UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Determinants of the spatial dynamics of population movements within Bangladesh Haq, Ziaush Shams 1974

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1974_A1 H36_9.pdf [ 16.19MB ]
Metadata
JSON: 831-1.0093159.json
JSON-LD: 831-1.0093159-ld.json
RDF/XML (Pretty): 831-1.0093159-rdf.xml
RDF/JSON: 831-1.0093159-rdf.json
Turtle: 831-1.0093159-turtle.txt
N-Triples: 831-1.0093159-rdf-ntriples.txt
Original Record: 831-1.0093159-source.json
Full Text
831-1.0093159-fulltext.txt
Citation
831-1.0093159.ris

Full Text

DETERMINANTS OF THE SPATIAL DYNAMICS OF POPULATION MOVEMENTS WITHIN BANGLADESH by Z I A U S H S H A M S HAQ B . S c . ( H o n s . ) i n G e o g r a p h y , D a c c a U n i v e r s i t y , 1 9 6 5 M . S c . i n G e o g r a p h y , D a c c a U n i v e r s i t y , 1 9 6 7 A T H E S I S S U B M I T T E D I N P A R T I A L F U L F I L M E N T OF T H E R E Q U I R E M E N T S FOR T H E D E G R E E OF DOCTOR OF P H I L O S O P H Y i n t h e D e p a r t m e n t o f GEOGRAPHY We a c c e p t t h i s t h e s i s a s c o n f o r m i n g t o t h e r e q u i r e d s t a n d a r d T H E U N I V E R S I T Y OF B R I T I S H C O L U M B I A O c t o b e r 1 9 7 4 In presenting th is thesis in par t ia l fu l f i lment of the require-ments for an advanced degree at the Univers i ty of B r i t i s h Columbia, I agree that the L ibrary shal l make i t f ree ly ava i lab le for reference and study. I further agree that permission for extensive copying of th is thesis for scholar ly purposes may be granted by the Head of my Department or by his representat ives . It is understood that copying or publ icat ion of th is thesis for f inanc ia l gain shal l not be allowed without my written permission. Department of Geography  The Univers i ty of B r i t i s h Columbia Vancouver, Canada Date U^L October 1974 ABSTRACT This d i sser ta t ion demonstrates that the spat ia l pattern of migration within Bangladesh can be 'expla ined' in a 'push-pu l l ' framework. As the 'push-pu l l ' theory of migration has been developed in the context of the developed world, the f indings of th is research va l idate the app l i c -a b i l i t y of the 'push-pu l l ' theory cross-cultural ly . However, the p a r t i c u l a r 'mix' of 'push' and ' p u l l ' factors was found to be d i f fe rent from what is normally expected. By employing pr inc ipa l axis type factor ana lys i s , i t was observed that the ' p u l l ' factors were dominant in determining the overal l spat ia l pattern of i n t e r - d i s t r i c t migration within Bangladesh. As Bangladesh is overwhelmingly rural and population pressure on agr i cu l tura l land is excessive, i t was expected that migrants would be compelled to leave the i r places and 'push' factors would be dominant. The analysis involved the fol lowing steps: (1) l i t e r a t u r e survey of theoret ica l and empirical approaches to migration to j u s t i f y concentration on the 'push-pu l l ' theory, ( 2 ) se lect ion of s ix kinds of migration streams such i i as (a) Rural-to-Rural Out-Migration Streams, (b) Rural-to-Rural In-Migration Streams, (c) Rural-to-Urban Out-Migration Streams, (d) Urban-to-Rural OUt-Migration Streams, (e) Urban-to-Urban Out-Migration Streams and (f) Urban-to-Urban In-Migration Streams; (3) Generation of a priori hypotheses descr ibing the d i f f e r e n t responses of 'push' and ' p u l l ' factors to the migration streams and the roles of (the intervening var iab le) distance on the migration streams as w e l l ; (4) estimation of migration data from ind i rec t source such as 'place of b i r th data' in the absence of d i rec t data on migrat ion; (5) se lect ion of explanatory var iab les ; (6) appl icat ion of p r inc ipa l axis type factor analysis to each migration streams (within each kind of migrat ion); (7) se lect ion of ' c r i t i c a l f a c t o r s ' ( C F . ) from factor matr ices; (8) preparation of tables (99 of them) incorporating factor loadings on the v a r i a b l e s , factor scores on the subjects ( d i s t r i c t s ) , numerical data on migrat ion, migration data expressed as ' v e l o c i t y of migrat ion' (V.-.), and distance separating d i s t r i c t s of o r ig in ( i ) and d i s t r i c t s of dest ina-t ion ( j ) ; (9) in terpretat ion of the tables in terms of the 'push-pu l l ' theory; (10) test ing the v a l i d i t y of the working hypotheses. The analysis revealed that while population pressure on agr i cu l tu ra l land, RD/SQMCA, and the presence of too many i i i landless a g r i c u l t u r a l labourers , LLALF/TALF, acted as the 'push' fac tors ; the a v a i l a b i l i t y of more land under r i c e and j u t e , Rice/TCA, Jute/NCA, and better opportunit ies for the expansion of cu l t ivated land due to low intens i ty of cropping, TCA/NCA, acted as the ' p u l l ' factors most f requent ly . It was hypothesised, a p r i o r i , that while the 'push' factors would be dominant in Rural-Rural out, Rural-Urban out and Urban-Urban out-migration streams, the ' p u l l ' factors would be more important in other streams such as Rural-Rural i n , Urban-Rural out, and Urban-Urban in-migration streams. In the hypotheses i t was also l a i d down that distance w i l l play a more important ro le in Rural-Rural i n , Urban-Rural out and Urban-Urban in-migrat ion streams than in Rural-Rural out, Rural-Urban out and Urban-Urban out-migration streams. But from the resul ts i t was discovered that the hypotheses were va l id only p a r t i a l l y . While the inf luence of the ' p u l l ' factors were more than the 'push' factors in explaining the total migration system of Bangladesh, in 1951, the role of distance on migration was as expected. In 1961, the importance of ' p u l l ' as well as distance increased. This study implies that the 'push-pu l l ' theory which has been developed based on one or two migration streams and thus considered incomplete, is also va l id when the whole migration system is considered. This w i l l serve as an impor-tant contr ibut ion to the f i e l d of demography. This research i v is also expected to be helpful in population planning of Bangladesh. From the f indings of th is study, i t appears that the majority of the migrants acted as ' ra t iona l economic men' and went to the d i s t r i c t s where more (economic) oppor-tun i t i es were ava i lab le to improve the i r standard of l i v i n g . Thus to have a balanced regional d i s t r i b u t i o n of population growth, the best way l e f t perhaps is to increase economic opportunit ies in the rural areas by bringing more land under i r r i g a t i o n , providing f e r t i l i z e r and better seeds, e t c . , so that the majority of the migrants may always be ' p u l l e d ' to the rural areas rather than to the c i t i e s . The improvement in the general economic conditions of the rural area w i l l , at the same time, also lessen the poss i -b i l i t i e s of large scale exodus of people. v TABLE OF CONTENTS Page ABSTRACT i i LIST OF TABLES x LIST OF FIGURES x i i i LIST OF MAPS xv ACKNOWLEDGEMENTS xvi Chapter 1 INTRODUCTION. 1 1 .0 The Problem. 1 1.1 Objective and Strategy of this Study . . . . 5 2 THEORETICAL AND EMPIRICAL APPROACHES TO MIGRATION 10 2.0 Introduction 10 2.1 Migration as a Demographic Phenomenon 10 2.2 Major Approaches to the Study of Internal Migration 12 2.3 L imitat ions of the Ex ist ing Studies 27 2.4 Migration Studies on the Indian Sub-Continent 32 v i Chapter Page 2.5 Summary 4 0 2.6 The Working Hypotheses 41 3 THE STUDY AREA 44 3.0 Introduction 44 3.1 Geographical Location and Physical Sett ing 4 5 3.2 Economic Sett ing of Bangladesh 48 3.3 Demographic Aspects 54 3.4 L i teracy 6 0 3.5 H i s t o r i c a l Aspects of Migration in Bangladesh 6 1 4 THE SOURCE OF DATA AND THE METHODOLOGY AND RATIONALE FOR SELECTING THE VARIABLES . . . . 63 4.0 Introduction 63 4.1 Source, Nature and Accuracy of Data 63 4.2 Estimation of Migration Figures from the 'Place of B i r t h ' Data 66 4.3 Estimation of Rate of Internal Migration by Using Modified Gravity Model 6 8 4.4 C l a s s i f i c a t i o n of the D i s t r i c t s into Rural and Urban D i s t r i c t s and the Select ion of Migration Streams 70 4.5 Select ioneof the Explanatory Variables 75 4.6 Rationale for Select ing the Explanatory Variables 76 vi i Chapter Page 5 DETERMINANTS OF THE SPATIAL PATTERN OF MIGRATION WITHIN BANGLADESH, 1951 & 1961 90 5.0 Introduction 90 5.1 I d e n t i f i c a t i o n of the ' C r i t i c a l Factors ' 93 5.2 Ident i f i ca t ions and Def in i t ions of the 'Push' and ' P u l l ' Factors in the 'Factor Models' 107 5.3 Determinants of the Spatial Pattern of Rural-to-Rural Out-Migration Streams, 1951 110 5.4 Determinants of the Spatial Pattern of Rural-to-Rural In-Migration Streams, 1951 129 5.5 Determinants of the Spatial Pattern of Rural-to-Urban Out-Migration Streams, 1951 148 5.6 Determinants of the Spatial Pattern of Urban-to-Rural Out-Migration Streams, 1951 166 5.7 Determinants of the Spat ia l Pattern of Urban-to-Urban Out-Migration Streams, 1951 . 175 5.8 Determinants of the Spatial Pattern of Urban-to-Urban In-Migration Streams, 1951 184 5.9 Summary Results of A l l the Migration Streams and V a l i d i t y of Hypothesis 4 in 1 951 1 91 5.10 Determinants of the Spat ia l Pattern of Rural-to-Rural Out-Migration Streams, 1 961 . . 209 5.11 Determinants of the Spatial Pattern of Rural-to-Rural In-Migration Streams, 1961 229 v i i i Chapter Page 5.12 Determinants of the Spatial Pattern of Rural-to-Urban Out-Migration Streams, 1961. 247 5.13 Determinants of the Spatial Pattern of Urban-to-Rufcaili Out-Migration Streams, 1 961 258 5.14 Determinants of the Spatial Pattern of Urban-to-Urban Out-Migration Streams. . . 267 5.15 Determinants of the Spatial Pattern of Urban-to-Urban In-Migration Streams, 1 961 277 6 SUMMARY AND CONCLUSIONS 278 6.0 Summary . 278 6.1 CConeiliusdons . 282 6.2 Mmpl i.cati>ons 284 SELECTED BIBLIOGRAPHY 287 APPENDICES I Chronology of P o l i t i c a l Events 308 II ;Measurement of D.y: 'from the Geometric Centres of the D i s t r i c t s . . 310 III Kushtia D i s t r i c t , Ca lcu lat ion of E3 (An Example) 311 IV Regression Estimates of Two Migration Streams (Example) 316 V Rest of the Tables Prepared for Chapter 5 . . 319 ix LIST OF TABLES Table Page 3.1 Aspects of Land Tenure, Bangladesh 50 3.2 Population Density and Percentage Var iat ion of Growth of Populat ion, 1951-61 58 4.1 T h e o r e t i c a l l y Possible Number of Independent Migration Streams 73 4.2 Migration Streams Selected for the Present Analysis 74 4.3 VariabilesVSel ebted s f6r5Ana lys i s , . „ „ . . 1951 77 4.4 Variables Selected for A n a l y s i s , 1961 78 5.1 Factor Loadings on the Factors , Rural-to-Rural Out-Migration Streams Or ig inat ing from B a r i s a l , 1951 (An Example) 94 5.2 Corre la t ion of the Factors 95 5.3 The Expected Nature of Factor Loadings on the Variables 96 5.4 Determinants of the R-R Out-Migration Streams Or ig inat ing from B a r i s a l , 1951 I l l 5A Summary Results of R-R Out-Migration Streams, 1951 126 x Table Page 5.17 Determinants of the R-R In-Migration Streams Coming to Bogra, 1951 131 5B Summary Results of R-R In-Migration Streams, 1951 144 5.30 Determinants of the R-U Out-Migration Streams Or ig inat ing from Comi l la , 1951 153 5C Summary Results of R-U Out-Migration Streams, 1951 , 163 5.40 Determinants of the U-R Out-Migration Streams Or ig inat ing from Chittagong, 1951 167 5D Summary Results of U-R Out-Migration Streams , 1951 172 5.47 Determinants of the U-U Out-Migration Streams Or ig inat ing from Kushtia, 1951 179 5E Summary Results of U-U Out-Migration Streams, 1951 182 5.49 Determinants of the U-U In-Migration Streams Coming to Chittagong, 1951 185 5F Summary Results of U-U In=Migration Streams, 1951 189 5.56 Determinants of the R-R Out-Migration Streams Or ig inat ing from Far idpur, 1961 196 5G Summary Results of R-R Out-Migration Streams, 1961 ' . . 204 5.69 Determinants of R-R In-Migration Streams Coming to Jessore, 1961 215 5H Summary Results of R-R In-Migration Streams , 1961 224 5.84 Determinants of R-U Out-Migration Streams Or ig inat ing from Pabna, 1961 238 xi Summary Results of R-U Out-Migration Streams, 1961 Determinants of U-R Out-Migration Streams Or ig inat ing from Dacca, 1961 Summary Results of U-R Out-Migration Streams , 1 961 Determinants of U-U Out-Migration Streams s 0r ig inat ing from Khulna, 1961 Summary Results of U-U Out-Migration Streams,1961 Determinants of U-U In-Migration Streams Coming to Chittagong H.T. 1961 .' Summary Results of U-U Out-Migration Streams,1961 x i i LIST OF FIGURES Fi gure Page 1.1 The Arrangement of the Chapters . . 9 5.1 'Distance E l a s t i c i t y ' of R-R Out-Migration Streams, 1951 128 5.2 'Distance E l a s i t i c i t y ' of R-R In-Migration Streams, 1951 146 5.3 'Distance E l a s t i c i t y ' of R-U Out-Migration Streams, 1951 165 5.4 'Distance E l a s t i c i t y ' of U-R Out-Migration Streams, 1951 173 5.5 'Distance E l a s t i c i t y ' of U-U Out-Migration Streams, 1951 183 5.6 'Distance E l a s t i c i t y ' of U-U In-Migration Streams, 1951. 190 5.7 'Distance E l a s t i c i t y ' of R-R Out-Migration Streams, 1 961 206 5.8 'Distance E l a s t i c i t y ' of R-R In-Migration Streams, 1 961 226 5.9 'Distance E l a s t i c i t y ' of R-U Out-Migration Streams, 1 961 246 x i i i Figure Page 5.10 'Distance E l a s t i c i t y ' of U-R Out-Migration Streams, 1 961 256 5.11 'Distance E l a s t i c i t y ' of U-U Out-Migration Streams, 1 961 266 5.12 'Distance E l a s t i c i t y ' of U-U In-Migration Streams, 1 961 274 xi v LIST OF MAPS Map Page 3.1 Location of Bangladesh 46 3.2 Physical Sett ing of Bangladesh 47 4.,1> D i s t r i c t s and Four Major Regions 65 5.1 RsR Out-Migration Stream, 1951 (An Example) 99 5.2 R-R In-Migration Streams, 1951 (An Example) 100 5.3 R-U Out-Migration Streams, 1951 (An Example) 101 5.4 U-R Out-Migration Streams, 1951 (An Example) 102 5.5 U-U Out-Migration Streams, 1951 (An Exampl e) 103 5.6 U-U In-Migration Streams, 1951 (An Example) . . 104 xv ACKNOWLEDGEMENTS The study has l e f t the writer in debt to many persons. I would l i k e to take this opportunity to express my thanks and apprec iat ion: To: Doctors K.G. Denike and R. North, for guidance, ass istance and construct ive c r i t i c i s m , but most of a l l , for unfa i l ing f a i t h , encouragement and understanding; To: Dr. B.Morrison, for i n t e r e s t , encourage-ment, construct ive c r i t i c i s m , and for allowing me to use his personal copies of census materials on East Pakistan (now Bangladesh). To: Dr. K.S. Sandhu, for his guidance, advice and encouragement at the i n i t i a l period of preparation of th is t h e s i s ; To: Dr. G. Gates, for i n t e r e s t , and valuable suggestions; To: Prof . Sharom Ahmat and Dr. G. E l l i s t o n of the Univers i ty , Sains Malaysia, Penang, Malaysia, for understanding and help; To: Miss Sharon Hal ler for the f i n a l typed copy of the thes i s ; To: My father and wife , for moral support and understanding under try ing circumstances. xvi Chapter 1 INTRODUCTION 1 .0 The Problem A major fact of our era is rapid population growth. The world population growth rate rose from four per thousand per year , three centuries ago, to 10 per thousand per year between the F i r s t and Second World Wars (Hauser, 1964:16). It continued to accelerate a f ter World War I I , so that in 1963, by United Nations estimate, i t approximated 20 per thousand. The present world average rate is more or less the same, but there has been a further increase of 2-3 per thousand per year in some underdeveloped countr ies , p a r t i c u l a r l y in South and Southeast Asia (excluding Japan). Consequences for these countries have been severe. Indeed, the problems of rapid population growth and the problems of most of the countries in the Third World have of late become synonymous. In many cases, moreover, the s i tuat ion is made worse by an _ Rate of growth of population i s not the highest in South or Southeast A s i a . In some of the countries of Lat in America, the rate of growth is much higher than in South A s i a . But the s i tuat ion is more alarming in South Asia as the total population there is more than three times than Lat in America. 1 2 uneven d i s t r i b u t i o n of population growth. The countries in the Third World are character ized by a wide d i f ference between rural and urban regions in terms of economic condit ions (Bogue and Zachariah, 1962; Alonso, 1969). Though the majority of the poeple l i v e in rural areas, the l i v i n g condit ions are gett ing worst as the r i g i d structure of property re la t ions in land unresponsive to the unprecedented increase of popu-l a t i o n i s increasing landlessness and rural poverty (Gunatil;i ke, 1972/73). The expansion of population is also exerting new pressure on an a g r i c u l t u r a l sector which is s t i l l at a low level of technology thereby creat ing new st imul i for migrat ion. On the otherrhanxl., the l i o n ' s share of the benef i ts of economic development, whatever has been achieved so f a r , has gone to the urban areas. This has only helped to widen the already ex is t ing i n e q u a l i t i e s . As a r e s u l t most of the developing countries are experiencing pers is tent and increasing streams of migrants to urban areas in search of * work and better l i v i n g condit ions (Hauser, 1957:9). In Ind ia , the largest developing country in terms of populat ion, i t appears that , from 1951 onwards, the rate of urban growth is slowing down and in most of the larger c i t i e s (having more than a m i l l i o n people) migration from the rural areas is p i t t e r i n g of f (Brush, 1970; Taneja, 1971). But the s l i g h t decrease in the rate of growth of urban popu-la t ion has not reduced the problem in any way as the addit ion of 16.2 m i l l i o n people (and most of them are s t i l l from the rural regions) in one decade, 1951-61, is anything but a manageable f i g u r e . Moreover the rural-urban migration remains as the most important factor in swell ing up the population of the medium class c i t i e s ( i . e . below 100,000 population) [Taneja, 1971:18], and the 1971 census revealed that urban i n -migration has accelerated again during the past decade. 3 Migration has reached such proportions in some of the developing countr ies , p a r t i c u l a r l y in South A s i a , that the c i t i e s can no longer cope with the i n f l u x . People are f l ee ing from ' rura l poverty' only to encounter 'urban misery' and the l i v i n g conditions of both c i t y and country are deteriorating. Also over burdening the economic i n f r a s t r u c t u r e of the urban areas holds back economic growth, s ince the c i t i e s contain most of the modern and dynamic sectors of the economy. Very recent ly , in countries such as Mexico (Barkin, 1972) and India (White, 1973), the respect ive governments have been try ing desperately to stem the flow of people from rural areas by creating new growth centres. They have had l i t t l e success, which is not s u r p r i s i n g . Measure l i k e these can of fer no guarantee of solv ing the problem since the process of migration i t s e l f is yet to be f u l l y understood in the developing world context. Our understanding of the factors inf luenc ing migration in underdeveloped countries is in fac t very l i m i t e d , for two reasons: F i r s t l y , scholars and government agencies, r e a l i s i n g the grave consequences of population growth out-s t r ipp ing economic growth, have proposed and implemented innumerable remedial measures, but they have paid comparatively l i t t l e attent ion to the spat ia l d i s t r i b u t i o n and r e d i s t r i b u -t ion of population growth. As a r e s u l t , among the demographic components of population growth and change, the least is 4 known about migrat ion. Secondly, up to the present, attent ion has been confined mainly to two types of migrat ion, namely, (1) Rural-to-Urban Migration and, (2) Urban-to-Urban Migrat ion. Attempts have been made to answer two basic question regarding these two streams of migrat ion, namely, (1) why do people move, and (2) how far do they move. From these attempts several theories have been developed, a l l based on developed world data. Out of these theories only 'push-pu l l ' theory appears to be concerned with the migration process p e r se and does not suffer from most of the l i m i t a t i o n s from which other theories suf fer (the l i m i -tat ions of migration theor ies , including 'push-pu l l ' theory, are discussed in deta i l in the next chapter) . Thus the present study has been conducted in the 'push-pu l l ' framework. We shal l out l ine the theme of the theory in the next paragraph. People might leave one region and go to another for two reasons: f i r s t l y to look for an opportunity to improve the i r standard of l i v i n g , and secondly to avoid adverse economic or soc ia l cond i t ions . In the f i r s t case, the place of dest inat ion exerts a ' p u l l ' on the migrant and in the second case he is . 'pushed' away from the place of o r i g i n . This 'push-pu l l ' theory had been o r i g i n a l l y enunciated by Ravenstn'ien (1 855 and 1889). Recently i t has been modified ' by Lee (1966). It i s , however, incor rect 5 to assume that the flow of people is simply due to the 'push' and the ' p u l l ' f a c t o r s . As Lee (1966) emphasized, there are intervening obstacles to migrat ion. The most important and most studied of these var iables is d istance. The i m p l i c i t assumption is that the ind iv idual , an economic rat iona l man, w i l l t ry to minimize the costs of migration and thus w i l l also try to minimize the distance t r a v e l l e d . In the r e l a t i v e l y few studies carr ied out in the developing countr ies , the general tendency has been simply to apply the theories which have been developed*in the developed-world .context. We may well question the log ic of assuming that the factors which lead to migration in two e n t i r e l y d i f f e r e n t s o c i a l , cu l tura l and economic sett ings are much the same. At the very l e a s t , there is a need to test the v a l i d i t y of such c ross -cu l tu ra l a p p l i c a t i o n s . The purpose of th is study is to meet th is need. 1 .1 Objectives and Strategy of this Study The present study seeks to analyze the streams of internal migration in Bangladesh for the years 1951 and 1961. The two basic object ives of this study are, (1) to e luc idate the spat ia l pattern of migration and to ident i f y and analyse the determinants of such movements, and (2) to d iscover , through pr inc ipa l axis type factor a n a l y s i s , the extent to which the 'push-pu l l ' theory, or at least the p r i n c i p l e s 6 involved in that theory, are appl icable to Bangladesh.* Instead of dealing with only one or two types of migration streams, as have most s tud ies , th is study recognizes s ix mutually independent types of stream (Tables 4.1 and 4.2) and attempts to i s o l a t e the i r p a r t i c u l a r 'mix' of 'push' and ' p u l l ' factors and d i f f e r e n t i a l response to the intervening distance as w e l l . Even by South Asian standards Bangladesh i s over-whelmingly r u r a l . Only 5.19% of the tota l population was urban in 1961 compared to about 20% in India and Pakistan. It may also be noted that more than 90% of the gross migration took place within the rural reg ion. Apparently most of the migrants were going to other rural areas since c i t i e s or towns were less a t t r a c t i v e . Thus, i t is hoped that th is study w i l l not only improve our understanding of the app l i c -a b i l i t y of the 'push-pu l l ' theory in a developing country, but w i l l also give an ins ight into a s i tuat ion where r u r a l -urban migration has been a small proportion of the whole. This may help in population planning of Bangladesh and other developing countries in order to achieve a better balanced d i s t r i b u t i o n of population growth between the urban and rural regions and also in the regional d i s t r i b u t i o n of population growth. It is understood that the lack of d i rec t data on migration may have been one of the main reasons for the paucity of studies in developing countr ies , but there has _ See Table 5.3 to understand the c r i t e r i a of i d e n t i -fying the determinants of migration ( v e l o c i t y ) . 7 also been a tendency to assume that no proof is needed before t rans fer r ing resu l ts obtained from developed world studies to s i tuat ions in the developing world. Direct data on migration are not ava i lab le for Bangladesh a l s o , but we shal l try to show that studies of th is nature can be carr ied out with reasonable accuracy. Natural ly the f indings must be considered highly tenta t ive . This research in no way pretends to be the l a s t word on the subject , rather i t opens the way for further research. A major advantage of th is approach is that i t examines the process of migration in i t s t o t a l i t y , taking a l l the poss ib le types of mutually independent migration streams. It provides a conceptual framework for conducting migration studies which should enable us e i ther to amend the 'push-pu l l ' theory or to disprove i t a l together . Agreements or disagreement on the a p p l i c a b i l i t y of a certa in theory c r o s s - c u l t u r a l l y may be best expressed as a ser ies of hypotheses sett ing out points of agreement or d i f ference and a l te rnat ive i n t e r p r e t a t i o n s . The Hypotheses can then be empir i ca l ly tested and ex is t ing conclusions and theory modified in the l i g h t of resu l ts obtained. However, before such hypotheses can be formulated, the status and trend of ava i lab le research concerning internal migration should be more thoroughly d iscussed. To th is end, a review of l i t e r a t u r e on the a l te rnat ive approaches 8 to migrat ion, both theoret ica l and empi r i ca l , is presented in the next chapter. The arrangement of the chapters is presented in Figure 1.1. 9 The Problem, Evaluation of the 'Push-Purpose and P u l l ' Theory in Relat ion Strategy of v to the Other Approaches th is Study ^ to Migrat ion, Chapter 2 Chapter 1 ; A Generation of the Working Hypotheses Based on the 'Push-P u l l ' Theory, Chapter 2 The Study Area, Accuracy and Nature of Data. Rationale for Select ing the Var iab les , Chapters 3 and 4 Appl icat ion of Factor Analysis Chapter 5 Conclusions and Imp!i cat ions , Chapter 6 Results from A n a l y s i s , Synthesis , V a l i d i t y of the Working Hypotheses Chapter 5 Figure 1.1 The Arrangement of the Chapters in the Present Study. Chapter 2 THEORETICAL AND EMPIRICAL APPROACHES TO MIGRATION 2.0 Introduction The primary object of this chapter is to define migration as a demographic phenomenon and to c r i t i c a l l y review the a l te rnat ive approaches to migrat ion. The review covers the works of Ravenstein, Lee, Gallaway, Sjaastad, Schul tz , S touf fer , Z i p f , Firey, Schwind and Morri l l . It is hoped that the l i t e r a t u r e survey w i l l help to evaluate the strength and weaknesses of the 'push-pu l l ' theory so that we can j u s t i f y concentrating on the 'push-pu l l ' approach in the present study. After the review the a priori hypotheses are generated based on the 'push-pu l l ' theory and on previous works which attempt an empirical explanation of migration streams using the theory. 2.1 Migration as a Demographic Phenomenon Migrat ion, as defined by demographers, is a response of the human organism to economic, cu l tura l and soc ia l forces in the environment (Bogue, 1959). According to him " . . . some 10 11 of the most acute soc ia l problems are associated with migra-t ion" (Bogue, 1969:752). The same author maintains that i f the problems of human f e r t i l i t y were not so c r i t i c a l at the present time, then human migrat ion, mainly internal migrat ion, and the p l ight of migrants (espec ia l ly in less developed countr ies) would be l i s t e d as a top p r i o r i t y for research and ac t ion . Migrat ion, however, is one of the most complex com-ponents of population change and growth. The other components such as b i r ths and deaths are phenomena that involve physio logic as well as socio-economic processes. Migrat ion, on the other hand, has no physio logic component, and hence patterns of migration are not so gradual and orderly as those of f e r t i l i t y and morta l i ty . As such, determinants of migrat ion, s p e c i a l l y internal migrat ion, are harder to i so la te and analyze. 2.1.1 Internal migration as a demographic process Internal migration,, that i s , permanent or semi-permanent movement within the bounds of one p o l i t i c a l un i t , is the process by which a given population is r e - d i s t r i b u t e d (Wilson, 1968:110). The study of migration is more i n t r i c a t e within a country than among countries since i t is not merely concerned with one d e f i n i t e stream from one country to another but with whole ser ies of streams such as: (1) urban-urban migration streams, (2) urban to rural migration streams, 12 (3) rural to rural migration streams and (4) rural to urban migration streams. These streams may be either completely independent or d i r e c t l y or i n d i r e c t l y re lated with each other. They may be again c l a s s i f i e d into i n- e.aridoout-mi grat i on streams. In-migration is a process by which a region gets people from other regions arid by out-migration a given region loses i t s populat ion. In addit ion there are two other kinds of migrat ion, namely (a) net-migration and (b) gross-migration which do not represent the flow of people. While net-migration indicates whether a given region is a net gainer or loser of population due to the resu l t of i n - and out-migrat ion, gross-migration indicates the total volume of migrants coming in and going out from a given region. 2.2 Major Approaches to the Study of Internal Migration Students interested in internal migration approach the phenomenon in several d i f fe rent ways. Migration has been lookeddat both as, (1) an independent system such as in (a) the 'push-pu l l ' theory or (b) the 'grav i ty model, 1 or (2) as a part of some economic theories such as (a) the theory of consumer demand or (b) the general investment theory (more s p e c i f i c a l l y , the theory of human capi ta l formation). Most geographers' works, however, have been empirical in nature. 13 A complete recap i tu la t ion of a l l relevant l i t e r a -ture is not attempted here. The works that are discussed are f e l t to be d e f i n i t e l y representative of d i f fe rent approaches to the problems. Migration studies on the Indian Sub-Continent are discussed separately since the present study is based on one of the countries of the sub-continent. 2.2.1 Ravenstein and Lee - 'Push-Pul l ' approach Ravenstein (1885 and 1889) was the f i r s t person to enunciate the 'push-pu l l ' re la t ionsh ip in migration though he did not present the theory in a formal way and the condit ions of the ' push-pul1 ' re la t ionsh ip were not l a i d down e x p l i c i t l y . But he recognized the essent ia l part of the hypothesis that ind iv idua ls may be provoked to migrate by pos i t ive (pul l ) and negative (push) forces . People may migrate to improve the i r l o t in l i f e . In th is case the place of dest inat ion exerts a ' p u l l ' on the migrant. People may also migrate to escape aft undesired soc ia l or economic s i t u a t i o n in t h e i r place of residence. These s i tuat ions const i tute an expulsive 'push' from the place of o r i g i n of migrat ion. The process, however, is quite a complicated one since in each case of migrat ion, ind iv idual or c o l l e c t i v e , several var iables of both types may be operating and i n t e r a c t i n g , so that the move cannot be att r ibuted wholly e i ther to 'push' or ' p u l l ' factors 14 alone. A p a r t i c u l a r 'mix' of push and pul l factors may be operat ing. But the p a r t i c u l a r 'mix' of push and pul l factors that leads to migration in one person or a group of persons may be very d i f fe rent from that which leads to migration of another. Bogue l i s t s as many as 50 socio-economic and demo-graphic factors which may inf luence ind iv idua ls e i ther in choosing a dest inat ion or leaving a place or both (Duncan and Hauser eds . , 1959:499-500). Ravenstein's work, however, was bereft of these points . These have been added much la ter by soc io log i s t s l i k e Thomas (1957,1960,1964) and Bogue (1959). Lee (1966) presented the hypothesis in a more formal way and added the notions of intervening obstac les . According to his model migration is a resul tant of 'pushes' or ' p u l l s ' or ' a t t r a c t i o n s ' or ' repu ls ions ' at both the place of o r i g i n and the place of des t inat ion . It is balanced in the context of r e l a t i v e e f f o r t or cost of overcoming the obstac les . Among the intervening factors he recognized distance separating the place of o r ig in and the place of dest inat ion as the most important one. The i m p l i c i t assumption in his model is that a migrant, an economic rat ional man, w i l l t ry to minimize the costs of overcoming the intervening obstac les , whatever they are and however they are measured. Thus, in summary, we may state that the 'push-pu l l ' theory is an abstract ion which is made in order to c l a s s i f y the s p e c i f i c forces or factors that may be at work in a given 15 migration system, and, as such, the problem of understanding any given migration system may become the problem of determining these forces at work. One of the ways by which these factors of migration may be determined is by examining a large number of persons and by estab l i sh ing the common pos i t ive or negative stimulants to movement. These common stimulants may be regarded as independent or explanatory var iables whose corre lates with migration (rate , volume, rat io or ve loc i ty) tend to account for the pattern or behaviour of migrat ion, which is the de-pendent var iab le . Thus, the problem of ident i fy ing factors underlying migration becomes one of d iscover ing , (a) whether or not a s i g n i f i c a n t s t a t i s t i c a l re la t ionsh ip exists between a dependent var iable (migration) and one or more independent var iab les , (b) i f so, what the d i rect ion and nature of the associat ion i s , and (c) whether or not there is a theoret ica l j u s t i f i c a t i o n , in the context of the 'push-pu l l ' theory for the observed r e l a t i o n s h i p . 2.2.2 Gallaway - Economic Theory of Labour Mobi l i ty Approach The formal economic theory of labour mobi l i ty may be perceived as a special case of consumer demand (Gallaway et al. , 1967). A worker's choice of a place of dest inat ion may, e s s e n t i a l l y , depend on the varying amounts of work-related 16 income that may be obtained in d i f f e r e n t l oca t ions . Let us consider two geographic labour markets. A worker, t ry ing to maximize his benefits from movement, would e lect to o f fer his labour in the market with the higher wage ra te , provided there are no unique costs associated with employment in both the regions and no costs of movement or barr iers between them. The decis ion making process, however^ would be more complex i f there are employment or movement costs . For example, i f the distance between the two regions is great the costs of moving (both money and non-money costs) would be important f a c t o r s . Money costs include transportat ion costs and non-money costs consist of , (a) income foregone during the period spent on t r a v e l l i n g , (b) the time taken to search f o r , and learn a new job. (Sjaastad, as we shal l see l a t e r , c a l l s them opportunity costs) etc . Such costs (both money and non-money) make an area less ' a t t r a c t i v e ' as an employment p o s s i b i l i t y and, consequently, when ind iv idua ls ( in this case only workers) evaluate various possible regional employment opportunit ies they discount the offered wage rate by whatever amount is necessary to compensate for these costs . Let us assume'" that a worker is already i n , say, i , and considering migrating to JAWhe re the wage rate is higher ( i . e . V. > W-..). Then the ' a t t r a c t i v e n e s s ' of j may be expressed mathematically as ' fo l lows: 17 Aj - (Wj. - W,) - C i d 2.1 where A. is the attract iveness of a region j W. is the actual (offered) wage rate at j W. is the current wage rate at i and C . are the costs (money and non-money) 1 J incurred in moving from i to j Assuming that D ^ - J i - e . distance separating i and j is great . In other words, the equation 2.1 is v a l i d only at the macro l e v e l . Thus, ceteris paribus, the greater the distance between the regions the less the value of the wage rate d i f f e r e n t i a l and the less a t t r a c t i v e the higher rate regions becomes to a worker considering migrating to i t . The r e l a -t i o n s h i p , in regression model;, may be expressed as fo l lows: n f J - f[(Hj - « , ) . D , ~ ] 2.2 where M.. is the number of migrant labour in J j from i . The expected signs of the p a r t i a l der ivat ions are: <'•': " »i> > 0 • M i / D i j < 0 2.2.3 Sjaastad and Schultz - Migration as a Human Capital Decision Approach Sjaastad and Schul tz , belonging to the so-ca l led Chicago School of Economics, postulate that migration flows are d i r e c t l y re lated to net present values of the investment 18 dec is ions . The economic aspects of migration decis ion are viewed, (a) as a part of general investment theory, and more s p e c i f i c a l l y (b) as a part of the theory of human capi ta l formation. Schultz stresses that the value of th is p a r t i c u l a r type of human capita l formation " . . . can be determined by discounting the addit ional future earnings i t y ie lds just as the value of a physical cap i ta l good can be determined by d i s -counting i t s income stream (Schultz , 1962:8). Returns from migration between regions, say, i and j , consist of a stream of expected income d i f f e r e n t i a l s between the twofareas. In order to r e a l i z e these returns , the migrants must invest in moving to the regions of higher income. constant over time and allow those earnings to prevai l over an i n d e f i n i t e per iod, the net present value (NPV) of a migra-t ion investment for the average migrant may be expressed mathematically as fo l lows: If we assume that earnings in the two regions are NPV = " E i J / D R i 2.3 where are the average annual earnings in i and j and where E. > E. is the cost (money and non-money) of moving from i to j 19 R. is the rate of discount applied to 1 future rece ipt in i and i & j are the regions of o r i g i n and dest inat ion respect ive ly Since most of the costs , both money-costs and non-money cost (or opportunity costs as Sjaastad c a l l s them), increase with d istance, i t may be postulated that c i j = f ( D u ' 2 We may expect ~ the flow of migrants from i to j (or more p r e c i s e l y , of migrants moving for economic reasons) to be p o s i t i v e l y re lated to NPV. Using -2.3 and 2.4, the flow of migration motivated by economic reasons may be stated as the fo l lowing: M . . = f i J or in regressionimodel Y M = a 0 + ai M i j / n R i ° « D 1 J 2.5 20 where Y M is the estimate of migration flow i j between i and j and oo, ai and a 2 are constants. Recently, Todaro (1969) has emphasized that the probab i l i t y of obtaining employment inf luences the decis ion to migrate. His argument" i's that the earning var iab le should be an expected value. It would be determined by both earnings and the p r o b a b i l i t y of gett ing a job . Following Todaro we may define the p robab i l i t y of obtaining employment at time ' t ' as the ra t io of employment growth to the pool of unemployed. Symboli c a l l y 2.6 where P . J is the probab i l i t y of obtaining employment in region j E . J is the employment in j , and L . J is the labour force in j Now 2.7 may be expressed as Y M . = a o + «i A E j / L j - E. J 2.7 assuming again the wage rate in j is greater than the wage rate in i , i . e . (W. > W^ .) and D.. is great. U • 21 2.2.4 Stouffer - Theory of Intervening Opportunities Approach Stouf fe r ' s main contention was that migration analyses which are based on 'push-pu l l ' theory would not be f r u i t f u l unt i l the distance component is conceptual ly and empir i ca l ly i so la ted (Stouf fer , 1941). He also observed that , in the extensive l i t e r a t u r e which demonstrated a close re la t ionsh ip between mobi l i ty and distance (Thomas, 1938; Robinson, 1938), there had been l i t t l e e f f o r t to analyze the ways in which distance operated to determine the d i s t r i b u t i o n of population movements. The theory that he proposed assumed that there was no necessary re la t ionsh ip between mobi l i ty and d is tance. Instead, he introduced the concept of ' intervening opportun i ty . ' According to his postulates the number of persons going a given distance would be d i r e c t l y proportional to the number of opportunit ies at that distance and inversely proportional to the number of intervening opportun i t ies . Another way of s tat ing the same hypothesis would be to say that the number of persons going a given distance would be d i r e c t l y propor-t ional to the per centage of opportunit ies at that distance (Stouf fer , 1941:846). Symbolical ly the theory had been expressed as fo l lows: Ay/As = aAx/xAs 2.8 22 where Ay is the number of persons moving from an o r i g i n to a c i r c u l a r band of width As, i t s inner boundary being s-l/2As units of distance from the centre of the c i r c l e and i t s outer boundary being s + l/2As units from the o r i g i n . [Distance may be measured in units of space, or even of time or cost (Stouf fer , 1941:846)] x is the number of intervening opportun i t ies , that i s , the cumulated number of oppor-t u n i t i e s between the o r i g i n and distance s IncJ a is a constant In other words, the theory was a 'bas ic concept in handling movement and distance expressed in terms of a rat io of opportunit ies in the region of dest inat ion to the i n t e r -vening opportun i t ies . It can, however, be noted that equation 2.8 does not speci fy a d i rect r e l a t i o n between mobi l i ty and d istance. "Rather i t postulates a d i rec t r e l a t i o n between mobi l i ty and opportun i t ies . The re la t ion between mobi l i ty and distance may be said to be on an a u x i l i a r y r e l a t i o n s h i p , which expresses the cumulated ( intervening) opportunit ies as a funct ion of distance" (Stouf fer , 1941:847). He also mentioned that i t was not necessary to assume that i t was a continuous funct ion , s ince , in a c t u a l i t y , the d i s t r i b u t i o n of opportunit ies over space varies from place to place. If i t i s , however, assumed to be continuous, or 23 x - f ( s ) 2.9 where x is the number of opportunit ies and s is the distance by subst i tut ing d i f f e r e n t i a l s in equation 2.8 we get dy/ds = a/x • dx/ds 2.10 again by subst i tut ing 2.9 in 2.10 and i n t e g r a t i n g , we get y = a log f ( s ) + c 2.11 where y is the cumulated number of movers between the o r ig in and a c i r c l e of radius s, f ( s ) is the cumulated number of oppor-t u n i t i e s within that c i r c l e , and c i s a constant (x/s) . Thus equation 2.11 says that the total number of movers who stop at any point within the c i r c l e is d i r e c t l y proportional to the logarithm of the number of opportunit ies within the c i r c l e (Stouf fer , 1941:848). 24 2.2.5 Z ipf - PiP 2/D Hypothesis Approach As opposed to Stouf fer , who proposed that there is no necessary re la t ionsh ip between mobi l i ty and d is tance , Z ipf hypothesized that ". . . the number of persons that move between any two communities, in the United States , whose respect ive populations are Pi and P 2 and which are separated by the shortest transportat ion d istance, D, w i l l be propor-t ionate to the r a t i o , P iP 2 /D, subject to the e f fec t of mod-i fy ing factors" (Z ipf , 1946:677). In other words, the i n t e r -community movement of persons (and also of goods by value) between any two communities, Pi and P 2 , that are separated by a- transportat ion d istance, D, w i l l be d i r e c t l y proportionate to the product, Pi x P 2 , and inverse ly proportionate to the d is tance , D, c e t e r i s paribus. One of the main assumptions has been that the re la t ionsh ip holds good only when employ-ment and income are uniformly d i s t r ibuted over the areas (which of course happens rare ly in real world s i t u a t i o n s ) . 2.2.6 Firey and Schwind - The Ecological Approach In this approach migration is perceived as a process by which the population of a reg ion, general ly of a c i t y , adjusts to or modifies i t s physical environment. According to Firey the average c i t y resident tends to f i l t e r , according to socio-economic successes, into better r e s i d e n t i a l d i s t r i c t s 25 (F i rey , 1947). Most of the la ter studies using th is approach to analyse North American c i t i e s have been found to be con-cerned with (a) ethnic groups and in p a r t i c u l a r the i r spat ia l d i s t r i b u t i o n and related economic a c t i v i t i e s , (b) the i r (ethnic group's) soc ia l condi t ions , or factors inherent in the i r adjustment to sub-areas and, (c) how the d i f fe rent ethnic groups in teract with one another and with the population of the c i t y as a whole (Mcentire, 1960). In some of the studies mul t ivar ia te s t a t i s t i c a l techniques such as factor analys is are used to e l i c i t the ' f a c t o r i a l ecology' of r e s i d e n t i a l locat ion (Murdie, 1967), while others also incorporate h i s t o r i c a l - c u l t u r a l approaches (Cho, 1970). Recently, Schwind, a geographer, has made an attempt to deal with migration patterns and the i r e f fect on regional development taking an ecological approach at' the macro-scale (Schwind, 1971). Through factor analys is and regression techniques he i d e n t i f i e d components of regional development and regional attract iveness in the United States . The indices of regional at t ract iveness i d e n t i f i e d by f a c t o r i a l ecology are used in a modified grav i ty model to estimate the volume and d i rec t ion of migration f lows. Though the macroscopic nature of the analysis prohib i ts the s p e c i f i c i d e n t i f i c a t i o n of the mechanisms that determine migrational flows and modify regional c h a r a c t e r i s t i c s , Schwind's work is a good example of the way in which an ecologica l approach may further the understanding of in ter - reg iona l migrat ion. 26 2.2.7 M o r r i l l - A Behavioural Approach In th is type of approach students belonging to the behavioural school of thought suggest that , as in terac t ion very often is preceded by the establishment of personal con-t a c t s , each would-be migrant tends to behave according to a migration p r o b a b i l i t y f i e l d and chooses his place of dest ina-t ion in a random order ( M o r r i l l , 1965b). The probab i l i t y f i e l d may be derived e i ther from observation of migration distance or through methods such as Monte Carlo and Markov Chain sim-u l a t i o n . As for example, rather than determining that a s p e c i f i c migrant moves from, say, one p a r t i c u l a r house to another in a given c i t y , the probab i l i t y of a typ ica l migrant's move from a house to another, within the same block or other b locks, is determined, and random numbers are used to decide which d e s t i n a t i o n , among the many poss ib le , he chooses. Thus a spat ia l pattern of moves is obtained which spreads settlements into new blocks and i n t e n s i f i e s i t in the old block. This may be described as a ' spat ia l d i f f u s i o n process' in which migrants of any occupation or ethnic group gradual ly penetrate the surrounding area at the micro-scale ( M o r r i l l , 1965b). At the meso or macro-scale, migrat ion, as Brown has shown, may be perceived as a type of d i f f u s i o n process, in which some members of a population at time ' t ' change the i r locat ions from time ' t ' to time ' t + 1' (Brown and Moore, 1967). 27 In studies of d i f f u s i o n of innovat ions, such as those of Hagerstrand (1967), i t has been rea l i zed that one of the usual impediments to the spread of the item in question is d i f f e r e n t i a l res i s tance , of e i ther a psychological or an economic nature, to adoption. This aspect, however, has received r e l a t i v e l y l i t t l e attent ion in geographical research (Brown and Moore, 1969). The more usual approach has been to assume that res istance is normally or randomly d i s t r ibuted through a population and hence that an id iographic approach is more than appropriate to describe and ' exp la in ' the spat ia l pattern of migrat ion. 2.3 L imitat ions of the Ex ist ing Studies From the above discussions i t may be observed that current studies on migration have rather cons is tent ly been s e l f - 1 i m i t i n g in terms of the emphasis given on the forces or factors of migrat ion. For example, in the 'economic theories of migrat ion' such as 'the economic theory of labour mob i l i t y ' and 'the theory of human capi ta l investment, ' i t has been assumed that a s p e c i f i c migrant being a ' ra t iona l economic man1 is a maximize^ of economic benefits and not a s t a i s f i e r and as such w i l l always tend to migrate where the p robab i l i t y of gett ing higher income is greater than his place of residence. But resu l ts from studies have often been contrad ic tory . While 28 the studies carr ied out by Lorgren (1956), Nelson (1959), Becker (1962), Speare (1971) and Trott (1972) substantiated the theme of the 'economic theories of m i g r a t i o n , ' Raimon (1962) and Schwind (1972) found that the economic var iab les do not adequately account for the migration f lows. This shows that the determinents of migration are too complex to be 'expla ined' only by economic var iables and also shows that the ' ra t iona l economic man' does not always ex ist in real world si tuat ions . Perhaps people's response to the economic factors also depends on the type of stream of migration to which they belong, e . g . r u r a l - r u r a l , rura l -urban, e tc . If that is so, the contradictory resu l ts are not surpr is ing because neither in the 'economic theories of migrat ion ' nor in the empirical studies which employed them, there has been an attempt to understand the d i f f e r e n t i a l response of migrants to economic var iables in d i f f e r e n t streams. In f a c t , as we shal l see l a t e r , this appl ies to a l l the theories of migration that have been developed so f a r . The other approaches also suf fer from too much emphasis on a p a r t i c u l a r set of v a r i a b l e s . S touf fe r ' s theory of intervening opportunit ies and Z i p f ' s P1P.2/D hypothesis (or the grav i ty model), for example, are s a t i s f i e d with deriv ing a function re la t ing migration and d istance. It is understood that as any migration invar iab ly involves a certa in d istance, i t may be considered as one of the most 29 important var iab les , however that may be measured, but i t cannot necessar i ly be considered as the only explanatory v a r i a b l e . Again the 'distance e l a s t i c i t y ' may vary from one type of migration stream to another, which has of course not been e x p l i c i t l y defined in these theor ies . In the ecological approach, migration is general ly considered as an independent (explanatory) var iable rather than a dependent var iab le . Thus primary importance has not been given to understand ., the factors of migrat ion. But th is approach is better of f than the 'economic theor ies ' or 'grav i ty model' in that i t has the scope to analyze migration as an independent system - incorporating a l l the possible types of streams and analyzing the i r d i f f e r e n t i a l response to a l l types of var iable including the economic one (Schwind, 1972). If the 'economic theor ies ' suffer from attaching too much importance to economic var iables in explaining migration pattern, 'the behavioural approach' also does not suffer less from attaching too much importance to non-rational var iab les . Non-economic var iables l i k e personal contacts may be impor-tant in some instances, but i t is again a very narrow view to perceive that people migrate due to th is factor alone. Another possible weakness of the theories l i e s in the fact that they may not be applied with the same e f f e c t i v e -ness at the two scales — micro and macro. Let us, for example, consider the case of the 'grav i ty model' i . e . Z i p f ' s PiP 2/D — 'Distance e l a s t i c i t y ' of migration may be defined as the rate of change of migration per unit change of d istance. 30 hypothesis. Here emphasis is given to the distance fac tor . It has been observed that distance tends to exert less i n f l u -ence in determining dest inat ions of migrants at the micro-level than at the macro- level . If the migration pattern within a c i t y , for example, were independent of distance (as there is almost a complete lack of ' f r i c t i o n of d i s t a n c e ' ) , the denominator of a gravity model, would tend to approach zero, thereby rendering the formula meaningless as the resu l t would tend to be For this reason Z i p f ' s hypothesis or gravity model may not be appropriate for descr ibing intra-urban migration and is general ly avoided. We may general ize about the scale a p p l i c a b i l i t y of certa in types of approaches. Thomas (1941) and Isaac (1947) observed that migra-t ion at the inter- reg iona l level is considerably more sens i -t ive to economic factors than migration at the intra-urban l e v e l . On the other hand behavioural (probab i l i ty ) models are much more appropriate at the micro-scale than at the macro-scale, and ecological studies are also found to be more appropriate in loca l migration studies than in regional level s tudies . A l l th is suggests that the importance of non-rational or at least non-economic f a c t o r s , such as personal contact, is greater at the micro-sca le . Concerning the ecological studies i t might also be noted that the greater the d istance, the less l i k e l y the migrant would be aware of the wide range of factors usual ly measured by the f a c t o r i a l eco log i s t s . 31 It appears that only the 'push-pu l l ' theory does not suffer from most of the l i m i t a t i o n s , from which the other theories s u f f e r . It is the only ' theory' which has not been developed as a subset of a general theory and does not attach undue importance to one set of va r iab les . The concept of ' ra t iona l economic man' is i m p l i c i t in Lee's formulat ion, but there is no s p e c i f i c statement that only the economic var iables can be considered as e i ther 'push' or ' p u l l ' factors of migra-t i o n . However, in th is theory a l s o , not a l l the possible independent migration streams have been considered, and there is no mention whether or not migrant's response to the explanatory var iables w i l l vary depending on which type of stream they belong to . But the structure is open enough to incorporate t h i s , i f anyone wants to do so. This theory may be considered as ' incomplete' since i t is f e l t that the study of one or two streams of migration may be useful and s u f f i c i e n t i f migration streams are employed as independent (or explanatory) var iables to understand i t s e f fect on phenomena l i k e urbanizat ion, economic development, popula-t ion r e - d i s t r i b u t i o n or soc ia l change. But i t is perhaps not s u f f i c i e n t i f the object is to understand the process of internal migration per se and analyze the factors or forces behind such movements. The factors regulat ing d i f fe rent types of migration streams are expected to be general ly d i f fe rent and the streams may respond d i f f e r e n t i a l l y to various f a c t o r s , whether they are 'push' or ' p u l l ' or combinations of both. 32 2.4 Migration Studies on the Indian Sub-Continent 2.4.1 The B r i t i s h India Studies about internal migration of B r i t i s h India have been done by Davis (1951) and Zachariah (1964). Both of these studies were concerned with the period before 1931. Davis did an important analysis but he described and discussed the problems in very broad terms. His general conclusion was that the population of the Indian-Sub-Continent is compara-t i v e l y immobile. He was the f i r s t author to use the 'place of b i r t h ' data as the source of ca l cu la t ing migration f igures for th is region and opined that they were reasonably correct est imat ion. No other study of the h i s t o r i c a l aspects of internal migration of India was published unt i l 1964, when Zachariah completed his book "A H i s t o r i c a l Study of Internal Migration in the Indian Sub-Continent, 1 901 -1 931 ." His study was much more deta i led than that of Davis. The primary object of that study was to measure and describe the internal migration of that sub-continent during the period 1901-1931. Migration estimates by age and sex had been prepared for regions, s ta tes , and to a l imi ted extent for towns with populationscof 20,000 or more, from i n d i r e c t sources such as 'place of b i r th data' and 'age-sex d a t a . ' These estimates were then used to describe and indicate the areas of popula-t ion gain and l o s s , to determine the d i rect ion and magnitude 33 of population migration streams, and to analyze the demographic c h a r a c t e r i s t i c s of net migrat ion. The main drawback of these studies seems to be the se lect ion of states as the spat ia l un i t s . They were not only too large in s ize but also d i f fe red appreciably from one another in terms of populat ion. For example, Burmah was at least ten times larger in s ize than Kerala and Uttar Pradesh had six times more population than Assam. These wide d i f -ferences, in terms of population and area, among the states d e f i n i t e l y reduced the accuracy of the a n a l y s i s , for i t did not take into account the substant ia l amount of i n t e r - d i s t r i c t migration within the s tates . In addit ion i t may be noted that these two studies were concerned only with the descr ip-t ion of the spat ia l pattern of migration within India and did not attempt to analyse the factors behind such movements. 2.4.2 Present Day India Gosal (1961), who carr ied out his analys is by using d i s t r i c t data, found that tthough the absolute number of migrants was la rge , i t was r e l a t i v e l y small in r e l a t i o n to tota l populat ion. His study revealed that , apart from refugee movement, people moved from densely populated or drought ridden a g r i c u l t u r a l t racts to areas newly won for agr i cu l ture or where i r r i g a t i o n offered fresh p o s s i b i 1 i t e s , or to the towns or c i t i e s . He also noticed that movements were general ly 34 over short d istances , and economic motivations predominated. He accounted for the r e l a t i v e l y low migration rates by lack of opportunity, ignorance, l i n q u i s t i c problems and caste and j o i n t family t i e s . He divided India into three groups of regions and ca l led them, (1) regions of high mob i l i t y , (2) regions of medium mobi l i ty and, (3) regions of low mobi l i ty respect ive ly on the basis of how many people were born outside a given d i s t r i c t but enumerated in that d i s t r i c t . According to th is c l a s s i f i c a t i o n high mobi l i ty areas (with over 16% not born in the d i s t r i c t s ) included Assam, the HooghTy side conurbation ( i . e . the g.rea<ter Cal cutta region) and the Damodar c o a l f i e l d , northwestern India ( i . e . Delhi and surroundings), south and southwest Mysore ( including Bangalore and Kolar Gold F i e l d s ) , western tndia ( inc luding the c i t i e s of Bombay andwAhmedabad) and the i so la ted urban foc i of immigration namely Madras, Hyderabaidi, :Gwalior, Nagpur, Jabalpur, Indore, U j j a i n , Kanpur, Lucknow and Dehra Dun. The regions of medium mobi l i ty (with 8-16% of the people not born in the d i s t r i c t s ) included the Upper Gangetic P l a i n , with short to medium range movement to new a g r i c u l t u r a l land; movements to acreas of expanding w e l l - i r r i g a t i o n in Rajasthan and local movement to towns in Madhya Pradesh. 35 The regions of low mobi l i ty (with under 8% not born in the d i s t r i c t s ) included the ' saturated ' areas in the Lower Gangetic p la in and the coastal p l a i n s , and sparsely populated h i l l areas. Gosal 's work suffers from two l i m i t a t i o n s . F i r s t l y the c l a s s i f i c a t i o n is misleading since the c r i t e r i a used did not necessar i ly represent mobi1ity. The percentage of people born outside a given d i s t r i c t but enumerated in that d i s t r i c t merely showed how many people came in to those d i s t r i c t s from other d i s t r i c t s . In other words, they showed the in tens i ty of in-migrat ion to those d i s t r i c t s . Mobi l i ty obviously includes both in and out-migrat ion. Thus the basis of c l a s s -i f i c a t i o n was not cor rect . Secondly, though his work, using d i s t r i c t data, was in much deta i l and had the advantage over the other two studies of r e l a t i v e l y uniform c r i t e r i a in r e l a -t ion to the s ize and population of the spat ia l u n i t s , i t was again e s s e n t i a l l y an id iographic study and no attempt was made e i ther to analyze the factors of migration or to examine theoret ica l j u s t i f i c a t i o n s of the resu l ts obtained. The l i s t of such id iographic works could be expanded. But the works carr ied out by Bhargava (1964), Bose (1967, 1 970), Bulsara (1 964), Desai (1 964), Kumar (1 967), Lai 1 (1961), Mitra (1967) and Narain (1967) do not add much new things to our knowledge about Indian interna l migrat ion. Bulsara, however, showed that the majority of the people 36 going to the nine urban areas of Western India were from the rural areas and most of them were e i ther farmers or a g r i c u l t u r a l labourers. According to his f i n d i n g s , ". . . adverse and unsat is factory economic condit ions in the o r ig ina l habitat formed the largest s ingle 'push' factor to dr ive or st imulate the movement of people to c i t i e s " (Ib id: 37). Recently the f indings of Bulsara have been, to some extent, sbustantiated by Greenwood (1971). By using the migration data in the 1961 census he examined the rural-to-urban and urban-to-urban migration streams of India. His main object ive was to d iscover , as have others with respect to both developing and developed countr ies , the extent to which the degree of urbanization determines the cityward migration in India. Through mult ip le regression model, which included the var iables l i k e income of workers at both the rural and urban areas, percentage di f ferences of urbanizat ions, and d istance, he showed that purely economic factors ' e x p l a i n ' 54% of the var ia t ion in migration in rural to urban areas in India , and 63% of the variance in migration for urban ito urban areas. From this resu l t he concluded that *Sovani (1966) and Bose (1970) argued that the importance of 'push' and ' p u l l ' factors in determining the spat ia l pattern of internal migration in India has been overemphasized. They attempted to show that they (push and pu l l ) do not e x i s t . However, the i r works lacked adequate substant iat ion as none of them attempted a deta i led study, both in terms of geographical coverage of India and the inc lus ion of a l l kinds of migration streams. 37 "something more than br ight l ights is c r i t i c a l in the potent ia l migrant's decis ion to move to a c i t y " (Greenwood, 1971:262). Perhaps the most deta i led study on internal migra-t ion of India is being carr ied out as a part of the V i l l a g e Study Project (VSP) [Dasgupta, 1974]. Though the study is far from complete, Dasgupta reported some of the tentat ive f i n d i n g s . Based on the data coming from 200 v i l l a g e s of India, the study could "\ . . succeed in ident i fy ing some patterns in the re la t ionsh ip between propensity to migrate, migrant c h a r a c t e r i s t i c s , and v i l l a g e level 'push' and ' p u l l ' factors" ( Ib id :10) . It was discovered that ". . . the pro-pensity to out-migrate can be explained by both 'push' (land shortage, low f e r t i l i t y of land, skewed d i s t r i b u t i o n of land, and the resu l t ing proportion of landless a g r i c u l t u r i s t s ) and ' p u l l ' (commercialization of a g r i c u l t u r e , extent of cash cropping, and proximity of town and main roads) factors [ Ib id : 12]. 2.4.3 Bangladesh On Bangladesh, Obaidullah (1967) did an important study regarding the internal migration of people. He estimated the migration f igures from 'place of b i r th data' and t r i e d to analyze the spat ia l pattern of migrat ion. He took the 17 d i s t r i c t s of Bangladesh as the spat ia l un i t s . 38 F i r s t l y , he prepared four tables showing, (1) the net rate of migration from d i s t r i c t to d i s t r i c t , (2) the percentage d i s t r i b u t i o n of migrants by 'categories of reg ions , ' (c) the density of migrants per square mile in d i f fe rent d i s t r i c t s and (4) the number of migrants by d istance. Net migration per square mile and density of popula-t ion f igures when converted to ranks, appeared to be c lose ly re lated (Spearman's rank cor re la t ion c o - e f f i c i e n t s have been found to be 0.84 for 1951 and 0.87 for 1961). Secondly, with the help of a s t a t i s t i c a l model an attempt was made to discover whether the re la t ionsh ip was assstrong as suggested by the c o - e f f i c i e n t of r a n k - c o r r e l a t i o n . It was found from the model that net migration of people in that country was p o s i t i v e l y re lated with density of population and negatively re lated with distance - thereby substant iat ing the resu l ts that were found from the tables and the c o - e f f i c i e n t of rank c o r r e l a t i on. Carried out by a s t a t i s t i c i a n , the study was more concerned with deriv ing the parameters for his s t a t i s t i c a l model rather than understanding the migration process per se. Select ion of net-migration as the dependent var iab le did not allow him to know the behaviour of migration streams at a l l . That is because net-migration represents the conse-quences from migration (both in and out) rather than the migration i t s e l f . The inc lus ion of only population density 39 and distance in the model also indicated that he was not very much interested in discovering a l l the important determinants of migrat ion. His f indings are also p a r t i a l l y va l id since population density has been found to be p o s i t i v e l y re lated with out-migration but negatively re lated with in-migration in the majority of s tud ies . 2.4.4 Pakistan and Ceylon (Sri Lanka) Afzal (1967) in his study of intercensal net migra-t ion to urban areas of Pakistan, made an attempt to analyze the spat ia l pattern of rates of urbanward migration from the rural areas. He employed the ' surv iva l ra t io method' to estimate the rate of rural-urban net-migrat ion. From his analys is i t was found that the volume of rural-urban migration in West Pakistan was almost four times that of E. Pakistan (now Bangladesh) during the intercensal period of 1951-1961. He concluded, though without substant ia t ion , that r e l a t i v e l y more ' p u l l ' factors such as, (a) more indust r ia l e s t a b l i s h -ments and (b) a larger number of urban centres were the reasons for this greater number of people migrating to the urban areas of W. Pakistan. Hashmi (1964), in an e a r l i e r study concerning the soc ia l aspect of the people of Karachi , in West Pakistan, however, gave an ind icat ion that in the la ter part of the 50's there was a tendency for rural-urban migration to s lacken. 40 Vamathevan (1961), in his monograph concerning internal migration in Ceylon found that the apparent pattern of internal movements was not one of d i s t i n c t l y rural-urban type. His study revealed that most of the people went to the rural areas of the country to reclaim agr i cu l tu ra l land and in response to other rural programmes. 2.5 Summary In sum, close examination of the theories of migra-t ion revealed that of a l l the theories only ' push-pu l l ' theory has been developed to understand the process of migration . pj3*» se. It may also be observed from the migration studies on the Indian subcontinent that there are none try ing to examine the a p p l i c a b i l i t y of the theories of migration e x p l i c i t l y . We therefore suggest that the 'push-pu l l ' theory should be tested c r o s s - c u l t u r a l l y and also that in the 'push-p u l l ' framework a l l the poss ib le streams should be inc luded, so that the theory may be considered as an approach which describes the migration system in tqto_. We may now formally state our in terpretat ion of the theory as a ser ies of hypotheses which can be tested in the context of Bangladesh. 41 2.6 The Working Hypotheses It is hypothesized that in Bangladesh, the migration streams that have been recognised in the study w i l l have d i f f e r e n t responses to 'push' and ' p u l l ' f a c t o r s , and the e f fec t of the intervening obstac le , distance separating o r i g i n and dest inat ion of migrat ion, w i l l also vary from one kind of migrat ion: stream to another. Since six kinds of migration streams are recognized in th is study we may generate the fol lowing hypotheses, a p r i o r i , based on the kind of response we would expect from these streams to 'push' and ' p u l l ' factors and to the intervening f a c t o r , d istance. Hypothesis 1 In the case of R-R migration streams, i t is hypothesised that both 'push' and ' p u l l ' factors would be operating in regulat ing the spat ia l pattern of Rural-to-Rural migration streams but 'push' factors would be dominant in R-R out-migration streams and ' p u l l ' factors would be dominant in R-R in-migrat ion streams. It is also expected that the i n t e r -vening f a c t o r , distance separating the o r ig in and dest inat ion of mi grat i on, woul d be a more important factor in determining the spat ia l pattern of R-R in-migrat ion streams than in that of R-R out-migration streams. 42 Hypothesis- 2 It is hypothesised that both 'push' and ' p u l l ' factors would be operating in determining the spat ia l pattern of Rural-to-Urban out-migration streams and Urban-to-Rural out-migration streams but 'push' factors would be dominant in R-U out-migration streams and ' p u l l ' factor would be dominant in U-R out-migration streams. In terms of the e f fect of the intervening obstac le , d is tance, on these streams, i t is expected that distance would have more e f fect on U-R out-migration streams than on R-U out-migration streams. Hypothesis S In the case of Urban-'toUrban migration streams i t is hypothesised that both 'push' and ' p u l l ' factors would be operating in determining the spat ia l pattern of these migration streams but 'push' would be dominant in U-U out-migration streams while ' p u l l ' w i l l be dominant in U-U i n -migration streams. The e f fect of distance would also be d i f f e r e n t . Distance w i l l have more e f fect on U-U in-migrat ion streams than on U-U out-migration streams. Hypothesis 4 Population pressure on a g r i c u l t u r a l land is very high in Bangladesh and as such per capita a v a i l a b i l i t y of 43 * land to the peasants is very low. This w i l l , in most of the cases, create a desperate s i tuat ion under which the peasants would leave the i r p laces. In other words, they w i l l be 'pushed 1 from the i r place of residence. It has already been mentioned that the overwhelming majority of the people l i v e in rural areas and over 90% of the tota l (gross) migration takes place within the rural d i s t r i c t s . Thus i t may be hypothesised that 'push' factors would be dominant in deter-mining the overal l spat ia l pattern of migration streams in Bangladesh. It is the purpose of the fol lowing chapters to test the above hypotheses by examining the spat ia l pattern of migration in Bangladesh for the period 1951-1961. However, before discussing the resu l t of the a n a l y s i s , i t has been f e l t necessary to introduce the study area to the reader and also to ver i f y the accuracy, source and nature of data on which the study is based. It has also been f e l t necessary to discuss the rat iona le on which the 'explanatory v a r i a b l e s ' have been se lec ted . Thus, while the next two chapters w i l l be devoted to discussing the study area and the nature of data the other chapters w i l l contain the resu l ts from the analys is and the conclusions. For d e t a i l s see Chapter 3. Chapter 3 THE STUDY AREA 3.0 Introducti on Bangladesh, on which the present study is based, * emerged as an independent nation in the year 1971 . Our purpose is to fami la r i ze the reader with those aspects and circum-stances of this region under which migration has been taking p lace. The primary motivation behind migrat ion, as discussed e a r l i e r , is said to be economic. Thus the materials that are covered here mainly describe the economic sett ings of the country. In add i t ion , an attempt has also been made to discuss the demographic, l i t e r a c y and physical aspects of th is part of the world. It may be noted here that unl ike India and Pakistan, Bangladesh is inhabited by one race — A chronological account of p o l i t i c a l events, which lead to the d i s in tegrat ionof Pakistan and emergence of Bangladesh is presented in Appendix I. Chapter 3 covers the basic aspects of the economy only . A more deta i led account may be found in Chapter 4. 44 45 Bengal is . Thus, l i n g u i s t i c a l l y and c u l t u r a l l y , i t is a homogenous region where neither physical (since i t is e s s e n t i a l l y a f l a t p la in) nor cu l tura l nor l i n g u i s t i c barr iers hinder movement from one place to another.* 3.1 Geographical Location and Physical Sett ing Bangladesh l i e s between 20° 75! and 26° 75' North l a t i tude and 80° 30' and 92° 75' East longitude and is s i tuated on the north-eastern side of the Indian sub-continent (Map 3 .1) . It has an area of 55,201 square miles and had a population of 50.84 m i l l i o n , according to the 1961 census. The decinial population census due to be taken in 1971 could not be conducted due to the p o l i t i c a l s i t u a t i o n . However, the most recent census was taken during February and March of th is year. The tota l population of Bangladesh is 71,316,517, according to the provis ional resu l ts released by the Census Commission on June 8, 1974 {Bangladesh - a f o r t n i g h t l y news b u l l e t i n issued by the Embassy of Bangladesh, Washington, D . C , June 25, 1 974, p. 3) . It indicated that the population of th i s r reg ion increased by 40.27 per cent during the l a s t 13 years ( i . e . about 31 per thousand per y e a r ) . Bangladesh comprises almost 4/5 of the Bengal Basin (see Map 3 .2) , which l i e s between the Chhota Nagpur Plateau and the western spurs of the Arakan Yomas (Rashid, 1965, 5) . It is e s s e n t i a l l y a filat p la in formed by the a l l u v i a l deposits * Chittagong H i l l Tracts is an exception where the majority of the people is t r i b a l . 46 Map Location No. 3.1 of Bangladesh Map No. 3.2 Physical Sett ing of Bangladesh 48 of the r i vers Ganges and Brahmaputra, mainly during the Pleistocene per iod. Besides the Ganges and the Brahmaputra (which are l o c a l l y known as the Padma and the Jamuna respec-t i v e l y ) the whole country is c r i sscrossed by innumerable t r i b u t a r i e s and d i s t r i b u t a r i e s forming a large r i v e r network system t o t a l l i n g at least f i f t e e n thousand miles in length. 3.2 Economic Sett ing of Bangladesh Bangladesh's economy is pr imar i ly a g r i c u l t u r a l . The dominant pos i t ion of agr i cu l ture can be gauged by the fact that i t contributes more than 55 per cent to Gross Domestic Product (GDP) and absorbs more than 75 per cent of the labour fo rce . In addit ion the processing of domestical ly produced a g r i c u l t u r a l goods accounts for well over half the value added in manufacturing industr ies (Khan, 1972). 3.2.1 Land Tenure and Ownership D i s t r i b u t i o n In Bangladesh the land/man ra t io is very low, and the recent 'Master Survey' conducted by the East Pakistan Agr i cu l tu ra l Min istry in 1968 shows a general decl ine in average farm s i z e . In 1960 the average farm s ize was 3.54 acres but by 1968 i t was reduced to 2.59 per a g r i c u l t u r a l household. The s t r i k i n g feature of the land ownership pattern is that the proportion of households without any land 49 whatsoever is very low. According to the 'Master Survey' only 8.6% of the rural population of ten and above belonged to the landless a g r i c u l t u r a l labour fo rce . For a land-poor country l i k e Bangladesh, the ra t io is very low indeed. Share-cropping, which is often i d e n t i f i e d with i n e f f i c i e n c y , is not very widely pract iced e i t h e r . From Table 3.1 i t may be observed that an overwhelming proportion; of farms are e i ther owner operated (61%) or owner-cum-tenant operated (37%). A n e g l i g i b l e proportion of farms are operated by landless farmers. From Table 3.1, sect ion ( c ) , i t appears that the larger owners tend to l e t out for share-cropping to small farmers, and that such arragenemtns cover a s i g n i f i c a n t , though not overwhelming, proportion of the cu l t i va ted land. 3.2.2 Land U t i l i z a t i o n and Cropping Pattern The remarkable aspect of the pattern of cropping and land use is that in sp i te of a great scarc i ty of land, cropping in tens i ty is not high. Though i t is t h e o r e t i c a l l y possible to ra ise as many as three crops per year from the same piece of land (in other words in tens i ty of cropping of 300%), due to the absence of any f r o s t or very dry periods and high s o i l f e r t i l i t y , on the average only 35% of the net c u l t i v a b l e area (N.C.A.) is cu l t i va ted more than once, since the d i s t r i b u t i o n of r a i n f a l l in uneven and the i r r i g a t i o n Table 3.1 ASPECTS OF LAND TENURE (A) (B) (C) Type of Tenure Percentage of Farms Type of Tenure Percentage of Area Average Size of Acres Owner operated farms 61 Owner operated farms 82 Owner operated farms 3.1 Owner-cum-Tenant farms 37 Tenant operated: S h a r e c r o p p i ng ... C 6 ) C a s h R e n t i n g ( 2 ) 18 Owner-cum-Tenant farms 4.3 Tenant farms 2 Tenant farms 2.4 Source: Compiled by Khan (1972:41) from the i960 Pakistan Census of A g r i c u l t u r e , A Summary of East Pakistan Data. 51 f a c i l i t i e s are not adequate. Ra in fa l l is very scanty, par-t i c u l a r l y during the winter [rabi) crop season. In general , the in tens i ty of c u l t i v a t i o n , however, var ies from 101% to 180% (Morrison, 1973:246). On the other hand, the scope for horizontal expansion of net cu l t ivated area (N.C.A-.) has almost been exhausted. Rice overwhelmingly dominates the agr i cu l ture of th is country. It covers almost 80% of the cropped acreage. Three kinds of r i c e are grown, namely Aus 3. Aman and Boro. Aman is the main crop and covers about 50% of the tota l cropped acreage of Bangladesh. It is sown in the rainy season (June-July) and harvested in October to November. Aus i s an ear ly rainy season crop and also matures ear ly . It takes only 80 to 120 days to r ipen . Seed is broadcasted in March, Apr i l or May and growth depends very great ly on the monsoon r a i n s . In many areas Aus and Aman are sown together to ensure that at least one crop is harvested. After the aus i s harvested, the Aman continues to grow and is harvested several months l a t e r . Boro or Spring r i c e is a transplanted kind of r i c e . Seeds are sown in October and November and seedlings are then transplated within f i v e to s ix weeks. It is harvested in A p r i l . Since i t is a spring crop when ra in is scanty i t grows well only in some parts of Bangladesh such as in Sylhet and the north-eastern parts of Mymensingh, where there i s plenty of water ava i lab le for i r r i g a t i o n from the lowlying areas l o c a l l y ca l l ed Haors or B i t s « 52 Recently the output from a l l these three types of r i c e stood in the neighbourhood of 12.5 m i l l i o n tons (Far EasternEoonomio Review, Apr i l 29, 1974:52). Though the output has increased s tead i ly over the l a s t few years i t could not cope with the increase of populat ion, p a r t i c u l a r l y from the l a t t e r part of the 1960's and onward. Every year more than a imi l l ion.- tons of fcice has to be imported from abroad to meet the d e f i c i t . The scope for increase in the y i e l d per acre i s , however, s u b s t a n t i a l , since the present y i e l d per acre (about 800 l b s . ) is much lower than in many of the South and South-east Asian countr ies , with s imi la r geographical and human circumstances, such as China, Taiwan, South Vietnam or Ceylon. For example, i f Bangladesh can approach the per acre y i e l d in war-torn Vietnam, the ent i re d e f i c i t can be wiped out. Jute , though not as important as r i c e in terms of share of acreage, is not less important in a d i f f e r e n t context. It is the pr inc ipa l source of foreign exchange for the country and the main cash crop for the peasants, and the money earned from i t is the s ing le most important determinant of the non-subsistence consumption of the rural populat ion. On the average i t occupies about two m i l l i o n acres of c u l t i v a b l e land as opposed to 21 m i l l i o n aeises occupied by r i c e . The sowing season is the same as that of Aus and thus jute and Aus compete for the same land in most of the areas of th is country. 53 The present anxiety is that i f the d e f i c i t of food continues unabated, the farmers may ult imately have to switch to Aus from Jute i r revocab ly . This w i l l have a serious impact on the economy of th is country since jute and jute products are the most important sources of foreign exchange. Among other crops,rmentiondmay be made of sugarcane, rapeseed and pulses . They are , however, dwarfed by r i c e and jute in terms of share of acreage. 3.2.3 Transportat ion Networks There were 1,712 miles of ra i lways, about 2,000 miles of metalled road ( l o c a l l y ca l l ed paoca road) , and 3,000 miles of a l l weather navigable waterways in Bangladesh in the early 1960's (Rashid, 1965). As the whole country is c r i sscrossed by innumerable r i vers and i t s t r i b u t a r i e s and d i s t r i b u t a r i e s , the main mode of t ransportat ion — for both passengers and cargoes, is waterways. During the rainy season the mileage increases to 5,000. In genera l , the north-western region or 'North Bengal is less access ib le by waterways than the rest of the country. But roads and railways do connect air the d i s t r i c t s of North Bengal. The slow going country boats are s t i l l the main mode of r i v e r t ransportat ion . Internal Water Transpor-ta t ion Authority (IWTA) — an autonomous corporation of E. Pakistan, estimated that there were at least 203,072 passenger 54 carrying countryboats in the early 1960's (Rashid, 1965:274). In addit ion there were 172 steamers and 8887 motor vessels (Ibid.:-274). Narayanganj, Dacca Ban" sal , Chandpur, Goalanda, Bharib Bazar and S i rajganj are some of the most important r i v e r ports of the country. There are about 106 stat ions ( l o c a l l y ca l led ghat), with a network of 124 routes. Railways operate over 63 routes and a l l the d i s t r i c t s are connected ( including subdiv is ion headquarters) except Bar isal and chittagong H.T. THere are 417 railway s t a t i o n s , 30 of which are railway junct ions . Rashid estimates that they serve an area of 20,000 square miles (Ibid:268). Roads are the only transportat ion network which connects a l l the d i s t r i c t s of the country and the mileage is increasing most r a p i d l y . In addit ion to the metalled roads, there are also 25,000 miles of major and 33^000 miles of minor unmetalled roads ( l o c a l l y c a l l e d katoha raasta). They connect almost a l l the thana headquarters of the country. There are regular bus services connecting a l l the d i s t r i c t s of Bangladesh. 3 .3 Demographic Aspects Bangladesh is indeed a very crowded land. In 1961 the density of population of th is region was 922 per square mi le , and i t has increased to 1291 per square mile in 1974. 55 This is the highest density for any country in the world except for some of the is land or c i t y states l i k e Hong Kong or Singapore. The nearest comparisons are to be found in the Netherlands (1065 in 1969), Taiwan (1039 in 1968) and Belgium (890 in 1969) [Khan, 1972:10]. The true magnitude of the problem cannot, however, be understood from such a comparison. Unlike in the Netherlands or Taiwan about 75% of the popula-t ion of Bangladesh depends d i r e c t l y on a g r i c u l t u r a l land as a source of income. 3.3.1 Population Growth The f i r s t census of 1872 put the population of the area of the then Bengal Province which now comprises Bangladesh at 22 m i l l i o n (Rashid, 1965:328). The population rose at an average rate of 8-10% at every census unt i l 1941 when the increase was 17,9% (Rashid, 1965:337). There may be two reasons for th is sudden r i s e in the growth of populat ion. F i r s t l y , unt i l 1931 both the f e r t i l i t y (about 50 per thousand) and morta l i ty rate (about 40 %) were high. Thus the growth rate was low. But from 1931 onward due to the improvement of medical f a c i l i t i e s the m o r t a i l i t y rate was sharply reduced while the f e r t i l i t y rate remained the same. As a r e s u l t , the growth rate was more than double that of the previous decades. Secondly, the 1941 f igure was i n f l a t e d by 56 enumerators of d i f fe rent communities (Hindus and Muslims) who wanted to show the i r community in a majority as the P a r t i t i o n of Bengal on re l ig ious grounds was imminent. Due to a larger exodus of people belonging to the Hindu community after the creat ion of East Pakistan in 1947, the 1951 popula-t ion census showed an increase of only 100,000 people during the decade 1941-51. (The total population was 42 m i l l i o n and 42.1 m i l l i o n in 1941 and 1951 r e s p e c t i v e l y . ) But the resp i te gained in the 1941-51 decade was foiliiliowed by a huge increase of 20.9% which pushed the population up to 50.84 m i l l i o n in 1961. As there was no large scale emigration and the morta l i ty rate continued to decl ine sharply while the f e r t i l i t y rate remained more or less constant,^the gap between the v i t a l rates was high and as such growth rate was higher. The recent census of 1974 revealed that the population has increased by 40.27 percent during the las t 13 years . In other words, the rate of growth continued to increase during the recent times (3.1% increase per year during 1961-74 compared to 2.1% increase per year during the decade 1951-61). This disproves Roger Revel le 's predict ions (Revel le , 1973). He argued that Bangladesh is in the las t leg of i t s 'demo-graphic t r a n s i t i o n . ' According to him the rate of growth of population w i l l at least remain constant for the next few decades i f not ac tua l ly dropping down. The present w r i t e r , however, bel ieves that Bangladesh is s t i l l in the _ Population growth 1941-51 was also severely affected by famine as well as disturbances and refugee movements at the p a r t i t i o n of India and Pakistan. 57 middle of the 'demographic t r a n s i t i o n ' and as such the rate of growth w i l l continue to increase unt i l the end of the century. The argument is that though both morta l i ty and f e r t i l i t y are decreasing, the rate of decrease of f e r t i l i t y w i l l be much slower than the rate of decrease of m o r t a l i t y . As a resu l t the gap between the v i t a l rates w i l l continue to increase unt i l the end of the century when the f e r t i l i t y rate may drop sharply i f the family planning measures are successful and people get more educated and adopt urban ways of l i f e . 333.2 Spat ia l D i s t r i b u t i o n of Population Growth In the decade 1951-1961, the spat ia l d i s t r i b u t i o n of the growth has been uneven. There has come a slowing of the rate of growth in the more populous parts such as in the d i s t r i c t s of B a r i s a l , Comi l la , Noakhali e tc . and an abrupt increase in that of the less populous areas such as in Rajshahi , Jessore, Kushtia etc . Table 3.2 shows the per-centage var ia t ion by d i s t r i c t s and the i r respect ive popula-t ion d e n s i t i e s . As the v i t a l rates are more or less the same through-out the country, i t is apparent that the var ia t ion in the growth of population within Bangladesh is la rge ly due to the loss of people from more out-migration than in-migration in 58 Table 3.2 POPULATION DENSITY AND PERCENTAGE VARIATION OF GROWTH OF POPULATION, 1951-61 Di s t r i cts Densi ty Percentage Var iat ion 1951 1961 Ban* sal 859 1 ,005 17 Bogra 851 1 ,048 23 Chi ttagong 929 1 ,103 18 Chittagong H.T. 56 76 34 Comilla .-~ 1 ,462 1 ,693 16 Dacca 1 ,413 1 ,768 25 Di naj pur 544 655 24 Fan' dpur 1 ,030 1 ,180 14 Jessore 643; 860 34 Khulna 446 526 18 Kus hti a 645 851 32 Mymeiisi ngh 090 1 ,103 21 Noakhali i , 11 7 1 ,285 15 Pabna 844 1,044 23 Raj sha hi 603 769 27 Rangpur 787 1 ,025 38 Sylhet 639 729 14 Sources: Rashid (1965:329 and 331) and Census of Pakistan, 1961, East Pakistan. 59 the populous d i s t r i c t s and gain of people from more in-migrat ion * than out-migration in the less populous d i s t r i c t s . It is also apparent that due to the absence of la rge-sca le urbani-zation most of the migration has taken place between the rural regions. 3.3.2 Urban-Rural D i s t r i b u t i o n As mentioned e a r l i e r , Bangladesh i s one of the least urbanized regions of South A s i a . Only 5.19% of i t s population l i v e s in urban areas. There were only four c i t i e s in 1961, namely, Dacca, Chittagong, Narayanganj and Khulna which had more than 100,000 people. In addit ion there were 73 towns of various sizes in that year. The urban population numbered about 2.7 m i l l i o n in 1961. Dacca was the largest c i t y with a population of 556,712 and Chittagong came next with 364,205. According to the provis ional resu l ts of the recent census of 1974, while the population of Dacca c i t y increased by more than 200%, Chittagong's population increased very s lowly. In f a c t , Khulna replaced Chittagong as the V i t a l rates s t a t i s t i c s are not ava i lab le in the census. However, when the 'dependency r a t i o s ' of the d i s t r i c t s were compared with each other, less than 2% devia-t ions were found. This i n d i r e c t l y shows that the v i t a l rates do not vary much from one d i s t r i c t to another. 60 second largest c i t y where population increased by more than 400% in 13 years . 3.4 Li teracy While the l i t e r a c y rate in 1951 was 21.1% i t was 17.6% in 1961. The decrease was for two reasons. F i r s t l y , the standard of attainment of a l i t e r a t e was changed. In 1951, a b i l i t y to read was s u f f i c i e n t , whereas in 1961 one had to be able to read and write a simple l e t t e r . Secondly, almost a l l the l i t e r a t e Hindus l e f t the reg ion. The spat ia l d i s -t r i b u t i o n of percentage of l i t e r a c y was, however, quite even. The range of d i f ference was not more than 2-3% among most of the d i s t r i c t s . Though the l i t e r a c y rate was very low by world standards i t was more than in Pakistan and almost equal to that of India . The present l i t e r a c y is estimated to be nearly 30%. In terms of the d i s t r i b u t i o n of l i t e r a c y by occu-pation i t may be noted that the peasants are almost 100% i11i te ra te . Another important feature is the d i f ference in the proportion of l i t e r a t e amongst males and females. Whereas 26% of the males were l i t e r a t e in 1961, the per-centage among females was only 8 .6 . The main language of l i t e r a c y is Bengal i , in which more than 90% of the l i t e r a t e s can read and wrfte. About 61 5% are l i t e r a t e only in Urdu. About 20% of the l i t e r a t e s are b i - l i n g u a l in Bengali and English and Arabic and Persian are known by about 3% of the l i t e r a t e s , in addit ion to Bengali or Eng l i sh . 3.5 H i s t o r i c a l Aspects of Migration in Bangladesh The geographical area which now comprises Bangladesh gained people from other parts of the Indian sub-continent unt i l the early 20th century. Substantial numbers of people started to move out from th is part of the world af ter World War I . Among them Mymensias and Noakhalias (people from the d i s t r i c t s of Mymensingh and Noakhali) were dominant. The main place of dest inat ion was the Brahmaputra va l ley of Assam, Goalpara d i s t r i c t in the west and Cachar d i s t r i c t in the east . Within the then undivided Bengal, Calcutta was the main place of des t inat ion . After the P a r t i t i o n of 1947, the whole s i t u a -t ion changed, however. Not only were the ' f r o n t i e r s ' in Assam no longer there, Bengal i t s e l f was divided and as a resu l t Calcutta was also l o s t . There were some migrations of Hindus to West Bengal just af ter p a r t i t i o n but that can hardly be described as a 'normal' case of migrat ion. The main reason was p o l i t i c a l and emigration almost completely stopped af ter 1951. Thus from 1951 the geographical region for which the present research has been carr ied out can be termed a 62 'c losed r e g i o n . ' Some people did emigrate to West Pakistan but the total number was not very la rge . [See Ahmed, 1958, Rashid, 1965 and Chapter V in E. Bengal; Tables and Report 1951, Census of Pakistan, Karachi .] Chapter 4 THE SOURCE OF DATA AND THE METHODOLOGY AND RATIONALE FOR SELECTING THE VARIABLES 4.0 Introduction This chapter w i l l pr imar i ly be devoted to discussing the nature, source and accuracy of data on which the analysis is based. The rat iona le for se lect ing the var iables w i l l also be d iscussed. 4.1 Source, Nature and Accuracy of Data Except for the var iables Z $ (index of cumulative connect iv i ty) and D.. [Distance separating d i s t r i c t of o r i g i n ( i th d i s t r i c t ) and d i s t r i c t of dest inat ion ( j th d i s t r i c t ) ] , which have been ca lculated by the present wr i te r , a l l other var iables have been gathered or ca lculated from secondary sources. There have been two main sources — the Census Reports of Pakistan from the years 1951 and 1961 and Harouner-Rashid 1 s text on Bangledesh (1965). 63 64 Regarding the nature of data, they are in aggre-gated form and most of the mater ia l s , at least those which have been found to be useful for the present research, are not ava i lab le beyond the d i s t r i c t l e v e l . The d i s t r i c t s are administrat ive units and they are second in the * hierarchy (see Map 4 . 1 ) . This is one of the reasons for se lect ing the d i s t r i c t s as. the spat ia l u n i t s . According.to some of the authors, such as Davis (1958), and Zachariah (1968) the data in the census are reasonably accurate. Rashid (1965) and Khan (1973) suggest that they suffer from some underenumeration rather than overestimation which is supposed to be better of the two e v i l s . Recently Levy (1972) used both one year migration (gathered from primary sources) and l i f e time migration (calculated from place of b i r th data) data of Venezuela to examine whether the explanatory var iab les (var iables measuring costs and benef its from migration) have d i f f e r e n t ef fects on the two kinds of migrat ion. But the resu l t showed that the determinants of 1961 migration and l i f e time migration in 1961 are b a s i c a l l y s i m i l a r . From the * The hierarchy of administrat ive units from top to bottom is as fo l lows: (1) D i v i s i o n s , (2) D i s t r i c t s , (3) Sub-d iv i s ion , (4) Thanas, (5) Union Councils and (6) V i l l a g e s . In 1961, there were four d i v i s i o n s , 17 d i s t r i c t s , 59 sub-d iv i s ions , 411 thanas, 4,055 union counci ls and 64,493 v i l l a g e s (Census of Pakistan, Vo l . 2, 1 961 , 1-22). 65 Map N o . 43.1 66 resu l ts of the study he also inferred that the estimation of migration data from the place of b i r th data was reasonably accurate (he used the same method that has been employed in th is research to ca lcu late the migration f i g u r e s ) . Thus, we may also consider our estimated f igures on migra-t ion from place of b i r th data to be reasonably cor rec t . Ind i rec t l y i t also provides the basis for considering our migration f i g u r e s , l i f e time migration in 1951 and in 1961, to represent the migration f igures for 1951 and 1961. 4.2 Estimation of Migration Figures (OM,. . and IM.. .) from  the Place of B i r th Data The areal units for which migration estimates are ca lcu lated are ca l led d i s t r i c t s ; they are 17 in number. In the census schedule (1951 and 1961); the s t a t i s t i c s for b i r th place were obtained by requir ing each person to report the d i s t r i c t in which he was born. If he was born outside Bangladesh (the then E. Pakistan) he was also asked to give the province of his b i r th (for W. Pakistan) or state (for India) . 4.2.1 The Method If IM stands for the number of persons enumerated in a d i s t r i c t ( jth d i s t r i c t jkbutbborn elsewhere ( i th d i s t r i c t s ) 67 in Bangladesh and OM stands for the number of persons born in that d i s t r i c t (now i th d i s t r i c t ) but enumerated in another d i s t r i c t ( j th d i s t r i c t ) , we may c a l l IM..' the l i f e -time in-migrants to the d i s t r i c t j from the i t h d i s t r i c t s and OM;:. . the l i f e time out-migrations from the i th d i s t r i c t to the j th d i s t r i c t s . The d i f ference between these two numbers (IM.. - OM. . = NM...) may be ca l led the l i f e - t i m e i j i j I j net-migrants for that p a r t i c u l a r d i s t r i c t . In our study, however, only in and out-migration have been c a l c u l a t e d , since we are concerned with the streams of migration not the net f igure on migrat ion. Thus the in and out migration f igures which have been ca lcu lated from the 1951 census place of b i r th data have been ca l l ed as ' i n or out migration streams in the year 1951' and the migration f igures which have been ca lcu lated from the 1961 census place of b i r th data have been c a l l e d as ' i n or out migration streams in the year 1961.' 4.2.2 Evaluation of the Method The method, as given above, has certa in drawbacks. These drawbacks may be traced to , (a) errors in the basic data, and (b) to the e f fect of d i f f e r e n t i a l mor ta l i ty . For i f the death rates are high, migration estimates obtained without morta l i ty correct ion may be in error by a s i g n i f i -cant extent. 68 In addit ion to the inaccuracies caused by errors in the basic data and morta l i ty there w i l l also be some d i s t o r t i o n s in migration estimates because p i a c e - o f - b i r t h data are stock rather than flow data. For instance neither mult ip le movements nor c i r c u l a r movements can be i d e n t i f i e d in census data. If a person moves from a d i s t r i c t and returns to i t within his l i f e time, e i ther d i r e c t l y or a f ter moving on to another d i s t r i c t , he w i l l be regarded as a non-migrant. Therefore, estimates of migration between two d i s t r i c t s (or any other spat ia l units) as obtained from p i a c e - o f - b i r t h data should be interpreted with caut ion . 4.3 Estimation of Rate of Internal Migration (IV. . or OV..-.:) by Using Modified Gravity Model Since the spat ia l units of th is study ( d i s t r i c t s ) vary widely in terms of population number, instead of using simple migration f i g u r e s , ' v e l o c i t y of migrat ion ' (V . . ) has been estimated by using the modified grav i ty model based on Bogue, Shrycock et al. (1957). This measure expresses the r e l a t i v e degree of in tens i ty of a migration stream as a rate by e l iminat ing the e f fect of population s ize of both the sending ( i th) and receiv ing ( jth) areas, as given by the formula: V, . = M. . • P./P. P . 4 i j I J t i j 69 where V . . = the rate of flow of the migration J stream from area i to area j , M . . = the number of migrants in the stream J from area i to area j , P. = population of the place of o r ig in ( i ) , P. = population of the place of dest inat ion ( j ) , and P. = tota l population of a l l potent ia l areas of dest inat ion inc luding the area of o r i g i n or the population of Bangladesh. From this general formula ' v e l o c i t y of in-migrat ion streams' ( I V . . ) and ' v e l o c i t y of out-migration streams ( O V . . ) , can • J * 3 be ca lcu lated by replacing V . . by IV . . or OV.. and M - . by I M ^ or O M ^ J J A H others remaining the same. Migration rate expressed in th is form serves two purposes: (a) i t el iminates the e f fec t of population s ize of both the sending ( i th) and receiv ing ( jth) d i s t r i c t s , and (b) i t also el iminates the e f fect of the physical s ize of the i th and j t h d i s t r i c t s since the population s i ze and the physical s ize of the d i s t r i c t s have been found to be p o s i t i v e l y corre lated (r = .8135). Ca lcu lat ions have been made for a l l types of streams and for a l l the possible pairs of sending and receiv ing d i s t r i c t s in them. In this research six types of migration streams have been recognized. Depending on the number of rece iv ing and sending d i s t r i c t s in the streams in 1951 and 1961 r e s p e c t i v e l y , ca l cu la t ions have been made for 440 pairs 70 of d i s t r i c t s for 1951 and 424 pairs of d i s t r i c t s for 1961. In other words, in t o t a l , ca l cu la t ions have been made for 862 pairs of d i s t r i c t s . 4.4 C l a s s i f i c a t i o n of the D i s t r i c t s into 'Rura l ' and 'Urban'  D i s t r i c t s and the Select ion of Migration Streams In the present research, internal migration has been looked at in terms of the streams of migration rather than gross or net f igures of migration for each of the d i s t r i c t s as i t is bel ieved that net or gross f igures of migration do not represent the movement of people. Thus there was a need to ident i fy the d i s t r i c t s of o r i g i n and d i s t r i c t s of dest inat ion (as any movement involves a place of o r i g i n and a place of dest inat ion) for each of the streams. The migration streams, in turn , may be c l a s s i f i e d into d i f f e r e n t types depending on the nature and kind of place of o r i g i n and place of d e s t i n a t i o n . The general tendency is to c l a s s i f y the regions into rural and urban areas and name d i f f e r e n t types of migration streams according to the i r place of o r i g i n and place of des t inat ion . In a d d i t i o n , migration streams may also be c l a s s i f i e d into in-migrat ion streams and out-migration streams depending on whether migrants are coming into the region or going out of the reg ion. 71 Accordingly , in the present research the 17 d i s t r i c t s of Bangladesh were also c l a s s i f i e d into ' r u r a l ' and 'urban' d i s t r i c t s . The c r i t e r i a for the c l a s s i f i c a t i o n was the level of urbanization of the respect ive d i s t r i c t s . D i s t r i c t s having greater than the national average of urbanization (percentage of people l i v i n g in the urban areas of the country) have been defined as 'urban' while the others (having less than the national average of urbanization) have been defined as ' r u r a l . ' Since in the d i s t r i c t s which have been defined as 'urban' the majority of the people l i v e in the rural areas of the d i s t r i c t they should not be taken as urban d i s t r i c t s in the l i t e r a l sense. The c l a s s i f i c a t i o n is appl icab le or v a l i d only for the s p e c i f i c purpose of the present research. It was, however, necessary to examine whether the urban population of the 'urban' d i s t r i c t s (which nowhere exceeds 12% of the total population of the respect ive d i s t r i c t s ) has had any e f fect on migration in order to j u s t i f y the c l a s s i f i c a t i o n . To meet this purpose a ' p i l o t p ro ject ' was undertaken. In that project a p r inc ipa l axis type factor analys is was employed with six var iables including v e l o c i t y . . . . . . of migration (OV..) and percentage of urbanization((U . . ) . Other selected var iab les , which are supposed to be p o s i t i v e l y or negatively re lated with migrat ion, were as fo l lows: (a) D... (Distance separating the i th and j th d i s t r i c t s ) , (b) TCA.. (Difference of tota l cu l t i va ted area between the 72 i t h and j t h d i s t r i c t s ) , (c) P.. (Difference in population s ize between the i th and j th d i s t r i c t s ) and (d) R.. (Difference in r i c e production between the i th and j t h d i s t r i c t s ) . Since there are 272 (17 x 16) possible pairs of sending and receiv ing d i s t r i c t s , ca l cu la t ions were made for a l l these 272 p a i r s . In other words, factor analys is was carr ied out with six var iables and each var iab le had 272 readings. From the factor analys is i t was found that on IL^ . loading was s i g n i f i c a n t and i t was s i g n i f i c a n t in factors where the loading on OV^ was also s i g n i f i c a n t . Thus i t may be in ferred that U. . was not independent of OV . . . It may, however, be mentioned that loading on OV.. was lower in 1961 that in 1951 where loading on U..was also s i g n i f i c a n t . It indicated that the e f fect of urbanization may have decreased during the decade instead of increased despite increase in the percentage of urbanization of the 'urban' d i s t r i c t s . In other words, urbanization of the d i s t r i c t s tended to be i d e n t i f i e d as a separate ' f a c t o r ' independent of OV . . . Thus, taking two c r i t e r i a , (a) the types of place of o r i g i n and place of dest inat ion (urban and rura l ) and (b) the types of migration (in and out) , t h e o r e t i c a l l y i t i s possible to have eight migration streams as shown in Table 4 .1 . However, in th is study only s ix independent migration streams have been recognized (Table 4.2) for i t was found that urban to rural in-migration streams and Table 4.1 Theoret i ca l ly Possible Types of Migration Streams Types of place of o r ig in and place of dest inat ion Types of Migration Types of Migration Streams O U T ^ : ^ ^ *»1 . Rural to Rural In-Migration Streams **2. M*Urban to Rural In-Migration Streams ^ 3 . RuRural to Rural Out-Migration Streams ""M. *Rural to Urban Out-Migration Streams U R B A N-^1^^^ ^ ^ * * ^ 0 U T ^ 1 ^ ^ *~5. Urban to Urban In-Migration Streams •^6. *Rural to Urban In-Migration Streams 0*7. Urban to Urban Out-Migration Streams **8. **Urban to Rural Out-Migration Streams The same stream with two d i f fe rent names. ^ co The same stream with two d i f fe rent names. Table 4.2 Migration Streams Recognized in th is Study Rural to Rural Out-Migration Streams Rural to Rural In-Migration Streams Rural to Urban Out-Migration Streams Urban to Rural Out-Migration Streams Urban to Urban Out-Migration Streams Urban to Urban In-Migration Streams 75 rural to urban out-migration streams are not independent but rather represent the same streams described as urban to rural out-migration streams and rural to urban in-migrat ion streams r e s p e c t i v e l y . 4.5 Select ion of the Explanatory Variables The major source of data for the explanatory var ib les i s the census of Pakistan. Other secondary sources such as Rashid's and Ahmed's texts on Bangladesh (1965 and 1958) were also h e l p f u l . The data are obviously not ideal but are the only ones a v a i l a b l e . Rarely any of the v a r i a b l e s , however, have been taken in the i r o r i g i n a l form. They have been modified to present them in a form which would be more mean-ingful for ournpurpose. For example population density of the respect ive d i s t r i c t s has been expressed in terms of the ' rura l population density per square mile of c u l t i v a t e d area' (RD/SQMCA), tota l cu l t ivated area of the respect ive d i s t r i c t s has been expressed in terms of 'the tota l c u l t i -vated area as a ra t io of net cu l t i va ted area' (TCA/NCA), r i c e acreage of the respect ive d i s t r i c t s has been expressed in terms of 'the percentage of area under r i c e in r e l a t i o n to the tota l cu l t i va ted area' (Rice/TCA) and so on. F i n a l l y 9 (nine) independent var iables have been selected for 1951 and 11 (eleven) for 1961. Two v a r i a b l e s , namely, (1) 'percentage of share croppers in r e l a t i o n to the tota l 76 number of a g r i c u l t u r a l labourers ' (SC/TALF) and (2) 'percentage of share croppers plus landless a g r i c u l t u r a l labourers in r e l a t i o n to the tota l number of a g r i c u l t u r a l labourers ' (SC+ LLALF/TALF) which have been selected for the 1960 analys is could not be included in the 1951 var iab le matrix since data for these are ava i lab le only for the 1961 census. (From the resu l t of the analys is i t was, however, discovered that the exclusion of these two var iables from the 1951 analysis did not matter much since in most of the cases they were found to be p o s i t i v e l y corre lated with each other and, moreover, LLALF represents the majority of the TALF.) Since a l l analyses have been performed for the pairs of d i s t r i c t s , each of them comprising the d i s t r i c t of o r i g i n ( i th d i s t r i c t ) and the d i s t r i c t of dest inat ion ( j th d i s t r i c t s ) , the f i n a l data for each of the var iables (excluding D.^ .) have been expressed in terms of the d i f ference between the i th and j t h d i s t r i c t s . The fol lowings are the var iab les which have been selected for analys is (Tables 4.3 and 4 . 4 ) . 4.6 Rationale for Select ing the Explanatory Variables The se lect ion of explanatory var iables has been based on three premises, namely: (1) the economic-demographic condit ions of Bangladesh, (2) the distance involved in the migration and the r e l a t i v e a c c e s s i b i l i t y of the d i s t r i c t s 77 Table 4.3 VARIABLES SELECTED FOR ANALYSIS, 1951 1. OV.. or IV . . Ve loc i ty of Out-Migration Streams or 1 J 1 J Ve loc i ty of In-Migration Streams between the i th and the j th D i s t r i c t s 2. RD/SQMCA Difference in the Rural Population Density/Square Mile of Cul t ivated Area between the i th and j th D i s t r i c t s . 3. Rice/TCA Difference in the Rice Acreage/Total Cul t ivated Area between the i th and j th D i s t r i c t s TCA/NCA Difference in the Total Cul t ivated Area/Net Cult ivated Area between the i th and j th D i s t r i c t s Jute/NCA Difference in the Jute Acreage/Net Cult ivated Area between the i th and j th D i s t r i c t s LLALF/TALF Difference in the Landless Agr i cu l tu ra l Labour Force/Total Agr i cu l tu ra l Labour Force between the i t h and j t h D i s t r i c t s 7. ze Difference in the Cumulative Index of Connectiv ity or A c c e s s i b i l i t y between the i th and j th D i s t r i c t s Difference in the Percentage of L i teracy between the i th and j th D i s t r i c t s U. . 1J Difference in the Percentage of Urbani-zation between the i th and j t h D i s t r i c t s 1 0 . o u Airplane Distance Separating the i th and j t h D i s t r i c t s in Miles 7 8 Table 4.4 VARIABLES SELECTED FOR ANALYSIS/ 3*961 (Including the Dependent Variable - Migration) OV.. or IV . . (Veloc i ty of Out-Migration Streams or Ve loc i ty of In-Migration Streams between the i th and the j t h D i s t r i c t s RD/SQMCA Difference in the Rural Population Density/Square Miles of Cul t ivated Area between the i th and j th D i s t r i c t s 3. Rice/TCA Difference in the Rice Acreage/Total Cul t ivated Area between the i th and j th D i s t r i c t s TCA/NCA Difference in the Total Cul t ivated Area/Net Cult ivated Area between the i th and j t h D i s t r i c t s Jute/NCA Difference in the Jute Acreage/Net Cult ivated Area between the i th and j th D i s t r i c t s Difference in the Cumulative Connec-t i v i t y between the i t h and j t h D i s t r i c t s SC+LLALF/TALF Difference in the Share Choppers and Landless Agr i cu l tu ra l Labour Force/ Total Agr i cu l tu ra l Labour Force between the i th and j th D i s t r i c t s 8 . SC/TALF Difference in the Share Croppers/ Total Agr i cu l tura l Labour Force between the i th and j t h D i s t r i c t s 9. LLALF/TALF Difference in the Landless Agr i cu l tura l Labour Force/Total Agr i cu l tu ra l Labour Force between the i th and j t h D i s t r i c t s CONTINUED 79 Table 4.4 (Continued) Diffierenee in the Percentage of L i teracy between the i th and j th D i s t r i c t s Difference in the Percentage of Urbanization between the i th and j t h D i s t r i c t s Airplane Distance Separating the i th and j th D i s t r i c t s in M i les . 80 and (3) the trend of the se lect ion of explanatory var iables in previous internal migration studies both theoret ica l and empi r i ca l . The selected var iables are discussed below. 4.6.1 Variables-Representing or Indicating the D i f f e r e n t i a l  Agr i cu l tu ra l Endowments of the i t h and j th D i s t r i c t s (Rice/TCA, Jute/NCA and TCA/NCA) Since the l i v e l i h o o d of the bulk of the population of the country depends on r i c e , and jute is the main source of non-subsistence income of the peasants, these two crops have been selected to examine whether the r e l a t i v e posit ions of the d i s t r i c t s in terms of the a v a i l a b i l i t y of land under r i c e or jute a f fec t people's movement. In the absence of any data on the d i f f e r e n t i a l income of the people of the 17 d i s t r i c t s of Bangladesh, jute has also been selected as a surrogate measure of income of the people of the d i f f e r e n t d i s t r i c t s . Rice acreage has been expressed in r e l a t i o n to the tota l cu l t i va ted area (T .C .A . ) because r i c e is cu l t i va ted more than once (in a given piece of land) in a year. On the other hand jute acreage has been expressed in r e l a t i o n to the net cu l t i va ted area (N.C.A.) since jute is cu l t i va ted only once in a year. Both r i c e and jute have been expressed in terms of the acreage that they occupy because the production 81 of r i c e and jute is p o s i t i v e l y re lated with the acreage under them. The y i e l d or acre does vary from d i s t r i c t to d i s t r i c t but the var ia t ion i s marginal . Thus land under r i c e and jute gives us an idea of the d i s t r i c t s in terms of the i r r e l a t i v e a g r i c u l t u r a l endowments. It is expected that , in order to search for work or better standards of l i v i n g , people w i l l tend to leave the d i s t r i c t s where the r i c e acreage is r e l a t i v e l y lower and try to migrate to the d i s t r i c t s where the a v a i l a b i l i t y of land under r i c e is r e l a t i v e l y higher. In the case of jute two kinds of move may r e s u l t . Since jute c u l t i v a t i o n is more labour- intens ive than r i c e c u l t i v a t i o n * i t is expected that the d i s t r i c t s which have more land under jute w i l l tend to a t t r a c t more people (maybe mainly landless a g r i c u l t u r a l labourers) from those d i s t r i c t s where jute c u l t i v a t i o n is comparatively l e s s . It is also expected that few people w i l l tend to leave the d i s t r i c t s l i k e Mymensingh, Rangpur, Dacca, Pabna and Comilla which are the main jute growing areas and go to the d i s t r i c t s l i k e Sylhet , Chittagong, Dinajpur, Rajshahi where jute c u l t i v a t i o n is r e l a t i v e l y l e s s . In other words peasants in d i s t r i c t s which grow r e l a t i v e l y more jute w i l l have a r e l a t i v e l y higher income and thus w i l l not tend to go to the d i s t r i c t s where land under jute is r e l a t i v e l y l e s s . 82 The rat io of the total cu l t i vated area (T .C .A . ) to the net cu l t i va ted area (N.C.A.) of the d i s t r i c t s has been selected to serve two purposes: (1) to indicate the r e l a t i v e in tens i ty of cropping of the respect ive d i s t r i c t s and also (2) to ind icate the r e l a t i v e proportion of the winter {rabi) crops to the T .C .A . In the rainy season {Kharif crop season) almost a l l parts of the country can grow either r i c e {Aus and Aman) or Jute , but only in some parts of the country can rabi crops such as o i l seeds, tobacco, Boro r i c e and wheat be grown. Two kinds of movements are expected: (a) People may tend to leave the d i s t r i c t s where the ra t io TCA/NCA is high and thus less opportunit ies for further expansion of cu l t i vated area, and go to the d i s t r i c t s with the low rat io which may of fer more potent ia l for expansion and may therefore be more a t t r a c -t i v e , or (b) people may also tend to leave the d i s t r i c t s with low TCA/NCA ra t io and thus less p o s s i b i l i t i e s for gett ing jobs , and go to the d i s t r i c t s with high r a t i o where immediate jobs may be a v a i l a b l e . When looked at in terms of the 'push-pu l l ' theory, i t is expected that a l l the var iables discussed in th is sec t ion , except TCA/NCA, w i l l act as ' p u l l ' factors to the j t h d i s t r i c t s i f the rat ios are greater than in the i th d i s t r i c t s , in both * in and out-migration streams. Since jute does not occupy much land, i t s e f fec t w i l l d e f i n i t e l y be much less than the *For TCA/NCA the condit ions w i l l be jus t opposite in most of the cases though i t w i l l also act as a ' p u l l ' i f the rat io is greater in the j th d i s t r i c t s than in the i th d i s t r i c t s . 83 other two var iab les in determining the spat ia l pattern of migration on a l l the streams. 4.6.2 Variables Representing or Indicating the Population  Pressure on the Agr i cu l tura l Land and the Supply of  Agr i cu l tu ra l Labour (RD/SQMCA, LLALF/TALF, SC/TALF and SC+LLALF/TALF) As the majority of the people in Bangladesh depends d i r e c t l y on agr i cu l tu ra l land as a source of income, instead of general population density ! : rural population density per square mile of cu l t i va ted a r e a , ' RD/SQMCA, has been selected to indicate the r e l a t i v e posit ions of the d i s t r i c t s in terms of population pressure on a g r i c u l t u r a l land. Urban population has been excluded from the numerator because only people l i v i n g in the rural areas are d i r e c t l y dependent on agr i cu l tu ra l land. Since i t is assumed that one of the main motivations of migration is to increase income or , in other words, increase the standard of l i v i n g , i t is expected that people w i l l move away from those d i s t r i c t s where the population pressure on a g r i c u l t u r a l land is greater and go to those d i s t r i c t s where i t is l e s s . Thus the var iab le RD/SQMCA has been selected to f ind out whether the r e l a t i v e posi t ions of the d i s t r i c t s in terms of the population pressure on a g r i c u l t u r a l land a f fec t people's movements. 84 The percentage of ' landless a g r i c u l t u r a l labourers in r e l a t i o n to the total a g r i c u l t u r a l labour f o r c e , ' LLALF/TALF, has been selected as one of the explanatory var iables to examine whether the a g r i c u l t u r a l labour supply posi t ions of the respect ive d i s t r i c t s a f fec t the magnitude or d i rec t ion of i n t e r - d i s t r i c t movements of people. Theoret i ca l l y i t may be assumed that in the d i s t r i c t s where the rat io LLALF/TALF is greater , the supply of a g r i c u l t u r a l workers is also greater . Thus the chances or opportunit ies for gett ing work in those d i s t r i c t s w i l l be lower than in those d i s t r i c t s where' the ra t io is lower. Under these circumstances, i t is expected that the a g r i c u l t u r a l labourers w i l l tend to move away from those d i s t r i c t s where they are more in number and go to those d i s t r i c t s where they are l e s s . The rat io LLALF/TALF also indicates the posit ions of the d i s t r i c t s in terms of the i r percentage of people who c u l t i v a t e the i r own land. Since i t may be assumed that the people who c u l t i v a t e the i r own land have more a f f i n i t y to the i r land than the people who c u l t i v a t e others land, i t may be expected that people w i l l migrate away in greater number from those d i s t r i c t s where the LLALF/TALF ra t io is greater and go to those d i s t r i c t s where i t is l e s s . The other two v a r i a b l e s , the rat io of 'share croppers plus landless a g r i c u l t u r a l labourers to tota l a g r i c u l t u r a l labour f o r c e , SC + LLALF/TALF, and the r a t i o 85 of 1 share croppers to total a g r i c u l t u r a l labour f o r c e , 1 SC/TALF, represent more or less the same condit ions of the d i s t r i c t s as expressed by LLALF/TALF except for the addit ion of share croppers. Level of e f f i c i e n c y of farming is supposed to be associated with share cropping for i t is assumed that share croppers tend to take less care of the land than c u l t i v a t o r s owning the land. Thus, we may expect that these two v a r i a b l e s , p a r t i c u l a r l y the l a t t e r one, w i l l enable us to examine whether the r e l a t i v e posi t ions of the d i s t r i c t s in terms of the i r e f f i c i e n c y of c u l t i v a t i n g land would lead the people of those d i s t r i c t s to make any move. It is expected that people w i l l tend to move away from those d i s t r i c t s where the rat ios SC/TALF and/or SC+LLALF/ TALF are greater and go to those d i s t r i c t s where they are l e s s . When looked at in terms of the 'push-pu l l ' theory, i t is expected that a l l the var iables discussed in th is sect ion w i l l act as a 'push' factor in the i th d i s t r i c t s i f the ra t ios are greater than in the j th d i s t r i c t s and th is w i l l happen for both in and out-migration streams. 86 4 . 6 . 3 Variables representing or ind icat ing the distance  involved in migration and a c c e s s i b i l i t y of the  d i s t r i c t s ( P . . and E3 ) •Any migration involves a certa in d istance. The general tendency is that the greater the distance between the o r i g i n and the des t inat ion , the smaller the migrat ion. In the 'push-pu l l ' theory, distance is considered to be the most important ' intervening v a r i a b l e . 1 Thus, we have selected 'the distance separating the d i s t r i c t of o r i g i n ( i ) and the d i s t r i c t of dest inat ion ( j ) , D . . , to examine the ro le of distance in a f fec t ing people's movement from one d i s t r i c t to another. However, measurement of distance separating the d i s t r i c t s has been found to be a l i t t l e complicated, for two reasons. F i r s t l y , in the case of i n t e r - d i s t r i c t migrat ion, the spat ia l units have been found to be large enough to generate several sets of 'migrat ion-d istances ' of var iab le lengths depending on the place of o r ig in and the place of dest inat ion within the d i s t r i c t s . For example, i f we consider the i n t e r - d i s t r i c t migration between Dacca and Mymensingh, the distance covered by migrants from the northern part of Dacca to the southern part of Mymensingh (about 2 0 miles) would be s i g n i f i c a n t l y less than, say, the distance covered by migrants from the southern part of Dacca to the northern part of Mymensingh (about 1 2 0 miles) [See Appendix I I ] . This kind of s i tuat ion would ar ise for almost a l l the pairs of 87 d i s t r i c t s , but the s i tuat ion would be more pronounced for contiguous than for non-contiguous d i s t r i c t s . Thus i t was f e l t necessary to determine the 'migration d istances ' which could reasonably represent or indicate the 'mean d istances ' t r a v e l l e d by i n t e r - d i s t r i c t migrants. Secondly, since i t is possible to travel by any of three modes of t ransportat ion ( r a i l , road and r i v e r ) , the distance covered by migrants would also vary according to the mode or modes of t ransporta-t ion involved in i n t e r - d i s t r i c t migrat ion. In order to tackle these problems the geometric centres of the respect ive d i s t r i c t s have been determined. Af ter locat ing the geometric centres of the d i s t r i c t s , the distances separating those 'nodes' ( in miles) have been ca lcu lated from a map of Bangladesh (Appendix I I ) . As these distances may be ca l led 'the a i r -d is tances separating the centres of the d i s t r i c t s , ' D.. has been designated as the ' a i r -d i s tances separating the i th and j t h d i s t r i c t s . ' Real iz ing that the a c c e s s i b i l i t y of any two places may sometimes be as important as the physical distance separating them, the 'cumulative connect iv i ty or a c c e s s i b i l i t y index, ' £3, of the respect ive d i s t r i c t s has also been selected as one of the explanatory v a r i a b l e s . £3 has been designated as a 'cumulative index' since the a c c e s s i b i l i t y of the d i s t r i c t s in terms of road, r a i l and r i v e r has been c a l -culated separately and then added to represent the 'cumulative 88 connect iv i ty index. ' The formula used to ca lcu late the 3 index has been the fo l lowing: 3 = v/c; where 3 = connect iv i ty index, c = number of nodes in the network ( i . e . number of thana, subdiv is ion and d i s t r i c t headquarters connected by any of the three transportat ion network: ,. roads.,~rai Iways and r iverways) , and v = number of v e r t i c e s , i . e . l i nes jo in ing any two nodes, in the network system. 3 index for each of the modes of t ransportat ion for each of the d i s t r i c t s has been ca lculated separately and then added to represent the 'cumu-l a t i v e connect iv i ty index' (Please see Appendix II for d e t a i l s ) . It may be expected that , c e t e r i s -paribus, the magnitude of out-migration from the i th d i s t r i c t s would be greater i f the Z3 is greater and v i c e - v e r s a . S i m i l a r l y , for in-migrat ion streams, the magnitude of migration would be greater in the j t h d i s t r i c t s i f the S3 i s greater and vice versa . 4.6.4 Other Variables (L . . and U,.) The other two variables,namely (1) 'the d i f ference in the percentage of l i t e r a c y between the d i s t r i c t s of o r i g i n ( i ) and the d i s t r i c t s of dest inat ion ( j ) , L.•, and • J (2) 'the d i f ference in the percentage of urbanization between the d i s t r i c t s of o r i g i n ( i ) and the d i s t r i c t s of dest inat ion (j ) , U. n., have been selected for they have been 89 found to be included in the explanatory var iab le matrix in almost a l l the previous studies on migrat ion. In most of these studies i t has been found that while the magnitude of migration for in-migrat ion streams varies p o s i t i v e l y with the percentage of l i t e r a c y and urbanization in the j th d i s t r i c t s i t varies negatively with them (% l i t e r a r c y and % of urbanization) for out-migration streams in the j t h d i s t r i c t s . Thus we may expect that the se lect ion of these two var iables would enable us to examine whether the d i f -ference of urbanization and l i t e r a c y between the d i s t r i c t s would a f fect the movement of people. When looked at in terms of the 'push-pu l l ' theory i t may be expected that both of these var iables would act as ' p u l l ' factors in both in and out-migration streams. People would be ' p u l l e d ' to the j th d i s t r i c t s i f 0 . . and/or U. . are greater than in the i th d i s t r i c t s . Chapter 5 DETERMINANTS OF THE SPATIAL PATTERN OF MIGRATION STREAMS WITHIN BANGLADESH, 1951 AND 1961 5.0 Introduction The primary object of th is chapter is to discuss the resu l ts obtained from the ana lys i s . The analys is has been performed by using a pr inc ipa l axis type factor analysis with the data matrix (which includes both dependent and independent var iab les) selected and discussed in the previous chapter.* Though most of the previous migration studies carr ied out by Sahota (1968), Greenwood (1968, 1969a, 1969b, 1970 and 1971), Labor (1971), Gallaway et al. (1967, 1968), and Logan (1970) used mult ip le regressiong techniques, p r i n c i p l a axis type factor analysis has been found to * The actual computations in th is study have been performed by "UBC FACTO" package programme in The Univers i ty of B r i t i s h Columbia's IBM 360 computer system. " T h e o r e t i c a l l y , the method of pr inc ipa l com-ponent cannot f ind a so lut ion to the factor analys is model 90 91 be more useful for th is study for the fol lowing reasons. When the mult ip le regression technique was employed with OV.. or IV . . as dependent var iab le and the other var iables as independent v a r i a b l e s , the resu l ts were not en l ightening. Though the R2 for the regression equations were high, cor re la t ion c o - e f f i c i e n t s of the v a r i a b l e s , r, were not general ly high, except for D . . . However, the high R2 helped us to know that the var iables included in the analys is do ' e x p l a i n ' a high percentage of the total var ia t ion of migration ( v e l o c i t y ) . In add i t ion , incorporat ion of a l l independent var iables was not possible for streams such as Urban to Urban and Rural to Urban where the number of unless the main diagonal of the cor re la t ion matrix is replaced by the communalities of the v a r i a b l e s . Thus pur ists l i k e to have the communalities estimated in some way before having cor re la t ion matrix fac tored . " (B jer r ing , James H., UBC FACTO: FACTOR ANALYSIS, UBC Computing Centre, May, 1969, pp. 4-5) . The FACTO, in our case, s tarts with the c o r r e l a -t ion matrix having 1's on the diagonal r e a l i z i n g ". . . that the factor space i t so obtain accounts for a greater per-centage of the tota l variance than the one i t would obtain i f i t placed community estimates on the main diagonal" [ibid.j p. 5) . After that , i t determines the communality estimates by computing squared mult ip le cor re la t ion c o e f f i c i e n t s . It may also be noted that the data in th is study comes from 'universe ' rather than any sample" and as such no confidence l i m i t has been set in the a n a l y s i s . * Regression estimates of two migration streams are attached in Appendix IV. 92 observations was less than the number of v a r i a b l e s , thus a f fect ing the degrees of freedom. Mult ip le regression was also considered to be a less neutral technique than factor analys is since in that ve loc i ty of migration (V.-.) would be considered as a dependent v a r i a b l e s , a p r i o r i . Thus the measures of re la t ionsh ip between the independent and depen-dence var iables would be biased to the extent that the corre lat ions would be measured in r e l a t i o n to the dependent var iab le and dependent var iab le only. But in factor analys is a l l the var iab les ( including migration) are considered inde-pendent. In i t , the re la t ionsh ip of the var iables is important, no one being selected as s p e c i a l , and successive factors are extracted by operating upon the basic data matrix, which are independent of each other. A l g e b r i c a l l y , each factor is a function of a set of unknown v a r i a b l e s , but is re lated to each known var iab le by an empirical parameter, the loading of the known var iables on the f a c t o r . The factors are said to be independent^of each othertwhenethey-are un-c o r r e c t e d and the pr inc ipa l components method assures t h i s . The factor scores produced as construct data on a factor are derived for each observations as the sum of the products of each var iab le value and the loading of that var iab le on the f a c t o r . [Each var iab le i s weighted proport iona l ly to i t s involvement in a pattern (described by the f a c t o r ) . The more involved the v a r i a b l e , the higher the weight. The sum of these weight-times-data products for a l l the var iables 9 3 y i e l d s the factor score] . Besides i t s more neutral character , the in terpretat ion of factor loadings and factor scores in the factor analysis leaves more room for inductive arguments than the in terpretat ion of co r re la t ion c o - e f f i c i e n t s which are deductive in nature. The a b i l i t y re late data in a meaningful fashion is a prime aspect of induct ion , and for t h i s , factor analys is is useful and e f f i c i e n t . As the 'push-pu l l ' theory is more of an inductive type, factor analys is was found to be more'helpful to ident i fy and define the 'push' and ' p u l l ' f a c t o r s . 5.1 I d e n t i f i c a t i o n of the ' C r i t i c a l Factors ' From the factor matrix, which consisted of 2-4 independent f a c t o r s , only that factor has been selected for in terpretat ion where the ' loading on 'the ve loc i ty of out-migration (OV..) or 'the v e l o c i t y of in-migrat ion ' ( I V . . ) , 1J * 3 has been highest ( level of s i g n i f i c a n c e > .400). The selected factor has been designated as the ' c r i t i c a l f a c t o r ' since i t describes the migration pattern most f u l l y . In some instances, however, loading on migration was s i g n i f i c a n t in more than one f a c t o r . It indicated that more than one 'system' described the migration pattern for that case and both the factors were interpreted though the emphasis was given on the ' c r i t i c a l f a c t o r . ' For example, in the case 94 of B a r i s a l , R-Rout migration streams, 1951, the fol lowing were the loadings on the factor matrix. Table 5.1 Loadings on the Factors* Variables Factor 1 Factor 2 Factor 3 1 . -.9117 .0236 -.0005 2. RD/SQMCA -.8135 .0457 -.3742 3. Rice/TCA .5999 .0794 -.6408 4. Jute/NCA -.1074 -.1983 .8160 5. TCA/NCA -.2836 -771153 .0699 6. LLALF/TALF -.7151 -.6145 .0729 7. .1053 -.8332 -.4190 8. .2623 -.3600 -.8124 9. .1661 -.9142 .2307 10. .6935 .3683 -.4991 S i g n i f i c a n t loadingss are underl ined. 95 Table 5.2 Corre lat ion of the Factors Factor 1 Factor 2 Factor 3 -Factor 1 1.0000 Factor 2 .00002 1.0000 Factor 3 .00002 .00001 1.0000 From Table 5.1 i t may be found that loading on 0V. . is highest and s i g n i f i c a n t only in factor 1. Thus, factor 1 was designated as the ' c r i t i c a l f a c t o r ' and other factors were considered to be redundant for our purpose. We could se lect the ' c r i t i c a l f a c t o r ' without any hes i tat ion since from Table 5.2 i t may be found that the factors are uncorrelated and thus independent of each other. In the tables only those var iables were mentioned on which loadings were s i g n i f i c a n t (see Table 5 .4) . We have recognized six kinds of migration streams but within a p a r t i c u l a r kind of migration stream there are a number of migration streams or ig inat ing from or coming to the d i s t r i c t s . Thus factor analys is has been performed for a l l the i n t e r - d i s t r i c t migration streams within a p a r t i c u l a r kind of migrat ion. As a resu l t factor analys is should have been performed as many as 102 times;.51 times Table 5.3 THE EXPECTED NATURE OF FACTOR LOADINGS ON THE EXPLANATORY VARIABLES AND MIGRATION VELOCITY a b c Explanatory Variables (push or pu l l ) Nature of Loadings on the Explanatory Variables Nature of Loadings on o v i . i o r I v i . i 1. RD/SQMCA ~ push -} 0 V 1 J i v i . t + +• •> 2. Rice/TCA ~ pul l + + J + -3. TCA/NCA - pull -} o r +} + 1 + • or _- + J +j 4. Jute/NCA ~ pull + +-} 5. E3 ~ intervening var iab le + -} + -6. SC+LLALF/TALF - p u s h > + '"J CONTINUED Table 5.3 (Continued) a b c Explanatory Variables (push or pu l l ) Nature of Loadings on the Explanatory Variables Nature of Loadings on o v i j °'- I v i j 7. SC/TALF ~ push -} + 8. LLALF/TALF ~ push -} + 9. L... ~ pul l • J -} 10. U. . ~ pull -} +J -} 11. D.. ~ intervening var iab le -} 98 for 1951 and 51 times for 1961. But in p r a c t i c e , factor analys is was not performed for Chittagong H.T. as the number of people coming out or going to that d i s t r i c t in 1951 was very small ( less than 100). Thus, the number of analyses carr ied out in 1951 was 48 instead of 51.. and the tota l number of analyses was 99 instead of 102. To f a c i l i t a t e understanding these i n t e r - d i s t r i c t migration within a par-t i c u l a r kind of migrat ion, six examples, one each from six kinds of migrat ion, are included in the sections descr ib ing d i f f e r e n t kinds of migration streams (see Maps 5.13to 5 .6) . The number of analyses performed under each category of migrat ion, both for 1951 and 1961, is given below: 1. R u r a l t o R u r a l O u t - M i g r a t i o n S t r e a m s , 195 1 = 1 2 2. R u r a l t o R u r a l O u t - M i g r a t i o n S t r e a m s , I 961 = I 2 3. R u r a l t o R u r a l I n - M i g r a t i o n S t r e a m s , 195 1 = 12 4. R u r a l t o R u r a l I n - M i g r a t i o n S t r e a m s , 1961 = 1 2 5. R u r a l t o U r b a n O u t - M i g r a t i o n S t r e a m s , I 95 I = 12 6. R u r a l t o U r b a n O u t - M i g r a t i o n S t r e a m s , •I 961 = 12 • 7. U r b a n t o R u r a l O u t - M i g r a t i o n S t r e a m s , 1951 =• 4 99 Map No. 5.1 Rural-to-Rural Out-Migration Streams, 19 51 O u t - M i g r a t i o n S t r e a m s f r o m M y m e n s i n g h D i s t r i c t o f O r i g i n ( i ) ~ M y m e n s i n g h D i s t r i c t s o f D e s t i n a t i o n ( j ) - O t h e r R u r a l D i s t r i c t s ( ' U r b a n 1 D i s t r i c t s a r e S h a d e d ) . 1 00 Map No. 5.2 Rurali-To-Rural In-Migration Streams, 1951 I n - M i g r a t i o n S t r e a m s t o M y m e n s i n g h D i s t r i c t o f D e s t i n a t i o n ( j ) ~ M y m e n s i n g h D i s t r i c t s o f O r i g i n ( i ) ~ O t h e r R u r a l D i s t r i c t s ( ' U r b a n D i s t r i c t s a r e S h a d e d ) . 1 01 Map No. 5.3 Rural-to-Urban Out-Migrat ion, 1951 O u t - M i g r a t i o n S t r e a m s f r o m M y m e n s i n g h D i s t r i c t o f O r i g i n ( i ) ~ M y m e n s i n g h D i s t r i c t s o f D e s t i n a t i o n s ( j ) ~ U r b a n D i s t r i c t s ( S h a d e d ) 102 B A N G L A D E S H ^ s R A N G P U R j | | D I S T R I C T S A N D T H E F O U R M A J O R J \ rfc R E G I O N S Map No. 5.4 Urban-to-Rural Out-Migration Streams, 1951 -O u t - M i g r a t i o n S t r e a m s F r o m D a c c a . D i s t r i c t o f O r i g i n ( i ) ~ D a c c a D i s t r i c t s o f D e s t i n a t i o n ( j ) ~ A l l R u r a l D i s t r i c t s ( non-Shaded ) 1 03 Map No. 5.5 Urban-1o-Urban 0ut- Migrat ion, 1951 O u t - M i g r a t i o n S t r e a m s f r o m D a c c a D i s t r i c t o f O r i g i n ( i ) ~ D a c c a D i s t r i c t s o f D e s t i n a t i o n ( j ) ~ O t h e r U r b a n D i s t r i c t s Map No. 5.6 Urban to Urban In-Migration Streams, 1951 -I n - M i g r a t i o n S t r e a m s t o D a c c a D i s t r i c t s o f O r i g i n ( i ) ~ O t h e r U r b a n D i s t r i c t s D i s t r i c t o f D e s t i n a t i o n ( j ) ~ D a c c a 1 05 8. U r b a n t o R u r a l O u t - M i g r a t i o n S t r e a m s , I 961 = 5 ( T h e n u m b e r o f u r b a n d i s t r i c t s i n c r e a s e d f r o m 4 i n 1951 t o 5 i n 1 9 6 1 ) 9. U r b a n t o U r b a n O u t - M i g r a t i o n S t r e a m s , 1951 = 4 10. U r b a n t o U r b a n I n - M i g r a t i o n S t r e a m s , 1961 = 5 1 1 . U r b a n t o U r b a n I n - M i g r a t i o n S t r e a m s , 19 5.1 =• 4 12. U r b a n t o U r b a n O u t - M i g r a t i o n S t r e a m s , 1961 = 5 T o t a l n u m b e r o f a n a l y s e s p e r f o r m e d f o r 1951 = 48 T o t a l n u m b e r o f a n a l y s i s p e r f o r m e d f o r 196.1 = 5 1 Based on the resu l ts of factor analys is 99 tables have been prepared. In addit ion to incorporporating the factor scores on the subjects ( d i s t r i c t s ) from the c r i t i c a l factor and the factor loadings on the var iables in the ' c r i t i c a l f a c t o r , ' the fol lowing information has also been * included to f a c i l i t a t e descr ibing the migration pattern. 1. N u m b e r o f M i g r a n t s i n t h e S t r e a m ( i n c o I urn n ' a ' ) . 2. V e l o c i t y o f O u t - M i g r a t i o n S t r e a m s , OV.. o r V e l o c i t y o f I n - M i g r a t i o n Streams, I V . . ( i n c o l u m n ' b ' ) . i J and D i s t a n c e ( a i r - d i s t a n c e s e p a r a t i n g t h e E)i s t r i c t o f O r i g i n ( i ) a n d t h e D i s t r i c t o f D e s t i n a t i o n ( j ) , D.. ( i n c o l u m n ' ' c ' ) "IT ' J A l l data included in the columns, ' a , ' 'b ' and ' c ' of the tables have been ca lcu lated by the author. See Chapter 4 for d e t a i l s . I 106 As there are as many as 99 tables only a few are included in the text while the rest are attached in Appendix V. However, the analyses of a l l the tables are included in the text . The analys is of every table consists of four paragraphs. In the f i r s t paragraph the spat ia l pattern of migration streams is d iscussed. In the second paragraph the ' C F . ' loadings are interpreted in terms of the 'push-pu l l ' theory and the dominance of 'push' or ' p u l l ' factor in determining the spat ia l pattern of migration stream has been ascerta ined. In the th i rd paragraph the e f fec t of distance on the migration pattern is determined from the ' C F . ' , and in the las t paragraph the factor score pattern of the subjects is compared with the actual spat ia l pattern of migration streams to determine the extent to which the system that has been recognized in the ' C F . ' to be regulat ing the flow of migrants contributes to the 'explanat ion' of the actual migration pattern . This arrangement of analys is of the tables enables us to meet our two main object ives : (a) to describe the spat ia l pattern of migration streams of Bangladesh on a d i s t r i c t l e v e l , (h) to test the v a l i d i t y of the hypotheses 1, 2, 3 and 4 set * out in Chapter 2. * It should be noted that as the data matrix contains only one var iab le representing 'urban' s i t u a t i o n s , the analys is may not indicate f u l l y the ' p u l l ' of the 'urban' 107 5.2 I d e n t i f i c a t i o n s and Def in i t ions of the 'push' and  'pu l l ' Factors in the ' fac tor models' In the present study the 'push' and 'pu l l factors were i d e n t i f i e d in terms of the magnitude and nature of the ' c r i t i c a l f a c t o r ' loadings on the independent var iab les . For example, to c i t e a hypothetical case, i f loadings were s i g n i f i c a n t ( level of s ign i f i cance > .4000) on both the ve loc i ty of migrat ion, U. . , and the index of the population pressure on a g r i c u l t u r a l land, RD/SQMCA, we would say that the population pressure on a g r i c u l t u r a l land acted as the 'push' factor in that p a r t i c u l a r case. The nature of load-i n g , pos i t i ve or negative, however, would vary depending on whether the 'push' factor had been defined for in-migration streams or out-migration streams. In the case of out-migra-t ion streams loadings on both 'the ve loc i ty of out-migration streams,' OV, . , and the rat io RD/SQMCA should be of the * 3 d i s t r i c t s . Thus the scope for the test of the v a l i d i t y of the Hypotheses 2 and 3 would be rather l i m i t e d . However, as stated e a r l i e r , Bangladesh is overwhelmingly rural and more than 90% of migration took place within rural d i s t r i c t s , both in 1951 and 1961, the thrust of the hypotheses test ing may be put on the R-R streams (out and i n ) . The analys is has been performed with a l l the six kinds of streams since the emphasis in th is study is to make an attempt to understand the factors of migration process in i d t o . Unlike previous s tud ies , th i s study attempts to analyse a l l the d i f f e r e n t kinds of streams and then lays down the basis for der iv ing conclusions based on two kinds of R-R streams. Thus the resu l ts of th is analys is cannot be termed as ' p a r t i a l . ' 108 same nature. In other words, i f the loading on OV.. was • I j negative i t should be negative also on RD/SQMCA and vice versa. If loadings were pos i t ive on both, i t would ind icate that the greater the population pressure on the i th d i s t r i c t , the greater would be the volume of out-migration from that d i s t r i c t and vice versa. It would be interpreted in the same way i f the loadings were negative on both. In case of in-migrat ion streams, loadings on 'the ve loc i ty of i n - m i g r a t i o n , ' I V . . , and on RD/SQMCA should be of opposite nature. In other words, i f loading was pos i t i ve on I V . . , i t should be negative on RD/SQMCA or any other var iab les which acted as the 'push' f a c t o r s . S i m i l a r l y i f the loading was negative on 11/^ i t should be p o s i t i v e on the 'push' factor var iables l i k e LLALF/TALF. In th is case, i t would ind icate that as the population pressure was greater in the d i s t r i c t s of o r i g i n ( i th d i s t r i c t s ) than in the d i s t r i c t of dest inat ion ( j th d i s t r i c t ) people were 'pushed' from the i t h d i s t r i c t s and came to the j t h d i s t r i c t . Following this schema i t was found that both the population pressure on a g r i c u l t u r a l land, RD/SQMCA, and the presence of too many landless a g r i c u l t u r a l labourers , LLALF/TALF, acted as the 'push' factors in the present study (see Table 5 .3) . On the other hand, the p o s s i b i l i t i e s for further expansi of cu l t i va ted land due to low intens i ty of cropping, TCA/NCA, in the j t h d i s t r i c t s w i l l act as a ' p u l l ' f a c t o r . The 1 09 a v a i l a b i l i t y of more land under r i c e , Rice/TCA, and j u t e , Jute/NCA, in the j t h d i s t r i c t s w i l l also act as the ' p u l l ' f a c t o r s . They were defined as ' p u l l ' factors when they were found to be s i g n i f i c a n t l y related with OV^. or IV^. and the nature of loading was of a ' spec ia l t ype . ' For example, in the case of out-migration streams, i f the loading was pos i t i ve on OV. . , i t should be p o s i t i v e , l e t us say, on TCA/NCA and vice versa. But in the case of in-migrat ion streams, i f the loading on IV^. was pos i t i ve i t should be negative on TCA/NCA and vice versa. The expected nature of loading on var iables l i k e Ri'ce/TCA and others to be ca l led as ' p u l l ' factors are shown in Table 5.3. The magnitudes of loading on these var iables were taken as the c r i t e r i a to determine whether 'push' or ' p u l l ' was stronger. To c i t e a hypothetical ease, i f i t was found that loadings on, say, TCA/NCA and RD/SQMCA were .7035 and * .8231 respect ive ly when the loading on OV., was high and p o s i t i v e , we would in terpret that the 'push' was stronger than the ' p u l l . ' Since the sections discussing the i n t e r - d i s t r i c t migration streams within each kind of migration are quite long (Sections 5 .3-5.14) , the reader may go through the summary of each sect ion before going through the deta i l s of In the case of IV^. the nature of loadings on TCA/NCA and RD/SQMCA w i l l be d i f f e r e n t . 110 each of the sub-sections within the sect ions . The summaries are presented in the l a s t sub-section of each of the sect ions . 5.3 Determinants of the Spat ia l Pattern of Rural to Rural  Out-Migration Streams, OV.. . , 1951 5.3.1 Determinants of the Spat ia l Pattern of Out- Migration from Barasal . The majority of out-migrations from B a r i s a l , in 1951, chose adjoining d i s t r i c t s as the i r places of dest inat ion (see Table 5.4, column ' a ' ) . Among them Khulna was the most popular followed by Noakhali and Faridpur. In terms of OV. . , however, the place of Khulna remained unchanged, but the far away d i s t r i c t s of Rajshahi and Rangpur replaced Noakhali and Faridpur as the second and t h i r d largest receivers of people from B a r i s a l . This change i s i l l u s t r a t e d in columns ' a ' and 'b ' of Table 5.4. Examining the factor loadings on the ' c r i t i c a l f a c t o r ' of the factor matrix i t may be interpreted that both 'push' and ' p u l l ' factors were operat ing, in 1951, though the 'push' factors were dominant. The population pressure on a g r i c u l t u r a l land and the presence of more land-less a g r i c u l t u r a l labourers in Bar isa l than in a given j t h *The factor loading on OV.. in the ' C . F . ' was - .9117. Table 5.4 R-R OUT-MIGRATION STREAMS, 1951 SPATIAL PATTERN OF O V u FROM BARISAL, 1951 Out-Migration from the i t h D i s t r i c t " Barisal n II a II c i. D i s t r i c t s of Dest inat ion ( j th D i s t r i c t s ) No. of Migrants in the Stream Veloc i ty of Out-Migration Streams Di stance Separati ng i and j ( in mi les) Factor Scores on the Subject from the ' C r i t i c a l Factor 1 Mi j Rank Bogra 245 C h i t t . H.T. 94 (12) Comi11a 1 ,169 (5) Fari dpur 6,141 (3) J es sore 1 ,968 (4) Khulna 24,864 (1) Mymensingh 1 ,086 (6) Noakhali 9,890 (2) Pabna 381 (10) R a j s h a h i 737 (8) Rangpur 798 (7) Sylhet 577 (9) OVij Rank 0022 0018 0015 0026 0023 0138 0021 0028 0008 0038 0030 0010 7 9 (10) (5) (6) (1) (8) (4) (12) (2) (3) (11) Di j Fj Rank 175 0.4744 (9) 132 0.3876 (8) 88 -0.2454 (4.) 55 0.0274 (6) 87 0.9856 (11) 50 -2.6283 (1) 1 50 0.3s CO (7) 62 -0.6663 (2) 1 25 -0.1350 (5) 175 -0.5044 (3) 220 0.6767 (10) 175 1 .1875 (12) Factor 1 C r i t i c a l Factor ( C F . ) OVij -0.91175 RD/SQMCA -0.81359 LLALF/TALF -0.71517 Dij 0.69358 Rice/TCA 0.59998 N. B. Only var iab les having s i g n i f i c a n t loadings are mentioned here. (S ign i f i cance leve l is > .4000). C r i t i c a l factor has been defined as the fac to r having the highest loading on OVij or I V i j . Source: Ca lcu lated by the author. 112 d i s t r i c t , (RD/SQMCA (-.8135) and LLALF/TALF (- .7151), acted as the 'push' f a c t o r s . On the other hand, people were ' p u l l e d ' to those d i s t r i c t s where land under r i c e , Rice/TCA (0.5999), was more than in B a r i s a l . The nature of factor loadings on OV^. (-.9117) and D.. (-.6935) suggested that people went to those d i s t r i c t s , in greater number, which were nearer to B a r i s a l . The factor score pattern on the receiv ing d i s t r i c t s suggested that in the majority of the d i s t r i c t s , the ve loc i ty of migration from Bar isa l was as expected. This can be seen by comparing rank order of the d i s t r i c t s in columns ' d ' and 'b ' of the same tab le . 5.3.2 Determinants of the Spatial Pattern of Out- Migration from Bogra As indicated by column ' a ' in Table 5.5, neighbour-ing d i s t r i c t s of Rangpur, Rajshahi and Pabna received the l i o n ' s share of the migrants from Bogra. By examining the factor loadings in both factor 1, which is the c r i t i c a l factor ( C F , ) , aarid factor 3, i t may be said that both 'push' and ' p u l l ' factors were operating but ' p u l l ' factors were dominant in the C F . In the other Loading on OV.. in the ' C F . 1 was .6216. 113 f a c t o r , i . e . factor 3, i t was ' p u l l ' only. No 'push' factors were operat ing. From the C F . i t was found that while migratns were ' p u l l e d ' to those d i s t r i c t s where land under j u t e , Jute/NCA (- .6899), was greater than in Bogra; the presence of more landless a g r i c u l t u r a l labourers in Bogra than in a given j t h d i s t r i c t , LLALF/TALF (.4724), acted as a 'push. ' From the factor 3 i t was also found that probably some of the migrants were "pul led" to those d i s t r i c t s where the in tens i ty of cropping was low, TCA/NCA (- .5138), as there was more opportunit ies for further expansion of cu l t i vated area. D.. had a var iab le r o l e . It was s i g n i f i c a n t in factor 3 (-.9287) but not in the C F . Thus i t may be i n t e r -preted that distance was an important consideration for those who were looking for cu l t i va ted land (supposedly for c u l t i v a t i n g crops other than j u t e , such as r i c e but not for those who were looking for jute c u l t i v a t i o n only. The factor score pattern of the j th d i s t r i c t s revealed that while in the major receiv ing d i s t r i c t s the ve loc i ty of migration was as expected, in others i t was not (compare rank orders in columns "d" and "b") . 114 5.3.3 Determinants of the Spatial Pattern of Out- Migration from Chittagong H.T. The number of people coming out from Chittagong H in 1951 was so small that i t was thought not be worthwhile to run factor analys is with the migration f igures (only 99 people came out from that d i s t r i c t in 1951). 5.3.4 Determinants of the Spat ia l Pattern of Out- Migration from Comill a As indicated by column ' a ' in Table 5.6, the d i s -t r i b u t i o n of out-migration from Comilla was quite widely spread ,though the majority of the migrants went to the contiguous d i s t r i c t s . Far away d i s t r i c t s l i k e Rajshahi , Bar isa l and Rangpar also received respectable numbers of migrants. By interpret ing the factor loadings in the ' C F . , ' of the factor matrix, i t was found that only "push" factors were operat ing. The people were 'pushed' from Comilla i f the presence of landless a g r i c u l t u r a l labourers , LLALF/TALF (.6835), was more than in a given j th d i s t r i c t . The nature of factor loading on d is tance , D.. . (- .6627), suggested that people had the tendency to go to t nearer j th d i s t r i c t s in greater number. Loading on OM. . in the C F . was .6371 . 11 5 The factor score pattern of the j t h d i s t r i c t s sug-gested thatthough in the majority of the d i s t r i c t s the ve loc i ty of migration was as expected, in some d i s t r i c t s l i k e Rajshahi, Rangpur and Pabna i t was more than expected (compare rank orders in columns "d" and "b") . 5.3.5 Determinants of the Spt ia l Pattern of Out- Migration from Faridpur Table 5.7 indicates that though neighbouring d i s t r i c t s received the majority of out-migrants from Faridpar quite a large number of people did go to non-neighbouring di s t r i c t s . On 0V..J factor loading was s i g n i f i c a n t in two * f a c t o r s . From the ' C F . 1 i t was observed that people were mainly ' p u l l e d ' to those j t h d i s t r i c t s where land under jute was more than in Far idpur, Jute/NCA (-.7370), but a c c e s s i b i l i t y , ZB (- .8820), appeared to be more important and distance was not s i g n i f i c a n t (.2819). From Factor 1, i t was found, however, that both 'push' and ' p u l l ' factors were operating but ' p u l l ' was dominant. People were ' p u l l e d ' to those j t h d i s t r i c t s which had greater ' l eve l of urbanizat ion, U. . (.9209), than Far idpar; while the presence of more landless Loading on OV.. in the ' C . F . ' was - .6836. 116 a g r i c u l t u r a l labourers in Far idpur, LLALF/TAL (.7041), than in a given j th d i s t r i c t acted as 'push . ' In th is factor a l s o , distance was not s i g n i f i c a n t . The factor score pattern revealed that in as many as six rece iv ing d i s t r i c t s the ve loc i ty of migration was not as expected (compare rank orders in columns ' d ' and ' b ' ) . This confirmed the operation of other var iables also in addit ion to the var iables that were found to be s i g n i f i c a n t l y re lated with OV.. in the 1 C . F . 1 i J 5.3.6 Determinants of the Spat ia l Pattern of Out-Migration  from Jessore The bulk of the but-migrations from Jessore, as indicated by Table 5.8, column ' a ' , went to two continguous d i s t r i c t s , namely, Khulna and Jessore. * It appeared from the ' C F . , ' that the migrants were pr imar i ly looking for land to c u l t i v a t e r i c e , Rice/TCA (-.7553), and they were ' p u l l e d ' to those d i s t r i c t s where the land for r i c e c u l t i v a t i o n was ava i lab le more than in Jessore. It Was also found that the presence of more land-less a g r i c u l t u r a l labourers in Jessore than in any given j t h d i s t r i c t , LLALF/TALF (.5987) acted as a 'push , ' but the 'push' was milder than the ' p u l l . ' *Loading on 0V i . in the ' C . F . ' was - .8586. 117 D.. (.7910)' was so re lated with OV.. that i t suggested that the tendency of the migrants was to go to the nearer d i s t r i c t s of des t inat ion . The factor score pattern of the j th d i s t r i c t revealed that in the majority of the d i s t r i c t s , the ve loc i ty of migration from Jessore was as expected (compare rank orders in columns ' d ' and ' b ' ) . 5.3.7 Determinants of the Spatial Pattern of Out- Migration from Khulna As indicated in Table 5.9, column ' a , 1 the bulk of the out-migrants went to the continguous d i s t r i c t s . In other words, the d i s t r i b u t i o n of migrants from this d i s t r i c t was skewed towards the adjacent d i s t r i c t s ind icat ing a strong inf luence of d is tance. * The loadimg c h a r a c t e r i s t i c s in the ' C F . ' con-firmed the e a r l i e r notion of the strong inf luence of distance ( D . . , .8473). It also suggested that no ' p u l l ' or 'push' factors were regulat ing the flow of migrants from Khulna since loading was not s i g n i f i c a n t in any of the explanatory var iables except D • . . The factor score pattern of the j th d i s t r i c t s revealed that the v e l c o i t y of migration was as expected in *Loading on 0V^. in the ' C . F . ' was - .8463. 118 the majority of the d i s t r i c t s (compare rank orders in columns ' d ' and ' b ' ) . 5 ? "5.3?8 Determinants of the Spatial Pattern of Out- Migration from Mymensingh The majority of the migrants, as indicated by Table 5.10, column ' a , 1 coming out from Mymensingh went to neighbouring d i s t r i c t s of Rangpur, Sylhet , Comilla and Pabna. * It appeared from the ' C . F . ' that the migrants were "pul led" to those d i s t r i c t s where there were more opportunit ies for the expansion of cu l t i vated land, TCA/NCA (-.7496), due to the low in tens i ty of cropping Mymensingh is the largest producer of jute in Bangladesh, both in quantity and q u a l i t y . Pos i t ive loading on j u t e , Jute/NCA (.5074), while i t was negative on OV^ ^ , indicated that i t did not encourage people (at least those who were c u l t i v a t i n g jute) to go out. D.. was marginally s i g n i f i c a n t (.4209) in the ' C F . ' suggesting that distance had a mild e f fect on OV. . (- .8873). The factor score pattern of the j t h d i s t r i c t s suggested that in the majority of the nectepient d i s t r i c t s , Loading on OV. . in the ' C F . ' was -.8873. 119 the ve loc i ty of migration from Mymensingh was not as expected (compare rank orders in columns ' d ' and ' b ' ) . 5.3.9 Determinants of the Spatial Pattern of Out- Migration from Noakhali Table 5.11, column ' a , ' indicates that out-migrants from Noakhali moved in a l l d i rect ions and at a l l distances at quite a s i g n i f i c a n t rate . It implied that distance did not play a very c ruc ia l ro le in th is case. * It appeared from the ' C F . ' that people were ' p u l l e d ' to those d i s t r i c t s where more opportunit ies were ava i lab le for the expansion of the tota l cu l t ivated area, TCA/NCA (- .5847), due to low in tens i ty of cropping and land under r i c e , Rice/TCA (.9448), were more than in Noakhali . There was no 'push , ' however, since loading on var iables l i k e RD/SQMCA or LLALF/TALF was not s i g n i f i c a n t . S i g n i f i c a n t loading on D.. (-.5847) disproved the e a r l i e r notion of the non-importance of distance in inf luencing the flow of migratns from this d i s t r i c t . The factor score pattern of the j t h d i s t r i c t s suggested that in most of the d i s t r i c t s , the ve loc i ty of migration from Noakhali was not as expected (compare rank orders in columns ' d ' and ' b ' ) . *Loading on OV^. in the ' C . F . ' was -.7761 . 1 20 5.3.10 Determinants of the Spatial Pattern of Out- Migration from Pabna D i s t r i c t s adjoining to Pabna such as Rajshahi , Bogra, Mymensingh and Faridpur received the majority of out-migrants from Pabna (Table 5.12). The factor loading c h a r a c t e r i s t i c s on OV.j (- .8906), D . . (.8801 ) and ZB (.6724) suggested that people went to those d i s t r i c t s , in greater number, which were nearer to i t and which had better a c c e s s i b i l i t y . It was also found that the greater ' l eve l of urbanizat ion ' of a given j t h d i s t r i c t r e l a t i v e to the d i s t r i c t of o r ig in ( i . e . Pabna), U, . (.4305), J J acted as a ' p u l l ' to the migrants from Pabna but the ' p u l l ' was mi ld . There was, however, no 'push. ' The factor score pattern of the j t h d i s t r i c t s indicated that in the majority of the d i s t r i c t s , the ve loc i ty of migration from Pabna was as expected (compare rank orders in columns ' d ' and ' b ' ) . 5.3.11 Determinants of the Spat ia l Pattern of Out- Migration from Rajshahi Though the majority of the migrants went to the two d i s t r i c t s , " B o g r a and Pabna, contiguous to Rajshahi; migrants did also go to far f lung d i s t r i c t l i k e Khulna in respectable numbers (Table 5.13, column ' a ' ) . 121 From the ' C . F . ' i t appeared that the a v a i l a b i l i t y of land under r i c e in a given j t h d i s t r i c t more than in Rajshahi, Rice/TCA (.7737), acted as a ' p u l l ' to the migrants. There was no 'push . ' Distance, D.. (.3386), was not a s i g n i f i c a n t factor ind icat ing that the ve loc i ty of out-migration from Rajshahi was independent of distance separating the i th and j t h di s t r i c t s . The factor score pattern of the j th d i s t r i c t s indicated that the expected ve loc i ty of migration from Rajshahi was only in the major receiv ing d i s t r i c t s (compare rank orders in columns ' d ' and 1 b 1 ) . This indicated that the marginal d i s t r i c t s might have been inf luenced by var iables other than which are included in our ana lys i s . 5.3.12 Determinants of the Spatial Pattern of Out- Migration from Rangpur Out-migrants from Rangpur, as indicated in Table 5.14, column ' d , ' did not move very far in great numbers and have chosen the i r places of dest inat ion within neighbouring di s t r i c t s . ** From the ' C F . 1 i t appeared that both 'push' and 1 p u l 1 1 factors were operating but 'push' factors were * Loading on in the ' C F . ' was - .7817. Loading on OV,. . in the ' C F . ' was - .7190. 1 22 dominant. While the population pressure on a g r i c u l t u r a l land, RD/SQMCA (-.7182) acted as a 'push , ' people were ' p u l l e d ' to those j th d i s t r i c t s where the ' leve l of urbani-z a t i o n , ' U. . (.4168), was more than in Rangpur but the pul l was not very strong. Since loading on d is tance , D.. (.8522), was high and s i g n i f i c a n t i t indicated that the migrants tended to go to the nearer d i s t r i c t s in greater numbers to minimize the distance t r a v e l l e d . The factor score pattern of the j t h d i s t r i c t s suggested that in most of the d i s t r i c t s the ve loc i ty of migration from Rangpur was as expected (compare rank orders i n columns ' d ' and ' b ' ) . 5.3.13 Determinants of the Spatial Pattern of Out- Migration from Sylhet Though Mymensingh — a d i s t r i c t having a common border with Sy lhet , received the largest number of people, migrants from this d i s t r i c t went to other d i s t r i c t s a l s o , in good numbers, i r respect ive of the distance involved (Table 5.14, column' 'b ' and ' c ' ) . * From the ' C F . , ' i t appeared that the population pressure on a g r i c u l t u r a l land, RD/SQMCA (- .7934), acted as Loading on OV.. in the ' C . F . ' was - .9457. 123 the 'push' factor while migrants were ' p u l l e d ' to those d i s t r i c t s where the scope for further expansion of c u l t i v a b l e land, TCA/NCA (- .7004), due to low in tens i ty of cropping was more than in Sylhet . The 'push' f a c t o r , however, was stronger than the ' p u l l ' f a c t o r . Distance, D.. (.1536), was not s i g n i f i c a n t , i n d i c a t -i i J ing that the out-migrantion from Sylhet was independent of distance separating the i th and j t h d i s t r i c t s . The factor score pattern of the j th d i s t r i c t s revealed that in most of the d i s t r i c t s , the v e l o c i t y of migration from Sylhet was expected (compare rank orders in columns ' d ' and ' b ' ) . 5.3.14 Summary of Section 5.3 From the discussion of the spat ia l pattern of out-migration streams or ig inat ing from 13 d i f fe rent rural d i s t r i c t s within the Rural-Rural Out-Migration stream, i t was discovered that the R-R out-migration was e s s e n t i a l l y a local phenomenon and thus the spat ia l d i s t r i b u t i o n s of migration streams were skewed towards the nearer d i s t r i c t s . The bulk of the migrants from a l l but 5 d i s t r i c t s namely, Comilla^ Far idpur, Noakhali , Rajshahi and Sylhet went to the contiguous and neighbouring d i s t r i c t s (see Table 5A and texts in sect ion 5.3) . 124 From the pr inc ipa l components type factor analys is i t was observed that both 'push' and ' p u l l ' factors were operating in determining the spat ia l pattern of out-migration streams. People were 'pushed' from the d i s t r i c t s of o r i g i n ( i th d i s t r i c t s ) i f the population pressure on a g r i c u l t u r a l land, RD/SQMCA, was more than in a given j t h d i s t r i c t , and the presence of landless a g r i c u l t u r a l labourers , LLALF/TALF, was more than in a given j t h d i s t r i c t . At the same time, i f the opportunit ies for further expansion of cu l t i va ted land, TCA/NCA, was more in the j t h d i s t r i c t s than in the i t h d i s t r i c t s due to the lower in tens i ty of cropping; i t acted as a ' p u l l ' f a c t o r . The a v a i l a b i l i t y of more land under r i c e , Rice/TCA, and j u t e , Jute/NCA, in the j th d i s t r i c t s also ' p u l l e d ' people away from the i th d i s t r i c t s . Thus, these f i v e v a r i a b l e s , out of nine explanatory var iables in the data matrix, were found to be most ac t i ve ly operating as e i ther 'push' or ' p u l l ' factors in determining the flow of migrants from the rural d i s t r i c t s to other rural d i s t r i c t s within the R-R out-migration stream. The p a r t i c u l a r 'mix' of 'push' and ' p u l l ' factors was, however, d i f fe rent in each of the i th d i s t r i c t s . While 'push' factors were dominant in determining the spat ia l pattern of out-migration from B a r i s a l , Comi l la , Rangpur and Sylhet; ' p u l l ' factors were dominant in others except Khulna where there was neither 125 any ' p u l l ' nor any 'push . 1 There distance was the only s i g n i f i c a n t var iab le (see texts in sect ion 5.3 and Table 5A). Distance had a var iab le ro le in determining the spat ia l pattern of out-migration streams or ig inat ing from 12 d i s t r i c t s . It was s i g n i f i c a n t l y re lated with migration ve loc i ty in 9 cases. Thus, on an average, i t palyed an important r o l e . The extreme case, as stated before, was for the migration streams or ig inat ing from Khulna. There distance was the only determinant of the migration pattern. There were neither any ' p u l l ' nor any 'push . ' In the migration streams or ig inat ing from Jessore, Pabna and Rangpur both 'push-pu l l ' factors and distance were operating but distance was the most important f a c t o r . The opposite also happened to the migration streams or ig inat ing from Bogra, Far idpur, Rajshahi and Sylhet , where distance was found to be of no importance at a l l . However, a d i f fe rent p icture emerged when simple regression technique was employed with OV. . and D . . . It was discovered that distance was s i g n i f i c a n t l y re lated with OV.. ( i . e . r > .4000) in determining the spat ia l pattern of migration streams or ig inat ing from a l l the rural d i s t r i c t s . It may be of in teres t to note that , D.. was p o s i t i v e l y re lated with OV.. in the cases of Bogra and Pabna (see Figure 5.1) . In other words, for these two d i s t r i c t s , Table 5A DETERMINANTS OF THE SPATIAL PATTERN OF R-R OUT-MIGRATION STREAMS, 19511 a b c d e . . . OV. . from the p i t h d i s t r i c t s Was the spat ia l di s t r i bution skewed towards the nearer di s t r i cts? Which of the factors were dominant -'push 1 or ' pull 1 ? Was distance s ign i f i cant? Was the ve loc i ty of migration as expected to the majority of j th d i s t r i c t s ? 1. Bar isal Yes Push (RD/SQMCA and LLALF/TALF) Yes Yes 2. Bogra Yes Pull (Jute/NCA) No * No 3. Comilla No Push (LLALF/TALF) Yes Yes 4. Faridpur No Pull (Jute/NCA) No * No 5. Jessore Yes Pull (Rice/TCA) Yes Yes ^Jhe columns re fer only to the in terpretat ion of the ' C . F . ' 2 Factor analys is has not been performed for Chittagong H.T. (see subsection 5.3.3 in section 5 .3) . OV.. . to the top rec ip ient d i s t r i c t s was as expected (see section 5.3 or Tables 5.5, 5.7, 5.10, 5.11 and 5.1-3). CONTINUED ro Table 5A (Continued) a b c d e OV. . from the i J i t h d i s t r i c t s Was the spat ia l di s t r i buti on skewed towards the nearer di s t r i cts? Which of the factors were dominant -'push ' or 1 pul1'? Was distance s ign i f i cant? Was the v e l o c i t y of migration as expected to the majority of j t h di s t r i c t s ? £6. Khulna Yes No push or pull Yes (Distance was the only s i g -n i f i c a n t vari abl e) Yes 7. Mymensingh Yes Pull (TCA?NCA) Yes * No 8. Noakhali No Pull '(Ri ce/TCA & TCA/JNCA) • "No * No 9. Pabna Yes Very mild pull (U . . ) , no push Yes Yes 10. Rajshahi No Pull (Ri ce/TCA) No * No 11. Rangpur Yes Push (RD/SQMCA) Yes Yes 12. Sylhet No Push (RD/SQMCA) No Yes Figure 5.1. 'Distance E l a s t i c i t y ' of R-R Out-Migrations Streams, 1 951 . 1 29 contrary to the normal expectat ions, the ve loc i ty of out-migration increased as the distance between the i th and j th d i s t r i c t s increased. From the factor score pattern of the subjects ( j th d i s t r i c t s ) i t was found that the ve loc i ty of migration was as expected to the majority of the recepient d i s t r i c t s from seven out of twelve d i s t r i c t s of o r i g i n such as B a r i s a l , Comi l la , Jessore, Khulna, Pabna, Rangpur and Sy lhet . It may be noted that even in the rest of the d i s t r i c t s where OV.. to the majority of the j th d i s t r i c t s was not as expected; in the top three or four j th d i s t r i c t s the OV. . was as expected (see Tables 5.5, 5.7, 5.10, 5.11, and 5.13). As in p r a c t i c e , the top three or four j th d i s t r i c t s received the majority of the migrants from each of the i th d i s t r i c t s , i t may be stated that the var iables that were s i g n i f i c a n t l y re lated with OV.. in the ' C . F . ' 'expla ined' a high proportion of the tota l variance of the spat ia l pattern of migration streams or ig inat ing from al1 the twelve rural d i s t r i c t s and going to the other rural d i s t r i c t s (see sect ion 5 .3) . 5.4 Determinants of the Spat ia l Pattern of Rural To Rural  In-Migration Streams, IV i . , 1951 5.4.1 Determinants of the Spatial Pattern of Iri- Migration to Bar isal As indicated by Table 5.16, columns ' a 1 and ' c , ' most of the in-migrants to Bar isal were from neighbouring 130 d i s t r i c t s of Far idpur, Noakhali and Comi l la . Looking at the magnitude and nature of factor * loading in the ' C F . , ' i t may be in ferred that greater population pressure on a g r i c u l t u r a l land in a given i t h d i s t r i c t than in B a r i s a l , RD/SQMCA (.8739), acted as the 'push' f a c t o r , whereas people were ' p u l l e d ' to Bar isal i f the in tens i ty of cropping was less than in a given i t h d i s t r i c t , TCA/NCA (.5600), as there was more scope for further expansion of cu l t i vated land. But the ' p u l l ' was mi 1der. Distance was s i g n i f i c a n t , D. . (.8752), and people * 3 preferred to move to the nearer d i s t r i c t s . The factor score pattern of the i th d i s t r i c t s indicated that the ve loc i ty of migration was not as expected to Bar isal from the major sending d i s t r i c t s (compare rank orders in columns ' d ' and ' b ' ) . 5.4.2 Determinants of the Spatial Pattern of In- Migration to Bogra As indicated by Table 5.17, column ' a ' and ' b , ' the d i s t r i c t s which are contiguous to Bogra send the la rgest number of migrants to th is d i s t r i c t . Loading on IV . . in the ' C . F . ' was - .8445. Table 5.17 R -R I N - M I G R A T I O N S T R E A M S , 1 9 5 1 S P A T I A L PATTERN OF I V U TO BOGRA , 1 9 5 1 D i s t r i c t s of Or ig in ( i th D i s t r i c t s ) D i s t r i c t of Destination "a" No. of Migrants in the Stream "b " Veloc i ty of Out-Migration Streams (jth D i s t r i c t ) ~ Bogra "c" Separati ng Distance i and j , (in miles) "d Factor Scores on the Subjects from the ' C r i t i c a l Factor ' IMij Rank IVi j Rank D i J Bar isa l C h i t t . H.T Comi11 a Faridpur Jessore Khulna Mymens i ngh N o a k h a 1 i Pabna Raj shahi Rangpur Sylhet 245 23 435 546 204 100 2,025 472 12,702 11,569 3,879 93 (8) (12) (7) (5) (9) (10) (4) (6) (1) (2) (3) (11) 0022 0024 0037 0064 0041 0016 001 5 0075 2639 1726 ,0437 ,0019 (10) (9) (8) (6) (7) (12) (4) (5) (1) (2) (3) (11) 175 250 120 113 105 175 75 175 45 42 50 150 Fj r 0.8164 •0.3514 0.7130 0.7003 0.8534 •0.8680 •0.2036 •0.0645 •2.0217 •1 .3085 0.4318 1.3028 Rank (10) (4) ( 9 ) (8) (11) (3) (5) (6) (1) (2) (7) (12) Factor 1 ( ' C F . ' ) IV i j Ri ce/TCA Dij-TCA/NCA •0.8444 .0.9558 0.2935 (not s i g n i f i c a n t ) 0.7678 1 32 * By analyzing the ' C F . , ' i t appeared that only the ' p u l l ' factor was in operat ion. There was no 'push . ' It appeared that people were ' p u l l e d ' to Bogra i f more land under, r i c e , Rice/TCA (-.9558), was ava i lab le there than in a given i th d i s t r i c t . People were also ' p u l l e d ' to th is d i s t r i c t i f the in tens i ty of cropping, TCA/NCA (.7678), was less than in a given i th d i s t r i c t as more scope was there for further expansion of cu l t ivated land. It was also found that the aforementioned factors were operating e f f e c t i v e l y i r respect ive of the distance involved in migration (loading on D.. . was .2935 and thus not s i g n i f i c a n t ) . The factor score pattern revealed that , except from the two top sending d i s t r i c t s , the ve loc i ty of migration to Bogra was not as expected from the other i t h d i s t r i c t s (compare rank orders in columns ' d ' and ' b ' ) . 5.4.3 Determinants of the Spatial Pattern of In- Migration to C h i t t . H.T. Factor analysis has not been performed with the in-migration of Chittagong H.T. since the number of people coming into that d i s t r i c t was found to be very smal l . Loading on IV . . in the ' C . F . ' was - .8445. 1 33 5.4.4 Determinants of the Spat ia l Pattern of In- Migration to Comilla As indicated by Table 5.18, columns ' a ' and ' c , ' the d i s t r i b u t i o n of d i s t r i c t s sending migration to Comilla was comparatively more even than the other receiv ing d i s t r i c t s in the R-R in-migration streams, though as usual , the majority of migrants did come to Comilla from neighbouring d i s t r i c t s . * From the ' C . F . ' i t may be in ferred that the presence of more people per square mile of cu l t i va ted area in a given i th d i s t r i c t , RD/SQMCA (- .4580), than in Comilla acted as a 'push' factor (loading on OV^. was - .8837) . There was no ' p u l l . ' Distance was the most important factor and the majority of the migrants tend to come from the nearer d i s t r i c t s (loading on D .^ was 0.7711)) From the factor score pattern of the i th d i s t r i c t s i t was found that the ve loc i ty of migration to Comilla was as expected from almost a l l the sending d i s t r i c t s (compare rank orders in columns ' d ' and ' b ' ) . 5.4.5 Determinants of the Spatial Pattern of Ih- Migration to Faridpur The number of migrants coming to Faridpur from a l l the neighbouring d i s t r i c t s was comparatively more than * Loading on IV^ in the ' C F . ' was - .8410. 1 34 from the far away d i s t r i c t s (Table 5.19, column ' a ' and * Analyzing the ' C F . , ' i t could be inferred that there was neither any 'push 1 nor any ' p u l l ' in the system determining the flow of migrants from other rural d i s t r i c t s . Only d istance, D.. (.7748) was the s i g n i f i c a n t factor and the majority of the people tend to come from the neighbour-ing d i s t r i c t s . The factor score pattern of the i th d i s t r i c t s further indicated that D .^ did act strongly as a factor in determining the spat ia l pattern of in-migration to Faridpur from other rural d i s t r i c t s (compare rank orders in columns 'b ' and ' d ' ) . 5.4.6 Determinants of the Spat ia l Pattern of In- Migration to Jessore As indicated by Table 5.20, columns ' a ' and ' c , ' though the majority of the migrants came to Jessore from the contiguous d i s t r i c t s of Khulna and Far idpur, s i g n i f i c a n t numbers of people also came from more d istant d i s t r i c t s such as Noakhali and Comi l la . *Loading on IV . . in the ' C . F . ' was - .9319. 135 Looking at the nature of loadings in the ' C . F . ' i t may be said that both 'push' and ' p u l l ' factors were operating but the 'push' was stronger. It appeared that population pressure on agr i cu l tu ra l land, RD/SQMCA (.7927), and the presence of more landless a g r i c u l t u r a l labourers in a given i th d i s t r i c t than in Jessore, LLALF/TAL (.5268), acted as the 'push' f a c t o r s . On the other hand, people here ' p u l l e d ' to Jessore i f more scope was there to expand cu l t ivated due to the lower in tens i ty of cropping, TCA/NCA (.5094) than in any given i t h d i s t r i c t . Distance, D,. (.7183), was s i g n i f i c a n t and the ' 3 migrants tend to prefer to come from the adjoining d i s t r i c t s . From the factor score pattern i t was revealed that the ve loc i ty of migration to Jessore was as expected from the majority of i th d i s t r i c t s (compare rank orders in columns ' d ' and ' b ' ) . 5. J5..4.7 Determinants of the Spatial Pattern of In- Migration to Khulna From Table 5.21, columns ' a ' and ' c , ' i t could be found that the majority of the migrants came to th is d i s t r i c t from the neighbouring d i s t r i c t s . Since in the ' C F . , ' there was a s i g n i f i c a n t loading only on IV, . (-.9389) and D. . (.8710), i t may be J 3 *Loading on IV i n . in the ' C . F . ' was -.8524. 1 36 said that neither ' p u l l ' nor 'push' factors were operat ing. Distance was the only determining factor and people tend to come from the nearer d i s t r i c t s . The factor score pattern of the i th d i s t r i c t s revealed that the ve loc i ty of migration to Khulna was as expected from the top three sending d i s t r i c t s only (compare rank orders in columns 'b ' and ' d ' ) . 5.4.8 Determinants of the Spatial Pattern of In- Migration to Mymensingh As indicated by Table 5.22, column ' a , ' the highest number of migrants came to Mymensingh from Pabna. Comilla and Sylhet came next in order. A l l of these three d i s t r i c t s are neighbouring to Mymensingh. By examining factor loadings of the var iables in * the ' C F . ' of the factor matrix i t may be interpreted that migrants were ' p u l l e d ' to Mymensingh i f more land under r i c e , Rice/TCA (-.8052), was ava i lab le than in a given i th d i s t r i c t . There was no 'push. ' The nature of loading on distance, D.. (0.781), indicated that distance did not act as a constra int in determining the flow of migrants to Mymensingh. Loading on IV . . in the ' C . F . ' was - .5900. 137 The factor score pattern of the i th d i s t r i c t s indicated that the ve loc i ty of migration to Mymensingh was as expected from the loading sending d i s t r i c t s (compare rank orders in columns ' d ' and ' b ' ) . 5.4.9 Determinants of the Spat ia l Pattern of In- Migration to Noakhali As indicated in Tab fee. 5.23, column ' a , ' the highest number of migrants came to Noakhali from B a r i s a l . Comi l la , Faridpur and Sylhet came next in order. By examining factor loadings of the var iables in * the ' C . F . ' of the factor matrix, i t may be said that people were ' p u l l e d ' to Noakhali i f more land under r i c e , Rice/TCA (.8816), was there than in a given i t h d i s t r i c t . People were also ' p u l l e d ' to th is d i s t r i c t i f the in tens i ty of cropping was more than in a given i th d i s t r i c t , TCA/NCA (.7116), and thus more scope for gett ing work in the f i e l d s were there. The nature of loading on d istance, D.. (.4822), indicated that the migrants tend to come from the nearer d i s t r i c t s to minimize the distance t r a v e l l e d . In addit ion to the ' C . F . ' loading on IV ^  ^ was also s i g n i f i c a n t in Factor 1. Loading on IV . . in the ' C . F . ' was .6008. 138 The factor score pattern of the i th d i s t r i c t s revealed that the ve loc i ty of migration to Noakhali was not as expected from the majority of the sending d i s t r i c t s (compare rank orders in columns ' d ' and ' b ' ) . 5.4.10 Determinants of the Spatial Pattern of In- Migration to Pabna As indicated by Table 5.24, column 'a,' the highest number of migrants came to Pabna from Rajshahi. Mymensingh, Faridpur and Bogra came next. A l l of these d i s t r i c t s are neighbouring to Pabna. By examining factor loadings of the var iables in the ' C F . ' of the factor matrix i t may be interpreted that people were ' p u l l e d ' to Pabna i f the in tens i ty of cropping was l e s s , TCA/NCA (.7555), and as such more scope for the expansion of cu l t i va ted land was there than in a given i th d i s t r i c t . However, there was no 'push . ' The nature of loading on d is tance , D.. (.8311), indicated that i t was the most important factor in regulat ing the flow of migrants to Pabna. The migrants tend to come from the nearer d i s t r i c t s in greater number to minimize the distance t r a v e l l e d . Loading on IV. . in the ' C . F . ' was - .8561. 139 The factor score pattern of the i th d i s t r i c t s revealed that the ve loc i ty of migration to Pabna was as expected from a l l the three lending/sending d i s t r i c t s (compare rank orders in columns ' d ' and ' b ' ) . 5.3.11 Determinants of the Spat ia l Pattern of In- Migration to Rajshahi The highest number of people come to Rajshahi from Pabna, which is contiguous to i t . Bogra, Comilla and Mymensingh came next in order. Except Comilla the other two d i s t r i c t s are also neighbouring to Rajshahi (Table 5.25, columns 'a 1 and 'c ' ) . By examining factor loadings of the var iables in * the ' C F . ' 1 of the factor matrix i t may be interpreted that the migrants were ' p u l l e d ' to Rajshahi i f more land under r i c e , Rice/TCA (.8577), was ava i lab le and the rate of l i t e r a c y , L . . (.4098), was higher than in a given i th d i s t r i c t . There was no ' 'push . ' The nature of loading on d istance, D.. (- .2878), indicated that distance did not play an important ro le in determining the flow of migrants to Rajshahi. The factor score pattern of the i th d i s t r i c t s revealed that except for Comilla and Khulna, more or less Loading on IV . . in the ' C . F . ' was .8343. 140 the v e l o c i t y of migration to Rajshahi was as expected from a l l other sending d i s t r i c t s (compare rank orders in columns ' d ' and ' b ' ) . 5.4.12 Determinants of the Spat ia l Pattern of In- Migration to Rangpur As indicated by Table 5.26, columns ' a ' and ' c , 1 more than 40% of the total migrants came from Mymensingh, which is neighbouring to i t . Substantial numbers of migrants came from Bogra, Pabna and Noakhali . Except Noakhali the other two are also neighbouring to Pabna. By examining factor loadings of the var iables in * the ' C F . ' of the factor matrix i t may be interpreted that people were ' p u l l e d ' to Rangpur i f more land under j u t e , Jute/NCA (-.8084), was ava i lab le than in a given i th d i s t r i c t . There was no 'push . ' Since loadings on both a c c e s s i b i l i t y , £B (- .5566), and d istance, D.. (38662), were s i g n i f i c a n t , i t indicated that people came to Rangpur in greater number not only from those d i s t r i c t s which were nearer to i t but also from those which were more access ib le . The factor score pattern of the i th d i s t r i c t s revealed that except for Barisal and Noakhali , the ve loc i ty Loading on IV . . in the ' C . F . ' was - .7679. 141 of migration to Rangpur was as expected from almost a l l other sending d i s t r i c t s (compare rank orders in columns ' d ' and ' b ' ) . 5.4.13 Determinants of the Spatial Pattern of In- Migration to Sylhet As indicated by Table 5.27, column ' a , ' the highest number of migrants came to Sylhet from Mymensingh, which is contiguous to i t . In almost equal numbers of migrants came from another contiguous d i s t r i c t , Comi l la . Substantial numbers of migrants also came from Noakhali . By examining factor loadings of the var iables in * the ' C F . ' of the factor matrix i t may be said that neither any ' p u l l ' nor any 'push' factors were operat ing. Only d istance, D.. (.8972), was the determining f a c t o r . The nature of loading on distance indicated that the migrants tend to come to Sylhet from nearer d i s t r i c t s in greater number to minimize the distance t r a v e l l e d . The factor score pattern of the i th d i s t r i c t s revealed that except for a few d i s t r i c t s l i k e Bogra and Chittagong H.T. , the ve loc i ty of migration to Sylhet was as expected from a l l other sending d i s t r i c t s (compare rank orders in columns ' d ' and ' b ' ) . *Loading on IV. . in the ' C F . ' was - .9145. 142 5.4.14 Summary of Section 5.4 From the discussion of the spat ia l pattern of i n -migration streams coming in to 12' d i f fe rent rural d i s t r i c t s within the R-R In-Migration Streams, i t was observed that the bulk of the migrants came from the neighbouring d i s t r i c t s to a l l but three d i s t r i c t s , namely Jessore, Rangpur and Sylhet . In other words, R-R in-migration was e s s e n t i a l l y a local phenomenon and thus the spat ia l d i s t r i b u t i o n of i n -migration streams were skewed towards the neighbouring di s t r i c t s . When the loadings on the ' C . F . ' s of the factor matrices were looked at in terms of the 'push-pu l l ' theory, both 'push' and ' p u l l ' factors were found to be operating in determining the spat ia l pattern of in-migration streams coming to the rural d i s t r i c t s from other rural d i s t r i c t s . People were 'pushed' from the i th d i s t r i c t s i f the popula-t ion pressure on agr i cu l tu ra l land, RD/SQMCA, was found to be more than in the j th d i s t r i c t s (see Tables 5.16, 5.18 and 5.20 and subsections 5 .4 .1 , 5.4.4 and 5.4.6 in sect ion 5.4) . On the other hand, people were ' p u l l e d ' to the j t h d i s t r i c t s i f land under r i c e and/or j u t e , Rice/TCA and Jute/NCA, was more than in a given i th d i s t r i c t (see Tables 5.17, 5.22, 5.23, and 5.26 and subsections 5.4.2, 5 .4 .8 , 5 .4 .9 , 5.4.10, 5.4.11 and 5.4.12 in section 5 .4) . People were also ' p u l l e d ' to th j t h d i s t r i c t s i f the in tens i ty of 143 cropping, TCA/NCA, was less than in a given i th d i s t r i c t as there would be more opportunit ies for the expansion of cu l t i va ted land by ra i s ing more than one crop in a year ^ from the same piece of land (see Tables 5.18, 5.23 and 5.24 and subsections 5 .4 .2 , 5.4.9 and 5 .4 .10) . The ' p u l l ' factors were found to be dominant in s ix out of twelve d i s t r i c t s such as Bogra, Mymensingh, Noakhal i , Pabna, Rajshahi and Rangpur. In a l l of these six d i s t r i c t s except Bogra, ' p u l l ' was the only factor and there was no 'push' (see subsections dealing with these d i s t r i c t s in section 5.4 or Table 5B). At the same time there were also occasions where there was neither any 'push' nor any ' p u l l ' in determining the flow of in-migrants to such d i s t r i c t s l i k e Far idpur, Khulna and Sylhet (see sect ion 5.4 and Table 5B). In the rest of the three d i s t r i c t s such as B a r i s a l , Comilla and Jessore, 'push' factor was dominant in determining the spat ia l pattern of in-migrat ion to these d i s t r i c t s (Table 5B). Thus from the above discussions i t may be said that both 'push' and ' p u l l ' factors were operating but the ' p u l l ' was dominant in determining the overa l l spat ia l pattern of R-R in-migration streams in 1951. On the average, distance was an important factor and i t appeared that the migrants tended to minimize the distance t rave l led but i t had a var iab le ro le to play in Table 5B DETERMINANTS OF THE SPATIAL PATTERN OF IN-MIGRATION STREAMS, 1951 a b c d e I V i j t 0 Was the spat ia l d i s t r i b u t i o n skewed towards the nearer d i s t r i c t s ? Which of the factors were dominant -1 push ' or ' p u l l 1 ? Was distance s i g n i f i c a n t ? Was the ve loc i ty of migration as expected from the majority of i th d i s t r i c t s ? 1. Bar isa l Yes Push (RD/SQMCA) Yes * No 2. Bogra Yes Pull (Ri ce/TCA & TCA/NCA) No * No 3. Comi 11 a Yes A mild push (RD/ SQMCA), no pull Yes Yes 4. Faridpur Yes No push or pull Yes [(.Distance was the only s i g -n i f i c a n t factor) Yes 5. J essore Yes Push (RD/SQMCA) Yes Yes The ve loc i ty of migration was as expected CONTINUED from the top two or three i th d i s t r i c t s . -. - f a . Table 5B (Continued) a b c d e .IV,j to Was the spat ia l di s t r i bution skewed towards the nearer di s t r i cts Which of the factors were dominant -'push 1 or ' p u l l 1 ? Was distance s ign i f i cant? Was the ve loc i ty of migration as expected from the majority of i th di s t r i cts? 6. Khulna Yes No push or pull Yes (Distance was the only s i g -n i f i cant facotr) * No &'7u Mymensingh Yes Only pull , no push No * No 8. Noakhali Yes Only pull (Rice/ TCA & TCA/NCAO, no push yes * No 9. Pabna Yes Only pull (TCA/NCA), no push Yes * No 10. Rajshahi Yes Only pull (Rice/TCA), no push No Yes 11. Ra ngpur Yes Only pull (Jute/NCA), no push Yes Yes 12. Sylhet Yes No push or pull Yes (Distance was the only sigs ni f i cant factor) Yes o o LO OH o o 1 CO OH o LO CM OH o o CM oH o LO OH o o OH o LO o 0' 040 ,030 ,020 .010 • -0 146 \ \ \ \ S N \ — ^ = mm \ \ T — i — r - T -50 i i T -i—i—r 100 150 200 250 0ISTRNCE IN MILES. DIJ 300 T — i — i — ] — r 350 7 - l 400 Figure 5.2. 'Distance E l a s t i c i t y ' of R-R In-Migration Streams,1951. 147 determining the spat ia l pattern of in-migration streams coming to 12 d i s t r i c t s . In the ' C F . ' factor loading on D. . was s i g n i f i c a n t in the cases of B a r i s a l , Comi l la , Far idpur, Jessore, Khulna, Noakhali , Pabna, Rangpur and Sylhet (see sect ion 5.4 and Table 5B). Distance was found to be of no importance to migration streams coming to Bogra, Mymensingh and Rajshahi ( i . e . loading on D. . was not s i g n i f i c a n t in the 1 C . F . 1 descr ibing the spat ia l pattern of migration streams of these d i s t r i c t s ) . There were also occasions such as in the case of migration streams coming to Far idpur, Khulna and Sylhet where distance was the only important f a c t o r . There were neither any 'push' nor any ' p u l l . ' Whereas in migration streams coming to Comi l la , Pabna and Rangpur both 'push-pu l l ' factors and distance were operating but distance was the most important factor (see Figure 5B). However, when D.. was regressed against I V . . , c e t e r i s parebus, distance was found to be s i g n i f i c a n t l y re lated with IV . . in a l l the j t h d i s t r i c t s though in the case of Rajshahi the re la t ionsh ip was pos i t i ve (see Figure 5 .2) . The factor score pattern of the i t h d i s t r i c t s revealed that i n s>ix<;outof twel ve d i s t r i c t s such as Comi l la , Far idpur, Jessore, Rajshahi , Rangpur and Sylhet the ve loc i ty of migration was as expected from the majority of the sending d i s t r i c t s . But even in the other d i s t r i c t s the ve loc i ty of 148 migration from the top sending d i s t r i c t s was as expected (see sect ion 5.4) . V a l i d i t y of Hypothesis 1 Based on the resu l ts of the analysis discussed in sections 5.3 and 5.4 i t may be said that the Hypothesis 1 is part ly v a l i d and part ly not v a l i d . As we have found that ' p u l l 1 factors were dominant in both R-R in-migration streams and R-R out-migration streams, in 1951, i t disproves the v a l i d i t y of the 1st part of the Hypothesis. But the second part of the Hypothesis has been found to be va l id since distance was found to be more important in R-R in-migrat ion streams than in R-R out-migration streams. 5.5 Determinants of the Spatial Pattern of Rural-to-Urban  Out-Migration Streams (0V i . ) , 1951 5.5.1 Determinants of the Spatial Pattern of Out- Migration from Bar isal As indicated by Table 5.28, column ' a , ' most of the people from Barisal went to Dacca and Chittagong. Dinajpur and Kuchtia received much l e s s . By examining factor loadings in the ' C F . '* 0 f the factor matrix, i t may be interpreted that the presence of too *Loading on OV.. in the ' C . F . 1 was - .9080. 149 many landless a g r i c u l t u r a l labourers , LLALF/TALF (- .7561), acted as the 'push' in the i th d i s t r i c t s . On the other hand, people were ' p u l l e d ' to the d i s t r i c t s of dest inat ion ( j th d i s t r i c t s ) i f more land under r i ce was a v a i l a b l e . But the 'push' was dominant. The nature of loading on d is tance , D .^ (.6610), suggested that the migrants going out from Bar isal preferred to move to the nearer urban d i s t r i c t s . The factor score pattern of the j th d i s t r i c t s confirmed that the majority of people from Bar isal went to the nearer d i s t r i c t s (compare rank orders in columns ' d , ' 'b ' and 'a ' ). 5.5.2 Determinants of the Spatial Pattern of Out- Migration from Bogra Dinajpur, as indicated by Table 5.29, c o l u m n ' ' a , ' received most of the migrants from Bogra. Dacca came next in order while others received much l e s s . * By examining factor loadings in t h e ' C . F . ' of the factor matrix i t may be interpreted that people were ' p u l l e d ' to those j t h d i s t r i c t s from Bogra where the in tens i ty of cropping was l e s s , TCA/NCA (.6456), as there was more oppor-tun i t i es for the expansion of cu l t i va ted land. There was no 'push. ' * Loading on 0V-. in the ' C . F . ' was - .8130. 1 50 The nature of loading on d istance, D.. (.8230), i 3 indicated that the migrants from Bogra tended to minimize the distance t r a v e l l e d . The factor score pattern of the j th d i s t r i c t s suggested that the veilocity of migration from Bogra was as expected to a l l the rec ip ient d i s t r i c t s (compare rank orders in columns 1 d , ' and ' b ' ) . 5.5.3 Determinants of the Spatial Pattern of Out- Migration from Chittagong H.T. Factor analysis has not been performed with the Chittagong H.T. migration data since the number of people coming out from that d i s t r i c t was very smal l . 5.5.4 Determinants of the Spat ia l Pattern of Out- Migration from Comilla As indicated by Table 5.30, column ' a , ' about 80% of the tota l out-migrants went to only the one d i s t r i c t — Dacca. By examining factor loadings of the var iables in * the ' C F . ' of the factor matrix, i t may be interpreted that the a v a i l a b i l i t y of more land under j u t e , Jute/NCA (.8244), in a given j t h d i s t r i c t acted as the ' p u l l 1 f a c t o r . On the other hand, the presence of too many landless a g r i c u l t u r a l *Loading on OV. . in the ' C . F . ' was - .8409. 151 labourers , LLALF/TALF (- .7601), in the i th d i s t r i c t acted as the 'push. ' But the ' p u l l ' was dominant. Low and thus not s i g n i f i c a n t loading on d is tance , D.. (.3165), indicated that distance did not act as a con-s t r a i n t in determining the spat ia l pattern of migration coming out from Comi l la . The factor score pattern of the j th d i s t r i c t s revealed that in the two top receiv ing d i s t r i c t s , Dacca and Kushtia, the ve loc i ty of migration from Comilla was as expected (compare rank orders in columns ' d ' and ' b ' ) . 5.5.5 Determinants of the Spatial Pattern of Out- Migration from Faridpur As indicated by the Table 5.31, column ' a , ' though the majority of the people from Faridpur went to Dacca, other j th d i s t r i c t s also received substant ia l amounts of migrants. Thus, the d i s t r i b u t i o n was quite even. By examining factor loadings of the var iables in * the ' C . F . ' of the factor matrix i t may be interpreted that the presence of too many landless a g r i c u l t u r a l labourers in Far idpur, LLALF/TALF (- .8461), acted as the 'push' f a c t o r , while people were ' p u l l e d ' to the j t h d i s t r i c t s i f land under j u t e , Jute/NCA (.6571), was more than in Far idpur. But the 'push' was dominant. *Loading on OV.. in the ' C . F . ' was - .8356. 1 52 Since loading on distance was not s i g n i f i c a n t , D.. 1 J (- .1561), i t indicated that people did not tend to minimize the distance t r a v e l l e d . From the factor score pattern of the j t h d i s t r i c t s i t was found that the ve loc i ty of migration from Faridpur was as expected to the two top rec ip ient d i s t r i c t s ~ Dacca and Chittagong (compare rank orders in columns ' d ' and ' b ' ) . 5.5.6 Determinants of the Spat ia l Pattern of Out- Migration from Jessore More than 60% of the tota l out-migrants from Jessore went to Kushtia which is contiguous to i t . Comparatively much fewer went to the other j t h d i s t r i c t s (Table 5.32, column 1 a 1 ) . By examining factor loadings of the var iables in * the ' C F . ' of the factor matrix i t may be interpreted that migrants were 'pushed' from Jessore i f landless a g r i c u l t u r a l labourers , LLALF/TALF (- .7800), were more. There was ho ; ' pu l l ' S i g n i f i c a n t loading on d istance, D.. (.6138), indicated that the migrants tended to minimize the distance travel 1ed. The nature and magnitude of factor score on the j t h d i s t r i c t s revealed that the ve loc i ty of migration to a l l * Loading on OV.. in the ' C . F . [ was - .7561. Table 5.32 R-U 0UT-MI6RATI0N STREAMS FROM JESSORE, 1951 Out-Migration from the i t h D i s t r i c t ~ Jessore it a" "b" "c" "d M Di s t r i cts of Dest inat ion ( j th D i s t r i c t s ) No. of i n the Migrants Stream Ve loc i ty of Out-Migration Streams Distance Separati ng i and j (in miles) Factor scores on the Subjects from the ' C r i t i c a l Factor 1 OMij Rank OVij Rank Di j Fj* Rank Chi ttagong 416 (3) .0043 (4) 185 0.0561 (4) Dacca 1 ,191 (2) .0072 (2) 84. -0.5131 (2) Di naj pur 327 (4) .0055 (3) 175 -0.1245 (3) Kus hti a 6.786 (1) .0188 (1) 45 -.8553 (1) * ' C F . OVij -0.7561 LLALF/TALF -0.7800 Uij -0.1158 (not s i g n i f i c a n t ) Dij 0.6138 Co 1 54 the j t h d i s t r i c t s were as expected (compare rank orders in columns 1 d ' and 1 b' ). 5.5.7 Determinants of the Spat ia l Pattern of Out- Migration from Khulna As indicated by Table 5.33, column ' a ' , about 55% of the people went to Dacca. Others received a much smaller proport ion. Looking at the magnitude and nature of factor * loadings of the var iables in the ' C F . ' of the factor matrix, i t may be interpreted that people were 'pushed' from Khulna i f too many landless a g r i c u l t u r a l labourers were present (loading on LLALF/TALF was - .7888) . There was no ' p u l l , ' however. S i g n i f i c a n t loading on d istance, D.. (.7605), indicated that the migrants tended to go to the nearer d i s t r i c t s ^ in greater numbers in order to minimize the distance travel 1ed. The nature and magnitude of factor scores on the j t h d i s t r i c t s suggested that the v e l o c i t y of migration from Khulna was as expected to a l l the rec ip ient d i s t r i c t s (compare rank orders in columns ' d ' and ' b ' ) . Loading on OV.. in the ' C . F . ' was -.8856. 1 55 5.5.8 Determinants of the Spatial Pattern of Out- Migration from Mymensingh As indicated by Table 5,34, column ' a , ' about 88% of the tota l out-migrants went to only one d i s t r i c t - Dacca. By examining the factor loading c h a r a c t e r i s t i c s * of the var iables in the ' C F . ' 1 of the rotated factor matrix i t may be interpreted that the presence of too many a g r i c u l -tural labourers , LLALF/TALF (- .7581), and population pressure on a g r i c u l t u r a l land, RD/SQMCA (- .6466), acted as 'push' f a c t o r s . On the other hand, people were ' p u l l e d ' to those j th d i s t r i c t s where land under j u t e , Jute/NCA (.7135), was more. But the 'push' was dominant. S i g n i f i c a n t loading on d istance, D.. (.8636), 1 3 indicated that people tended to go to the nearer d i s t r i c t s in greater numbers to minimize the distance t r a v e l l e d . From the nature and magnitude of factor scores on the subjects ( j th d i s t r i c t s ) i t may be said that except for Dinajpur the ve loc i ty of migration from Mymensingh was as expected to a l l the receiv ing d i s t r i c t s (compare rank orders in columns ' d ' and 'b'*). 5.5.9 Determinants of the Spatial Pattern of Out- Migration from Noakhali As indicated by Table 5.35, column 'a ;J though the majority of out-migrants went ot Chittagong (56%) and l o a d i n g on OV.. in the ' C . F . ' was - .8449. 13 1 56 Dacca (29%), other d i s t r i c t s also received substant ia l numbers of migrants. In other words, the d i s t r i b u t i o n of out-migration from Noakhali was quite even compared to other i th d i s t r i c t s in the R-U out-migration stream. By examining factor loadings of the var iables in * the ' C . F . ' of the rotated factor matrix i t may be said that people were ' p u l l e d ' from Noakhali to those j t h d i s t r i c t s where land under r i c e , Rice/TCA(. 6616 ), was more than in Noakhali . On ther other hand, the presence of too many landless a g r i c u l t u r a l labourers , LLALF/TALF 9-.6461 ) , and population pressure on a g r i c u l t u r a l land, RD/SQMCA (- .5134), in Noakhali acted as the 'push' f a c t o r s . Both 'push' and ' p u l l ' factors appeared to be equally strong. S i g n i f i c a n t loading on d istance, D.. (.8134), suggested that the migrants wentoto the nearer d i s t r i c t s in greater numbers to minimize the distance t r a v e l l e d . The factor score pattern of the j t h d i s t r i c t s indicated that the ve loc i ty of migration from Noakhali was as expected to a l l the receiv ing d i s t r i c t s (compare rank orders in columns ' d ' and ' b ' ) . 5.5.10 Determinants of the Spat ia l Pattern of Out- Migration from Pabna As indicated by Table 5.36> column ' a , ' the maximum number of people from Pabna went to Dacca c lose ly followed by *Loading on OV.. in the "C ,F . " was - .7165. 1 57 Kushtia. Dinajpur and Chittagong also received substant ia l numbers of migrants. In other words, the d i s t r i b u t i o n of out-migration from Pabna was quite even. By examining the factor loadings of the var iables * in the ' C . F . ' of the factor matrix i t may be interpreted that the presence of too many landless a g r i c u l t u r a l labourers , LLALF/TALF (- .8565), in Pabna acted as the 'push' f a c t o r . On the other hand, people were ' p u l l e d ' to those j th d i s t r i c t s i f land under j u t e , Jute/NCA (.7534), was more than in Pabna. But the 'push' was stronger. S i g n i f i c a n t loading on d is tance , D.. (.7665), "1 3 indicated that the migrants tended to go to the nearer d i s t r i c t s in greater numbers to minimize the distance t r a v e l l e d . The factor score pattern of the j t h d i s t r i c t s revealed that except for Chittagong the v e l o c i t y of migration from Pabna was as expected to a l l d i s t r i c t s of dest inat ion (compare rank orders in columns 'd ' iand ' b ' ) . 5.5.11 Determinants of the Spatial Pattern of Out- Migration from Rajshahi As indicated by Table 5.37, column ' a , ' out of the total out-migrants from Rajshahi, Dinajpur received the most c lose ly followed by Kushtia. Other received many fewer. * Loading on OV.. in the ' C . F . ' was - .8558. 158 By examining factor loadings of the var iables in * the ' C . F . ' of the factor matrix i t may be interpreted that people were 'pushed' from Rajshahi i f the presence of landless a g r i c u l t u r a l labourers, LLALF/TALF 9-.8134), was more than in a given j t h d i s t r i c t . On the otherhand, people were ' p u l l e d ' from Rajshahi i f more land under r i c e , Rice/TCA (.6314), was ava i lab le in a given j th d i s t r i c t than in Rajshahi. But the 'push' was dominant. S i g n i f i c a n t loading on d istance, D.. (.7663), indicated that the migrants tended to go to the nearer d i s t r i c t s in greater numbers to minimize the distance t r a v e l l e d . The factor score pattern of the j t h d i s t r i c t s re-vealed that the ve loc i ty of migration from Rajshahi was as expected to a l l the d i s t r i c t s of dest inat ion (compare rank orders in columns ' d ' and ' b ' ) . 5.5.12 Determinants of the Spat ia l Pattern of Out- Migration from Rangpur As indicated by Table 5.38, column ' a , ' out of the tota l migrants about 84% went to only d i s t r i c t , Dinajpur, which is contiguous to i t . Others received much l e s s . Thus the spat ia l pattern of d i s t r i b u t i o n of out-migrants was very skewed. * Loading on OV.. in the ' C . F . ' was -.8396. 1 59 By examining factor loadings of the var iables in * the ' C F . ' of the rotated factor matrix i t may be i n t e r -preted that the presence of too many landless a g r i c u l t u r a l labourers , LLALF/TALF (- .8111), in Rangpur compared to a given j t h d i s t r i c t acted as the 'push' f a c t o r . On the other hand, people here ' p u l l e d ' to those j th d i s t r i c t s where the in tens i ty of cropping was less and thus more p o s s i b i l i t i e s were there to increase total cu l t i va ted land, TCA/NCA (.6616). More land under j u t e , Jute/NCA (.5614), in a given j th d i s t r i c t than in Rangpur also acted as ' p u l l . ' But the 'push' was domi nant. S i g n i f i c a n t loading on d istance, D. . (.7105), indicated that the migrants went to the nearer d i s t r i c t s in greater numbers to minimize the distance t r a v e l l e d . The factor score pattern of the j t h d i s t r i c t s revealed t h a t ' i n the leading two d i s t r i c t s the v e l o c i t y of migration from Rangpur was as expected (compare rank orders in columns ' d 1 and ' b 1 ) . 5.5.13 Determinants of the Spat ia l Pattern of Out- Migration from Sylhet As indicated by Table 5.3.9, Column ' a , ' the majority of migrants from Sylhet went to Dacca and Chittagong, while 8 Loading on OV. . in the ' C . F . ' was - .8310. 1 60 the number of people going to Dinajpur and Kushtia was nomi nal . By examining factor loadings of the var iables in * the ' C F . ' of the factor matrix i t may be interpreted that the presence of too many landless a g r i c u l t u r a l labourers , LLALF/TALF (- .7310), in Sylhet acted as the 'push' f a c t o r . There was no_'pull , ' however. S i g n i f i c a n t loading on d istance, D.. (.8838), indicated that the migrants tended to go to the nearer d i s t r i c t s in greater numbers to minimize the distance t r a v e l l e d . The factor score pattern of the j th d i s t r i c t s indicated that while in Chittagong and Kushtia the v e l o c i t y of migration from Sylhet was more than expected, in Dacca i t was l e s s . Dinajpur was the only d i s t r i c t where OV.. was 13 as expected (compare rank orders in column ' d 1 and ' b ' ) . 5.5.14 Summary of Section.' 5.. 5. The spat ia l d i s t r i b u t i o n of Rural-to-Urban out-migration streams was d i f f e r e n t from that of Rural-to-Rural migration streams. In th is case substant ia l numbers of migrants did go to the nearer d i s t r i c t s but Dacca, as expected, attracted the highest number of migrants from 7 out of 12 rural d i s t r i c t s (excluding Chittagong H . T . ) , though not a l l *Loading on OV^ in the ' C F . ' was -.9134. 161 of them were located near Dacca (see Table 5C, column ' c ' ) . Thus, on an overal l bas is , i t may be said that the migrants going out from rural d i s t r i c t s in the R-U out-migration streams went to only one 'urban' d i s t r i c t , Dacca. By examining the factor loadings of the var iables in the C . F . ' of the factor matrix i t was found that both 'push' and ' p u l l ' factors were operating in determining the spat ia l pattern of out-migration streams going out from the rural d i s t r i c t s to the 'urban' d i s t r i c t s . People were 'pushed' from the i th d i s t r i c t s i f there was too many land-less a g r i c u l t u r a l labourers, LLALF/TALF, and/or the popula-t ion pressure on a g r i c u l t u r a l land, RD/SQMCA, was excessive (see section 5.5 and Table 5C, column ' c ' ) . At the same time, people were ' p u l l e d ' to the j t h d i s t r i c t s i f the in tens i ty of cropping was low, TCA/NCA, and thereby more opportunit ies were there for the expansion of cu l t i va ted land ( l i k e in Dinajpur) . [See subsection 5.5.2 in sect ion 5.5.] People were also sometimes ' p u l l e d ' to the d i s t r i c t s of dest inat ion i f land under j u t e , Jute/NCA, was more than in a given i t h d i s t r i c t . It may be noted that loading on II.. was not s i g -n i f i c a n t in any of t h e ' C . F . ' s ind icat ing that the ' p u l l ' from the urban areas of the 'urban' d i s t r i c t s was lacking in 1951. The 'push' factors were dominant in nine out of twelve d i s t r i c t s such as B a r i s a l , Far idpur, Jessore, Khulna, 162 Mymensingh, Pabna, Rajshahi, Rangpur and Sylhet in eight of them the presence of too many landless a g r i c u l t u r a l labourers was acting as the 'push' (see Table 5C). There was only one occasion when the ' p u l l ' factor was dominant (such as in Comilla) and in the case of Noakhali 'push' and ' p u l l ' were almost equally strong. Thus from the above discussions i t may be said that both 'push' and ' p u l l ' factors were act ive in determining the spat ia l patterns of R-U out-migration streams but the 'push' factors were overwhelmingly dominant. Distance played a s i g n i f i c a n t ro le in determining the spat ia l pattern of R-U out-migration streams coming out from the majority of the d i s t r i c t . The general tendency of the migrants was to go to the nearer d i s t r i c t s in greater numbers to minimize the distance t r a v e l l e d . In the ' C . F . ' s , factor loading on D.j was s i g n i f i c a n t in ten out of twelve d i s t r i c t s such as Ban"sal , Bogra , Jessore, Khulna, Mymensingh, Noakhali , Pabna, Rajshahi, Rangpur and Sylhet . Only in the cases of migration streams or ig inat ing from Comilla and Faridpur distance was of no importance at a l l as the .d is tance separating these d i s t r i c t s and the d i s t r i c t s of dest inat ion did not a f fect the ve l i co ty of out-migrat in . However, when D.. . were regressed against OV.^, ceteris -paribus, distance was found to be s i g n i f i c a n t l y re lated with 0\l. . in a l l the i th d i s t r i c t s though in the Table 5C DETERMINANTS OF THE SPATIAL PATTERN OF R-U OUT-MIGRATION STREAMS, 1951 a b c d e OV.. from 1 J Wastifche spat ia l di s t r i bution skewed towards the nearer di s t r i cts? Which of the factors were dominant -'push' or ' p u l l 1 ? Was distance si gni f i cant? Was the ve loc i ty of migration to Dacca highest? 5 1 . Bar isal Yes Push (LLALF/TALF) Yes Yes 2. Bogra Yes Only pull (TCA/NCA), no push Yes No 3. Comi11 a Yes Pull (Jute/NCA) No Yes 4. Faridpur Yes Push (LLALF/TALF) No Yes 5. Jessore Yes Only push (LLALF/TALF) ; no pul l Yes No 6. Khulna No Only push (LLALF/TALF) no push Yes Yes 7. Mymensingh Yes Push (RD/SQMCA and LLALF/TALF) Yes Yes CONTINUED CM CO Table 5C (Continued) a b c d e OV.. from * J Was the spat ia l d i s t r i b u t i o n s kewed towards the nearer di s t r i cts? Which of the factors were dominant -1 push ' or 1 pul1 1 ? Was distance s i g n i f i c a n t ? Was the ve loc i ty of migration to Dacca highest? 8. Noakhali Yes Push (RD/SQMCA and LLALF/TALF) and pull (Rice/TCA) equally strong Yes No 9. Pabna Yes Push (LLALF/TALF) Yes Yes 10. Raj shahi Yes Push (LLALF/TALF Yes No 11 . RAngpur Yes Push (LLALF/TALF) Yes No 12. Sylhet Yes Push (LLALF/TALF) Yes Yes o O - i LO O co O H § 4 o LO CM oH o o CM oH o LO oH o o oH o LO o 0-040 030 .020 ,010 - f l -ies n T f= |—i -r T- r—| T T* r~i -j r r i i - | i i l T * T * I *r 50 100 150 200 250 DISTRNCE IN MILES. DIJ T* | i i r i—|—r 300 350 T — I — | 400 Figure 5.3. 'Distance E l a s t i c i t y ' of R-U Out-Migration Streams,1951. i 166 case of Bar isal the re la t ionsh ip was pos i t i ve (see Figure 5.3) . The factor score pattern of the j t h d i s t r i c t s revealed that the ve loc i ty of migration to Dacca was highest from seven out of twelve i th d i s t r i c t s . This indicated that most of the people in the R-U out-migration streams was going to Dacca, despite the fact that the 'push' from the rural d i s t r i c t s was stronger than the ' p u l l ' from Dacca. 5.6 Determinants of the Spatial Pattern of Urban-to-Rural  Out-Migration Streams, 1951 5.611 5.,.Determi!hanifeSF0fefcthe^Sp Migration from Chittagong As indicated from Table 5.40, column ' a , 1 Chittagong H.T. got the l i o n ' s share of the migrants from Chittagong, which is contiguous to i t . B a r i s a l , Comilla and Noakhal i , a l l neighbouring d i s t r i c t s , came next in order. Others got much l e s s . By examining factor loadings of the var iab les in the ' C . F . ' of the factor matrix i t may be said that people were ' p u l l e d ' from Chittagong i f more r i c e land, Rice/TCA (.7300), was ava i lab le in a given j t h d i s t r i c t . There was no 'push. ' Loading on OV.. in the ' C . F . ' was -.7564. Table 5.40 U R B A N T O R U R A L 0 U T - M I 6 R A T I 0 N S T R E A M S , 1 9 5 1 S P A T I A L P A T T E R N OF O V I J FROM CHITTAGONG D i s t r i c t of Origin ( i t h D i s t r i c t ) ~ Chittagong n _ II a "b" "c" "d" D i s t r i c t s of Desti nation (jth D i s t r i c t s ) No. of Migrants in the Stream Ve loc i ty of Out-Migration Streams Di stance Separati ng i and j (in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor 1 Bar isa l Bogra C h i t t . H.T. Comi11a Faridpur Jessore Khulna Kymensingh Noakhali Pabna Rajshahi Rangpur Sylh e t OMij - Rank OVij - Rank Di j Fj* ~ Rank 2.005 (2) 144 (13) 13,025 (1) 1,749 (3) 500 (9) 233 (11) 1.006 (5) 927 (6) 1,716 (4) 223 (12) 253 (1 0) 720 (8) 858 (7) .0099 (3) .0026 (10) .8690 (1) .0083 (4) .0031 (8) .0022 (12) .0082 (5) .0028 (9) .0736 (2) .0024 (11) .0019 (13) .0044 (7) .0051 (6) 112 250 45 112 150 185 210 200 70 200 265 288 175 -.0482 (6) .5105 (9) -2.4946 (1 ) -.0068 (7) -.1505 (5) .1440 (8) -.4626 ( 3 ) . 8 1 8 3 (12) -1.047 (2) .7813 ( I D .6550 (10) 1.5457 (13) -.2440 (4) *Factor 1 ( ' C F . ' ) OVij -.7564 Rice/TCA .730 Dij .8806 1 68 S i g n i f i c a n t loading on d istance, D.. (.8806), indicated that the migrants went to the nearer d i s t r i c t s in greater numbers to minimize the distance t r a v e l l e d . The factor score c h a r a c t e r i s t i c s of the j t h d i s t r i c t s revealed that except for the two loading rec ip ient of migrants from Chittagong (Chittagong H.T. and Noakhali) in a l l other d i s t r i c t s the ve loc i ty of migration was not as expected (compare rank orders in columns ' d ' and ' b ' ) . 5.6.2 Determinants of the Spatial Pattern of Out- Migration from Dacca As indicated by Table 5.41, column ' a , ' Mymensingh and Faridpur got the majority of migrants from Dacca. B a r i s a l , Rajshahi, Sylhet and Rangpur also got substant ia l numbers of migrants. It may be noted that while the former two d i s t r i c t s are neighbouring to Dacca, the l a t t e r four are not. Thus the d i s t r i b u t i o n of out-migration was not l o c a l i z e d but quite widespread. By examining factor loadings of the var iab les in * the . C . F . ' of the factor matrix i t may be said that the migrants were ' p u l l e d ' from Dacca i f in a given j t h d i s t r i c t the in tens i ty of cropping, TCA/NCA (.5536), was less as more opportunit ies were there to expand cu l t ivated land. There was no 'push' f a c t o r , however. *Loading on OV^. in the ' C . F . ' was - .8795. 169 The nature of loading on d istance, D. . (.6691 ) , indicated that the migrants tended to go to the nearer d i s t r i c t s in greater numbers to minimize the distance t r a v e l l e d . The factor score pattern of the j t h d i s t r i c t s con-firmed that in far away d i s t r i c t s l i k e Rajshahi 9 and Rangpur the ve loc i ty of migration was more than expected (compare rank orders in columns 'b ' and ' d ' ) . 5.6.3 Determinants of the Spatial Pattern of Out- Migration from Dinajpur As indicated by Table 5.42, column ' a , 1 substant ia l numbers of migrants went to only three d i s t r i c t s , Rangpur, Bogra and Rajshahi. They a l l neighbour Dinajpur. A nominal number of migrants went to other d i s t r i c t s . By examining factor loadings of the var iables in * the ' C . F . ' of the factor matrix i t may be interpreted that there was only ' p u l l , ' no 'push . ' The migrants were ' p u l l e d ' from Dinajpur i f more land under j u t e , Jute/NCA (.8032), was ava i lab le in a given j t h d i s t r i c t . The nature of loading on d istance, D.. (.8692), indicated that the migrants tended to go to the nearer d i s t r i c t s in greater numbers to minimize the distance t r a v e l l e d . The factor score pattern of the j th d i s t r i c t s indicated that in a l l the leading rec ip ient d i s t r i c t s the *Loading on OV.,. in the ' C . F . ' was - .7427. 170 ve loc i ty of migration from Dinajpur was as expected (compare rank orders in columns ' d ' and ' b ' ) . 5.6.4 Determinants of the Spatial Pattern of Out- Migration from Kushtia As indicated by Table 5.43, column ' a , ' the majority of migrants went to the neighbouring d i s t r i c t s of Jessore, Pabna and Faridpur and Rajshahi. Others got much l e s s . By examining factor loadings of the var iables in * the ' C F . 1 of the factor matrix, i t may be interpreted that there was only a ' p u l l 1 factor operat ing, no 'push . 1 The migrants were ' p u l l e d ' from Kushtia i f more land under r i c e , Rice/TCA (-.8514), was ava i lab le in a given j t h d i s t r i c t . The nature of loading on d istance, D.. (- .5672), indicated that the migrants tended to go to the nearer d i s t r i c t s in greater numbers to minimize the distance travel 1ed. The factor score pattern of the j th d i s t r i c t s revealed that in the top rec ip ient d i s t r i c t s the ve loc i ty of migration from Kushtia was as expected, (compare rank orders in columns 'd ' and ' b ' ). Loading on OV.. in the ' C . F . ' was - .8630. 171 5.6.5 Summary of Section 5.6 From the examination of the spat ia l d i s t r i b u t i o n pattern of U-R out-migration streams i t was observed that the majority of the migrants went to the neighbouring d i s t r i c t s o r ig inat ing from a l l the four 'urban' d i s t r i c t s (see sect ion 5.6 and Table 5D). When the loadings on the ' C . F . ' s of the factor matrices were'slooked at in terms of the 'push-pu l l ' theory, i t was found that only ' p u l l ' factors were operating in determining the spat ia l pattern of U-R out-migration streams. There was no 'push. ' People were ' p u l l e d ' to the j t h d i s t r i c t s i f the land under r i c e , Rice/TCA, or j u t e , Jute/NCA, was more than in the i t h d i s t r i c t s (see sectionl '5.6 and Table 5D). Apparently the a v a i l a b i l i t y of more r i c e and jute land in those d i s t r i c t s offered more job opportunit ies and thus was considered to be more a t t r a c t i v e to the migrants. People were also ' p u l l e d ' to the j j th d i s t r i c t s i f the in tens i ty of cropping, TCA/NCA, was less than in the i th d i s t r i c t s . Probably the lower in tens i ty of cropping in those d i s t r i c t s offered more opportunit ies for the expansion of cu l t i va ted land and thus considered to be a t t r a c t i v e to the migrants coming out from the i th d i s t r i c t s . On D.jj factor loading was s i g n i f i c a n t in a l l the ' C . F . ' s of the respect ive d i s t r i c t s (see Table 5D). The nature of loadings on and OV... suggested that distance Table 5D DETERMINANTS OF THE SPATIAL PATTERN OF U-R OUT-MIGRATION STREAMS, 1951 a b c d e OV.. from Was the spat ia l d i s t r i b u t i o n skewed towards the nearer d is t r ic ts?^ Which of the factors were dominant -'push 1 or ' p u l l ' ? Was distance s i g n i f i c a n t ? Was the ve loc i ty of migration as expected to the majority of j t h d i s t r i c t s ? 1 . Chittagong Yes only pul1 (Rice/TCA), no push Yes * No 2. Dacca Yes only pull (TCA/NCA) , no push Yes Yes 3. Dinajpur Yes only pull (Jute/NCA), no push Yes Yes 4. Kushtia Yes only pul1 (Rice/TCA), no push Yes Yes Distance was the most important f a c t o r . o o O-i 173 o LO OH \ o o CO OH o LO CM OH o o CM OH \ \ \ \ \ \ \ \ \ \ N I * \ ^ \ o LO OH o o .N o LO o OH 040-030-020-010-- 0 - i—i—r I 1 1 1 1 I 1 50 100 1 1 I ' 150 ~ r — ] — i — i — i — r — j — i — i — i — i — | — I — I I I |—I—I—I—I—| 200 250 300 350 400 DISTANCE IN MILES. D U Figure 1 5.4. 'Distance E l a s t i c i t y ' of U-R Out-Migration Streams, 1951. 174 played an important role in determining the spat ia l pattern of U-R out-migration and the migrants had the tendency of minimizing the distance t r a v e l l e d (see sect ion 5.6 for d e t a i l s ) . When D. . was regressed against OV. . , c e t e r i s parebus3 the strong inf luence of distance on migration streams was confirmed (see Figure 5.4) . The tendency of people not to go away too far from Dacca may be of in teres t to note. The factor score pattern of the i th d i s t r i c t s revealed that ; the ve loc i ty of migration from three out of four i th d i s t r i c t s was as expected to the majority of the j t h d i s t r i c t s (see Table 5D). Even in the case of Chittagong the ve loc i ty of migration from this d i s t r i c t was as expected to the top rec ip ient d i s t r i c t s where the majority of migrants went. V a l i d i t y of Hypothesis 2 Based on the resu l ts of the analysis discussed in sections 5.4 and 5.6 i t may be said that the Hypothesis 2 is p a r t i a l l y v a l i d . It was observed that , in 1951, though both 'push' and ' p u l l 1 factors were operat ing, contrary to our expectat ion, 'push' factors were dominant in both Rura l - to-Urban out-migration streams and Urban-to-Rural out-migration streams. Thus the f i r s t part of the Hypothesis 2 has been found to be not v a l i d . The second part of the Hypothesis, however, has been found to be va l id f o r , as expected, distance 175 had more e f fect on Urban-to-Rural out-migration streams than on Rural-to-Urban out-migration streams. 5.7 Determinants of the Spatial Pattern of Urban-to-Urban  Out-Migration Streams, 1951 5.7.1 Determinants of the Spatial Pattern of Out- Migration from Chittagong As indicated by Table 5.44, column ' a , ' 87% of the total out-migrants from Chittagong went to only one d i s t r i c t -Dacca. Dinajpur and Kushtia received almost the same number of migrants. Thus the d i s t r i b u t i o n of the spat ia l pattern out-migration from Chittagong was skewed. By examining factor loadings of the var iab les in * the ' C F . ' of the factor matrix i t may be interpreted that people were 'pushed' from Chittagong i f population pressure on a g r i c u l t u r a l land, RD/SQMCA (- .9811), was more than in a given j t h d i s t r i c t . On the other hand, r e l a t i v e l y greater ' l eve l of urbanizat ion ' in a given j t h d i s t r i c t ' p u l l e d ' people away from Chittagong. But the 'push' was stronger. S i g n i f i c a n t loading on d istance, D.. (.9629), indicated that the migrants from Chittagong tended to go to the nearer 'urban' d i s t r i c t s in greater numbers to minimize the distance t r a v e l l e d . • *Loading on OV,... in the ' C . F . ' was - .9511. 176 From the factor score c h a r a c t e r i s t i c s of the j t h d i s t r i c t s i t may be said that the v e l o c i t y of migration from Chittagong was as expected to a l l d i s t r i c t s of dest inat ion (compare rank orders in columns ' d ' and ' b ' ) . 5.7.2 Determinants of the Spatial Pattern of Out- Migration from Dacca As indicated by Table 5.45, columns ' a ' and ' b , 1 almost the same number of migrants went to the j th d i s t r i c t s but the rank order was a ltered when the number of migrants was expressed in terms of v e l o c i t y of migrat ion. By examining factor loadings of the var iables in the * ' C . F . ' of the rotated matrix i t may be interpreted that people were ' p u l l e d ' to those j th d i s t r i c t s where the in tens i ty of cropping, TCA/NCA (-.9852), was less than in Dacca as there was more p o s s i b i l i t i e s to increase cu l t i va ted land. On the other hand, the presence of too many landless a g r i c u l t u r a l labourers, LLALF/TALF (- .6426), in Dacca than in a given j t h d i s t r i c t acted as the 'push' f a c t o r . But the ' p u l l ' was stronger. S i g n i f i c a n t loading on d istance, D.. (.9878), indicated that the migrants from Dacca tended to go to the nearer d i s t r i c t s in greater numbers to minimize the distance travel 1ed. *Loading on OV^. in the ' C F . ' was - .7489. 177 The factor score pattern of the j th d i s t r i c t s revealed that the ve loc i ty of migration from Dacca was as expected only in Kushtia,(compare rank orders in columns 'b ' and ' d ' ) . 5.7.3 Determinants of the Spat ia l Pattern of Out- Migration from Dinajpur As indicated by Table 5.46, column ' a , 1 very few people went out from Dinajpur as i t was mainly a receiver of migrants. Out of the total number of out-migrants the majority went to Dacca followed by Chittagong and Kushtia. By examining factor loadings of the var iab les in the * ' C F . 1 of the factor matrix i t may be interpreted that people were ' p u l l e d ' to those j th d i s t r i c t s i f more cu l t i va ted land was a v a i l a b l e , TCA/NCA (.9768). The higher level of urbanization in a j t h d i s t r i c t , U. . (.8122), also ' p u l l e d ' people from Dinajpur. There was no 'push . ' Low and thus non-s ign i f i cant loading on d is tance , D.. (- .2871), indicated that distance separating the i th and j t h d i s t r i c t s did not act as a const ra in t . The factor score pattern of the j th d i s t r i c t s indicated that in a l l the j t h d i s t r i c t the v e l o c i t y of migra-t ion from Dinajpur was as expected. Loading on OV.. in the ' C . F . ' was - .9875. 178 5.7.4 Determinants of the Spatial Pattern of Out- Migration from Kushtia As indicated by Table 5.47/, column ' a , ' the highest numbers of migrants went to Dacca followed by Dinajpur and Chi ttagong. By examining factor loadings of the var iables in * the ' C . F . ' of the factor matrix i t may be interpreted that people were ' p u l l e d ' to those j t h d i s t r i c t s in greater number where land under j u t e , Jute/NCA (-.9561), was greater than in Kushtia. On the other hand, the presence of more landless a g r i c u l t u r a l labourers , LLALF/TALF (-.921:1), in Kushtia than in a given j th d i s t r i c t acted as the 'push' f a c t o r . But the ' p u l l ' was dominant. S i g n i f i c a n t loading on d is tance , D.. (.9831), indicated that the migrants tended to go to the nearer d i s t r i c t s in greater numbers to minimize the distance involved in migrat ion. The factor score pattern of the j t h d i s t r i c t s revealed that the v e l o c i t y of migration from Kushtia was as expected only in Chi ttagong., (compare rank orders in columns ' d ' and ' b ' ) . * Loading on OV.. in the ' C . F . ' was - .8339. Table 5.47 U R B A N T O U R B A N O U T - M I G R A T I O N S T R E A M S / 1 9 5 1 Spat ia l Pattern of OVij from Kushtia, 1951 D i s t r i c t of Origin (ith d i s t r i c t ) - Kushtia II a" "b" "c" " d D i s t r i c t s of Dest inat ion ( j th d i s t r i c t s ) No. of in the Mi grants Stream V e 1 o c i ty Mi grat i on of OUt-Streams Di stance Separati ng i and j (in miles) Factor scores on the Subjects from the ' C r i t i c a l Factor ' OM i j ~ Rank OVij Rank Di j Rank Chittagong 497 (3) .0097 (2) 225 -0.5262 (2) Dacca 819 (1) .0091 (3) 95 -0.6269 (1) Dinajpur 690 (2) ".0240 (1) 140 1 .1532 (3) *Factor 1 ( C F . ) OVij -0.8339 Dij 0.9831 Jute/NCA -0.9561 LLALF/TALF 0.9211 1 8 0 5 . 7 . 5 Summary of Section 5.7 The spat ia l d i s t r i b u t i o n s of Urban-to-Urban out-migration streams were unique in that sense that , out of four 'urban' d i s t r i c t s , Dacca got the highest number of migrants from the other three 'urban' d i s t r i c t s . Thus i t may be said that , on an overal l bas i s , the d i s t r i b u t i o n of U-U out-migration was heavily skewed towards Dacca. Both 'push' and 1 pul1 ' factors were found to be operating in determining the spat ia l pattern of U-U out-migration streams when the loadings on the ' C . F . ' s of the factor matrices were looked at in terms offthe 'push-pu l l ' theory. People were 'pushed' from the i th d i s t r i c t s i f the population pressure on agr i cu l tu ra l land,( RD/SQMCA, was found to be more than in the j t h d i s t r i c t s . On the other hand, people were ' p u l l e d ' to the j t h d i s t r i c t s i f the in tens i ty of cropping, TCA/NCA, was e i ther less or more than in a given i th d i s t r i c t . In the case of migrants coming out from Dacca people went to the d i s t r i c t s l i k e Kushtia and Dinajpur where the in tens i ty of cropping was less because more opportunit ies were there to expand cu l t ivated land. But in the case of migrants coming out from Dinajpur, people went to Dacca and Kushtia because the in tens i ty of cropping was more. Apparently the c u l t i v a t i o n of land more than once in those d i s t r i c t s offered more jobs in the f i e l d s and thus was considered to be a t t r a c t i v e to the migrants 4. People 181 were also ' p u l l e d ' to the j t h d i s t r i c t s i f the ' l eve l of urbanizationJJ U^., was more than in a given i th d i s t r i c t such as in the case of people coming from Dinajpur to Dacca or Chittagong. The a v a i l a b i l i t y of more land under j u t e , Jute/NCA, in a given j th d i s t r i c t also ' p u l l e d ' people to that d i s t r i c t as the jute c u l t i v a t i o n is more labour intensive and thus needed more a g r i c u l t u r a l labourers . The ' p u l l ' factors were found to be dominant in three out of four d i s t r i c t s such as Dacca, Dinajpur and Kushtia (see Table 5E). Thus we may conclude that , on an overal l bas is , ' p u l l ' factors were dominant in determining the spat ia l pattern of U-U out-migration streams. Loading on D.^  was found to be s i g n i f i c a n t in the ' C . F . ' s of three out of four d i s t r i c t s (see table 5E). In migration streams coming out from Dacca and Kushtia, though 'push' and ' pul1 ' factors were also operat ing, distance was the most important f a c t o r . Thus we may say that distance played a very strong role in determining the spat ia l pattern of U-U out-migration streams. The strong role of distance was again confirmed when D.j . was regressed against OV^y c e t e r i s parebus. From the simple regression analysis i t was found that distance was s i g n i f i c a n t l y re lated with migration v e l o c i t y in-migrat ion streams coming out from a l l the i th d i s t r i c t s (Figure 5 .5) . Table 5E D E T E R M I N A N T S OF T H E S P A T I A L P A T T E R N OF U - U O U T - M I G R A T I O N S T R E A M S , 1 9 5 1 a b c d e OV.. from Was the spat ia l d i s t r i b u t i o n skewed toward the nearer d i s t r i c t s ? Which of the factors were dominant - 'push' or ' p u l l 1 ? Was distance s ign i f i cant? Was the ve loc i ty of migration as expected -• to • the j t h d i s t r i c t s ? 1. Chittagong Yes Push (RD/SQMCA Yes A l l 2. Dacca No Pull (TCA/NCA) Yes only in Kushtia 3. Dinajpur No Pull (TCA/NCA and U., v.' J No A l l 4. Kushtia Yes Pull (Jute/NCA) Yes only in Chittagong CO ro 5 I \ \ I I \ \ \ \ \ \ \ \ I I I \ \ I I \ \ \ \ \ \ I I \ \ \ \ \ I I \ \ \ \ 183 » i "7 T 1 i i i i r I i r i i ^ i i1 i i * | i i i i | i i i i r i r i t | i i i i i 50 100 150 200 250 300 350 400 DISTANCE IN MILES. DIJ Figure 5.5. 'Distance E l a s t i c i t y ' of U-U Out-Migration Streams, 1951. 184 The general tendency of the migrants were to minimize the distance involved in migrat ion. The factor score pattern of the j th d i s t r i c t s revealed that the ve loc i ty of migration from two out of four d i s t r i c t s such as Chittagong and Dinajpur was as expected to a l l the j t h d i s t r i c t s (see Table 5E). 5.8 Determinants of the Spatial Pattern of Urban-to-Urban  In-Migration Streams, 1951 5.8.1 Determinants of the Spatial Pattern of In- Migration to Chittagong As indicated by Tables 5.48, column ' a , 1 more that 90% of the total migrants came from Dacca. By examining factor loadings of the var iables in * the ' C F . ' of the fa Eton, matrix i f may Be interpreted that the population pressure on a g r i c u l t u r a l Inad, RD/SQMCA (.6633), acted as the 'push' f a c t o r . There was no ' p u l l , ' however. The nature of loading on d istance, D.. (.8446), indicated that i t was the most important factor and people came to Chittagong from nearer d i s t r i c t s in greater numbers to minimize the distance t r a v e l l e d . The factor score pattern of the i th d i s t r i c t s indicated that the ve loc i ty of migration to Chittagong was Loading on I V . . i n the ' C F . ' was .8644. 185 as expected from a l l the sending d i s t r i c t s (compare rank orders in columns ' d ' and ' b ' ) . 5.8.2 Determinants of the Spatial Pattern of In- Migration to Dacca As indicated by Table 5.49, column ' a , ' more than 65% of the tota l migrants to Dacca came from Chittagong. By examining factor loadings of the var iables in * the ' C F . 1 of the factor matrix i t may be interpreted that only ' p u l l ' factors were operat ing. There was no 'push 1 . The a v a i l a b i l i t y of more land under j u t e , Jute/NCA (- .6498), and the higher ' l eve l of urbanizat ion ' in Dacca ' p u l l e d ' people from the i th d i s t r i c t s . The nature of loading on d istance, D.. (.6390), indicated that the migrants tended to come from nearer d i s t r i c t s in greater numbers to minimize the distance t r a v e l l e d . The factor score pattern of the i th d i s t r i c t s indicated that the ve loc i ty of migration to Dacca was as expected fromsall the i th d i s t r i c t s (compare rank orders in columns ' d ' and 'b 1 ) . 5.8.3 Determinants of the Spatial Pattern of In- Migration to Dinajpur As indicated by Table 5.50, column ' a , ' more than 90% of the total migrants to Dinajpur came from Dacca. *Loading on IV. . in the ' C . F . ' was - .7211. 186 By examining factor loadings of the var iables in * the ' C F . ' of the factor matrix i t may be interpreted that the presence of too many landless a g r i c u l t u r a l labourers , LLALF/TALF (.7678), and population pressure on a g r i c u l t u r a l land, RD/SQMCA (.6614), in a given i th d i s t r i c t acted as the 'push' f a c t o r s . And there was no_ ' p u l l ' from Dinajpur. Loading on d istance, D.. (.4513), was marginal ly s i g n i f i c a n t ind icat ing that the e f fect of distance was not very strong. The factor score pattern of the i th d i s t r i c t s i n -dicated that the ve loc i ty of migration to Dinajpur was as expected from a l l the d i s t r i c t s of o r i g i n (compare rank orders in columns ' d ' and ' b ' ) . 5.8.4 Determinants of the Spat ia l Pattern of In- Migration to Kushtia As indicated by Table 4.5T, column ' a , 1 about 90% of the tota l migrants to Kushtia came from Dacca. By examining factor loadings of the var iables in ick the ' C . F . ' of the factor matrix i t may be said that there were neither any ' p u l l ' nor any 'push' factors operat ing. Loading on IV^- in the ' C F . ' was - .7400. Loading on IV^. in the ' C F . ' was - .8266. 187 only d is tance, D.. (.8401), was s i g n i f i c a n t l y re lated with v e l o c i t y of in-migrat ion ( I V . . ) . Thus, i t may be said that almost a l l the migrants came to Kushtia from Dacca to minimize the distance t r a v e l l e d . The factor score pattern of the i th d i s t r i c t s indicated that the ve loc i ty of migration to Kushtia was as expected from a l l the i t h d i s t r i c t s (compare rank orders in columns 1 d ' and ' b ' ) . 5.8.5 Summary of Section 5.8 Like Urban-to-Urban out-migration streams in Urban-to-Urban in-migrat ion streams also Dacca overshadowed a l l other 'urban' d i s t r i c t s in terms of the number of people coming to them. In a l l the three rec ip ient d i s t r i c t s such as Chittagong, Dinajpur and Kushtia, the majority of im-migrants was from Dacca. When the factor loadings of the var iables in the ' C . F . ' were interpreted in terms of the 'push-pu l l ' theory, i t was observed that both 'push' and ' p u l l ' factors were operat ing. People were 'pushed' from the i th d i s t i r c t s i f the population pressure on agr icutural land, RD/SQMCA, was greater than in a given j t h d i s t r i c t or i f there were more landless agr i cu l tu ra l labourers, LLALF/TALF. On the other hand, the a v a i l a b i l i t y of more land under j u t e , Jute/NCA, and the greater level of urbanizat ion, U..3 in a given j t h 188 d i s t r i c t ' p u l l e d ' people to that d i s t r i c t . The 'push' factors were found to be dominant in determining the flow of migrants coming to Chittagong and Dinajpur while ' p u l l ' was dominant in the case of Dacca. Neither any 'push' nor any ' p u l l ' was operating in determining the flow of migrants to Kushtia. Thus, i t may be said that , on the average, 'push' factors were dominant in determining the overa l l spat ia l pattern of U-U in-migration streams in .1951. Distance played a strong ro le in determining the spat ia l pattern of U-U in-Migrat ion streams. From the ' C . F . ' s i t was found that factor loading was s i g n i f i c a n t in a l l four 'urban' d i s t r i c t (see Table 5F). The strong inf luence of distance was also confirmed when D^ was regressed against IV.... In a l l the four 'urban' d i s t r i c t s distance was s i g -n i f i c a n t l y re lated with migration v e l o c i t y (Figure 5.6) . The general tendency of the migrants was to minimize the distance involved in migrat ion. The factor score pattern of the i th d i s t r i c t s revealed that the ve loc i ty of migration to a l l the four 'urban' d i s t r i c t s was as expected from a l l the i th d i s t r i c t (Table 5F). V a l i d i t y of Hypothesis S Based on the resu l ts of the analysis discussed in sections 5.7 and 5.8 i t may be said that the Hypothesis 3 is p a r t i a l l y v a l i d . It was observed that , in 1951, though both Table 5 F D E T E R M I N A N T S O F T H E S P A T I A L P A T T E R N O F U - U I N - M I G R A T I O N S T R E A M S , 1 9 5 1 a b c d e IV. . from Was the spat ia l d i s t r i b u t i o n skewed toward the nearer d i s t r i c t s ? Which of the factors , were dominant - 'push 1* or ' p u l l 1 ? Was-distance s ign i f i cant? Was the v e l o c i t y of migration as expected from the i th d i s t r i c t s ? 1. Chi ttagong Yes Push (RD/SQMCA) Yes A l l 2. Dacca No Bull (Jute/NCA and U 1 3 ) Yes A l l 3. Di naj pur No Push (LLALF/TALF) Yes A l l 4. Kushti a Yes No 1 push' or ' p u l l ' Yes A l l CO o 1 90 ^ ^  -i r i t i — i — i — r t i l l T ~ -" I — I — I — I I - l I I I T 1 t I t T ] I I I I I 100 150 200 250 360 350 D15TFNCE IN MILES. DIJ 1—1—1—r 401 Figure 5.6. 'Distance E l a s t i c i t y ' of U-U In-Migration Streams, 1951. 191 'push' and ' p u l l ' factors were operat ing, contrary to our expectat ion, ' p u l l ' factors were dominant in Urban-to-Urban out-migration streams and 'push' in Urban-to-Urban in-migrat ion streams. Thus, the f i r s t part of the Hypothesis 3 has been found to be not v a l i d . The second part of the Hypothesis i s , however, v a l i d as distance had more e f fect on Urban-to-Urban in-migration streams than in Urban-to-Urban out-migration streams. 5.8.6 Summary Results of a l l Migration Streams, in 1951 z~a'nd/'thesV^lidityf of v Hypbthesi s 4 From the previous discussions of the resu l ts of factor analysis of a l l the six kinds of migration streams, i t may be said that though both 'push' and ' p u l l ' factors were operating in regulat ing migration streams in Bangladesh in 195(1, ' p u l l ' factors were stronger than 'push' factors in a majority of the cases. This resu l t i s , however, contrary to our expectations and thus disproves the v a l i d i t y of Hypothesis 4. 5.9 Determinants of the Spatial Pattern of Rural-to-Rural  Out-Migration Streams (0V i •.), 1961 5.9.1 Determinants of the Spatial Pattern of Out- Migration from Barisal 1 92 The neighbouring d i s t r i c t s of Noakhali , Faridpur and Jessore received the majority of migrants from Barisal (Table 5.52, column ' a ' ) . * From the ' C F . ' i t appeared that the presence of more land under r i c e in a given j t h d i s t r i c t than Bar isal acted as a ' p u l l , ' Rice/TCA (.8631). On the other hand the presence of too many landless a g r i c u l t u r a l labourers and share croppers, LLALF/TALF (-.6185) and SC+LLALF/TALF (-.4977), in Bar isal compared to a given j th d i s t r i c t 'pushed' people away from th is d i s t r i c t . The ' p u l l ' appeared to be dominant. Dli:s-1anee 1Wa.saf• s.i• gr i i f i cant factor D. . (.7977), and people try ing to minimize the distance t rave l l ed f locked to the neighbouring d i s t r i c t s . The factor score pattern of the j t h d i s t r i c t s suggested that the ve loc i ty of migration from Barisal was as expected to the majority of the j t h d i s t r i c t s (compare rank orders in columns 'b ' and ' d ' ) . 5.9.2 Determinants of the Spatial Pattern of Out- Migration from Bogra The majority of the migrants from Bogra seemed to have gone to the neighbouring d i s t r i c t s (Table 5.53, column ' a ' ) . Factor loading on OV.. was -.8975. 193 From the ' C F . 1 i t appeared that population pressure on agr i cu l tu ra l land, RD/SQMCA (-.9342), and the presence of too many agr i cu l tu ra l labourers , LLALF/TALF (- .8856), 'pushed' people away from Bogra. On the other hand, people were ' p u l l e d ' to those j t h d i s t r i c t s where c u l t i v a b l e land, TCA/NCA (.7085), was more than in Bogra. But the 'push' was stronger. Since loading on d istance, D.. (.8567), was s i g -n i f i c a n t i t indicated that the migrants tneded to go to the nearer d i s t r i c t s in greater number to minimize the distance i nvolved. From the factor score pattern of the j t h d i s t r i c t s i t was found that the ve loc i ty of migration from Bogra was as expected to a l l the major rec ip ient d i s t r i c t s (compare rank orders in columns 'b ' and ' d ' ) . 5.9.3 Determinants of the Spatial Pattern of Out- Migration from Comilla As indicated by Table 5.54, column 'a,' the d i s t r i b u -t ion of out-migrants from Comilla was quite even though the neighbouring d i s t r i c t s received the highest number of migrants. * From the ' C . F . ' i t appeared that people were ' p u l l e d ' away to those j th d i s t r i c t s where more land under r i c e , *Loading on OV.. in the ' C . F . ' was - .8410. 1 94 Rice/TCA (.7959), was ava i lab le than in Comi l la . On the other hand people were 'pushed' from this d i s t r i c t i f popula-t ion pressure on a g r i c u l t u r a l land, RD/SQMCA (-.6242), was more than in any given j th d i s t r i c t . But the ' p u l l ' was domi nant. As loading on D.. was not s i g n i f i c a n t (i?0799), dt c appeared that people moved away from Bogra as soon as these condit ions were met i r respect ive of the distance i n -volved in the journey. 5.9.4 Determinants of the Spatial Pattern of Out- Migration from Dinajpur As indicated by Table 5.55, column 'a,' out-migrants from Dinajpur concentrated in three contiguous d i s t r i c t s , namely Rangpur, Bogra and Rajshahi. It appeared that the high in tens i ty of cropping, TCA/NCA (.9719), in a given j t h d i s t r i c t ' p u l l ' e d people away from Dinajpur as more jobs were ava i lab le due to the a v a i l a b i l i t y of more cu l t ivated land. At the same time the presence of more landless a g r i c u l t u r a l labourers , LLALF/TALF (- .8041), in Dinajpur than in a given j t h d i s t r i c t 'pushed' people away from that d i s t r i c t . The ' p u l l ' was stronger, however, Loading on OV.. in the ' C . F . ' was - .9737. 195 Distance involved in the journey was an important f a c t o r , D.. (.7469), and the migrants tended to minimize the distance t r a v e l l e d by going to the hearer d i s t r i c t s in greater numbers. By examining factor scores on the subjects in column ' d ' and comparing their rank order with that of in column ' b,' i t could be inferred that the v e l o c i t y of migration from Dinajpur was as expected to the majority of the j t h d i s t r i c t s . 5.9.5 Determinants of the Spatial Pattern of Out- Migration from Faridpur The neighbouring d i s t r i c t s received the majority of migrants from Faridpur (Table 5.56, column ' a ' ) . * As indicated from the ' C . F . ' the presence of more land under r i c e in a given j t h d i s t r i c t , Rice/TCA (.5322), than Faridpur acted as a ' p u l l . 1 That was the only factor operat ing. Therevwas f lnop'J:push.' The low and non-s ign i f i cant loading on D.. (00816) suggested that the migrants were not being inf luenced by the distance involved in the t r a v e l . From the factor score c h a r a c t e r i s t i c s of the receiv ing d i s t r i c t s i t could be found that the ve loc i ty of migration from Faridpur was as expected to the majority of the j th d i s t r i c t s (compare rank orders in columns 'b ' and ' d ' ) . *Loading on OV,.. in the ' C . F . ' was - .9319. Table 5.56 R - R O U T - M I G R A T I O N S T R E A M S , 1 9 6 1 S P A T I A L P A T T E R N OF O V I J FROM F A R I D P U R , 1 9 6 1 Out-•Migration from the i t h D i s t r i c t ~ Faridpur a" "b" 11 c" II d " D i s t r i c ts of Dest inat ion (jth D i s t r i c t s ) No. of in the Mi grants Stream Ve loc i ty of Out-.Migration Streams D i stance Separati ng i and j ( in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor ' Mi j Rank OVij Rank Dij Fj* - Rank Bar isa l Bogra Comi 11 a Dinajpur Jessore Mymensi ngh Noakhali Pabna Rajshahi Rangpur Sylhet 18,022 1 ,168 10,407 197 26,908 2,322 1,046 4,745 2,511 1 ,865 1 ,196 (2) (9) (3) (11) (1) (6) (10) . (4) (5) (7) (8) .0665 .0115 .0372 .0006 .1932 .0052 .0008 .0380 .0139 .0077 .0053 iii (4) (11) (1) (8) (10) (3) (5) (7) (9) 55 113 62 195 50 100 75 64 125 162 145 -0.3815 0.5517 -0.9109 1 . 5330 -1 .4231 0.1666 -0.6207 -1.3563 0.5603 1.0071 0.8737 (3) (11) (1) (6) (4) (2) (8) (10) (9) *Factor 1 ( ' C . F. 1 ) OVij Rice/TCA Dij -0.9319 0.5322 0.0816 (not s i g n i f i c a n t ) 10 197 5.9.6 Determinants of the Spatial Pattern of Out- Migration from Jessore From Table 5.57, column ' a , ' i t could be found that more than 50% of the out-migrants from Jessore went to only one d i s t r i c t - Far idpur. * From the loading c h a r a c t e r i s t i c s in the ' C . F . ' of the factor matrix i t may be inferred that both 'push' and ' p u l l ' factors were operating but the ' p u l l ' was stronger. It appeared that the migrants were pr imar i ly looking for land to growrrice and were ' p u l l e d ' to those j th d i s t r i c t s where land under r i c e was more than in Jessore, Rice/TCA (+.9075). On the other hand population pressure on a g r i c u l t u r a l land, RD/SQMCA (-.4524) acted as the 'push' in Jessore. Distance was a s i g n i f i c a n t f a c t o r , D., (.5958), and i t appeared that the migrants from Jessore tended to minimize the distance t r a v e l l e d . By examining the factor score pattern on the subjects in column ' d ' and comparing the i r rank order with that of in column 'b ' i t was found that the v e l o c i t y of migration from Jessore was not as expected in the majority of the j th d i s t r i c t s . "Loading on OV.. in the ' C F . ' was -.8524. 198 5.9.7 Determinants of the Spat ia l Pattern of Out- Migration from Mymensingh From Table 5.58, column ' a , ' i t may be observed that the majority of the migrants coming out from Mymensingh went to neighbouring d i s t r i c t s of Rangpur, Sy lhet , Comilla and Pabna. By examining the loading c h a r a c t e r i s t i c s in the * ' C F . ' of the factor matrix i t may be inferred that only ' p u l l ' factors were operat ing. There was no 'push . ' It appeared that people were pr imar i ly looking for cu l t i va ted land and were ' p u l l e d ' to the j t h d i s t r i c t s i f the in tens i ty of cropping, TCA/NCA (- .7996), was less as more p o s s i b i l i t i e s were there to expand cu l t i va ted land. People were also ' p u l l e d ' to the j th d i s t r i c t s i f more land under r i c e , Rice/TCA (.6515), was ava i lab le than in Mymensingh. Low and thus non-s ign i f i cant loading on D.. (.1256) indicated that people tended to go to the i r d i s t r i c t s of dest inat ion without considering the distance involved in the travel . The factor score pattern of the j th d i s t r i c t s revealed that the ve loc i ty of migration from Mymensingh was as expected to the majority of the receiv ing d i s t r i c t s (compare rank orders in columns 'b ' and ' d ' ) . Loading on OV.. in the ' C F . ' was high and negative (- .8615). J 1 99 5.9.8 Determinants of the Spatial Pattern of Out- Migration from Noakhali Unlike other d i s t r i c t s , the highest and second highest number of people from th is d i s t r i c t went to far away d i s t r i c t s l i k e Jessore and Rangpur. Migrants went out in every d i rect ion and at every distance (Table 5.59, column 'a').. By examining factor loadings of the var iables in * the ' C F . ' of the factor matrix i t may be said that people were ' p u l l e d ' to those j th d i s t r i c t s where more land under r i c e , Rice/TCA (.9069), was ava i lab le than in Noakhali . There was no 'push . ' Since loading on d istance, D.. (.5820), was s i g n i f i -cant i t indicated that the migrations tended to go to the nearer d i s t r i c t s in greater number to minimize the distance travel 1ed. From the factor score pattern of the j t h i t was found that the ve loc i ty of migration from Noakhali was not as expected.to the majority of the receiv ing d i s t r i c t s (compare rank orders in columns 'b ' and ' d ' ) . This suggested that the inf luence of Factor 4, where also the loading on OV.. was s i g n i f i c a n t , was quite strong. Loading on OV.. in the ' C . F . ' was - .6366. 200 5.9.9 Determinants of the Spatial Pattern of Out- Migration from Pabna In 1961, d i s t r i c t s adjoining to Pabna continued to receive the majority of people from Pabna (Table 5.60, column 'a 1 ) . By examining the factor loading c h a r a c t e r i s t i c s * in the ' C F . ' of the factor matrix i t was found that only a ' p u l l ' factor was operat ing. There was no_ 'push . ' It appeared that people were pr imar i ly looking for cu l t i va ted land and thusj was ' p u l l e d ' to the j th d i s t r i c t s i f the cropping in tens i ty was l e s s , TCA/NCA (-.8791), as there was more opportunit ies to expand cu l t ivated land. S i g n i f i c a n t loading on D. . (.8877) indicated that people tended to minimize the distance t r a v e l l e d . The factor score pattern of the subjects confirmed that the majority of the people moved to the nearer d i s t r i c t s only for the aforementioned factors (compare rank o r d e r s i i n columns 1 d ' and 1 b 1 ) . 5.9.10 Determinants of the Spatial Pattern of Out- Migration from Rajshahi The majority of people continued to go to the neighbouring d i s t r i c t s in 1961 also (Table 5.61 ,<*• col umn cU-'). * 'Factor loading on OV ,^ in the ' C . F . ' was - .8373. 201 By examining the factor loading c h a r a c t e r i s t i c s of * the ' C F . ' i t appeared that both 'push' and ' p u l l ' factors were operating but the ' p u l l ' was stronger. It appeared that the migrants were pr imar i ly looking for cu l t ivated land and thus were p u l l e d 1 to the j th d i s t r i c t s i f the in tens i ty of cropping, TCA/NCA (- .8534), was less as there was more opportunit ies to expand cu l t ivated land. People were also 'pushed' from Rajshahi i f the presence of share croppers and landless agr i cu l tura l labourers, SC+LLALF/TALF (- .4530), was more than in a given j th d i s t r i c t . High and pos i t ive loading on D.. (.8380) indicated that people tended to minimize the distance t r a v e l l e d . By examining the factor score pattern of the j t h d i s t r i c t s i t may be said that the ve loc i ty of migration from Rajshahi was as expected to the majority of the receiv ing d i s t r i c t s (compare rank orders in columns 'b ' and ' d ' ) . 5.9.11 Determinants of the Spatial Pattern of Out- Migration from Rangpur In 1961, most of the out-migrants from Rangpur, continued to go to the neighbouring d i s t r i c t s (Table 5.62, columns 'a 1 ) . Looking at the factor loading c h a r a c t e r i s t i c s in * the ' C . F . ' of the factor matrix i t may be interpreted that * Loading on OV.. in the ' C F . ' was - .8651. 202 both 'push' and ' p u l l ' factors were operating but the ' p u l l ' was dominant. High and pos i t ive loadings on Rice/TCA (.7357) and marginally s i g n i f i c a n t loading on TCA/NCA (-.4560) indicated that people were ' p u l l e d ' to those j th d i s t r i c t s where land under r i c e was more and in tens i ty of cropping was lesssthan in Rangpur. On the other hand, population pressure on the agr i cu l tu ra l land in Rangpur, RD/SQMCA (-. 641 3) , etacted as the 'push. ' High loading on d istance, D.. (.9306), indicated that the migrants tended to minimize the distance t r a v e l l e d . The factor score pattern of the j t h d i s t r i c t s indicated that the ve loc i ty of migration from Rangpur was as expected to the top receiv ing d i s t r i c t s (compare rank orders in columns 1 b'aand 1d ' ). 5.9.12 Determinants of the Spatial Pattern of Out- Migration from Sylhet As indicated in Table 5.63, column ' a ' , the majority of out' mi grants from Sylhet went to the neighbouring d i s t r i c t s of Mymensingh and Comil la . * From the ' C F . , ' i t appeared that the migrants were pr imar i ly looking for land to grow r i c e , Rice/TCA (.8515), and, as such, were ' p u l l e d ' to those j th d i s t r i c t s Factor loading on OV.. in the ' C . F . ' was - .9487. 203 where such land was ava i lab le more than in Sylhet . On the other hand, the presence of too many landless a g r i c u l t u r a l labourers, LLALF/TALF (-.4032), acted as the 'push' factor but the 'push' was very mi ld . S i g n i f i c a n t loading on D.. (.8321) indicated that the migrants tended to minimize the distance t r a v e l l e d . The nature and extent of factor scores on the j th d i s t r i c t s indicated that the ve loc i ty of migration from Sylhet was as expected to the majority receiv ing d i s t r i c t s (compare rank orders in columns ' d ' and ' b ' ) . 5 5.9.13 Summary of Section 5.9 From the discussion of the spat ia l pattern of R-r out-migration streams in 1961, i t was discovered that the R-R out-migration continued to be a loca l phenomenon and thus the spat ia l d i s t r i b u t i o n s of migration streams were skewed towards the nearer d i s t r i c t s . In f a c t , i t appeared that more people were going to the neighbouring d i s t r i c t s in 1961 than 1951. In 1951, the majority of migrants from a l l but f i ve d i s t r i c t s went to the contiguous d i s t r i c t s (see Table 5A and section 5.3) , but in 1961, there were only two d i s t r i c t s from where the majority of the migrants did not go to the neighbouring d i s t r i c t s (Table 5G). This indicated that 'd istance e l a s t i c i t y ' of migrat ion, i . e . rate of change of migration ve loc i ty per unit change of d is tance, increased by 1961 rather than decreasing. Table 5G D E T E R M I N A N T S O F T H E S P A T I A L P A T T E R N OF R - R O U T - M I G R A T I O N S T R E A M S , 1 9 6 1 a b c d e OV. . from Was the spat ia l di s t r i buti on skewed towards the nearer d i s t r i c t s ? Which of the factors were dominant -1 push 1 or ' p u l l 1 ? Was distance s i g n i f i c a n t ? Was the ve loc i ty of migration as expected to the majori ty of j th d i s t r i c t s ? 1. Bar isa l Yes Pull (Rice-TCA) Yes No I 2- Bogra Yes Push (RD/SQMCA) Yes No 1 ! 3 -Comi 11 a No Pull (Rice/TCA) No * No •4 t Di naj pur Yes Pull (TCA/NCA Yes * No t !5. i Faridpur Yes Pull (Rice/TCA) No Yes i •6. Jessore Yes Pull (Rice/TCA) Yes No I ! 7 - Mymensi ngh Yes Only pull (TCA/NCA and Rice/TCA) No Yes 8. Noakhali No Only pull (Rice/TCA Yes No The v leoc i ty of migration was as expected to the top rec ip ient d i s t r i c t s . ^ CONTINUED 2 Table 5G (Continued) a b c d e OV.. from Was the spat ia l d i s t r i b u t i o n skewed towards the nearer d i s t r i c t s ? Which of the factors were dominant -1 push' or ' p u l l ' ? Was distance s ign i f i cant? Was the ve loc i ty of migration as expected to the majority of j t h di s t r i cts? 9. Pabna Yes Only pull (TCA/NCA) Yes * No 10. Raj shahi Yes Pul l (TCA/NCA) Yes * No 11 . Rangpur Yes Pull (Rice/TCA) Yes * No 12. Sylhet Yes Pull (Rice/TCA) Yes * No ro o cn o O-i o LO ro OH o o i co oH o LO OJ OH o o OJ OH o LO OH o o oH o LO o 0 040 030 020 010 \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ 3 L \ \ -^ — v 206 \ DISTANCE IN MILES. 01J Figure 5.7. 'Distance E l a s t i c i t y ' of R-R Out-Migration Streams, 1961. 207 From the pr inc ipa l axis type of factor analys is i t was observed that both 'push' and ' p u l l ' factors were operat-ing in determining the spat ia l pattern of out-migration streams. People were 'pushed' from the i th d i s t r i c t s i f the presence of landless a g r i c u l t u r a l labourers, LLALF/TALF, was more than in a given j th d i s t r i c t and/or i f the population pressure on a g r i c u l t u r a l land, RD/SQMCA, was more than in a given j th d i s t r i c t . At the same time, the a v a i l a b i l i t y of more land under r i c e , Rice/TCA, in the j th d i s t r i c t s ' p u l l e d ' people away from the i th d i s t r i c t s . There were also occasions when the migrants were ' p u l l e d ' to the j t h d i s t r i c t s i f the p o s s i b i l i t i e s of expansion of cu l t i va ted land were more due to the low intens i ty of cropping, TCA/NCA. Thus these four v a r i a b l e s , out of eleven, were found to be most ac t i ve ly operating e i ther as 'push' or ' p u l l . ' However, the par-t i c u l a r 'mix' of 'push' and ' p u l l ' factors was d i f f e r e n t . From Table 5G, i t may be observed that the ' p u l l ' factors were dominant in determining the spat ia l pattern of out-migration from a l l but one d i s t r i c t - Bogra. This shows that the ' p u l l ' was the dominant factor in determining the overal l spat ia l pattern of R-R out-migration streams in 1961. Distance had a var iable ro le in determining the spat ia l pattern of out-migration streams or ig inat ing from 12 d i s t r i c t s in 1961 a lso . It was s i g n i f i c a n t l y re lated with OV,. in nine cases (Table 5G, column ' d ' ) . Thus, on the average, i t played an important r o l e . 208 However, a l i t t l e d i f fe rent p icture emerged when D. . was regressed against OV. . , c e t e r i s paribus. It was discovered that distance was s i g n i f i c a n t l y re lated with ve loc i ty of migration in a l l but two d i s t r i c t s - Bar isal and Noakhali . In other words, for these two d i s t r i c t s , the ve loc i ty of out-migration was independent of distance separating the i th and j th d i s t r i c t s . It may be noted that from factor analysis i t was found that loading on D.. was "13 not s i g n i f i c a n t in the ' C F . ' descr ibing the determinants of migration or ig inat ing from Comi l la , Faridpur and Mymensingh -not from Bar isal and Noakhali . This indicated that distance acting as the intervening var iab le had a d i f fe rent role than distance acting as the explanatory v a r i a b l e , c e t e r i s parebus. By comparing Figures 5.1 and 5.7 i t was confirmed that the 'distance e l a s t i c i t y ' of R-R out-migration increased during the decade, 1951-1961. From the factor score pattern of the j t h d i s t r i c t s i t was revealed that , unl ike 1951, in 1961, the ve loc i ty of migration to the majority of j th d i s t r i c t s was not as expected from ten out of 12 d i s t r i c t s (Table 56, column e ) . However, the top rec ip ient d i s t r i c t s continued to get the expected numbers of migrants from the majority of the i th d i s t r i c t s . 209 5.10 Determinants of the Spatial Pattern of Rural-Rural  In-Migration Streams (IV^j), 1961 5.10.1 Determinants of the Spatial Pattern of In- Migration to Bar isal Like 1951, in 1961 also Bar isal received most of her in-migrants from the neighbouring d i s t r i c t s such as Faridpur, Naokhali and Comil la . Though other d i s t r i c t s also send considerable numbers of people, in general , the d i s t r i -bution was skewed to the neighbouring d i s t r i c t s (see column ' a ' in Table 5.64)• * By examining loading c h a r a c t e r i s t i c s in the ' C . F . ' of the factor matrix i t may be said that population pressure on a g r i c u l t u r a l land, RD/SQMCA (.6242), and the presence of too many landless agr i cu l tura l labourers, LLALF/TALF (.4250), 'pushed' people away from the i th d i s t r i c t s . On the other hand, people were ' p u l l e d ' to Bar isal i f r i c e land, Rice/TCA (-.6654), was ava i lab le there more than in a given i th d i s t r i c t . The ' p u l l ' was a l i t t l e stronger than the 'push . ' The nature of loading on d istance, D.. (.8746), indicated that people tended to come from the nearer i th d i s t r i c t s in greater numbers and the inf luence of distance was very strong. Loading on IV . . in the ' C F . ' was -.9502. 210 The nature and extent of factor scores on the i th d i s t r i c t s indicated that the ve loc i ty of migration to Bar isal was as expected from the top sending d i s t r i c t s (compare rank orders in columns ' d ' and ' b ' ) . 5.10.2 Determinants of the Spatial Pattern of In- Migration to Bogra Migrants to Bogra continued to come, in greatest numbers, from the neighbouring d i s t r i c t s of Pabna, Rajshahi and Rangpur (see column ' a ' of Table 5.65). Looking at the loading c h a r a c t e r i s t i c s of the * var iables in the ' C F . ' of the factor matrix i t may be i n t e r -preted that people were 'pushed' from those i th d i s t r i c t s where population pressure on agr i cu l tu ra l land, RD/SQMCA (.9433), and the presence of landless agr i cu l tu ra l labourers , LLALF/TALF (.8220), were more than in Bogra. On the other hand, the migrants were ' p u l l e d ' to Bogra i f the opportunit ies were greater for the expansion of cu l t i va ted land due to the low in tens i ty of cropping, TCA/NCA (.7578). But the 'push' was stronger. The nature and magnitude of loading on d istance, D.. . (.8902), indicated that the migrants tended to minimize the distance t r a v e l l e d . "Loading on IV. . in the ' C F . ' was - .8445. 211 The factor score pattern of the i th d i s t r i c t s indicated the ve loc i ty of migration to Bogra was as expected from the top sending d i s t r i c t s (compare rank orders in columns 'b ' and 'd ' ) . 5.10.3 Determinants of the Spatial Pattern of In- Migration to Comilla As indicated by Table 5.66, columns ' a ' and ' c , ' d i s t r i b u t i o n of in-migration to Comilla had been quite even, though the neighbouring and contiguous d i s t r i c t s such as Noakhali , Faridpur and Sylhet did send the majority of migrants to this d i s t r i c t . Looking at the loading c h a r a c t e r i s t i c s of the * var iables in the ' C F . ' of the factor matrix, i t may be interpreted that people were 'pushed' from the sending d i s t r i c t s i f the presence of landless agr i cu l tu ra l labourers , LLALF/TALF (.4483), was more than in Comil la . There was no 'pu l l . 1 The nature and magnitude of loading on d istance, D. . (.8286), indicated that distance was the most important var iab les . Thus, we may say that most of the migrants came from the neighbouring d i s t r i c t s in order to minimize the distance t r a v e l l e d . Loading on IV. . in the ' C . F . ' was (- .8410). 212 The factor score pattern of the i th d i s t r i c t s indicated that the ve loc i ty of migration to Comilla was as expected from the top sending d i s t r i c t s (compare rank orders in columns 'b ' and ' d ' ) . 5.10.4 Determinants of the Spatial Pattern of In- Migration to Dinajpur As indicated by Table 5.67,column ' a , 1 the largest numbers of people came from Mymensingh, followed by Rangpur, Domilla and Rajshahi. Among them only Rangpur is contiguous to Dinajpur. Thus the d i s t r i b u t i o n pattern of in-migration to Dinajpur was not skewed to the neighbouring d i s t r i c t s . Looking at the loading c h a r a c t e r i s t i c s of the var iables in the ' C F . ' 1 of the factor matrix, i t may be interpreted that the p o s s i b i l i t i e s of further expansion of cu l t i va ted land due to the low in tens i ty of cropping, TCA/NCA (-.8944)", in Dinajpur ' p u l l e d ' people to th is d i s t r i c t . At the same time the population pressure on a g r i c u l t u r a l land, RD/SQMCA (- .5133), 'pushed' people from the i t h d i s t r i c t s . But the ' p u l l ' was stronger. The negative and high loading on j u t e , Jute/NCA (-.8303), was harder to i n t e r p r e t . The nature of loading suggested that the more jute c u l t i v a t i o n there was in a given i th d i s t r i c t , the less was the migration to Dinajpur from that d i s t r i c t - which is contrary to our expectat ion. As the ra t io Jute/NCA was also used as a *Loading on IV in the ' C . F . ' was .6984. i j 21 3 surrogate measure of income of the farmers, we expected that less people would go out;from those d i s t r i c t s where there was more jute c u l t i v a t i o n . Low and thus non-s ign i f i cant loading on d istance, D.. (- .2742), indicated that distance separating i th and j th d i s t r i c t s was not an important factor in regulat ing in-migration to Dinajpur. On the other hand, high and pos i t ion loading on a c c e s s i b i l i t y , E3 (.8221), suggested that access i -b i l i t y was the important f a c t o r , not d istance. The factor loading c h a r a c t e r i s t i c s of the i th d i s t r i c t s revealed that except for the top sending d i s t r i c t s the ve loc i ty of migration to Dinajpur was not as expected from the majority of the sending d i s t r i c t s (compare rank orders in columns ' d ' and ' b ' ) . 5.10.5 Determinants of the Spatial Pattern of In- Migration to Faridpur Like 1951, in 1961 aos most of the migrants con-tinued to come from the neighbouring d i s t r i c t s , namely, Comi l la , Jessore, Pabna and Barisal (Table 5.68, column ' a ' ) . By examining loading c h a r a c t e r i s t i c s of the var iables * in the ' C . F . ' of the factor matrix i t may be interpreted that the a v a i l a b i l i t y of more land under r i c e , Rice/TCA (-.5322), in Faridpur than in a given i th d i s t r i c t ' p u l l e d ' migrants to this d i s t r i c t . There was no 'push. ' Loading on IV... in the ' C . F . ' was - .9319. 214 The nature and magnitude of loading on d istance, D. . (.9426), indicated that distance was the most important factor and people came from the neighbouring d i s t r i c t s to Far idpur, in greater number, in order i to minimize the distance t r a v e l l e d (compare rank orders in columns ' d ' and ' b ' ) . . 5.10.6 Determinants of the Spatial Pattern of In- Migration to Jessore As indicated by Table 5.69, column ' a , ' the majority of migrants came from Noakhali , Comi l la , Far idpur, Bar isal and Pabna. Only Faridpur is contiguous to Jessore among them. Thus i t may be said that the spat ia l d i s t r i b u t i o n of in-migration to this d i s t r i c t was quite even as migrants came from a l l d i rect ions i r respect ive of the distance involved in migrat ion. From the loading c h a r a c t e r i s t i c s of the var iables * in the ' C F . ' of the factor matrix i t may be said that people were ' p u l l e d ' to Jessore i f more land under r i c e , Rice/TCA (-.9075), was ava i lab le than in a given sending ( i th) d i s t r i c t On the other hand, the population pressure on a g r i c u l t u r a l land, RD/SQMCA (.4543), 'pushed' people from the i t h d i s t r i c t s but the 'push' was mi ld . Loading on IV... in the ' C . F . ' was -.8524. Table 5.69 R - R I N - M I G R A T I O N S T R E A M S , 1 9 6 1 S P A T I A L P A T T E R N OF I V I J TO J E S S O R E , 1 9 6 1 D i s t r i c t of Destination (jth D i s t r i c t ) - Jessore "a" "c" "d D i s t r i c t s of Or ig in ( i th D i s t r i c t s ) No. of Migrants in the Stream Ve loc i ty of Out-Migration Streams Separati ng Distance i and j (in mi 1es ) Factor Scores on the Subjects from the ' C r i t i c a l Factor ' IMij Rank IV i j Rank Di j Bar isa l Bog ra Comil la Dinajpur Fari dpur Mymensi ngh Noakhali Pabna Rajshahi Rangpur Sylhet 6,385 298 30,207 177 25,908 1 ,458 31 ,903 2,259 911 406 295 (4) (9) (2) (10) (3) (6) (1) (5) (7) (8) (11) .0346 .0043 .1 592 .0024 .1958 .0048 .3096 .0266 .0075 .0025 .0019 (4) (8) (3) (10) (2) (7) (1) (5) (6) (9) (11) 87 105 113 175 50 125 126 65 90 162 182 Fj Rank 0.1778 (6) 0.2295 (4) 1 .3514 (2) 0.9585 (9) 1 .1866 (3) 0.4201 (7) 1.4897 (1) 0.2229 (5) 0.4820 (8) 1.4451 (11) 0.9970 (10) Factor 1 ( ' C.F . ' ) IV i j -0.8524 Rice -0.9075 Dij 0.5958 RD/SQMCA 0.4543 216 The nature and magnitude of loading on d is tance , D. . (.5958), indicated that i t was an important factor but not as important as the ' p u l l ' f a c t o r . The factor score pattern of the i th d i s t r i c t s indicated that the ve loc i ty of migration to Jessore was as expectef rom the top senders of migrants (compare rank orders in columns 1b 1 a nd ' d ' ). 5.10.7 Determinants of the Spatial Pattern of In- Migration to Mymensingh As indicated by Table 5.70, columns ' a ' and ' c , ' the three neighbouring d i s t r i c t s of Pabna, Comilla and Sylhet continued to be the main sources of in-migrat ion to Mymensi ngh. The factor loading c h a r a c t e r i s t i c s of the var iables * in the ' C . F . ' of the factor matrix indicated that the migrants were ' p u l l e d ' to Mymensingh i f land under r i c e , Rice/TCA (-.4035), was more than in a given i th ;di .str i c t . But the ' p u l l ' was very mild andhthere was no 'push . 1 The nature and magnitude of factor loading on d istance, D.. (.5996), indicated that distance was the most important factor in determining the flow of in-migrants to th is d i s t r i c t and people, t ry ing to minimize the distance * Loading on IV . . in the ' C . F . ' was -.9714. 217 involved in the journey, came from the neighbouring d i s t r i c t s in greater numbers. The factor score pattern of the i th d i s t r i c t s confirmed that the ve loc i ty of migration to Mymensingh was highest from the neighbouring d i s t r i c t s (compare rank orders in columns ' d , ' 'b ' and ' c ' ) . 5.10.8 Determinants oftthe Spatial Pattern of In- Migration to Noakhali As indicated by Table 5.71 , col urnns ' a ' and 1 c , ' most of the in-migrants to Noakhali came from the neighbouring d i s t r i c t s of B a r i s a l , Comilla and Faridpur. Thus the d i s t r i -bution of in-migration was skewed to the neighbouring d i s t r i c t s . From the factor analysis i t was found that in the rotated factor matrix loading on ve loc i ty of in-migrat ion (IV.^.) was s i g n i f i c a n t on two f a c t o r s . In factor 1, the ' C F . , 1 loading on IV . . was -.6366. S i g n i f i c a n t loadings on other var iables were as fo l lows: Rice/TCA (-.9062), D.. (.5820) and U., (- .4705). In the other factor ( factor 4) loading on IV . . was less than in factor 1, but s i g n i f i c a n t (.5148). Other var iables having s i g n i f i c a n t loadings were the fo l lowing: RD/SQMCA (-.8406), TCA/NCA (.7941), LLALF/TALF (-.7605) and D .^ (-.6489):. Thus i t may be interpreted that as many as seven var iables had s i g n i f i c a n t re la t ionsh ip with I V ^ . It may also be said that both 'push' and ' p u l l ' 218 factors were operating but the ' p u l l ' was r e l a t i v e l y stronger. From the ' C F . ' i t appeared that the migrants were pr imar i ly looking for land to grow r i ce and, as such, ' p u l l e d ' to Noakhali when such land was ava i lab le more than in a given i th d i s t r i c t . There was no 'push. ' On the other hand, from factor 4, i t appeared that the migrants were 'pushed' from those i th d i s t r i c t s where population pressure on agr i cu l tu ra l land and the presence of landless a g r i c u l t u r a l labourers was more than in Noakhali . Some of the migrants were also ' p u l l e d ' to th is d i s t r i c t i f more c u l t i v a b l e land was ava i lab le than in a given i t h d i s t r i c t . The nature and magnitude of loading on D .^ indicated that in a l l the cases, however, i t was necessary that the distance separating Noakhali from the i th d i s t r i c t s was not great . The factor score pattern of the i th d i s t r i c t s revealed that the ve loc i ty of migration to Noakhali was as expected from the major sending d i s t r i c t s (compare rank orders in columns 'b ' and ' d ' ) . 5.10.9 Determinants of the Spatial Pattern of In- Migration to Pabna Most of the migrants, as indicated in Table 5.72, columns ' a ' and ' c , ' came from the neighbouring d i s t r i c t s of Mymensingh, Rajshahi, Faridpur and Bogra. 219 By examining factor loadings of the var iables * in the ' C F . ' of the factor matrix, i t may be interpreted that people were ' p u l l e d ' to Pabna i f the p o s s i b i l i t i e s of further expansion of cu l t i vated land were there due to the low intens i ty of cropping, TCA/NCA (.8796). There was no 'push. ' High and pos i t ive loading on D. . (.8877) suggested that the ' p u l l ' to Pabna was e f fec t i ve only when the d i s t r i c t s of o r i g i n ( i th d i s t r i c t s ) were not very far from Pabna. The factor score pattern of the i th d i s t r i c t s confirmed that the above factors operated e f fec t i ve (compare rank orders . 5.10.10 Determinants of the Spatial Pattern of In- Migration to Rajshahi As indicated by Table 5.73, columns ' a ' and ' c , ' people came to Rajshahi in good numbers not only from the neighbouring d i s t r i c t s but also from the far away d i s t r i c t s . From the factor analys is i t was found that load-ing on ve loc i ty of in-migrat ion ( I V , . ) was s i g n i f i c a n t in three factors of the factor matrix. Factor 4 was considered Loading on IV. . in the ' C . F . ' was - .8407. 220 as the ' C F . , ' since loading on IV. . was the highest. By 13 examining the factor loading c h a r a c t e r i s t i c s of the var iables i t may be interpreted that three systems were operating in determining the spat ia l pattern of in-migration to Rajshahi and both 'push' and ' p u l l ' factors were operat ing. But the ' p u l l ' was stronger. While in the ' C . F . ' there was only ' p u l l , ' in factor 1 and factors 3, both ' p u l l ' and 'push' were operat ing. From the ' C . F . ' i t may be said that people were ' p u l l e d ' to Rajshahi for i t s r e l a t i v e l y higher rate of urbanization (loading on U. . was .6135). S i g n i f i c a n t loading on cumulative a c c e s s i b i l i t y index, Eg (.7070), but non-s ign i f i cant loading on d istance, D.. (- .0424), indicated that in th is case a c c e s s i b i l i t y was a more important explana-tory var iable than the distance. From factor 1 and fac tor 3 i t may be said that some people did also tend to minimize the distance t r a v e l l e d (loadingsvonCD. . were -.5803 and -.6968 r e s p e c t i v e l y ) . • 3 3 In these cases people were ' p u l l e d ' to Rajshahi i f more cu l t i vab le land was ava i lab le than in a given i th d i s t r i c t (loading on TCA/NCA was .9402 in factor 1) or i f the l i t e r a r y rate , was more than in a given i th d i s t r i c t (loading on -Loading on IV . . in the ' C . F . ' was .5930 but loadings ' 3 in factor 1 and factor 3 were -.5228 and .5O038respectively. Percentage of l i t e r a c y is p o s i t i v e l y re lated with the number of higher education centres in Bangladesh. 221 L. . was .5806 in factor 3). On the other hand, the presence of more share croppers and landless agr i cu l tu ra l labourers in a given i th d i s t r i c t than Rajshahi acted as the 'push' factors (loadings on SC+LLALF/TALF and SC/TALF were .8715 and .8959 r e s p e c t i v e l y ) . From the factor score pattern of the i t h d i s t r i c t s in column ' d ' i t may be found that the rank order of the i th d i s t r i c t s has been appreciably a l tered when compared with the rank order of the i th d i s t r i c t s in column ' b . 1 It indicated that the other v a r i a b l e s , which were found to be s i g n i f i c a n t along with IV . . in factor 1 and factor 3, have also inf luenced the spat ia l pattern of in-migration to Rajshahi. Since magnitude of loading on IV^. in those factors was marginally lower, inf luence of those var iables should have been appreciable indeed. 5.10.11 Determinants of the Spat ia l Pattern of  In-Migration to Rangpur As indicated by Table 5.74, column ' a , ' though the bulk of the people continued to come from Mymensingh, in 1961, substant ia l numbers of people did also come from almost a l l other rural d i s t r i c t s . In other words, the d i s t r i b u t i o n of in-migration to Rangpur was quite even. 222 By examining factor loading c h a r a c t e r i s t i c s in * the ' C F . ' of the factor matrix i t may be interpreted that the presence of too many share croppers and landless a g r i c u l -tural labourers in the sending ( i t h ) u d i s t r i c t s (loadings on LLALF/TALF and SC/TALF were -.6436 and -.4892 respect ive , acted as the 'push' factors while people were 'pu l led to Rangpur i f land under j u t e , Jute/NCA (.4323), was more than in a given i th d i s t r i c t . But the ,';push' was dominant. Since loading on distance, D.. (- .1256), was not s i g n i f i c a n t i t may be said that the above factors were operating e f f e c t i v e l y i r respect ive of the distance separating Rangpur from the i th d i s t r i c t s . From the factor score pattern of the i th d i s t r i c t s i t was apparent that the ve loc i ty of migration to Rangpur was as expected from the majority of the i th d i s t r i c t s (compare rank orders in columns ' d ' and ' b ' ) . 5.10.12 Determinants of the Spatial Pattern of In- Migration to Sylhet As indicated by Table 5.75, columns ' a ' and ' c , ' though more than 60% of the total in-migrants to Sylhet came from two contiguous d i s t r i c t - Mymensingh and Comi l la , considerable numbers of people did also come from non-contiguous d i s t r i c t s , namely, Noakhali , Bar isal and Faridpur. * Loading on IV. . in the ' C . F . ' was .8607. 223 By examining factor loading c h a r a c t e r i s t i c s in * the ' C F . ' of the factor matrix i t may be interpreted that people were ' p u l l e d ' to Sylhet i f more land under r i c e , Rice/TCA (- .8546), was ava i lab le than in a given i th d i s t r i c t . On the other hand, the presence of too many landless a g r i -cu l tura l labourers, LLALF/TALF (.4047), acted as the 'push' f a c t o r . But the ' p u l l ' was stronger. High and pos i t ive loading on d istance, D.. (.8309), 3 indicated that the migrants tended to minimize the distance travel 1ed. From the factor score pattern of the i th d i s t r i c t s i t may be said that the ve loc i ty of migration to Sylhet was as expected from the major sending d i s t r i c t s (compare rank orders in columns ' d ' and ' b ' ) . 5.10.13 Summary of Section 5.10 From the discussion of the spat ia l pattern of i n -migration streams coming to 12 d i f fe rent rural d i s t r i c t within the R-R In-Migration Streams, i t was discovered that , except for Dinajpur, Jessore, Rajshahi and Rangpur, the majority of migrants continued to come from the neighbouring d i s t r i c t s (see Table 5H and section 5.10). Thus i t may be said that the spat ia l d i s t r i b u t i o n of R-R in-migrat ion was skewed heavi ly towards the neighbouring d i s t r i c t s . _ Loading on IV,.,. in the ' C . F . ' was - .9475. Table 5H D E T E R M I N A N T S O F T H E S P A T I A L P A T T E R N O F R - R I N - M I G R A T I O N S T R E A M S , 1 9 6 1 a b c d e IV t J . to Was the spat ia l d i s t r i b u t i o n skewed towards the nearer di s t r i cts? Which of the factors were dominant -1 push' or ' p u l l ' ? Was distance s ign i f i cant? Was the ve loc i ty of migration as expected from the majority of i th d i s t r i c t s ? 1. Bar isal Yes Pull (Rice/TCA) Yes Yes 2. Bogra Yes Push (RD/SQMCA & LLALF/TALF) * Yes ** No 3. ComiUa Yes Only push (LLALF/ TALF) Yes ** No 4. Dinajpur No Pull (TCA/NCA) Yes ** No 5. Faridpur Yes Only pull (Rice/ TCA * Yes Yes 6. Jessore No Pull (Rice/TCA) Yes ** No Distance was the most important f a c t o r . The ve loc i ty of migration was as expected from the top sending d i s t r i c t s . . CONTINUED Tab!e 5H (Continued ) a b -j c d e Was the spat ia l d i s t r i bution skewed towards the nearer di s t r i cts? Which of the factors were dominant -' push' or ' p u l l ' ? Was distance s ign i f i cant? Was the v e l o c i t y of migration as expected from the majori ty of i th d i s t r i c t s ? 7. Mymensi ngh Yes Pull (Rice/TCA) * Yes Yes 8. Noakhali Yes Pull (Rice/TCA) Yes ** No 9. Pabna Yes Only pull (TCA/NCA) * Yes ** No 10. Rajshahi No Pull (TCA/NCA) No No 11. Rangpur No Push (LLALF/TALF) No ** No 12. Sylhet Yes Pull (Rice/TCA) Yes Yes ro ro cn o 0-o LO CO OH o o co oH o LO CVJ oH o o CVJ o LO o o OH o LO o 0-040 030 ,020 ,010 -0-226 \ V 0 x \ v\^\ % ^ - - - N \ \ x . ^ $ . s \ \ \ N x ^ = =r ~ ~i—r ' 1 I 1 100 T — i — r 250 • [J -r—|—i i i — i — | — i — i — i — i — | 300 350 400 50 150 200 DISTANCE IN MILES. Figure 5.8. 'Distance E l a s t i c i t y ' of R-R In-Migration Streams, 1961. 227 The pr inc ipa l axis type factor analysis revealed that both 'push' and ' p u l l ' factors were operating in determin-ing the spat ia l pattern of in-migration streams. People were 'pushed' from the i th d i s t r i c t s i f the population pressure on agr i cu l tu ra l land, RD/SQMCA, was more than in a given j th d i s t r i c t , and/or the presence of landless agr i cu l tu ra l labourers, LLALF/TALF, was more than in a given j th d i s t r i c t . At the same time, i f the a v a i l -a b i l i t y of land under r i c e , Rice/TCA, was more in the j th d i s t r i c t s than iii a given i th d i s t r i c t , people were ' p u l l e d ' to these d i s t r i c t s . The p o s s i b i l i t i e s of increasing c u l t i -vated land in the j t h d i s t r i c t s due to the low in tens i ty of cropping, TCA/NCA, also acted as the ' p u l l ' f a c t o r . Thus these four var iab les , out of el eveniexplanatOry v a r i a b l e s , were found to be most ac t i ve ly operating e i ther as ' p u l l ' or 'push' factors in determining the flow of migrants within this kind of migration stream. The p a r t i c u l a r 'mix' of 'push' and ' p u l l ' factors was, however, d i f fe rent in each of the j t h d i s t r i c t s . While 'push' factors were dominant in determining the spat ia l pattern of in-migration to Bogra, Comilla and Rangpur; ' p u l l ' factors were dominant in a l l other d i s t r i c t s (see Table 5H). It may be noted that in the cases of migrants coming to Faridpur and Pabna the ' p u l l ' was not only dominant but also the only factor operating (Table 5H). 228 Thus we may conclude that the ' p u l l ' was dominant in determin-ing the overal l spat ia l pattern of R-R in-migration streams in 1961 a l so . Distance played a var iab le r o l e . It was s i g n i f i -cantly re lated with IV . . in ten out of twleve cases (table 5H, column d)'!.) Thus, on the average, i t played an important rol e. In the cases of migration streams terminating in Bogra, Far idpur, Mymensingh and Rajshahi both 'push-pu l l ' factors and distance were operating but distance was the most important f a c t o r . In others , distance played a secondary rol e. When D..J was regressed against IV. ^ , c e t e r i s parebus, distance was found to be s i g n i f i c a n t l y (negatively) re lated with migration ve loc i ty in a l l the cases except for Rangpur and Noakhali (Figure 5.8). In other words, for these two d i s t r i c t s , the ve loc i ty of migration was independent of distance separating the i th and j th d i s t r i c t s , c e t e r i s parebus. From the factor score pattern of the j th d i s t r i c t s i t was found that the ve loc i ty of migration was as expected only from the top sending d i s t r i c t s in the majority of cases (Table 5H, column 3). As in p r a c t i c e , the top three or four i th d i s t r i c t s send the majority of migrants to each of the j th d i s t r i c t s , i t may be stated that the var iables that were s i g n i f i c a n t l y re lated with IV in the 'C . F . ' Cexplained' 229 a high proportion of the tota l variance of the spat ia l pattern of R-R in-migration streams. V a l i d i t y of Hypothesis 1 Based on the resul ts of analysis discussed in sections 5.9 and 5.10 i t may be said that the Hypothesis 1 was v a l i d p a r t i a l l y in 1961 a l s o . As the ' p u l l ' factors were found to be dominant in both the R-R out-migration streams and R-R in-migration streams, in 1961, the f i r s t part of the Hypothesis was not found to be v a l i d . On the other hand, distance continued to be more important a factor in R-R in-migration streams than in R-R out-migration streams. Thus, the second part of the Hypothesis was v a l i d in 1961 a l so . 5.11 Determinants of the Spatial Pattern of Rural-to-Urban  Out-Migration Streams ( O V ^ ) , 1961 5.11.1 Determinants of the Spatial Pattern of Out- Migration from Bar isal As indicated by Table 5.76, column ' a , ' about 65% of the total out-migrants from Bar isal went to the contiguous 'urban' d i s t r i c t of Khulna. Dacca and Chittagong also received substant ia l numbers of migrants. 230 By examining factor loading of the var iables in the ' C . F . ' of the factor matrix i t may be said that popula-t ion pressure on agr i cu l tu ra l land, RD/SQMCA (.9808), and presence of too many share croppers and landless a g r i c u l t u r a l labourers, SC+LLALF/TALF (.8213), acted as the 'push' f a c t o r s . On the otherhand, the higher level of urbanizat ion, U. . (- .5697), acted as the ' p u l l . ' But the 'push' was domi nant. Since loadiirig on d istance, D.. (- .0436), was not s i g n i f i c a n t , i t suggested that distance did not act as a constra int in determining the d i s t r i c t of des t inat ion . The factor score pattern of the j th d i s t r i c t s suggested that while the ve loc i ty of migration to Dacca and Chittagong H.T. was as. expected from B a r i s a l , in Chittagong and Khulna i t was more in Kushtia less (compare rank orders of columns 1 d ' and 1b ' ). 5.11.2 Determinants of the Spatial Pattern of Out- Migration from Bogra As indicated by Table 5.77, column ' a , ' out of the tota l number of migrants going out from Bogra, Dacca received more than 50%, while others received much l e s s . Loading on 0 V i . in the 1 C . F . 1 was .9070. 231 By examining factor loading c h a r a c t e r i s t i c s in * the ' C . F . 1 of the factor matrix i t may be interpreted that the migrants were ' p u l l e d ' to the j th d i s t r i c t s i f : (a) more land under r i c e was a v a i l a b l e , Rice/TCA (.9563), (b) total cu l t i vated land was more, TCA/NCA (.9013), and (c) i f the ' l eve l of urbanization was higher, U.. . (. 9557). On the other hand, population pressure on a g r i c u l t u r a l land, RD/SQMCA (-.6915), acted as the 'push' f a c t o r . But the ' p u l l ' was domi nant. Since loading on d istance, D.. (.3898), was not s i g n i f i c a n t i t indicated that distance did not act as a constra int in regulat ing the ve loc i ty of out-migration from Bogra. From the factor score pattern of the j th d i s t r i c t s i t may be found that except for Dacca in no other d i s t r i c t s the ve loc i ty of migration was as expected from Bogra (compare rank orders in columns 'b ' and ' d ' ) . 5.11.3 Determinants of the Spatial Pattern of Out- Migration from Comilla Though the bulk of out-migrants went to Dacca (62%) and Chittagong (27%), the d i s t r i b u t i o n of out-migration from Comilla was quite even (Table 5.78, column ' a ' ) . Loading on OV. . in the ' C F . ' was - .9570. 232 By examining factor loading c h a r a c t e r i s t i c of -* the var iables in the ' C . F . ' of the factor matrix, i t may be interpreted that the a v a i l a b i l i t y of more land under j u t e , Jute/NCA (.9585), in a given j t h d i s t r i c t acted as the ' p u l l . ' On the other hand, the presence of too many landless a g r i c u l t u r a l labourers and share croppers, SC+LLALF/ TALF (- .8698), in the i th d i s t r i c t s acted as the 'push . ' but the ' p u l l ' was dominant. Since loading on d istance, D.. (- .0445), was not s i g n i f i c a n t i t indicated that distance did not play an important r o l e . From the factor score c h a r a c t e r i s t i c of the subjects i t was found that the ve loc i ty of migration from Comilla was as expected to a l l the j th d i s t r i c t s (compare rank orders in columns ' d ' and ' b ' ) . 5.11.4 Determinants of the Spat ia l Pattern of Out- Migration from Dinajpur As indicated by Table 5.79, column ' a , 1 the bulk of the tota l out-migrants went to Dacca (55%). Others received far fewer. By examining factor loading c h a r a c t e r i s t i c s of ** the var iables in the ' C . F . ' of the factor matrix i t may *Loading on OV^. in the ' C F . ' was - .8915. •k k Loading on O M . . in the ' C . F . ' was - .9957. 233 be interpreted that people were ' p u l l e d ' to a given j t h d i s t r i c t i f l i t e r a c y , L. . (.9842), was more than in Dinajpur. There was JTO ' push . 1 S i g n i f i c a n t loading on both a c c e s s i b i l i t y , E3 (- .9908), and d istance, D.. (.8785), indicated that people went to that j th d i s t r i c t in greater numbers which was more access ib le and at the same time nearer. From the factor score pattern of the j th d i s t r i c t s i t may be said that the ve loc i ty of migration from Dinajpur was as expected to Dacca and Kushtia only (compare rank orders in columns 'b ' and ' d ' ) . . 5.11.5 Determinants of the Spatial Pattern of Out- Migration from Faridpur The bulk of the migrants went to only one d i s t r i c t , Dacca (66%). Others received far fewer (Table 5.80, column 1 a ' ) . By examining factor loading c h a r a c t e r i s t i c s of the * var iables in the ' C . F . ' of the factor matrix i t may be interpreted that people were ' p u l l e d ' to that j th d i s t r i c t where land under j u t e , Jute/NCA (-.7588), was more than in Faridpur. There was no_ 'push. ' S i g n i f i c a n t loading on both a c c e s s i b i l i t y , (- .8545), and d istance, D. . (.9015), indicated that migrants * Loading on OV.. in the ' C . F . ' was - .9112. 234 from Faridpur went to that j t h d i s t r i c t in greater numbers which was access ib le and also nearer. From the factor score pattern of the subjects i t may be said that except for Chittagong the ve loc i ty of migration from Faridpur was as expected to a l l otherjth d i s t r i c t s (compare rank orders in columns ' d ' and ' b ' ) . 5.11.6 Determinants of the Spatial Pattern of Out- Migration from Jessore As indicated by Table 5.81, column ' a , 1 89% of the tota l out-migrants went to two d i s t r i c t s , Khulna and Khustia, which are contiguous to i t . By examining factor loading c h a r a c t e r i s t i c s of * the var iables in the ' C . F . ' of the factor matrix i t may be interpreted that people were ' p u l l e d ' to that j th d i s t r i c t where land under j u t e , Jute/NCA (.8911), was more than Jessore. At the same time, people were 'pushed' from this d i s t r i c t i f share croppers and landless a g r i c u l t u r a l labourers , SC+LLALF/TALF (- .8403), were more than in a given j t h d i s t r i c t . But the ' p u l l ' was dominant. S i g n i f i c a n t loading on d istance, D.. (.9073), indicated that people tended to go to nearer d i s t r i c t s in order to minimize the distance t r a v e l l e d . *Loading on OV... in the ' C . F . ' was -.8592. 235 The nature and magnitude of factor scores of the subjects indicated that except for Khulna and Kushtia, the ve loc i ty of migration from Jessore was as expected to a l l other j t h d i s t r i c t s (compare rank orders in columns ' d ' and ' b ' ) . 5.11.7 Determinants of the Spatial Pattern of  Out-Migration from Mymensingh As indicated by Table 5.82, column ' a , ' a l l but a few migrants wenttto Dacca (more than 90%). Thus the d i s t r i b u t i o n of out-migration from Mymensingh was heavi ly skewed towards Dacca. By examining factor loading c h a r a c t e r i s t i c s of * the var iables in the '''.'C . F.'' of the factor matrix i t may be interpreted that the presence of too many share croppers and a g r i c u l t u r a l labourers , SC+LLALF/TALF (- .8878), in Mymensingh acted as the 'push' f a c t o r . At the same time, people were ' p u l l e d ' to those j th d i s t r i c t s where land under j u t e , Jute/NCA (.6468), was more than in Mymensingh. However, the 'push' was dominant. S i g n i f i c a n t loadings on a c c e s s i b i l i t y , E(3 (- .9204), and d istance, D. . (.9871 ) , indicated that people tended to go to those j th d i s t r i c t s in greater numbers which were access ib le and also nearer. That possibly 'exp la ins ' why * Loading on OV^. in the ' C . F . ' was -.7684. 236 people went to Dacca in such a big number which was con-tiguous to Mymensingh and was also better connected. From the factor score pattern of the subjects i t was found that the rank of the j t h d i s t r i c t s was a l t e r e d , except for Dacca (compare rank orders in columns ' d ' and ' b ' ) . This further suggested that the system that was recognized in the ' C F . ' was e f fec t i ve only for Dacca. For other j th d i s t r i c t s , maybe, var iables in factor 3 operated more e f f e c t i v e l y . Factor 1 is considered to be s a t i s f a c t o r y for our purpose since i t ' exp la ins ' for more than 90% of out-migrants from Mymensingh. 5.11.8 Determinants of the Spatial Pattern of Out- Migration from Noakhali As indicated by Table55883, column ' a , ' though the majority of the migrants went to Chittagong (42%) and Dacca (39%), substant ia l numbers of migrants did also go to otherjth d i s t r i c t s . In other words, the d i s t r i b u t i o n of our-migration from Noakhali was quite even compared to that of other i th d i s t r i c t s in the R-U out-migration stream. By examining factor loading c h a r a c t e r i s t i c s of the * var iab les in the ' C . F . ' of the factor matrix i t may be interpreted that people were ' p u l l e d ' to those j t h d i s t r i c t s i f land under r i c e , Rice/TCA (.8648), and total cu l t i va ted *Loading on OV^. in the ' C . F . ' was -.9952. 237 land, TCA/NCA (.5294), were more than in Noakhali . On the other hand, the presence of too many landless a g r i c u l t u r a l labourers, LLALF/TALF (-.7131), and population pressure on agr i cu l tu ra l land, RD/SQMCA (-.6520), acted as the 'push' f a c t o r s . But the ' p u l l ' was stronger. S i g n i f i c a n t loading on d istance, D. . (.6731 ), indicated that people tended to go to the nearer j t h d i s t r i c t s in greater numbers to minimize the distance t r a v e l l e d . The factor score pattern of the subjects revealed that the ve loc i ty of migration from Noakhali was as expected to a l l the j th d i s t r i c t s (compare rank orders in columns ' d ' and ' b ' ) . 5.19.9 Determinants of the Spatial Pattern of Out- Migration from Pabna As indicated by Table 5.84, column ' a , ' the largest number of migrants from Pabna went to Dacca c lose ly followed by Kushtia and Khulna. In other words, the spat ia l pattern of d i s t r i b u t i o n of out-migration from Pabna was quite even compared to that of other i th d i s t r i c t s in the R-U out-migration stream. By examining the factor loadings of the var iables * in the ' C . F . ' of the factor matrix i t may be said that Loading on OV. . in the ' C F . ' was -.8438. Table 5.84 R-U OUT-MIGRATION STREAMS FROM PABNA, 1961 Out-Migration from the i t h D i s t r i c t ~-'Pabna II a" " b " " d D i s t r i c t s of Dest inat ion ( j th D i s t r i c t s ) No. of in the Migrants Stream Ve loc i ty of Out-Migration Streams Distance Sepa ra t i ng i and j (in miles) Factor scores on the Subjects from the ' C r i t i c a l Factor ' OMij Rank OVij Rank Dij Fj Rank Chi ttagong 1 ,049 (4) .0089 (4) 200200 0.5834 (4) Chittagong H.T. 85 (5) .0053 (5) 225 1 .3102 (5) Dacca 3,471 (1) .0173 (3) 65 -1 .1501 (1) Khulna 2,200 (3) .0229 (2) 125 0.0426 (3) Kushti a 2,993 (2) .0652 (1) 50 -0.7862 (2) ' C F . ' OVij -0.8438 SC+LLALF/TALF -0.8942 LLALF/TALF -0.7121 Jute/NCA 0.7109 Ui j -0.0821 (not Di j 0.0966 (not s i g n i f i c a n t ) s i g n i f i c a n t ) co oo 239 the presence of too many share croppers and landless a g r i -cu l tura l labourers , SC+LLALF/TALF (- .8942), in Pabna compared to any given j t h d i s t r i c t acted as the 'push 1 f a c t o r . On the other hand, people were ' p u l l e d ' to the j t h d i s t r i c t s i f land under j u t e , Jute/NCA (.7109), was more than in Pabna. But the 'push' was dominant. Low and thus non-s ign i f i cant loading on d istance, D. . (.0966), indicated that distance separating the i th and j th d i s t r i c t s did not act as a constra int in determining the flow of migrants from Pabna to the 'urban' d i s t r i c t s . From the factor score pattern of the subjects i t was found that while in Dacca the ve loc i ty of migration from Pabna was much less than expected, in Khulna and Kushtia i t was more than expected (compare rank orders in columns ' d ' and 'b ' ). 5.11.10 Determinants of the Spatial Pattern of Out- Migration from Rajshahi As indicated from Table 5.85, column ' a , ' an almost equal number of migrants went to Dacca and Kushtia, Others received far fewer. From the factor loading c h a r a c t e r i s t i c s of the * var iables in the ' C . F . ' of the factor matrix i t may be "Loading on OV^. in the ' C . F - ' was - .8189. 240 interpreted that people were 'pushed' from Rajshahi i f the presence of landless agr i cu l tu ra l labourers and share croppers, SC+LLALF/TALF (-.9195), were more in Rajshahi than in a given j t h d i s t r i c t . On the other hand, people were ' p u l l e d ' to those j t h d i s t r i c t s where land under j u t e , Jute/NCA (.7681), was more than in Rajshahi. But the 'push' was domi nant. S i g n i f i c a n t loading on d istance, D.. (.9045), 3 indicated that people tended to go to the nearer d i s t r i c t s in greater numbers to minimize the distance t r a v e l l e d . The factor score pattern of the j th d i s t r i c t s revealed that while in Kushtia the ve loc i ty of migration from Rajshahi was more than expected, in Dacca i t was l e s s . Other received as expected (compare rank orders in columns ' d ' and 'b ' ) . 5.19.11 Determinants of the Spatial Pattern of  Out-Migration from Rangpur As indicated by Table 5.86, column ' a , ' the majority of migrants went to Dacca and Chittagong. Other received far fewer. By examining factor loadings of the var iables in * the ' C . F . ' of the factor matrix i t may be interpreted that *Loading on OV^. in the ' C . F . ' was -.8834. 241 the presence of too many share croppers and landless a g r i c u l -tural labourers , SC+LLALF/TALF (- .8250), in Rangpur than in a given j t h d i s t r i c t 'pushed' people away from th is d i s t r i c t . There was JTO ' p u l l . ' Since loadings on both a c c e s s i b i l i t y , E3 (- .9510), and d istance, D. . (.8687), were s i g n i f i c a n t , i t indicated that people went to those j th d i s t r i c t s in greater numbers which were nearer to i t and also better connected. From the factor score pattern of the j t h d i s t r i c t s i t was found that while in Dacca, Khulna and Chittagong H.T., the ve loc i ty of migration was as expected from Ranpur, in Chittagong i t was more and Kushtia 1 ess»(compare rank orders in columns ' d ' and 1 b ' ) . 5.11.12 Determinants of the Spatial Pattern of  Out-Migration from Sylhet As indicated by Table 5.87, column ' a , ' the largest number of migrants from Sylhet went to Dacca followed by Chittagong and Chittagong H.T. Other received far fewer. By examining factor loadings of the var iables in * the ' C . F . ' of the factor matrix i t may be interpreted that people were 'pushed' from Sylhet i f the presence of landless a g r i c u l t u r a l labourers, LLALF/TALF (-.8775), was more than in a given j th d i s t r i c t . There was, however, n£ ' p u l l . ' *Loading on OV.. in the ' C . F . ' was - .9707. 242 S i g n i f i c a n t loadings on both a c c e s s i b i l i t y index, E3 (- .9342), and distance, D.. (.9758), indicated that people tended to go to those j th d i s t r i c t s in greater numbers which were nearer and better connected. From the factor score pattern of the subjects i t may be said that while the ve loc i ty of migration from Sylhet was more than expected to Chittagong H.T. and Khulna, in others i t was less (compare rank orders in columns ' d ' and 5.11.13 Summary of Section 5.11 Dacca overshadowed a l l other 'urban' d i s t r i c t s in terms of gett ing migrants from rural d i s t r i c t s in the R-U out-migration stream. The largest number of migrants from a l l but three d i s t r i c t s such as B a r i s a l , Jessore and Noakhali , went to Dacca. Thus, the spat ia l d i s t r i b u t i o n of migrants in the R-U out-migration streams was heavi ly skewed towards only one d i s t r i c t - Dacca. From the pr inc ipa l axis type of factor analys is i t was observed that both 'push' and ' p u l l ' factors were operating in determining the spat ia l pattern of out-migration streams. People were 'pushed' from the rural d i s t r i c t s ( i th d i s t r i c t s ) i f the presence of landless a g r i c u l t u r a l labourers and/or share croppers, SC+LLALF/TALF, LLALF/TALF, 243 was more than in a given j t h d i s t r i c t and/or i f population pressure on agr i cu l tu ra l land, RD/SQMCA, was more than in a given i th d i s t r i c t . At the same time, the a v a i l a b i l i t y of more land under j u t e , Jute/NCA, or r i c e , Rice/TCA in the j t h d i s t r i c t s ' p u l l e d ' people away from the i th d i s t r i c t s . There were also occasions when the higher in tens i ty of cropping of a j th d i s t r i c t , TCA/NCA, or the higher level of l i t e r a c y , L..., or the higher level of urbanizat ion, U.^ . , acted as the ' p u l l ' factors (see Table 51, c o l u m n c ) . Thus, these six var iables were found to be most ac t ive ly operating e i ther as a 'push' or ' p u l l ' f a c t o r . However, the p a r t i c u l a r 'mix' of 'push' and ' p u l l ' was d i f f e r e n t . From Table 51, column c , i t may be observed that while the ' p u l l ' factors were dominant in Bogra, Comi l la , Dinajpur, Far idpur, Jessore and Noakhali , the 'push' factors were dominant in the rest of the six d i s t r i c t s . As there were also two occasions such as in Dinajpur and Faridpur, when 'push' factors were not operating at a l l , we may conclude that , on the average, the ' p u l l ' factors were dominant in the R-U out-migration streams, but the dominance was marginally stronger. Distance had a var iab le ro le to play as an i n t e r -vening v a r i a b l e . It was s i g n i f i c a n t in eight out of twelve cases (Table 51, column d) . Thus, on the average, i t played an important r o l e . There were occasions when distance was found to be the most important factor though the 'push-pu l l ' Table 51 DETERMINANTS OF THE SPATIAL PATTERN OF R-U OUT-MIGRATION STREAMS 1 9 6 1 a b c d e OV,. from I Was the spat ia l di s t r i bution skewed towards the nearer di s t r i cts? Which of the factors were dominant -1 push 1 or ' p u l l ' ? Was distance s ign i f i cant? Was of nr.-. j the ve loc i ty migration to "•"•'.Dacca .3 r high est? 1 . Bar isal Yes Push (RD/SQMCA and LLALF/TALF) No No 2. Bogra Yes Pull (Rice/TCA, U., . & TCA/NCA No Yes 3. Comi11 a Yes Pull (Jute/NCA) No Yes 4. Dinajpur No Only pull ( L . j ) * Yes Yes 5. Fai rdpur Yes Only pull (Jute/NCA) * Yes Yes 6. J essore Yes Pull (Jute/NCA) Yes No Both distance and a c c e s s i b i l i t y were s i g n i f i c a n t . CONTINUED £ Table 51 (Continued) a b c d e OV., from Was the spat ia l di s t r i bution skewed towards the nearer di s t r i cts? Which of the factors were dominant -'push' or 'pul l '? Was distance s ign i f i cant? Was the ve loc i ty of migration to Dacca highest? 7. Mymensingh Yes Push (SC+LLALF/TALF) * Yes Yes 8. Noakhali Yes Pull (Rice/TCA) Yes No 9. Pabna Yes Push (SC+LLALF/TALF) Yes No 10. Rajshahi Yes Push (SC+LLALF/TALF) Yes No 11. Rangpur No Push (SC+LLALF/TALF) * Yes Yes 12. Sylhet Yes Push (LLALF/TALF) * Yes Yes O - i o LO r o oH o o co OH o LO CM oH o o CM OH o LO oH o o o L O o 0 040 030 020 010 \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ N >s N . ^* _ N A 246 | ' ' ' ' | i i i i | i i r*i^ -| i i ^ i i i T | i i ~t ^\aKi T ^'-| i i i i I 0 50 100 150 200 250 300 350 400 DISTANCE IN MILES. DIJ Figure 5.9. 'Distance E l a s t i c i t y of R-U Out-Migration Streams , 1961 . 247 factors were also operat ing. This happened in the case of migrants coming out from Faridpur, Jessore, Mymensingh, Rangpur and Sylhet . It may also be noted that a c c e s s i b i l i t y , another intervening v a r i a b l e , was also found to be s i g n i f i -cant in f i v e out of twelve d i s t r i c t s along with distance (see Table 51, column d) . In other words, i t may be inferred that the intervening var iables gained importance during the decade, 1951-1961, p a r t i c u l a r l y in the case of R-U out-migrat ion. When D .^ was regressed against OV.. . , c e t e r i s parebus3 distance was found to be s i g n i f i c a n t l y re lated with migra-t ion ve loc i ty in a l l but one d i s t r i c t - Bogra (Figure 5.9) . This indicated that though the inf luence of distance as an intervening var iable increased in 1961, distance acting as the explanatory v a r i a b l e , c e t e r i s parebus, was s t i l l stronger. From the factor score pattern of the j t h d i s t r i c t s i t was found that the ve loc i ty of migration to Dacca was the highest in seven out of twelve cases (Table 51, column e ) . This confirmed the e a r l i e r notion that Dacca overshadowed a l l ohter 'urban' d i s t r i c t s in terms of receiv ing migrants from the rural d i s t r i c t s . 5.12 Determinants of the Spat ia l Pattern of Urban-to-Rura1  Out-Migration Streams, 1961 5.12.1 Determinants of the Spatial Pattern of Out- Migration from Chittagong 248 As indicated by Table 5.88, column ' a , ' ' t h e highest number of migrants from Chittagong went to Jessore. Comi l la , Noakhali , Bar isal and Sylhet also received substant ia l numbers of migrants. Thus the d i s t r i b u t i o n of out-migrat io-was quite even. By examining factor loadings of the var iables in * the ' C F . 1 of the factor matrix i t may be interpreted that the presence of too many a g r i c u l t u r a l labourers , LLALF/TALF (- .7998), and population pressure on agr i cu l tu ra l land, RD/SQMCA (- .7875), acted as the 'push' f a c t o r s . On the other hand, people were ' p u l l e d 1 to the j th d i s t r i c t s i f more land under r i c e , Rice/TCA (.8182), was ava i lab le than in Chittagong. The ' p u l l ' was marginally stronger. Since loading on distance was s i g n i f i c a n t , D. . (.8781), i t indicated that the migrants tended to go to the nearer d i s t r i c t s in greater number to minimize the distance ** travel 1ed. From the factor score pattern of the subjects i t was revealed that the ve loc i ty of migration from Chittagong was as expected to the majority of the j th d i s t r i c t (compare rank orders in columns ' d ' and ' b ' ) . Loading on 0 V i . in the ' C . F . ' was - .6785. ick Jessore appeared to be an exceptional case. Though i t was at quite a d istance, the largest number of people went there. 249 5.11.2 Determinants of the Spatial Pattern of Out- Migration from Chittagong H.T. As indicated by Table 5.89, column ' a , ' very few people migrated away from Chittagong H.T. Out of the total out-migrants more than 60% went to only one d i s t r i c t , Comi l la . Other received far fewer. By examining factor loadings of the var iables in * the ' C F . ' of the factor matrix i t may be interpreted that the population pressure on agr i cu l tu ra l land, RD/SQMCA (-.8805), and the presence of too many landless a g r i c u l t u r a l labourers, LLALF/TALF (- .6940), acted as the 'push' f a c t o r s . On the otherhand, people were ' p u l l e d ' to those j t h d i s t r i c t s where land under r i c e , Rice/TCA (.6112), was more. However, the 'push' was dominant. S i g n i f i c a n t loading on d istance, D.. (.9015), indicated that people tended to go to the nearer d i s t r i c t s in greater numbers to minimize the distance t r a v e l l e d . The factor score pattern of the subjects revealed that the ve loc i ty of migration from Chittagong H.T. was not expected to the majority of the j t h d i s t r i c t s (compare rank orders in columns ' d ' and ' b ' ) . Loading on OV.. in the ' C . F . ' was - .7881. 250 5.11.3 Determinants of the Spatial Pattern of  Out-Migration from Dacca As indicated from Table 5.90, column ' a , ' the highest number of migrants from Dacca went to Mymensingh, which is contiguous to i t . Far idpur, Rajshahi and Rangpur came next in order. It may be noted here that unlike other d i s t r i c t s of o r i g i n in the U-R out-migration stream, substant ia l numbers of migrants went to a l l the j th d i s t r i c t s from Dacca. By examining factor loadings of the var iables in * the ' C . F . ' of the factor matrix i t may be interpreted that people were ' p u l l e d ' from Dacca i f land under j u t e , Jute/ NCA (.8694), and r i c e land, Rice/TCA (.7616), were more in a given j th d i s t r i c t . There was no 'push . ' S i g n i f i c a n t loading on d istance, D.. (.7558), indicated that the migrants tended to go to the nearer d i s t r i c t s in greater numbers to minimize the distance travel 1ed. From the factor score pattern of the subjects i t was revealed that the ve loc i ty of migration from Dacca was as expected to the majority of the j th d i s t r i c t s (compare rank orders in columns ' d ' and ' b ' ) . *Loading on OV.... in the ' C . F . ' was -.6432 Table 5.90 U R B A N TO R U R A L O U T - M I G R A T I O N S T R E A M S , 1 9 6 1 S P A T I A L P A T T E R N OF O V I J FROM DACCA D i s t r i c t of Origin (ith D i s t r i c t ) ~ Dacca H _ n a "b II "c" "d II D i s t r i c t s of Desti nation ( j th D i s t r i c t s ) No. of in the Migrants Stream Veloci ty Migration of Out-Streams Distance Separati ng i and j (in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor ' OMi j Rank OVij Rank Di j Fj Rank Bar isa l Bogra Comi11 a Dinajpur Fari dpur Jessore Mymensi ngh Noakhali Pabna Rajshahi Rangpur Sylhet 8 , 9 3 6 2 , 1 6 0 11 , 8 97 8 , 6 4 2 2 7 , 4 0 2 1 0 , 0 4 1 4 7 , 8 3 5 1 ,1 65 11 , 5 9 5 1 7 , 7 8 0 1 6 , 5 7 9 1 6 , 1 4 0 ( 9 ) ( 1 1 ) ( 6 ) ( 1 0 ) ( 2 ) ( 8 ) ( 1 ) ( 1 2 ) ( 7 ) ( 3 ) ( 4 ) ( 5 ) . 0 2 9 0 .01 34 . 0 2 6 5 . 0 4 9 6 . 0 8 4 6 . 0 4 4 9 . 0 6 6 8 . 0 0 4 7 . 0 5 8 0 . 0 6 2 0 . 0 4 2 8 . 0 3 6 0 ( 9 ) ( 1 1 ) ( 1 0 ) ( 5 ) ( 1 ) ( 6 ) ( 2 ) ( 1 2 ) ( 4 ) ( 3 ) ( 7 ) ( 8 ) 90 100 45 150 47 84 80 100 65 130 168 n o 1 . 2753 - . 3 4 4 3 - . 1 2 3 1 - . 8 9 9 1 - 1 . 2 8 2 1 . 3570 - 1 . 0 5 0 0 1 .2191 - 1 . 5 5 4 0 - . 3 1 6 2 - . 5 3 4 7 1 . 6 8602 ( 1 1 ) ( 6 ) ( 8 ) ( 4 ) ( 2 ) ( 9 ) ( 3 ) ( 1 0 ) ( 1 ) ( 7 ) ( 5 ) ( 1 2 ) *Factor 2 ( ' C F . ' ) OVij - . 6 4 3 2 Rice/TCA . 7 6 1 6 Jute/NCA . 8 6 9 4 Dij . 7 5 5 8 4 252 5.11.4 Determinants of the Spatial Pattern of Out- Migration from Khulna As indicated by Table 5.91, column 'a,''more than 70% of the total migrants from Khulna went to only one d i s t r i c t , Jessore, which is contiguous to i t . B a r i s a l , Faridpur and Comilla came next in order. Other received far fewer. By examining factor loadings of the var iables in * the ' C . F . ' of the factor matrix i t may be interpreted that there was neither any ' p u l l ' nor any 'push. ' Only access i -b i l i t y and distance were the regulat ing f a c t o r s . The nature of loadings on the index of a c c e s s i b l i t y , £ 3 (.8740), and d istance, D.. (- .7918), indicated that the migrants did not only go to theenearer d i s t r i c t s in greater numbers but also to those which were more^aceessible. The factor score pattern of the subjects indicated that the ve loc i ty of migration from Khulna was as expected to the majority of the j th d i s t r i c t s (compare rank orders in columns ' d ' and 1 b-'). 5.21.5 Determinants of the Spatial Pattern of Out- Migration from Kushtia As indicated by Table 5.92, c o l u m n . ' a , 1 more than 40% of the tota l migrants from Kushtia went to only one *Loading on 0 V i . in the 1G ,F . ' was .9496. 253 d i s t r i c t , Jessore, which is contiguous to i t . Far idpur, Rajshahi , Pabna and Comilla also got substant ia l numbers of migrants. Others received far fewer. By examining factor loadings of the var iab les in * the ' C . F . ' of the factor matrix i t may be said that only distance was s i g n i f i c a n t l y re lated with v e l o c i t y of migration (OV . , ) . There was neither any 'push' nor any ' p u l l . ' The nature of loading on d istance, D.. (.9052), indicated thattthe migrants tended to go to the nearer d i s t r i c t s in greater numbers to minimize the distance t rave led . The factor score pattern of the subjects indicated the ve loc i ty of migration from Kushtia was as expected to the majority of the j t h d i s t r i c t s (compares rank orders in columns 'd 1 and 1 b ') . 5.12.6 Summary of Section 5.12 From the discussions of the spat ia l d i s t r i b u t i o n of U-R out-migration streams in sect ion 5.12, i t was d i s -covered that the majority of the migrants went to the nearer d i s t r i c t s from a l l but one d i s t r i c t of o r i g i n - Chittagong (see Table 5J, column b) . This indicated the strong * Loading on OV. . in the ' C F . ' was -.8666. 254 inf luence of distance on this kind of migration streams. The pr inc ipa l axis type of factor analys is revealed that both 'push' and ' p u l l ' factors were operating in determining the spat ia l pattern of migration streams. If the presence of landless a g r i c u l t u r a l labourers , LLALF/TALF, was more in the i th d i s t r i c t s than in a given j th d i s t r i c t , and/or i f the population pressure on a g r i c u l t u r a l land, RD/SQMCA, was more than in a given j t h d i s t r i c t , people were pushed' from the d i s t r i c t s of o r i g i n . At the same time, people were ' p u l l e d ' to the j t h d i s t r i c t s i f the a v a i l a b i l i t y of land under r i c e , Rice/TCA, and/or j u t e , Jute/NCA, was more than in the i th d i s t r i c t s . Thus these four var iables were found to be most a c t i v e l y operating as e i ther the 'push' or 'pu l l ' , f a c t o r s . But these were also occasions, such as in the case of Khulna and Kushtia, when neither the 'push' nore the ' p u l l ' was operating (Table 5J, column c ) . The p a r t i c u l a r 'mix' of 'push' and ' p u l l ' factors was also d i f f e r e n t . While 'push' factors were dominant in Chittagong H.T. , ' p u l l ' factors were dominant in Chittagong and Dacca. Thus, i t may be said that the ' p u l l ' factors were dominant in determining the overal l spat ia l pattern of U-R out-migration streams in 1961. Distance was a s i g n i f i c a n t factor in determining the spat ia l pattern of U-R out-migration streams coming out from a l l the f i v e d i s t r i c t s of o r i g i n (see Table 5J, Table 5J DETERMINANTS OF THE SPATIAL PATTERN OF U-R OUT-MIGRATION STREAMS, 1961 a b c d e 0 V . . from Was the spat ia l di s t r i buti on skewed toward the nearer d i s t r i c t s ? Which of the factors were dominant - 'push' or 'pul l '? Was distance s ign i f i cant? Was the ve loc i ty of migration as expected to the . Jth di s t r i cts? 1 . Chi ttagong No Pull (Rice/TCA * Yes Yes 2. Chittagong H.T. Yes Push (LLALF/TALF, RD/SQMCA) * Yes No 3. Dacca Yes Only Pull (Jute/NCA, and Rice/TCA) Yes Yes 4. Khulna Yes No ' p u l l ' or ' push1 ** Yes Yes 5. Kushti a Yes No ' pul1 ' or ' push ' ** Yes Yes * Distance was the most important f a c t o r . Distance was the only explanatory f a c t o r . ro cn O H o LO ro OH o o ro OH o LO CVJ OH o o CVJ o'H o LO oH o o OH o LO o 0 040 030 020 010 \ \ \ \ \ > \ * \ N \ V \ w \ \ \ \ \ \ \ \ \ \ ^ 256 -0 i i i i i i i i ' I ' ' ' " I 1 1 1 1 I 1 1 1 1 1 1 1 1 ** I 1 1 ' 1 "*! 1 1 1 1 I 0 50 100 150 200 250 300 350 400 DI5TRNCE IN MILES. O I J Figure 5.10. 'Distance E l a s t i c i t y 1 of U-R Out-Migration Streams, 1961. 257 column d) . It may also be noted that in the cases of Khulna and Kushtia, distance was the only explanatory factor and in the cases of Chittagong and Chittagong H.T. , , though 'push' and ' p u l l ' factors were operat ing, distance was the most important f a c t o r . This indicated that the 'd istance e l a s t i c i t y ' of migration has decreased in 1961. In other words, the importance of distance in regulat ing migration n a s i ncreased. From the factor score pattern of the subjects i t was revealed that the v e l o c i t y of migration was as expected to the majority of the j t h d i s t r i c t s from a l l but one d i s t r i c t - Chittagong H.T. (Table 5J, column e) . V a l i d i t y of Hypothesis 2 As the ' p u l l ' factors were found to be dominant in both R-U out-migration streams and U-R out-migration streams, the f i r s t part of the Hypothesis 2 was found to be not v a l i d . On the other hand, the second part of the Hypothesis is va l id because, in 1961, i t was observed that distance continued to play a more important ro le in U-R out-migration streams than in R-U out-migration streams. 258 5.13 Determinants of the Spatial Pattern of Urban-to- Urban Out-Migration Streams, 1961 5.13.1 Determinants of the Spat ia l Pattern of Out- Migration from Chittagong As indicated by Table 5.93, column ' a , ' about 67% of the tota l out-migrations from Chittagong went to Chittagong H.T. which is contiguous to i t . Dacca and Khulna also received s i g n i f i c a n t numbers of migrants. By examining factor loadings of the var iables in * the ' C . F . ' of the factor matrix i t may be interpreted that people were ' p u l l e d ' to the j th d i s t r i c t s i f the p o s s i b i l i t i e s of growing more crops were there due to the lower in tens i ty of cropping, TCA/NCA (.9419), and land under r i c e , Rice/TCA (-.9031), was more than in Chittagong. The r e l a t i v e l y higher ' leve l of urbanizat ion 1 of any j th d i s t r i c t , U. . (-.8375), also acted as the ' p u l l . ' At the same time, people were 'pushed' from Chittagong i f the presence of share croppers and landless a g r i c u l t u r a l labourers , SC+LLALF/TALF (.6836), was more than in a given j t h d i s t r i c t . But the ' p u l l ' was domi nant. Low and thus non-s ign i f i cant loading on d is tance , D.. (-.2830), indicated that distance did not act as a constra int in regulat ing flow of out-migrations from Chi ttagong. *Loading on OV.. in the ' C . F . ' was .6842. Table 5.93 U R B A N T O U R B A N O U T - M I G R A T I O N S T R E A M S Spat ia l Pattern of OVij from Chittagong, 1961 "a" "b" "c" "d" D i s t r i c t s of Dest inat ion ( j th d i s t r i c t s ) No. of Migrants in the Stream Ve loc i ty of Out-Migration Streams Distance Separating i and j (in miles) Factor scores on the Subjects from the ' C r i t i c a l Factor ' Chittagong H.T. Dacca Khulna Kushtia Mij ~ Rank OVij - Rank Dij Fj* - Rank 22,529 m (67%) U ) 8,061 (2) 2,481 (3) 403 (4) .9808 (1) .0265 (2) .0069 (3) .0032 (4) 35 150 162 225 0.9886 (1) -0.0566 (3) -1.3562 (4) 0.4242 (2) Factor 2 OVij -0.6020 SC+LLALF/TALF -0.9587 LLALF/TALF -0.9670 Dij 0.8593 Ui j 0.5423 'Factor 3 (C . F . ) OVij SC+LLALF/TALF Rice/TCA TCA/NCA Ui j Dij 0.6842 0.6836 -0.9031 -0.9419 -0.8375 -0.28307 (not s i g n i f i c a n t ) cn 260 The factor score pattern of the subjects revealed that the ve loc i ty of migration from Chittagong was not as expected to a l l but Chittagong H.T. (compare rank orders in col urnns 1d• and 1 b 1 ) . 5.13.2 Determinants of the Spat ia l Pattern of Out- Migration from Chittagong H.T. As indicated by Table 5.94, column ' a , ' very few people went put from Chittagong H.T. Most of these migrants went to Chittagong, which is contiguous to i t . Thus the d i s t r i b u t i o n of out-migration was skewed. By examining factor loadings of the var iables in * the ' C F . 1 of the factor matrix i t may be said that people were pul led tocthose j t h d i s t r i c t s i f tota l cu l t i va ted land, TCA/NCA (.9610), was more than in Chittagong H.T. On the other hand, the presence of too many share croppers, SC/TALF in Chittagong H.T. acted as the 'push' f a c t o r . But the 1 pul1 1 was domi nant. S i g n i f i c a n t loading on both d is tance , D.. (.6380), * 3 and a c c e s s i b i l i t y , E3 (- .6292), indicated that the migrants were not only going to the nearer d i s t r i c t s , in greater number, but also to those which were better connected. From the factor score pattern of the subjects i t was revealed that while the v e l o c i t y of migration from * Loading on OV.. in the ' C . F . ' was - .7711. 261 Chittagong and Khulna, in Dacca and Kushtia i t was not (compare rank orders in columns ' d ' and ' b ' ) . 5.13.3 Determinants of the Spat ia l Pattern of Out- Migration from Dacca As indicated by Table 5.95, column ' a , 1 though the highest number of migrants from Dacca went to Khulna, other d i s t r i c t s also received substant ia l numbers of migrants. Thus, the d i s t r i b u t i o n of out-migration was quite even. By examining factor loadings of the var iables in * the ' C F . ' of the rotated factor matrix i t may be i n t e r -preted that only 1 push 1 ' factor was operat ing. There was no ' p u l l . ' The population pressure on a g r i c u l t u r a l land, RD/SQMCA (.9631), acted as the 'push' f a c t o r . Low and thus non-s ign i f i cant loading on d istance, D.. . ( .0022), indicated that distance did not act as an obstacle in determining the flow of migrants from Dacca. The factor score pattern of the subjects revealed that while the ve loc i ty of migration from Dacca was as expected to Kushtia and Chittagong, i t was not as expected to Chittagong H.T. and Khulna (compare rank orders in columns "d ' and 'b'O':. Loading on OV.. in the ' C . F . ' was .9928. 262 5.13.4 Determinants of the Spatial Pattern of Out- Migration from Khulna The number of people going out from Khulna was not very large for i t was pr imar i ly a rec ip ient d i s t r i c t (Table 5.96, column ' a ' ) . Out of the total number of migrants, more than 70% went to only one d i s t r i c t , Dacca, which is not contiguous to i t . By examining factor loadings of the var iables in * the ' C F . ' of the factor matrix i t may be interpreted that the presence of too many share croppers and landless a g r i c u l -tural labourers in Khulna, SC+LLALF/TALF (- .9748), acted as the 'push' f a c t o r . On the other hand people were ' p u l l e d ' to the j th d i s t r i c t s i f land under j u t e , Jute/NCA (-.8367), was more than in Khulna. But the 'push' was stronger. Since loadings on both a c c e s s i b i l i t y index, Eg (-.9041), and d istance, D.. (.9422), were s i g n i f i c a n t , i t indicated that the migrants went to those d i s t r i c t s in greater number which were not only nearer but also more access ib le . The factor score pattern of the subjects indicated that thet ve loc i ty of migration from Khulna was as expected to a l l j t h d i s t r i c t s (compare rank orders in columns ' d ' and ' b ' ) . Loading on OV.. in the ' C . F . ' was - .9897. 263 5.13.5 Determinants of the Spatial Pattern of Out- Migration from Kushtia As indicated by Table 5.97, column ' a , ' most of the migrants from Kushtia went to Dacca and Khulna. Others got far fewer. By examining factor loadings of the var iables in the * ' C F . ' of the factor matrix i t may be interpreted that only 'push' factor was operating and there was JTO ' p u l l . ' The presence of too many share croppers and landless a g r i c u l t u r a l labourers, SC+LLALF/TALF (- .9270), in Khulna compared to a given j th d i s t r i c t 'pushed' people away from th is d i s t r i c t . Since loadings were s i g n i f i c a n t on both the a c c e s s i b i l i t y index, Eg (- .8061), and d istance, D.. (.9436), ' 3 i t indicated that the migrants went to those j th d i s t r i c t s in greater number which were not only nearer but also more access i b le . The factor score pattern of the subjects revealed that while the ve loc i ty of migration from Kushtia was as expected to Chittagong and Chittagong H.T., i t was not in Dacca and Kushtia (compare rank orders in columns ' d ' and ' b ' ) . 5.13.6 Summary of Section 5.13 From the discussion of spat ia l d i s t r i b u t i o n of U-U out-migration streams, i t appeared that the importance of *Loading on 0V,.. in the ' C . F . ' was - .7228. 264 Dacca as a major rec ip ient of migrants decreased in 1961. Unlike 1951, the majority of migrants went to Dacca from only two d i s t r i c t s such as Khulna and Kushtia. When the factor loadings on the ' C . F . ' s were looked at in terms of the 'push-pu l l ' theory, i t was observed that 'push' and ' p u l l ' factors were operat ing. People were 'pushed from the i th d i s t r i c t s i f the presence of share croppers and landless a g r i c u l t u r a l labourers , SC+LLALF/TALF, was more than in a given j t h d i s t r i c t or the population pressure on a g r i -cu l tura l land, RD/SQMCA, was more in the i th d i s t r i c t s (see Table 5K, column c ) . At the same time, people were ' p u l l e d ' to the j t h d i s t r i c t s i f , (a) the total cu l t i va ted land, TCA/NCA, was more, (b) the land under r i c e , Rice/TCA, was more, and (c) the ' leve l of u rban iza t ion , ' U , , , was more (than the i th d i s t r i c t s ) . The p a r t i c u l a r 'mix' of 'push' and 'pu l l ' factors -was, however, d i f f e r e n t . From the analysis i t was discovered that while ' p u l l ' factors were dominant in Chittagong and Chittagong H.T. , 'push' factors were dominant in Dacca, Khulna and Kushtia (Table 5K, column c ) . Thus, i t may be concluded that , in 1961, on the average, 'push' factors were dominant in determining the spat ia l pattern of U-U out-migration streams. Distance was s i g n i f i c a n t in three out of f i v e d i s t r i c t s (Table 5K, column d ) . It may also be noted that a c c e s s i b i l i t y was also s i g n i f i c a n t in these three d i s t r i c t s . Table 5K DETERMINANTS OF THE SPATIAL PATTERN OF U-U OUT-MIGRATION STREAMS, 1961 a b c d e OV. . from Was the spat ia l di s t r i bution skewed toward the nearer d i s t r i c t s ? Which of the factors were dominant - 'push' or 'pul l '.? Was distance s ign i f i cant? Was the ve loc i ty of migration as expected to the j th d i s t r i c t s ? 1. Chi ttagong 2. Chi ttagong D.T. Yes Yes Pull (TCA/NCA, Rice/ TCA & U . . Pull (TCA/NCA) No * Yes ** No ** No 3. Dacca No Push (RD/SQMCA) No ** No 4. Khulna No Push (SC+LLALF/TALF) * Yes Yes 5. Kushtia Yes Push(SC+LLALF/TALF) * Yes ** No A c c e s s i b i l i t y was also s i g n i f i c a n t . The majority of migrants went to only one or two d i s t r i c t s . ro cn -.1 I - \ . \ \ 266 ^ = - - V T — i — i f | — i — i — i — i — | — i T r i — j — r 200 250 300 350 50 100 "1-r"T 150 400 DISTANCE IN MILES. DIJ Figure 5.11. 'Distance E l a s t i c i t y ' of U-U Out-Migration Streams,1961. 267 When D.. was regressed against OV. . , cetevis parebus, i t was observed that except for Dacca, distance was s i g -n i f i c a n t l y negatively re lated with migration v e l o c i t y . The 'distance e l a s t i c i t y ' of migration seems to have also i n -creased in the majority of the cases (Figure 5.11). The factor score pattern of the subjects revealed that the v e l o c i t y of migration was not as expected to the majority of j t h d i s t r i c t s from four out of f i v e i th d i s t r i c t s (Table 5K, column e) . 5.14 Determinants of The Spatial Pattern of Urban-to-Urban  In-Migration Streams, 1961 5.14.1 Determinants of the Spat ia l Pattern of In- Migration to Chittagong As indicated by Table 5.98, column ' a , ' the majority of migrants (about 70%) came to Chittagong from Dacca. By examining factor loadings of the var iables in * the ' C . F . ' of the factor matrix i t may be interpreted that the presence of too many share croppers and a g r i c u l t u r a l labourers , SC+LLALF/TALC (.7514), in a given i th d i s t r i c t acted as the 'push' f a c t o r . On the other hand, people were ' p u l l e d ' to Chittagong i f the ' l eve l of u r b a n i z a t i o n , ' U . . . (- .5100), was greater than in a given i t h d i s t r i c t . But the 'push' was dominant. _ — — Loading on IV . . in the ' C . F . ' was - .6719. 268 The nature of loading on d istance, D.. (.2614), indicated that distance did not act as a constra int in deter-mining the flow of migrants to Chittagong. Factor score pattern of the i th d i s t r i c t s indicated that the ve loc i ty of migration to Chittagong was as expected from a l l the d i s t r i c t s except Kushtia (compare rank orders in columns ' d ' and 1 b 1 ) . 5.14.2 Determinants of the Spatial Pattern of In- Migration to Chittagong H.T. As indicated by Table 5.99, column ' a , ' more than 90% of the migrants came to Chittagong H.T. from Chittagong, which is contiguous to i t . By examining factor loadings of the var iables in the ' C F . 1 of the rotated factor matrix i t may be interpreted that the population pressure on a g r i c u l t u r a l land, RD/SQMCA (.5639), and the presence of too many landless a g r i c u l t u r a l labourers, LLALF/TALF (.4311), in a given i th d i s t r i c t acted as the 'push 1 f a c t o r s . But there was no ' p u l l ' from Chittagong H.T. The nature of loading on d istance, D.. (.8237), con-firmed that i t acted as the most important factor in deter-mining the flow of migrants to Chittagong H.T. Loading on IV . . in the ' C F . ' was - .7311. Table 5.99 U R B A N T O U R B A N I N - M I G R A T I O N S T R E A M S , 1 9 6 1 S P A T I A L P A T T E R N OF I V I J TO CHITTAGONG H . T , D i s t r i c t of Destination (jth D i s t r i c t ) ~ Chittagong H.T. II _ H a "b i " c " "d D i s t r i c t s of Or ig in ( i th D i s t r i c t s ) No. of in the Mi grants Stream Veloci ty Migration o f ' I n -Streams Distance Separati ng i and j (in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor ' IMij Rank IV i j Rank Di j Fj Rank Chi ttagong 22,529 (1) .9965 (1) 35 -1 .2135 (1) Dacca 1 ,397 (2) .0361 (2) 185 -.7311 (2) Khulna 69 (3) .0037 (4) 195 .6133 (4) Kushti a 44 (4) .0049 (3) 260 -.0269 (3) Factor 2 ( 'C .F . ' ) IVij -.7311 Dij .8237 RD/SQMCA .5639 LLALF/TALF .4311 270 The factor score pattern of the i th d i s t r i c t s indicated that the ve loc i ty of migration to Chittagong H.T. was as expected from a l l the d i s t r i c t s of o r i g i n (compare rank orders in columns 'd'aand ' b ' ) . 5.14.3 Determinants of the Spatial Pattern of In- Migration to Dacca As indicated by Table 5.100, column ' a , ' the highest number of migrants came to Dacca from Chittagong. Khulna and Kushtia came next. By examining factor loadings of the var iables in * the ' C . F . ' of the factor matrix i t may be interpreted that only ' p u l l ' factors were operat ing. There was no 'push . ' The higher ' l eve l of u rban iza t ion , ' U . . (- .7613), and the a v a i l a b i l i t y of more land under j u t e , Jute/NCA (-.5113), in Dacca acted as the ' p u l l ' f a c t o r s . The nature of loading on d istance, D.. (.2156), indicated that distance did not act as a constra int in determining the flow of migrants to Dacca. The factor score pattern of the i th d i s t r i c t s i n d i -cated that the ve loc i ty of migration to Dacca was as expected from a l l the d i s t r i c t s of o r ig in (compare rank orders in eolumns ' d ' and ' b ' ) . * Loading on IV . . in the ' C . F . ' was - .7915. 271 5.14.4 Determinants of the Spat ia l Pattern of In- Migration to Khulna As indicated by Table 5.101, column ' a , ' the highest number of people came from Dacca. Chittagong and Kushtia came next. By examining factor loadings of the var iables in the * ' C . F . ' of the rotated factor matrix i t may be interpreted that the p o s s i b i l i t i e s of expansion of cu l t i va ted land due to the low in tens i ty of cropping, (TCA/NCA (.6109), in Khulna than in a given i th d i s t r i c t ' p u l l e d ' migrants to th is d i s t r i c t . On the other hand people were 'pushed' from the i th d i s t r i c t s i f the presence of landless a g r i c u l t u r a l labourers , LLAILF/TALF, was more than in Khulna. The nature of loading on d istance, D.. (.3144), i n d i -cated that the distance was not an important f a c t o r . The factor score pattern of the subjects indicated that while the ve loc i ty of migration to Khulna was not as expected from Chittagong and Dacca, i t was as expected from thetbther~twoddiistr<ictstof o r i g i n (compare rank orders in columns 'b ' and ' d ' ) . * Loading on IV . . in the ' C . F . ' was -.8266. 272 5.14.5 Determinants of the Spatial Pattern of  In-Migration to Kushtia As indicated by Table 5.102, column ' a , ' more than 75% of the tota l migrants came from Dacca. By examining factor loadings of the var iables in * the ' C F . ' of the factor matrix i t may be said that the a v a i l a b i l i t y of more land under r i c e , Rice/TCA (- .4336), in Kushtia ' p u l l e d ' people to th is d i s t r i c t . But there was no 'push. ' The nature of loading on d istance, D.. (.8669), indicated that the migrants tended to come from the nearer d i s t r i c t s in greater number to minimize the distance t r a v e l l e d . The factor score pattern of the i th d i s t r i c t s indicated that the ve loc i ty of migration to Kushtia was as expected from a l l the i th d i s t r i c t (compare rank orders in columns ' d ' and 'b ' ). 5.14.6 Summary of Section 5.14 From the discussion of the spat ia l d i s t r i b u t i o n of migration in the U-U in-migration streams, i t was observed that the majority of migrants did not necessar i ly come from the nearer d i s t r i c t s in most of the j t h d i s t r i c t (Table 5L, column a ) . It indicated that , compared to 1951, in 1961, the spat ia l d i s t r i b u t i o n of migration became more even. Loading on IV . . ihethe ' C . F . ' was -.7814. Table 5L DETERMINANTS OF THE SPATIAL PATTERN OF U-U IN-MIGRATION STREAMS, 1961 a b c d e OV. . from i j Was the spat ia l di s t r i buti on skewed toward the nearer d i s t r i c t s ? Which of the factors were dominant - 'push' or ' p u l l ' ? Was distance s i g n i f i c a n t ? Was the v e l o c i t y of migration as expected from the majority of i th d i s t r i c t s ? 1. Chi ttagong No Push (SC+LLALF) No Yes 2. Chittagong H.T. Yes Push (RD/SQMCA, LLALF/TALF) * Yes Yes 3. Dacca No Only Pull (IKj and Jute/NCA) No Yes 4. Khulna No Pull (TCA?NCA) No Yes 5. Kushtia Yes Only Pull (Rice/TCA) Yes Yes Distance was the most important f a c t o r . ro co O-i o LO o 4 o o LO OH o o OvJ oH o LO oH o o oH o LO o 0 040 030 020 ,010 -0 274 J.. " A - - _ \ -i—i—i—i—|—i—i—i—i—|—i—i—i—i—| i i—i—i—|—i—i—i—i—1—i * T v f*"| i i—i—r*-|—i—i—i—i—| 50 100 150 200 250 300 350 400 0I5TRNCE IN MILES. 01J Figure 5.12. 'Distance E las t i c i ty . ' of U-U In-Migration Streams , 1961 . 275 When the factor loadings of the var iables in the ' C . F . ' s were looked at in terms of the 'push-pu l l ' theory i t was observed that both 'push' and ' p u l l ' factors were operat ing. People were 'pushed' from the i th d i s t r i c t s , (a) i f the population was more, (b) i f the presence of share croppers and/or landless agr i cu l tu ra l labourers , SC+LLALF/TALF, was more than in a given j th d i s t r i c t (Table 5L, column c ) . At the same time, the a v a i l a b i l i t y of more land under r i c e , Rice/TCA, or j i i te , Jute/NCA, acted as the ' p u l l ' factors in the j th d i s t r i c t s . There were also cases when the poss i -b i l i t i e s of further expansion of cu l t i va ted land due to the lower in tens i ty of cropping, TCA/NCA, ' p u l l e d ' migrants to that d i s t r i c t (such as in Khulnal)c. However, the p a r t i c u l a r 'mix' of 'push' and ' p u l l ' factors were d i f f e r e n t . While the 'push' was dominant in Chittagong and Chittagong H.T. , the ' p u l l ' was dominant in Dacca, Khulna and Kushtia (Table 5L, column c ) . Thus, i t may be concluded that t h e t ' p u l l ' was dominant in determining the overal l spat ia l pattern of U-U out-migration streams in 1961 . Distance was s i g n i f i c a n t in two out of f i v e cases. There was, however, one case such as in Chittagong H.T. , where distance was the most important f a c t o r . This indicated that though in 1961, the inf luence of distance on th is kind 276 i of migration streams decreased, the inf luence was s t i l l more compared to U-U out-migration streams. When . was regressed against IV^., c e t e r i s paribus i t was found that the ve loc i ty of migration was s i g n i f i c a n t l y negatively re lated with distance in a l l the cases. However, the 'distance e l a s t i c i t y of migrat ion' seems to have increased (Figure 5.12). The factor score pattern of the s u b j e c t - d i s t r i c t s revealed that the ve loc i ty of migration from the majority of the i t h d i s t r i c t s was as expected to a l l the i th d i s t r i c t s (Table 5L, column e) . V a l i d i t y of Hypothesis S For the f i r s t time Hypothesis 3 was found to be t o t a l l y v a l i d . As i t was observed from the discussions of the resu l ts of analysis in sections 5.23 and 5.25 that 'push' factors were dominant in U-U out-Migration streams and ' p u l l ' in the U-U in-migration streams the 1st part of the Hypothesis was proved to be v a l i d . On the otherhand, the second part of the Hypothesis was also proved to be v a l i d since distance played a more important role in U-U in-migration streams than in U-U out-migration streams. 277 5.15 Summary Results of A l l Migration Streams in 1961 and the  V a l i d i t y of Hypothesis 4 From the above discussions of factor analys is of a l l the six kinds of migration streams i t may be said that though both 'push' and ' p u l l ' factors were operating in determining the spat ia l pattern of migration streams in Bangladesh in 1961, ' p u l l ' factors were stronger than 'push' factors in a majority of the cases. As th is resu l t is contrary to our expectations i t disproves the v a l i d i t y of Hypothesis 4 in 1961 a l so . 5.16 Conclusions from the Results of the Analysis Based on the results of the analysis for 1951 and 1961 i t may be conculluded that the p r inc ip les of ' push-pu l l ' theory as such were va l id in Bangladesh but responses of the 'push' or ' p u l l ' factors to the d i f f e r e n t streams were d i f fe rent than what is normally expected. There was a dominance of ' p u l l ' factors in the total system descr ibing the migration pattern. As a by-product of the analysis i t was also d i s -covered that in the low income subsistance economy of Bangladesh, the migrants, who were mostly landless a g r i c u l t u r a l labourers, behaved as ' ra t iona l economic men' and went to the nearer d i s t r i c t s in greater numbers to minimize the distance travel 1ed. Chapter 6 SUMMARY AND CONCLUSIONS 6.0 Summary The present study reveals that the i n t e r - d i s t r i c t migration within Bangladesh, both in 1951 and 1961, was e s s e n t i a l l y a loca l phenomenon and as such the spatial d i s t r i -bution of migration was heavi ly skewed towards the neigh-' bouring d i s t r i c t s . Through the appl icat ion of p r inc ipa l axis type factor analys is i t was found that the 'push-pu l l ' theory was v a l i d , both in 1951 and 1961, in ' exp la in ing ' the spat ia l pattern of i n t e r - d i s t r i c t migration in that reg ion , but the p a r t i c u l a r 'mix' of 'push' and ' p u l l ' factors was d i f f e r e n t from what is normally found in the l i t e r a t u r e descr ibing 'push-pu l l ' empirical explanation of migration streams. From the analysis i t was found that out of a l l the explanatory var iables included in the factor a n l a y s i s , Rural Density per Square Mile of Cult ivated Area (RD/SQMCA), Land-less Agr i cu l tu ra l Labour Force per Total Agr i cu l tura l Labour Force (LLALF/TALF), Land under Rice per Total Cu l t ivate Area (Tice/TCA), Total Cult ivated Area per Net Cult ivated Area 278 279 (TCA/NCA), Land under Jute per Net Cult ivated Area, Jute/NCA, and Distance (in miles) separating the origin ( i ) and desination (j) of migration (D. , ) were most strongly re lated with e i ther Ve loc i ty of In-Migration Streams ( I V . . ) or 1 3 Veloc i ty of Out-Migration Streams (OV..) most f requent ly . Other var iables such as L i teracy ( L . . ) , Urbanization (U. . ) * 3 '3 and A c c e s s i b i l i t y (Ef3) were found to be re lated with migration only occas iona l l y . The nature of loadings on the former var iables indicated that while the population pressure on agr i cu l tu ra l land (expressed by the r a t i o RD/SQMCA) and the presence of too many landless a g r i c u l t u r a l labourers and share croppers (expressed by the rat io LLALF/TALF and SC+LLALF/TALF) acted as the 'push' f a c t o r s ; a v a i l a b i l i t y of more land under r i c e and jute (expressed by the ra t io Rice/TCA and Jute/NCA) and p o s s i b i l i t i e s of expansion of cu l t i va ted land due to the low in tens i ty of cropping (expressed by the ra t io TCA/NCA) acted as the ' p u l l ' f a c t o r s . The nature of loadings on these var iables also suggested that probably most of the migrants in Bangladesh were the landless a g r i -cu l tura l labourers and they l e f t (pushed from) those d i s t r i c t s where the population pressure on agr i cu l tura l il=and was more and the i r presence was also more and went (pulled) to those d i s t r i c t s where more land under r i c e was ava i lab le and there was more c u l t i v a b l e land. As the importance of ' p u l l ' factors increased in 1961, i t indicated that , perhaps, the 280 dependency on r i c e also increased and more land was brought under double cropping or mult ip le cropping through increased i r r i g a t i o n f a c i l i t i e s during the decade 1951-1961. It was hypothesised, a p r i o r i , that while the 'push' factors would have more e f fect on R-R out-migration streams, R-U out-migration streams and U-U out-migration streams, the ' p u l l ' factors w i l l be more important in other streams such as R-R in-migration streams, U-R out-migration streams and U-U in-migration streams. In the hypotheses i t was also incorporated that the impact of distance w i l l be more in R-R i n , U-R out, and U-U in=migration streams than in R-R out, R-U out and U-U out-migration streams. But from the factor analys is i t was found that , in 1951, while ' p u l l ' factors were dominant in R-R out, R-R in and U-U out-migration streams; 'push' factors were dominant in the other three streams. Since ' p u l l ' factors were dominant a majority of the cases ( d i s t r i c t s ) in both R-R in and R-R out-migration streams i t was concluded that the inf luence of ' p u l l ' was more than 'push' in determining the tota l migration system in Bangladesh in 1951. In 1961, the inf luence of ' p u l l ' factors increased. It was dominant in a l l but U-U in-migrat ion streams. However, from the a n a l y s i s , the impact of the intervening var iab le distance on d i f f e r e n t streams was found to be as expected both in 1951 and 1961. Thus, the hypotheses 281 1, 2 and 3 w e r e found to b e v a l i d only p a r t i a l l y . In t h e hypothesis 4 i t was l a i d down that since Bangladesh i s an overpopulated country the pressure on agr i cu l tu ra l land would be intense and as such 'push' factors would be more important than ' p u l l ' f a c t o r s . But from the r e s u l t s , as stated e a r l i e r , i t was found that ' p u l l ' factors were dominant in determining the overa l l i n t e r - d i s t r i c t migration both in 1951 and 1961. Thus the v a l i d i t y of the hypothesis 4 was not estab l i shed . When the factor scores of the subjects ( d i s t r i c t s ) were examined by comparing i t s ( factor scores) rank order with the rank order of the d i s t r i c t s in terms of the actual ve loc i ty of in or out-migration streams (OV. . or I V . . ) ( i . e . comparison between the rank orders in columns 'b ' and ' d ' of each t a b l e ) , i t was discovered that there was not much dev ia t ion . This confirmed that the r e l a t i o n of the explanatory var iables with migration (OV^. or I V ^ - O J w a s not spurious and the ' leve l of explanations' by the four or f i v e var iables (which were i d e n t i f i e d e ither as 'push' or ' p u l l ' factors or as intervening var iab le distance) were quite high. Perhaps the most important feature of the analys is was that though factor analysis was performed as many as 99 times, the resu l ts were consistent enough to get a c lear ins ight into the mechanics of the migration process in a l l the streams. 282 6.1 Conclusions The two basic aims of th is thesis have been, (1) to analyze the spat ia l pattern of migration within Bangladesh for the years 1951 and 1961, and (2) to discover the extent to which the 'push-pu l l ' theory, or at least the p r i n c i p l e s involved in that theory, is appl icable to that region through a push-pull empirical explanation of migration streams by employing pr inc ipa l axis type factor a n a l y s i s . Several theories have been developed, to date, to explain the migra-t ion process but-only 'push-pu l l ' theory appears to be most complete and concerned with the migration process per se and does not suffer from most of the l im i ta t ions from which others su f fe r . However, the 'push-pu l l ' theory has also been developed in the context of the developed world and hardly any work has been undertaken to prove the v a l i d i t y of th is theory in the developing world. The present study has been undertaken on the r e a l i z a t i o n that the factors which lead to migration in two e n t i r e l y d i f f e r e n t socio-economico-cu l tura l sett ings are hardly the same and there isaanneed to test the v a l i d i t y of Ipush-pul l ' theory in c ross -cu l tura l appl i cat i ons. Apparently both the object ives have been met. The analys is shows that Bangladesh migration is e s s e n t i a l l y a loca l phenomenon and the bulk of the migration is within 283 rural d i s t r i c t s . From the factor analys is i t is also proved that though the p r i n c i p l e s of the 'push-pu l l ' theory are appl icable in th is region the p a r t i c u l a r 'mix' of 'push' and ' p u l l ' factors are d i f fe rent from what is normally found in the developed world. As a by product, i t was also discovered that inaa rural subsistence economy l i k e Bangladesh, the migrants behaved l i k e "rat ional economic men" and migrated to the nearest d i s t r i c t s to minimize the distance t r a v e l l e d . This study confirms that 'push-pu l l ' formulation is b a s i c a l l y strong and f l e x i b l e enough for c ross -cu l tura l app l i cat ion but more studies should be undertaken to determine the p a r t i c u l a r 'mix' of 'push' and ' p u l l ' factors in tjheedeyeiliopri ng world context. However, there are some l im i ta t ions of th is study. F i r s t l y the pol i t ico-economic s i tuat ion of the study area has changed much from what i t was i n M Q G l . Unti l 1961, Dacca c i t y , was a sleepy prov inc ia l town of 500,000 people. Suddenly ini l971 i t was turned into the capi ta l of the new nation - Bangladesh. Natural ly i t w i l l be more act ive than before and may a t t rac t more people. As we have found in our analys is that the ' p u l l ' f rom th.e 'urban' areas were very weak, the f indings of our study should be confined to the period 1951 -1961 only. Secondly, th i s migration study is on a d i s t r i c t l e v e l . So nothing could be learned about the migrants who 284 changed the i r locat ions within the d i s t r i c t s . Perhaps more study is necessary in micro- level to confirm the resu l ts of th is study. However, the importance of this thesis cannot be underrated as the number of i n t e r - d i s t r i c t migrants is probably greater than the number of i n t r a - d i s t r i c t migrants and the factors of migration are also probably the same. In addit ion i t is expected that th is w i l l provide a spr ing-board from which more studies can be i n i t i a t e d . 6.2 Imp!i cations This study has important impl icat ions on two l e v e l s . F i r s t l y the f indings that the 'push-pu l l ' theory is also appl icable with marginal modif icat ion in determining the spat ia l pattern of migration streams in the developing world context w i l l serve as a contr ibut ion to the f i e l d of demography. Perhaps the more important contr ibut ion is that the study has been conducted with a l l the possible types of independent streams and i t is proved that the 'push-pu l l ' theory which has been developed based on one or two migration streams and thus considered incomplete, is also va l id when the whole migration system is considered. Secondly, i t is expected to have important i m p l i c a -t ions for population planning of Bangladesh. Like a l l other South Asian countries (except Nepal) the author i t ies and planners in Bangladesh are very worried by the alarming rate 285 of cityward migration at the present time. Unt i l 1961 the urban population increased at a manageable r a t e . The popula-t ion of Dacca c i t y , the biggest urban area in that reg ion, increased by 65.72% during the decade 1951-1961. But the provis ional estimate from the census of February and March, 1974, shows that the population of Dacca, now cap i ta l of * Bangladesh, increased by 262% in 13 years . Does th is s ign i fy that the ' p u l l ' factors in the rural areas that was found to be dominant in determining the d i rec t ion of migration in 1951 and in 1961 are no longer a t t rac t ing more people than the c i t i e s could at t ract? In other words, are the poss i -b i l i t i e s for further expansion of i r r i g a t i o n f a c i l i t i e s e tc . by which more land could be brought under double or mult ip le cropping exhausted? It is apparent that people are coming to the c i t i e s only in desperation since the l i v i n g condit ions are no better than before i f not worse. Thus to stem the ever- increasing number of migrants towards c i t i e s and to have a balanced d i s t r i b u t i o n bfi population growth, the only way l e f t perhaps is to a t t rac t people to the rural areas again. In th i s regard the present study may be u s e f u l . This study shows that people w i l l d e f i n i t e l y migrate to the rural areas rather than to the c i t i e s provided Bangladesh, a f o r t n i g h t l y news b u l l e t i n issued from the Embassey of Bangladesh, Washington, D.C. , June 10, 1974. 286 the opportunit ies for growing crops and s e t t l i n g down there are adequate. It appears that the migrants, the majority of whom are landless agr i cu l tu ra l labourers, w i l l be happy to take agr icu l ture as the i r profession where they are s k i l l e d rather than going to the c i t i e s where they w i l l be treated as unsk i l led labourers. BIBLIOGRAPHY I LITERATURE CITED Afzal , M. 1967. "Migration to Urban Areas in Pak is tan ," Contributed Papers, IUSSP, Sydney Conference, A u s t r a l i a , pp. 684-695. Alonso, Wi l l iam. 1969. "Urban and Regional Imbalances in Economic Development," Eco. Dev. and Cult. Change, 1969, pp. 1-12. Barkin, David. 1972. "The Demographic Impact of Regional Development: A Mexican Case Study," Growth and Change, V o l . 3, No. 4, pp. 15-22. Becker, Gary. 1962. "Investment in Human C a p i t a l : A Theoret ical A n a l y s i s , " J.P.E., Vo l . 70, pp. 9-49. Beals, R .E . , M.B. Levy and L.N. Moses. 1967. "Rat iona l i ty and Migration in Ghana," Review of Economics and S t a t i s t i c s , Vo l . 49, pp. 480-86. Bhargava, Jagadish, P. 1964. "Urbanization Migration and Housing: A Case Study of Ind ia . " Unpublished M.A. Thes is , U .B .C . , Canada. Bose, A. 1967. Patterns of Population Change in India, 1951-61;~Bombay. Bogue, Donald J . 1959. "Internal Migrat ion ," The Study of Population, Hauser and Duncan, Ed i tor , Univers i ty of Chicago Press, pp. 606-7. 1969. P r i n c i p l e s of Demography, John Wiley M _ . . \l I and Sons, New York. 287 288 Bogue, Donald J . and K;C. Zachariah. 1962. "Urbanization and Migration in Ind ia , " India's Urban Future, Roy Turner, Editor (University of C a l i f o r n i a Press) , Berkeley. Bogue, Donald J . and H.S. Shryock et al. 1957. "Subregional Migration in the United States , " 1935-40, Scripps Foundation Studies in Population D i s t r i b u t i o n , No. 5, 1957, pp. 48-49. Bose, Ashish. 1970. Urbanization in India: An Inventory of Source Materials3 Academic Books, Bombay. Borowski, S. 1967. "Factors of Labour Mobi l i ty in an Advanced Technological Soc iety , " Contributed Papers3 IUSSP, Sydney Conference, pp. 1035-46. Brown, L.A. and E.G. Moore, 1967. "Di f fus ion Research in Geography: A Perspect ive ," Progress in Geography-13 pp. 119-57. Brown, L.A. and E.G. Moore. 1969. "Di f fus ion Research in Geography," Progress in Geography-13 1969, pp. 119-1 57. Brush, John E. 1970. "Some Dimensions of Urban Population Pressure in Ind ia , " in Geography and Crowding World3 Wilbur Zel insky et al. ( e d s . ) , Oxford Univers i ty Press, New York, 1970. Bulsara, Jal F. 1964. Problems of Rapid Urbanization in India, Popular Prakshan, Bombay, Indian Cho, George. 1970. "Chinese Residential Patterns, Vancouver, B.C." Unpublished M.A. Thes is , U.B.C. Dasgupta, B ip lab. 1974. "Labour U t i l i z a t i o n in Poor Country A g r i c u l t u r e , " I n s t i t u t e of Development Studies B u l l e t i n 3 V o l . 5, No. 4, pp. 8-12. Davis, Kingsly . 1951 The Population of India and Pakistan. Princeton Press. 289 Desai , L.P. 1 964. The Pattern of Migration and Occupation in a South Gujrat V i l l a g e . Poona, Deccan Col lege. F i rey , N. 1947. Landuse in Central Boston, Harvard Univers i ty Press, Cambridge, Mass. Gallaway, L .E .R .F . G i lber t and P.E. Smith. 1967 and 1968. "The Economics of Labour Mob i l i t y : An Empirical A n a l y s i s , " Western Economic Journal, 5 (1967), 211-223 and "Reply" 6, 1968. Gosal , G.S. 1961. "Internal Migration in India - A Regional A n a l y s i s , " Indian Geographical Journal, pp. 106-21. Greenwood, Michael J . 1968. "The Economics of Labour M o b i l i t y : An Empirical Ana lys i s : Comment," Western Eco. Journal, Vol . 6, pp. 243-44. . 1 969 a. "The Determinents of Labour Migrants in Egypt," Journal of:Regional Science, V o l . 9, pp. 283-90. : —. 1 969 b. "An Analysis of the Deter-minants of Internal Labour Mobi l i ty in the United States , " Review of Economics and S t a t i s t i c s , 51 , pp. 189-94. 1970. "Lagged Response in the Decision to Migrate," Journal of Regional Science, 10 (3) , pp. 375-84. 1971. "A Regression Analysis of Migration to Ujrban Areas of Less-Developed Country: The Case of Ind ia , " J. of Regional Science, Vo l . 11 , No. 2, pp. 253-62 . Gunat i l i ke , Godfrey. 1972/73. "Development and the Rural-Urban Balance: The Experience in Ceylon," Asia, Winter, No. 28, pp. 10-60. Hashmi, Sultan S. 1965. The People of Karachi: Demographic C h a r a c t e r i s t i c s , PIDE. 290 Hauser, P h i l i p M. 1957. Urbanization in Asia and the Far East, Unesco Pub l i ca t ion , Ca l cut ta , India . . 1 964. "The Population of the World: Recent Trends and Prospects ," Population: The V i t a l Revolution, Ronald Freedman, Editor (Anchor Books, New York), pp. 15-29. Isaac, J . 1947. Economics of Migration. Kegan Paul , London. Khan, A.M. 1972. The Economy of Bangladesh, McMillan Press, London, U.K. K r o t k i , K .J . 1964. "The F e a s i b i l i t y of an E f fec t i ve Population Pol icy in Pak is tan ," PDR, pp. 283-311. Kro tk i , K .J . and N^Alimed. 1 964. "V i ta l Rates in East and West Pakistan Tentat ive Results from P6E Exper i -ment," PDR, pp. 377-406. Kumar, J . 1967. "The Pattern of Internal Migration in India during 1951-61," Contributed Papers, IUSSP, Sydney Conference, 1967, pp. 623-34. Labor, Gene and Richard X. Chase. 1972. "Interprov ine ia l Migration in Canada as a Human Capita l Dec i s ion , " J. of Pol. Eco., pp. 795-804. L a l l , A. 1961. "Patterns of In-Migration in Indian C i t i e s , " Geographical Review of India, pp. 16-23. Lee, E.S. 1966. "A Theory of Migrat ion ," Demography, Vo l . 3, pp. 47-57. Levy, Mildred B. et al. 1972. "Lifet ime Versus One Year Migration in Venezuela," Journal of Regional Science, Vol . 12, No. 3, pp. 407-1 5. Logan, M.I. 1970. "On Measuring Geographic M o b i l i t y , " Journal of American S t a t i s t i c a l Association, Vol . 65, pp. 1195-1204. 291 Lowry, I .S . 1966. Migration and Metropolitan Growth: Two A n a l y t i c a l Models, Chandler Publishing Company, San Francisco. Lovgren, E. 1956. "The Geographical Mobi l i ty of Labour,^ Geografiska Annaler, Vo l . 38, No. 4, pp. 344-94. Mcentire, D. 1960. Residence and Race, Univers i ty of C a l i f o r n i a Press, Berkeley. M i t ra , A. 1967. "Internal Migration and Urbanization in Ind ia , " Contributed Papers, IUSSP, Sydney Conference, 1967, pp. 607-614. Morri l l , Richard L. 1 965. "The Negro Ghetto: Problems and A l t e r n a t i v e s , " Geographical Revi-ew, V o l . 55, pp. 339-61. Morrison, Barri e . . . 19.73 . "Regions of High Agr i cu l tu ra l Product iv i ty and t h e i r Socio-Economic Basis in Bangladesh," West Bengal and Bangladesh: Perspec-t i v e s from 1972, Thomas E. & Lavan G . , Ed i tors . Asian Studies Center, Michigan State Un ivers i ty , South Asia Ser ies , Occasional Paper no. 21, pp. 320-35. Murdie, R. 1967. F a c t o r i a l Ecology of Metropolitan Toronto, 1951-61, Chicago, Dept. of Geography and Research Paper no. 116, pp. 17-23. Narain, Vatsa la . 1967. "Migrants in the Metropolitan Areas of Ind ia , " Contributed Papers, IUSSP, pp. 729-36. Nelson, P. 1959. "Migrat ion, Real Income and Information," Journal of Regional Science, Vo l . 1, pp. 43-47. Obaidul lah, M. 1967. "Internal Migration in East Pak is tan ," Oriental Geographer, A p r i l , pp. 10-21. Raimon, Robert L. 1962. "Interstate Migration and Wage Theory," The Review of Eco. and S t a t i s t i c s , 54, pp. 428-38. 292 Rashid, Harouner. 1965. East Pakistan: A Systematic Regional Geography and i t s Development Planning Aspects3 Sh. Ghwam A l i and Sons, Lahore. Ravenstein, E.C. 1855 and 1889. "The Laws of M igrat ion , " Journal of the Royal S t a t i s t i c a l Scoeity, 48, 167-227 and 52 (1889), 241-304. Revel le , Roger. 1973. "Possible Futures for Bangladesh," Asia, Vo l . 29, pp. 34-54. Robinson, H.W. 1938. "Studies in Mobi l i ty of Labour: A Tentative S t a t i s t i c a l Measure," Oxford Economic Papers, pp. 39-62. Sahota, Gian S. 1968. "An Economic Analysis of Cultura l Migration in B r a z i l , " Journal of P o l i t i c a l Economy, V o l . 76, No. 2, pp. 218-46. Sanders, Thomas G. 1971. Internal Migration B r a z i l , Colombia, Chile, Peru. American Un ivers i t i es F ie ld Staf f Reports, West Coast South America Ser ies , V o l . XIX, No . 1 . \ \ Schul tz , T.W. 1962. "Ref lect ions on Investment in Man," Journal of P o l i t i c a l Economy, LXX, No. 5, Part , pp. 1-9. Schwind, Paul J . 1972. Migrationand Regional Development in the United States, 1950-60, Univers i ty of Chicago, Dept. of Geography. Sjaastad, L.A. 1962. "The Costs and Returns of Human Migrat ion , " Journal of P o l i t i c a l Economy, LXX, No. 5. Speare, Alden, J r . 1971 "A cost Benefit Model of Rural to Urban Migration in Taiwan," J. of Pol. Eco., pp. 117-30. Sovani, N.V. 1966. Urbanizationcand Urban India, Asia Pub-l i s h i n g House, Bombay, India. 293 Stouf fer , S.A. 1940. "Intervening Opportunit ies: A Theory Relating Mobi l i ty and Distance," American Soc. Review, 5, 845-67. Taneja, Kusum Lata. 1971. Morphology of Indian C i t i e s , S i l v e r Jubi lee P u b l i c a t i o n , Varanasi , India. Thomas, Dorothy S. 1938. Research Memorandum on Migration D i f f e r e n t i a l s , Social Science Research Council B u l l e t i n , No. 43, New York, pp. 420-21. Thomas, Warren S. 1965. Population Problems. McGraw-Hill Book Company, New York, 593 pp. Thomas, Dorothy, Simon Kuznets et al. 1957, 1960 and 1964. "Population Redistr ibut ion and Economic Growth in the United States , 3 Vo ls . (Ph i lade lph ia , The American Phi losophical Scoei ty , 1957, 1960, 1964). Thomas, D. 1941. Social-Economic Aspects of Swedish Popula-tion Movement, 1750-1933, Stockholm Economic Studies, No. 9, New York, McMil lan. Todaro, M.P. 1969. "A Model of Labour Migration and Urban Unemployment in Less Developed Countr ies ," American Eco. Review, 59, 138-48. T r o t t , Charles E. 1972. " D i f f e r e n t i a l Responses in the Decision to Migrate," Regional Science Association Papers, Vo l . 28, pp. 203-22. Vamathevan, S. 1961. Internal Migrattion in Ceylon, 1946-54, Department of Census and S t a t i s t i c s Monograph No. 13, Colombo. White, Peter T. 1973. "Ca lcut ta , " National Geographic Magazine, Vo l . 143, A p r i l , 1973, pp. 534-63. Wilson, M.G.A. 1969. Population Geography. Thomas Nelson L t d . , Sydney, A u s t r a l i a . 294 Zachariah, K.C. 1964. A H i s t o r i c a l Study of Internal Migration in the Indian Sub-Continent, 1901-1931, Asia Publishing Co . , London. Z i p f , George K. 1946. "The P1.P2/D Hypothesis: On I n t e r c i t y Movement of Persons," American Sociological Review, V o l . 1 1 , pp. 677-686. BIBLIOGRAPHY II LITERATURE NOT CITED BUT FOUND USEFUL Abu-Lughod, Janet. 1961. "Migrant Adjustment ot C i ty L i f e : The Egyptian Case," AJS, Vo l . 67, pp. 22-32. . 1 964. "Urban-Rural Difference as a Func-t ion of the Demographic T r a n s i t i o n ; Egyptican Data and an Ana ly t i ca l Model," American Journal of 'Sociolegys-Vol . 69, No. 5, pp. 476-90. Adams, J . S . 1969. "Direct ional Bias in Intra-Urban Migra-t i o n , " Economio Geography, Vo l . 46, pp. 302-23. Ahmed, N. 1958. An Economic Geography of East Pakistan, Oxford. Ahmed, N. and K. K r o t k i . 1963. "Simultaneous Estimates of Population Growth: The Pakistan Experience," PDR, pp. 37-65. A lan, S . I . 1968. "Age at Marriage in Pakistan," PDR, pp. 489-506. Alonso, Wi l l iam. 1971. The System of Inter Metropolitan Population Flows, Ins t i tu te of Urban and Regional Development (mimeographical), Univers i ty of C a l i f o r n i a . Alverson, Hoyt, S. 1967. "Time Series Analysis of Migratory S t a b i l i z a t i o n : A Research Technique for Qual i fy ing Individual and Group Pattern of Cyc l i c Migration with Special Reference to Sub-Sahara A f r i c a , " African Studies (Johannesburg), Vo l . 26, No. 3, pp. 139-44. 295 296 Anderson, Isabel B. 1966. Internal Migration in Canada, 1921-61, Staf f Study No. 13, Economic Council of Canada, Ottawa, Queens' P r i n t e r . Anderson, T.R. 1955. "Intermetropolitan Migrat ion: A Comparison of the Hypotheses of Z ipf and Staufer , " Amer. Soc. Rev., Vo l . 20, pp. 287-91. Anderson, Theodre A. 1956. "Interpo l i tan Migrat ion; A Corre lat ion A n a l y s i s , " AJS, pp. 459-62. Aslam, M. et al... 1967. "Abridged L i fe Tables of Pakistan and Provinces by Sex, 1962," PDR, pp. 66-106. Bachi, R. "Effects of Migration on the Geographical D i s t r ibut ion of the Population of I s r a e l , " Contributed Papers, IUSSP. Sydney Conference, A u s t r a l i a , pp. 737-51. Barclay, George W. 1966. Techniques of Population Analysis, John Wiley and Sons, I n c . , New York, Seventh Pr int ing Beale, C .L . 1964. "Rural Depopulation in the United States: Some Demographic Consequences of Agr i cu l tu ra l Adjustments," Demography, Vo l . I , pp. 264-72. B la lock, H.M. 1963. "Correlated Independent Var iab les : The Problem of M u l t i c o l 1 i n e a r i t y , " Social Forces, 42, pp. 233-37. Bose, S.R. 1968. "Trend of Real Income of the Rural Poor in East Pakistan, 1949-66," PDR, pp. 452-88. . 1963. "Labour Force and Employment in Pakistan, 1961-81 ," PDR, pp. 371-398. Bone, Robert M. 1960. "ffike E air 1 y 'S'ijb eriva nf iM/irgjr a tA-o.n," Journal of Geography, pp. 112-17. Bose, A. 1967. "Migration Streams in Ind ia , " Contributed Papers, IUSSP, Sydney Conference, A u s t r a l i a , pp. 597-606. 297 Bowles, G. et al. 1966. "Net Migration of the Populat ion, 1 950-60, by Age, Sex and Color," 2, V o l s . , Economic Research Service, U.S. Department of A g r i c u l t u r e . Br ight , M.L. and M.S. Thomas. 1941. "Interstate Migration and Intervening Opportuni t ies ," American Sociological Review, Vo l . VI, No. 6, pp. 773-83. Brown, L .A . , J . Odland and R.G. Col ledge. 1970. "Migration Functional Distance and the Urban Hierarchy," Economic Geography, Vo l . 46, No. 3, pp. 472-86. Brown, M. L . , gF,iE£. Horfcon , and R.I. Wi t t i ck . 1 970. "On Place U t i l i t y and the Normative A l l o c a t i o n of Int ra-urban Migrants," Demography, Vo l . 7, pp. 175-83. Brown, L.A. and D.B. Longbrake. 1970. "Migration Flows in Intra-Urban Space: Place U t i l i t y Cons iderat ion," A.A.A.G., V o l . 60, 368-84. Brown, M.L. and E.G. Moore. 1970. "The Intra-Urban Migration Process: A Perspect ive ," Geografiska Annaler, Series B, Vo l . 52, pp. 1-13. Brown, L.A. and J . Holmes. 1971. "Intra-Urban Migration L i f e -time: A Spatial View," D-emography, Vo l . 8, pp. 103-22. Brock, Jan O.M. and John W. Webb. 1968. A Geography of Mankind, McGraw-Hill Book Company, New York. Brush, John E. 1968. "Spatial Patterns of Population in Indian C i t i e s , " Geographical Review, Vo l . 58, No. ' pp. 362-91. Claeson, C F . 1971. "A Two-Stage Model of In-Migration to Urban Centres: Deductive Development of a Variant of the Gravity Formulation," Geografiska Annaler, 50 B, V o l . 2, pp. 133-41. Clark, C. and G.R. Peters. 1965. "The Intervening Opportun-i t i e s , Methods of T r a f f i c A n a l y s i s , " T r a f f i c Quarterly, Vo l . 19, pp. 104-15. 298 Clark, W.A.V. 1970. "Measurement and Explanation in In t ra -urban Residential M o b i l i t y , " T.E.S.G., 61, pp. 49-57. C lark, John I. 1965. Population Geography, Pergamon Press, Oxford, England, p. 123. . 1971. Population Geography and the Developing Countries, Pergamon Press, Oxford, England, 1971, 273 pages. DeCani, J . S . 1961. "On the Construction of Stochast ic Models of Population Growth and Migrat ion ," Journal of Regional Science, Vo l . 3, No. 2, pp. 1-13. Drewe, P. 1970. "Steps Towards Action - Oriented Migration Research," Regional Science Association Papers, Vo l . XXVI (European Congress, London, 1970), pp. 145-64. E ldr idge , H. and Dorothy Swaine. 1964. Demographic Analysis and I n t e r r e l a t i o n s (American Phi losophical Soc iety , Phi 1adelphia). E ldr idge , H. 1965. Net Intscensal Migration for States and Geographic Divisions of the United States, 1940-60 (Ph i lade lph ia , Population Studies Center, Univers i ty of Pennsylvania, Report No. 5, 1965). Fried! ander, S .L . 1 965. Labour Migration and Economic Growth: A Case Study of Puerto Ricog, Cambridge, Mass.: M.I.T. Press. G a l l e , O.R., Karl E. Tauber. 1966. "Metropolitan Migration and Intervening Opportuni t ies ," A.S.R., V o l . 31, No 1, pp. 5-13. Gavett, T.W. 1967. Migration and Changes in the Quality of Labour Force, West V i r g i n i a Un ivers i ty . George, M.V. 1961. Internal Migration in Canada - Demographic Analysis, 1961, Census Monograph, D.B.S. 299 Get i s , A. 1969. "Residential Location and the Journey to Work," Proc Ass. Amer. Geog., No. 1, pp. 55-59. Gregory, J . 1971. "Migration in Upper V o l t a , " African Urban Notes, 6, pp. 44-52. Gould, Peter, P. 1 969. "Spatial D i f f u s i o n , " A.A.A.G.., Washington, U.S . , p. 15. Hagerstrand, T. 1967. Innovation of Diffusion as a Spatial Process, Univers i ty of Chicago Press, Chicago. Hakala, Juhar ie . 1970. "Migration as a Geographical Indicator of Social Change in E l e l a , - Pohijanamaa, F in land ," Fennia, 100, paper - 1, pp. 1-132. Hart, R.A. 1970. "A Model of Inter-regional Migration in England and Wales," Regional Studies, Vo l . 4, No. 3, pp. 279-96. Herr ick, Bruce H. 1966. "Urban Migration and Economic Development in C h i l e , " Cambridge, Mass., M.I.T. Press. Hoover^ E.M. and M-?.rP:er'l:man. 1 966. "Measuring the Ef fects of Population Control on Economic Development: A Case Study of Pakistan," PDR, pp. 545-66. Hussain, S.M. 1964. "A Note on Farmer Response to Rice in East Pakistan," PDR, pp. 223-60. Isard, W. 1960. Methods of Regional Analysis, M.I.T. Press, pp. 493-566. Islam, M.A. 1966. "The Future Relationship between Rural Population and Agr icu l ture in East Pak is tan ," Pakistan Geographical Review, V o l . 21, No. 2, pp. 39-52. Johnson, B.L .C. 1962. "Rural Population Densit ies in East Pakistan," P a c i f i c View Point, V o l . 3, pp. 61-69. 300 Johnston, R.J . 1969. "Some Tests of a Model of Int ra-urban Population Mob i l i t y : Melbourne, A u s t r a l i a , " Urban Studies, Vo l . 6, pp. 34-57. . 1971. "Resistance to Migration and the Mover/ Stayer Dichotomy: Aspects of Kingship and Popula-t ion S t a b i l i t y in an English Rural Area ," Geografiska Annaler, V o l . 53 B., No. 1, pp. 16-27. Keown, P.A. "Carr ier Cycle and the Stepwise Migration Process," New Zealand Geographer, Vo l . 27, No. 2, pp. 175-84. K e y f i t z , Nathan and Wilhelm F l e i g e r . 1971. Population: Facts and Methods of Demography, W.H. Freeman and Company, San Francisco. Khan, I.M. 1963. "A Note on Consumption Patterns in the Rural Areas of E. Pak istan," PDR, pp. 399-413. Khazaneh, H.T. 1967. "Urbanization Process in Iran During 1956-66," Contributed Papers, IUSSP, Sydney Con-ference, A u s t r a l i a , pp. 752-61. Kohn, Clydef . 1970. "The 1960's: A Decade of Progress in Geographical Research and Development," A.A.A.G., V o l . 60, pp. 211-219. K r o t k i , K .J . 1964. "The F e a s i b i l i t y of an E f fec t ive Popula-t ion Pol icy in Pakistan," PDR, pp. 283-311. Kro tk i , K .J . and Re. Hashim. 1 962. "Report on Census Enumeration," PDR, pp. 377-406. Kuper, Hilda ( e d . ) . 1965. Urbanization and Migration in West AFrica, Berkeley, Univers i ty of C a l i f o r n i a Press. Kuroda, T. 1967. "Recent Trends and Prospects of Internal Migration in Japan," Contributed Papers, IUSSP, Sydney Conference, A u s t r a l i a , pp. 708-714. 301 Laquian, Apwdieis A. 1971. Rural-Urban Migrants and Metro-p o l i t a n Development, Internal Metropolitan Studies Ser ies , Toronto, Canada. Lewis, Kurt. 1951 F i e l d Theory in Social Science, New York, Harper & Row. Lewis, L.A. 1969. "The Spatial Propert ies of Population Mobi1ity within Puerto R ico ," Journal of Tropical Geography, Vo l . 29, pp.. .33.-38. Lydolph, P.E. 1964. Geography of the U.S.S.R., John Wiley & Sons. Mangalam, J . et al. 1970. "Some Theoret ical Guidel ines Towards a Sociology of Migrat ion ," The International Migration Review, Vo l . 4, No. 11, pp. 5-21. McGinnis, R. 1968. "A Stochast ic Model of Social M o b i l i t y , " American Sociological Review, V o l . 33, pp. 712-22. Mera, Ko i ch i . 1973. "On the Urban Agglomeration and Economic E f f i c i e n c y , " Eco. Dev. and Cult. Change, V o l . 21 , pp. 309-24. Mere, D. 1971. "Direct ionary Travel Behavior and the Urban iMental Map," A.G.S., Vo l . IX, No. 2 , 133-43. Miernyk, W.H. 1955. "Labour Mobi l i ty and Regional Growth," Eco. Geog., XXXI, No. 4, pp. 321-30. M i l l e r , Edward. 1973. "Return and Non-return In-Migrat ion ," Growth and Change, Vo l . 4, No. 1, pp. 3-9. M i l l e r , Norman N. 1971. The Dynamics of Population in Uganda, American Un ivers i t ies F ie ld Staff Reports, East A f r i c a SEr ies , Vo l . X, No. 5. Mohammed, G. 1967. "Development of I r r igated Agr icu l ture in East Pakistan: Some Basic Cons iderat ion," PDR, pp. 315-375. 302 Mohammed, G. 1964. "Some Strategic Problems in A g r i c u l t u r a l Development in Pakistan," PDR, pp. 223-60. Mobugunje, Akin L. 1970. "Systems Approach to a Theory of Rural-Urban Migrat ion ," Geographical Analysis, Vo l . 2, No. 1, pp. 1-18. Morri l l , R.L. 1965. Migration and the Spread and Growth of Urban Settlement (Land Studies in Geography, Series B, Human Geography, No. 26), Lund, Gleerup. Moore, E.G. 1 969. * The Nature of Intra-Urban Migration and Some Relevant Research S t r a t e g i e s , " Proc. Assoc. Am. Geog., Vo l . 1, pp. 113-17. Myrdal , Gunnar. 1 957. Rich Lands and Poor, Harper and Row, New York. Myers, George, C R . McGinnis and George Masnick. 1967. "The Duration of Residence Approach to a Dynamic Stochastic Model of Internal Migrat ion: A Test of the Axiom of Cumulative I n e r t i a , " Eugemics Quarterly, Vo l . 14, pp. 121-26. Nelson, P. 1957. "A Study in the Geographic Mobi l i ty of Labour," Unpublished Ph.D. Thes is , Columbia Un ivers i ty , New York, 1957. NG, Ronald, 1969. "Internal Migration Regions in Scot land," Geografiska Annaler, pp. .1 39-47. Nystuen, J . D . and H.F. Dacey. 1961. "A Graph Theory Inter-pretat ion of Nodal Regions," Papers of the Regional Science Association, Vo l . 7, pp. 29-42. Okun, B. and Richardson. 1964. "Regional Income INequality and Internal Population Migrat ion ," Eco . Dev. and Cult. Change, IX, No. 1, pp. 128-43. Olsson, Gunnar. 1965 a. Distance and Human Interaction, Regional Science Research I n s t i t u t e , Ph i lade lph ia , -Pennsylvania. 3 0 3 Y Olsson, Gunnar. 1965b. "Distance and Interact ion: A Migration Study," Geografiska Annaler, V o l . 47B, pp. 3-43. O ' S u l l i v a n , P. 1970. "Variat ions in Distance F r i c t i o n in Great B r i t a i n , " Area, No. 2, pp. 36-40. Papageorgiou, G . J . 1969. "Descript ion of a Basis Necessary to the Analysis of Spatial Systems," Geographical Analysis, V o l . 1, No. 2, pp. 213-15. Parr , J . B . 1966. "Out-Migration and the Depressed Area Problem," Land Economics, V o l . XIII , No. 2, pp. 149-59. Peterson, Wi l l iam. 1958. "A General Typology of Migrat ion , " ASR, V o l . 23, No. 3, pp. 256-66. Pokshishevsky, V . V . , V.V. Vorob'yev et al. 1 964. "On Basic Migration Patterns ," Population Geography: A A Reader, Demkorj 5George J . et al. ( e d s . ) , pp. 318-31. Pryor, Robin J . 1969. "Laws of Migrat ion: The Experience of Malaysia and other Countr ies ," Geographica, V o l . 5, pp . 65-76. . 1 969. "The Residential Background of Urban Fringe Dwellers," A.G.S., V o l . VII , No. 2, pp. 159-66. R idde l , J . Barry. 1970. "On Structur ing a Migration Model," Geographical Analysis, pp. 403-9. . 1 970. The Spatial Dynamics of Moderniza-tion in Sierra Leone, Evanston, N.W. Un ivers i ty . R idde l , J . Barry and E^HarveylMi Tton. 1972. "The Urban System in the Migration Process: An Evaluation of Stepwise Migration in S ier ra Leone," Eco. Geog., Vo l . 48, No. 3, pp. 270-83. Rogers, A. 1966. "A Markovian Pol icy Model of Interregional M igrat ion , " Regional Science Association, Paper, V o l . XVII, pp.2205-24. 304 Ronald, N.G. 1970. "A Study of Recent Internal Migration In Tha i land," Journal of Tropical Geography, Vol . 31',• 65-78. Roof, Michael X. and A. Fredr ick . 1 959. "Population Re-d i s t r i b u t i o n in the Soviet Union," Geographical Review, pp. 208-24. Rummel, R .J . 1967. "Understanding Factor A n a l y s i s , " The Journal of C o n f l i c t Resolution, Vo l . XI, No. 4, pp. 444-80. Rutman, G i lber t B. 1970. "Migration and Economic Oppor-t u n i t i e s in West V i r g i n i a : A S t a t i s t i c a l A n a l y s i s , " Rural Sociology, Vo l . 35 (2), pp. 206-17. Shyrock, H. J r . 1964. Population M o b i l i t y within the United States (Chicago, Community and Family Study Center) . Simmons, J.W. 1968. "Changing Residence in the C i t y , " Geographical Review, Vo l . 58, pp. 622-51. Somermeyer, W.H. 1971. "Multipolan Human Flow Models," Regional Science Association Papers, V o l . XXVI, pp. 131-44. Stouf fer , S.A. 1971. "Intervening Opportunit ies and Competing Migrants," Journal of Regional Science, V o l . I I , No. 1 , ^pp—:il;§269 Stegman, M.A. 1969. " A c c e s s i b i l i t y Models and Residential Locat ions ," J. Am. Inst, of Planners, V o l . 35, pp. 22-29. Stewart, John Q. 1951. "Empirical Mathematical Reviews Concerning the D is t r ibut ion and Equi l ibr ium of Populat ion," Demographic Analysis, pp. 344-71. Stewart, Charles T. 1960. "Migration as a Function of Population and Distance," American Sociological Reivew, V o l . 25, No. 3, pp. 347-57. 305 Stone, L.O. 1970. Migration in Canada-Regional Aspects, 1961, Census Monograph, D.B.S. Taafe, E . J . , B .J . Garner, and M.H. Yeats. 1963. The Peripheral Journey to Work: A Geographic Consideration, Evanston, III N.W. Univers i ty Press, Transportation Center. Tay le r , P . J . 1971. "Distance Transformation and Distance Decay Funct ions," Geographical Analysis, Vo l . 3, pp. 221-38. Tekse, K. 1967. "On the Internal Migration of Jamaica's Populat ion," Contributed Papers, IUSSP, Sydney Conference, pp. 647-58. Treadgold, D.W. 1957. The Great Siberian Migration, Princeton Univers i ty Press. Wendel , B. 1953. A Migration Scheme: Land Studies in Geography, Series B, HumanGeog., No. 9. . 1 957. "Regional Aspects of Internal Migration and Mobi l i ty in Sweden, 1946-50," Migration in Sweden, Lund. Weinard, H.C. "A Spatio-Temporal Model of Economic Development, A.G.S., Vo l . X, A p r i l , No. 1, pp. 95-100. Whitelaw, J . S . 1971. "Migration Patterns and Resident ia l Select ion in Auckland, N.Z. ," AGS, V o l . IX, A p r i l , No. 1 , 61-76. . 1 972. "Scale and Urban Migration Behavior," A.G.S., Vo l . X, A p r i l , No. 1, 101-6. Whitt ing, R.F. 1952. "Home to Work Relationships of Workers L iv ing in Public Housing Projects in Chicago," Land Economics, Vo l . 28, pp., 283-90. W i l l i s , K.G. 1968. Vector Analysis and the Representation of Movement: The Case of Migration, Newcastle, England, Dept. of Geography. 306 Winkler, W. 1967. "Some New Measures of Aimed Spat ia l M o b i l i t y , " Contributed Papers, IUSSP, Sydney Conference, pp. 762-66. Wolforth, John. 1965. Residential Location and Place of Work, Tantalus , Vancouver, Canada. Wolpert, J u l i a n . 1965. "Behavioral Aspects of the Decision to Migrate," Pap. Proc. Regional Science Assoc., pp. 159-69. . 1 966. "Migration as an Adjustment to Environmental S t r e s s , " Journal of Social Issues, Vol . XXII, 92-1 02. — . 1 969. "The Basis for S t a b i l i t y of Inter-regional Transact ions ," Geographical Analysis, Vo l . ii], No. 2, A p r i l , pp5 1 52-80. Yapa, L. , M. Polese and J . Wolpert. 1971. "Interdepen-dencies of Communiting, Migration and Job S i te Re locat ion ," Eco. Geography, V o l . 47, pp. 49-72. Ze l insky , W. 1971. "The Hypothesis of the Mobi l i ty T r a n s i t i o n , " Geog. Rev., Vo l . 61, pp. 217-49. Zuiches, James J . 1970. "In-Migration and Growth of Non-Metropolitan Urban P laces , " Rural Sociology, Vol . 35(3), pp. 410-20. 307 ABBREVIATIONS O.P.E . Journal of P o l i t i c a l Economy IUSSP International Union for the S c i e n t i f i c Study of Population PIDE Pakistan Inst i tute of Development Economics PDR Pakistan Development Review AJA.A.G. Annals of the Associat ion of American Geographer A.G.S . Austra l ian Geographic Studies APPENDIX I CHRONOLOGY OF POLITICAL EVENTS The contiguous Muslim majority parts of the B r i t i s h Indian provinces of Bengal and Assam are carved out to form the province of East Bengal , which is united with the geographical ly separated Muslim provinces of the north-western part of B r i t i s h India in a federal Muslim state ca l led Pakistan. The United Front of the opposit ion p a r t i e s , demand-ing maximum regional atuonomy for East Bengal among other th ings, defeats the ru l ing Muslim League by capturing 97 per cent of the seats in the f i r s t e lect ions to the prov inc ia l l e g i s l a t u r e . The United Front Government is dismissed within weeks of assuming power by the Central Pakistan Government on the pretext of ant i -nat iona l a t t i t u d e s . The f i r s t Const i tut ion of Pakistan is adopted, forming a highly centra l i sed federat ion between East Bengal (renamed East Pakistan) and the provinces of West Pakistan (soon afterwards un i f i ed into one province) . East Pakistan is given fewer representat ives per head of e lectorate in the National Assembly than West Pakistan. Scheduled national eject ions are cance l led . General Ayub Khan takes power through a m i l i t a r y coup and imposes a v i r t u a l l y unitary form of government giving the author i tar ian Central Government the power to appoint and dismiss the Prov inc ia l Government, which was not responsible to the l e g i s l a t u r e elected on the basis of very l imi ted f ranch ise . 308 309 1966 Sheikh Mujibur Rahman of the Awami League demands the autonomy of East Pakistan on the basis of a s ix -po int programme which envisaged a central government with r e s p o n s i b i l i t i e s for defence and foreign a f f a i r s only while giving a l l taxation and economic powers to the prov inc ia l governments. 1968 The Pakistan Government attempts to crush the autonomy movement by i n s t i t u t i n g a conspiracy t r i a l against Sheikh and others. 1969 Nationwide p o l i t i c a l upsruge topples Ayub Khan, who is succeeded by General Yahya Khan. The l a t t e r promises e lect ions to t ransfer power to the popular representat ives . 1970 Sheikh's Awami League wins a massive e lect ion v ic tory in December, capturing a l l but two seats in the National Assembly from East Pakistan. 1971 On 25-6 March the Pakistan Army cracks down on the autonomist forces who were demanding immediate t ransfer of power to the elected party in East Pakistan on the basis of complete regional atuonomy. Mass k i l l i n g and destruct ion in Dacca and e lse-where in East Pakistan fol low the arrest of Sheikh by the army. The Government of the People's Republic of Bangladesh is proclaimed in Apr i l by the n a t i o n a l i s t res istance led by the Awami League and spearheaded by the rebe l l i ous Bengalis in the Pakistan army. India provides sanctuary to Bangladesh g u e r i l l a s . Tension bui lds up between India and Pakistan leading to the outbreak of f u l l - s c a l e war in early December. The j o i n t Indian and Bangladesh Command accepts the surrender of the Pakistan army in Dacca on 16 December 1971. 1972 Bangladesh is recognized by most European and Asian countries by February. Indian army with-drawal from Bangladesh completed on 12 March. 'Bangladesh' is the t r a d i t i o n a l local- language name of the l i n g u i s t i c area cons ist ing of the present state of BAngladesh and the present Indian state of West Bengal. Source: Khan (1972: XV-XVI). APPENDIX II Map No. 4.1 Measurement of D.. from the Geometric Centres of the ' D i s t r i c t s . 31 0 APPENDIX III KUSHTIA DISTRICT, CALCULATION OF E3 (An Example) 311 312 M I L E S . 1 0 . ^ ?° S C A L E 7 I N D I A s D a u l a t p u t a'miilcdia ^fy; •tBiXeramar^ Y a g o t j j ^ o r a d j • G a n g n i Jtumarkhali / ksa , M e h e r p u r . ' y C h u a t i a j i g a • \ D s f a u r h u d a D a r s h a n ? J E S S O R E D I S T R I C T I N D I A a o ® x K E Y d i s t r i c t H e a d q u a r t e r S u b d i v i s i o n a l H e a d q u a r t e r T h a n a H e a d q u a r t e r T h a n a c u m S u b d i v i s i o n a l H e a d q u a r t e r s R a i l w a y J u n c t i o n s R i v e r s R o a d s R a i l w a y l i n e T h a n a B o u n d a r y S u b d i v i s i o n a l B o u n d a r y Map Mo. 4. Kushtia D i s t r i c t , Transportat ion Networks Connecting  Administrat ive Headquarters. 313 Figure 4 . 1 . Kushtia D i s t r i c t , Rail Networks Administrat ive Headquarters. Connecti ng 314 Figure 4.2. Kushtia D i s t r i c t , River Networks Connecting Administrat ive Headquarters. 31 5 J ibannagar Figure 4.3. Kushtia D i s t r i c t , Road Networks Administrat ive Headquarters. Connecti ng APPENDIX IV REGRESSION ESTIMATES OF TWO MIGRATION STREAMS (Two Examples) 316 Table 5.103 REGRESSION ESTIMATES OF RURAL-RURAL OUT-MIGRATION STREAMS, 1951* E x p l a n a t o r y C o r r e l a t i o n C o - e f f i c i e n t ( r ) v a r i d D l e s V e l o c i t y of O u t - M i g r a t i o n From: B a r i s a l Bogra C h i t t a g o n g H.T. Comi11 a F a r i d p u r J e s s o r e Khulna Mymensi ngh No a k h a l i Pabna R a j s h a h i Rangpur S y l h e t RD/SQMCA .685 .378 -.340 .265 .001 .418 .308 .012 .089 .522 .229 .233 .726 Rice/TCA - .487 -.282 -.388 -.368 -.286 .430 -.255 -.431 -.780 -.229 -.266 -.168 .060 TCA/NCA .073 .211 .190 .265 .006 .240 .085 .322 .123 .083 .315 .273 -.573 Jute/TCA .294 .400 .197 .213 .268 .072 .036 -.565 .062 .031 .128 .213 .161 LLALF/TALF .598 .310 .086 .379 .407 .473 .423 .175 .176 .010 .011 .027 .192 16 -.078 .151 -.017 -.346 .519 -.241 -.085 -.351 .301 -.057 -.181 -.144 -.198 L i j -.309 -.048 -.298 -.328 -.388 -.070 .032 -.098 -.287 .283 .085 .063 .140 U i j -.220 -.538 .061 .200 -.198 -.341 .103 -.334 .1 72 -.221 -.166 -.136 -.342 °ij -.433 .508 -.262 -.421 -.436 -.675 -.521 -.496 -.435 -.608 -.552 -.612 -.324 Standard E r r o r .0029 .0103 .0020 .0251 .0934 .0075 .0291 .0253 .0178 .049: .0581 .0087 .0060 R 2 .8561 .9312 .6679 .8824 .9759 .7072 .9231 .9177 .9359 .9461 .7505 .9045 .9609 * F v a l u e s have not been shown as the data come from ' u n i v e r s e * r a t h e r than any 'sample'. Table 5.104 REGRESSION ESTIMATES OF RURAL-RURAL IN-MIGRATION STREAMS, 1961* E x p l a n a t o r y C o r r e l a t i o n C o - e f f i c i e n t ( r ) V a r i a b l e s V e l o c i t y of O u t - M i g r a t i o n From: B a r i s a l Bogra C ;omil l a D i n a j p u r F a r i d p u r J e s s o r e Mymensi ngh No a k h a l i Pabna Raj shahi Rangpur S y l h e t RD/SQMCA :' -.598 -.737 -.309 -.'470 -.072 -.161 -.026 -.288 -.017 -.445 -.277 -.254 Rice/TCA . .567 .089 .647 .049 .333 .673 .452 .381 .011 .062 .239 .704 TCA/NCA -.036 -.206 :-.244 -.536 -.053 -.160 -.213 .441 -.623 -.609 -.272 -.096 Jute/NCA . -.037 .'.080 : .220 .589 .311 .108 .091 .344 .286 .070 .385 -.126 SB .097 -.150 .087 -.602 .282 .205 .100 .172 .097 .167 .107 .261 SC+LLALF/: TALF + .012 -.383 -.298 -.372 -.100 -.212 -.270 -.722 -.190 -.325 -.251 -.107 SC/TALF -.029 -.744 .-.391 -.110 -.182 -.296 -.057 -.045 -.245 -.350 -.1 39 -.322 LLALF/TALF -.066 -.363 -.112 -.175 -.035 -.104 -.297 -.277 .067 -.369 -.466 -.374 L . . i j .108 .213 .352 .263 .003 .148 .401 .228 .252 .021 .296 -.446 U i j .001 • + .035 .339 .264 .147 .131 .261 .371 .222 .200 -.180 .294 °1j -.709 -.699 -.606 -.710 -.789 -.285 -.574 -.708 -.699 -.511 -.088 -.828 Standard E r r o r o f Es t i m a t e .0169 .0566 .0128 .0389 .0232 .081 ) .011 .0133 .0163 .0734 .0780 .01 52 R 2 .0619 .6 535 .8141 .7200 .6226 .583 , .4611 .7707 . 5533 .5134 .4461 .9449 *F v a l u e s have not been shown as the data come from ' u n i v e r s e ' r a t h e r than any 'sample' APENDIX V REST OF THE TABLES PREPARED FOR CHAPTER 5 31 9 Table 5.5 R-R O U T - M I G R A T I O N S T R E A M S , 1 9 5 1 S P A T I A L PATTERN OF O V I J FROM BOGRA , 1 9 5 1 Out-Migration from the i t h D i s t r i c t ~ Bogra II , n a "b" "c" "d" D i s t r i c t s of Dest inat ion ( j th D i s t r i c t s ) No. of Migrants in the Stream Ve loc i ty of Out-Migration Streams Distance Separati ng i and j (in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor ' Mi j Rank Bar isa l 64 (9) C h i t t . H.T. 8 (11) Comil la 140 (7) Faridpur 161 (6) Jessore 63 (10) Khulna 125 (8) Mymens i ngh 961 (4) Noakhali 6 (12) Pabna 3,321 (3) Rajshahi 8,516 (2) Rangpur 17,990 (1) Sylhet 177 (5) OVij Rank .0004 .0001 .0010 .0017 .0009 .0019 .0054 .0001 .0684 .1 263 .201 5 .0018 (10) (11) (8) (7) (9) (5) (4) (12) (3) (2) (1) (6; Dij Rank 175 -0.5042 (8) 250 -2.3688 (12) 120 0.4373 (5) 113 -0.1885 (7) 105 -0.1703 (6) 175 -0.6145 (10) 75 -0.6443 (4) 175 -0.6443 (11) 45 0.7062 (2) 42 0.6414 (3) 50 1.5786 (1) 150 -0.5219 (9) Factor 1 (C .F . ) OVij 0.62169 Jute/NCA -0.68993 LLALF/TALF 0.47247 Factor 3 OVij Dij TCA/NCA 0.52309 -0.92871 -0.51383 Table 5.6 R-R O U T - M I G R A T I O N STREAMS/ ' 1 9 5 1 S P A T I A L P A T T E R N OF O V I J FROM C O M I L L A , 1 9 5 1 Out--Migration from the i t h D i s t r i c t ~ Comilla a" "b II "c" II d" D i s t r i c t s of Dest inat ion ( j th D i s t r i c t s ) No. of in the Migrants Stream Veloci ty Migration of Out-Streams Distance Separati ng '. and j (in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor ' Mi j Rank OVij Rank Di j Fj* Rank Bari sal Bogra C.H.T. Faridpur Jessore Khulna Mymens i ngh Noakhali Pabna Raj shahi Rangpur Sylhet 6,111 435 549 7,149 4,962 1 ,425 10,603 7,916 3,034 6,502 2,083 28,181 (6) (12) (11) (4) (7) (10) (2) (3) (8) (5) (9) (1) .0186 .0037 .0021 .0290 .0320 .0078 .0202 .0384 .0210 .0324 .0078 .1164 (8) (11) (12) (5) (4) (10) (7) (2) (6) (3) (9) (1) 88 1 20 113 62 114 125 87 45 102 105 170 80 -1 .5193 -0.3611 0.8184 0.6900 0.4710 -0.7457 0.0069 1 .1478 -0.1246 -0.0264 -1 .7970 1 .4399 (11) (10) ' (3) (5) (6) (4) (7) (2) (9) (8) (12) (1) Factor 1 ( ' C F . 1 ) OVij 0.6371 LLALF/TALF 0.6835 Dij -0.6627 CO ro Table 5 . 7 . R-R O U T - M I G R A T I O N S T R E A M S , 1 9 5 1 S P A T I A L P A T T E R N OF O V I J FROM F A R I D P U R , 1 9 5 1 Out-•Migration from the i t h D i s t r i c t ~ Faridpur II a" "b "c" d" D i s t r i c t s of Dest inat ion ( j th D i s t r i c t s ) No. of in the Mi grants Stream Veloci ty Mi grat i on of Out-Streams Distance Separating i and j (in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor 1 Mi j Rank OVij Rank Dij Fj* Rank Bar isa l Bogra C.H.T. Comi11 a Jessore Khulna Mymens i ngh Noakhali Pabna Raj shahi R ngpur Sylhet 21,348 546 34 4,471 17,391 11 ,998 1 ,668 1 ,413 3,690 1 ,563 1 ,697 562 (1) (11) (12) (4) (2) (3) (7) (9) (5) (8) (6) (10) .0940 .0063 .0001 .0182 .1574 .0819 .0045 .0095 .0359 .0107 .0089 .0027 (2) (9) (12) (5) (1) (3) (10) (7) (4) (6) (8) (ID 55 113 1 55 62 50 75 100 75 64 125 102 145 -2.2024 0.4837 1 . 6292 0.6140 -0.1947 -0.0422 1 . 4564 0.0882 0.3051 -0.2144 0.7267 0.2191 (1) (8) (12) (9) (3) (4) (11) ( 5 ) (7) (2) (10) (6) Factor 1 *Factor 2 ( C F . ) OVij Ui j LLALF/TALF Dij •0.5167 0.9209 •0.7041 0.2819 (not s i g n i f i c a n t ) OVi j Jute/NCA Dij •0.6836 •0.8820 •0.7370 0.0450 (not s i g n i f i c a n t ) C O ro ro Table 5.8 R-R O U T - M I G R A T I O N S T R E A M S , 1 9 5 1 S P A T I A L PATTERN OF O V I J FROM J E S S O R E , 1 9 5 1 Out--Migration from the ith d i s t r i c t ~ Jessore II a" "b" "c" II d" Di s t r i cts of Dest inat ion ( j th D i s t r i c t s No. of in the Mi grants Stream Ve loc i ty of Out-Migration Streams D i stance Separati ng i and j (in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor ' Mi j Rank OMij Rank Di j _ . * Fj Rank Bar isa l Bogra C.H.T. Comil la Faridpur Khulna Mymensi ngh Noakhali Pabna Rajshahi Rangpur Sylhet 694 204 2 177 13,601 17,294 203 55 440 579 681 88 ( 3 ) 7) (12) (9) (2) (1) (8) (11) (6) (5) (4) (10) .0047 .0036 .0001 .0011 .01 23 .0204 .0008 .0005 .0066 .0063 .0056 .0006 (6) (7) (12) (8) (2) (1) (9) (11) (3) (4) (5) (10) • 87 1 05 205 113 58 62 1 25 128 65 90 162 182 -0.1250 0.1895 0.6965 0.7758 -0.4990 -2.2020 0.5525 0.1897 -1.042C -0.7231 0.6344 1 .5525 (5) (6) (10) (11) (4) (1) (8) (7) (2) (3) (9) (12) Factor 1 ( ' C F . ' ) OVij -0.8586 Dij 0.7910 Rice/TCA -0.7353 LLALF/TALF -0.5987 CO ro co Table 5.9 R-R O U T - M I G R A T I O N S T R E A M S , 1 9 5 1 S P A T I A L P A T T E R N OF O V I J FROM KHULNA , 1 9 5 1 Out- Migration from the i t h D i s t r i c t ~ Khulna II a" "b "c" II d" D i s t r i c t s of Dest inat ion ( j th D i s t r i c t s ) No. of in the Migrants Stream V e1o c i ty Migration of Out-Streams Di stance Separati ng i and j ( in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor ' Mij Rank OMij ' Rank . Dij Fj* Rank Bar isa l Bogra C.H.T. Comil la Faridpur Jessore Mymensi ngh Noakhali Pabna Raj s ha hi Rangpur Sylhet 4,097 100 2 191 1 ,871 17,852 91 69 143 300 254 88 (2) (8) (12) (6) (3) (1) (9) (11) (7) (4) (5) (10) .0226 .001 5 .0001 .0010 .01.37 .1574 .0005 .0004 .0018 .0026 .0016 .0005 (2) (7) (12) (8) (3) (1) (9) (11) (5) (4) (6) (10) 50 175 188 1 25 75 62 170 n o 125 158 225 21 2 -1.1243 0.4458 0.8372 0.4364 -0.4651 -2.4630 0.5345 0.1996 -0.3258 -0.0345 0.9073 1 .05178 (2) (8) (10) (7) (3) (1) (9) (6) (4) (5) (11) (12) Factor 1 ( ' C F . ' ) OVij -0.8463 Dij 0.8473 TCA/NCA -0.2228 (not s i g n i f i c a n t ) CO ro Table 5.10 R-R.OUT-MIGRATION STREAMS, 1 9 5 1 S P A T I A L PATTERN OF O V I J FROM MYMENSINGH, 1 9 5 1 Out-Migration from the i t h D i s t r i c t ~ Mymensingh n. n a •I5 •• 11 c" "d" Di s t r i cts of Dest inat ion ( j th D i s t r i c t s ) No. of Migrants in the Stream Ve loc i ty of Out-Migration Streams D i stance Separating i and j ( in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor 1 Bar isa l Bogra C.H.T. Comi11 a Faridpur J essore Khulna Noakhali Pabna Raj shahi Rangpur Sylhet Mij ~ Rank OMij - Rank Di j Fj* ~ Rank 704 (9) 2,205 (6) 645 (12) 4,823 (4) 984 (7) 322 (11) 444 (10) 722 (8) 4,300 (5) 5,328 (3) 40 ,541 (1 ) 29,113 ( 2 ) .0014 (11) .0114 (5) .0032 (7) .0092 (6) .0026 (8) .0013 (12) .0015 (10) .0022 (9) .0196 (3) .0175 (4) .1007 (1 ) .0690 (2) 150 . 75 200 87 75 125 170 138 80 112 87 80 .54530 (10) .53932 (9) .60087 (11) .7083 (2) .4930 (7) .5160 (8) .4611 (6) .7467 (12) -.3729 (3) .2948 (5) -2.3594 (1) 0.2433 (4) *Factor 1 ( ' C . F . ' ) OVij -0.8873 TCA/NCA -0.7596 Jute/NCA 0.5074 Dij 0.4209 OO ro Table 5.11 R - R O U T - M I G R A T I O N S T R E A M S , 1 9 5 1 S P A T I A L P A T T E R N O F O V I J F R O M N O A K H A L I , 1 9 5 1 Out-Migration from the ith District ~ Noakhali II a" "b II "c" - II d" Di s t r i cts of Dest inat ion ( j th D i s t r i c t s ) No. of in the Migrants Stream V e 1 o c i ty Mi grat i on of Out-Streams Distance Sepa ra t i ng i and j (in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor ' Mi j Rank OMij Rank Dij Fj Rank Bar isa l Bogra C.H.T. Comil la Faridpur Jessore Khulna Mymens i ngh Pabna Raj shahi Ra ngpur Sylhet 1 6 , 0 1 9 472 846 ' 1 5 , 8 0 2 3 , 7 3 4 9 , 5 6 0 1 , 1 74 6 , 4 0 6 6 . 9 2 , 4 1 3 11 , 1 6 8 1 0 , 7 8 7 ( 1 ) ( 1 2 ) ( 1 0 ) ( 2 ) (7) ( 5 ) ( 9 ) ( 6 ) ( 1 1 ) ( 8 ) ( 3 ) ( 4 ) . 0 8 1 0 . 0 0 6 5 . 0 5 3 6 . 0 7 6 8 . 0253 . 0 8 4 9 . 0 2 5 9 . 0 2 0 4 . 0071 . 0 2 0 0 . 0 7 0 3 . 0 6 4 4 ( 2 ) ( 1 1 ) ( 6 ) ( 3 ) ( 8 ) ( 1 ) (7), ( 9 ) ( 1 2 ) ( 1 0 ) ( 4 ) ( 5 ) 62 175 80 45 64 65 n o 138 1 37 185 21 5 130 - 0 . 9 5 2 2 0 . 1 4 6 9 0 . 7 6 0 3 -1 . 2 436 - 0 . 4 1 0 9 - 1 . 0 0 9 4 0 . 6 4 3 0 0 . 6 8 3 0 1 . 4606 1 . 4318 - 0 . 2 2 7 6 -1 . 2 820 ( 4 ) (7) ( 1 0 ) ( 2 ) (5) ( 3 ) ( 8 ) (9) ( 1 2 ) (11) ( 6 ) ( 1 ) * ' C F . ' OVij Rice/TCA - 0 . 7 7 6 1 0 . 9 4 4 8 TCA/NCA Dij • 0 . 5 8 4 7 .6611 CO ro Table 5.12 R - R O U T - M I G R A T I O N S T R E A M S , 1 9 5 1 S P A T I A L P A T T E R N O F O V I J F R O M P A B N A , 1 9 5 1 Out-•Migration from the i t h D i s t r i c t ~ Pabna a "b" "c" II d" D i s t r i c t s of Dest inat ion ( j th D i s t r i c t s ) No. of Migrants in the Stream Ve loc i ty of Out-Migration Streams Di stance Sepa ra t i ng i and j (in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor ' Mij ~ Rank OMij ~ Rank Dij Rank Bar isa l Bogra C.H.T. Comi11 a Faridpur Jessore Khulna Mymensingh Noakhali Rajshahi Ra ngpur Sylhet 149 12,702 19 342 5,440 465 3,039 12,131 1 6 18,063 13,999 101 .0010 .2627 .0002 .031 5 .0528 .0070 .0902 .0554 .0001 .2164 .1264 .0930 1 25 45 21 2 102 125 90 1 25 80 137 62 1 00 137 0.9551 -2.0428 1 .8603 0.0528 -0.3788 0.1305 -0.1314 -0.1680 0.3699 -1.3849 0.2232 0.3958 *Factor 1 ( 'C .F . ' ) OVij -0.8906 Dij 0.8801 IB 0.6724 Ui j 0.4305 co ro Table 5 .1 3 R-R O U T - M I G R A T I O N S T R E A M S , 1 9 5 1 S P A T I A L P A T T E R N OF O V I J FROM R A J S H A H I / 1 9 5 1 Out-Migration from the i t h D i s t r i c t ~ Rajshahi n_ II a "c" D i s t r i c t s of Dest inat ion ( j th D i s t r i c t s ) No. of Migrants in the Stream Ve loc i ty of Out-Migration Streams Di stance Separati ng-i and j (in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor ' Mij Rank Bar isa l n o (9) Bogra 11,569 (1) C.H.T. 41 (12) Comil la 1 55 (8) Faridpur 388 (5) Jessore 280 (6) Khulna 1 ,296 ( 4 ) Mymensi ngh 269 (7) Noakhali 45 (11) Pabna 4,343 (2) Rangpur 1 ,955 (3) Sylhet 96 (10) OVij Rank Dij .0005 .1714 .0015 .0017 .0019 .0029 .0275 .0009 .0002 .0517 .0126 .0006 (11) (1) (8) (7) (6) (5) (3) (9) (12) (2) ( 4 ) (10) 175 42 275 102 125 90 158 112 185 62 84 187 Fj 0.7720 -2 0651 0.3694 1.0256 0.2211 0.6117 0.6938 0.3105 0.1449 1.5445 0.4174 1.1696 Rank (10) (1) ( 4 ) (11) ( 6 ) (9) ( 3 ) (7) (5) (2) ( 8 ) (12) Factor 1 ( ' C F . ' ) OVij Rice/TCA Dij •0.7817 0.7736 0.3386 (not s i g n i f i c a n t ) CO ro Co Table 5.14 R-R O U T - M I G R A T I O N S T R E A M S , 1 9 5 1 S P A T I A L P A T T E R N OF O V I J FROM RANGPUR, 1 9 5 1 Out- Migration from the i t h D i s t r i c t ~ Rangpur II a" "b n "c" n d" D i s t r i c t s of Dest inat ion ( j th D i s t r i c t s ) No. of i n the Migrants Stream Veloci ty Migration of Out-Streams Separati ng Di stance i and j (in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor ' Mi j Rank OVij • Rank Dij Fj Rank Bar isa l Bogra C.H.T. Comil la Fai rdpur Jessore Khulna Mymens i ngh Noakhali Pabna Raj shahi Sylhet 483 3,879 24 131 188 148 194 2,532 121 544 1 ,104 62 (5) (1) (12) (9) (7) (8) (6) (2) (10) (4) (3) (11) .0005 .0434 .0011 .0006 .0009 .0012 .0030 .0062 .0009 .0041 .0071 .0003 (11) (1) (7) (10) (8) (6) (5) (3) (9) (4) (2) (12) 220 50 282 170 162 165 110 90 21 5 100 84 150 -0.0574 -2.0500 1 .01375 0.6890 0.1148 -0.3877 1 .4765 -0.4672 1 .3024 -0.4700 -0.8395 -0.3247 (7) (1) (10) (9) (8) (5) (12) (3) (11) (4) (2) (6) *Factor 1 ( ' C F . ' ) OVij -0.7190 Dij 0.8522 RD/SQMCA -0.7182 Uij 0.4168 0 0 ro CO Table 5.15 R-R O U T - M I G R A T I O N S T R E A M S , 1 9 5 1 S P A T I A L P A T T E R N OF O V I J FROM S Y L H E T , 1 9 5 1 "a" Out-Migration from the i t h D i s t r i c t - Sylhet " c" D i s t r i c t s of Dest inat ion ( j th D i s t r i c t s ) No. of Migrants in the Stream Ve loc i ty of Out-Migration Streams Separati ng Distance i and j ( in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor ' Bar isa l Bogra C.H.T. Comi11 a Faridpur Jessore Khulna Mymens i ngh Noakhali Pabna Raj shahi Rangpur Mi j Rank OVij 249 (6) .0009 93 (10) .0008 156 (7) .0068 7,011 (2) .0109 1 29 (8) .0006 114 (9) .0008 85 (11) .0002 10,537 (1) .0249 1 ,086 (5) .0064 64 (12) .0001 6,502 (3) .0402 2,083 (4) .0097 Rank (7) (9) (5) (3) (10) (8) (11) (2) (6) (12) (1) (4) Dij Fj Rank 175 0.2488 (6) 1 50 0.9684 (12) 162 -0.1891 (5) 80 -0.6409 (3) 145 0.9091 (11) 182 0.7288 (9) 21 2 0.7755 (10) 80 -0.8581 (2) 130 0.5356 (8) 1 37 0.3582 (7) 187 -2.8512 (1) 1 50 -0.3180 (4) Factor 1 ( 'C .F . 1 ) OVij RD/SQMCA TCA/NCA Dij •0.9457 •0.7934 •0.7004 0.1536 (not s i g n i f i c a n t ) . OJ CO o Table 5.16 R-R I N - M I G R A T I O N S T R E A M S , 1 9 5 1 S P A T I A L P A T T E R N OF I V I J TO B A R I S A L , 1 9 5 1 "a" D i s t r i c t of Destination (jth D i s t r i c t ) ~ Barisal "c" Di s t r i c t s of Or ig in ( i th D i s t r i c t s ) No. of Migrants in the Stream Ve loc i ty of Out-Migration Streams Separati ng Distance i and j ( in miles) Factor scores on the Subjects fromthe ' C r i t i c a l Factor ' IMij Rank Bogra 64 (11) C h i t t . H.T. 10 (12) Comi11 a 6,111 (3) Faridpur 21,348 (1) Jessore 694 (6) Khulna 4,097 (4) Mymens i ng h 704 (5.) Noakhali 16,019 (2) Pabna 149 (8) Raj shahi n o (9) Rangpur 83 (10) Sylhet 249 (7) IV i j Rank Dij 0006 ,0001 0186 0089 0049 0227 0014 0893 0010 0006 0004 0009 (10) (12) (4) (2) (5) (3) (6) (1) (7) (9) (11) (8) 175 132 88 55 87 50 1 50 62 125 175 220 175 Fj Rank 0.3631 (6) 0.5616 (8) 0.6110 (4) 1.0989 (3) 0.5542 (7) 1.6903 (1) 0.6767 (10) 1.6239 (2) 1.7028 (11) 0.2305 (5) 1.3497 (12) 0.5852 (9) Factor 1 ( ' C F . ' ) IV i j -.7771 Dij 0.8752 RD/SQMCA 0.8739 TCA/NCA 0.56009 OO OO Table 5 . 1 8 R - R I N - M I G R A T I O N S T R E A M S , 1 9 5 1 S P A T I A L P A T T E R N O F I V I J T O C O M I L L A , 1 9 5 1 D i s t r i o t of Destination II _ II o (jth D i s t r i o t ) ~ Comilla " d D i s t r i c t s of Or ig in ( i th D i s t r i c t s ) No. of Migrants in the Stream Ve loc i ty of Out-Migration Streams Separating Distance i and j (in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor ' Bar i sa l Bogra C h i t t . H.T. Faridpur Jessore Khulna Mymensi ngh Noakhali Pabna Rajshahi Rangpur Sylhet IMij Rank IV i j 1 , 1 6 9 (5 ) ' . 0 0 3 5 140 ( 1 0 ) . 0 0 0 9 47 ( 1 2 ) . 0 0 1 8 4 , 4 7 1 ( 4 ) .01 78 177 ( 8 ) . 0 0 1 2 191 ( 7 ) . 0 0 1 0 4 , 8 2 3 ( 3 ) . 0 0 9 2 1 5 , 8 0 2 ( 1 ) . 0 8 4 6 342 ( 6 ) . 0 0 2 4 152 ( 9 ) . 0 0 0 8 121 ( 1 1 ) . 0 0 0 5 7 ,011 ( 2 ) . 0 2 0 0 Rank ( 5 ) ( 1 0 ) ( 7 ) (3 ) ( 8 ) ( 9 ) ( 4 ) (1 ) (6 ) ( 1 1 ) ( 1 2 ) ( 2 ) Dij Fj Rank 88 1 . 0 1 6 7 ( 1 1 ) 120 0 . 3 9 7 8 ( 9 ) 113 - 0 . 2 3 9 9 ( 5 ) 62 - 0 . 8 0 1 8 ( 3 ) 114 0 .1281 (6 ) 125 0 . 2 8 1 0 ( 8 ) 87 - 0 . 2 9 6 9 ( 4 ) 45 - 2 . 6 0 7 9 ( 1 ) 102 0 . 2 2 7 5 ( 6 ) 105 0 . 5 3 0 2 ro o 170 1 . 3 3 7 3 ro o 80 - 0 . 8 2 7 7 ( 2 ) Factor 2 ('C . F . ' ) IV i j - 0 . 8 8 3 9 Dij 0 . 7 7 1 1 RD/SQMCA 0 . 4 5 8 0 CO CO ro Table 5.19 R - R I N - M I G R A T I O N S T R E A M S , 1 9 5 1 S P A T I A L P A T T E R N O F I V I J TO F A R I D P U R , 1 9 5 1 D i s t r i c t of Destination (jth D i s t r i c t ) ~ Faridpur a" "b "c" II d" D i s t r i c t s of Or ig in ( i th D i s t r i c t s ) No. of in the Migrants Stream Vel oci ty Migration of Out-Streams Separati ng Di stance i and j (in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor 1 IMij Rank IV i j Rank Dij Fj Rank Bar isa l Bogra C h i t t . H.T. Comil la Jessore Khulna Mymensi ngh Noakhali Pabna Rajshahi Rangpur Syl het 6,141 161 4 7,149 13,601 1 ,871 984 3,734 5,140 308 188 129 (3) (11) (12) (2) (1) (6) (7) (5) (4) (8) (9) (10) .0255 .0019 .0001 .0285 .1259 .0137 .0026 .0273 .0492 .0021 .0009 .0006 (5) (8) (12) (3) (1) (6) (7) (4) (2) (11) (9) (10) 55 113 1 55 62 50 75 100 75 64 125 102 145 -0.2291 0.5922 0.6770 -0.4713 -2.5882 -0.0901 0.7657 -0.3462 -0.8048 0.6134 0.9763 0.7248 (5) (9) (8) (3) (1) (6) (11) (4) (2) (7) (12) (10) * ( ' C F . ' ) IV i j Dij -0.9786 0.7748 OO OO OO Table 5.20 R-R I N - M I G R A T I O N S T R E A M S , 1 9 5 1 S P A T I A L PATTERN OF I V I J TO J E S S O R E , 1 9 5 1 D i s t r i c t of destination "a" (pth D i s t r i o t ) - Jessore D i s t r i c t s of Or ig in ( i t h D i s t r i c t s ) No. of Migrants in the Stream Ve loc i ty of Out-Migration Streams Sepa r a t i ng Distance i and j (in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor ' IMij Rank Bar isa l 1 ,968 (5) Bogra 63 (11) C h i t t . H.T. 2 (12) Comil la 4,962 (4) Faridpur 17,391 (2) Khulna 17,858 (1) Mymensi ngh 322 (7) Noakhali 9,560 (3) Pabna 465 (6) Rajshahi 280 (8) Rangpur 148 (9) Sylhet 114 (10) I Vi j Rank Dij .0138 .0012 .0001 .0336 .1609 .2209 .0014 .1185 .0075 .0032 .001 3 .001 2 (5) (11) (12) (4) (2) (1) (8) (3) (6) (7) (9) (10) 87 105 205 113 50 62 125 128 65 90 162 183 Fj -2 0.2113 0.1987 0.9155 0.0103 0.8341 2666 0.7627 •0.9855 •0.1140 0.2547 1 . 2708 0.9989 Rank (4) (7) (11) (6) (3) (1) (9) (2) (5) (8) (12) (10) >actor 1 ( ' C . F. 1 ) IV i j -0.8965 RD/SQMCA 0.7927 Dij 0.7183 LLALf/TALF 0.5268 TCA/NCA 0.5094 Table 5.21 R-R I N - M I G R A T I O N S T R E A M S , 1 9 5 1 S P A T I A L PATTERN OF I V I J TO KHULNA , 1 9 5 1 D i s t r i c t of Destination (jth D i s t r i o t ) ~ Khulna I I , 11 a Di s t r i cts of Or ig in ( i th D i s t r i c t s ) No. of Migrants in the Stream Veloc i ty of Out-Migration Streams Di stance Separati ng i and j (in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor ' IMij Rank IVi j Rank Dij Bar isal Bogra C h i t t . H.T. Comi11 a Faridpur Jessore Mymensi ngh Noakhali Pabna Raj shahi Rangpur Sylhet 24,864 125 3 1 ,425 11 ,998 17,294 431 2,944 253 384 188 193 (1) " 1 12) (5) (3) (2) (6) (4) (8) (7) (10) (9) .1383 .0198 .0001 .0076 .0876 .2139 .001 5 .0288 .0022 .0035 .0013 .0012 (2) , ( 5 1 12 (6) (3) (1) (9) (4) (8) (7) (11) (10) 50 175 188 125 75 62 170 110 125 158 225 212 Fj Rank 1 .6251 4 (2) 0.3530 , ( 7 ) 0.8957 10 0.6235 (9) 0.7609 (3) 2.0850 (1) 0.5687 (8) 0.3333 (6) 0.1556 (4) 0.0447 (5) 0.8001 (11) 1 .0074 (12) Factor 1 ( 'C .F . ' ) IVi j -0.9389 Dij 0.8710 CO CO cn Table 5.22 R-R I N - M I G R A T I O N S T R E A M S , 1 9 5 1 S P A T I A L P A T T E R N OF I V I J TO MYMENS INGH , 1 9 5 1 D i s t r i c t of Destination ( j t h D i s t r i c t ) ~ Mymensingh M n a "b" "c" 11 d" D i s t r i c t s of Or ig in ( i th D i s t r i c t s ) No. of Migrants in the Stream Ve loc i ty of Out-Migration Streams Separating Distance i and j (in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor 1 Bari sal Bogra C h i t t . H.T. Comi11 a Faridpur Jessore Khulna Noakhali Pabna Rajshahi Rangpur Sylhet * IMij Rank IV i j Rank Dij Fj* Rank 1,086 (7) 961 (8) 4 (12) 10,603 (2) 1,688 (5) 203 (9) 91 (11) 6,406 (4) 12,131' (1 ) 155 (10) 2,532 (6) 10,537 (3) .0021 (8) .0055 (6) .0001 (12) .0203 (4) .0032 (7) .0009 (9) .0003 (11) .0224 (3) .0556 (1) .0005 (10) .0063 (5) .0250 (2) 1 50 75 200 87 75 125 170 138 80 112 87 80 j — i - .5858 (5) - .4836 (6) - .1498 (7) .0281 (8) - .8026 (4) -1.1568 (2) 1.2324 (12) .4092 (1 1 ) -2.2903 (1) 0.1206 (9) .2780 (10) -1.1801 (3) *Factor 1 ( ' C F . ' ) IV i j -0.5900 Rice/TCA -0.8052 Dij 0.0781 (not s i g n i f i c a n t ) CO CO cn Table 5.23 R - R I N - M I G R A T I O N S T R E A M S , 1 9 5 1 S P A T I A L P A T T E R N O F I V I J T O N O A K H A L I , 1 9 5 1 D i s t r i o t of Destination "a" (jth D i s t r i o t ) ~ Noakhali D i s t r i c t s of Or ig in ( i th D i s t r i c t s ) No. of Migrants in the Stream Ve loc i ty of Out-Migration Streams Sepa r a t i ng Di stance i and j (in miles) Factor Scores on the Subjects from the ' C r i t i c a l FActor ' IMij Rank Bar isa l 9.890 (1) Bogra 6 (11) C.H.T. 5 (12) Comi 11 a 7,916 (2) Faridpur 1 ,413 (3) Jessore 55 (8) Khulna 69 (7) Mymens ingh 722 (5) Pabna 16 (10) Rajshahi 45 (9) Rangpur 121 (6) Sylhet 1 ,086 (4) IV i j Rank .0551 .0001 .0001 .0423 .0103 .0007 .0006 .0025 .0002 .0004 .0008 .0072 (1) (11) (12) (2) (3) (7) (8) (5) (10) (9) (6) (4) Dij Fj Rank 62 .97629 (3) 175 .30496 (6) 80 -1.12240 (10) 45 1.28145 (1) 64 0.78682 (4) 65 0.57442 (5) 110 -0.48740 (8) 138 -0.63472 (9) 137 -1 .30309 (11) 185 rl .47009 (12) 21 5 -0.16182 (7) 130 1.25557 (2) Factor 1 IVi j LLALF/TALF RD/SQMCA Uij •0.5333 0.9523 .7151 .61 50 r Factor 3 ( ' C F . ' ) IV i j .6008 Rice .8818 TCA/NCA -.7116 L i j .7101 Dij -.4822 CO co Table 5.24 R-R I N - M I G R A T I O N S T R E A M S , 1 9 5 1 S P A T I A L PATTERN OF I V I J TO P A B N A , 1 9 5 1 D i s t r i c t of destination (jth D i s t r i c t ) ~ Pabna a" "b "c" d" Di s t r i cts of Or ig in ( i t h D i s t r i c t s ) No. of in the Mi grants Stream Veloci ty Migration of Out-Streams Separati ng Di stance i and j (in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor ' IMij Rank IV i j Rank Dij Fj Rank Bar isa l Bogra C h i t t . H.T. Comilla Fari dpur Jessore Khulna Mymens ingh Noakhali Rajshahi Rangpur Sylhet 381 3,321 7 3,034 3,690 440 143 4,300 619 4,343 544 64 (8) (4) (12) (5) (3) (10) (9) , (2) (6) (1) (7) (11) .0027 .0690 .0001 .0212 .0353 .0071 .0018 .0197 .0081 .0523 .0049 .0005 (9) (1) (12) (4) (3) (7) (10) (5) (6) (2) (8) (11) 125 45 21 2 102 1 25 90 125 80 137 62 100 137 0.5613 -2.0150 1 .01 5 0.3561 -0.7134 0.6111 0.7156 -0.4561 0.7999 -1 .8165 -0.3915 0.8566 (7) (1) (12) (6) (3) (8) (9) (4) (10) (2) (5) (11) *Factor 1 ( ' C . F . ' ) IV i j TCA/NCA Dij -0.8561 0.7555 0.8311 -CO OO Table 5.25 R-R I N - M I G R A T I O N S T R E A M S , 1 9 5 1 S P A T I A L P A T T E R N OF I V I J TO R A J S H A H I , 1 9 5 1 D i s t r i o t of Destination (jth D i s t r i c t ) ~ Rajshahi "a" "b" "d" D i s t r i c t s of Or ig in ( i t h D i s t r i c t s ) No. of Migrants in the Stream Ve loc i ty of Out Migration Streams Separati ng Di stance i and j (in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor 1 Bar isa l Bogra C h i t t . H.T Comi11 a Faridpur Jessore Khulna Mymens i ngh Noakhali Pabna Rangpur Sylhet IMij Rank IV i j 737 (8) .0038 8,516 (2) .1 271 6 (12) .0001 6,502 (3) .0327 1 ,563 (6) .0107 579 (9) .0067 300 (10) .0027 5,328 (4) .0175 2,413 (5) .0222 18,063 (1) .2175 1 ,016 (7) .0101 173 (11) .0010 Rank Dij (9) (2) (12) (3) (6) (8) (10) (5) (4) (1) (7) (11) 175 42 275 102 1 25 90 158 112 185 62 84 187 Fj •0.8400 0.9115 0.2877 •0.8804 0.3888 •0.7263 0.5932 •0.3022 •0.0359 2.3894 •0.5951 •1 .1906 Rank (10) (2) (5) (11) (4) (9) (3) (7) (6) (1) (8) (12) Factor 2 (.'C.F. ' ) IV i j 0.8343 Rice 0.8577 L i j 0.4098 Dij -0.2878 (not s i g n i f i c a n t ) C O C O V O Table 5.26 R-R IN-MIGRATION STREAMS, 1951 S P A T I A L PATTERN OF I V I J TO RANGPUR, 1951 D i s t r i c t of Destination II _ II a "b" (pth D i s t r i c t ) - Rangpur "c" D i s t r i c t s of Or ig in ( i t h D i s t r i c t s ) No. of Migrants in the Stream Ve loc i ty of Out-Migration Streams Separati ng Distance i and j (in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor ' IMij Rank Bar isa l 998 (8) Bogra 17,990 (2) C.H.T. 6 (12) Comi 11 a 2,083 (5) Faridpur 1 ,697 (7) Jessore 681 (9) Khu1na 254 (10) Mymensi ngh 40,541 (1) Noakhali 11,168 (4) Pabna 13,999 (3) Rajshahi 1 ,955 (6) Sylhet 182 (11) IV i j Rank Dij .0039 .2026 .0001 .0079 .0088 .0060 .0018 .1009 .0776 .1272 .0127 .0009 (9) (D (12) (7) (6) (8) (10) (3) (4) (2) (5) (11) 220 50 282 170 162 165 210 90 21 5 100 84 150 Fj' 5242 6731 6400 •0.3346 •0.2097 •0.0617 0.6406 •1 .291 2 0. 2922 •0.9903 0.1763 0.2603 Rank (4) (1) (11) (4) (5) (6) (10) (2) (9) (3) (7) (8) Factor 1 ( ' C F . ' ) IV i j -0.7679 Dij 0.86 62 Jute -0.8084 16 -0.5566 Table 5.27 R - R I N - M I G R A T I O N S T R E A M S , 1 9 5 1 S P A T I A L P A T T E R N O F I V I J T O S Y L H E T , 1 9 5 1 D i s t r i c t of Destination n I I d (jth District) ~ Sylhet "c" Di s t r i cts of Or ig in ( i t h D i s t r i c t s ) No. of Migrants in the Stream Ve loc i ty of Out-Migration Streams Separating Distance i and j ( in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor ' IMij Rank Bar isa l . 577 (4) Bogra 177 (6) C.H.T. 18 (12) Comil la 28,181 (2) Faridpur 562 (5) Jessore 88 (9) Khulna 88 (10) Mymens i ngh 29,113 (1) Noakhali 10,787 (3) Pabna 101 (7) Raj shahi 96 (8) Rangpur 62 (11) IV i j Rank Dij ,0021 ,0019 ,0003 ,1020 0028 0007 0003 0690 0714 0009 0007 0005 (5) (6) (12) (1) (4) (8) (ID (3) (2) (7) (9) (10) 175 1 50 162 80 145 182 21 2 80 130 317 187 150 Fj' -2 1 0.3231 0.7537 0.0703 0089 0.5478 0.6765 2637 1 .6626 0.801 2 0.4634 0.5809 0.0663 Rank (6) (11) ( 4 ) (1) (8) (10) (12) (2) (3) (7) (9) (5) 'Factor 1 ( ' C F . ' ) IV i j -.9145 Dij .8972 Table 5.28 R-U OUT-MIGRATION STREAMS FROM BARISAL 1951 Out Migration from the i t h D i s t r i c t ~ Barisal II a" "b II "c" "d D i s t r i c t s of Dest inat ion ( j th D i s t r i c t s ) No. of in the Migrants Stream Veloci ty Mi grat i on of Out-Streams Distance Separating i and j ( in miles) Factor scores on the Subjects from the ' C r i t i c a l Factor ' OMi j Rank OVij Rank Dij Fj Rank Chi ttagong 1,372 (2) .0067 (2) 112 -0.6130 (2) Dacca 5,523 (1) .0076 (1) 90 -1 .6130 (1) Di naj pur 570 (3) .0016 (4) 250 0.1040 (3) Kushti a 233 (4) .0028 (3) 125 0.7510 (4) " C F . " ( C r i t i c a l Factor) N.B. OVij -0.9080 Level of s i g n i f i c a n c e > .4000 LLALF/TALF -0.7561 Rice/TCA 0.6430 Uij -0.1350 (not s i g n i f i c a n t ) Dij 0.6610 Table 5.29 R-U OUT-MIGRATION STREAMS FROM BOGRA, 1951 Out-Migration from the i t h D i s t r i o t ~ Bogra II a" "b II "c" "d" D i s t r i c t s of Dest inat ion ( j th D i s t r i c t s ) No. of in the Migrants Stream Ve loc i ty Migration of Out-Streams Distance Separati ng i and j (in miles) Factor scores on the Subjects fro.:: the ' C r i t i c a l Factor ' OMij Rank OVij Rank Dij Fj Rank Chittagong 406 (3) .0057 (4) 250 0.8560 (4) Dacca 955 (2) .0076 (3) 100 0.5600 (3) Di naj pur 2,506 0 ) .0595 (1) 80 -1.015 (1) Kushti a 243 (4) .0088 (2) 82 -0.4431 (2) * ' C F . OVij -0.8130 Dij 0.8230 TCA/NCA. -0.6456 Uij -0.1815 (not s i g n i f i c a n t ) Table 5.30 R-U OUT-MIGRATION STREAMS FROM COMILA, 1951 Out-Migration from the i t h D i s t r i c t - Comilla "a" " b" "c" M d " D i s t r i c t s of Dest inat ion ( j th D i s t r i c t s ) No. of Migrants in the Stream Ve loc i ty of Out-Migration Streams Distance Separating i and j ( in miles) Factor scores on the Subjects from the ' C r i t i c a l Factor ' Chi ttagong Dacca Di naj pur Kushtia OMij ~ Rank OVij - Rank Dij Fj* - Rank 1,033 (4) 25,306 (1) 2,074 (3) 2,856 (2) .0527 (4) .6850 . (1) .1660 (3) .1990 (2) 112 45 212 137 0.1561 (3) -1.016 (1) 0.4151 (4) -0.6150 (2) * ' C . F . ' OVij -0.8409 LLALF/TALF -0.7601 Dij 0.3165 (not s i g n i f i c a n t ) U i j -0.0156 Jute/NCA 0.8244 co Table 5.31 R-U OUT-MIGRATION STREAMS FROM FARIDPUR, 1951 Out-•Migration from the i t h D i s t r i o t -- Faridpur a" "b n "c" "d i i D i s t r i c t s of Dest inat ion (jth D i s t r i c t s ) No. of in the Migrants Stream V e 1 o c i ty Migration of Out-Streams Distance Separati ng i and j ( in miles) Factor scores on the Subjects from the ' C r i t i c a l Factor ' OMi j Rank OVij Rank Dij Fj Rank Chittagong 1 ,596 (3) .0106 (4) 1 50 0.2463 (3) Dacca 8,622 (1) .0325 (2) 47 -1.056 (1) Di naj pur 1 ,041 (4) .0114 (3) 195 0.5551 (4) Kushtia 3,439 (2) .0597 (1) 80 -0.8051 (2) * ' C F . OVij -0.8356 LLALF/TALF -0.8461 Jute/NCA 0.6571 Uij -0.1561 (not s i g n i f i c a n t ) Dij 0.7544 Table 5.33 R-U OUT-MIGRATION STREAMS FROM KHULNA, 1951 Out-Migration from the i t h D i s t r i c t - Khulna M _ n d "b" "c" "d" D i s t r i c t s of Dest inat ion ( j th D i s t r i c t s ) No. of Migrants in the Stream Ve loc i ty of Out-Migration Streams Distance Separati ng i and j ( in miles) Factor scores on the Subjects from the ' C r i t i c a l Factor ' Chi ttagong Dacca Di najpur Kushtia OMij ~ Rank OVij ~ Rank Dij Fj* ~ Rank 576 (2) 1,163 (1) 146 (4) 267 (3) .0030 (3) .0056 (2) .0020 (4) .0059 (1) 162 115 237 105 0.1618 (3) -0.9134 (2) 0.7134 (4) -1.5618 (1) * ' C F . OVij -0.8856 Di j 0.7605 LLALF/TALF -0.7888 U i j -0.1105 (not s i g n i f i c a n t ) Table 5.34 R-U OUT-MIGRATION STREAMS FROM MYMENSINGH, 1951 Out-Migration from the i t h D i s t r i c t ~ Mymensingh M. H a "b" "c" 11 d" D i s t r i c t s of Dest inat ion ( j th D i s t r i c t s ) No. of Migrants in the Stream Ve loc i ty of Out-Migrat ion Streams Di stance Separati ng i and j (in miles) Factor scores on the Subjects from the ' C r i t i c a l Factor ' Chi ttagong Dacca Dinajpur Kushtia OMij ~ Rank OVij - Rank, Dij Fj* ~ Rank 1 ,045 (3) 22,977 (1) 1,507 (2). 431 (4) .0036 (3) .0409 (1) .0079 (2) .0035 (4) 200 80 127 112 -.6615 (2) -1.5884 (1) 0.0564 (3) 0.5618 (4) * ' C F . OVij -0.8449 Dij 0.8636 LLALF/TALF -0.7581 Ui j -0.1615 (not s i g n i f i c a n t ) Jute/NCA 0.7135 RD/SQMCA -0.6466 Table 5.35 R-U OUT-MIGRATION STREAMS FROM NOAKHALI, 1951 "a" "b" "c" "d" D i s t r i c t s of Dest inat ion ( j th D i s t r i c t s ) No. of Migrants in the Stream Ve loc i ty o f Out-Migration Streams Di stance Separati ng i and j (in miles) Factor scores on the Subjects from the ' C r i t i c a l Factor ' Chi ttagong Dacca Dinajpur Ku s hti a OMij ~ Rank OVij ~ Rank Dij Fj* - Rank 22,362 (1) 12,819 (2) 2,525 (3) 1,174 (4) .1781 (1) .0477 (2) .0337 (3) .0242 (4) 70 82 250 156 -2.0009 (1) -1.0913 (2) 0.5614 (3) 1.003 (4) 'Factor 1 ( C F . ) OVij -0.7165 Dij 0.8134 Rice 0.6616 LLALF/TALF -0.6461 Ui j -0.1109 RD/SQMCA -0.5134 (not s i g n i f i c a n t ) Factor 2 OVij Dij Ui j RD/SQMCA TCA/NCA 0 -0 -0 0 -0 5613 3145 0135 6613 681 5 (not s i g n i f i c a n t ) (not s i g n i f i c a n t ) Table 5.36 R-U OUT-MIGRATION STREAMS FROM PABNA, 1951 Out-Migration from the i t h D i s t r i c t ~ Pabna "a" " b " "c" " d " D i s t r i c t s of Dest inat ion (jth D i s t r i c t s ) No. of Migrants in the Stream Ve loc i ty of Out-Migration Streams Di stance Separating i and j (in miles) Factor scores on the Subjects from the ' C r i t i c a l Factor ' Chi ttagong Dacca Dinajpur Kushti a OMij ~ Rank OVij ~ Rank Dij Fj* ~ Rank 903 (4) 3,320 (1) 1,647 (3) 3,039 (2). .0160 (4) .0214 (3) .0315 (2) .0902 (1) 200 65 125 50 0.8833 (4) -1.6134 (1) 0.5614 (3) -0.8339 (2) OVij -0.8558 LLALF/TALF -0.8565 Dij 0.7665 Jute/NCA 0.7534 Uij -0.1366 CO Table 5.37 R-U OUT-MIGRATION STREAMS FROM RAJSHAHI, 1951 Out-Migration from the i t h D i s t r i c t ~ Rajshahi II a" "b II "c" "d II D i s t r i c t s of Dest inat ion ( j th D i s t r i c t s ) No. of in the Mi grants Stream V e 1 o c i ty Migration of Out-Streams Di stance Separating i and j ( in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor 1 OMi j Rank OVij Rank Dij Fj Rank Chi ttagong 352 (4) .0027 (4) 265 0.5510 (4) Dacca 766 (3) .0035 (3) 120 0.1534 (3) Dinajpur 1 ,990 (1) .0272 (2) 90 -.5310 (2) Kushtia 1 ,296 (2) . .0275 (1) 60 -.8563 (1) * 'C. F. OVij -0.8396 Dij 0.7663 LLALF/TALF -0.8134 Uij -0.1334 (not s i g n i f i c a n t ) Rice/TCA 0.6314 OJ cn o Table 5.38 R-U OUT-MIGRATION STREAMS FROM RANGPUR, 1951 Out-Migration from the i t h D i s t r i c t ~ Rangpur II a" "b II "c" "d D i s t r i c t s of Dest inat ion ( j th D i s t r i c t s ) No. of in the Migrants Stream Veloci ty Mi grat i on of Out-Streams Distance Separating i and j (in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor ' OMi j Rank OVij Rank Dij * Fj Rank Chi ttagong 473 (3) .0028 (3) 288 0.6610 (4) Dacca 1 ,236 (2) .0043 (2) 168 0.0145 (2) Di najpur 10,119 (1) .1052 (1) 58 -2.1444 (1) Kushti a 194 (4) .0030 (4) 132 0.3140 (3) * ' C F . OVij -0.8310 LLALF/TALF -0.8111 Dij 0.7105 Uij -0.1212 (not s i g n i f i c a n t ) Jute/NCA 0.5614 TCA/NCA -0.6616 C O cn Table 5.39 R-U OUT-MIGRATION STREAMS FROM SYLHET, 1951 Out- Migration from the i t h D i s t r i c t ~ Sylhet II a" "b II "c" "d it Di s t r i cts of Des t i nation ( j th D i s t r i c t s ) No. of in the Migrants Stream Veloci ty Migration of Out-Streams Distance Separating i and j (in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor ' OMij Rank OVij Rank. Dij * Fj Rank Chi ttagong 1 ,685 (2) .0098 (1) 175 -0.3416 (2) Dacca 2,581 (1) .0086 (2) 100 -1.6031 (1) Dinajpur 96 (3) .0009 (4) 205 0.8134 (.3) Kushti a 85 (4) .0011 (3) 187 0.8568 (4) * ' C F . OVij -0.9134 Dij 0.8838 LLALF/TALF -0.7310 Uij -0.1565 (not s i g n i f i c a n t ) C O cn Table 5.41 U R B A N T O R U R A L O U T - M I G R A T I O N S T R E A M S , 1 9 5 1 S P A T I A L PATTERN OF O V I J FROM DACCA D i s t r i o t of Origin (ith D i s t r i o t ) ~ Daooa "a" "b" 11 c" "d" D i s t r i c t s of Dest inat ion ( j th D i s t r i c t s ) No. of Migrants in the Stream Veloc i ty of Out-Migration Streams Distance Separati ng i and j ( in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor ' OMi j Rank OVij Rank Bar isa l Bogra C h i t t . H.T Comi11 a Fari dpur Jessore Khulna Mymens i ngh Noakhali Pabna Raj shahi Rangpur Sylhet 14,905 1 ,392 309 7,864 24,486 2,925 3,027 38,125 1 ,540 7,703 12,320 10,165 11 ,098 (3) (14) (13) (7) (2) (10) (9) (1) (11) (8) (4) (6) (5) .0419 .0110 .0010 .021 2 .9252 .0176 .0149 .0676 .0068 .0409 .0572 .0356 .0372 (4) (11) (13) (8) (1) (9) (10) (2) (12) (5) (3) (7) (6) Dij Fj Rank 90 -.8477 (2) 100 .0714 (7) 110 .4772 (ID 45 -.0563 (5) 47 -2.9306 (1) 84 .0059 (6) 120 1 .384 (9) 80 -.121 5 (4) 100 .4171 (10) 65 -.2431 (3) 130 .7204 (12) 168 1 .2854 (13) 110 .1338 (8) Factor 1 ( ' C F . ' ) OVij -.8795 TCA/NCA .5536 Dij .6691 on CO Table 5.42 U R B A N T O R U R A L O U T - M I G R A T I O N S T R E A M S , 1 9 5 1 S P A T I A L P A T T E R N O F O V I J F R O M D I N A J P U R D i s t r i c t of Origin (ith D i s t r i c t ) ~ Dinajpur II , n a "b" 11 c" "d" D i s t r i c t s of Dest inat ion (jth D i s t r i c t s ) No. of Migrants in the Stream Veloc i ty of Out-Migration Streams Distance Separating i and j (in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor ' Bar isa l Bogra C h i t t . H.T. Comilla Fari dpur J essore Khulna Mymensi ngh Noakhali Pabna Rajshahi Rangpur Sylhet OMij ~ Rank OVij - Rank Dij Fj* ~ Rank 21 (11) 1,639 (2) 2 (12) 61 (9) 91 (6) 55 (11) 125 (7) 161 (5) 2 (13) 256 (4) 1,148 (3) 4,584 (1) 75 (8) .0002 (11) .0385 (2) .0001 (1 2) .0003 (10) .0008 (6) .0017 (5) .0008 ( 8 ) .0008 (7) .0001 (13) .0046 (4) .0156 (3) .0479 (1) .0007 (9) 250 80 312 21 2 195 175 237 137 250 125 90 58 205 .6323 (9) -.9967 (2) 1.4721 (1 3) -.0004 (6) .1 239 (7 ) .3337 (8) 1.2339 (1 2) -.7229 (3) 1.1645 (11) -.2449 (5) -.2661 (4) -2.091 (1) .6384 (10) *Factor 1 ( ' C F . ' ) OVij -.7427 Jute/NCA .8032 Dij .8692 co cn Table 5.43 U R B A N TO R U R A L O U T - M I G R A T I O N S T R E A M S , 1 9 5 1 S P A T I A L P A T T E R N OF O V I J FROM K U S H T I A D i s t r i o t of Origin (ith D i s t r i o t ) ~ Kushtia "a" | "b" "c" "c" D i s t r i c t s of Dest inat ion ( j th D i s t r i c t s ) No. of Migrants in the Stream Ve loc i ty of Out-Migration Streams Di stance Separating i and j ( in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor ' Bar isa l Bogra C h i t t . H.T. Comilla Faridpur Jessore Khulna Mymensi ngh Noakhali Pabna Raj shahi Rangpur Sylhet OMij ~ Rank OVij ~ Rank Dij Fj* ~ Rank 53 (10) 643 (6) 1 (13) 153 (8) 3,285 (3) 6,785 (1) 619 (7) 117 (9) 3 (12) 5,728 (2) 3,134 (4) 684 (5) 45 (11) .0006 (10) .0037 (7) .0001 (1 2) .0017 (8) .0573 (3) .0188 (4) .0141 (5) .0007 (9) .0001 (13) .1705 (1) .0667 (2) .0107 (6) .0004 (11) 1 25 82 237 137 80 45 1 05 112 156 50 60 132 187 -.4511 (7) .0156 (5) -.9150 (11) -1.2573 (13) .2710 (4) -.2170 (6) .7204 (3) -.5585 (9) -.5449 (8) 2.3570 (1) 1.3688 (2) -.7988 (10) -1.055 (12) Factor 3 ( ' C . F . ' ) OVij .8630 Rice/TCA -.8514 Dij -.5672 CO cn cn Table 5.44 URBAN TO URBAN OUT-MIGRATION STREAMS, 1951 Spat ia l Pattern of OVij from Chittagong, 1951 D i s t r i c t of Origin (ith D i s t r i c t ) ~ Chittagong "a" "b" "c" "d ii D i s t r i c t s of Dest inat ion ( j th d i s t r i c t s ) No. of in the Migrants Stream V e1o c i ty Migration of Out-Streams Di stance Separating i and j (in miles) Factor scores on the Subjects from the ' C r i t i c a l Factor ' OMij Rank OVij Rank Dij Fj* - Rank Dacca 3,872 (1) .0170 (1) 150 -1.0861 (1) Di naj pur 267 (2) .0034 (3) 330 0.8826 (3) Kushtia 200 (3) .0039 (2) 225 0.2035 (2) * ' C . F . ' (Factor 2) OVij -0.9511 Dij 0.9629 RD/SQMCA -0.9811 Uij 0.8629 CO cn cr> Table 5.45 URBAN TO URBAN OUT-MIGRATION STREAMS, 1951 Spat ia l Pattern of OVij from Dacca, 1951 H _ ii o "b" "c" "d" D i s t r i c t s of Dest inat ion ( j th d i s t r i c t s ) No. of Migrants in the Stream Ve loc i ty of Out-Migration Streams Distance Separating i and j ( in miles) Factor scores on the Subjects from the ' C r i t i c a l Factor ' Omij ~ Rank OVij ~ Rank Dij Fj* ~ Rank Chi ttagong Di naj pur Kushtia 4,704 (1) 3,902 (3) 4,503 (2) .0207 (3) .0290 (2) .0521 (1) 150 175 95 0.0660 ( 2 ) 0.9653 (3) -1.0313 (1) Factor 2 (C .F . ) OVij -0.7489 TCA/NCA -0.9852 LLALF/TALF -0.6426 Dij 0.9878 Factor 1 OVIJ RD/SQMCA Rice/TCA Ui j Di j •0.6626 •0.9633 •0.8702 0.9535 0.1553 (not s i g n i f i c a n t ) CO cn Table 5.46 URBAN TO URBAN 0UT-MI6RATI0N STREAMS, 1951 Spat ia l Pattern of OVij from Dinajpur, 1951 D i s t r i c t of Origin (ith d i s t r i c t ) ~ Dinajpur "a" "b" "c" "d II D i s t r i c t s of Dest inat ion ( j th d i s t r i c t s ) No. i n of Migrants the Stream Veloc i ty Mi grat i on of Out-Streams Di stance Separati ng i and j (in miles) Factor scores on the Subjects from the ' C r i t i c a l Factor ' OMi j ~ Rank OVij Rank Dij Fj* ~ Rank Chi ttagong 219 (2) .0028 (2) 330 -0.1351 (2) Dacca 41 2 (1) .0030 (1) 175 -0.9255 (1) Kushtia 89 (3) .0024 (3) 140 1.0607 (3) . *Factor 2 ( C F . ) OVij -0.9875 TCA/NCA -0.9768 Ui j • 0.81226 L i j 0.7516 Dij -0.2871 (not s i g n i f i c a n t ) CO cn CO Table 5.48 U R B A N T O U R B A N I N - M I G R A T I O N S T R E A M S , 1 9 5 1 S P A T I A L P A T T E R N OF I V I J TO CHITTAGONG D i s t r i c t of Destination (jth D i s t r i o t ) ~ Chittagong II _ II a "b" "c" "d" D i s t r i c t s of Or ig in ( i th D i s t r i c t s ) No. of Migrants in the Stream Veloc i ty of In-Migration Streams Distance Separati ng i and j ( in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor ' Dacca Di najpur Kushtia IMij Rank IV i j Rank Dij Fj* Rank 4,701 (1) 219 (3) 497 (2) .0203 (1 ) .0027 (3) .0094 (2) 150 330 225 -1.1130 (1) .5912 (3) .1667 (2) *Factor 1 ( ' C F . ' ) IVij -.8644 Dij .8446 RD/SQMCA .6633 Table 5.49 U R B A N T O U R B A N I N - M I G R A T I O N S T R E A M S , 1 9 5 1 S P A T I A L PATTERN OF I V I J TO DACCA D i s t r i c t of Destination (jth D i s t r i c t ) ~ Dacca n _ a a "b i "c" "d D i s t r i c t s of Or ig in ( i t h D i s t r i c t s ) No. of in the Migrants Stream Veloci ty Mi grat i on of In-Streams Distance Separati ng i and j ( in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor ' IMij Rank IV i j Rank Dij _ .* Fj Rank Chi ttagong 3,872 0 ) .0159 (1) 150 -1.235 (1) Di naj pur 412 (3) .0031 (3) 175 .8630 (3) Kushtia 819 (2) .0095 (2) 95 .1469 (2) Factor 1 ( ' C F . ' ) IV i j -.7211 Dij .6390 Jute/NCA -.6498 Uij -.5661 Table 5.50 U R B A N T O U R B A N I N - M I G R A T I O N S T R E A M S , 1 9 5 1 S P A T I A L PATTERN OF I V I J TO D I N A J P U R D i s t r i c t of Destination (jth D i s t r i c t ) ~ Dinajpur II n a " b "c" "d II Di s t r i cts of Or ig in ( i th D i s t r i c t s ) No. of in the Migrants Stream V e1o c i ty Mi grat i on of In-Streams Di stance Separati ng i and j (in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor ' IMij Rank IVi j Rank Dij Fj* - Rank Chi ttagong 267 (2) .0032 (3) 330 .5411 (3) Dacca 3,902 (1) .0297 (1) 175 -.8630 (1) Kushti a 146 (3) .0242 (2) 140 . 3669 (2) Factor 2 ( ' C F . ' ) IVij -.7400 LLALF/TALF .7678 RD/SQMCA .6614 Dij .4513 CO C T t Table 5.51 U R B A N TO U R B A N I N - M I G R A T I O N S T R E A M S , 1 9 5 1 S P A T I A L PATTERN OF I V I J TO K U S H T I A D i s t r i c t of Destination (jth D i s t r i o t ) ~ Kushtia n - II a "b i "c" "d D i s t r i c t s of Or ig in ( i th D i s t r i c t s ) No. of in the Mi grants Stream Veloc i ty Migration of In-Streams Distance Separati ng i and j (in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor ' IMij Rank IVi j Rank Dij Fj Rank Chittagong 201 (2) .0038 (2) 225 -.4963 (2) Dacca 4,503 (1) .0525 (1) 95 -1 .6319 (1) Di naj pur 89 (3) .0032 (3) 140 1.2314 (3) *Factor 1 ( ' C F . ' ) IVi j -.8266 Dij .8401 co ro Table 5.52 R - R O U T - M I G R A T I O N S T R E A M S , 1 9 6 1 S P A T I A L P A T T E R N O F O V I J F R O M B A R I S A L , 1 9 6 1 Out-Migration from the i t h .District ~ Barisal -rrg-rr D i s t r i c t s of Dest inat ion ( j th D i s t r i c t s ) ! No. of Migrants in the Stream Ve loc i ty of Out-Migration Streams Di stance Separating i and j (in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor 1 Mi j Rank OVij Rank Bog ra Comil la Di naj pur F e " i d p u r Jessore Mymensi ngh Noakhali Pabna Raj shahi Rangpur Sylhet 455 2,989 929 9,487 6,387 1 ,024 12,960 473 1 ,066 1 ,013 2,01 9 (11) (4) (9) (2) (3) (7) (1) (10) (6) (8) (5 ) ,0032 0080 0063 0350 0341 0025 0638 0027 0044 0030 0053 (8) (4) (5) (2) (3) (11) (1) (10) (7) (9) (6) Dij 7C Fj Rank 175 0.7228 (8) 88 -0.6267 (3) 250 0.9223 (10) 55 -1.8806 (1) 87 0.6976 (7) 150 1.0556 (ID 62 0.0774 (6) 125 -0.2578 (4) 175 0.8365 (9) 220 -1.5242 (2) 175 -0.0228 (5) Factor 1, ' C . F . ' OVij -0.8975 Rice/TCA 0.8631 Dij 0.7977 LLALF/TALF -0.6185 SC+LLALF/TALF -0.4977 C O C T I C O Table 5.53 R - R OUT-MIGRATION S T R E A M S , 1 9 6 1 S P A T I A L P A T T E R N O F O V I J F R O M B O G R A , 1 9 6 1 Out-Migration from the i t h D i s t r i c t ~ Bogra "a" " d " D i s t r i c t s of Dest inat ion (jth D i s t r i c t s ) No. of Migrants in the Stream Ve loc i ty of Out-migration Streams Distance Separating i and j (in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor ' Bar isa l Comi11 a Di naj pur F a i r d p u r Jessore Mymensi ngh Noakhali Pabna Raj shahi Rangpur Sylhet Mi j Rank OVij 111 (11) .0007 1 ,200 (6) .0085 6,258 (3) .11 52 298 (8) .0025 554 (7) .0040 3,256 (5) .0146 1 76 (10) .0021 3,689 (4) .0596 13.075 (2) .1476 28,882 (1) .2417 203 (9) .0014 Rank (11) (6) (3) (8) (7) (5) (9) (4) (2) (1) (10) Dij Rank 175 0.4788 (8) 120 0.6231 (9) 80 -0.7753 (3) 113 0.3918 (7) 105 0.3940 (6) 75 0.2105 (5) 175 1.1048 (10) 45 -1.9094 (1) 42 -1 .4493 (2) 50 -0.6948 (4) 1 50 1.3256 (11) "Factor 1 , ' C F . ' OVij -0.8728 RD/SQMCA -0.9342 LLALF/TALF -0.8856 Dij 0.8567 TCA/NCA -0.7085 CO Tahle 5.54 R - R O U T - M I G R A T I O N S T R E A M S , 1 9 6 1 S P A T I A L P A T T E R N OF O V I J FROM C O M I L L A , 1 9 6 1 Out-Migration from the i t h D i s t r i c t ~ Comilla II _ n a "c" "d" D i s t r i c t s of Dest inat ion ( j th D i s t r i c t s ) No. of Migrants in the Stream Ve loc i ty of Out-Migration Streams Distance Separating i and j (in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor ' Mi j Rank OVij Rank Dij Bar isa l Bogra Dinajpur Fari dpur Jessore Mymensi ngh Noakhali Pabna Raj shahi Rangpur Sylhet 212 792 284 178 190 17,046 12,738 1 ,739 12,482 555 16,440 (6) (10) (5) (7) (8) (1) (3) (9) (4) (11) (2) .0391 .0056 .0485 .0370 .0140 . 0276 .0608 .0100 .0506 .0016 .0419 ( 5 ) (10) (3) (6) (8) (7) (1) (9) (2) (11) (4) 88 1 20 212 62 113 87 45 102 102 170 80 Fj 0.1664 0.4121 1.2227 •1 .4539 0.0430 •0.4206 •1 . 9992 0.0774 0.8985 1 .0961 •0.0426 Rank (7) (8) (11) (2) ( 5 ) (3) (1) (6) (9) (10) (4) Factor II ( C F . ) OVij Rice/TCA RD/SQMCA Uij Dij -0.8410 0.7959 -0.6242 -0.5807 0.0792 (not s i g n i f i c a n t ) Table 5.55 R - R O U T - M I G R A T I O N S T R E A M S , 1 9 6 1 S P A T I A L P A T T E R N OF O V I J FROM D I N A J P U R , 1 9 6 1 Out-•Migration from the i t h D i s t r i o t ~ Dinajpur II a" "b" "c" II d" D i s t r i c t s of Dest inat ion ( j th D i s t r i c t s ) No. of in the Mi g rants Stream Ve loc i ty of Out-. Migration Streams Di stance Separating i and j (in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor ' Mi j Rank OVij Rank Dij r- •* Fj Rank Bar isa l Bogra Comi11 a Faridpur Jessore Mymens i ngh Noakhali Pabna Raj shahi Rangpur Sylhet 71 2,382 1 ,219 228 177 365 59 382 2,000 9,871 89 (10) (2) (4) (7) (8) (6) (ID (5) (3) (1) (9) .0002 .0442 .0080 .0010 .0009 .0038 .0002 .0056 .0020 .0759 .0002 (10) (2) (3) (7) (8) (4) (11) (5) (6) (1) (9) 250 80 212 195 175 137 250 125 90 58 205 0.2830 0.0852 0.1438 0.5151 0.6059 -3.8066 0.5245 0.3865 0.4024 0.0041 0.8556 (5) (3) (4) (8) (10) (1) (9) (6) (7) (2) (11) Factor 1 ( ' C F . ' ) OVij -0.9737 TCA/NCA -0.9719 LLALF/TALF -0.8041 Dij 0.7469 C O cn cn Table 5.57 R - R O U T - M I G R A T I O N S T R E A M S , 1 9 6 1 S P A T I A L P A T T E R N OF O V I J F R O M J E S S O R E , 1 9 6 1 Out--Migration from the i t h D i s t r i c t ~ Jessore a" "b" "c" II d" D i s t r i c t s of Dest inat ion ( j th D i s t r i c t s ) No. of in the Migrants Stream Ve loc i ty of Out-Migration Streams Distance Separati ng i and j (in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor ' Mi j Rank OMij Rank Dij Fj* ~ Rank Bar isa l Bogra Comil la Dinajpur Faridpur Mymens ingh Noakhali Pabna Rajshahi Rangpur Sylhet 616 482 1 ,489 658 12,246 329 326 492 873 662 1 33 (6) (8) (2) (5) (1) (9) (10) (7) (3) (4) (11) .0032 .0069 .0077 .0085 .0879 .0010 .0030 .0055 .0069 .0039 .0007 (8) (5) (3) (2) (1) (10) (9) (6) (4) (7) (11) 87 105 112 175 50 125 1 26 65 90 162 182 0.1778 -0.2295 -1.3519 0.9585 -1 .1866 0.4201 1.4897 -0.2229 0.4820 1.4445 0.9970 (5) (3) (1) (8) (2.) (6) ( I D (4) (7) (10) (9) Factor 1 ( ' C.F, OVij Rice/TCA Dij RD/SQMCA -0.8524 -0.9075 0.5958 -0.4524 CO CTi - ^ 1 Table 5.58 R-R O U T - M I G R A T I O N STREAMS/ 1 9 6 1 S P A T I A L P A T T E R N O F O V I J F R O M M Y M E N S I N G H , 1 9 6 1 Out-Migration from the ith Distriot - Mymensingh II _ II a "b" II c i . 11 d" D i s t r i c t s of Dest inat ion ( j th D i s t r i c t s ) No. of Migrants in the Stream Ve loc i ty of Out-Migration Streams Di stance Separating i and j (in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor ' Bar isa l Bog ra Comilla Di naj pur Faridpur Jessore Noakhali Pabna Rajshahi Rangpur Sylhet Mi j ~ Rank OMij ~ Rank Dij Fj* ~ Rank 1,377 (10) 3,493 (7) 6,379 (5) 31,29 2 (3) 1,402 (9) 1,458 (8) 838 C U ) 6,159 (6) 12,061 (4) 74,363 (1) 44,446 (2) •0022 (11) .0157 (6) .0101 (7) .1451 (2) .0031 (9) .0036 (8) .0023 (1 0) .0205 (5) .0410 (4) .1 834 (1 ) .0721 (3) 1 50 75 87 1 37 1 00 1 25 138 80 112 87 80 .8891 (1 1 ) -.4010 (5) .1022 (7) -1.2114 (2) .6641 (10) .2465 (8) .6134 (9) -.3688 (6) -.6614 (4) -1.6034 (1 ) -.9876 (3) *Factor 1 ( ' C F . ' ) OVij -.8615 TCA/NCA -.7990 Dij .1256 (not s i g n i f i c a n t ) Rice/TCA .6515 OJ cn co Table 5.59 R - R O U T - M I G R A T I O N S T R E A M S , 1 9 6 1 S P A T I A L P A T T E R N OF O V I J FROM N O A K H A L I , 1 9 6 1 Out-Migration from the i t h D i s t r i c t ~ Noakhali "a" D i s t r i c t s of Dest inat ion ( j th D i s t r i c t s ) No. of Migrants in the Stream Ve loc i ty of Out-Migration Streams Distance Separati ng i and j ( in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor ' Mi j Rank OMij Rank Bar isa l Bogra Comi 11 a Di najpur Faridpur Jessore Mymensi ngh Pabna Rajshahi Rangpur Sylhet 9,346 763 17,872 5,798 6,561 31 ,903 6,529 624 5,989 19,884 18,982 (5) (10) (4) (9) (6) (1) (7) (11) (8) (2) (3) .0459 .0101 .0854 .0709 .0433 .3056 .0194 .0066 .0446 .1098 .0907 (6) (10) (4) (5) (8) (1) (9) (11) (7) (2) (3) Dij Fj Rank 62 -0.5771 (8) 175 -1 .1219 (10) 45 -1 .9709 (11) 250 0.9685 (2) 75 -0.3691 (7) 125 -0.6486 (9) 138 -0.0954 (5) 137 -0.2039 (6) 185 0.8563 (3) 21 5 0.4302 (4) 130 1.4347 (1) Factor 1 ( ' C F . ' ) OVij -0.6366 Rice 0.9069 Di j 0.58200 Ui j 0.4705 Factor IV OVij RD/SQMCA TCA/NCA LLALF/TALF Dij 0.5148 0.8406 •0.7941 0.76052 •0.6489 CO CO Table 5.60 R - R O U T - M I G R A T I O N S T R E A M S , 1 9 6 1 S P A T I A L P A T T E R N OF O V I J FROM P A B N A . 1 9 6 1 Out-Migration from the i t h D i s t r i c t ~ Pabna II _ II d D i s t r i c t s of Dest inat ion ( j th D i s t r i c t s ) No. of Migrants in the Stream Mi j Bar isa l 328 Bogra 14,313 Comil la 1,415 Dinajpur 3,407 Faridpur 10,179 Jessore 2,259 Mymensi ngh 11,096 Noakhali 57 Rajshahi 32,098 Rangpur 18,506 Sylhet 577 Factor 1 ( ' C . F . 1 ) Rank (10) (3) (8) (6) (5) (7) (4) (11) (1) (2) (9) Ve loc i ty of Out-.Migration Streams OVij Rank .0019 .2319 .0081 .0507 ,0816 ,0261 0402 0004 2914 1243 0032 (10) (2) (8) (5) (4) (7) (6) (11) (1) (3) (9) "c" Di stance Separating i and j (in miles) Dij 125 45 102 125 64 65 80 137 62 100 137 Factor Scores on the Subjects from the ' C r i t i c a l Factor ' Fj 1 1 1123 1.7548 0.1548 0.4513 0962 0.1396 0.1337 0.6126 1 .0106 0.1630 1.6711 Rank (10) (1) *4) (8) (2) (6) (5) (9) (3) (7) (11) Table 5.61 R-R O U T - M I G R A T I O N S T R E A M S , 1 9 6 1 S P A T I A L PATTERN OF O V I J FROM R A J S H A H I , 1 9 6 1 Out- Migration from the i t h D i s t r i c t ~ Raj shahi II a" "b "c" II d " D i s t r i c t s of Dest inat ion ( j th D i s t r i c t s ) No. of in the Mi grants Stream Veloci ty Migration of Out-Streams Separati ng Di s tance i and j (in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor ' Mi j Rank OVij Rank Dij F j* Rank Bar isa l Bogra Comilla Di naj pur Faridpur Jessore Mymensi ngh Noakhali Pabna Rangpur Syl het' 213 15,950 1 ,222 7,166 703 911 607 51 5,365 2,347 1 56 (9) (1) • (5) (3) (7) (6) (8) (11) (2) (4) (10) .0008 .1800 .0049 .0745 .0039 .0073 .001 5 .0003 .0486 .0109 .0005 (9) (1) (6) (2) (7) (5) (8) (11) (3) (4) (10) 175 42 102 90 1 25 90 112 185 62 84 187 0.2372 -1.5464 -0.0255 -0.2387 0.4029 0.5463 2.0698 0.4765 -0.3557 0.9081 0.9431 (5) (1) (4) (3) (6) - (8) (11) (7) (2) (9) (10) Factor 1 ( 'C .F . ' ) OVij -0.8373 TCA/NCA 0.8534 Dij 0.8380 SC+LLALF/TALF -0.4530 Table 5.62 R - R O U T - M I G R A T I O N S T R E A M S . 1 9 6 1 S P A T I A L P A T T E R N OF O V I J FROM RANGPUR, 1 9 6 1 Out- Migration from the i t h D i s t r i c t ~ Rangpur a" "b II "c" d" D i s t r i c t s of Dest inat ion ( j th D i s t r i c t s ) No. of i n the Mi grants Stream Veloc i ty .Migrat ion of Out-Streams Separating Distance i and j (in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor 1 Mi j Rank OVij Rank Dij Fj* Rank Bar isa l Bogra Comil la Dinajpur Fari dpur J essore Mymens i ngh Noakhali Pabna Rajshahi Sylhet 148 6,275 1 ,399 19,022 312 406 3,685 91 734 2,114 357 (10) (2) (5) (1) (9) (7) (3) (11) (6) (4) (8) .0004 .0524 .0041 .1466 .0012 .0024 .0069 .0004 .0048 .0985 .0013 (10) (3) (6) (1) (9) (7) (4) (ID (5) (2) (8) 220 50 1 70 58 162 164 87 21 5 100 84 150 0.3840 -0.6022 1 .3481 -1 . 5060 -0.5089 0.4092 -1 .8786 0. 5623 0.4043 0.6417 0.7460 (5) (3) ( 1 1 ) (2) (4) (6) (1) (8) (7) (9) (10) Factor 1 ( ' C F . 1 ) OVij -0.8651 Dij 0.9306 Rice/TCA 0.7357 TCA/NCA -0.4560 RD/SQMCA -0.6413 CO ro Table 5.63 R-R O U T - M I G R A N T I O N S T R E A M S , 1 9 6 1 S P A T I A L PATTERN OF O V I J FROM S Y L H E T , 1 9 6 1 M -. II a Out-Migration from the i t h D i s t r i c t - Sylhet II c i. "d D i s t r i cts of Dest inat ion ( j th D i s t r i c t s ) No. of Migrants in the Stream Ve loc i ty of Out-Migration Streams Separati ng Distance i and j (in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor ' Bar isa l Bogra Comi11 a Dinajpur Faridpur Jessore Mymens i ngh Noakhali Pabna Rajshahi Rangpur Mi j Rank OVij 31 3 (7) .0008 426 (5) .0030 9,731 (2) .0252 197 (10) .001 2 329 (6) .0009 239 (9) .0010 11,354 (1) .0184 439 (4) .0020 143 (11) .0008 268 (8) .0010 555 (3) .0016 Rank Dij Fj : (10) (3) (1) (6) (9) (7) (2) (4) (ID (8) (5) 175 1 50 80 205 145 182 80 130 1 37 187 1 50 .27813 .66976 •1 .8062 0.4499 0.0J83 1.031] •1 •1 •0. 0 1 1465 1633 4191 81 68 1408 Rank (6) (8) (1) (7) ( 5 ) (10) (2) (3) (4) ( 9 ) (11) 'Factor 1 ( ' C F . 1 ) OVij -0.9487 Rice/TCA 0.8515 Dij 0.8321 LLALF/TALF -0.4032 Table 5.64 R - R I N - M I G R A T I O N S T R E A M S , 1 9 6 1 S P A T I A L P A T T E R N O F I V I J TO B A R I S A L , 1 9 6 1 D i s t r i c t of Destination II a" "b II "c" II d" D i s t r i c t s of Or ig in ( i t h D i s t r i c t s ) No. of in the Mi grants Stream V e1o c i ty Migration of Out-Streams Separati ng Distance i and j (in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor 1 IMij Rank IV i j Rank Dij Fj Rank Bogra Comil la Dinajpur Faridpur Jessore Mymensi ngh Noakhali Pabna Rajshahi Rangpur Sylhet 111 5,212 71 18,022 61 6 1 ,377 9 ,346 328 213 1 48 313 (10) (3) (11) (1) (5) (4) (2) (7) (8) ( 9 ) (6) .0018 .0141 .0005 .0675 .0033 .0023 .0467 .0020 .0009 .0004 .0011 (7) (3). (10) (1) (4) (5) (2) (6) (9) (11) (8) 175 88 250 55 87 1 50 62 125 175 220 175 0.3040 -0.2629 1 . 2076 -2.0976 0.1127 0.3261 r l .5956 0.1357 0.4856 0.9181 0.4661 (6) (3) (11) (1) (4) (7) (2) (5) (9) (10) (8) *Factor 1 ( 1 C . F . 1 ) IV i j D i j Rice/TCA RD/SQMCA LLALF/TALF -0.9502 0.8746 -0 .6654 0.6242 0.4250 - - - - -(jth D i s t r i c t ) - Barisal 0 0 Table 5.65 R - R I N - M I G R A T I O N S T R E A M S , 1 9 6 1 S P A T I A L P A T T E R N OF I V I J TO BOGRA , 1 9 6 1 D i s t r i o t of Destination "a" (jth D i s t r i o t ) ~ Bogra "d D i s t r i cts of Or ig in ( i th D i s t r i c t s ) No. of Migrants in the Stream Ve loc i ty of Out-Migration Streams Separati ng Distance i and j (in miles) Factor Scores on the Subjects from the . ' C r i t i c a l Factor 1 Bar isa l Comilla Dinajpur Faridpur Jessore Mymensi ngh Noakhali Pabna Raj shahi Rangpur Sylhet IMi j Rank 455 (10) 792 (7) 2,383 (5) 1 ,168 (6) 482 (9) 3,493 (4) 763 (8) 14,318 (2) 15,950 (1) 6,275 (3) 300 (11) IV i j Rank .0034 .0056 .0450 .0118 .0070 .0160 .0103 .2359 .1831 .0533 .0027 (10) (9) (4) (6) (8) (5) (7) (1) (2) (3) (11) Dij _ .* Fj Rank 175 1 .0776 (9) 120 0.0587 (5) 80 -0.8272 (3) 113 0.4288 (8) 105 0.1829 (6) 75 0.2804 (7) 175 1.1350 (10) 45 -1.9449 (1) 42 -0.9859 (2) 50 -0.6070 (4) 150 1.2015 (11) 'Factor 2 ( ' C . F . ' ) IV i j -0.8445 RD/SQMCA 0.9433 Dij 0.8902 LLALF/TALF 0.8220 TCA/NCA -0.7578 CO cn Table 5.66 R - R I N - M I G R A T I O N S T R E A M S , 1 9 6 1 S P A T I A L P A T T E R N OF I V I J TO C O M I L L A , 1 9 6 1 D i s t r i c t of destination " a " (jth D i s t r i c t ) - Comilla "c" "d Di s t r i cts of Or ig in ( i th D i s t r i c t s ) No. of Migrants in the Stream Ve loc i ty of Out-Migration Streams Separati ng Distance i and j (in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor ' IMij Rank IV i j Rank Dij Bar isa l Bogra Dinajpur Fari dpur Jessore Mymensi ngh Noakhali Pabna Rajshahi RangDur Sylhet 2,989 1 ,200 1 ,219 10,407 1 ,489 6,379 17,872 1 ,41 5 1 ,222 1 ,399 9,731 (5) (11) (10) (2) (7) (4) (1) (6) (9) (8) (3) .0081 .0088 .0082 .0378 .0078 .0105 .0868 .0083 .0050 .0043 .0322 (8) (5) (7) (2) (9) (4) (1) (6) (10) (11) (3) 88 1 20 21 2 62 113 87 45 102 102 170 80 Fj" 0.1664 0.4121 1 . 2227 •1 .4539 0.0430 •0.4206 •1 . 9992 0.0774 0.8985 1.0961 •0.0426 Rank (7) (8) (11) (2) (6) (3) (1) (5) (9) (10) (4) Factor 1 ( ' C F . ' ) IV i j -0.8410 Dij 0.8286 LLALF/TALF 0.4483 Table 5.67 R - R I N - M I G R A T I O N S T R E A M S , 1 9 6 1 S P A T I A L P A T T E R N OF I V I J TO D I N A J P U R , 1 9 6 1 D i s t r i o t of destination (jth D i s t r i o t ) - Dinajpur "a "b I I "c" d" Di s t r i cts of Or ig in ( i t h D i s t r i c t s ) No. of Migrants in the Stream V e1o c i ty Migration of Out-Streams Separati ng Distance i and j (in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor ' IMij Rank IV i j ' Rank Dij Rank Bar isa l Bogra Comi11 a Faridpur Jessore Mymens i ngh Noakhali Pabna Raj shahi Rangpur Sylhet 929 6,258 7 ,284 2,014 658 31 ,292 5,798 3,407 7,146 19,022 197 (9) (5) (3) (8) (10) (1) (6) (7) (4) (2) (11) .0073 .1181 .0493 .0188 .0089 .1 324 .0722 .0516 . 0755 .1488 .0017 (10) (3) (7) (8) (9) (2) (5) (6) (4) (1) (11) 250 80 21 2 195 175 1 37 250 1 25 90 58 205 -0.4392 0.5169 0.7896 0 .4000 1 .1957 2.0584 -0.3988 -0.8105 -0.6469 0.8160 -1.089 (7) (4) (3) (5) ( 1 1 ) ( 1 ) ( 6 ) (9) (8) (2) (10) Factor 1 *Factor IV ( ' C F . 1 ) IV i j RD/SQMCA LLALF/TALF TCA/NCA Dij -0.4900 0.9286 0.7455 +0.7 371 0.8339 IVi j Jute/NCA EB Rice/TCA/NCA Dij 0.6984 -0.8308 0.8221 --.8856 -0.2742 C O Table 5.68 R - R I N - M I G R A T I O N S T R E A M S , 1 9 6 1 S P A T I A L P A T T E R N O F I V I J TO F A R I D P U R , 1 9 6 1 D i s t r i c t of Destination II _ n d (jth D i s t r i c t ) ~ Faridpur "d D i s t r i c t s of Or ig in ( i th D i s t r i c t s ) No. of Migrants in the Stream Ve loc i ty of Out-Migration Streams Separati ng Di stance i and j (in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor 1 Bari sal Bogra Comil la Dinajpur Jessore Mymensi ngh Noakhali Pabna Raj shahi Rangpur Sylhet IMij Rank IV i j 9,487 (4) .0356 554 (8) .0056 15,910 (1) .0579 228 (11) .0021 12,246 (2) .0891 1 ,402 (6) .0032 6,561 (5) .0439 10,179 (3) .0830 703 (7) .0040 31 2 (10) .0013 329 (9) .0015 Rank Dij (5) (6) (3) (9) (1) (8) (4) (2) (7) (11) (10) 55 113 62 195 50 100 75 64 125 162 145 Fj •0.3815 0.5517 •0.9109 1 . 5330 •1 .4231 0.1666 •0.6207 •1 .3563 0.5603 1 .0071 0.8737 Rank (5) (7) (3) (11) (1) (6) (4) (2) ( 8 ) (10) (9) Factor 1 ' C F IV i j Dij Rice/TCA •0.9319 0.9426 •0.53225 Table 5.7.0 R-R I N - M I G R A T I O N S T R E A M S , 1 9 6 1 S P A T I A L P ATTERN OF I V I J TO M YMENSINGH, 1 9 6 1 D i s t r i c t of Destination (jth D i s t r i c t ) - Mymensingh II a" "b II "c" II d" D i s t r i c t s of Or ig in ( i th D i s t r i c t s ) No. of in the Migrants Stream V e1o c i ty Migration of Out-Streams Separati ng D i s tance i and j (in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor ' IMij Rank IVi j • Rank Dij Fj* Rank Bar isa l Bogra Comi11 a Dinajpur Fari dpur J essore Noakhali Pabna Rajshahi Rangpur Sylhet 1 ,531 3,256 17,046 365 2,322 329 6,529 11 ,096 607 3,685 11,354 (8) (6) (1) (10) (7) (11) (4) (3) '9) (5) (2) .0026 .0149 .0281 .001 5 .0053 .0010 .0197 .0409 .001 5 .0070 .0235 (8) (5) (2) (10) (7) (ID (4) (1) (9) (6) (3) 1 50 75 87 137 100 125 138 80 112 87 80 0.4470 -0.3614 -0.9926 0.7863 0.4784 1 .3573 -0.1190 -2.0982 0.7595 0.5666 -0.8239 (6) (4) (2) (10) (7) (11) (5) (1) (9) (8) (3) *Factor 2 ( ' C F . ' ) IVi j -0.9714 Dij 0.5996 Rice/TCA -0.4035 Table 5.71 R-R I N - M I G R A T I O N S T R E A M S , 1 9 6 1 S P A T I A L P A T T E R N OF I V I J TO N O A K H A L I , 1 9 6 1 D i s t r i c t of Destination (jth D i s t r i c t ) ~ Noakhali a" "b H c n- II d " Di s t r i cts of Or ig in ( i t h D i s t r i c t s ) No. of in the Mi grants Stream Veloc i ty Migration of Out-Streams Separati ng Distance i and j (in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor ' IMij Rank IV i j ' Rank Dij Fj Rank Bar isa l Bogra Comil la Di naj pur F a i r d p u r Jessore. Mymens i ngh Pabna Raj shahi Rangpur Sylhet 12,960 176 12,783 59 1 ,046 326 838 57 51 91 439 (1) (7) (2) (9) (3) (6) (4) (10) (11) (8) (5) .0648 .0024 .0618 .0007 .0070 .0031 .0025 .0006 .0004 .0005 .0026 (1) (7) (2) (8) (3) (4) (6) (9) (11) (10) (5) 62 175 45 250 75 125 138 137 185 21 5 130 -0.5771 -1.1219 -1 .970 0.9685 -0.3691 0.6486 -0.0954 -0.2039 0.8563 0.4302 1.4347 (3) (2) (1) (10) ( 4 ) (8) (6) (5) (9) (7) (11) *Factor 1 ( ' C . F . ' ) Factor 4 IV i j Rice Di j U i j -0.6366 -0.9062 0.5820 -0.4705 IVi j RD/SQMCA TCA/NCA LLALF/TALF Dij 0.5148 -0.8406 -0.7941 -0.7605 -0.6489 Table 5.72 R-R I N - M I G R A T I O N S T R E A M S . 1 9 6 1 S P A T I A L P A T T E R N OF I V I J TO P A B N A , 1 9 6 1 II . H a D i s t r i c t of Destination (jth D i s t r i c t ) ~ Pabna " d D i s t r i c t s of Or ig in ( i t h D i s t r i c t s ) No. of Migrants in the Stream Ve loc i ty of Out-Migration Streams Separati ng Distance i and j (in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor 1 Bar isa l Bogra Comi11 a Di najpur Pari dpur Jessore Mymensi ngh Noakhali Raj shahi Rangpur Sylhet IMij Rank IV i j 473 (8) .0029 3,689 (4) .0607 143 (10) .0008 382 (9) .0057 4,745 (3) .0387 492 (7) .0006 6,1 59 (1) .0227 1 ,739 (5) .0189 5,365 (2) .0494 734 (6) .0050 143 (ID .0010 . Rank (8) (1) (10) (6) (3) (ID (4) (5) (2) (7) (9) Dij Rank 125 1 .1123 (10) 45 -1.7548 (1) 102 -0.1548 (4) 125 0.4513 ( 8 ) 64 -1 .096 (3) 65 0.1396 (6) 80 -0.1337 (5) 137 0.6126 (9) 62 -1.0106 (2) 100 0.1630 (7) 137 1 .6711 (11) Factor 1 ( ' C F . ' ) IV i j -0.8407 TCA/NCA -0.8791 Dij 0.8877 CO CO Table 5.73 R-R I N - M I G R A T I O N S T R E A M S , 1 9 6 1 S P A T I A L PATTERN OF I V I J TO R A J S H A H I , 1 9 6 1 D i s t r i o t of Destination (jth. D i s t r i o t ) ~ Rajshahi II a" "b "c" "d D i s t r i c t s of Or ig in ( i th D i s t r i c t s ) No. of i n the Migrants Stream Veloci ty Mi grat i on of Out-Streams Separati ng Distance i and j ( in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor ' IMij Rank IV i j •Rank Dij Fj Rank Bar isa l Bogra Comi11 a Di najpur Faridpur Jessore Mymens i ngh Noakhali Pabna Rangpur Sylhet 1 ,066 13,075 12,482 1 ,999 2,511 873 12,061 5,989 32,098 2,114 268 (9) (2) (3) (8) (6) (10) (4) (5) (1) (7) (11) .0045 .1501 .0513 .0211 .0042 .0071 .0310 .0454 .2961 .0001 .0054 (10) (2) (3) (6) (7) (8) (5) (4) (1) (11) (9) 175 42 102 90 125 90 112 185 62 84 187 -1.9859 -0.9517 0.9433 -1.092 1.2374 0.9802 -1.9236 2.3805 C.2457 0.1564 0.1664 (11) (8) (4) (9) (2) (3) (10) (1) (5) (6) (7) Factor 1 Factor 3 *Factor 4 ( 1 C F . ' ) IV i j TCA/NCA SC+LLALF/TALF Dij L i j -0.5228 -0.9402 +0.8715 0.8216 -0.5803 IV i j 0.5003 SC/TALF -0.8959 L i j +0.5806 Dij -0.6968 IV i j SB . Uij Dij 0.5930 0.7070 0.6135 -0.0424 (not s i g n i f i c a n CO CO ro Table 5.74 R - R " I N - M I G R A T I O N S T R E A M S , 1 9 6 1 S P A T I A L P A T T E R N OF I V I J TO RANGPUR, 1 9 6 1 II . II a D i s t r i c t of Destination (,jth D i s t r i c t ) ~ Rangpur "b" "c" M d l , D i s t r i c t s of Or ig in ( i t h D i s t r i c t s ) No. of Migrants in the Stream Ve loc i ty of Out-Migration Streams Separati ng Di stance i and j (in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor ' IMij Rank IV i j Rank Dij Bar isa l Bogra Comi11 a Di naj pur Faridpur Jessore Mymensi ngh Noakhali Pabna Rajshahi Sylhet 1 ,013 28,882 4,683 9,871 1 ,865 662 74,363 19,884 18,506 2,347 555 (9) (2) (6) (5) (8) (10) (1) (3) (4) (7) (11) .0031 .2455 .0142 -.0772 .0078 .0040 .1417 .1116 .1264 .0111 .0021 (10) (1) (6) (5) (8) (9) (2) (4) (3) (7) (11) 220 50 170 58 162 164 87 21 5 100 84 150 Fj •0.5904 2.0510 0.0903 •0.3766 •0.2617 •0.6479 0.0969 1 .3869 0.4895 •1 .2957 •0.9423 Rank (8) (1) (5) (7) (6) (9) (4) (2) (3) (ID (10) 1 ( ' C . F . ' ) I V i j LLALF/TALF SC/TALF Dij Jute 0.8607 -0.6436 -0.4892 -0.1256 (not s i g n i f i c a n t ) 0.4323 CO CO CO Table 5.75 R-R I N - M I G R A T I O N S T R E A M S , 1 9 6 1 S P A T I A L P A T T E R N OF I V I J TO S Y L H E T , 1 9 6 1 D i s t r i o t of Destination n - n a (jth D i s t r i o t ) ~ Sylhet "c" Di s t r i cts of Or ig in ( i th D i s t r i c t s ) No. of Migrants in the Stream Ve loc i ty of Out-Migration Streams Separati ng Distance i and j (in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor ' IMij Rank IV i j Rank Bar isa l Bogra Comi11 a Dinajpur Faridpur Jessore Mymens i ngh Noakhali Pabna Raj shahi Rangpur 2,019 203 43,574 89 1 ,195 133 44,446 18,982 577 156 357 (4) (8) (2) (11) (5) (9) (1) (3) (6) (10) (7) .0069 .0019 .1445 .0007 .0054 .0009 .0921 .1159 .0428 .0008 .0013 (5) (7) (1) (10) (6) (9) (3) (2) (4) (11) (8) Dij Rank 175 0.2744 (6) 150 0.6611 (8) 80 -1.8074 (1) 205 0.5616 (7) 145 0.0208 (5) 182 1.0397 (10) 80 -1 .1399 (3) 130 -1 .1605 (2) 137 -0.4205 (4) 187 0.8153 (9) 1 50 1.1554 (11) Factor 1 ( ' C F . 1 ) IV i j -0.9475 Rice -0.8546 Dij 0.8309 LLALF/TALF 0.4047 CO co -p. Table 5.76 R-U OUT-MIGRATION STREAMS FROM BARISAL. 1961 Out-Migration from the i t h D i s t r i c t ~ Barisal a" "b 11 c" "d" D i s t r i c t s of Dest inat ion ( j th D i s t r i c t s ) No. of in the Migrants Stream Veloci ty Migration of Out-Streams Distance Separating i and j ( in miles) Factor scores Subjects from ' C r i t i c a l F a c on the the , j r' OMij Rank OVij Rank • Dij Rank Chi ttagong 4,642 (3) .0182 (3) 112 0.2236 (4) Chittagong H.T. 6 26 (5) .0019 (5) 132 -2.0078 (5) Dacca 17,430 (2) .0401 (2) 90 0.4261 (2) Khulna 42,635 (1) .2042 (1) 50 0.2138 (3) Kushtia 629 (4) .0062 (4) 125 0.7075 (1) C . F . ' N.B. OVij 0.9070 Level of s i g n i f i c a n c e > .4000. RD/SQMCA 0.9808 Only var iab les which are s i g n i f i c a n t SC+LLALF/TALF 0.8213 are mentioned here (except Di j and U i j ) . Uij -0.5697 Dij -0.0436 (not s i g n i f i c a n t ) co CO cn Table 5.77 R-U O U T - M I G R A T I O N S T R E A M S F R O M BOGRA, 1 9 5 1 Out-Migration from the i t h D i s t r i c t ~ Bogra II n a 11 b" "c" "d" D i s t r i c t s of Dest inat ion (jth D i s t r i c t s ) No. of Migrants in the Stream Veloc i ty of Out-Migration Streams Distance Separati ng i and j ( in miles) Factor scores on the Subjects from the ' C r i t i c a l Factor ' Chittagong Chittagong H.T. Dacca Khulna Khus t i a OMij ~ Rank OVij ~ Rank Dij Fj* - Rank 606 (2) 115 (5) 2,217 (1) 496 (3) 430 (4) .0063 (4) .0097 (2) .0135 (1) .0062 (5) .0064 (3) 250 265 100 175 82 1.2032 (5) -0.7914 (2) -0.0868 (3) 0.8033 (4) -1.1 282 (1 ) * ' C F . OVij -0.9570 Rice 0.9563 TCA/NCA -0.90131 Uij 0.9557 Dij 0.3898 (not s i g n i f i c a n t ) RD/SQMCA -0.6915 CO oo cn Table 5.78 R-U OUT-MIGRATION STREAMS FROM COMILLA. 1961 Out-Migration from the i t h d i s t r i c t - Comilla "a" "b" "c" "d" Di s t r i cts of Dest inat ion ( j th D i s t r i c t s ) No. of Migrants in the Stream Ve loc i ty of Out-Migration Streams Distance Separati ng i and j (in miles) Factor scores on the Subjects from the ' C r i t i c a l Factor ' Chittagong Chittagong H.T. Dacca Khulna Kushtia OMij - Rank OVij - Rank Dij Fj* ~ Rank 29,416 (2) 3,494 (5) 71,117 (1) 5,054 (4)' 5,893 (3) 1.1238 (2) .1030 (3) 1.5906 (1) .0234 (5) .0574 (4) VI 2 130 45 125 137 0.2426 (4) 0.4018 (3) -1.7795 (1) 0.5123 (4) 0.6227 (5) ' C F . OVij -0.8915 Jute/NCA 0.9585 SC+LLALF/TALF -0.8698 SC/TALF -0.4561 Dij -0.0445 (not s i g n i f i c a n t ) Ui j -0.1952 (not s i g n i f i c a n t ) OJ 0 0 Table 5.79 R-U OUT-MIGRATION STREAMS FROM DINAJPUR, 1961 Out-Migration from the i t h D i s t r i c t ~ Dinajpur a" "b "c" "d II D i s t r i c t s of Dest inat ion ( j th D i s t r i c t s ) No. of in the Migrants Stream V e 1 o c i ty Migration of Out-Streams Distance Separati ng i and j (in miles) Factor scores on the Subjects from the ' C r i t i c a l Factor ' OMi j Rank OVij Rank Dij Fj* ~ Rank Chi ttagong 471 (2) .0035 (3) 295 0.6307 (4) Chittagong H.T. 25 (5) .0002 (5) 320 0.7723 (3) Dacca 1 ,382 (1) .0079 (1) 1 50 -2.0644 (1) Khulna 322 (3) . .0028 (4) 237 0.6616 (5) Kus hti a 222 (4) .0045 (2) 200 -0.0002 (.2) * ' C F . OVij -0.9957 1 6 -0.9908 L i j 0.9842 Dij 0.8785 Ui j 0.0009 (not s i g n i f i c a n t ) CO CO CO Table 5.80 R-U O U T - M I G R A T I O N STREAMS FROM F A R I D P U R , 1 9 6 1 Out-Migration from the i t h D i s t r i c t ~ Faridpur "a" "b" "c" "d" D i s t r i c t s of Dest inat ion ( j th D i s t r i c t s ) No. of Migrants in the Stream Ve loc i ty of Out-Migration Streams Distance Separati ng i and j ( in miles) Factor scores on the Subjects from the ' C r i t i c a l Factor ' Chi ttagong Chittagong H.T. Dacca Khulna Kushtia OMij ~ Rank OVij - Rank Dij Fj* ~ Rank 2,804 (4) 400 (5) 48,188 (1) 8,949 (2) 6,048 (3) .0147 (5) .0163 (4) .1487 (1) .0574 (3) .0815 (2) 1 50 175 47 75 80 0.8049 (4) 1.2267 (5) -1.2133 (1) -0.2675 (3) -0.5507 :2) * ' C F . OVij -0.9112 Dij 0.9015 13 ,-0.8545 Jute , 0.7588 Ui j 0.08201 (not s i g n i f i c a n t ) CO co to Table 5.81 R-U OUT-MIGRATION STREAMS FROM JESSORE. 1961 Out-Migration from the i t h D i s t r i c t ~ Jessore a" "b I I "c" "d D i s t r i c t s of Dest inat ion ( j th D i s t r i c t s ) No. of in the Mi grants Stream Veloci ty Mi grat i on of Out-Streams Distance Separating i and j (in miles) Factor scores on the Subjects from the ' C r i t i c a l Factor ' OMij Rank OVij Rank Dij Fj Rank Chi ttagong 524 (4) .0039 (4) 185 0.4932 (4) Chittagong H.T. 45 (5) .0023 (5) 205 1 .4508 (5) Dacca 2,727 (3) .0121 (3) 84 -0.1505 (3) Khulna 19,864 (1) .1852 (1) 62 -0.7949 (2) Kushtia 7,230 (2) .1414 (2) 45 -0.9985 (D 'C . F . OVij -0.8592 Dij 0.9073 Jute/NCA 0.8911 SC+LLALF/TALF -0.8403 SC/TALF -0.5220 L i j 0.6613 Uij -0.0134 (not s i g n i f i c a n t ) CO CD O Table 5.82 R-U O U T - M I G R A T I O N STREAMS FROM MYMENS INGH . 1 9 6 1 "a" "b" "c" "d" D i s t r i c t s of Dest inat ion ( j th D i s t r i c t s ) No. of Migrants in the Stream Ve loc i ty of Out-Migration Streams Distance Separati ng i and j (in miles) Factor scores on the Subjects from the ' C r i t i c a l Factor ' Chi ttagong Chittagong H.T. Dacca Khulna Kushti a OMij - Rank OVij ~ Rank Dij Fj* - Rank 2,576 (2) 325 (5) 41.907 (1) 1,501 • ' (3) 781 (4) .0061 (2) .0059 (3) .0585 (1) .0043 ' (5) .0047 (4) 200 225 80 170 112 0.5890 (4) 1.2158 (5) -1.3790 (1) 0.0887 (3) -0.5146 (2) Factor 1 ( C r i t i c a l Factor) OVij -0.7684 Dij 0.9871 SC+LLALF/TALF -0.8878 Eg -0.9204 Uij 0.0009 Jute/NCA 0.6468 LLALF/TALF -0.7785 (not s i g n i f i c a n t ) Factor 3 OVij RD/SQMCA SC/TALF L i j Uij Dij 0.6377 •0.6028 0.9*69 0.9808 •0.2250 0.0607 } not s i g n i f i c a n t CO 0 Table 5.83 R - U O U T - M I G R A T I O N S T R E A M S F R O M N O A K H A L I , 1 9 5 1 Out-Migration from the i t h D i s t r i c t ~ Noakhali a" "b "c" "d D i s t r i c t s of Dest inat ion ( j th D i s t r i c t s ) No. of in the Migrants Stream Veloci ty Migration of Out-Streams Distance Separating i and j (in miles) Factor scores on the Subjects from the ' C r i t i c a l Factor ' OMij Rank OVij Rank Dij Fj Rank Chi ttagong 47,598 (1) .3347 (2) 70 -1 . 7325 (1) Chittagong H.T. 8,219 (4) .4471 (1) 90 0.5604 (4) Dacca 44,056 (2) .1814 (3) 100 0.0038 (2) Khulna 8,896 (3) .0761 (4) n o 0.5561 (3) Kushtia 3,295 (5) .0572 (5) 156 0.6121 (5) * ' C F . OVij -0.9952 Rice/TCA 0.8648 Dij 0.67 31 RD/SQMCA -0.6520 TCA/NCA -0.5294 U i j 0.2198 (not s i g n i f i c a n t ) LLALF/TALF' -0.7131 CO CD ro Table 5.85 R - U O U T - M I G R A T I O N S T R E A M S F R O M R A J S H A H I , 1 9 6 1 Out-Migration from the i t h D i s t r i c t ~ Rajshahi a" "b I I "c" "d II D i s t r i c t s of Dest inat ion ( j th D i s t r i c t s ) No. of in the Mi grants Stream V e 1 o c i ty Migration of Out-Streams Di stance Separati ng i and j (in mi 1ps) Factor Scores on the Subjects from the ' C r i t i c a l Factor ' OMij Rank OVij Rank Dij Fj Rank Chittagong ' 587 (4) .0034 (4) 265 0.5415 (4) Chittagong H.T. 39 (5) .0009 (5) 285 1.3921 (5) Dacca 2,479 (2) .0085 (2) 130 -1.0150 (1) Khulna 762 (3) .0055 (3) 1 58 -0.0615 (3) Kushtia 2,493 (1) .0454 (1) 60 -0.8570 (2) Factor 1 ( C r i t i c a l Factor) OVij -0.8189 Dij 0.9195 SC+LLALF/TALF -0.9105 LLALF/TALF -0.6402 Jute/NCA 0.7681 Uij -0.0711 (not s i g n i f i c a n t ) Factor 2 OVij 0.5102 Uij -0.9513 Rice/TCA-0.9196 TCA/NCA -0.9458 Dij -0.2298 (not s i g n i f i c a n t ) CO <£> CO Table 5.86 R-U O U T - M I G R A T I O N STREAMS FROM RANGPUR. 1 9 6 1 Out -Migration from the i t h D i s t r i o t - Rangpur II a" "b "c" "d Di s t r i c t s of Dest inat ion ( j th D i s t r i c t s ) No. of in the Migrants Stream V e 1 o c i ty Migration of Out-Streams Distance Separati ng i and j (in mi 1es ) Factor Scores on the Subjects from the ' C r i t i c a l Factor ' OMij Rank OVij Rank Dij Fj* Rank Chi ttagong 1 ,279 (2) .0056 (2) 288 0.7000 (4) Chittagong H.T. 63 (5) .0021 (5) 300 1 .0452 (5) Dacca 2,974 (1) .0077 (1) 168 -1 .4405 (1) Khulna 664 (3) .0036 (3) 225 0.2321 (3) Kushtia 222 (4) .0025 (4) 132 -0.5368 (2) * ' C F . OVij -0.8834 Dij 0.8687 IB -0.9510 LLALF/TALF -0.8467 SC+LLALF/TALF -0.8250 Uij -0.0873 (not s i g n i f i c a n t ) CO 4^ Table 5.87 R-U 0 U T - M I 6 R A T I 0 N STREAMS FROM S Y L H E T , 1 9 6 1 Out-Migration from the i t h D i s t r i c t ~ Sylhet II _ I I a " c " D i s t r i c t s of Dest inat ion ( j th D i s t r i c t s ) No. of Migrants in the Stream Ve loc i ty of Out-Migration Streams Distance Separating i and j (in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor ' Chittagong Chittagong H.T, Dacca Khulna Kushtia OMij Rank 3,113 (2) 1 ,310 (3) 6,769 (1) 465 (4). 206 (5). OVij Rank Dij .0118 .0384 .01 51 .0038 .0019 (3) (1) (2) (4) (5) 175 185 110 21 2 187 Fj 0.2946 0.4360 •1 .7670 0.6932 0.3438 Rank (2) (4) (1) (5) (3) ' C F . ' OVij Dij IB LLALF/TALF Uij -0.9707 0.9758 -0.9342 -0.8775 0.1254 (not s i g n i f i c a n t ) C O C D cn Table 5.88 URBAN TO RURAL O U T - M I G R A T I O N S T R E A M S , 1 9 6 1 S P A T I A L PATTERN OF O V I J FROM CHITTAGONG D i s t r i c t of Origin (ith D i s t r i c t ) ~ Chittagong "a" "b" "c" 11 d" D i s t r i c t s of Dest inat ion ( j th D i s t r i c t s ) No. of Migrants in the Stream Veloc i ty of Out-Migration Streams Distance Separating i and j (in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor ' Bar isa l Bogra Comi11 a Di naj pur Faridpur Jessore Mymensi ngh Noakhali Pabna Raj shahi Rangpur Sylhet OMij - Rank OVij ~ Rank Dij Fj* ~ Rank 1,494 (5) 426 (11) 5,373 (2) 446 ( 9 ) 565 (8) 31,903 (1) 1,364 (6) 3,681 (3) 216 (12) 445 (10) 1,170 (7) 3,455 (4) .0044 (8) .0057 (7) .0248 (2) .0043 (9) .0035 (11) .3100 (1) .0058 (5) .0117 (3) .0026 (12) .0057 (6) .0039 (10) .0114 (4) 112 250 • 112 295 150 185 200 70 200 265 288 175 -.8100 (3) .1360 (7) -1.3657 (2) 1.5073 (12) -.6199 (4) .3869 (8) -.0515 (5) -1.6859 (1 ) 1.0840 (11) 1.0456 (10) .6966 (9) -.0312 (6) *Factor 1 ( ' C F . ' ) OVij -.6785 RD/SQMCA -.7875 Rice/TCA .8182 LLALF/TALF -.7998 Dij .8781 co co cn Table 5.89 U R B A N T O R U R A L O U T - M I G R A T I O N S T R E A M S , 1 9 6 1 S P A T I A L P A T T E R N OF O V I J FROM CHITTAGONG H . T . D i s t r i o t of Origin (ith D i s t r i o t ) - Chittagong H.T. II n a " b " "c" " d " D i s t r i c t s of Desti nation ( j th D i s t r i c t s ) No. of Migrants in the Stream Veloc i ty of Out-Migration Streams Distance Separati ng i and j ( in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor ' Bar isa l Bogra Comi11 a Dinajpur Faridpur J essore Mymensi ngh Noakhali Pabna Raj shahi Rangpur Sylhet OMij ~ Rank OVij ~ Rank Dij Fj* - Rank 25 (5) 196 (2) 1,108 (1) 13 (7) 10 (8) 6 (9) 43 (4) 21 (6) 5 (10) 5 (11) 1 (12) 69 (3) .0007 (7) .0164 (2) .0333 (1 ) .0010 (5) .0004 (8) .0003 (9) .0008 (6) .0011 (4) .0003 (10) .0002 (11) .0001 (12) .0026 (3) 132 265 130 320 175 205 225 90 225 285 300 185 -.4211 (5) .5124 (8) -.4562 (4) 1 .4332 (1 2) -.6203 (3) .1033 (7) -.1858 (6) -2.2755 (1) .8467 (10) .6863 (9) 1.0340 (1 1 ) -.6569 (2) Factor 1 ( ' C F . ' ) OVij -.7881 RD/SQMCA -.8805 Rice/TCA .6112 LLALF/TALF -.6940 Dij .9015 0 0 vo Table 5.91 URBAN TO RURAL O U T - M I G R A T I O N S T R E A M S , 1 9 6 1 S P A T I A L PATTERN OF O V I J FROM KHULNA "a" 11 b" "c" "d" D i s t r i c t s of Desti nation ( j th D i s t r i c t s ) No. of Migrants in the Stream Veloc i ty of Out-Migration Streams Distance Separating i and j ( in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor ' Bar isa l Bogra Comi11 a Di naj pur Faridpur Jessore Mymensi ngh Noakhali Pabna Raj shahi Rangpur Sylhet OMij ~ Rank OVij ~ Rank Dij Fj* ~ Rank 4,118 (2) 246 (10) 1,571 (4) 612 (5) 2,805 (3) 36,018 (1) 222 (11) 134 (12) 248 (9) 478 (7) 396 (8) 494 (6) .0196 (2) .0031 (7) .0072 (5) .0079 (4) .0157 (3) .3359 (1) .0006 (12) .0010 (11) .0025 (8) .0033 (6) .0020 (10) .0023 (9) 50 175 1 25 237 75 62 170 110 125 158 225 212 1.013 (1) -.4139 (6) -.2356 (5) .6636 (3) -.1031 (4) .8539 (2) -2.1001 (12) -1.2439 (9) -.8391 (8) -.6831 (7) -1.4466 (10) -1.8936 (11) Factor 2 ( ' C F . ' ) OVij .9496 E6 .8740 Dij .7918 C O CO oo Table 5.92 U R B A N T O R U R A L O U T - M I G R A T I O N S T R E A M S , 1 9 6 1 S P A T I A L P A T T E R N OF O V I J FROM K A S H T I A D i s t r i c t of Origin (ith D i s t r i c t ) - Kushtia II . n Q II B . . "c" "d" D i s t r i c t s of Desti nation ( j th D i s t r i c t s ) No. of Migrants in the Stream Veloci ty of Out-Migration Streams Distance Separati ng i and j ( in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor ' Bar isa l Bogra Comi11 a Dinajpur Faridpur Jessore Mymensi ngh Noakhali Pabna Rajshahi Rangpur Sylhet OMij ~ Rank OVij ~ Rank Dij Fj* ~ Rank 167 (11) 799 (8) 1,485 (5) 822 (7) 5,097 (2) 15,932 (1) 235 (10) 26 (12) 3,742 (4) 4,947 (3) 825 (6) 510 (9) .0016 (10) .0212 (5) .0144 (7) .0205 (6) .0685 (4) .3117 (1) .0013 (11) .0001 (12) .0818 (2) .0753 (3) .0092 (8) .0048 (9) 125 82 137 200 80 45 112 156 50 60 132 187 .2139 (7) -.2246 (5) -.1254 (6) 1 .7399 (1 2) -.8956 (3) -1.5318 (1) .8568 (10) .4233 (8) -1.4973 (2) -.4987 (4) .5215 (9) 1.1017 (11) Factor 4 ( ' C F . ' ) OVij -.8666 Dij .9052 CO VO VO Table 5.94 URBAN TO URBAN OUT-MIGRATION STREAM Spat ia l Pattern of OVij from Chittagong H.T. , 1961 D i s t r i c t of Origin (ith D i s t r i c t ) ~ Chittagong H.T. I I a" "b" "c" "d II Di s t r i c t s of Dest inat ion ( j th d i s t r i c t s ) No. of in the Migrants Stream V e 1 o c i ty Migration of Out-Streams Di stance Separati ng i and j (in miles) Factor scores on the Subjects from the ' C r i t i c a l Factor ' Mi j ~ Rank OVij Rank Dij Fj* - Rank Chi ttagong 670 (1) .0296 (1) 35 -1 .1427 (1) Dacca 138 (2) .0003 (3) 185 -0.1234 (2) Khulna 6 (3) .0002 (4) 195 1 .2953 (4) Kushti a 11 (4) .0011 (2) 260 -0.0292 (3) *Factor 3 ( C F . ) OVij -0.7711 TCA/NCA 0.9610 SC/TALF -0.8398 Dij 0.6380 IB -0.6292 Table 5.95 URBAN TO URBAN OUT-MIGRATION STREAM Spat ia l Pattern of OVij from Dacca, 1961 D i s t r i o t of Origin (ith D i s t r i c t ) ~ Dacca a" "b" "c" "d D i s t r i c t s of Dest inat ion ( j th d i s t r i c t s ) No. of in the Migrants Stream V e 1 o c i ty Migration of Out-Streams Distance Separating i and j ( in miles) Factor scores on the Subjects from the ' C r i t i c a l Factor ' OMij ~ Rank OVij Rank Dij Rank Chi ttagong 8,433 (2) .0277 (4) 150 -2.6274 (4) Chittagong H.T. 1 .395 (4) .0351 (3) 185 0.9052 (2) Khulna 8,949 (1) .0358 (2) 190 0.6057 (3) Kushtia 7,353 (3) .0859 (1) 95 1 .1165 (1) Factor 2 (C . F.) OVij 0.9928 RD/SQMCA 0.9631 Table 5.96 U R B A N T O U R B A N O U T - M I G R A T I O N S T R E A M S , 1 9 6 1 Spat ia l Pattern of OVij from Khulna, 1961 D i s t r i c t of Origin (ith D i s t r i c t ) " Khulna H. n Q "b" "c" "d" D i s t r i c t s of Dest inat ion ( j th d i s t r i c t s ) No. of Migrants in the Stream Ve loc i ty of Out-Migration Streams Di stance Separati ng i and j (in miles) Factor scores on the Subjects from the ' C r i t i c a l Factor ' OMij ~ Rank OVij ~ Rank Dij Fj* ~ Rank Chittagong Chittagong H.T. Dacca Kushtia 614 (2) 69 (4) 2,950 (1) 491 (3) .0041 ' (3) .0002 (4) .0117 (1) .0087 (2) 162 195 115 105 0.4943 (3) • 1.1277 (4) -1.1010 (1) -0.5209 (2) Factor 1 ( C F . ) OVij -0.9897 Dij 0.9422 IB -0.9041 SC+LLALF/TALF -0.9748 LLALF/TALF -0.8790 Jute/NCA -0.8367 Table 5.97 URBAN TO URBAN OUT-MIGRATION STREAM Spat ia l Pattern of OVij from Kushtia, 1961 D i s t r i o t of Origin (ith D i s t r i c t ) - Kushtia II a" "b" "c" "d D i s t r i c t s of Dest inat ion ( j th d i s t r i c t s ) No. of in the Mi grants Stream Veloci ty Migration of Out-Streams . Distance Separating i and j ( in miles) Factor scores on the Subjects from the ' C r i t i c a l Factor ' OMij - Rank OVij Rank Dij Fj* Rank Chittagong 755 (3) .0106 (3) 225 0.6751 (3) Chittagong H.T. 44 (4) .0044 (4) 260 1.0016 (4) Dacca 1 ,892 (1) .0158 (2) 95 -1.0980 (1) Khulna 1 ,648 (2) .0287 (1) 105 -0.5787 (2) Factor 1 (C . F.) OVij -0.7228 Dij 0.9436 SC+LLALF/TALF -0.9270 IB -0.8061 Table 5.98 U R B A N T O U R B A N I N - M I G R A T I O N S T R E A M S . 1 9 6 1 S P A T I A L P A T T E R N OF I V I J TO CHITTAGONG D i s t r i c t of Destination (jth D i s t r i c t ) ~ Chittagong "a" "b "c" "d II D i s t r i c t s of Or ig in ( i th D i s t r i c t s ) No. of in the Migrants Stream Veloci ty Migration of In-Streams Distance Separati ng i and j ( in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor ' IMij Rank IVi j Rank Dij Fj Rank Chittagong H.T. 670 (3) .0296 (2) 35 .21 56 (2) Dacca 8,433 (1) .0903 (1) 1 50 -1 .6844 (1) Khulna 614 (4) .0042 (4) 162 .5989 (3) Kushti a 755 (2) .0110 (3) 225 .7385 (4) *Factor 1 ( ' C F . ' ) N.B. IVij -.6719 Level of S ign i f i cance RD/SQMCA .7514 > .4000 Uij -.5100 Dij .2614 (not s i g n i f i c a n t ) Table 5.100 URBAN TO URBAN I N - M I G R A T I O N S T R E A M S . 1 9 6 1 S P A T I A L PATTERN OF I V I J TO DACCA D i s t r i c t of Destination (jth D i s t r i c t ) - Dacca II _ II a "b i "c" "d D i s t r i cts of Or ig in ( i th D i s t r i c t s ) No. of i n the Mig rants Stream Veloc i ty Mi gration of .In-St reams Di stance Separating i and j (in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor ' IMij Rank IV i j Rank Dij Fj Rank Chittagong 8,061 (1) .0269 (1) 150 -1 .4561 (1) Chittagong H.T. 138 (4) .0035 (4) 185 .2561 (4) Khulna 2,950 (2) .0120 (3) 185 -.5611 (3) Kushti a 1 ,829 (3) .01 56 (2) 95 -.8819 (2) Factor 1 ( ' C . F . ' ) IVi j -.7915 Uij -.7613 Dij .2156 (not s i g n i f i c a n t ) Jute/NCA -.5113 Table 5.101 U R B A N T O U R B A N I N - M I G R A T I O N S T R E A M S , 1 9 6 1 S P A T I A L P A T T E R N O F I V I J TO K H U L N A D i s t r i c t of Destination (jth D i s t r i o t ) ~ Khulna n _ n a "b" "c" "d" D i s t r i c t s of Or ig in ( i th D i s t r i c t s ) No. of Migrants in the Stream Veloc i ty of In-Migration Streams Distance Separati ng i and j ( in miles) Factor Scores on the Subjects from the ' C r i t i c a l Factor 1 Chi ttagong Chittagong H.T. Dacca Kushtia IMij ~ Rank IVi j ~ Rank Dij * Fj ~ Rank 4,668 (2) 10 (4) 8,949 (1 ) 1,648 (3) .0461 (1) .0005 (4) .0365 (2) .0293 (3) 162 195 115 105 -1.5539 (2) .8914 (4) -1.6614 (1) .0134 (3) Factor 1 ( ' C F . ' ) IVi j -.8266 TCA/NCA .6109 LLALF/TALF .569