Islamabad in West Pakistan (now Pakistan). The two provinces were separated by about 1,000 miles of the Indian territory. The physical separation induced to open embassies and consulates of a number of countries in Dhaka. Almost all commercial firms that are worth the name in the province had their offices in the city. Today’s three well defined industrial areas: Tejgaon, Postogola, and Hazaribagh were established in the Pakistani period. Also, a number of industrial establishments, both large and small 67 had increased. Besides, Dhaka occupied a key position in Pakistan as a major center for imparting technical training and university teaching and research. Also, the Pakistani government developed a large number of residential and commercial areas. In this time, large-scale immigration of Muslims from India contributed to the rapid population increase. The establishment of jute and other manufacturing industries has also contributed to the city's growth. This multifarious growth of administrative-commercial activities became pull factors for the migrants from the rural areas and smaller towns. During this time, population had increased by 64 percent. Dhaka has been experiencing its rapid growth since its independence in 1971. It is the tenth-ranking city in South Asia and is also one of the fastest-growing cities in the region. Khulna: The positive index value .760 for Khulna indicates a steady growth. Also, as seen in the Table 4.3, in the year 1901, this city had occupied 23rd place in the urban system and in the year 1981 it came to a third place. This overall increase in status may be attributed to the development of this city as an important sea port and an industrial center by the Pakistani government. This is evident from the increased index (.263) for the period 1951-61 (Table 4.4). 68 Debhata: This city has the highest negative rank-mobility index (-.863) which suggests highest downward movement in the urban system. It is seen from the Figure 4.5 that it has experienced a continuing decline in the rank order irrespective of the regime in power. This severe decline was due to lack of industrial, commercial, and administrative functions. Barisal: Though its rank-mobility index is zero, it has experienced both rise and fall in its rank status as may be noted from Table 4.3. From 1911 to 1921, it had moved from rank 7 to rank 3. It maintained its rank till 1951 and started to decline reaching the rank 7 again in the year 1974 and thereafter maintained the same status till 1981. The Census of Pakistan (1961) also noted this declining trend of Barisal. The fall in rank in 1951 when the Pakistan government took over, could be mainly because it remained in the shadow of Khulna a city 50 miles away from Barisal, which caused out-migration. However, it moved back to its original rank 7 in 1981. It is not very clear as to what factors have contributed to this upward movement. An analysis of change of rank-order or the growth pattern, using the ‘rank-order mobility index' (Table 4.3) has shown that stability varied from city to city. The 69 shift in rank-order for small cities is considerably higher compared to that for larger cities. It is also seen that the variations in the stability have some correspondence with the change of regime that came into power over time. Nearly 37 percent of the cities have shown a declining trend (index below -.300) over the years. It is also noted that 57 percent of them have severely declined (index below.-.500) and are mostly in the western part of Bangladesh, situated between Dhaka and Calcutta cities. Islam (1971) through his study of growth of the urban centers also found that the number of declining centers have generally been in the western part than in the other areas. Islam suggests that one reason for this pattern is that these centers are located in low-lying and ill-drained areas. This phenomena may also be explained by the disruption of nodality of communication which existed during the British time. CHAPTER V SUMMARY AND CONCLUSION The purpose of this study was to analyze the structural changes of the urban system in Bangladesh over time, and to examine whether the rank-size trend is toward more concentration (primacy) or deconcentration (convergence in rank-size) 1901-1981. The analysis of the growth pattern was done using the data available from decennial censuses I____ ! conducted by the governments of India, Pakistan, and Bangladesh from 1901 to 1981. The criteria for defining urban centers are the same as those used by the Bangladesh government. Analysis was conducted using the following qualitative and quantitative methods: (1) rank-size plots, (2) rank-size regression, and (3) rank-mobility index (M). Graphics and tables have been used to present the results on the rank-size plots for nine census years. Regression results were based on log population of central places used as the dependent variable and log rank as the independent variable. The findings of this study are presented in three sections: (1) historical development of rank-size relationships of urban centers in the urban system, (2) structural Changes in the urban system, and (3) growth patterns of selected urban centers. 70 71 t o eve o n o ank-Si e elat ons 21.nxhsn_§!zssn / During the years 1901-1981 Bangladesh went through three major political changes: (1) the British period (1901-1947), (2) the Pakistani period (1947-1971), and (3) the independent Bangladeshi period (1971-Present). The study has revealed that the city-size distribution pattern is well marked by the above political changes. The British period is characterized by dual primacy of Dhaka and Chittagong the two largest cities of Bangladesh. During the British time, Dhaka was used as a center for collecting raw materials. Chittagong as a sea port for trading of rice, jute, and tea. It is also observed that there was no medium size city during this period which had resulted in an uneven urbanization. These observations confirm that colonial rule had laid a foundation of primate city development by putting more emphasis one or two cities. Such institutional and functional barriers hampered a balanced growth of the urban system in Bangladesh. During the period of Pakistani occupation, dual primacy diminished and a third city Khulna started emerging as an important city in the urban system. Population concentration in Dhaka began to take place. From 1961 onwards, a tendency towards a ”balanced" growth of urban centers, i.e., existence of secondary cities, had emerged. This observation may be explained by the fact that during the Pakistan period importance for smaller urban centers was growing resulting in a situation that narrowed down the gap 72 between the expected and the actual population in those centers. This study has also revealed that during the Pakistan period the growth of large cities was greater than the corresponding growth during the British period. This may be seen as a result of the Pakistan government investing heavily for development of industries in Dhaka, Chittagong, and Khulna. The study has indicated that during the Bangladeshi period the relative growth of population in the second city Chittagong decreased compared to Dhaka resulting in a " further decline of the dual primacy. This trend had continued till 1981. From 1974 to 1981, a trend toward deconcentration had set-in. However, the population of Dhaka city had increased during the period and at the same time a trend toward a higher spatial integration began to appear. Reduction of the duality during the Bangladeshi period may be explained by the fact that in this period Dhaka became the capital city of independent Bangladesh as opposed to being only a provincial capital during the earlier regimes. During the Pakistan time, Islamabad, the capital of Pakistan was the focus of international interests and it eclipsed the importance of Dhaka. With the emergence of independent Bangladesh, Dhaka began to receive an international status. Nonetheless, concentration slackened during this period owing to the growth of several urban centers under the decentralization policy of the Bangladeshi 73 government covered by the Second Five Year Plan (1980-1985). Under the decentralization policy, the government of Bangladesh has given special attention to the development of small urban centers which serves the purpose of promoting grass-roots development. The historical development of city rank-size relationships shows clearly the structural effect of the policies of the British, the Pakistani and independent Bangladeshi over the years beginning in 1901. The major observations noted in this area relate to ggnggntzatign, dualitx. rankzsizs. and desensentratien- This study has revealed that during the British period a tendency towards ggnggntrgtign was in the process of emergence which truly manifested itself in the Pakistani period and continued till the early stages of the Bangladeshi period. Duality which was strong during the British regime, diminished during the Pakistani regime and further diminished during the Bangladeshi period. Bankgsigg or "balanced" growth of the urban system was practically absent during the British period but had begun to manifest itself in the Pakistani period and became prominent in the Bangladesh period. A process of ggggnggntratign had set-in in 1981 as the concentration trend began to diminish with the emergence of Bangladesh as an independent nation. This deconcentration may be attributed to the decentralization policy of the Bangladeshi government. 74 c C a e U n s This study has revealed that the urban system in Bangladesh had gone through three important structural changes during the period 1901-1981. These changes are punctuated by dual primacy, primacy, and decentralization. During the census years 1901 to 1941, the Bangladeshi urban system manifested dual primacy. This confirms the findings by Elahi (1972). Berry (1972), however, found no duality in the Bangladeshi urban system. Historically, Dhaka has always been the main city of Bangladesh due to its advantageous geographical location in the country and also due to its proximity to natural resources like the tea plantations. In addition, the fine cotton fabric industry located in the city has added to its prominence. Chittagong has always been a port city even before 1901, but increased its function when the British regime developed it for shipping rice, jute, and tea. It became a port officially in 1928. As such, Chittagong's growth was boosted from 1930 onwards contributing to the duality of the urban system. It is only in the 19508 that Chittagong's importance diminished when Khulna's growth became rapid. A move toward primacy in the Bangladeshi urban system is observed after 1950. Dhaka became a rapidly growing metropolis in Bangladesh and some of the services like administration, communication links, educational facilities, and centers of cultural interest were located more in Dhaka than in other centers. Another possible reason for the 75 increasing of primacy after 1950 was that Chittagong which was the major port, started loosing its prominence after the establishment of another sea port near Khulna and the expansion of the river-ports of Narayanganj and Chandpur. This study has noted that during the period 1974 to 1981 there was a growth of a large number of small urban centers causing deconcentration. Deconcentration may be attributed to the policy of the government of Bangladesh as outlined in their Second Five Year Plan (1980-1985). This policy was based on restraining the rapid growth of large cities so as to ensure the distribution of development benefits to the rural areas. The policy aimed at mobilizing local communities to participate in national development. e o es - This study has shown that rank-stability varied from city to city. It is also seen that the variations in the rank-stability have some correspondence with the change in political regimes that came into power in Bangladesh from 1901 to 1981. Dhaka, Chittagong, Rajshahi, Comilla, Barisal, and Patuakhali have experienced no net change in terms of rank-mobility during the period 1901-1981. But Dhaka and Chittagong maintained first and second ranks throughout these periods. As primate and second largest cities, Dhaka and Chittagong had always higher populations than other cities. As such, even when there was some decline in the growth of these cities, the smaller cities 76 were unable to overtake them in terms of their population. Mymensingh and Khulna showed a high degree rank-mobility. Development of Khulna as a port city and Mymensingh's metropolitan developments explain this change. Nearly 40 percent of the cities have a downward rank mobility trend over the years. It is also noted that about 60 percent of them have showed a severe downward trend. Host such cities are located in the western part of Bangladesh between Dhaka and Calcutta cities. Four cities were selected to demonstrate some typical 'L patterns of rank-mobility. Dhaka continued to dominate the urban system by maintaining its first position; Khulna showed highest upward mobility; Debhata showed highest downward mobility; and Barisal which had experienced both upward and downward mobility during the study period finally attained the rank it had in 1901. Rank-mobility appears to be dependent primarily on the investment policy of the government which is often biased in favor of large cities as noted by Wellisz (1971). It appears that the growth of large cities is unlikely to decrease in the near future. This observation seems to agree with the observation made by the World Bank that the largest cities in developing countries will continue to grow even if the national policy favours decentralization. 77 221W Policy implications may be discussed at two levels, general and specific. At the general level, this study gives an understanding of the Bangladeshi urban structure using urban hierarchical theory - the city rank-size rule. It also gives some directions on how to restructure the urban system in order to bring about an improved spatial organization of urban/rural relationships. However, a comprehensive spatial development of a country requires the knowledge of the structure and process of socio-economic dynamics and locational decisions of firms (Wubneh, 1982). More specifically, the study yields several implications. The first, has to do with decentralization. Brutzkus (1975) suggests that the total number of urban centers should not exceed the necessities to service rural areas adequately and to ensure spread effects of development benefits. Brutzkus indicates that a basic rule for the spacing of larger growth poles is for each to cover a complementary region from which people are able to commute at least on a weekly basis. Ideas like this based on studies of the urban structure would have greatly improved the government’s decentralization plan and prevented a drain of development resources which the 383 new ill chosen centers are currently creating. The purpose of creation of these new urban centers was to ensure a grass-root development by providing development facilities in all urban centers. The inappropriate choice of centers has led to 78 political tension, thereby threatening the stability of the country. The rank-size shift analysis suggests that some towns have no potential for growth. For example, Debhata shows a declining trend over the study period. Its declining trend caused it to be dropped out from the urban centers list in 1974 (Table 4.3). But vigorous decentralization policy of the Bangladeshi government picked up this city in 1981. It seems unrealistic to revive this city which simply costs additional resources. Debhata is situated in the extreme south-west corner of the country which is less populated. With better knowledge of the urban structure, potential growth centers could have been strategically located on rural urban migration paths and thereby serving the goals of decentralization. Studying relative performance of existing urban centers would have also helped to revive some apparently stagnant towns thereby saving resources. Secondly, studies of the urban structure could aid in reshaping the transportation and communication system of the country. This could be designed to link potential growth centers and incorporate underdeveloped peripheral regions into the national and development process. For example, in the south-eastern part of the country, there is only one large city, i.e., Chittagong. Being in a mountainous densely forested region, its transportation system is underdeveloped. But as it is a center for timber, fruits and other forest products, it should have been better 79 connected with other parts of the country to improve regional trade. Thirdly, rank-size plot of 1981 indicates a break in the fourth and fifth ranking cities, i.e., Rajshahi and Mymensingh which demand a special attention in terms of localizing development efforts of the Bangladeshi government. It is specially true for Rajshahi which is the leading city of the Northern part of Bangladesh. The regional development of this part solely depends on the improvement of this city. It is basically a university town. However, it has rich resources which should be attractive to industrial development if only the physical infrastructures were adequate. Lastly, the control of growth of Dhaka, the primate city of Bangladesh is also an important consideration. The study shows that Bangladesh does not have high primacy like most other less developed countries. Its central function could be enhanced by more development expenditure to improve the depressing conditions in the city, while at the same time steaming migration into it. This could be done by strengthening urban centers like Comilla, Noakhali, and Faridpur districts, the main areas from where the Dhaka city receives most of its migrants. APPENDIX Population and Rank of Bangladesh's Urban Centers 80 .m 88.. 8 88. NN 88. ..N 88. ..N 88. 8 88. .n 2.... .n 8.... 9. 88 28.82 cm .88 an 88. ..N 8.8 ..N “88 8. .88 2 828 a ..88 N. 88. o. 88. 85...: 8 8.8 - - - - . - - - - - - - - - - 8...... ..N 88.. ..N 88.. ..N 9...: 8 N8: 3. n8. 8 8.... - - - - - 8.8... 8 88.. NN N88 NN N88 2. ..88 2 .88 .N 8.8. a. 88. ..N a... ..N 88.. 58...... 8N 8.8 .N 8.8 a. .88 a. 888 .N .88 8 a8. 8 88. ..N a... 8 8.... 5.... 8 .88 8 SEN 8 88 - - - - - - - - - - - . 885330 ..N .8: - - - . - . - . . . - - - - - - 8.2.8980 8 8.8 8 88. 8.. 8n: - - - - - - - . - - - 8...... NN 88. ..N 8.8 8 N8... - - - - - - - - - - - . - .8552 .N N88 8 88m 8 N88 8 8.8 8 N8. .N 8.... o... as... 8 88 8 88 8.82 ..N 88.. 2N .28. 8 8.... 8 e82 - - - - - - - - - - 8.88.0 a. 2.8. a. 8.... 8 .88 NN an: ..N 88 a. N8... 8 88. n 88. N. 88. .85... 2. 888 ..N 888 2. as... 2. 2.8. n. 8...... 2 .89. 2 88. NN :8. 8 88 8.5.0 .. 28.. v. 8.8 N. 3...: a 88.. a $8.. 2 N88 2 3.8 2 .88 n “.8. 2.85822 8. ..NE. 8 88.. .N 8.8 .N .88 a. 8.8 ..N 88. 8 88. n NNNNN .. 88... .8288... n. “2.... s .88 m. 8...... 2 N88 N. 8.8 c. .88 e. 2.8 a 8.8 a 88. .88.... z 2.8.. a. 828 .. :5. .. .83. v. 8.8 n. 88. z 88. a. 88. 2. 688 8...... m. 88... .. .88 c. 8.8 a. 88. 8 8N... .. 8.8 n 288 n Ea N 3.8 82.28 N. .88. n. «88 N. N88 2 8.... .. 88. .. .88 N. 88. .. 8N... .. 8...: as... .. 288. a N28 8 88.. v .88 8 8.8. 2. 28. NN 2.x. 8 88 NN 88 25...... o. 88: o. 8.... : 88.. 8 8.8 8 28. ..N 88.. 8 8.... ..N ..8 8 88 85. a 8.8. N. 88. v. «8.... a. .88 o. .88 n. 8.8 n. 88... z 88. z 88. .8282 2 .88. 2. 688 a. 88. N. 88.. n. 8.8 N. “88 8. N8. 2. 83.. 2 88. 658 s 88: h 8.8 n 888 N 88.. n 8:... n .58 m 32.8 N 8.8 p 2.8. .88.. a 8...... a 38.. .. 3.8 n 8:... n N38 n 8.: v «88 8 N88 8 88. 2.280 n :8... n 8.8. a 888 8 88.. v 688 a 83m .. ..88 c. 88. n. 88. 888...: .. 888 8 S82 .. 88.. .. N88 .. 8...... .. «88 2 .88 .e 8.8 c .88 2.8.5. n 888 n 388 e 688. .. 8.... N. 9...... a. 8.... .. 88. .N 88. NN 83. 68.0. N .882 N 888 N .88.. N .288 N 888 N 88... N 8.8. N 888. N 88... 2828.8 . 8.9.3. . N88... . N58 . 888 . n88 . :8... . 2.8. . 888. . .88. 5...... 2 2 2 2 2 2 2 .. 2 2 2 2 2 .. 2 2 2 2 .8. 8... 8... .8. .3. .8. .N... .3. .8. E0 287.8. .556 .8... 22.82.82 .6 .282 .0... 8.5.8.. NuDZmN—Nz 81 8 8.3. - - - - - - - . - . - 82:2 8 ..8. - . - . - - - . - - - 8.3882 8. N8... 8 .88 - - - - - - - - - - - .8..... 8 ...N.. 8 8.... .. ..8. .. 8.... 8 N8... 8 ..8 .N ..3. 8 .8. 8 88 29.2 No 8.... .N .8.. . . . - . . - . . . - . - . 2.2.8.5 8 ..8. N. .88 8 8.... .. 8.. .. 8.. .. ..8 .. 8.. 8 ..8 .. ..8. 88...... 8 ..8. - - - - - - . - - - - - - - - - .3283. .8 ..8. .. ..8. .8 ..8. ..8 ..8 8 .8. 8 ..8 8 88 - - «.2... .8 38. 8 ..8. 8 ..8 8 .8. - - - - - - 5.58... 8 ..8. - . - - - - . - - - - - 8.32 8 ..8. - - - - - - - - - - .. $5.... 8 88. 8 .88 8 ..8 . - . . . . - 8.8.5.2 8 ..8. 8 .38 8 .88 8 ..8 8 8.... 8 ..8 - - a: .8..... 8 8.... .. .8.. 8 3.. 8 .8. 8 8.. 8 88 - - - .8225... 8 8... 8 ..8. . - . - . . - - - - - 3.5.... 8 N.... 8 .88 8 .8.. - - - - - - . - - - - .8352 ..8 .8.. 8 ..88 8 .8.. 8 ..8. 8 8... .N 8.... 8 .8. .. 8... 8 .8.. 53.82 8 .8.. 8 N.... 8 ..8. 8 8.. N. ..8. N. ..8. - . . - - 53...... 8 .8.. - - - - - . . . . - - - - 5...... .8 8.... 8 88. 8 88. 8 .8.. .. .3. .. 3.. 8 .8. - - - 8.2.2 8 3... - - - - . - - - - - - - - 8832 8 .8.. - . - - - . . - - - - - - - 2.39... 3. .8.. 8 .8.. 8 8... 8 8... .. 8... 8 ..8 8 ..8 8. 88 8 .8. £93.... 8 N88 8 .8. N. .8.. - - - - - . - - - 1585...... 8 8.8 . - . - - . . . . . - - . - - 5.....2 8 .8... .. N... 8 88 - - - - - - - - . - - 888...... 8 .8... .. 8.... 8 8.. 8 .8. 8 .8. 8 ..8 8 88 8 ..8 .8 2... 58...... .. 8.... 8 88. 8 8... 8 ..8 - . - - - - . - - .889... .. 8.8 8 8.8 8 88. .. ..8. 8 .8... .. 88 8 8... 8 ..8 .. 88 8.3.... .. ..8... .. ..8. 8 8.8 8 N8.. .. ..8... e. .8.. n. .8.. .. ..8. .. 88. 2...... in... 8 8.8 - - - . - . - - - - - - - - - 3.8.82 8 8.... 8 .88 .. 8.8 8 8.... 8 8.... 8 .8.. 8 ..N... 8 8.... .N 8.. 8.8... 8 88. 8 .88 .N .88 8 8.... 8 8.8 8 ..8. .. .8.. .. 8.... .. 8.8. 896...... 8 8.8 8 ..8. 8 .88 8 .8.. .. 88. 8 3.. . - - . - - 5...... 8 8.8 - - - - - - - . - . - - - . . . .32.... 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 8.. v... 8.. 8.. .3. .8. 8.. .... .8. Eu 82 2: $8 a a 5 - - - - - - . . - - 3w: 2: nnwon . - - - .. - - - - - .535 8 28a . - - - - - - - 5&5. ma 83m . - - - - - - - - - awn—Em 8 83 - - - - - - . - Bean: 3 mam—N - - - - . £835. 3 88a - - - - - - - gm 3 8mm" - - - - - - - - - 5&5. a 88" - - - - - - - - - - 58m 3 ofinu - - - - - - - =5 3 $32 - - - - . - - - Ewan—«3. co chums - - - - - - - - - Eco—£3 & an S 33 - - - . - - . - - . - . .5553 3 ER 8 3.2 8 $8 8 8:. 9 a: S 3% 8 8% R 6% a 8:. ES»: 8 ~88 - - - - - - - . - . . - . . - . .38 3 58 8 3a: 3. 582 on as. a on»: on 2.? a. 3% a 3% on can .898: B «88 - - - - - - - - - - - - - - - 33852 3 8an - - - .. - - . - - - .. 5:55 2 SSA - - - - - - - - - - - - - - 32.5. 3 58 a a; as 89. G 88 an and - - - - - 38:3 8 can" - - - - - - - . - - - - - «35: 8 82a - - - - - . - - - gm 3 «an . - . . . . . . . - - 85m 56 up 38 . - - - - - - - - - - . £593. 2. «$8 - - . . - - - - - - - - - - 8258 2 SE «5 8n: 2. 5n . - . - . - . - . - . Sims: 3 55 on «as a 35 a a: 8 :5 a as. a is an as: 2 2.: .35. 3. «as . . . - . . . . - . . - . . - :3. as Sea 8 88a 3 8% - - . . . - - - . - - .53.. 2. SEN . . . . - . . . . . . . - . . Ens: F 3.2" a. 53 R 8% an 3% 8 “82 - - - - - - - - E 2. Sea . - - . . . . . . . - - . - - .555 on ass 9. 83a - - - - - . . . - - . . - . 38.3 8 ”3% an Ba: 8 RR 3 PS 9. 8% a 83. c. in 8 23 9 88 553:2 5 zoom 3 85 R as a «an an 3% S 29. 9 «new 3 2.9 S a: 58.80 a a a m a m a m a .— u .— u a a m a a 82 E: 82 32 E: E: 82 :2 82 EU €025.80 - ESE—(v 83 n82 - - - - - - - - - - . - . . - gm :8: - - . - - - - - - . - .3355. 38. - - - - - . - - . - - - - - - Bea—530 «82 «5 so: an 38 a. as 3 E: a. as a 25 2 RS 8 8% 38;: 39: . - - - - - - - - - - - - . - - .255 an: - - - - - - - . - . . - . - - - 5&5 8a: 3 «8: mo 83 an 3% a an S :9. 3 as s. $8 3. 83 gas: 8n: - - - - - . - . - - - - - - - - sum «22 . - - - - - . . - - . - . was... 88. i 28. - - . . - . . . - - 2&3 8a: - - - - - - - - - - - - - .5925.— .82 - - - - - . - - 255 «a: - - - - - - - . - - - - Fania SN: - - - - - - - - - - - - - - - - .555 as: 3 g: a. ~82 : 8:: h an: a as: 2 8n: a «$2 a 8§ 3:8 3...: - - - - - - - - - - - - . - - gm «8: z. 29.. . - - - . . - - - - - - - - .33 «a: . . . . . . - . . . - . .335: was. 2. .82 an on: - - - - - - - - - - - - .aaaaoo :2: - . - - - - - - - - - - - - - €553 Rx: - - - - - - - . - - - - - - - . .5982... was: - . . - - - . - - . . - . - - 3&5 $9: - . - - . . - - - - - - - . - - e355. 8:: - . . - - - - - - - . - . - - 32.8 «E: 2. 2n: 3. 9.8 «e S: 8 3:. 9. 8% s. 38 an 83 9. 82 3.5% $2: - - . - - - . - - . - . . . - . Be...— 232 B 38. a 85 2 Ban 8 an. 9. =3 . . - . - . 3..-pan $.8— - . - . - - - . - - - - - - - - - 5:8 NE: . - - - - . . - - - . - - - - - =5. .8.. 3:: - . . - - - . . - - - . - . - - 52:53 8%. 8 as: . - - - - - - - . . - - - - as: «88 3. 2mm - - - . . . - - - - - - - - 89.3 33" 8 5.2 an :5 a 88 a. 23 a. as an 9.2. 8 £8 a 38 28.65. m a a a a a .— u a a a a .— u .— a .— 83 E: 32 Ge .2: .8. as :2 8.: r5 69.5.80 - van—295$ OKPPfltflDDK -(Runhnuxb 1981 1921 1931 1941 1951 1961 1974 P .P P 1911 1901 137 138 139 140 141 142 15086 15032 15004 15421 15273 14936 15630 78 13248 78 577 143 144 145 146 147 148 149 74 5892 53 7028 63 10020 42 to... d on 0.0.0:: can-08 6319 3335382 ghnnnn ~33333 150 an ‘5 151 13991 13979 13828 152 68 3885(51)60 5353 54 55 5582 362 3 39 3769 43 3763 46 4584 «a v1 vaaxun "fl“ 5138 r~r~en «avavu “F.— I II I II I I I III I II I I I I I I I II I I I I II I II II I I I I I I I I I I I I I 156 157 13643 13157 76 13147 12997 158 159 12974 160 161 12788 12646 12631 162 163 164 165 166 167 12589 12427 12351 12197 12175 168 169 12166 lhnhnchmr 12148 hflohongaqj Oo— r~r~ I-l—I I‘m MN c-IU-I NN u-oc-n a o o u o o o a u o u n o u a u u a 1 u n u I I n. n a a u o a Bakerganj 85 83 8n «n? 3a 33 8a $8 8" :8 an as . 8a 38 255.58: 9: Ex: - - 2.: ~32 R :22 5 2:: - - we 8:: no. 52 2: Sn: 8. mom: «a :9: 33.830 .2 name 9: can: - a: 82: - a: 82: - 2: ans - a: 8§ - n: 8:: - . a: 82: - $353.0 2: c8: - «a ”3: - z: E: - 9: c8: . - - - - - - . - . - - . - 35:8 a: GE 3 n3. 2. use an 3% on man an 39. 9 3% an 39. S 3:. ace 5928 a: mac: - - - - - - - . - . - . - - 5 Km: . a: as: n: 80: v: as: n: as: a: n8: 6% m a a m a a a a a m a a m a m m a .— SE 3.2 62 :2 $2 32 82 :3 82 €35.80 . Magus 86 Sn 2% .3: 8a was .395. am 9.2. Saga: an" 82 859 an 88 2.55. on" 88 3322 n8 8; i=8. ..8 a; £95»: an 88 - - - - - - - - - - - - - - - . 5338 an" an; 2: a: Q 2:. 8 as" R as an an 3 :3 s. .5. 9. 8:. 59:23: an 3.3 - - - - - - - . - . - - - - - . .322 as EV» EEE an 33 E532 ”an «an» - - 53586 an 32 8 8; :83 8a 32 . - 388 an «m8 369 EN $8 :55 am 85 Ea: «mm as £35 an 8: is: 8" 2% 3533a as 8% .35 «a 88 is: :u :8 .8580 2a 28 .33 a on; 5.835 v8 $8 3388.. an 3a 83 Na :8 as 5 R8 in: em :3 1.38m SN 33 .252 8a 33 3:55 8" 33 29.5 a a a a a .— u a a .— u a a a a m a a 82 3.2 82 Ga .3. .8. 82 :2 82 EU I 7 III? A3338. Enzmks 87 2." 68 5833. E :8 583m «8 88 9.8 2m 38 5380 NR 85 £38 5 3% . BE. :5 9% and. a." as 655?: 8a 2.3 5&3? SN as .5333. 8n 8% 5&9:— BN 2% 3:895 3" £8 $.85 Bu 2% 65°82 SN 2% 52.3 .8 8% Eu 8" ~85 86.38 on 32 again «a 8:. .2955 5 3:. 53380 on $9 .8589 «mm Sn. 593 X" a? 355— ng $42. as «a 82. ii: 5 an? 59.32 ca «.35 538m .3 :3 .853 3 «5 BE .52 3 S? E! 3 2.3 Sea 3 3:. .2320 3a RF Egan 9a «a: £3.53 3 82. use: so a .— m a a m a .— z m a m x a u a a a $2 3.2 Ge E: .3. 82 82 :2 82 E0 €95.80 . van—Ems SR - - - - - - - - - . . - - - . - Eng «no... - - - - - - - - - - . - - - . . .3553 $3 - - - - - - - . - - . - - - - . its... 2% - - - - . - - . - - - - - - - - “Ema—Em can - - - - - - - - - - - - - - . . 2&8 3n - - - - . - - - - - - - - - - - .25: 3% - . - - - - - - - - - - - - - - ..8»qu 8am - .. - - - - - . - - - - - - - . 32928. Ex - - - - - - - - - - - - - - . . inn—5293 3% - - - - - - - . - - - . - - . . 5.9.83.6 :8 . - - - - - - . - - - - . . . . 3293:: 88 . - - - . - - - - - - - - . - . §§< $8 - - - . . - - . . - - - - - . . raga as - - - - - - - . - - - - - - - - 25am BS . . . - . - . - . - . - - - - . g 88 - - - - . - . - - - - - . - . . 535m 88 - - - - . - - - - . . - - - - - Esau RS - . - . - . - . . - . - - - . . 3.9.80 88 - . - - . - - - . . . . - - . . Eon 2.8 - - - - . - - . - - - . . - . - :55 2.8 - - - . . - - . - - - - - - . - 55:5 88 - - - - . - . . - . - - - . . . isawfi 88 . - . . . . . - . . . . . . . . Bonsai 93 - . - - - - - . - . - . - . - . £95 £3 - . - - . - - . - - . . . . . . .32 SS . - - - - - - . - - - - - . . . 52a?— 83 . - . - - . . - . . . - - . . . .5338 28 - - - - - - - - - - - . - - - . .388 A M A A A a A m A ~— A m A A A x A 32 32 53 33 30— Ana ANS :2 53 EU goon - Egv .8.-—n .AA .83. 895 .8 :33— .33 £550 .633 £83355 .33 .. . .-8 .8 .v 0.3... ..8. ..an 8.358.. gang. 3.3. .8... s a... E .v 0...: ..8. 3...... .o 3.80 ”.878... 89 38...}... 8:8 .5. a .8338. .. as: ..8. .2583 son 5...... ..8“ 56.x. «..8 5.5... 8: .252... 3.3 39.55.55 2...“ - - IE... 8% 8 as... at... :3 . - 8%.. a.“ 3.55.... man 8.8.. E. .3538. .3 39.5.. $2 38.. an ..E... 9% 53.3 an 5.5.8.. ..8... 3:8 .. a a a .. a .. a A. a .. .8. «a. .3. .8. .3. .8. E 90.5950 - Egg LI ST OF REFERENCES LIST 0! REFERENCES Ahmad, Nafis. 1976. A flew Economic geography of fianglgggsh. New Delhi: Vikas Publishing House Pvt. Ltd. ' Ahmad, Qazi Kholiquzzaman (ed.). 1980. nggglgpmgnt ' e - e ' o £12g_2g§;_2;gn. Dhaka: Bangladesh Unnayan Parishad. Ahmad, Qazi S. 1967. "Distribution of city sizes in Pakistan". Eakistan Geographical Review. Vol. 22: 77-85. Atigullah, H. and F. Karim Khan. 1965. ngg§h_gfi_phgk§ ° ° ' 6 - . Dhaka: University of Dhaka. Berry, Brian J. L. 1971. "City size and economic development: conceptual synthesis and policy problems with special reference to South and Southeast Asia." in urbanizaLi2n_and_fla§ignal_nexelgemen_. Leo Jakobson and Ved Prakash (eds.). Beverly Hills, CA: Sage Publications. Boal, F. W. and D. B. Johnson. 1965. "The rank-size curve: a diagnostic tool?" 0 s 'o o . Vol. 17 (5): 21-23. Brutzkus, Eliezer. 1975. "Centralized Versus Decentralized patterns of urbanization in developing countries: An attempt to elucidate a guideline principle." Eggggmig Dexel22men§_and_§ultural_2hangeo Vol. 23: 637-40- Carroll, Glenn R. 1982. "National city-size distributions: what do we know after 67 years of research?" Egggrgsg in_numan_§eegranhx- Vol- 6: 1-43- Chowdhury, H.I. and Hohd. Asad-Uz-Zamman. 1976. "Settlement pattern and some special problems." Habitat. United Nations Conference on Human Settlements, Vancouver, May ll-June14, pp. 49-63. 90 91 Choudhury, A. H. N. A. and Anil Chandra Das. 1975. ”The role of settlements planning in national development policy." Habitat: United Nations Conference on Human Settlements, Regional Conference, Tehran, June 14-19, pp. 101-106. Choguill. Charles L- 1987. Hew.922munities.f2r_nrban 2L: —. -‘r.!‘ .-u 9‘ “.1 1 '9: 1 ial°_ad‘-1. New York: Plenum Press. Centre for Urban Studies, University of Dhaka. 1976. "Squatters in Bangladesh Cities." Centre for Urban Studies, University of Dhaka. 1989. "Experience of external assistance to urban planning activities in Bangladesh” Qg§_fiullg§in. No. 16 8 17: 10-11. Danta, Darrick Rollin. 1985. ”Identifying agglomerative/deglomerative trends in the Hungarian urban system, 1870-1980." Unpublished Ph.D. diss. The Ohio State University. Dickensen, J. and A. Hodgekiss. 1983. A_§gggrgphy_gfi_thg Ih1;g_flgzlg. London: Hethuen. Dutt, Ashok K. 1983. " Cities of South Asia” in Qitig§_gfi tne_H2rldi__H9rld_Besignal_nrban_nexelgnment- Brunn. Stanley D. and Jack F. Williams (eds.). New York: Harper & Row Publishers. Elahi, K. Haudood. 1972. "Urbanization in Bangladesh: a geodemoqraphic study-" Ine_inental_§eegranner- Vol. 16 (1): 1-15. Fan, C. Cindy. 1988. "The temporal and spatial dynamics of city-size distributions in China." EQlegtiQn_Bg§gargn and_EQlisx_Bexie2- 7: 123-157. Fisher. James s. 1989. Qe2granhx_and_nexelgnmenti__a_flgrld Begigggl_fipprgg§h. Columbus. Ohio: Herril Publishing Company. 92 Friedmann. John- 1966. Bsgi2nal_2exelsnmen§_zgli§21__a case_§tndx_9f_2snezusla- Cambridge: The MIT Prese- Government of the People's Republic of Bangladesh. The 0 v - . Dhaka: Planning Commission, Ministry of Planning. Government of the Peoples Republic of Bangladesh. 1985. e 9 - . Dhaka: Planning Commission, Ministry of Planning. Government of the People' s Republic of Bangladesh. 1981. aangladesn_zsnulatign_sen§uso Dhaka: Bangladesh Bureau of Statistics. Government of the People's Republic of Bangladesh. 1981. Bangladesn_292ulaLi2n_Qensus1__Bengrt_gn_nrban_Arsa. Dhaka: Bangladesh Bureau of Statistics. Harris. Chauncy D- 1970- Qities_9f_1ns_fisxist_nnien- Chicago: Rand HcNally and Company. Haggett, Peter and et a1. 1983. figgg;gphy;__5_flgggzn Synthesis. New York: Harper 8 Row, Publishers. Islam, Nazrul and Amanat Ullah Khan. 1988. "Increasing the absorptive capacity of metropolitan economics of Asia: A case study of Dhaka.” Begi2nal_Dexelsemsnt_nial2gue- Vol.9 (4): 107- -133. Islam, Nazrul. 1989. "Environment, economy and land use with special reference to slum and squatters in Dhaka city, Bangladesh." Paper Presented at the Division of Human Settlements Development Asian Institute of Technology, Bangkok, Thailand. Islam, Nazrul and Hemayet Hossain. 1975. "Relationship of urban centers with their rural hinterlands." Habitat: United Nations Conference on Human Settlements, Regional Conference, Tehran, June 14-19, pp. 51-106. Jensen, Lois. 1988. "Developing countries begin grappling with urban woes." flg;1§_ngxglgpm§n§. 1(4): 4-8. 93 Khan, Abdullah Al-Hamun. 1982. "Rural-urban migration and urbanization in Bangladesh." Ing_gggg;§phig§1_3gylgw. Vol. 72 (4): 379-394. Logan, H.I. 1972. The spatial system and planning strategies in developing countries. Thg_§eggzaphiga1 Bgyigg. Vol. 62: 229-244. Lyman, Brad. 1992. "Urban primacy and world-system position”. Uzban_5fifigiz§. Vol. 28 (1): 22-37. Marshall. John 0. 1989. The_§trusturs_gf_nrhan_§2stsm§- Toronto: University of Toronto Press. Richardson, Harry W. 1987. ”The cost of urbanization: a four-country cenparison-" Essnsmis_nexelssment_and Quitu;g1_gn§ngg. Vol. 35 (3): 561-579. Robson, Emma. 1991. "The life-saving mission of Bangladesh’s diarrhoea doctor." figures. 3(3): 4-7. Rizvi. 8- N. H- (ed-)- 1969. East_£akis§an_nistrist fiazg;;ggzg,_nngk§. Dhaka: Government Press. Robson, Brian T. 1973. n:bgn_§zgw§h;_§n_Apngggh. London: Hethuen & Co Ltd. Rondinelli. Dennis A. 1983. fies2ndarx_Qities_in.Dexelgsing Beverly Hills, California: SAGE Publications. Shakur, Hohammed Tasleem. 1987. ”An analysis of squatter settlement in Dhaka, Bangladesh." Ph.D. Thesis. University of Liverpool. Shibli, Khalid. 1980. "Pakistan and Bangladesh". . Arnold Whittick (ed.). New York: Robert E. Krieger Publishing Company. Sheppard, Eric. 1982. "City size distributions and spatial economic change." Eggigw. 7 (2): 127-151. 94 United Nations. 1983. "Urbanization." World population Trends, Population and Development Interrelations and Population Policies. Monitoring Report. Vol. 1, Population Trend, pp. 181-206. Wellisz, Stanislaw H. 1971. "Economic development and urbanization. " Wat Leo Jakobson and Ved Prakash (eds.). Beverly Hills, CA: Sage Publications. Williams, Jack F. et al. 1983. "World urban development." in C t es 0 ° d onal a Qggglgpmgnt. Brunn, Stanley D. and Jack F. Williams (eds.). New York: Harper 8 Row Publishers. World Bank. 1979. ”Urbanization: patterns and polices." W. Washington D- C- = Oxford University Press. World Bank- 1992- W WWW- Washington D.C.: Oxford University Press. Wubneh. Mulatu- 1982. WWW 199W. Foreign and Comparative Studies/African Series XXXIX. New York: Syracuse University. zipf. G. K. 1949. a e ' he r ' Leas; Efifioz . Cambridge. Massachusetts: Addison Wesley. HICHIGAN STATE UNIV. LIBRnRIEs \NllmlllHHIIHII\IllHIIIlIHIbIWIIIIlHHHIHIIIIIIHI 31293007914652