UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Urban planning in the developing countries: the use of quantitative planning methods Kumapley, Frank Titus Kofi 1981

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

Item Metadata

Download

Media
831-UBC_1981_A6_7 K94.pdf [ 5.04MB ]
Metadata
JSON: 831-1.0095403.json
JSON-LD: 831-1.0095403-ld.json
RDF/XML (Pretty): 831-1.0095403-rdf.xml
RDF/JSON: 831-1.0095403-rdf.json
Turtle: 831-1.0095403-turtle.txt
N-Triples: 831-1.0095403-rdf-ntriples.txt
Original Record: 831-1.0095403-source.json
Full Text
831-1.0095403-fulltext.txt
Citation
831-1.0095403.ris

Full Text

URBAN PLANNING IN THE DEVELOPING COUNTRIES: THE USE OF QUANTITATIVE PLANNING METHODS by Frank Titus Kofi/Kumapley B.Sc.(Hons.) University of Science and Technology Kumasi, Ghana, 1977 A THESIS PRESENTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDIES (School of Community and Regional Planning) We accept th is thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA MAY, 1981 © FRANK TITUS KOFI KUMAPLEY, 1981 I n p r e s e n t i n g t h i s t h e s i s i n p a r t i a l f u l f i l m e n t o f t h e r e q u i r e m e n t s f o r an advanced degree a t t h e U n i v e r s i t y o f B r i t i s h C o l u m b i a , I a g r e e t h a t t h e L i b r a r y s h a l l make i t f r e e l y a v a i l a b l e f o r r e f e r e n c e and s t u d y . I f u r t h e r agree t h a t p e r m i s s i o n f o r e x t e n s i v e c o p y i n g o f t h i s t h e s i s f o r s c h o l a r l y p u r p o s e s may be g r a n t e d by t h e head o f my department o r by h i s o r h e r r e p r e s e n t a t i v e s . I t i s u n d e r s t o o d t h a t c o p y i n g o r p u b l i c a t i o n o f t h i s t h e s i s f o r f i n a n c i a l g a i n s h a l l n o t be a l l o w e d w i t h o u t my w r i t t e n p e r m i s s i o n . Department o f Community and Regional P lann ing The U n i v e r s i t y o f B r i t i s h C o l u m b i a 2075 Wesbrook P l a c e V ancouver, Canada V6T 1W5 Date May, 1981 i i ABSTRACT The relevance of western planning models to urban processes in the developing countries i s a topic which has engaged.the attent ion of many planners. In th is study, three commonly used quant i tat ive planning methods are evaluated with respect to the circumstances under which they may be used, and what resul ts may be expected from applying them to urban planning operations in the developing countr ies. The three methods evaluated are: j - the Cohort-Survival Method of population forecast ing, - the Urban Economic Base model; and - the Gravity Models of Spat ia l Interaction The study uses the structura l analysis approach to model evaluat ion. In th is approach, the decis ive factor i s the a b i l i t y of the model being eva l -uated to adequately answer, or provide information on, relevant planning questions which are charac te r i s t i c of c i t i e s in the developing countr ies. i The author arr ives at the fol lowing major f indings and conclusions: iv) that the main deterrent to any possible use of the Cohort-Survival model in the developing countries i s the d i f f i c u l t y of estimating the migration vector. i i ) that the structure and the r e l a t i ve l y small base-ratios of urban economies in the developing countries great ly reduce the potential of the Economic Base model to the extent that i t can be considered i r re levant to c i t i e s in these countr ies. i i i ) that i n s t a b i l i t i e s in travel patterns, imperfections in urban housing markets and the existence of per iodic systems of marketing i i i in the developing countries create circumstances which are seem-ingly very d i f f i c u l t to handle within the presently known theore-t i c a l framework of the Gravity models of intra-urban t r i p d i s t r i -bution, res ident ia l locat ion and r e t a i l l oca t ion , respect ive ly. ; iv) that , in view of the above f ind ings, three main factors may be used to explain the d i f f i c u l t y of t ransferr ing western planning methods to the developing countr ies. These include: (a) d i f f e r -ences in cu l tura l and economic environments; (b) differences in rates of urban growth; and (c) differences in the degree of d i f f i -cu l ty in obtaining the necessary data. To overcome the d i f f i c u l t i e s of using the three models evaluated in the developing countr ies, a number of both short- and long-term measures are suggested. The short-term measures, which are spec i f i c to each of the three models, concentrate on ident i fy ing a l ternat ive approaches and sources of i n -formation for use by planners faced with the problems of c i t y development in the developing countr ies. F i na l l y , the author contends that i f long-run. solut ions are to be found to the problems of planning methodology transfer.by evolving and using l o ca l l y relevant planning techniques, then measures such as increased planning re-search c lose ly matched by improvements in planning data base and administra-t ion are required. i v TABLE OF CONTENTS ABSTRACT LIST OF TABLES LIST OF FIGURES ACKNOWLEDGEMENT CHAPTER I INTRODUCTION 1.1.1 - Models in Urban Planning 2 1.1.2 - Approach and Methodology 2 1.1.3 - Organization of the Study 4 II THE COHORT-SURVIVAL MODELS SECTION ONE - REVIEW 2.1.1 - Conceptual Framework 7 2.1.2 - Matrix Formulation 9 SECTION TWO - EVALUATION 2.2.1 - Introduction 13 2.2.2 - Survival Rates 13 2.2.3 - Estimating F e r t i l i t y Rates 20 2.2.4 - Estimation of the Migration Vector 27 2.2.5 - Conclusion 38 III THE ECONOMIC BASE MODEL SECTION ONE - REVIEW 3.1.1 - General Theory 42 3.1.2 - Base Area Del imitat ion 43 3.'1.3 - Basic/Non-Basic Ac t i v i t y C l ass i f i ca t i on 45 3.1.4 - L imitat ions of the Base Model 49 i i v i i v i i i V SECTION TWO - EVALUATION 3.2.1 - Background: The Urban Economic Base Structure 53 3.2.2 - The Base Ratio S t ab i l i t y Assumption 55 3.2.3 - The Base Theory as a Population Forecasting Device 59 3.2.4 - Problems Involving Base Measurement 61 3.2.5 - Conclusion 65 IV THE-GRAVITY MODELS SECTION ONE - REVIEW 4.1.1 - Conceptual Framework 66 4.1.2 - The Production-Attract ion Constrained Model 67 4.1.3 - The Attract ion-Constrained Model 69 4.1.4 - The Production-Constrained Model 70 SECTION TWO - EVALUATION 4.2.1 - Tr ip D is t r ibut ion 73 4.2.2 - Residential Location 76 4.2.3 - Retai l Location 78 4.2.4 - Conclusion 80 V CONCLUSIONS AND RECOMMENDATIONS SECTION ONE - SUMMARY OF FINDINGS 5.1.0 - Introduction 82 5.1.1 - The Cohort-Survival Model 82 5.1.2 - The Economic Base Model 83 5.1.3 - The Gravity Models 83 SECTION TWO - CONCLUSIONS 5.2.0 - Conclusions 85 vi SECTION THREE - RECOMMENDATIONS 5.3.0 - Introduction 87 5.3.1 - Approaches to Short-Term Planning: 87 A - Short-term Uses of the Cohort-Survival Model 88 B - Short-Term Alternat ives to The Economic Base Model 92 C - Short-Term Alternat ives to The Gravity Models 95 5.3.2 - Approaches to Evolving Appropriate Techniques 97 5.4.0 - Conclusions 103 BIBLIOGRAPHIC NOTES 105 BIBLIOGRAPHY 108 APPENDIX A - Def in i t ions of Population Terms Used in the Study 112 v i i LIST OF TABLES TABLE PAGE 2.2.2a Computation of Survival Rates For A Hypothetical Study Area 14 2.2.2b Cohort-Survival Rate D is t r ibut ion in L i be r i a , 1970 16 2.2.3a Natural Growth Indicators: A lger ia 1969-1971 22 2.2.3b L ibe r i a : Cohort F e r t i l i t y Rate D i s t r ibu t ion , 1970 22 2.2.4a Dis t r ibut ion by Decile of Average "Household" Consumption in Rural and Urban Areas, Five Countries 31 2.2.4b Gross Domestic Product Per Economically Act ive Person in Agr icu l tura l and Non-Agr icu l tura l Sectors: Five Countries 33 2.2.4c Tunis ia: Average Dai ly Calor ie and Protein Intakes Per Head in Urban and Rural Areas, 1964-1968 33 2.2.4d Ghana: Administrat ive Staf f D is t r ibut ion and Job Probab i l i t i es 34 2.2.4e Niger ia: Rural and .Urban Investment in Selected Sectors, 1970-1974 34 vi i i LIST OF FIGURES FIGURE PAGE 1.1.2a Organization of the Study 5 2.2.3a F e r t i l i t y Rate Project ion by Cohorts 24 2.2.3b F e r t i l i t y Path Determination 26 2.2.3c Graphical I l l u s t ra t i ons of F e r t i l i t y Assumptions 28 2.2.5a Components of Urban Population Growth in the Developing Countries 40 3.1.2a An I l l u s t r a t i on of Base Area Del imitat ion Using Three Variables 46 3.2.2a Mu l t i p l i e r Sens i t i v i t y to Changes in Base Ratio 57 i x ACKNOWLEDGEMENT I would l i k e to thank my supervisors, Dr. Michael Poulton and Professor Brahm Weisman. Their guidance, constructive c r i t i c i sm and encouragement are gra te fu l l y acknowledged. I am also very heavi ly i n -debted to Professor Henry Hightower who has generously contributed his time and ideas to the preparation of th is thes i s . F i na l l y , my special gratitude goes to the government of Ghana for sponsoring my graduate studies in Urban Planning at the University of B r i t i s h Columbia. CHAPTER ONE I N T R O D U C T I O N 2 1.1.1 - MODELS IN URBAN PLANNING One of the most fundamental changes in the recent h istory of urban and regional planning i s the development of quant i tat ive methodology. Primar-i l y , th i s change involves the development of mathematical models which attempt to simulate the structure and in ter re la t ionsh ips of urban and regional land uses (Wilson, 1.971). Although the tendency to use the new models as aids to planning decision-making has been world-wide, e f for ts leading to the i r development have been most extensive in the western coun-t r i e s . As a resu l t they are often designed.to address planning problems of local concern; and to use information that i s r e l a t i v e l y easy to co l l e c t in those countr ies. Crooks (1971), remarked that models constructed in the developed countries have not been very e f fec t ive when planning for the developing countr ies. He at t r ibutes th is to the probable inappropriateness of the assumptions, the var iab les , or the sub-models contained in the overal l model and to the planning problems pecul iar to the developing countr ies. These observations lead to the research object ive of th is study. ' 1.1.2 - APPROACH AND METHODOLOGY In th i s study, three quant i tat ive urban planning techniques are evalu-ated with respect to the i r possible use in the developing countr ies. The word 'eva luated ' , as used here, implies an assessment of the circumstances 3 in which i t might be appropriate to use the models; and the resul ts that one may expect from applying them to planning operations in the developing countr ies. The three models selected for th i s purpose are: a. The cohort-survival method of population forecast ing. Population analysis i s widely recognized as extremely important in regional and urban planning, (W i l l i s , 1972). The cohort-survival model has been iden t i f i ed by Rees and Wilson, (1975) as ". . . s t i l l the most common demographic model employed at nat iona l , regional and urban leve ls in many countr ies." b. The urban economic base model. Andrews, (1956), noted that: planners have a duty to understand the economies of the c i t i e s they p lan, i f they are to generate plans that w i l l promote socia l and physical development. Such an understanding w i l l make possible the estimation of secular trends in the urban economy; including the project ion of economic a c t i v i t y and employment. The urban economic base model i s one of the most common planning techniques used in the study of urban economies (see for example S i r k i n , 1959, p. 426; Andrews, 1968, p. 76). c. The gravity models of spat ia l in terac t ion . Models have been bu i l t to re late a wide range of spat ia l interact ion phenonmenon in planning. Examples include: journey to work, journey to shop, migrat ion, f re ight movement, telephone c a l l s , newspaper c i r cu l a t i on , the spread of innovation and so on. The tool that has been used most extensively by planners for the analys is of spat ia l in ter re la t ionsh ips i s the family of gravity models, (Black, 1973, p. 299; Wilson, 1971, p. 1). 4 This study uses the technique of "model evaluation by structural ana-lys is" as i ts methodology.' This involves two main steps: 1. An outline of the model to be evaluated; with emphasis on spell ing out its procedure, underlying assumptions, the measurement of var i -ables and their interrelationships; 2. An assessment of the relevance of the model based on: i) An identif ication of the problem or planning question on which the model is expected to provide information in the developing countries; and i i ) The characteristics of the model already outlined in step one. The main focus is on how the model is l ike ly to perform under the 'stresses' of the defined problems,, rather than on how i t can theoretical ly perform. This in turn implies that the evaluation" c r i ter ia are based essential ly on the problems that the model must answer, and how far i t can possibly answer them; including the poss ib i l i t ies of obtaining the relevant data to operate the models. Figure 1.1.2a i l lustrates how the model review and evaluation stages of the study relate to the various conclusions and recommendations outlined in the study. 1.1.3 - ORGANIZATION OF THE STUDY The study is organized into five chapters. Chapter 2 deals with the cohort-survival method of population forecasting. It is divided into two sections. The f i r s t section outlines the theory.of the cohort-survival model, while the second section evaluates i t in the l ight of the various factors that are l ike ly to influence urban population growth in the develop-ing countries. 5 COHORT MODEL ECONOMIC-BASE MODEL GRAVITY MODEL A REVIEW OF THE MODEL INCLUDING ITS THEORETICAL LIMITATIONS; A COMPARISON OF ONE ASPECT OF THE MODEL TCj THE RELATED URBAN CHARACTERISTIC IN THE DEVELOPING COUNTRIES 3. CONCLUSIONS ON AS-PECTS OF THE MODEL CONCLUSION ON MODEL SHORT-TERM APPROACHES TO PLANNING 6 GENERAL CONCLUSIONS < t 7 LONG-TERM RECOMMENDATIONS Figure 1 .1 .2a: Organization of the Study 6 Chapter 3, which i s devoted to the theory of the urban economic base, i s also divided into.two sections - theory and appraisal respectively. Chapter 4 focuses on the theory of the gravity models (- section one); and evaluation of i t s possible relevance to c i t i e s in the developing coun-t r i e s ; in as far as travel demand forecasting, residential and r e t a i l , loca-tions are concerned (section two). F i n a l l y , in chapter 5, sections one and two outline the findings and the general conclusions to be drawn from the analyses of.the three models respectively; while section three presents.a set ;of recommendations aimed at encouraging local research, and the development of l o c a l l y relevant plan-ning theory in the developing countries. 7a CHAPTER TWO T H E C O H O R T - S U R V I V A L A P P R O A C H TO U R B A N P O P U L A T I O N F O R E C A S T I N G SECTION ONE A REVIEW OF THE COHORT-SURVIVAL MODEL •2.1.1 - CONCEPTUAL FRAMEWORK1 Current population project ion e f for ts adopt variants of the cohort-component method. General ly, however, the cohort-component approach to urban population project ion involves carrying foreward the base-year urban population, (P*"), by age and sex, to one or more future dates by al lowing separately for f e r t i l i t y , mortal i ty ' and migration. It i s based on the con-cept of the "cohort", which in population models represents a group of peo-ple of the same age, or belonging to the same age-group. Given the base-year populat ion, (P^L of the n^1 cohort and the corresponding survival rate (S^) over the project ion period therefore, the number of people who w i l l survive to the next, cohort, (n + 1 ) , at time (t + d t ) , may be expressed as fo l lows: p ( t + dt) = c d t p t • m where the l e f t hand side of the equation represents the projected tota l -male and female - population of the n t ' 1 cohort. When disaggregated by sex, equation (1) may be rewritten as: p ( t + dt) = H S d t Hpt + p s d t t { ) n + 1 n n n n v ' where the H(male) and F (female) are introduced to indicate the sex - spec i f i -c i t y of the project ion var iab les . In a quite s im i la r approach, the number of new b i r ths , B^, (male and female) to chi ld-bear ing mothers ( i . e . 4 < n < 9, for f ive-year cohorts) , 8 may be e s t i m a t e d by m u l t i p l y i n g FP^ by t he c o r r e s p o n d i n g c o h o r t f e r t i l i t y ,dt r a t e , f , o r ; n - B d t =, f ^ . F P * (3) n ': n n T o t a l b i r t h s , ( T B d t ) , t o mothers i n ALL t he c h i l d - b e a r i n g c o h o r t s o v e r t he p e r i o d , d t , may then be o b t a i n e d by summing up t he i n d i v i d u a l c o h o r t b i r t h s i n the f o l l o w i n g manner : T Bd t = f d t . FP* . (4) ' r fs -4 n n The t o t a l new b i r t h s , t hus o b t a i n e d i s t hen d i s a g g r e g a t e d i n t o t o t a l f ema l e b i r t h s , FB , and t o t a l male b i r t h s , HB , by a p p l y i n g an e x o g e n o u s l y d e -d t d t t e r m i n e d i n f a n t s e x - r a t i o n . . The FB and t h e HB a r e t h en s u r v i v e d s e p a r a t e l y o v e r t h e p e r i o d , d t , by a p p l y i n g the f e m a l e - c h i l d s u r v i v a l , F S d t ; and t he m a l e - c h i l d s u r v i v a l , H B d t , r a t e s r e s p e c t i v e l y ; t o o b t a i n t he n e t number o f b i r t h s i n each sex c a t e g o r y . F i n a l l y , t h e t h i r d i n p u t t o t h e c oho r t - c omponen t a p p r o a c h , ne t i n - m i g r a -t i o n , must a l s o be e s t i m a t e d by age and by sex and then i n t r o d u c e d i n t o t h e c o h o r t " o p e r a t o r " . G e n e r a l l y , c o h o r t n e t i n - m i g r a t i o n , M d t , i s d e f i n e d as t h e a b s o l u t e number o f p e op l e i n t h e c o h o r t n , who have become r e s i d e n t s o f the s t u d y - a r e a o v e r t h e . p e r i o d d t , ( a f t e r a l l o w i n g f o r d e a t h s ) . * G e n e r a l l y t h e r e f o r e , t he p r o j e c t e d p o p u l a t i o n o f any o f t h e u sua l 18 c o h o r t s (0 - 4 , 5 - 9 , . 85+) f o r f ema l e s and ma les may be r e p r e -s e n t e d r e s p e c t i v e l y as f o l l o w s : J L ( , f dt _ F p t ( 1 _ m ) F S d t ) + F M d t 9 ( t o t a l f ema l e b a b i e s ) r * — T I n n i i t + dt_ F P n + l - 4 (5) F S d t . F P J + F M d t , 2 < n < 18 * T h i s w i l l be n e c e s s a r y i n a s i t u a t i o n where t he m i g r a n t s were c oun t ed as t h e y e n t e r e d t h e s t u d y a r e a . 9 HP t + dt n + 1 (fd t:LMS? t.FP t (m)) + n .1 n .dt HS^.HP* n n + HM dt (tota l male babies) . . . . (6) 2 < n < 18 where m i s the probab i l i ty that a newly born baby i s a male. 2.1.2 - MATRIX FORMULATION2 The procedure described above may be re-stated using matrix algebra. The fol lowing broad steps are necessary: 1) Assuming for the sake of s imp l i c i t y that we are only concerned with the project ion of the female population in our study area; l e t us f i r s t compile a colum vector, P*, to show the female age d i s t r i bu t i on in n (or 18) cohorts - 0 - 4, 5 - 9, . . . . . . . 85+, at the base year, t . (7) 2) Compile a matr ix, FS, n x n, showing the cohort f e r t i l i t y rates (f ) and the cohort surv ival rates, (S )> for females, simultaneously. 10 FS f l , 3 f l , 4 f 1 ,5 • 0 0. 0 0 0 0 0 0 0 0 0 4,3 0 0 0 1,9 *2,1 0 S 3,2 0 0 s 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 s v 0 n,n-2 (8) ,dt 3) Compile a second colum matrix, M , showing, in absolute numbers, the net in-migration of females into the study area over the project ion per iod, (after al lowing for deaths). i — m1 ,dt m (9) 4) By mult ip ly ing FS and P together, the projected female population plus babies (male and female), but without in-migration i s obtained as fol1ows: 11 ,t+dt 0 P. Natural Growth Babies 0 0 0 0 0 0 New Babies (10) 5) By mult ip ly ing the vector of new babies by (1-m), where m i s the probab i l i ty of a newly born baby being a male; and then by ch i l d (0-4) sur-v iva l rates, the number of female babies, P^, surv iv ing to the end of the period, dt, i s obtained and then added to the natural. growth matrix of fe-males . pt+dt (adjusted for sex and death of babies) 1 (11) 12 6) F i na l l y , the tota l projected female population in each cohort, TP is obtained by adding equations (9) and (11) together. T i + m7 P 2 + m2 t+dt TP t+dt P3 + m3 P 4 + m4 P + m n n (12) At the end of the 6th step, the female population of the study area is projected by sex and by age cohorts. The process may be repeated for the male population, the only differences being the absence of f e r t i l i t y rates in the FS, now S, matrix; and the fact that the number of babies from equation (10) times m, then mul t ip l ied by male ch i ld morta l i ty rates; and male in-migration w i l l be added to natural growth. 13 SECTION TWO  EVALUATION OF THE COHORT-SURVIVAL METHOD 2.2.1 - INTRODUCTION The potent ia l of the cohort-survival method of population forecast ing to y i e l d su f f i c i en t l y accurate resul ts for planning purposes is mainly de-pendent on the r e l i a b i l i t y with which the necessary inputs to the method can be estimated. These include: - the expected survival rates by age-cohorts and sex; (S*); - the expected f e r t i l i t y rates by cohorts, of chi ld-bear ing mothers; (F*); and - the expected net migration by sex and age-cohorts; (M*). The rest of th is chapter examines the r e l i a b i l i t y with which each of the above inputs could be obtained for c i t i e s in the developing countr ies; with respect to the fol lowing major areas of concern: - the need to obtain s t a t i s t i c s for. small areas as well as for aggre-gates of areas in the c i t y of in te res t ; and - the need to make provis ion for future changes in the inputs, given that under conditions of rap id.urbanizat ion, there can be consider-able changes in both the magnitude and d i s t r i bu t ion of the data within r e l a t i v e l y short periods of time. 2.2.2 - SURVIVAL RATES In population studies, the term "cohort surv ival rate" refers to the probab i l i t y that a person in a given age-cohort w i l l survive to the next 14 age-cohort. Survival rates show s ign i f i can t var iat ions in re la t ion to age, socio-economic status (e.g. occupation, l i t e r a c y ) ; and residence (rural or urban). They also vary depending on such environmental charac ter i s t i cs as c l imate, a l t i tude and qua l i ty of ava i lab le health f a c i l i t i e s , water sup-ply and so on. A l l . such variations, deserve consideration in the se lect ion and analysis of survival s t a t i s t i c s which are often avai lab le in the form of death records. For example i f the study area i s not homogeneous in terms of the socio-economic charac ter i s t i cs of. i t s r e s i den t s , . i t may then be found necessary to compile d i f fe rent survival rates for. each soc.i-economic group. Also, improvements in local. .health . f a c i l i t i e s , may, up to a point , lead to increases in surv ivorship. TABLE 2.2.2a Computation of Survival Rates for a Hypothetical Study Area POPULATION Deaths Death Survival Survival (a) (b) Ratio Ratio Rate (c-b/a) (1-b/a)  0 - 4 18000 1000 1/18 1 - 1/18 0.944 5 - 9 17000 1000 1/17 1 - 1/17 0.941 10 - 14 16000 1000 1/16 1 - 1/16 0.938 85 & over 1000 1000 1/1 1 - 1 0.000 Table 2.2.2a shows the computation of f ive-year survival rates for a hypothetical area. Columns a and b may.be obtainable. d i r ec t i y from such sources as census and death records, respect ive ly; or i nd i r ec t l y through intracensual or post-censual estimates. They may also be.obtained from 15 model l i f e tab les . In the developing countr ies, however, circumstances ca l l for a careful examination of the r e l i a b i l i t y of survival rates derived from these sources before the i r use in the project ion; since mortal i ty re-cords - pa r t i cu l a r l y those of infants - are notoriously unrel iable in many Afr ican countr ies. Levels of morta l i ty in A f r i ca are general ly very high; (Conde, 1971). Crude death rates range between 20 and 40 deaths per thousand population. As remarked by Cantre l le , (1971); morta l i ty leve ls in A f r i ca also vary be-tween d i f f e ren t regions of the continent; over time; and even between rural and urban areas in the same country. In the Cameroons for example, mortal-i t y is observed to be lower among the pastoral people (of Adamoua) than among neighbouring farmers. The data ava i lab le from d i f fe rent Afr ican countr ies, according to Cantre l le , also reveal that morta l i ty rates are general ly lower in the urban centres compared to those preva i l ing in the rural areas; (Table 2.2.2b). A comparison of diagrams; by Conde, (1971); showing observed morta l i ty rates in . ce r ta in Afr ican countries with those shown by the model l i f e tables of the United Nations leads him to three main conclusions. F i r s t , there ex is ts considerable deviations.between the observed information and those reproducible by the l i f e tables. Secondly, the use of . the U.N. model l i f e tables in A f r i ca leads to an underestimation of both infant morta l i ty and mortal i ty in the extreme old age groups. Third ly Conde concluded that the model l i f e tables overestimate morta l i ty rates for adul ts . Cantrel le (1971) explains that the above mentioned disagreements between the model l i f e 16 TABLE 2.2.2b Cohort Survival Rate Distribution in Liberia, 1970 Urban Rural Age Cohort Total Male Female Total Male Female 0 - 4 0.948 0.973 0.950 0.911 0.906 0.915 5 - 9 0.996 0.995 0.998 0.996 0.995 0.996 10 - 14 0.999 0.998 0.999 0.996 0.996 0.997 15 - 19 0.998 0.998 0.997 0.996 0.996 0.995 20 - 24 0.995 0.996 0.994 0.995 0.992 0.996 25 - 29 0.998 0.997 0.998 0.993 0.992 0.995 30 - 34 0.994 0.996 0.993 0.993 0.991 0.995 35 - 39 0.994 0.996 0.993 0.993 0.991 0.995 40 - 44 0.996 0.995 0.996 0.991 0.984 0.996 45 - 49 0.991 0.991 0.986 0.992 0.993 0.992 50 - 54 0.982 0.975 0.993 0.986 0.978 0.995 55 - 59 0.989 0.996 0.980 0.982 0.975 0.990 60 - 64 0.980 0.974 0.990 0.984 0.985 0.983 65 - 69 0.972 0.976 0.966 0.979 0.976 0.981 70 & over 0.873 0.859 0.893 0.959 0.956 0.962 All Ages 0.988 0.988 0.989 0.982 0.980 0.984 SOURCE: Computed from mortality rates, U.N. Demographic Yearbook, 1977. 17 table resu l ts and the actual morta l i ty trends in A f r i ca derive from the fact that the data which was used for the compilation of the tables do not in many cases re late to t rop ica l populations; pr imar i ly because they do not account for the s ign i f i can t d i spa r i t i e s that ex i s t in patterns of l i f e expectancy. I t is imperative that the population analyst in the developing coun-t r i e s obtains survival rates for his study area that actua l ly r e f l e c t the ex is t ing socio-economic mix; espec ia l l y because of the continuous in f lux of rural migrants into the urban areas. The use of national averages i s not recommended in th is par t i cu la r case. The reasons for th is are twofold. F i r s t , as already mentioned, d i f fe rent socio-economic groups in A f r i ca ex-perience wide var iat ions in morta l i ty rates (between 20 to 40 deaths per thousand). Secondly, the a v a i l a b i l i t y and level of service of medical f a c i l i t i e s vary substant ia l l y over geographic regions. Under these circum-stances, i t i s perhaps best to use d i f fe rent survival rates for d i f fe rent population groups in the study area. I t should be possible to obtain th is type of information from hospital records.and/or death r eg i s t r i e s . Another equally chal lenging question that faces population analysts in the developing countries i s how to forecast future survival, rates from the avai lab le information on morta l i ty trends. The need to forecast survival rates ar ises -because of recent^improvements in health care; which can have s ign i f i can t impacts on future, morta l i ty rates. The techniques often used for the project ion of survival rates include: a. an extrapolat ion of h i s t o r i c trends either, graphical ly or by f i t t i n g the appropriate mathematical curve. 18 b. an appl icat ion of standard percentage changes to the base-year surv ival rates, depending on the level reached at each successive project ion year; and c. subject ive ly establ ish ing survival rates for a d istant future date and then securing rates for intermediate dates by some form of in te rpo la t ion . The f i r s t two approaches l i s t e d above commonly involve the use of a progressive set of survival rates for each successive f ive-year (cohort) per iod. Their use in c i t y planning in the developing countries w i l l there-fore be reduced by d i f f i c u l t i e s in obtaining the kind of demographic data required. Furthermore, massive investments in medical: f a c i l i t i e s or d i s -ease eradicat ion programs, (as often observed in the developing countr ies) , may completely upset h i s t o r i c trends; and therefore render any resu l ts thus obtained grossly erronous. In th i s regard, the th i rd approach appears to be most appealing to the population, analyst in the developing countr ies, espec ia l ly when i t i s supplemented by the delphi technique; (see for example Ley and Anderson., 1975). I t can be based large ly on future plans and p o l i -c ies aimed at e l iminat ing certa in causes of death (e.g. malar ia) . In a l l the above cases, the projections w i l l generally provide for im-provements in surv ivorship. This therefore implies the existence of the r i sk of a r r i v ing at unreasonably high leve ls of surv ivorship, when d i rect extrapolat ion i s used. Accordingly, therefore, a l im i t should be set to the improvement assumed to occur, so that a f ter a given future date, the cohort surv iva l rates may be assumed not to change any further; e i ther a l l 19 at once, or separately at d i f fe rent dates. The target rates set to avoid the occurence of such unreasonable levels of survivorship may be derived in a number of d i f ferent ways. a. the use of rates already attained in some economically more ad-vanced areas of the country. b. the use of rates already attained in another urban area, not in the same country, but which has socio-economic character i s t i cs and other survivorship determinants s im i la r to those ant ic ipated in the study area ( i . e . use of a pattern area). c. an analysis of the cohort survival rates being projected in terms of the i r components (causes of death) for which judgemental projec-t ions could more conf ident ly be made. The f i r s t of the above three a l ternat ives i s theore t i ca l l y quite rea-sonable. I f town A now registers a high survival ra te , i t can be imagined that, given time, town B in the-same country can also reg is ter a s im i l a r l y high survival rate. In pract ice however, the question ar ises as to WHEN town B w i l l equalize A, and under WHAT condit ions. Even i f the two geogra-phic subdivis ions can enjoy equal * levels of medical care, there are other factors such as ethnic and occupational.dif ferences which can influence survival rates. In A f r i ca such d i spa r i t i e s can be very substant ia l . The obvious conelus ion then is that any imi tat ion of survival rates as target rates for forecast ing purposes cannot be without d i f f i c u l t i e s . One can expect even worse resu l ts when the rates are adopted from d i f fe rent coun-t r i e s . A point in support of th is view can be derived from Cantre l le ' s 20 (1971) explanation of why model l i f e tables do perform badly in A f r i c a . It therefore seems more appropriate for the population analyst in the developing countries to attempt to project his target survival rates from an analysis of the expected changes in the various causes of death. In Ghana for example, s t a t i s t i c s on causes of death are ava i lab le from Ghana Sta t i s t i ca l Yearbooks. One d i f f i c u l t y with th is approach however is the amount of research time that w i l l be necessary to achieve reasonably accur-ate resu l t s . 2.2.3 - ESTIMATING FERTILITY RATES Cohort f e r t i l i t y rates are the.number of l i v e births experienced by mothers in each of the seven chi ld-bear ing cohorts; (15 - 19, 20 - 24, . . . 45 - 49), over a f ive-year per iod. F e r t i l i t y rates are highly sens i t ive indices which can f luctuate con-s iderably during the planning horizon covered by most population project ions, (Masser, 1972). Such f luctuat ions may have cumulative adverse ef fects on project ion resul ts when over-estimated b i r ths , for example, begin to pro-duce the i r own babies in the l a te r stages of the project ion. Most popula-t ion project ion ef for ts w i l l therefore use projected f e r t i l i t y rates, es-pec ia l l y in cases when s t a b i l i t y in those observed h i s t o r i c a l l y cannot be predicted. According to Conde, (1971), a l l estimates ind icate/that f e r t i l i t y rates are current ly very high in A f r i c a , (about 47 per thousand women aged 15 -49 years, as compared to only 19 in Europe). Variat ions between countries range from 25 per thousand in Gabon to 62 per thousand in Guinea. 21 S im i l a r l y , age-speci f ic f e r t i l i t y rates are also known to be generally quite high in the developing countr ies; with the national maxima of the various countries occuring in the 20 - 24 age-cohort of mothers; Table 2.2.3 a and b. A number of reasons have been given for the persistence of r e l a t i ve l y high natural growth rates in the developing countr ies. Among these are: decl in ing infant and adult morta l i ty rates due to improvements in health care; increasing marriage rates , and the payment of family allowances, (based on number of ch i ld ren) , as pract ised in Togo for example. The impl icat ions of the above mentioned f e r t i l i t y trends for popula-t ion project ion for c i t i e s in the developing countr ies, re la te to the num-ber of assumptions that must be made in order to ar r ive at meaningful e s t i -mates of f e r t i l i t y rates for various stages of the project ion period. These w i l l centre mainly around (a) whether future f e r t i l i t y trends w i l l depend on past chi ld-bear ing preferences; or (b) whether i t i s more j u s t i f i a b l e to make calculated guesses as to what future rates might be. In resolv ing th is dilemma the two known methods of f e r t i l i t y rate fo recas t ing . - (the period-f e r t i l i t y method and the c oho r t - f e r t i l i t y method) - are examined. ( i ) The Per iods-Fert i l i ty Method This method examines the h i s to r i ca l data.on b i r th rates recorded for indiv idual age-groups, (or cohorts), of chi ld-bear ing women. It consists of: a. ca lcu la t ing and analyzing a time series of b i r th rates for each f ive-year age-group women. b. Project ing the rates e i ther graphica l ly or by f i t t i n g the appro-pr iate curve. 22 TABLE 2.2.3a  Natural Growth Indicators of A lge r i a , 1969-71 Crude General Gross Net Repro B i r th F e r t i l i t y Reproduction duction Area Rate Rate Rate Rate Urban Strata 0.0472 0.2191 3.59 2.85 Rural Strata 0.0513 0.2388 . 3.95 x 2.74 Total 0.0499 0.2321 3.83 2.76 Source: Negadi, Gourari; " F e r t i l i t y in A lge r i a , Population In Afr ican Development, Ordina, Dolhain, 1971, pp. 72. TABLE 2.2.3b Cohort F e r t i l i t y Rate Dist r ibut ion ,in L i be r i a , 1970 (Per Thousand) Age Cohort Urban Strata Rural Strata 15 - 19 226.9 212.7 20 - 24 ,267.3 251.5 25 - 29 265.0 226.4 30 - 34 218.0 253.6 35 - 39 144.5 166.4 40 - 44 61.3 115.1 45 - 49 and over 33.7 30.0 A l l Ages 217.6 198.1 SOURCE: U.N. Demographic Yearbook, 1977. 23 The charac te r i s t i c feature of th is method i s that the trend analysis is in terms of rates for a given age-group. A hypothetical example of the p e r i o d - f e r t i l i t y project ion approach i s shown graphica l ly in Figure 2.2.3a, for the two age-groups 1 5 - 1 9 and 2 0 - 2 4 . This method i s very simple to apply. Operat ional ly, the a l ternat ive projections i l l u s t r a t ed in dotted l ines in Figure 2.2.3a permits the popula-t ion analyst . to input any subjective judgement he might have regarding ex-pected future changes in p e r i o d - f e r t i l i t y trends. The r e l i a b i l i t y of the method however depends 1 argely on the a v a i l a b i l i t y of h i s to r i ca l data on p e r i o d - f e r t i l i t y . While such data may be obtained from the records of Births and Deaths reg i s t r i e s located in many c i t i e s of the developing coun-t r i e s , the analyst may have to allow a substantial amount of time for the extract ion of such data from the records (Gardiner, 1972, page 8) . A second and perhaps more important weakness of the p e r i o d - f e r t i l i t y method with regard to i t s appl icat ion in project ing urban population in the developing, countries derives from the f a c t that i t does not account for the postponement or otherwise of marriages, nor the increase or decl ine in pre-ferred family s i zes . In the developing countr ies, espec ia l l y in A f r i ca the chances of such changes occuring are very high; due to changes in income leve l s , standards of l i v i n g and health care. Conde (1971) for example, re-marked that a very high corre lat ion i s known to ex i s t between infant mor-t a l i t y and b i r th rates in A f r i c a . Since infant morta l i ty varies between 150 and 250 deaths per thousand b i r ths , women in most cases i n s i s t on having between s ix and seven births in order to be certa in of having at least two of the chi ldren survive to the age of twenty. This, therefore, implies Figure 2.2.3a The Cohor t -Fer t i l i t y Approach to the Projection of F e r t i l i t y Rates 25 that , with, dec! ining infant morta l i ty rates due to improved health care; preferred family s izes among women in A f r i ca might dec l ine. The usefulness of the p e r i o d - f e r t i l i t y method in the project ion of f e r t i l i t y rates for c i t i e s in the developing countries i s therefore quite dubious. (i i.) The Cohor t -Fe r t i l i t y Method The c oho r t - f e r t i l i t y rate approach examines the trend of f e r t i l i t y for separate birth-cohorts of women in terms of f ive-year (or durat ion-spec i f i c) b i r th - ra tes . I t employs data on the f e r t i l i t y h istory of chi ld-bear ing women as they progress through the chi ld-bear ing cohorts. The fact should not be overlooked that c oho r t - f e r t i l i t y data represents the cumulative fe r -t i l i t y of chi ld-bear ing women over successive cohorts. To i l l u s t r a t e the approach, l e t us assume that Figure 2.2.3b represents a p lot of the observed f e r t i l i t y h is tory of a given cohort of mothers. Given that the curve thus obtained represents the f e r t i l i t y path of any group of women entering the chi ld-bear ing ages, then the horizontal scale of Figure 2.2.3b may be replaced by a time-scale in years as shown in the horizontal scale-two, above the graph. Assuming that the current ly observed, c oho r t - f e r t i l i t y rate of mothers aged 15 - 19; whose f e r t i l i t y we want to project i s located at point A shown in the Figure; then the i r projected f e r t i l i t y rate in f i ve years time is given by point B; a l l other things remaining the same. One of the advantages in the use of the c oho r t - f e r t i l i t y approach i s that i t uses assumptions d i r ec t l y formulated in terms of the completed fe r -t i l i t y of real cohorts of women; so that unrea l i s t i c assumptions about im-p l ied family sizes are avoided. Furthermore, the approach makes possible 26. Figure 2.2.3b: Graph Plotted From Data on the F e r t i l i t y History of Chi ld Bearing Women 27 the use of addit ional information such as the expressed expectations of women regarding completed family s izes (as may be obtained in a local sample sur-vey). Changes in f e r t i l i t y trends based on expressed expectations of lower or greater family s izes w i l l lead to downward or upward SHIFTS in the o r i -ginal curve as shown in dotted l ines in the Figure 2.2.3c. The ef fects of the various combinations of other assumptions are also shown in the same f igure. The c oho r t - f e r t i l i t y method therefore has a number of important qua l i t i e s that must be of par t i cu la r in terest to the population analyst in the develop-ing countr ies. Apart from the fact that i t allows the introduction of r e l e -vant assumptions into the ana lys i s , i t s data requirements are also manageable. This i s because i t s implementation may be completely based.on cross-sect ional data, co l lected for the base-year only. Assumptions regarding preferred family s izes and,postponment or otherwise, of marriages, may be based on the resul ts of an opinion survey of mothers. In short, the c oho r t - f e r t i l i t y method should be more appealing to population analysts in the developing countries than the p e r i o d - f e r t i l i t y approach. 2 . 2 . 4 - ESTIMATION OF THE MIGRATION VECTOR "Formal demographic methodology i s least appl icable to the problems of project ing migrat ion, and conversely, th is i s the area in which the profes-sional judgement and deta i led knowledge of the planner is most relevant. For these reasons i t i s suggested thatmi.gration be introduced into the project ion by adding a vector . . . o f net in tegrat ion. . . " , (Hightower, 1968). 28 CONSTANT MARRIAGE AND BIRTH EXPECTATIONS. BUT A = desire for larger family s i zes . B = desire for smaller family s i zes . 30. 20 . 10 • i i i i i • 0 5 10 15 20 25 3 | • i i l | L _ ^ « k ' l * L . L r _ _ ! | ^ 1 / l > s 1 1 1 , . 1 - . X V l ; 1 i 1 1 • \ j • • . . 1 1 ' f*** CONSTANT FAMILY SIZES BUT C = postpones marriages/ b i r ths expected D = ea r l i e r marriages/ b i r ths expected F* 30 . 20 . 10 . 1 > 1 l - l 1 > • , 1 1 . 1 1 , 1 1 1 1 J i _ _ J 1 . . . . . . L . . | L - . - . j T — -H^.-j + u- -' i - -i - - • V & U ! : : 1 ! ^ LARGER FAMILY SIZES EXPECTED BUT E = postponed marriages/births F = Ear l i e r marriages/births F* 30 . 20 . 10 . i • : i : i U -*.V^ -% '^h;- n , j - - - r - - - -1 • i i i u____, r _ _ . ' ! I ! ! SMALLER FAMILY SIZES EXPECTED BUT H = postponed marriages/births G = e a r l i e r marriages/births F* 30 . 20 . 10 . ! ' ! i 1 1 i i I . I L . J 1 L ! L ^••••I '-A:--—-t I ' ' • i r f i ~ ' N: *~ I : s ! *k ! : ! ! ! 1 1 1 1 1 1 - - m i n i Figure 2.2.3c: Graphs Showing Various F e r t i l i t y Assumptions (In A l l Cases, the Bold-Type Curve Shows the Observed F e r t i l i t y Path) 29 As already mentioned in section 2.2.2 of th is chapter, the average an-nual growth rate of c i t i e s in A f r i ca is about 5 per cent. Todaro, (1976), remarked that, the continuing in-migration of rural people into urban areas in the developing countries i s a major cause of the rapid rates of urban pop-ulat ion growth. In A f r i c a , for example, rural-urban migration takes a 50 per cent share ( i . e . 2.5% out of 5%) of recorded urban growth rates. However, in some of the large and fast-growing c i t i e s , such as Cairo, Lagos and Ibadan, growth rates may be as high as 10%; with net in-migration accounting for well over 50% of tota l growth. The factors that influence the importance of rura.l-urban migration in the growth of c i t i e s in the developing countries are varied and complexly in te r re la ted . The migration decision process of the rural dwellers in these countries has been conceptualized as involv ing certa in s o c i a l , demographic, educational and economic considerations.; based on the character i s t i cs of both the or ig in and the dest inat ion areas. The overwhelming conclusion of almost a l l migration studies i s that the economic factors operate through the demo'-, graphic and the educational factors to const i tute the primary determinants of rural-urban, migrant f low, (Gugler, 1976; Clarke, 1971; Browning, 1971; Todaro, 1976). More s pe c i f i c a l l y , the studies maintain that: 1. migrants t yp i ca l l y do not represent a random sample of the population from which they or ig inate . Instead they tend to possess certa in demo-graphic and educational/professional at t r ibutes which in many cases rank above average compared to those of the population from which they or ig inate . This notion introduces the concept of a "migratory reservo i r " ; which refers co l l e c t i v e l y to the source areas of migrat ion. The s i ze , 30 qua l i ty and demographic character i s t i cs of th is ' reservo i r ' then con-s t i t u t e a determining factor of the rate of migrat ion. These are often ca l led "Push" fac tors . 2. That people migrate pr imar i ly for economic reasons. The greater the difference in economic opportunit ies between the source and the dest in-ation areas of migrat ion, the greater the flow of migrants. "While d i s -tance i s usual ly a s ign i f i can t intervening obstacle, i t s negative im-pact can be large ly of fset by s izeable income d i f f e r en t i a l s , espec ia l ly for the more educated migrants", (Todaro, 1976). The expected dest ina-t ion economic opportunit ies are often c l a s s i f i e d as (a) income d i f f e r -en t i a l s , (b) job p robab i l i t y , (c) the f r i c t i o n factor of migrat ion, and (d) urban serv ices. a. INCOME OR WAGE DIFFERENTIALS: Differences in.average income or wage levels between two places invar iab ly turns up among the most important explan-atory factors inf luencing the flow of migrants between them, (Todaro, 1976). In the developing countries where wide gaps are commonly i den t i f i ed between rural, and urban income d i s t r ibut ions therefore, the rate of rural-urban migra-t ion w i l l increase in d i rec t proportion to the s ize of the rural/urban income d i f f e r e n t i a l . It may also be possible to replace income d i f f e ren t i a l s with (or supple-ment them by) such variables as expenditure, product iv i ty or welfare d i f f e r -en t i a l s . It may be noted from Table 2.2.4a that, general ly, the proportion of poor households tends, to be larger in the rural areas'than in the urban centers. This is most l i k e l y due to the fact that product iv i ty tends to be lower in the. agr i cu l tura l sectors than in the non-agr icultural ones, espec ia l ly 31 TABLE 2.2.4a The D is t r ibut ion by Decile of Average "Household" Consumption in Rural and Urban Areas: Five Countries  Decile INDONESIA - MEXICO PAKISTAN TANZANIA TUNISIA Rupiahs/Mth./ Pesos/Mth./ Rupees/Mth./ Sh i l l i ngs / Dinars/Year/ Household Family Household Year/ Household Member Household Member Rural Urban Rural Urban Rural. Urban Rural. Urban.. Rural Urban I 389 552 294 493 86 103 359 903 13 28 II 566 836 469 /'808 N\ 113 151 657 1713 24 45 III 728 1032 523 :1051 141 173 851 ;*2238"- 30 56 IV 837 ;' 1208 \ 592 .: 1306 159 200 864 • 2652 : 3 6 66 V 971 ; 1410 676 1528 : 168 229 981 2963 : 41 ; ' " 7 3 (c) VI 1148 1581 871 1921 . 201 257 1021 : 3626 \ 49 ; 93 VII 1345 1981 1081 2374 225 306 \ 1431 : 4554 59 119 VIII 1604 2273 1296 2907 250 358 1705 4840 70 141 IX 2036 2645 1798 3879 289 443 2031 5648 84 184 X 3566 5249 2637 8572 484 875 3296 9653 163 441 "Pove r ty -^ne 1132 632 772 214 262 1440 1920 54 70 Percentage of Households Below "Poverty Line: 55 30 40 15 65 55 65 20 60 40 (c) SOURCE: NOTES: Ginneken, a 1976; p. 371 For Tanzania, an Ordinance of 1969 on basic minimum wages pro-vides a means of quantifying rural and urban poverty l i ne s . For Tun is ia , a poverty l i ne i s provided in a development plan in which a minimum of 70 dinars/head is set as target. The poverty contour. 32 in the developing countr ies. Ginneken (1976) concludes from his study of f i ve developing countries (Table 2.2.4b) that product iv i ty d i f f e r en t i a l s between agr i cu l tura l and non-agr icultural sectors play a s ign i f i can t - pos-s i b l y the pr inc ipa l - role in determining the extent of expenditure (or i n -come) inequa l i t i es between rural and urban areas. Table 2.2.4c shows a sam-ple set of welfare indicators for rural and urban areas in Tunis ia. b. URBAN JOB PROBABILITY: According to Todaro, (1976); ' job probabi l -i t y ' as a var iable in rural-urban migration flow functions appears to have an "independent" s t a t i s t i c a l s ign i f i cance in the sense that when i t i s i so -lated from other var iables such as absolute or re la t i ve income d i f f e r en t i a l s , i t improves the overal l explanatory power of the regression funct ion. In other words, the greater the l i ke l ihood of gett ing urban employment, the larger the flow of rural migrants to the c i t i e s . Table 2.2.4d shows adminis-t ra t i ve job** d i s t r i bu t ion indices for seven urban centres in Ghana. The differences in the p robab i l i t i e s r e f l e c t the inequa l i t ies in the spat ia l d i s t r i bu t ion of administrat ive jobs. S imi lar inequa l i t ies in rural/urban job opportunit ies may be inferred from the information presented in Table 2.2.4e. c. THE FRICTION FACTOR OF MIGRATION:. The negative ef fect of distance on migration as predicted by the t rad i t i ona l gravity models (Schultz, 1976)* i s pronounced in most migration studies. Migrants tend to move to the c i t i e s in the i r own state or region; but they w i l l move longer distances i f the des-t i na t i on ' s economic opportunit ies are comparatively higher, (House and *As quoted by Todaro, (1976). **'Administrat ive job ' as used here refers to jobs in government o f f i c e s . 33 TABLE 2.2.4b Gross Domestic Product Per Economically Act ive Person in Non-Agricultural Sectors: Five Countries Agr icu l tura l and (Indices , Countrywide (a l l sectors) Product iv i ty = 100) Country Agr ic . Sector (A) Non-Agric. B/A Sector (B) (.C) Urban/Rural Consumption Expenditure Indonesia (1971) 69 159 2.3 1.4 Mexico (1970) 27 153 5.7 2.3 Pakistan (1972) 65 152 2.3 1.5 Tanzania (1967) 48 689 14.2 2.9 Tunisia (1966) 39 151 3.8 2.2 SOURCE: Ginneken, 1976; p. 40. TABLE 2.2.4c Tunis ia: Average Daily Calor ie and Protein Intakes Per Rural Areas (1964-68) Head In Urban and Calor ie Intakes (Minimum Need i s 2220) Protein Intakes (Minimum Need i s 62 Grams.) Country Total 2365 65 Urban Areas 2550 68 Rural Areas 2315 64 SOURCE: Ginneken, 1976; p. 32. 34 TABLE 2.2.4.d Ghana: Administrat ive* Staf f D is t r ibut ion Indices (1960) Employment as % Population As % Administrat ive of National of National Job Urban Center Total (a) Total (b) P robab i l i t i es (a x b) 100^ Accra 30 5.0 0.01500 Kumasi 11 2.7 0.00297 Sekondi-Takoradi 6 1.1 0.00066 Cape Coast 1 0.6 0.00006 Tamale 1 0.6 0.00006 Koforidua 2 0.5 0.00010 Winniba 0.4 0.4 0.00002 Rest of Country 49 89.1 -SOURCE: Computed from 1960 Population Census of Ghana, Special Report A. TABLE 2. ,2.4e Niger ia: Rural Urban Investment in Selected Sectors, 1970-74 Urban Rural Total Investment Investment Planned N N Investment Sector M i l l i on % M i l l i on % (N M i l l i on ) Industry 77.7 91.2 8.4 8.8 86.1 E l e c t r i c i t y 40.3 89.0 5.6 11.0 45.3 Water & Sewage 42.2 71.6 9.5 18.4 51.7 Town & Country Planning 18.0 94.3 1.1 5.7 19.1 Education 98.4 70.9 40.5 29.1 138.9 Health 45.2 84.0 8.6 16.0 53.8 Social Welfare 11.0 91.7 1.0 8.3 12.0 TOTAL 322.9 81.8 74.1 18.2 406.9 SOURCE: M i t che l l , 1971. *The word ADMINISTRATIVE as used here refers to jobs in Government Of f ices . 35 Rempel, 1976)*. Destination contacts tend to const i tute a force act ing in a d i rect ion opposite to the " f r i c t i o n " of migration. Such contacts in form of fr iends and re la t ives provides important information on job opportunit ies and ass i s t in reducing the ef fect ive cost of job search by of fer ing cost less or low-cost accommodation to the migrants; (F ie lds , 1975)*. d. URBAN SERVICES AND AMENITIES: "The re la t i ve abundance of urban ser-vices and amenities do not seem to exert an independent pos i t ive ef fect on migration. The evidence on th is point, however i s very tentat ive and fuzzy since none of the current econometric studies measures a migrant's u t i l i z a -t ion of urban serv i ces , " Todaro (1976). Generally therefore, economic motivations are paramount in the migration decisions of rural-urban migrants in the developing countr ies. Future trends in such migration processes w i l l , in consequence, depend on such factors as: improvements in national economies; increasing population pressure on source areas; as well as changes in educational po l i c i e s . Other factors , which may be depressive in character.to rural-urban migrant f lows, are: increases in urban unemployment; the tendency for migrants to become permanent urban residents in order to ensure cont inuity in employment;, increases in rural employment opportunit ies, and f i n a l l y , d i rec t governmental po l i c i es that may be aimed at r e s t r i c t i ng the i n f l ux .o f rural dwellers into the c i t i e s . The impl icat ion then is that, any method directed at forecast ing r u r a l -urban migrant flows into c i t i e s of the developing countries must take due account of a l l the socio-economic and po l i t i c a l parameters. I t must be possible to introduce expected future changes in rural-urban inequa l i t ies *As quoted by Todaro, (1976). 36 (such as those measured by 'job p robab i l i t i e s ' and income d i f f e ren t i a l s ) into the ana lys is . This is pa r t i cu l a r l y imperative because of the highly s ign i f i can t ro le that rural-urban migration plays in urban growth in the developing countr ies. In urban population forecast ing, the migration vector, as suggested by Hightower (1968), may be obtained in two ways. The f i r s t involves a study of migrant flows in and out of the study area for a su f f i c i en t l y long period of time. The flows are then related to some var iable properties (job probab i l i t y , average income etc . ) of the urban area by means of regression analys is . In the second (cross-sect ional) approach, a number of urban des-t inat ions are sleeted and the flows between them and the rural or ig ins of the migrants are then analyzed for one or two periods to derive regression models (or equations) of rural-urban migrat ion. The time series approach to the project ion of migration accounts for only one aspect of the migration process - changes in the past volumes of migrants.. In the developing countr ies, however, i t i s not l i k e l y that past migration behaviour can be su f f i c i en t l y ind icat ive of future migrant be-haviour to j u s t i f y ; the use of time ser ies analysis to forecast migrat ion. In these countr ies, massive investments in agr icu l ture and industry can s i g -n i f i c an t l y a l t e r previous patterns of migrant movement i ns tan t l y ; espec ia l ly where such investments are labour- intens ive. It is for instance possible for the volume of migrant flow to a por t - c i t y to suffer a drast ic decline without changes in i t s economic a c t i v i t i e s , simply because.a new mine has been opened in the hinter land or because.a massive agr i cu l tura l project has been i n i t i a t e d . In both cases, labour"which should have migrated to the 37 por t - c i t y would be re-directed to the new areas of economic a c t i v i t y . A second d i f f i c u l t y with the t ime-series approach, to the forecast ing of migration i s that, i t does not provide for a complete understanding of the factors underlying the migration process, espec ia l ly those factors operating in the source areas. Examples of such cases, include the soc io-economic and demographic character i s t i cs of the potential migrants in the rural areas; and those related to urban-rural economic d i f f e r en t i a l s . In short the project ion of rural-urban migration based on past trends does not allow the integrat ion of the pul l and the push factors , (which have been shown to be important determinants of migrant movement in the developing countries) into the forecast ing process. A th i rd reason why the t ime-series approach to rural-urban migration forecast ing cannot be pa r t i cu l a r l y useful in the developing countries is the d i f f i c u l t y of obtaining the necessary h i s to r i c data on migrant f lows. F ewc i t i e s i n the world keep per iodic records of rural-urban migrant f lows. In th is regard, the cross-sect ional type of t ime-series analysis (-using point data); whose data requirements may be sa t i s f i ed through sample surveys con-ducted for the base-year;:may be more funct iona l , even though i t s usage i s not free from the foregoing l im i t a t i ons . Generally therefore, insofar as h i s to r i c migration.trends cannot be very ind icat ive of the effects of. massive and innovative changes which may occur in the project ion period on the volumes of rural-urban.migrant f lows; and insofar as an adequate understanding of the causative, var iables of r u r a l -urban flows i s essent ia l in forecast ing the migration vector, the time-ser ies technique seems to have very l i t t l e r e l i a b i l i t y as an e f f i c i en t 38 methodology in migration forecast ing. In the developing countr ies, h i s -t o r i c migration patterns are un l ike ly to have a bearing even on the imme-diate future; an assumption which i s implied in the use of t ime-ser ies. What matters instead, i s the future state of.rural-urban soci-economic and demographic d i f f e r en t i a l s . These comments are perhaps equally appl icable to most c i t i e s of the world; but the difference to be noted here i s that , where the migration element of urban growth is neg l i g ib l e , the need for more accurate methods of estimation become greatly reduced. For the develop-ing countries therefore, there i s need for a more-comprehensive migration forecast ing approach. 2.2.5 - CONCLUSION From the foregoing discuss ion, i t appears safe to conclude that there are many requirements to sa t i s fy in order to ensure that forecasts of urban population in the developing countries using the cohort-component approach are not so inaccurate as to be dangerously misleading. The t ime-series technique which i s known to be the simplest approach to forecasting the input variables of the cohort model is large ly based on the assumption that past trends can reasonably be expected to maintain the i r mo-mentum into the future. The i den t i f i c a t i on of regular trends from h i s to r i c data; for example decl in ing f e r t i l i t y and migration rates; may lead to the conclusion t h a t t h e yw i l l continue into the future.. However, the rates of change are not l i k e l y to remain the same over time. Where growth i s rap id, as in the developing countr ies, the po s s i b i l i t y of error in predict ing the future becomes great ly increased. Furthermore, the kinds of data which are 39 required for t ime-series analysis are r e l a t i ve l y d i f f i c u l t to come by in the developing countr ies. Although schemes for obtaining such s t a t i s t i c s by survey can be devised, they w i l l have to cover su f f i c i en t l y large samples to be of any value; in which case the i r costs may be p roh ib i t i ve . Theanswerto th is dilemma therefore seems to l i e in the use of po l i cy -oriented ( i . e . designed to take account of future changes in po l i c i es ) i n -stead of ana lyt ic (or regression) techniques. This can be done in a var iety of ways, inc luding a deta i led study of the indiv idual demographic components -f e r t i l i t y , morta l i ty , migration and so on; an invest igat ion of change i n d i -cators; and comparisons of experience. By studying the demographic components, a c learer picture can be ob-tained of how d i f fe rent factors contribute to the overal l population growth. At the f i r s t stage th is would simply be a breakdown of the urban growth into i t s components of natural growth and rural-urban migration as shown in Figure 2.2.5a. Further disaggregation of these factors can lead to progressive un-derstanding of the growth process.. Measures of f e r t i l i t y and morta l i ty can reveal the extent to which b i r th rates and death rates are dependent on struc-tural factors that w i l l change.as the urban area becomes more s t ab i l i z ed . I f some assessment can be made of the various contr ibutions to the growth rate in th i s way extrapolat ion into the future can be carr ied out with more con-fidence. Furthermore, i f knowledge of the demographic components can be supple-mented by information on what changes may be expected to take place, (as suggested in the c oho r t - f e r t i l i t y method discussed e a r l i e r ) , a stronger basis for estimating the inputs to the cohort-component model can be establ ished. 1 FRICTION FACTOR DESTINATION CONTACTS MINUS COST OF MIGRATION XT'- . URBAN JOB PROBABILITY —Z^+Z— URBAN/RURAL INCOME, WAGE, EXPENTURE DIFFERENTIALS T / / EDUCATION AND PROFESSIONAL ATTAINMENTS OF MIGRANT AGE, SEX-RATIO AND MARITAL STATUS OF MIGRANTS URBAN RATES FERTILITY AND TRENDS URBAN RATES SURVIVAL & TRENDS PUSH FACTORS RURAL-URBAN MIGRATION RATES! URBAN NATURAL GROWTH RATES URBAN POPULATION GROWTH RATES o Figure 2.2.5a The Components of Urban Population Growth in the Developing Countries -*• Affects Affects Ease of Movement Affects L ikel ihood of employment Affects Migrant's Perception of Urban/Rural D i f fe rent ia l s 41 Investigations of change indicators (or pre-condit ions for change), inc lud-ing governmental po l i c ies and publ ic values can be pa r t i cu l a r l y usefu l . These may appear in the att i tudes of publ ic agencies (e.g. family planning au thor i t i e s ) ; of parents toward family s i z e , and of the youth toward d i f -ferent professions. Even though the interpretat ion of such change ind ica-tors i s not l i k e l y to be easy, i t s completion w i l l provide a forewarning of the future. Data on demographic components and change indicators can be obtained in many ways, inc luding censuses and administrat ive records. However, the most important and f l e x i b l e source i s the special sample inquiry. Generally therefore, in spite of the foregoing def ic ienc ies connected with the use of the cohort-survival method of population forecast ing for c i t i e s in the developing countr ies; i t i s s t i l l possible to use the method, espec ia l ly i f a su f f i c i en t l y r e l i ab l e approach could be found to estimate the migration vector. 42a CHAPTER THREE T H E E C O N O M I C B A S E - M U L T I P L I E R M O D E L 42> SECTION ONE  A REVIEW OF THE ECONOMIC-BASE MODEL 3.1.1 - GENERAL THEORY In i t s simplest form, the theory of economic base states that the s ize and growth of the economy of an urban area depends upon the behaviour of a set of productive a c t i v i t i e s whose products are exported, (S iege l , 1967). This set i s ca l led the basic sector. Payments for those exports by part ies outside the study area, when translated into incomes for people and firms working in the basic (or export) sector, serve to support non-basic (or loca l ) a c t i v i t i e s - those directed at the demands of the inhabitants of the study area. Thus, an increase in exports w i l l stimulate growth in the basic a c t i v i t i e s , which w i l l then lead to increased basic employment. The increased income accruing to basic employees.through increased basic a c t i v i t i e s and employment w i l l , in turn, stimulate demand for non-basic goods and serv ices. These increases in both the basic and the non-basic employment, presumably through wage adjustments, w i l l lead to an i n f l u x of labour, so that the urban population ult imately increases. In short, the economic base-mult ip l ier methodology postulates that the tota l urban employment as well as the tota l urban population are predictable mult iples of the basic a c t i v i t y employment. When used for forecast ing pur-poses therefore, i t i s only necessary to: a. obtain a set of mu l t ip l i e rs (or rat ios) def ining the re lat ionship between each combination of the var iables - basic employment, non-basic employment, tota l employment, and tota l population; and 43 b. obta in, exogeneously, an estimate of future export demands for the basic industry products. Suppose that: Q(t) = tota l employment at time t , Qb(t) = employment in basic a c t i v i t i e s at time t , and Qs(t) = employment in the rest of the economy ca l led the non-basic sector, at time t . The economic base theory claims that Q(t) is proportional to Qb(t); so that Q(t) = m x Qb(t) (1) where m is a constant mu l t i p l i e r . Since Q(t) = Qb(t) + Qs(t); m = 1 + (Qs(t)/Qb(t)), (the mu l t i p l i e r ) (2) I f there is an appropriate evidence, i t may be desirable to make m = m(t), 3 a function of time . 3.1.2 - BASE AREA DELIMITATION The determination of the l im i t s of the urban economic base study area can have a s i gn i f i can t bearing on the d i rect ion of any resul ts obtained. I f the c i t y l im i t s are selected as the perimeter of the base area, the des-c r ip t ion of the base area thus obtained w i l l -be en t i re l y d i f fe rent from 4 what i t w i l l otherwise be i f a larger area were chosen... I t i s therefore important that the de l imi tat ion of. any economic base study area re f l ec t s the actual economic a c t i v i t y patterns of the metropolitan area or region. The base area de l imi tat ion process seeks to achieve th is purpose. As a s tar t ing point , i t may be necessary to examine the core* study area and i t s surrounding settlements in the l i gh t of the fol lowing tes ts : *The core study area i s used here to refer to the urban area of in terest only ( i . e . , excluding i t s sub-region). 44 "Where t h e sys tems o f t r a n s p o r t a t i o n b eg i n t o v e e r o f f t owa rd o t h e r m e t r o p o l i t a n c e n t e r s . "The r a d i u s s e r v e d by t he m e t r o p o l i t a n p r e s s and o t h e r a d v e r t i z i n g m e d i a . "The dependence o f o u t l y i n g f i n a n c i a l i n s t i t u t i o n s on t he c e n t e r f o r c l e a r a n c e s and r e s e r v e s . "Whether i t i s t h e c e n t e r f rom wh i ch t he r e t a i l e r s i n a b o r d e r l i n e town o b t a i n t h e i r s u p p l i e s . "Whether t he b o r d e r l i n e town i s i n dependen t o r dependen t upon t h e c e n -t e r f o r many o f t h e f o l l o w i n g f u n c t i o n s : 1. s t o r a g e f o r t h e c o n v e n i e n c e o f consumer , r e t a i l e r , w h o l e s a l e r , m a n u f a c t u r e r and s h i p p e r ; 2. whe the r t h e o u t l y i n g p r o d u c e r ma r ke t s d i r e c t l y t o t h e l o c a l c o n -sumer o r t h r ough the m e t r o p o l i t a n m a c h i n e r y ; 3. whe the r a b o r d e r l i n e communi ty communica tes by r a i l , t e l e p h o n e , e t c . , t h r ough t he c e n t e r o r i n d e p e n d e n t l y o f i t ; 4. where t h e b o r d e r l i n e town sends i t s s u r p l u s p r o d u c t f o r d i s p o s a l and s t o r a g e ; 5. whe t he r a f i r m o r i n d u s t r y w h i c h b o a s t s i t s i ndependence i n some one r e s p e c t , e . g . , t h e m a r k e t i n g o f i t s w a r e s , i s o r i s n o t d epen -den t upon t h e c e n t e r f o r i t s s u p p l i e s and f i n a n c e s ; and 6. whe the r t h e b o r d e r l i n e town i s t o o f a r . away t o a v a i l i t s e l f o f t h e c e n t r a l a s semb lage o f museums, t h e a t e r s , l i b r a r i e s , i n s t i t u t i o n s o f l e a r n i n g , and whe the r i t l o o k s f o r gu i d an ce i n f a s h i o n s , 5 t a s t e s and amusements" . 45 When th is kind of prel iminary survey of consumer and r e t a i l service establishments i s analyzed, a set of overlapping regions may be obtained as i l l u s t r a t ed in Figure 3.1.2a. There is no un iversa l agreement as to how the f ina l l im i t s of the local economic area are to be derived from th is set of overlapping regions. However, rel iance on inspection alone, per-haps aided by some weighting procedure, may lead to the i den t i f i c a t i on of a "consensus area", as i l l u s t r a t ed in Figure 3.1.2a. Whatever approach is used, attempts should be made to ensure that the f i na l i z ed l im i t s do not cut across s t a t i s t i c a l un i ts , so that standard sources of data may be used in the economic analyses. 3.1.3 - BASIC - NON-BASIC ACTIVITY -CLASSIFICATION Given the del imited urban economic area, the c l a s s i f i c a t i on of the tota l urban economy into basic and non-basic sectors s tar ts with a st ruc-tural breakdown of the economic a c t i v i t i e s . These structura l d i v i s i ons , which are normally formalized in a standard indust r ia l c l a s s i f i c a t i on sys-tem include for example: agr i cu l tu re , mining, construct ion, manufacturing, t ransportat ion, services and government. Each of these d iv id ions may again be broken down into major subgroups. For example, the broad group labe l led "serv ices" , may be subdivided .into hote ls , auto-repair , legal and medical. Most economic base studies adopt a two-level structura l , c l a s s i f i c a t i o n , even 6 though further sub-c las i f i ca t ions are possible . Beyond th i s stage of the c l a s s i f i c a t i on process, the next step i s de-termined by local object ives, problems, data a v a i l a b i l i t y and budget. The essent ia l quest ion, however, i s what c r i t e r i a must be used to separate the various structura l d iv i s ions in to .bas ic and non-basic sectors. Five ap-Figure 3 .1 .2a: Base Area Del imitat ions Using Three Variables FOR ILLUSTRATION ONLY 47 proaches of varying r e l i a b i l i t y are often used for th is purpose. These include the (1) Assumption Approach., (2) Location Quotients, (3) Minimum Requirements, (4) Direct Invest igat ions, and (5) The Dominants Approach. 1 T The Assumption Approach: This i s the oldest and the least r e l i^ able of the f i ve approaches. I t operates on the assumption that, broad economic sectors or d iv i s ions such as manufacturing, wholesal ing, and mining, are invar iab ly export (or basic) in nature whereas, r e t a i l i n g , government, education, and services are invar iab ly oriented toward the local market. It does not examine major groups within the d i v i s i on ; and i t does not recog-nize the fact that any one economic d iv i s ion may have major groups that are e i ther export or l o c a l . " 2 - The Location Quotient Approach: The c l a s s i f i c a t i on of the tota l urban economy into basic and non-basic sectors by the locat ion quotients approaches based on the assumption that national demand patterns are uniform. It therefore proceeds to argue that i f a certa in urban industry group ac>-counts for a greater percentage of. the c i t y ' s labour force than the same i n -dustry group does at the national l e v e l ; the excess percentage i s a t t r i bu t -able to the export sector. Tiebout (1962)* presents a formula for the locat ion quotient computa-t ion which i s as fo l lows: X National Employment in X Total Urban Employment ~ Total National Employment where X represents the major industry group employment which i s being s p l i t . A l te rna t i ve l y , the quotient may be computed by d iv id ing the urban ra t io (or percentage), by the national r a t i o . *As quoted by Andrews, B. Richard (1968). 48 Thus, by examining each of the various industry groups in th is man-ner, i t i s possible to produce a basic-nonbasic c l a s s i f i c a t i on of the urban economy, when the indiv idual employment categories are summed up. 3 - The Minimum Requirements Approach: This approach compares the c i t y being studies with a large number of other c i t i e s which are judged to be s im i la r t o . i t in some way. For each c i t y then,, a computation i s made of the percentage share which, each economic a c t i v i t y claims of the c i t y ' s tota l labour force. The technique next assumes .that, the lowest percentage share (among the universe of c i t i e s ) exhibited by each a c t i v i t y type indicates the min i -mum requirement of loca l demand, for,.the par t i cu la r good or service pro-duced. This minimum then serves as the benchmark for determining the de-gree of excess which must be assumed to represent export oriented employ-ment. Summation in employment terms;:of the mini mums and excesses; indus-try by industry, w i l l produce the tota l basic - nonbasic c l a s s i f i c a t i on de-s i red for the c i t y . 4 - Direct Invest igat ion: Accuracy in the basic - nonbasic c l a s s i f i c a -t ion of the urban economy, at any point in time can only be achieved by d i rec t invest igat ion^. This i s best accomplished by a combination of ques-t ionnaires and personal interviews. Small c i t i e s are the typ ica l users of th is approach, while large c i t i e s tend to use the der ivat ive ind i rec t d iv ices .out ! ined above. Besides the apparent fact that greater precis ion of c l a s s i f i c a t i on i s possible with the d i rec t . inves t iga t ion approach, the method also permits the co l l ec t ion of supplementary information; such as growth trends and 49 future devlopment plans for the various a c t i v i t y groups; which may aid in the forecast of economic a c t i v i t y , employment and population. 5 - The Dominants Approach: Essent ia l l y , th i s method i s a spec ia l i zed d i rec t sampling approach to the c l a s s i f i c a t i on process. I t extracts from the context of:the tota l urban economy those economic a c t i v i t i e s which dom-inate the economy. Dominance as used here i s a subjective concept. Common-l y , however, dominant economic a c t i v i t i e s are taken to be those which account for a s ign i f i can t amount of employment within the c i t y . Other possible mea-sures of dominance include payrol l volume, sales volume, plant investment, and s imi la r factors . Once the dominant economic a c t i v i t i e s are thus extracted., they are again r e - c l a s s i f i ed according to the i r degree of dominance. For example those a c t i v i t i e s which export 80% or more, of the i r products may be c l a s s i f i e d as heavy-exporters; while those exporting between 79% - 55% of the i r products may be c l a s s i f i e d as l ight -expor ters . In th is par t i cu la r example, any f irm that exports less than 55% w i l l therefore be considered a non-basic a c t i v i t y . Which of the above f i ve methods i s used for the functional c l a s s i f i c a -t ion of the urban economy depends en t i r e l y upon the object ives of the study, the budget avai lab le to the analyst, and the volume of ex i s t ing relevant i n -formation. "Data problems, . . . , are so universal, that they are usual ly 4 taken as g iven." 3.T.4 - LIMITATIONS OF THE BASE MODEL The economic base theory, as an urban planning t oo l , may be used to obtain aggregate forecasts of the various parts of the economy. Guided by the mu l t i p l i e r , ( tota l employment/basic employment), and a system of r a t i o s , 50 the forecast basic employment i s then manipulated to obtain a tota l pupula-t ion forecast for the project ion year. The system of rat ios normally used for such forecasts include: Basic Employment:, Nonbasic Employment Basic Employment: Total Employment (the mu l t i p l i e r ) Total Employment: Total Population Basic Employment: Total Population Generally however, the base theory i s c r i t i c i z e d for a number of d i f f i c u l t i e s inherent in i t s app l i ca t ion . These include d i f f i c u l t i e s of base measurement and assumptions regarding ra t io s t a b i l i t y . 1 - Base Measurement:. On putt ing the base theory to pract i ce , extreme importance must be attached to ident i fy ing the basic employment cor rec t l y . I f for example some basic a c t i v i t y escapes i den t i f i c a t i on , the mu l t i p l i e r w i l l be too large; (ref. equation 2); since Qs(t) w i l l be too large, and Qb(t) w i l l be too smal l . Ult imately therefore, forecast tota l employment and tota l population w i l l also be too large. The converse of th is system of errors i s also poss ib le. But there are a number of d i f f i c u l t i e s which reduce the r e l i a b l i t y of the f i ve base c l a s s i f i c a t i on methods discussed above. F i r s t , the urban economy may. have businesses that are actua l ly ind i rec t exporters, but which, by c l a s s i f i c a t i on may be labe l led "nonbasic". There i s l i t t l e doubt that the locat ion quotient method for example i s very un-sat i s factory for estab l i sh ing the true volume of basic a c t i v i t i e s .^ Values produced by th is method may be affected by d i f f e ren t i a l tastes between the study area and the region; by d i f f e ren t i a l product iv i ty and perhaps other 51 di f ferences. For th is reason, many invest igators supplement the use of g locat ion quotients with judgement. Secondly, the economic base technique, with i t s fundamental theoret ica l stress on exports and internal transact ions, ignores almost completely the import side of the economy*. The possible impact of th i s def ic iency on the operation of the method i s an over-estimation of the non-basic a c t i v i t i e s , tota l employment, the mu l t i p l i e r , and, u l t imate ly , the tota l population of the urban area. Th i rd ly , employment i s the.commonest measure of economic a c t i v i t y used by the base technique,.general ly because data is more complete for i t than could be obtained for other possible measures such as output and value added. Both short and long term errors may, however, ar ise when employment is used as the unit of measurement. If i t i s assumed that changes in employment w i l l pa ra l l e l changes in output, (which i s implied in the use of employment as a unit) there i s an imp l i c i t presumption that firms operating in the ur-ban economy do not experience diminishing returns. I f on the other hand the firms do in fact experience diminishing returns, as i s most probably the case, then increases in demand for export goods and services w i l l bear a non-l inear re lat ionsh ip to employment as a factor of production, instead of the d i rec t re lat ionsh ip commonly assumed by the base technique. Further-more, the method also ignores the po s s i b i l i t y that, in the long run, increases in demand for both basic and nonbasic products may be par t ly catered for by changes in product iv i ty and production technology; implying a disproport ion-ate change in employment. *(-by assuming that a l l goods consumed in the economy are produced l o c a l l y ) . 52 2 - Base Ratio S t ab i l i t y : One other fundamental assumption of the base theory i s that the base ra t io (or Basic Employment/Total Employment) stable over the project ion per iod. But Andrews (1968, p. 98) has establ ished that, apart from the po s s i b i l i t y of the base rat io changing with changes in pro-duc t i v i t y , and production technology; i t also does depend on changes in c i t y s ize and density of development; which are themselves d i f f i c u l t to predict without a knowledge of future tota l employment and population. What i s ca l led fo r , then, i s a base theory that y ie lds a curve re la t ing the s ize of economic base to s ize and density of the urban area. For forecast ing pur-poses, i t i s not su f f i c i en t to jus t know that the base w i l l change with the c i t y s ize and density. A knowledge of the magnitude of change is also essen-t i a l . In view of these def ic ienc ies ". . . , even i f the weaknesses of the i n i t i a l assumptions are overcome, there are a number of serious l im i ta t ions 9 involved in the use of th i s concept as a forecast ing device." In conclusion therefore, "The economic base technique i s . a device which i s very far from per fect ion, both in conception and in app l i ca t ion . Div is ion of a complex piece of urban economic machinery into two, and at most, three parts for purposes of descr ipt ion and ana lys is , i s r i gh t l y considered an over -s imp l i f i ca t ion l i k e l y to produce crude and inaccurate resul ts ."^ 53 SECTION TWO 3.2.1 -BACKGROUND: THE URBAN ECONOMIC STRUCTURE The "dua l i s t i c " nature of urban economies.; in the developing countries have been the focus of many researches over.the past decade; (examples i n -clude: Hart, 1973; Blandy and Wery, 1973; Weeks, 1975). According to the concept, of dualism, the urban economy i s d ist inguishable into two sectors; namely the formal, and the informal sectors. This c l a s s i f i c a t i on into fo r -mal and informal sectors is based on the organizational character i s t i cs of exchange re lat ionships and the posi t ion of the economic a c t i v i t y v i s -a -v i s the state; (Weeks, 1975). In th is sect ion, an understanding of the formal-informal'concept i s f i r s t developed before proceeding to discuss i t s imp l i -cations for possible uses of the economic base model in the. developing coun-t r i e s . i . The .'Formal Sector: Bas i ca l l y , the formal sector includes government a c t i v i t y , and those enterprises in the pr ivate sector which.are o f f i c i a l l y recognized and regu-lated by the state. Almost a l l the enterprises in th is sector run with some level of. bureaucracy and are therefore amenable to enumeration by surveys. They are characterized by r e l a t i v e l y large-scale operations, cap i t a l - i n ten -sive techniques and re l a t i ve l y stable wage rates. These production features imply that th is sector produces for national and/or internat ional markets; so that the level of employment, in the formal sector i s necessar i ly i n f l u -enced by consumption factors which are both internal and external to the urban area in which,they are located.. In conventional terminology, there-fore, the formal sector economic a c t i v i t i e s may be par t ly basic and part ly non-basic; (Sethuraman, 1976). 54 i i . The Informal Sector: The informal sector consists of enterprises operating outside the system of benefits (e.g. c red i t f a c i l i t i e s ) and regulation by government. In most developing countr ies, the economic agents in th is sector operate " i l l e g a l l y " , though they may be pursuing economic a c t i v i t i e s s im i la r to those in the formal sector. I l l e g a l i t y , then, i s not usual ly a consequence of the nature of the a c t i v i t y but a consequence of o f f i c i a l r e s t r i c t i ons on reg i s t ra t i on , locat ion and operation;.(Weeks, 1975). Because of i t s l im i ted s k i l l requirements and the f ree ly competitive nature of i t s product and factor markets; access to the informal sector by prospective investors i s generally f a i r l y easy. As a resu l t , the sector is characterized by a multitude of smal l-scale enterpr ises, operating v i r t u a l l y without capi ta l or\formal organization and using very simple technology; (Souza and Tokman, 1976). This in turn, implies that production tin the informal sector uses large ly local inputs and caters mostly for the low-income urban market. The a c t i v i t i e s in th i s sector often include smal l -scale manufacturing, transportat ion serv ices, construction and trade. Conventionally therefore, the informal sector of the urban economy in the developing countries i s essent ia l l y a non-basic economic a c t i v i t y ; (Sethuraman, 1976). In short, the urban economies of c i t i e s in the developing countries are dua l i s t i c both in terms of employment and economic operations. In the fo r -mal, sectors, which may be basic or non-basic, r e l a t i ve l y high wages and sa lar ies are paid. As a resu l t , labour, is avai lab le in. excess supply to these sectors. Growth in employment i s therefore governed by growth in 55 demand for labour; which is in turn a function of changes in output leve ls as well as production technology. In the informal sector, on the other hand, i n s t i t u t i ona l forces com-bined with minimal cap i ta l and s k i l l requirements provide an economic frame-work which permits the absorption of a l l avai lab le labour into productive a c t i v i t i e s , even at very low leve ls of product iv i ty and income. Hence, employment there is determined' not only by the demand for labour, but also by the supply of labour in the urban economy as a whole; less the labour demand of the formal sector. These patterns of economic operation explain why the informal sector i s var iously referred to in the l i t e ra tu re as; 'the low product iv i ty urban sec to r 1 , 'the reserve army of unemployed and under-employed', 'the urban t rad i t iona l sector ' and so on; (Hart, T973, p. 68). The rest of th i s chapter examines the probable impl icat ions of the dua l i s -t i c nature of urban economies in the developing countries for the use of the economic base theory as a forecast ing device. 3.2.2 - THE BASE-RATIO STABILITY ASSUMPTION One fundamental assumption underlying the use of the economic base theory for predict ion purposes i s that the base-rat io, (or Basic Employ-ment/total employment), w i l l remain stable over the project ion period. I f during th is per iod, however, the ra t i o of basic employment to tota l employment undergoes a change, then the projections of tota l employment and population, u t i l i z i n g a mu l t i p l i e r derived from the i n i t i a l ra t io w i l l be inaccurate. For example, taking equation (2) in section 3.1.1, i t can be deduced that: 56 m = Qb(t) + Qs(t) . QbTt) so that 1/m = Qb(t) Qb(t) + Qs(t) = Basic Employment = Base Ratio Total Employment Expressed in th i s form, i t becomes obvious that.the s ize of the mul-t i p l i e r is influenced by the re lat ionsh ip between basic and tota l employ-ments. In other words, the value of the mu l t i p l i e r can be seen to be ar i thmet ica l l y l inked to the proportion of tota l employment, that i s bas ic. The important ins ight given by Figure 3.2.2a i s the extent to which the mu l t i p l i e r can be sens i t ive to changes in the value of the base r a t i o . Where the base ra t io is greater than 0.5 any changes in i t s value w i l l produce r e l a t i ve l y small' changes in the mu l t i p l i e r ; and vice versa. i . Low Base Rat ios: Evidence from a ser ies of research studies conducted by the International Labour Organization, (IL0), on the employment potent ia ls of the informal sec-tor , (which in most cases i s about 95 per cent non-basic)., in selected A f r i -can countries show that.the informal urban economic sector i s an important source of employment in many Afr ican countr ies, and pa r t i cu l a r l y in c i t i e s where, (as in the case of Kumas.i) i t provides work for some 60 to 70 per cent of a l l employed persons, (Sethuraman, 1977). In contrast , Singer (1971, p. 31) noted that the formal sector employ-ment in the developing countries has, over the years, f a i l ed to increase in 57 BASE RATIO = BASIC EMPLOYMENT TOTAL EMPLOYMENT MULTIPLIER = TOTAL EMPLOYMENT BASIC EMPLOYMENT BASE RATIO X MULTIPLIER = 1 Figure 3.2.2a: Graph Showing the Sens i t i v i t y of the Mu l t i p l i e r to Changes in the Base Ratio 58 proportion to the demand for jobs. That i s , indust r ia l employment in the formal sector i's;not increasing at the. rate of 7-8 per cent often observed for indust r ia l production in the developing countr ies. Instead; he cont in-ued; i t increases at a far lower rate of 3%. Since tota l urban populations in these countries are increasing at rates close to 5% per annum, th is means that the bulk of the increase in urban labour force i s absorbed by the in fo r -mal sector. One conclusion that c lea r l y emerges from the above discussion is that the predominance of the informal sector over the formal one in terms of em-ployment necessar i ly places the base ra t io of c i t i e s in the developing coun-t r i e s in the lower ranges, (see for example World Bank, 1972; and Hart, 1973). Even i f i t i s generously assumed that the informal sector accounts for 60 non-basic jobs out of every 100 tota l urban employment; and that the formal sector contributes a further 10%. to th i s r a t i o , then.the base ra t io of c i t i e s in the developing countries cannot be greater than 30%. As a re-su l t of th is low base ratio.,(0.3 or less) therefore, any economic base mul-t i p l i e r derived for c i t i e s in the developing countries w i l l be overly sen-s i t i v e to small changes in the i r base ra t i o s ; as i l l u s t r a t ed in Figure 3.2.2a. i i . Unstable Base Ratios: As borne out by empirical studies conducted by the ILO, the informal sector, in sp i te of the d i f f i cu l t i e s .unde r which i t operates, has also been growing over.the years. The number of informal enterpr ises, according to the ILO study, has been increasing by about 12% per year in Kumasi, and 25%peryear in Freetown; (Sethuraman, 1977). 59 Secondly, Gabler (1971) also argues that for any c i t y , the base ra t io i s l i k e l y to undergo changes as a resu l t of increases in the s ize of urban population. He explained that, growth in c i t y s ize may lead to increasing employment e i ther because basic economic a c t i v i t i e s can then take advantage of the accumulated pool of s k i l l e d manpower; or simply because the increase in population has made possible an expanded and more varied non-basic sec-tor . Generally therefore, considerable i n s t a b i l i t i e s in the urban economic base of c i t i e s in the developing countries are to be expected; f i r s t , be-cause the i r base rat ios are invar iab ly low (0.3 or l e s s ) ; and secondly, because rapid increases in the i r informal sector employment (betwen 12% -25% per year) as well as tota l urban population (growing at about 5% per year) w i l l i n i t i a t e changes in the i r base ra t i o s . 3.2.3 - THE BASE THEORY AS A POPULATION .FORECASTING DEVICE S imi lar to the base ra t io s t a b i l i t y not ion,,the base theory also assumes that there is a constant.proportionate re lat ionship between the tota l urban population and the number of jobs that are ava i lab le to the c i t y . That i s , for example, i f Population in 1970 = b x Employment in 1970, then, Population in 1980 = b x Employment in 1980. When the base theory is used in population forecast ing therefore, i t i s only necessary to obtain an estimate of tota l future employment; s ince, according to the theory, the re la t ion (or e l a s t i c i t y ) b, between population and employ-ment w i l l remain constant over the predict ion period. Imp l i c i t in th i s population/employment ra t io s t a b i l i t y assumption i s the presumption that the project ion star ts from a point where there i s f u l l 60 employment in the urban economy; so that the supply of labour in response to the demand for i t w i l l almost always or ig inate from outside the c i t y boundaries. For c i t i e s in the developing countries however, the notion of f u l l employment i s hardly appl icab le. Scattered evidence suggests that unem-ployment*!^ the urban areas of these countries is more often about 10% than below; and that, in some major c i t i e s , unemployment* leve ls above the 30% mark are poss ib le; (World Bank, 1972). This being the case, i t can be expected that any increase in demand for labour in the urban economy w i l l most probably be f i l l e d up by those who are current ly unemployed in the urban economy rather than by persons migrating from the rural areas. The obvious conclusion then is that tota l urban population in these countries w i l l not necessar i ly bear a proportionate constant re lat ionsh ip to changes in urban labour demand. One could of course retor t that , as the unemployed urban residents be-come absorbed into the urban economy, the urban job probab i l i ty for poten-t i a l rural-urban migrants would increase; so that, u l t imately (or in the long-run) the urban population would be increasing in response to the i n -creased demand for urban labour (or increased job p robab i l i t y ) . While th is argument appears reasonable enough, i t i s however weakened by the point already made in subsection 2.2.4 of th is study that, apart from purely economic motivations (of which job probab i l i ty i s one), r u r a l -urban migration in the developing countries is also influenced by certain demographic and socia l fac tors; (re: section 2.2.4). In th i s regard, i t i s un l ike ly that population would remain a constant mult ip le of tota l em-*The term 'unemployment' as used here does not include underemployment. 61 ployment over the years. A second reason why urban population in the developing countries i s l i k e l y to increase independently of the demand for labour i s the almost unlimited a b i l i t y of the informal sector to continue absorbing whatever^ labour becomes avai lab le in the urban economy; even in the face of f a l l i n g product iv i ty and dec l in ing wages, (Sethuraman, 1976, p. 75). In short, even though changes in tota l urban employment do have a casual e f fect on urban population growth in the long-run, the re lat ionship between the two var iables cannot be one of constant propor t iona l i ty . The population/ employment ra t io i s necessar i ly influenced by a complex set of economic, demographic and soc ia l factors-; a l l ~of which may be operating in d i f fe rent d i rect ions at any one given time. The notion that population w i l l remain a constant mult ip le of tota l employment therefore f a l l s very short of what i s l i k e l y to be the case in the developing countr ies. 3.2.4 - PROBLEMS INVOLVING BASE MEASUREMENT: It is extremely important, in .put t ing the economic base theory to work, to measure the base ra t io with, reasonable accuracy. I f the re la t ion of basic employment to tota l employment i s not.measured accurately enough, a wrong value of the mu l t i p l i e r w i l l be obtained, pa r t i cu l a r l y in cases when the non-basic sector predominates over the basic sector. However, a number of conceptual and pract ica l problems often ar ise when attempts are made to measure the base ra t io of c i t i e s in general. Em-ployment data published in censuses, for example, are usual ly found to be ser ious ly lack ing in most of the deta i led information often required in 62 economic base ana lys i s . They provide no. measure of what i s basic and what i s to be termed non-basic economic a c t i v i t y . Secondly, they do not pro-vide the kind of information often needed to define urban economic regions in terms o f the geographic areas within which the c i t y ' s goods and services are so ld. In spi te of these l im i ta t i ons , however, census data s t i l l remains as one of the essent ia l sources of v i t a l inputs to urban economic base studies. i . S t a t i s t i c a l Base: The commonly known problem of inadequate s t a t i s t i c a l base in the develop-ing countries i s further aggravated, in urban economic ana lys i s , by the ex i s -tence of the informal sector. As has already been stated at the beginning of th is sect ion, a substantial proportion of urban workers in the developing countries are employed in smal l , own-account, mostly unregistered enterpr ises. It should further be recol lected that a disproport ionately large share of the additions to the urban labour force, resu l t ing mainly from rural-urban migrat ion, tend to be absorbed in such small enterpr ises. I t i s therefore very d i f f i c u l t to ident i fy the population employed in the informal sector and c l a s s i f y i t into basic and non-basic informal employments. In an attempt to f i l l in th i s s t a t i s t i c a l gap, an ILO urbanization and employment research project , in 1976, i n i t i a t ed a number of studies on the informal sector in selected major towns of the developing countr ies. The concepts and de f in i t i ons involved, and the methodology followed are d i s -cussed in an a r t i c l e published in the International Labour Review by Sethuraman, (1976). 63 i i . Measurement Techniques: The f i ve commonly used methods of basic/non-basic a c t i v i t y measurement have already been out l ined in sub-section 3.1.3. The essent ia l question raised here concerns the relevance of each of these methods to the pecu-l i a r charac ter i s t i cs of the urban economies in the developing countr ies. Apart from the theoret ica l l im i ta t ions of these methods, the i r use in the developing countries is l i k e l y to suffer from other de f i c ienc ies . To as-sess the nature of such de f i c i enc ies , however, factual quant i tat ive in fo r -mation w i l l inev i tab ly be needed on the re lat ionsh ip between estimates of basic and non-basic employments produced by each of the f i ve techniques; and the i r corresponding actual counts. Even i f such information could be obtained for the purposes of th is study, considerable conceptual and pract i ca l problems concerning d e f i n i -t ion and methods of measurement would most probably be encountered in i so -l a t i ng and interpret ing the re la t ionsh ips . Fortunately, however, such re lat ionships have already been quant i ta-t i v e l y invest igated by Brodsky and Safarty, (1977). The general picture that emerges from the i r assessment of the relevance of the various methods of base measurement to urban economic structures in .the developing coun-t r i e s i s that , as an ind i rec t method of measuring the economic base of c i t i e s , the minimum requirements technique has several advantages over the other three ind i rec t methods. In comparison with the locat ion quotient for example, the minimum requirements technique measures gross rather than net basic a c t i v i t y . A lso, unl ike the locat ion quotient, the minimum tech-64 nique uses the aggregate indust r ia l c l a s s i f i c a t i on so that i t s data re-quirements are r e l a t i v e l y modest. Brodsky and Sa fa r ty cautioned tha* neither the locat ion quotient nor the minimum method w i l l automatical ly adjust for sectoral var iat ions in income as commonly known to ex i s t between the formal and the informal sectors charac te r i s t i c of the c i t i e s in the developing countr ies. In such a case, a s ingle average economic base mu l t i p l i e r derived from the min i -mum or locat ion quotient technique i s also un l ike ly to r e f l e c t the income differences that may ex i s t between the two sectors. Several indiv idual sectoral mu l t ip l i e r s may therefore have to be derived by sca l ing the s i n -gle average mu l t i p l i e r upward or downward. "There i s probably no simple way to do th i s accurately, although i t i s reasonable,to assume that the re l a -t ive strength of a sectoral employment mu l t i p l i e r i s related to wage l eve l s " , (Brodsky and Safarty, 1977). In short, the d i f f i c u l t i e s , involv ing any attempt to measure the econo-mic base of c i t i e s in the developing countries are twofold. The f i r s t de-r ives from the imp l i c i t weaknesses of the s t a t i s t i c a l bases ava i lab le in these countr ies; which are further worsened.by the existence of the in fo r -mal sector. Secondly, the i n ab i l i t y of the minimum requirements technique (which has been, seen to be the most relevant of the four ind i rec t methods of base measurement), to y i e l d mu l t ip l i e r s which re f l e c t sectoral income differences const i tutes a further drawback on the use of the economic base theory in the developing countr ies. These def i c ienc ies , coupled with the fact that the base ra t io of c i t i e s in the developing countries often possess 65 base rat ios less than 30% imply that extremely wide margins or error should be expected in the derived mu l t i p l i e r s . 3.2.5 - CONCLUSION The assumptions of the base theory regarding the base r a t i o , and the population/employment ra t io s t a b i l i t i e s have been found to deviate s i gn i f i c an t l y from what appears to be the s i tuat ion in the developing countr ies. The reasons for th i s include the dua l i s t i c nature of the urban economies, the predominance of the non-basic over the basic employment, the r e l a t i v e l y fas t rates of urban population growths, and f i n a l l y , the continued persistence of high unemployment leve ls in c i t i e s of the de-veloping countr ies. In addit ion to these de f i c ienc ies , there are certa in pract ica l problems mainly involv ing the inadequacy o f t h e avai lab le s t a t i s -t i c a l bases as well as the base measurement techniques, which when taken together, w i l l d r a s t i c a l l y reduce the potent ia l of the base theory as a planning tool that could be used in the developing countr ies. General! therefore, any use of the base theory in i t s present formula-t i on , as a population forecast ing device in the developing countries should be subject to empirical research to ascertain i t s relevance to local condi-t ions . A s im i la r but more general view i s shared by Needham, (1974, p. 215), who advises against any attempt to bui ld general theories of the urban eco-nomy. 66a CHAPTER FOUR T H E G R A V I T Y M O D E L S 66b SECTION ONE  A REVIEW OF THE GRAVITY MODELS 4.1.1 - CONCEPTUAL FRAMEWORK The gravity model concept as commonly used in urban planning i s para l le l to Newton's Law of Gravitat ion which states that two bodies in the universe a t t rac t each other in proportion to the i r mass and in inverse proportion to the i r distance apart. The urban planning analogy i s that any two zones in a c i t y or metropolitan region interact, with each other in proportion to the i r s ize and in inverse proportion to the i r distance. Three main pr inc ip les thus underly the concept of gravity (or spat ia l interact ion) models of urban planning. ( i ) That the interact ion between an or ig in and a dest inat ion zone increases as the 'populat ion' (measure of t r i p generation) of the or ig in zone increases, ( i i ) That the interact ion between an or ig in and a dest inat ion zone i n -creases as the opportunit ies in the dest inat ion zone increases, ( i i i ) That the interact ion between the o r ig in zone and the dest inat ion zone decreases as the d i f f i c u l t y of travel between the zones i n -creases . In pract i ce , var iat ions in these assumptions produce a family of four gravi ty models - product ion-attract ion constrained; production constrained; a t t rac t ion constrained; and the unconstrained versions of the model. In planning the unconstrained version i s rare ly usedl^; and w i l l therefore not be elaborated upon in th is paper. 67 4.1.2 - PRODUCTION-ATTRACTION CONSTRAINED MODEL This type of gravity model i s used when the or ig in a c t i v i t y produc-ing the demand; and the dest inat ion a c t i v i t y a t t rac t ing the demand are both known. What i s required therefore i s to predict the resu l t ing i n -teract ion between the two zones. I t is commonly used in the Tr ip D i s t r i -bution stage of the transportat ion planning process. FORMULATION The gravity model of t r i p d i s t r i bu t ion is based on the hypothesis that t r i ps produced at an or ig in and attracted to a dest inat ion are d i rec t -ly proportional to the tota l t r i p production of the o r i g i n , the tota l t r i p a t t rac t ion at the dest inat ion, a ca l i b ra t ing term, and, possibly a soc io-economic adjustment factor . This re lat ionsh ip may be expressed as fo l lows: T i j =Ci;PiAjFijKij (1) where, T i j represents t r i ps produced at i and attracted at j ; Pi i s the tota l t r i p production at i ; Aj i s the tota l t r i p a t t rac t ion at j ; F i j i s the ca l ib ra t ing term for the interchange i j to represent some measure of the d i f f i c u l t y of t r ave l . K i j i s the socio-economic adjustment factor for interchange i j ; to represent the ef fect on travel patterns of socia l and economic character i s t i cs which are not otherwise accounted for in the model. C i s a constant of propor t iona l i ty , i i s an o r ig in zone number, i = 1, 2, . . . n j i s a dest inat ion zone number, j = 1, 2, . . . n n i s the tota l number of zones. 68 Since the sum of a l l T i j ' s for or ig in i must be equal to P i , i t fol lows from equation 1 that; n n Pi =2L T i j (.Ci Pi A j F i j K i j ) . (2) j =1 j = ] n = c i p i 2Z , (A j F i j K i j ) j = 1 So that Ci = 1 (3 ) n 21-, (A j F i j K i j ) j = 1 Hence, from equations 1 and 3 T i j = Pi A j F i j K i j (4) n _ (A j F i j K i j ) 3 which is the standard gravi ty model for t r i p d i s t r i bu t i on . Or ig ina l l y , the gravity model of t r i p d i s t r i bu t ion used distance to measure the d i f f i c u l t y of travel between two zones. This was on the assump-t ion that the further away a centre was, the less a t t rac t i ve i t w i l l seem to people wanting to go there. However, recent formulations have used travel time to measure the d i f f i c u l t y of t r ave l . There i s as yet i n s u f f i -c ient evidence to spec i f i c a l l y state that time i s the c r i t e r i on by which people judge the d i f f i c u l t y of t r a v e l . 1 1 There have been some attempts to use travel costs as the factor of impedance. I f used, costs should i dea l l y 6 9 include those due to travel time, parking and fuel in pr ivate transport, and fares and wait ing time in publ ic transport. 4.1.3 - ATTRACTION .CONSTRAINED MODEL The at t rac t ion constrained gravi ty model i s used when the locat ion of the dest inat ion zone or a c t i v i t i e s is known; (e.g. employment oppor-t un i t i e s ) ; and the locat ion of the or ig in zone or a c t i v i t y is required; (e.g. res ident ia l populat ion). It i s commonly used in simulating the locat ion of urban res ident ia l a c t i v i t i e s . FORMULATION The gravity model of res ident ia l locat ion assumes that people seek to l i v e in zones c losest to the i r work place. In th is regard, the l i k e l i -hood of bui ld ing in a zone is large ly a function of the zone's distance or travelt ime from places of employment or commerce. In. i t s most widely used form ( i . e . the Hansen Formulation); i t states that: ' i =. 1 where; Gj i s the res ident ia l growth increment in zone j ; Gt i s the exogenously determined tota l res ident ia l growth increment in. the c i t y being studies Li&Lj are measures of the amount of land avai lab le in zones i and j respect ive ly Ai&Aj are indices of a c ce s s i b i l i t y of zones i and j , respect ive ly , to a l l other zones. (5) 70 The a c ce s s i b i l i t y index most often used i s of the form n Aj =2Z (Ei / D*ij) i = 1 where; Ei i s t h e s i z e o f employment in zone i D i j is the distance (or some other form of travel impedance) between i and j . *is a parameter value to be determined by ca l i b r a t i on . Thus, the index, A j , i s an index of a c ce s s i b i l i t y to jobs; subject to the constraint that n LIMITATIONS "The model makes good sense - at l eas t , in terms of indiv idual beha-viour. People do make trade-off between land and travel costs, at least in theory, but there is good reason to believe that, many other things such as schools, neighbours, recreation, access and urban renewal also i n -fluence locat ion dec is ion, and these factors.are poorly explained by the 12 gravity model". 4.1.4. - PRODUCTION CONSTRAINED MODEL The production constrained model i s used when the or ig in a c t i v i t y i s known, (e.g. population) and i t i s required to locate the dest inat ion zone Gt j = 1 71 It i s commonly used in marketing studies to ident i fy potent ia l r e t a i l growth centres o r to estimate expenditure in r e t a i l centers. FORMULATION As formulated by Huff (1963)6^ the amount of business attracted to a r e t a i l ' centre ( j ) , from a res ident ia l zone ( i ) can be expressed as: Si j = Ei . Aj /d* i j (60 n (Aj / d*i j ) 1 J = where, S i j i s the amount of r e t a i l sales attracted to centre j from zone i ; Ei i s tota l r e t a i l expenditure in zone i Aj i s the attract iveness of centre j . d . . i s the distance from zone i to centre j "13 * i s a parameter value which i s to be estimated empir ica l ly to r e f l e c t the ef fect of travel distance, time, or cost on shopping t r i p s . Subject to the constra int that n n S i j j = 1 i = 1 LIMITATIONS Retai l models have frequently been questioned about why they f a i l to ©Quoted by McCabe(1974) 72 account for a l l factors that are known to af fect centre sales. The repu-tat ion of a par t i cu la r r e t a i l e r for example, may have a benef ic ia l e f -fect on sales in the centre; i f that r e t a i l e r locates in the centre. S im i l a r l y , the drawing power of a r e t a i l centre may be influenced by geo-graphic features, as well as the socio-economic charac ter i s t i cs of custo-mers. Secondly, r e t a i l models are macro^level.in nature. They do not attempt to predict the behaviour of ind iv idual customers. "Whatever'their real shortcomings, r e t a i l models have been developed in the past decade into a powerful technique for interpret ing the behaviour of aggregations of people at a macro sca le . "They s t i l l appear to be the best means avai lab le to date for evaluat-13 ing a number of shopping centre proposals at one t ime." 73 SECTION TWO  AN EVALUATION OF THE GRAVITY MODELS 4.2.1 - TRIP DISTRIBUTION From the gravity model formulation (equation 4.) shown above, four separate parameters are required before the t r i p interchange, ( T i j ) , can be estimated. Two of the basic parameters; the number of t r i ps "produced" (Pi) by zone i , and the number of t r i ps "attracted" (Aj) by zone j ; are normally related to the use of the land as well as to some measure of the socio-economic charac ter i s t i cs of the trip-makers. Both Pi and Aj are obtainable through a base year or ig in-and-dest inat ion survey. The th i rd parameter., Fij,.measure the ef fect of spat ia l separation on travel behaviour between zones. I t i s var iously referred to as the f r i c t i o n or the impedance fac tor . Once the appropriate measure of the f r i c t i o n factor has been chosen, the next step in the process i s to c a l i -brate the model to the base year, or igin-and -dest inat ion data. Ca l ibrat ion involves a repeated adjustment of an i n i t i a l set of F i j ' s unt i l a sa t i s fac -tory approximation resu l t s . F i na l l y , the zone-to-zone adjustment fac tor , K i j , must also be obtained as input to the grav i ty model of t r i p d i s t r i bu t i on . I t i s said to r e f l e c t the ef fects of soc ia l and economic character i s t i cs of pa r t i cu la r zones or portions of the study area on travel patterns. Unlike the F-factor (or f r i c t i o n factor) the K-factor is applied only to interchange i - j . They should be used only when s ign i f i can t socio-economic var iat ions in land use are distinguishable or expected.^ 74 Several measures are normally taken to ensure that the gravity model w i l l s u f f i c i en t l y simulate trip-making behaviour in the study area in which i t i s applied before i t s use. F i r s t , in view of the fact that var iat ions between ex i s t ing and estimated conditions are inherent in any approximation process, s t a t i s t i c a l tests are often made to determine the accuracy of the gravity model in forecasting future travel patterns, once i t has been c a l i -brated. Among the several kinds of tests which may be used are Chi-square and the t - t e s t . I f the re la t i ve error for each t r i p interchange (obtainable from the tests) i s within the l im i t s of accuracy the planner i s prepared to accept, then the model i s deemed s t a t i s t i c a l l y sa t i s fac tory ; otherwise the source of. error must be located before the model i s appl ied. In short, the processes of ca l ib ra t ion and test ing imp l i c i t in the appl icat ion of the model "provide an ana ly t i ca l framework for developing a gravi ty model for any c i t y . " 1 5 However, a number of reasons may.be given to explain why the use of constant F i j 1 s and K i j 1 s for c i t i e s in the developing countries should be cautioned. These include population and land use changes and increasing urban household incomes. 1 . Population and Land Use Changes Average i n t r a - c i t y t r i p lengths are known to be strongly related to 16 the areal s ize of c i t i e s , which in turn depends on population growth and density changes. In the more developed nations, annual increases in urban land use are usual ly very smal l , amounting to about 2% or less per year. A l so , in these countr ies, annual addit ions to ex i s t ing urban i n f r a -structure which i s already very extensive, are normally neg l i g ib l e . In 75 the c i t i e s of the developing countries however, urban land uses change many times more rap id ly , and urban, inf rastructure often doubles in less than 10 y ea r s .^ This has in many cases led to s ign i f i can t changes in travel pat-terns; as r e s i den t i a l , commercial and indust r ia l a c t i v i t i e s become re-organ-ized to take advantage of new transportat ion f a c i l i t i e s . Secondly, c i t i e s in the developing countries are characterized by ex-treme concentration of a c t i v i t i e s on l im i ted geographic areas, while func-t i ona l l y related land uses are in many cases separated by long distances. More pa r t i cu l a r l y , the central business d i s t r i c t s of these c i t i e s (where most of the urban employment i s invar iab ly located) sprawl in a l l d i rect ions and serve as the points of convergence of the avai lab le transport system, thus at t ract ing further concentration of a c t i v i t i e s . Mogridge (1975) for example estimated that CBD employment in Nairobi (Kenya) w i l l increase from 50,000 in 1970 to about 150,000 by the year 2,000. These extremely rapid changes in urban population followed by massive investments in urban inf rastructure which in turn leads to sprawling urban physical structure w i l l resu l t in drast ic increases in t r i p lengths. Any estimates of t r i p interchanges for c i t i e s in developing countries based on constant f r i c t i o n factors are therefore most l i k e l y to ser ious ly underest i-mate the actual future t r i p patterns. 2. Increasing Household Incomes The present high densi t ies of land development often observable in c i t i e s of many developing countr ies, which have tended to rest ra in i n -creases in average t r i p length, cannot be expected to continue at the i r 76 present l eve l s . With increases in incomes, lower densi t ies w i l l be pre-1 g ferred implying corresponding increases in average t r i p length. In add i t ion, r i s i ng incomes w i l l also have upward influences on t r a -vel demand; espec ia l ly because of the present highly skewed income st ruc-ture of most developing countries which points to a s i tuat ion in which more than proportionate increases in motorized travel can be expected from the s l i gh tes t increases in spending power. Even though urban car ownership rates are s t i l l very low in the developing countr ies, (0.36 cars per household in Nairobi in 1970, according to Mogridge,. 1975, p. 40), World Bank (1972) estimates show that automobile reg i s t ra t ion in many of these countries i s increasing by between 10% to 20% per year; and that nearly a l l the vehicles are concentrated: in the c i t i e s . More recent World Bank (1975) project ions also show that: urban car populations in these coun-t r i e s w i l l jump from about 17 m i l l i on in 1970 to about 117 m i l l i on by the turn of the century. Thus decreasing densit ies and, increasing re l iance on motorized travel are much more l i k e l y to induce, s i gn i f i can t changes in the F-factors as well as in the zone-to-zone adjustment factors (K-factors) of gravity models in use in the developing countries than in the more developed ones. 4.2.2 - RESIDENTIAL LOCATION The gravity model of res iden t i a l locat ion i s concerned with the d i s -t r ibut ion of the tota l number of households to res ident ia l areas. I t y i e lds a d i s t r i bu t ion of home-workplace travel patterns based on the as-sumption that distance or some other measure of . the d i f f i c u l t y of travel i s a major:determinant in res ident ia l . cho ice decision-making. The pre-77 sumption therefore i s that other d i f f i c u l t i e s of searching for houses such as housing shortage are su f f i c i en t l y i n s i gn i f i can t and that house-holds always act on the basis of adequate information when making re loca-19 t ion decis ions. In the. developing countr ies, urban, housing shortages .constitute one of the major, problems a r i s i ng out of the commonly known discrepancies be-tween rates of population concentration and cap i ta l formation in urban i n -f rast ructure. A 1965 United Nations housing study estimated that the housing needs of Afr ican c i t i e s would r i se from 5.9 m i l l i on dwell ing units in 1965 to 9.9 m i l l i on by 1975. In c i t i e s of the developing countries therefore, the aggregate cost of housing as determinable by the supply and demand functions are very l i k e l y to have s ign i f i can t influences on res ident ia l choice. In other words, for quite a. large number of households in such c i t i e s , the commut-ing costs to be incurred at a l ternat ive locat ions may not rea l l y be " v i s i -ble" at a time*a housing decision i s made. Housing cost, pressures w i l l tend to push households away from.centers, of a c t i v i t y . Brand (1972), for example supports th is notion when he concluded from an analysis of t h i r -teen indicators of spat ia l organization in Accra that . "at present cost- -benefit l e ve l s , sub-urbanization appears to be a t t rac t i ve to two quite d i f ferent , groups: t h e e l i t e who seek open-space and the amenities of su-burban, l i v i n g ; and the poor, who seek low rents and are w i l l i n g to walk long distances to work". In short, the existence of a varied range of influences on the oper-ation of the housing market in the developing countries which sometimes 78 lead to publ ic interventions in the form of rent controls and subsidies, makes i t more d i f f i c u l t to provide a simple and broad explanation of the behaviour of households in the urban housing markets. In fac t , i t has been argued that the influence of such interventions in the housing mar-kets are so great that i t i s worthless to attempt to apply economic ana-20 l y s i s to the housing market. 4.2.3 - RETAIL LOCATION Retai l forecast ing i s normally based upon a number of assumptions, such as the s i ze , locat ion and the operating schedule of competing centers. I t i s for example assumed that, at the time a buyer i s faced with the choice between possible destinations of his shopping t r i p , a l l competing r e t a i l centers are opened so that distance (or.cost of t rave l ) and some measure o f the attract iveness of the centers remain the only determining factors . I f for some reason therefore, a number of the shops are c losed, obviously, the probab i l i ty of the shopper preferr ing a par t i cu la r shop to 21 a l l others would have become a l te red . A pecul iar feature.of the market systems in A f r i ca i s the i r period-i c i t y o f occurence. This means that?buyers and se l l e r s converge on a given locat ion da i l y , or. every second, th i rd or "nth" day. The reasons behind the per iod i c i t y o f marketing a c t i v i t i e s in A f r i ca are not made c lear in the l i t e r a t u r e . However, according to Eager!und and Smith (1970), the 'distance minimizing c r i t e r i o n 1 should be useful, in explaining market per-i o d i c i t y patterns in A f r i c a . In other words, the buyer or se l l e r cons i -ders the re l a t i ve locat ion of a market in terms of both time and distance so that locat ional and temporal competition i s complementary in the per iodic 79 system of marketing. When therefore two markets are located near each other, opening days are normally seperated by two or more days. In the urban economies of the developing countr ies, where the ro le of t rad i t iona l markets in most cases exceed that of western-type r e t a i l f a c i l i t i e s , the incidence of pe r iod i c i t y i s s im i l a r l y not iceable. Every sector of the c i t y w i l l have i t s own market which i t s residents can use in addit ion to the more extensive 'urban Market 1. In Accra for example, one can mention Nima and Kaneshie Markets as sector markets; and the Markola Market as the urban Market. Since, in the per iodic market system, the indiv idual , markets open on d i f fe rent days, the spat ia l competition, which the gravi ty model assumes to ex i s t between competing r e t a i l i n g f a c i l i t i e s becomes d i f f i c u l t to v i sua l i ze . This i s because the per iodic system allows the shopper to buy in one market at one t ime,and then v i s i t another market on another occa-s ion. A lso, because se l l e r s are not f ixed in space ( i . e . they are free to move the i r merchandise from one market to another at d i f fe rent market days) the computation o f thie re la t i ve attract iveness of the various markets w i l l not be an easy task espec ia l ly i f i t should be based on the volume of sa les , tota l sales number o f r e t a i l e r s and the assortment of goods. In short, the basic set of store choice a l ternat ives for consumers in c i t i e s of the developing countries i s not the same at a l l times as implied in the theory of the gravity model of r e t a i l locat ion. And with such va r i a -t ions in the basic set of store choice, comes var iat ions in pr ices of com-modit ies, espec ia l l y staple items; as well as var iat ions in the periods of operation of the/retai l f a c i l i t i e s . 80 Generally therefore, where per iod ic i ty ex is ts in marketing a c t i v i t i e s , as i s commonly found in c i t i e s of the developing countr ies, corresponding per iodic var iat ions can be expected in both the spat ia l competiton and the re la t i ve attract iveness factors of the gravity model of r e t a i l locat ion as measured by d i j and Aj respect ive ly . This them implies that any attempt to use th is model in the developing countries w i l l be faced with d i f f i c u l t i e s of having to estimate var iables which can account for per iodic changes in spat ia l competitiveness of r e t a i l f a c i l i t i e s . 4.2.4- - CONCLUSION To summarize from the preceeding d iscuss ion, three main points have so far been made regarding the relevance of the gravity model concept to c i t i e s in the developing countr ies. These include: i ) the rap id i ty of changes in land use patterns, densit ies as well as increasing re l iance on motorized travel which could possibly i n -duce s ign i f i can t changes in the f r i c t i o n factor , (F - fac tor ) , and the zone-to-zone socio-economic adjustment factor , (K- factor) , of the gravity model when used in transportat ion planning. To assume therefore that these factors w i l l , remain constant over the planning horizon as implied by the use of the gravity model of travel demand is inappropriate to the developing countr ies. i i ) the common occurences of serious housing shortages often resu l t ing in v i r tua l lack, of competition between housing developers which re-presents a considerable deviat ion from a basic, assumption of the gravi ty model of res ident ia l locat ion ( i . e . competitive choice of loca t ion) . 81 i i i ) p e c u l i a r i t i e s i n the pat te rns o f r e t a i l i n g a c t i v i t i e s found i n c i t i e s o f the deve lop ing c o u n t r i e s , p a r t i c u l a r l y the occurence o f market p e r i o d i c i t i e s , which are qu i t e l i k e l y to produce d a i l y v a r i a t i o n s i n s p a t i a l compet i t i veness , (measured by d i j ) , and r e l a -t i v e a t t r a c t i v e n e s s , (measured by A j ) , between r e t a i l i n g f a c i l i -t i e s ; wh i le the g r a v i t y model o f r e t a i l l o c a t i o n assumes these v a r i -ab les to remain constant over the p lann ing ho r i z on . Genera l l y t h e r e f o r e , un less c e r t a i n t y about the re levance o f the g r a v i t y model cou ld be e s t ab l i s h ed through emp i r i c a l , research, , i t i s c l e a r from the above d i s cu s s i on tha t i t s use in the deve lop ing coun t r i e s cannot be expected to produce a s u f f i c i e n t l y sound bases f o r p lann ing dec i s i on -mak ing . 8 2 a CHAPTER FIVE S U M M A R Y O F F I N D I N G S , C O N C L U S I O N S A N D R E C O M M E N D A T I O N S 8% SECTION ONE  SUMMARY OF FINDINGS 5.1.0 - INTRODUCTION This section summarizes the spec i f i c f indings of the study. 5.1.1 - THE COHORT-SURVIVIAL MODEL The spec i f i c f indings with respect to th is model are: 1. That there are two major obstacles to the estimation of the inputs ( i . e . su rv i va l , f e r t i l i t y , and migration rates) of the cohort-sur-v iva l model when i t i s selected for use in the developing countr ies: (a) insuf f i c iency of demographic data pa r t i cu l a r l y in the form of time ser ies ; and (b) the need to estab l i sh future values of the inputs them-selves pr io r to the i r use in the model; because of the highly unstable growth pattern experienced by c i t i e s in the developing countr ies. 2. That i t i s poss ib le , however, to estimate su rv i va l rates by subject ive ly estab l ish ing target rates; based on an analysis of indiv idual causes of death as well as re lated health po l i c i e s . 3. That the problems of f e r t i l i t y rate estimation in the developing coun-t r i e s can be overcome by using the c oho r t - f e r t i l i t y approach. This uses the f e r t i l i t y h is tory of a given set of mothers to derive a f e r t i l i t y path, (- a curve). The curves can then be shi f ted upward or downward ( sens i t i v i t y analys is) to r e f l e c t increasing or decreasing family s i zes ; respect ive ly . S im i l a r l y , they can be sh i f ted to the l e f t or to the r ight to describe ear ly or postponed b i r th expectations by ch i l d -bearing mothers, respect ive ly . 4. That in the developing countries where migration patterns are highly unstable and are a major contr ibut ing factor to urban population 83 growth, the need to estimate expected future migration by age and sex remains the major deterrent to any possible use of the cohort -surv i -val method. 5.1.2 - THE ECONOMIC BASE MODEL The spec i f i c f indings with respect to th is model are: 1. That, in the developing countries where the base rat ios of c i t i e s are general ly low (about 30% or l e s s ) , and informal ( large ly non-basic) employment increases by about 20% per year, the difference between com-puted and actual mu l t ip l i e r s i s l i k e l y to be s i gn i f i c an t . 2. That when used in urban population forecast ing, the base theory assumes that population w i l l remain a constant mult ip le o f tota l employment. In the developing countries however, the existence of the informal sector makes i t possible for people to move into the urban areas, not necessar i ly because they expect to. get a formal job there. In other words, i t i s possible for urban population in the developing countries to increase without a corresponding increase in formal job opportunit ies. 3. That in the case of urban economic base studies, the general problem of planning information shortage in the developing- countries i s par-t i c u l a r l y aggravated, by the existence of the informal sector. 5.1.3 - THE GRAVITY MODELS The spec i f i c f indings on th is set of models are: 1. That because of frequent i n s t a b i l i t i e s in land use and travel behav-iour in the developing countr ies, any estimates of tr ip- interchanges based on the use, of constant f r i c t i o n factors (F-factors) and constant 84 zone-to-zone socio-economic adjustment factors (K-factors) are very l i k e l y to be inaccurate and therefore misleading. That in the developing countr ies, imperfections in the housing market over-shadow any probable ef fects of commuting costs on res ident ia l locat ion to the extent that i t i s v i r t u a l l y impossible to provide a simple and broad explanation of the behaviour of households in the urban housing market along the l ines implied by the use of the gravity model of res ident ia l locat ion . That in c i t i e s of the developing countries., pa r t i cu l a r l y those of Af-r i c a , the occurence o f . pe r i od i c i t i e s in the marketing systems makes the effects of spat ia l competition and re la t i ve a c ce s s i b i l i t y d i f f i c u l t to v i sua l i ze and.account for in the gravity model of r e t a i l locat ion . That, general ly therefore, inaccurate and therefore misleading resul ts can be expected from any appl icat ion of the gravity concept of spat ia l interact ion to c i t i e s in the developing countr ies. With improvements in data base and increased emphasis on planning research, however, i t may be possible to modify the family of models in order to increase i t s relevance to local condit ions. 85 SECTION TWO 5.2.0-- CONCLUSIONS It may be concluded from the above f indings with respect to the three models discussed in th is study that the d i f f i c u l t i e s of t ransfer ing western planning tools to the developing countries are to a large extent a t t r i bu t -able to the fo l lowing three factors: a. Social, and Economic Differences Since the ex i s t ing planning methodologies have!been constructed within the soc ia l and economic contexts of the western countr ies; a considerable number o f the; assumptions upon which the methods are based maybe inappropriate to the developing countries where the socia l and economic mil ieus are yet to become homogeneous and stable. b. Differences in Rates of Urban Growth The rate of urban.growth in the developing countries i t s e l f creates the need to use planning models which are "dynamic" (or which can respond to changing condi t ions). Since the extent.of the dynamism thus required i s often not necessary to simulate the r e l a t i v e l y more stable planning environments of c i t i e s in the developed coun-t r i e s ; i t fol lows that models developed in these countries may hot be en t i re l y useful in simulating the rapid changes in population and land uses which are current ly noticeable in the developing countr ies. 86 c. Lack of Planning Data The th i rd conclusion to be drawn with respect to the problems associated with the appl icat ion of western planning methodologies in the developing countries re lates to the data required for the operation of the models. Most of the planning models current ly ava i lab le require the use of considerable amounts of planning data of a spec i f i c nature. In most developing countries however, data shortage and un r e l i a b i l i t y , pa r t i cu l a r l y in the form of time series data; i s a common problem; either.because the necessary f inanc ia l and human resources cannot be obtained, or: simply because the im-portance of co l l ec t ing such data i s not yet recognized. Under such circumstances, any attempt to use even the most elementary planning models can create a l o t of d i f f i c u l t i e s . 87 SECTION THREE  RECOMMENDATIONS 5.3.0 - INTRODUCTION This section puts foreward a number of both short- and long-term recommendations. The short-term recommendations are intended for meet-ing the immediate planning technique requirements of.planners in the developing countr ies; while the long-term measures, are expected to u l t i -mately resu l t in the development o f loca l l y - re levant planning techniques. 5.3.1 - APPROACHES TO SHORT-TERM PLANNING Where conventional planning techniques.are not d i r e c t l y appl icable to planning problems in the developing countr ies, a number of rudimentary approaches may be found to be pa r t i cu l a r l y useful in the 'short-term' when planning decisions must be made. Short-term as used here i s defined as the period over which, ex i s t ing inadequacies in planning data can be eliminated to open the way to intensive planning research aimed at evolving l o ca l l y relevant planning methodologies. The short-term approaches suggested in th i s section are spec i f i c a l l y related to each of the three models evaluated; namely: a. The Cohort-Survival Model b. . The Economic Base Model c. The Gravity Models 88 A. SHORT-TERM USES OF THE COHORT-SURVIVAL MODEL 22 1. Sources of Information a. Special Surveys: Population data can be obtained from anc i l l a ry sources. These include surveys for par t i cu la r purposes, (usual ly economic), and ad-min is t rat ive records. The systematization and thorough exp lo i tat ion of such information i s a most promising means of quickly improving avai lable s t a t i s t i c s for use in urban growth management in the develop-ing countr ies. By the i r nature, these surveys and records are often concerned with the:population groups which are c r i t i c a l in the provis ion of ur-ban in f ras t ruc ture: Also, the data are often co l lected from a smaller number of units than are f u l l - s c a l e population censuses, and these units (business, schools, hosp i ta ls , e tc . ) are in general better able to prepare accurate records without the need for lengthy invest igat ion. F ina l l y the nature of the units from which the records are gathered may make i t easier to concentrate on par t i cu la r planning purposes. Thus the use of s t a t i s t i c s avai lab le from special surveys and adminis-t ra t i ve records is a possible short-term measure to deal with the con-s t ra in ts which inadequate demographic information imposes on the use of the cohort-survival model of population forecast ing in the develop-ing countr ies. b. Sample Surveys: Where economies in data co l l ec t ion is the major concern as i s often the case in most developing countr ies, the sample census i s another 89 possible means of obtaining planning information at a relat ively low cost. As vehicles for the collection of demographic s ta t i s t i c s , sam-ple surveys have certain advantages and disadvantages. The sample survey provides a means of investigating specif ic ques-tions required by a particular study and which would normally not have been covered by the census. Since national population censuses are multi-purpose stat is t ica l projects by their nature, a fa i r l y large number of different topics must be investigated and no one of them can be explored in great depth. In a sample survey, i t is possible to probe a particular topic in depth at a relat ively moderate additional cost. 2. Using the Bayesian Approach One possible short-term approach to the estimation and forecasting of the necessary demographic components of the cohort method in the developing countries is the use. of informed guesses; or what is often 23 referred to as "Bayesian Information Processing". The procedure in -volves three main steps: a. A Study of Demographic Components: By studying the demographic components separately, a clear picture can be obtained of how different factors affect their measurement, and thus a greater insight may be gained into the potential it ies of existing trends to continue or alter in the immediate future. For example, measures of f e r t i l i t y and mortality can reveal the extent to which birth and death rates are dependent on structural factors 90 that w i l l change as the urban area becomes more s t ab i l i z ed . In th is way the extremely general question, 'how i s the growth rate l i k e l y to change?' becomes converted into a ser ies of l imi ted ones such as 'w i l l women continue to begin chi ld-bear ing ea r l y ? 1 , 'are the newly provided'health services l i k e l y to reduce infant morta l i ty qu ick ly? ' and so on. The more extensive the knowledge of. the demographic com-ponents the easier i t w i l l become to make reasonable judgments about the future trends. b. An Investigation of Future Change Indicators: A strong basis may be establ ished for judgment by supplementing knowledge of demographic components by information on what changes may be expected to take place. The best source of such evidence i s the study of change ind icators . For example a lower f e r t i l i t y in par-t i c u l a r age ranges for the more urbanized sections of the population may be a pointer to future reduction in b i r th rate. S im i la r l y change indicators for other demographic components may be noticeable in the att i tudes of parents, (e.g. toward family s i ze , education). c. Estimation of Subjective P robab i l i t i e s : Given knowledge about the ex i s t ing s i tuat ion and the basis for future changes; estimates of the chances and magnitudes of expected changes in the various demographic variables can be estimated using expert opinion. Once these estimates are obtained they can be fed into the cohort model to obtain forecasts of future population. 91 3. Estimating Migration Roussel (1970) suggests a technique which could be used to answer some of the problems of migration estimation pointed out in section 2.2.5. The fundamental idea of the method which he tested in the Ivory Coast i s based on a "double-run" survey. In the f i r s t stage, a l i s t i s compiled of the names of a l l the persons making up the population of a given geo-graphical zone or housing sample. Then, a simple questionnaire i s f i l l e d in for each resident giv ing such standard information as age, sex, level of education, occupation and so on. After, a time interva l which may. vary from s ix months to one year, a second run i s made to check cont inui ty in residence of. the study population as well as the entry of new members into the study area. In th is way i t i s possible not only to measure the, magnitude of ex i t and entry into the study area during the.period covered by the runs, but also to describe the migrant population qua l i t a t i v e l y . The two main streams of movement to.be covered in th is type of study are: i . Ex i t - from one residence to another in the study area, - from the study area to another, community in the same c i t y , - from the study area to another community not in the same c i t y ; i i . Entry -from one community in the same c i t y into the study area, and -from outside the c i t y into the study area. Generally therefore, there are a number of short-term measures which can be used to avert the d i f f i c u l t i e s of using the cohort survival model 92 to forecast urban population in the developing countr ies. These include f i r s t the use of any previously completed special studies as well as sam-ple surveys as supplementary sources of information. Secondly, the Bayesian approach can be used to estimate future f e r t i l i t y and survival rates; while the migration vector of the cohort model can be estimated using the "double-run" survey approach suggested by Roussel (1970). B. SHORT-TERM ALTERNATIVES TO THE ECONOMIC BASE MODEL Given the d i f f i c u l t i e s of using the base model in the developing coun-t r i e s , planners interested in studying the urban economies of these coun-t r i e s may consider using the fo l lowing approaches as short-term a l te rna-t i ves : 1.. Reviewing the Ex ist ing Urban Economy Studying the economy of an urban area w i l l normally commence with an appraisal of the ex i s t ing urban economy: the nature of economic a c t i v i t i e s , production l e ve l s , employment and so on. Where the data necessary for such appraisals i s not ava i lab le , they may be obtained as fo l lows: 25 a. Information From "Key Informants": Co l lect ion of information from key informants such as factory i n -spectors, health inspectors, school inspectors, c i t y l i cens ing o f f i -cers and so on, should be of great use in f i l l i n g much of the major gaps that may be found, to ex i s t in s t a t i s t i c a l surveys in developing countr ies. The key informant idea i s based on a recognit ion of the fact that a considerable number of persons in key posit ions and with key r e spons i b i l i t i e s , both publ ic and pr ivate , possess a wide knowledge 93 about economic and employment patterns related to the i r day-to-day operations. Hence i t may be found feas ib le to co l l e c t the informa-t ion these people can provide on a comprehensive, systematic and regu-l a r basis; and piece i t together to form a meaningful mosaic at the urban l e ve l . In fac t , Richter, (1978), claims that th is has been proved to be re l i ab l e in India and Pakistan. b. Informal Sector Data Needs: The informal sector provides employment pr imar i ly for the urban poor, whether as employers, unpaid-family members, own-account workers or wage^earners.. I t i s therefore,.necessary when planning c i t i e s in the developing countries to seek information about i t s potential for ra i s ing incomes and promoting employment. The International Labour Organization (ILO) has already made several attempts to measure the informal sector in the developing countr ies. Below i s a descr ipt ion of one of the approaches used by the ILO as documented by Sethuraman, (1976). i . Because the universe consist ing of the informal sector enterprises i s a large one, the ILO adopts a sample survey approach. This con-s i s t s of f i r s t ident i fy ing areas of concentration of informal sec-tor a c t i v i t i e s in the urban area. i i . Secondly, a sampling frame is constructed for each of the one-d i g i t codes in the International Standard C l a s s i f i ca t i on of a l l Economic A c t i v i t i e s . (Mining, u t i l i t i e s , banking and insurance are often excluded from the scope of the survey on the assumption that v i r t u a l l y a l l such enterprises are pub l i c ly owned or are 94 large indust r ia l or commerical concerns belonging to the formal sector. Likewise, the agr i cu l tura l sector may also be excluded not only because i t i s less important in the urban areas, but also because the problems of the informal sector enterprises in ag r i cu l -ture are s i gn i f i c an t l y d i f fe rent and therefore are best i n ve s t i -gated separately). This leaves the fol lowing f i ve categories: manufacturing, construct ion, t ransportat ion, trade and services, i i i . F i na l l y data are co l lected for each of the selected enterprises by administering a questionnaire developed for the purpose with the help of trained interviewers. Information sought in the questionnaire w i l l vary depending on spec i f i c planning object ives. Generally, however, th is w i l l include information on: - the physical background of. the enterprise - locat ion in re la t ion to other a c t i v i t i e s in . the c i t y , space usage; - the structura l background of the enterprise - deta i led descr ipt ion of the main and subsidiary a c t i v i t i e s , nature of markets, etc . - h istory of the enterprise - age., changes in business a c t i v i t i e s , goods and services produced, volume of out-put, employment and technology. - operational charac ter i s t i cs - capacity u t i l i z a t i o n , extent of underemployment. 27 2. Labour Forecasts* On the basis of population forecasts and establ ished a c t i v i t y rates, *A11 references made to economic a c t i v i t i e s , employment, etc . in section 5.3.IB imply both formal and those informal economic a c t i v i t i e s which can be covered in a study. 95 (or the p ropo r t i on o f the t o t a l popu la t i on tha t i s economica l l y a c t i v e ) , es t imates can be made o f the number of c i t y r e s i den t s l i k e l y to be seek-ing employment over the plan p e r i o d ; ( i . e . Tota l Popu la t ion x A c t i v i t y Rate = Labour Supp ly ) . S i m i l a r l y , es t imates of t o t a l demand f o r . l a b o u r by economic a c t i v i -t i e s can be obta ined by s imply summing f o recas t s o f product ion and employ-28 ment made by va r ious f i rms in the-:urban economy. The in fo rmat ion necessary to do t h i s can be c o l l e c t e d through sample surveys o f the k ind descr ibed above. Thus, i n a s i t u a t i o n where the urban economic base model cannot be used w i th conf idence in the deve lop ing c oun t r i e s , measures such as sample surveys of economic a c t i v i t y may be used to prov ide the bas i s f o r es t imates of economic a c t i v i t y i n the urban economy over the p lan p e r i o d . C. SHORT-TERM ALTERNATIVES TO.THE GRAVITY MODELS 29 The concept o f 'bas i c -needs s t r a t e g y ' prov ides an a l t e r n a t i v e to the g r a v i t y model as a way o f answering.most of the quest ions regard ing the l o c a t i o n o f urban hous ing , t r an spo r t a t i o n and r e t a i l f a c i l i t i e s i n the deve lop ing c o u n t r i e s , a t l e a s t in the shor t run . ' Bas i c -needs ' as used here r e f e r s to hous ing, road-network and r e t a i l f a c i l i t y requirements e s -s e n t i a l f o r minimum human comfort . The process o f bas ic-needs p lann ing begins w i th an i d e n t i f i c a t i o n o f lags, i n the p rov i s i on , o f the f a c i l i t i e s con-cerned, f o l l owed by a de c i s i on as to where new f a c i l i t i e s should be l o c a t e d . i ) I d e n f i t i c a t i o n o f Lags In o rder to map the nature and ex tent o f lags i n the p r o v i s i o n of u r -ban f a c i l i t i e s , t h e i r s p e c i f i c q u a n t i t a t i v e and d i s t r i b u t i o n a l data w i l l 96 be required. This kind of information can be obtained through on-the-spot inventories of the f a c i l i t i e s as well as household surveys of basic 30 needs. For example, an or i gin-and-destination survey wil 1 normally pro-vide information on basic urban travel needs; while an urban road-network inventory w i l l provide a measure of what the c i t y current ly offers in the form of transportation f a c i l i t i e s . The difference between the iden t i f i ed urban travel needs, and the assessed level of serv ice-of the avai lab le transportation f a c i l i t i e s therefore becomesa measure of the lag in urban 31 travel f a c i l i t i e s . Through a s im i l a r process of basic r e t a i l needs survey and r e t a i l f a c i l i t i e s inventory, i t i s possible to establ ish lags in the provision of r e t a i l f a c i l i t i e s . Besides sample surveys, large-scale a r i a ! photographs - espec ia l ly on a t ime-series basis - supplemented by relevant f i e l d checks could be useful in the inventory of land-uses and f a c i l i t i e s . i i ) Location of F a c i l i t i e s Considerations regarding the locat ion of f a c i l i t i e s to sa t i s fy basic needs w i l l normally depend on the nature of the ex i s t ing land-use patterns as well as the type of u rban . fac i l i t y being planned. Generally however, i t may be found necessary to f i r s t c l a s s i f y the f a c i l i t i e s by service area or l e ve l ' o f service-; (e.g. neighbourhood as against regional shopping cen-t res ; primary as against secondary roads; low income as against medium income housing). Once th i s is done, each type of f a c i l i t y can then be located on the basis of establ ished convenience., a c ces s i b i l i t y and cost standards. For example, places of residence w i l l normally be located as 97 close to employment centres as poss ib le; while at the same time taking account of the a v a i l a b i l i t y and cost of land and transportat ion. Generally therefore, in s i tuat ions where the gravity model cannot be used in c i t i e s of the developing countr ies, the strategy of basic ur-32 ban f a c i l i t y needs planning may provide a short-term a l te rnat ive . 5.3.2 - APPROACHES TO EVOLVING APPROPRIATE TECHNIQUES On the basis of the f indings and the conclusions out l ined in the f i r s t two sections of th is chapter, three main steps may be recommended to u l t i -mately evolve appropriate planning methodology in the developing countr ies. These include: a. The Promotion of Planning Research: "Within the l as t decade, s c i e n t i f i c research and development a c t i v i t i e s of a l l kinds have expanded rapid ly in the developed countr ies. . . . The special problems of developing countries have been studied to some extent, and experts from the de-veloped countries have proposed various measures to solve them. Despite the progress achieved in such studies, i t has become increas ingly evident that research re la t ing to these problems 33 must be carr ied out in the developing countr ies." This study recommends that the developing countries should become ser ious ly involved with planning research. I f such research e f fo r ts can be continuously directed at achieving a greater understanding of the i n d i -genous planning environment, they could lead to the development of l o ca l l y relevant planning models. In short, the promotion of local planning re-98 search, with special emphasis on the social and economic factors operating in the planning environment should be considered as a necessary prerequisite to the development of l o c a l l y relevant methodology to aid planning decision-making in the developing countries. b. The Establishment of Planning Data Banks: "Conventionally, the i d e n t i f i c a t i o n , description and quantification of problems in the more developed worlds r e l i e s heavily on continuous analysis of existing trends against a r e l a t i v e l y fixed backdrop of incremental devel-opment and stable social and p o l i t i c a l attitudes. But in the developing countries, change i s , by d e f i n i t i o n , occuring at proportionately greater rates under the aegis of evolving p o l i t i c a l , administrative and social systems; . . . It can therefore be contended that the need to define the fi n a n c i a l resources and p o l i t i c a l means available f o r im-plementing solutions i s as important as the data required to define the scale and types of problems requiring 34 solution." This study recommends that any ef f o r t s to.undertake planning research in the developing countries w i l l have to be closely matched with correspond-ing e f f o r t s to improve the currently unreliable state of planning data in the developing countries. This implies that, while at the i n i t i a l stages, planning researchers w i l l have to operate within the scope of whatever 'skeletal' data i s currently available, ultimately, expansions in the f i e l d of planning, research w i l l depend on the a v a i l a b i l i t y of better organ-99 ized forms of planning data. This i s because once the hypotheses re la t ing to the local planning:environments are conceptualized by researchers, they must be tested against both quant i tat ive and qua l i ta t i ve .descr ip t ions of the real-world s i t ua t i on . The only conceivable way to make such descr ip-t ions in the form of planning data read i ly avai lab le to potential researchers i s to estab l i sh a central agency that i s charged with the respons ib i l i t y of co l l e c t i ng , processing and stor ing basic planning data. To keep the costs of co l l ec t ing and stor ing such data to a minimum, the fol lowing three guide-l ines may be found pa r t i cu l a r l y usefu l . i ) The data system should be re lated to the regular operating processes of the par t i c ipa t ing agencies. This c r i t e r i on i s very important in ensuring that only the minimal costs in data co l l ec t ion are incurred. For example, the type of information, to be derived from appl icat ions for planning permission should form an important source of the data to be co l l ec ted . In th is par t i cu la r case, the co l l ec t ion of informa-t ion i s under the d i rec t control of the local planning author i ty , and becomes a by-product of i t s day-to-day operat ion. The advantages i n -herent in th i s approach include the po s s i b i l i t y of regular updating and co-ordination of the data system with the d i rect ion of planning. i i ) Since the nature of planning i s such that the uses to which any planning information may be put are varied (e.g. transportat ion planning, t r ans i t planning, housing), i t i s important that the planning data sys-tem is capable of producing output in a var iety of forms. This could be achieved i f the system is capable of aggregating information from the lowest level at which data must be stored to the highest level 100 at which data i s l i k e l y to be required. In other words, the system must: a. permit each user or part ic ipant to have easy access to the in fo r -mation in the data system; b. provide for the manipulation and preparation of data in meaningful forms at the inst ruct ion of the user, and c. make i t possible for each user to independently determine the pur-pose of his data without s a c r i f i c i ng the level of deta i l necessary for his purpose. I t i s only when the system is adequately user-oriented that i t w i l l be able to e l i c i t the support and co-opera-t ion of a multi-agency c l i e n t e l e . i i i ) Sources of planning information often cover a wide range. They include reports and s t a t i s t i c s published by publ ic or pr ivate agencies; and o f f i c i a l census reports which may be used to supplement any informa-t ion that may be obtained d i r e c t l y by the planning authority in a var iety of ways. Any agency responsible for planning information management should therefore be able to extend i t s ' tentac les ' to a l l possible sources, while ensuring area and period compat ib i l i ty . I t should serve as a "clearing-house" for new surveys to avoid dup l ica-t ions . The data management agency should further be able to co-ordin-ate supplementary surveys; and to formalize the a r i a ! units of data co l l ec t i on e i ther by an establ ished grid-system or by using Census Enumeration Areas. c. Planning Administration Re-organization: To a large extent, the i n ab i l i t y of many planning author i t ies in the developing countr ies, pa r t i cu l a r l y those in A f r i c a , to solve the problems 101 of the i r urban centers can be traced to a general lack of an administra-t i ve framework capable of conceiving, implementing and monitoring urban development planning and programs. The management of c i t i e s in A f r i ca was t r ad i t i ona l l y the concern of local author i t ies whose prime objectives were the maintenance of law and order, inc luding emphasis on the achievement of adequate sanitary and safety condit ions. The i n ab i l i t y of these bodies to manage the urban centers effect ive ly.was not pa r t i cu l a r l y recognized under r e l a t i v e l y low levels of urbanization.. With the increasing pace of urbani-zation within the l as t two decades, however, i t became obvious that many of the local planning author i t ies do not have the administrat ive resources necessary to combat the der ivat ive problems. This therefore led to the set t ing up of urban planning author i t i es , presumably,, to undertake more e f f i c i en t management of urban growth. Unfortunately however, the Afr ican Urban Planning Author i t ies have, in most cases, operated as independent bodies, instead' of working in close col laborat ion with other loca l agencies in the e f f i c i e n t management of c i t y growth. They have also concentrated so le ly on development control with very l i t t l e attent ion to planning research. In recommending a course of action for promoting the e f f i c i e n t management of c i t i e s , and the develop-ment of planning methodology.in the developing countr ies, therefore, the need for a simultaneous reform in the organizational structures of the ex i s t ing planning author i t ies cannot be overlooked; espec ia l ly when i t i s needed to create an administrat ive set up for d i f ferent area l eve l s , as well as to provide a more adequate framework for planning research, po l i cy formulat ion, plan preparation, implementation and monitoring. 102 In a d d i t i o n t o w i d e n i n g t he scope o f o p e r a t i o n o f t he e x i s t i n g urban management a u t h o r i t i e s , s e r i o u s c o n s i d e r a t i o n s h o u l d a l s o be g i v e n t o t he c o - o r d i n a t i o n o f t he e x i s t i n g f r a g m e n t a r y p l a n n i n g a g e n c i e s . In o t h e r wo r d s , what i s needed i s a s e l f - s u f f i c i e n t m u l t i - p u r p o s e u n i t o f l o c a l p l a n -n i n g a d m i n i s t r a t i o n o f c omp rehens i v e u rban deve l opmen t p l a n n i n g and manage-ment ; p r e f e r a b l y d e f i n e d on t he b a s i s o f s o c i a l , economic and o t h e r f u n c -t i o n a l " r e l a t i o n s h i p s ; w i t h emphas i s on t he e s t a b l i s h m e n t o f d a t a management and r e s e a r c h u n i t s t o p r o v i d e t he r e l e v a n t f a c t u a l b a s i s f o r p o l i c y f o r m u -l a t i o n . A s s e s s i n g t he E f f e c t i v e n e s s o f : P I a n n i n g A d m i n i s t r a t i o n The f o l l o w i n g c r i t e r i a a r e . s u g g e s t e d f o r a s s e s s i n g t h e e f f e c t i v e n e s s o f e x i s t i n g u rban p l a n n i n g i n s t i t u t i o n s i n the d e v e l o p i n g c o u n t r i e s ; as a f i r s t s t e p i n any e f f o r t t o r e - o r g a n i z e them: a . The i n s t i t u t i o n a l s e t up o f any p l a n n i n g agency, s h o u l d be f l e x i b l e , and l e n d i t s e l f t o e x p a n s i o n o r r e d e s i g n i n t h e l i g h t o f i n c r e a s e d e x p e r i e n c e and c h ang i n g r e q u i r e m e n t s . T h i s c r i t e r i a i s p a r t i c u -l a r l y i m p o r t a n t because o f t h e i n e v i t a b l y e x p e r i m e n t a l n a t u r e o f p l a n n i n g i n s t i t u t i o n s i n t h e d e v e l o p i n g c o u n t r i e s a t t he i n i t i a l s t a g e s . b. The i n s t i t u t i o n a l d e s i g n must be w i t h i n t h e l i m i t s s e t by t h e a v a i l -a b i l i t y o f p e r s o n n e l , equ ipment and f i n a n c i a l r e s o u r c e s . c . The p l a n n i n g i n s t i t u t i o n must have t h e f u l l s u p p o r t o f t he c e n t r a l govenment and o t h e r r e l a t e d a g e n c i e s such as r e g i o n a l deve l opmen t c o r p o r a t i o n s . 103 d. F i na l l y , the functioning of the re-organized planning i n s t i t u -tions must take due account of the existence of national and regional planning a c t i v i t i e s . This w i l l avoid any dupl icat ion of development e f f o r t s , and ensure consistency in the promotion of order ly growth in general. 5.4.0 - CONCLUSION Generally therefore, th i s study recommends that, in dealing with the problems of planning methodology in the developing countr ies, planners must draw a d i s t i nc t i on between short-term and long-term st ra teg ies . More s pe c i f i c a l l y , i t i s suggested that, where the planner in a developing coun-try must immediately make planning decisions in the face of inadequate data and planning methodology, he could resort to the use of a number of tenta-t ive or short-term techniques. Thus, for the cohort, the economic base, and the gravity models which const i tute the focus of th i s study, a number of short-term approaches are out l ined. They are intended to overcome or a l l uv ia te the d i f f i c u l t i e s a r i s ing out of lack of planning data and/or l o ca l l y relevant planning tech-niques. F i na l l y , the study also recommends that the need to i den t i f y , adapt or evolve planning methodologies which can e f fec t i ve l y r e f l e c t the work-ings of urban systems,(an integrat ion of the phys ica l , socia l and economic environments), through research supported by corresponding improvements in planning data and administrat ion must become the primary focus of planning e f fo r ts in the developing countr ies. I n i t i a l l y , the resu l ts of such re-104 search e f f o r t s may be c rude l y framed. However, once the bas i c r e l a t i o n -sh ips are e s t a b l i s h e d , g rea te r e f f i c i e n c y in p r ed i c t i o n can be accompl ished w i th the help o f f u r t h e r data c o l l e c t i o n and re sea r ch . In o ther words, the process o f l o c a l l y r e l e van t p lann ing theory development in the deve lop-ing coun t r i e s should be t r ea t ed as a cont inuous process i n v o l v i n g compre-hensive feedback mechanisms to help avo id mistakes a l ready made. 105 BIBLIOGRAPHIC NOTES 1. The information presented in th i s section is an extract from two main sources; " i . Shryock, Henry S. et a l . (.1975); Methods and Materials of demography, U.S. Government Pr int ing Of f i ce , Washington, D .C , pp. 712-719. i i . Hightower, H.C. (1968); "Population Studies"; in W.I. Goodman and E.C. Freund. (eds.); Pr inc ip les and Pract ice of Urban Planning; International Ci ty Managers Associat ion; Washington, D .C: pp. 51-75. 2. See for example, Masser, Ian (1972), Analyt ica l Models for Urban and Regional Planning, David and Charles; Newton Abbot; pp. 29-42; and Rogers, A. (1966), "Matrix Methods of Population Ana lys i s" , JAIP, V. 32, n. 1, pp. 40-44. 3. Hel ly , Walter (1975). Urban Systems Models. Academic Press, New York, p. 64. 4. Chapin, F. Stuart, J r . (1976). Urban Land Use Planning (2nd ed i t i on ) , Univers i ty of I l l i n o i s Press, Chicago, pp. 135-140. 5. Gras, N.S.B., as quoted by F.S. Chapin, J r . (1976). Urban Land Use Planning. (2nd ed i t i on ) , Univers i ty of I l l i n o i s Press, Chicago, p. 140. 6. Andrews, B. Richard, (1968). "Economic Studies" in W.I. Goodman and E.C. Freund; Pr inc ip les and Pract ice of Urban Planning. International City Managers' Associat ion, Washington, D.C. 7. S iege!, Richard A. (1967). "The Economic Base and Mu l t i p l i e r Analy-s i s " . Urban A f fa i r s Quarterly, V. 2, No. 2, pp. 24-38. 8. Tiebout, CM . (1962). The Community Economic Base Study. Committee for Economic Development Supplementary Paper No. 16, New York. 9. Masser, Ian (1972). Analyt ica l Models for Urban.and Regional Planning. David & Charles: Newton Abbot. 10. Swan, T.R. (1976). An Introduction to Mathematical Planning Models. New South Wales Planning Commission, Technical Bu l l e t in No. 7, p. 18. 11. U.S. Department of Transportation. (1972). Urban Transportation Planning. Federal Highway Administrat ion, Washington, D .C , pp. 15-40. 106 12. Krueckeberg, D.A. and A.L. S i l v e r s , (1974) Urban Planning Analys is: Methods and Models. John Wiley & Sons, New York, p. 334. 13. McCabe, W.R. (1974), Planning Appl icat ions of Retai l Models. Min ist ry of Treasury Economics and Intergovernmental A f f a i r s , Ontario. 14. Cesario, F .J . (1974). "The Interpretat ion and Calculat ion of the Gravity Model Zone-to-Zone Adjustment Factors". Environment and  Planning A, V. 6, pp. 247-257. 15. Hansen, W.G. (1962) "Evaluation of the Gravity Model Trip D is t r ibut ion Procedures". Highway Research Board Bu l l e t i n , v. 347, pp. 67-76. 16. Chapin, F.S. J r . (1968). "Ac t i v i t y Systems and Urban Structure: A Working Scheme", JAIP, v. 34, n. 1, p. 11. 17. World Bank, (1975), The Urban Transport Sector. IBRD, Washington, D.C. 18. Amato, P.W. (1970) "A Comparison: Population Densit ies. Land Values and Socio-Economic Class in Four Latin American C i t i e s . " Land  Economics, V. 46, pp. 446-455. 19. Winger, A.R. (1970). "The V i s i b i l i t y of Commuting Costs and Residen-t i a l Locat ion," Environment and Planning, v. 2, pp. 89-74. 20. Kirwan, R.M. and D.B. Martin (1971), "Some Notes on Housing Market Models for Urban Planning.". . Environment and Planning, v. 3, pp. 243-252. 21. Agergard, E. et a l . (1969) "Interact ion Between Reta i l ing and the Urban Center Structure: A Theory of Spira l Movement". Environment  and Planning, v. 2, pp. 55-71. 22. Rees, G. and T.L. Rees, (1977) "Alternat ives to the Census". Town Planning Review,: V. 48, n. 1, pp. 123-140. Also see, Ian Masser, (1974), "Planning with Incomplete Data: Population Growth and Metropolitan Planning in the Third World," Town Planning Review, V. 45, no. 2, pp. 159-169. 23. - "Andriole, S.d.- (1979), "Comparative :Forecasting.",. Futures':,:-, The Journal of Forecasting and Planning, V. 11, August 1979, pp. 275-286. 24. Roussel, Louis, (1970), "Measuring Rural-Urban Dr i f t in Developing Countries: A Suggested Method: International 'Labour Review, V. 101, No. 3, pp. 229-246. 25. Richter, L. (1978), "New Sources of Manpower Information in the Developing Countries", International Labour Review, V. 117, n. 4, pp. 453-462. 107 26. Sethuraman, S.V. (1976), "The Urban Informal Sector: Concept, Measurement, and Po l i cy" , International Labour Review, V. 114, n. 1, pp. 69-81. 27. Wery, Rene (1978), "Manpower Forecasting and the Labour Market", International Labour Review, V. 117, no. 3, pp. 332-342. 28. Tanzania (1979) Par es Salaam Master Plan, (Technical Supplement 2, Population and Economic P r o f i l e ) , Min ist ry of Lands, Housing and Urban Development, Dar es Sallam, Tanzania. 29. Mayer, Jean (1979), "Spatial Aspects of Basic-Needs Strategy: The Dis t r ibut ion of Essential Services, International.Labour Review, V. 118, no. 1, pp. 59-74. 30. Radwan, S. and T. A l f than, (1978), "Household Surveys for Basic-Needs: Some Issues", International Labour Review, V. 117, n. 2, pp. 197-210. 31. Mahayni, R.G. (1977), "Reorienting Transportation, Planning Rationale in Developing Countr ies," T ra f f i c Quarterly, V. 31, pp. 351-365. 32. Cole, S. and H. Lucas (eds.) (1979), Models, Planning and Basic Needs, Pergamon Press, New York, pp. 148-153. 33. United Nations Industr ia l Development Organization (UNIDO), (1970) "The Need for Research Within Developing Countries"; Development Digest, V. 8, no. 3, p. 64. 34. Harrison, W.J. (1978) "The Information Problem and Planning: Making Good Use of Imperfect Data", in PTRC, Urban and Regional Planning  in Developing Countries, Proceedings of Seminar E, Ju ly , pp. 41-42. 108 •BIBLIOGRAPHY 1. Andrews, R.B. (1956) "Mechanics of the Urban Economic Base: The base Concept and the Planning Process", l and Economics. V. 32, pp. 68-84. 2. Andrews, R.B. (1968) "Economic Studies" in W.I. Goodman and E.C. Freund (eds.) Pr inc ip les and Pract ice of Urban Planning; International City Managers Assoc iat ion, Washington, D.C. pp. 76-79. 3. Black, W.R. (1973) "An Analysis of the Gravity Model Distance Exponents", Transportation, V. 2, pp. 299-312. 4. Blandy, R. and R. Wery (1973). . "Population and Employment Growth: Bachue-1". JJLR, V. 107, pp. 441-449. 5. Brand, R.R. (1972) "The Spatial Organization of Spatial Areas in Accra, Ghana., With Par t i cu la r Reference to Aspects of Modern-i z a t i on " , Economic Geography, V. 48, pp. 248-298. 6. Brodsky, H. and D.E. Safarty (1977) "Measuring the Urban Economic Base in a Developing Country". Land Economics. V. 53, No. 4, pp. 445-455. 7. Browning, H.L. "Migrant Se l e c t i v i t y and the Growth of Large C i t ies in Developing Soc ie t ies" . In Rapid Population Growth, sponsored by the National Academy of Sciences. The John Hopkins Press, London, 1971. pp. 273-314. 8. Cantre l le , P. (1971) "Is There a Standard Pattern of Tropical Morta l i ty?" Population in Afr ican Development., V. 1, Ordina Edi t ions, Dolihain. 9. Clarke,. J . I . (1971). Population Geography and the Developing Countries. Pergamon Press, New York. 10. Conde, J . & J . Boute (1971), "Introduction" Population in Afr ican Development. V. 1, Ordina Edi t ions, Dolhaih. 11. Crooks, R.I. (1971) "Planning for Developing Countries" Journal of the Town Planning Ins t i tu te , V. 57, no. 6, pp. 251-256. 12. Fagerlund, V.G. and R.H.T. Smith, (1970). "A Prel iminary Map of Market Pe r i od i c i t i e s in Ghana". Journal of Developing Areas. V. 4, pp. 333-348. 109 13. F ie lds , (1975) "Rural-Urban Migrat ion, Urban Unemployment, and Job Search Ac t i v i t y in the Less Developed Countries" Journal of  Development Economics, June, 1975, pp. 165-187. 14. Gabler, R. (1971). "Population Size as a Determinant of Ci ty Expen-ditures and Employment - Some Further Evidence'.1 Land Economics. V. 47, pp. 131-138. 15. Gardiner, P. (1972) An Estimation,of a Quasi-Stable Age-Sex D i s t r i -bution for Ghana, i i i 1960. Working Paper No. 33, U.S. Department of Commerce, Bureau of the Census. 16. Ginneken, W. (1976); Rural-Urban Income Inequal i t ies: International Labour Of f i ce , Geneva. 17. Ginsberg, R.B. (1971.).. Two Papers, on the Use and Interpretat ion of P robab i l i s t i c Models with Appl icat ion to the Analysis of Migrat ion. Centre for Environmental Studies; Working Paper No. 73. 18. Gugler, J . (1976). "Migrating to. Urban Centres of Unemployment in Tropical A f r i c a . " in A.H. Richmond (ed.); Internal Migrat ion; Sage; Ca l i f o rn i a . 19. Hart, Keith (1973), "Informal Income Opportunities and Urban Employment in Ghana." Journal of Modern Afr ican Studies.. V. 11, pp. 61-89. 20. Hightower, H.C. (1968).; "Population Studies"; in W.I. Goodman and E.C. Freund (eds.); Pr inc ip les and Pract ice of Urban Planning; International City Managers Assoc iat ion; Washington D .C ; pp. 51-75. 21. House, W.J. & H. Rempel; (1976) "Labour Market Pressures and Wage Determination in Less Developed Countries: The Case of Kenya." In Economic Development and Cultural Change, (Unpublished). 22. Ley, D.F. & G. Anderson (1975), "The Delphi Technique in Urban Fore-cast ing"; Regional Studies; Vol . 9, pp. 243-249. 23. Masser, Ian. (1972), Analyt ica l . Models for Urban and Regional Planning, David and Charles; Newton Abbot. 24. Mogridge, M.C. (1975). "Transportation Planning in Nairobi" T ra f f i c Engineering and Control , V. 16, n. 1, January. 25. Needham,. D.B. (1974.) "Three Ways of Studying the Urban Economy." Urban Studies, V. 11, n. 2, pp. 211-215. n o 26. Rees, P.H. and A.G. Wilson (1975) "A Comparison of Avai lable Methods of Population Change", Regional Studies, V. 9, pp. 39-61. 27. Rogers, Andrei (1966); "Matrix Methods of Population Ana lys is" . JMP , V. 32, N. 1, pp. 40-44. 28. Rogers, Andrei (1976) "Shrinking Large-Scale Population Project ion Models by Aggregation and Decomposition." Environment and Planning A. V. 8, N. 5, pp. 515-541. 29. Schultz, T.P. (1976). Notes on Migration Decision Functions. Papers Presented at the World Bank Research Workshop on Rural-Urban Labour Market Interact ions, Washington, D.C. 30. Sethuraman, S.V. (1976), "The Urban Informal Sector: Concept, Measure-ment and Po l i cy . " ILR, V. 113, pp. 69-81. 31. Sethuraman, S.V. (1977). ; "The Urban .Informal Sector in A f r i c a . " ILR, V. 116, pp. 343-352. 32. S iege l , R.A. (1967). "The Economic. Base and Mu l t i p l i e r Ana lys i s . " Urban A f fa i r s Quarterly. V, 2, N. 2, pp. 24-38. 33. Singer, H.W. (1971), "Employment Problems, in the Developing Countries". Manpower and Unemployment Research in A f r i c a . V. 4, N. 1, pp. 29-35. 34. S i r k i n , G. (1959) "The Theory of the Regional Economic Base", Review of Economics and S t a t i s t i c s , Nov. 1959, pp. 426-429. 35. Souza, P.R. & V.E. Tokman (1976). "The Informal Urban Sector in Latin America," ILR.. V. 113, pp. 355-365. 36. Tiebout, C M . (1962) The Community Economic Base Study. Committee for Economic Development Supplementary Paper No. 16, New York. 37. Todaro, M.P. (1976); Internal Migration in Developing Countries; I . L . C , Geneva. 38. Weeks, John (1975) "Po l i c ies for Expanding Employment in the Informal Urban Sector of Developing Countries". ILR (International Labour  Review). V. I l l , N. 1, pp. 1-13. 39. W i l l i s , K.G. (1972), "Population Studies in Planning" Planning Out-look, V. 12, pp. 51-57. I l l 40. Wilson, A.G. (1971) "A Family of Spatial Interaction Models and Associated Developments", Environment and Planning, V. 13, N. 1, pp. 1-32. 41. World Bank, (1972). Urbanization: Sector Working Paper, The John Hopkins Press. Baltimore. 112 APPENDIX A DEFINITIONS OF POPULATION TERMS USED IN THE STUDY Age-Sex Structure: The number of proportion of females and males in each age category of a population group. Age-Specif ic F e r t i l i t y Rate: The year ly number of l i v e b ir ths per thousand mothers in each of seven age groups (15-19, 20-24, . . . . 45-49). B i r th Rate: Total annual l i v e b i r ths per thousand population. Survival Rate: Yearly number of survival per thousand population. F e r t i l i t y Rate: Number of l i v e b i r ths experienced by thousand women of chi ld-bear ing age annually. Infant Morta l i ty Rate: Yearly number of deaths among chi ldren below the age of one year per thousand l i v e b ir ths in the same year. Rate of Natural Increase: Yearly percentage increase in population s ize due to excess of b i r ths over deaths. Rate of Population Growth: Annual percentage change in population s ize due to natural increase and migrat ion. 113 Sex Ratio: number of males in population number of females in population 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0095403/manifest

Comment

Related Items