Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Intra-urban industrial linkages and rail terminal location Torchinsky, Raymon L. 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_1982_A8 T67.pdf [ 6.17MB ]
Metadata
JSON: 831-1.0095538.json
JSON-LD: 831-1.0095538-ld.json
RDF/XML (Pretty): 831-1.0095538-rdf.xml
RDF/JSON: 831-1.0095538-rdf.json
Turtle: 831-1.0095538-turtle.txt
N-Triples: 831-1.0095538-rdf-ntriples.txt
Original Record: 831-1.0095538-source.json
Full Text
831-1.0095538-fulltext.txt
Citation
831-1.0095538.ris

Full Text

INTRA-URBAN INDUSTRIAL LINKAGES and RAIL TERMINAL LOCATION by RAYMON L. TORCHINSKY B.'A. M c G i l l U n i v e r s i t y , 1973 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS i n t he department o f GEOGRAPHY We ac c e p t t h i s t h e s i s as con f o r m i n g t o the r e q u i r e d s t a n d a r d THE UNIVERSITY OF BRITISH COLUMBIA November, 1981 © RAYMON L. TORCHINSKY, 1981 In presenting t h i s thesis i n p a r t i a l f u l f i l m e n t of the requirements for an advanced degree at the University of B r i t i s h Columbia, I agree that the Library 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 reference and study. I further agree that permission for extensive copying of t h i s thesis for s c h o l a r l y purposes may be granted by the head of my department or by h i s or her representatives. It i s understood that copying or p u b l i c a t i o n of t h i s thesis for f i n a n c i a l gain s h a l l not be allowed without my written permission. Department of MBoyftPH^ The University of B r i t i s h Columbia 2075 Wesbrook Place Vancouver, Canada V6T 1W5 ABSTRACT I t has been suggested t h a t d e c l i n i n g urban f r e i g h t t r a n s p o r t c o s t s have l e s s e n e d the i m p o r t a n c e of i n d u s t r i a l l i n k a g e as a f a c t o r i n i n t r a - u r b a n i n d u s t r i a l l o c a t i o n , r e s u l t i n g i n the suburb-a n i z a t i o n of i n d u s t r i a l a c t i v i t y . I n n e r c i t y i n d u s t r i a l a r e a s have been c h a r a c t e r i z e d as 'zones i n t r a n s i t i o n ' , i m p l y i n g t h a t such are a s do not c o n s t i t u t e e f f i c i e n t use of l a n d i n the c o n t e x t of modern urban s p a t i a l o r g a n i z a t i o n . . T h i s t h e s i s examines the i n d u s t r i a l l i n k a g e s a s s o c i a t e d w i t h Vancouver's i n n e r c i t y r a i l t e r m i n a l , l o c a t e d on the n o r t h shore of F a l s e Creek, i n order, to tes.t the h y p o t h e s i s t h a t the f i r m s r e m a i n i n g i n the v i c i n i t y of t h e . t e r m i n a l form a v i a b l e i n d u s t r i a l complex . focussed. on the t e r m i n a l o p e r a t i o n . Data g a t h e r e d i n 1975 f o r a comprehensive survey of Vancouver t r u c k -i n g o p e r a t i o n s i s adapted and a n a l y z e d to. determine t e r m i n a l -r e l a t e d l i n k a g e p a t t e r n s . .A. comparison w i t h the l i n k a g e s a s s o c i a t e d w i t h Vancouver's other, r a i l / t r u c k g e n e r a l f r e i g h t t e r m i n a l shows the s i m i l a r i t y of the two ..terminals' impact on the d i s t r i b u t i o n of l o c a l i n d u s t r i e s . In both .cases . t r a n s p o r t c o s t s are an impor-: t a n t f a c t o r i n e x p l a i n i n g l i n k a g e p a t t e r n s . The p e r s i s t e n c e of i n d u s t r i a l a c t i v i t y i n the i n n e r c i t y can thus be a t t r i b u t e d to the c o n t i n u i n g i n f l u e n c e . o f t r a n s p o r t - r e l a t e d f a c t o r s , s u p p o r t i n g the h y p o t h e s i s t h a t . a n i n n e r , c i t y l o c a t i o n .is not n e c e s s a r i l y t r a n s i t i o n a l or i n e f f i c i e n t , f o r i n d u s t r i a l f i r m s . However, non-t e r m i n a l l i n k a g e s of i n n e r c i t y f i r m s do.not d i m i n i s h w i t h d i s t a n c e ; the v i a b i l i t y of t h i s complex w i l l be a d v e r s e l y a f f e c t e d by the proposed r e l o c a t i o n o f . t h e F a l s e - C r e e k t e r m i n a l to a suburban s i t e . i i TABLE OF CONTENTS Page LIST OF FIGURES v CHAPTER ONE I n d u s t r i a l l i n k a g e s and r a i l t e r m i n a l s 1 A. INTRODUCTION * 1 B. i n d u s t r i a l L i n k a g e s 3 1. L i n k a g e s and l o c a t i o n t h e o r y 3 2. L i n k a g e s a t the l e v e l o f the f i r m 8 3. Approaches t o L i n k a g e R e s e a r c h 12 C. R a i l - t e r m i n a l l i n k a g e s 15 1. T e r m i n a l l i n k a g e s 15 2. R a i l t e r m i n a l s and the Vancouver i n d u s t r i a l s t r u c t u r e 18 3. The F a l s e - C r e e k t e r m i n a l 19 D. F o o t n o t e s t o Chapter One 23 TWO A n a l y s i s o f r a i l t e r m i n a l l i n k a g e s 25> A. I n t r o d u c t i o n 25 B. D e s c r i p t i o n o f da t a 25 1. The Vancouver urban t r u c k i n g s t u d y 25 2. Use o f VUT d a t a t o e s t i m a t e demand 28 i ) G e n e r a l i s s u e s 30 i i ) I s s u e s r e l a t e d t o VUT d a t a 41 a) N o n - S p a t i a l a s p e c t s o f the sample 41 b) A r e a l a g g r e g a t i o n 45 i i i -CHAPTER Page TWO c) C l a s s i f i c a t i o n o f commodity type 60 d) Land-use d a t a 62 e) D i s t a n c e measurement 63 C. R e g r e s s i o n a n a l y s i s 63 D. Component a n a l y s i s 71 1. D i s a g g r e g a t i o n o f f l o w d i s t r i b u t i o n s 71 2. Commodity a n a l y s i s 78 3. A n a l y s i s o f commodity s i z e d i s t r i b u t i o n s 82 4. S p a t i a l a n a l y s i s 97 i ) I n t r o d u c t i o n 97 i i ) A n a l y s i s o f a t t r a c t i o n 101 i i i ) A n a l y s i s o f d i s t a n c e decay 115 E. F o o t n o t e s t o Ch a p t e r Two 137 THREE C o n c l u s i o n 139 A. R a i l T e r m i n a l l i n k a g e s 139 B. F a l s e - C r e e k T e r m i n a l l i n k a g e s 142 C. Methodology 143 D. F o o t n o t e s t o Chapter Three 145 BIBLIOGRAPHY 146 i v LIST OF FIGURES FIGURE Page 1 I n t r a - U r b a n R a i l - T e r m i n a l L i n k a g e s 17 2 I m p o r t / E x p o r t S t r u c t u r e o f S e l e c t e d Vancouver I n d u s t r i e s '20 3 Vancouver Urban T r u c k i n g Study Q u e s t i o n n a i r e 29 4 R e l a t i o n s h i p Between Truck Routes and Consignment Flows 32 5 T r u c k i n g O p e r a t i o n s : Route B i a s 36 6 E f f e c t o f Route D i s t a n c e on Pick-Up and D e l i v e r y P a t t e r n s 46 7 Zone Map 47 8 Mean Zone S i z e by D i s t a n c e From Zone 6 49 9 C l a s s i f i c a t i o n o f Zones by T r a v e l - T i m e from Zone 6 51 10 T o t a l Urban A r e a , Number o f Zones and Consignments v s . A c c e s s t o Zone 6 52 11.. Urban A r e a , I n d u s t r i a l Land, Zones and Consignments (from Zones 6 and 17) v s . I n d u s t r i a l C o n c e n t r a t i o n 54 12 Mean I n t r a - Z o n e D i s t a n c e by D i s t a n c e from Zone 6 56 13 L o c a t i o n U n c e r t a i n t y v s . A c c e s s t o Zone 6 * .58 14 R e g r e s s i o n C o - e f f i c i e n t s . 67 15 Zone 6 R e g r e s s i o n 68 16 Zone 17: -Observed v s . P r e d i c t e d Flows 69 v FIGURE Page 17 Schematic R e p r e s e n t a t i o n o f shipments R e l a t e d t o a R a i l / T r u c k T e r m i n a l Zone 73 18 T r i p a r t i t e D i v i s i o n o f T e r m i n a l Zone shipments 76 19 Commodity Type D i s t r i b u t i o n T e s t s 81 20 I n t e r p r e t a t i o n o f K e n d a l l ' s Tau C . 86 21 Commodity Weight D i s t r i b u t i o n T e s t s 89 22 Commodity Weight - R a i l Components 91 23 Commodity Weight - R a i l - T e r m i n a l Components 92 24 Commodity Weight - Other Components 94 25 Commodity Weight - R a i l - T e r m i n a l G e n e r a l F r e i g h t Consignments ' 95 26 Commodity Weight - R a i l - T e r m i n a l F a b r i c a t e d M a t e r i a l Consignments 96 27 R e l a t i o n s h i p o f D e n s i t y w i t h D i s t a n c e from Urban C e n t r e 100 28 D e n s i t y o f I n d u s t r i a l Land-Use 102 29 Flow Component D i s t r i b u t i o n T e s t s : C l a s s e s A g g r e g a t e d by I n d u s t r i a l C o n c e n t r a t i o n 106 30 Zone 6 Component D i s t r i b u t i o n s v s . I n d u s t r i a l C o n c e n t r a t i o n 108 31 Zone 17 Component D i s t r i b u t i o n s v s . I n d u s t r i a l C o n c e n t r a t i o n 109 32 Zone 6 D e l i v e r i e s v s . I n d u s t r i a l C o n c e n t r a t i o n 110 33 R a i l - T e r m i n a l Components vs I n d u s t r i a l C o n c e n t r a t i o n 111 v i FIGURE Page 34 A d j u s t e d Component D i s t r i b u t i o n s vs I n d u s t r i a l C o n c e n t r a t i o n 112 35 Other Components v s . I n d u s t r i a l C o n c e n t r a t i o n 114 36 Commodity Type D i s t r i b u t i o n s by I n d u s t r i a l C o n c e n t r a t i o n : CN R a i l - T e r m i n a l Component 116 37 Commodity Type D i s t r i b u t i o n s by I n d u s t r i a l C o n c e n t r a t i o n : CP R a i l - T e r m i n a l Component 117 38 C l a s s i f i c a t i o n o f Zones by T r a v e l - T i m e from Zone 17 118 39 Flow Component D i s t r i b u t i o n T e s t s : C l a s s e s A g g r e g a t e d by T r a v e l - T i m e 120 40 R a i l - T e r m i n a l Component D i s t r i b u t i o n v s . T r a v e l Time 121 41 R a i l - T e r m i n a l Component D i s t r i b u t i o n s : E l a s t i c i t y w i t h Respect t o T r a v e l - T i m e 122 42 Other Component D i s t r i b u t i o n s : E l a s t i c i t y w i t h Respect t o T r a v e l - T i m e 123 43 I n d u s t r i a l Land v s . A c c e s s 126 44 I n d u s t r i a l C o n c e n t r a t i o n v s . T r a v e l - T i m e 127 45 P o t e n t i a l I n d u s t r i a l C o n t a c t s v s . T r a v e l - T i m e 128 46 E l a s t i c i t y o f P o t e n t i a l I n d u s t r i a l C o n t a c t s w i t h Respect t o T r a v e l - T i m e 129 47 D i s t a n c e Decay Exponent V a l u e s f o r T e r m i n a l Zone D i s t r i b u t i o n Components 131 48 L i n k a g e I n t e n s i t y o f R a i l - T e r m i n a l Component D i s t r i b u t i o n s w i t h Respect t o T r a v e l - T i m e 132 v i i FIGURE Page 49 L i n k a g e I n t e n s i t y o f Other Component D i s t r i b u t i o n s w i t h Respect t o T r a v e l - T i m e 133 50 Commodity Type v s . T r a v e l - T i m e : CP R a i l -T e r m i n a l Component 135 51 Commodity Type v s . T r a v e l - T i m e : CN R a i l -T e r m i n a l Component 136 v i i i 1 CHAPTER ONE  INDUSTRIAL LINKAGES AND RAIL TERMINALS A. INTRODUCTION The pattern of i n d u s t r i a l location in an urban area i s the re s u l t of a large number of separate location decisions made by in d i v i d u a l firms. As each decision i s made and implemented, the i n d u s t r i a l environment i n which subsequent decisions are made i s altered. The d i s t r i b u t i o n of industry at any given time i s thus both the re s u l t of and the precondition for an ongoing process. This dynamic concept can be extended to include other location decisions being made i n the c i t y ; the indus-t r i a l sector i n t e r a c t s , to varying degrees, with a l l aspects of i t s milieu. A dynamic framework acknowledges the i n t e r -dependent nature of urban growth, but i t s i m p l i c i t complexity makes analysis of the process very d i f f i c u l t . I t i s necessary to simplify the problem by i d e n t i f y i n g r e l a t i v e l y s e l f - c o n t a i n -ed components which can be analyzed, and re-integrating the set of results into an in c l u s i v e theory. For example, while an understanding of the i n d u s t r i a l location process i s a necessary input into a model of urban growth, a p r i o r stage i n the analysis requires interdependency within the i n d u s t r i a l sector be studied to arr i v e at a theory of i n d u s t r i a l location. The l o g i c of t h i s approach i s based on two arguments. The f i r s t i s one of p r a c t i c a l i t y , i n that an a l l - i n c l u s i v e theore-t i c a l structure requires the inc l u s i o n of a set of i n t e r - r e l a t e d variables too large for empirical analysis. Rogers states that: "In the area of urban s p a t i a l structure, the con-sequences of a series of i n t e r a c t i n g a c t i v i t i e s 2 which vary i n space as well as i n time must be i d e n t i f i e d and measured... The overwhelming complexities associated with analyzing human behaviour make f u t i l e a l l e f f o r t s to derive relationships between observable factors. The general r e s u l t i s a complex model of r e a l i t y which defies empirical v e r i f i c a t i o n . " ! The second argument i s based on the assumption that there exists a causal l i n k between the l o c a t i o n a l pattern of industries and the d i s t r i b u t i o n of other urban a c t i v i t i e s . This has been suggested by Kain: "The location of manufacturing i s e s p e c i a l l y c r i t i c a l i n determining metropolitan s p a t i a l structure, since the l o c a t i o n a l decisions of most manufacturing firms are l a r g e l y unaffected by the d i s t r i b u t i o n of metro-p o l i t a n population. Manufacturing determines the l o c a t i o n a l decisions of urban households, not vice versa."2 It i s l i k e l y that the d i s t r i b u t i o n of labour, viewed as a factor of production, i s of more importance for i n d u s t r i a l location than Kain suggests. However, the influence of i n d u s t r i a l d i s t r i b u t i o n on that of other sectors i s of primary si g n i f i c a n c e . This study focuses on certain aspects of i n t e r a c t i o n among i n d u s t r i a l establishments i n order to gauge the r e l a t i v e importance of interdependency factors on location patterns. I n d u s t r i a l i n t e r a c t i o n can be broken down into two components: physical linkages (goods flow) and non-physical linkages (information flow). The analysis w i l l be l i m i t e d to a consider-ation of physical linkages. The movement of f r e i g h t between firms incurs costs which can be d i r e c t l y attributed to s p a t i a l separation, whereas the s p a t i a l component of e l e c t r o n i c inform-ation flow i s minor. It may be noted that many aspects of face-to-face communication and "paper" information flow are s i m i l a r 3 to goods flow, but t h e i r e f f e c t on location i s d i f f i c u l t to assess. In any case, i t i s reasonable to assume that goods movement i s the major linkage factor i n the i n d u s t r i a l sector, whereas information flows are more s i g n i f i c a n t for the location of o f f i c e functions. B. INDUSTRIAL LINKAGES 1 . LINKAGES AND LOCATION THEORY P a r t i a l equilibrium theories of i n d u s t r i a l location have been developed along three main l i n e s , according to whether emphasis i s placed on production factors, demand factors, or competition for land. There have recently been attempts to synthesize these approaches into models which adhere more close l y to real-world conditions. These have more relevance to the question at hand; they s p e c i f i c a l l y set out to discern the r e l a t i v e importance of the various factors to i n d u s t r i a l location. Due to t h e i r increased complexity, the composite. models tend to be more descriptive than a n a l y t i c a l . Market or demand oriented theories of location (based on 3 work by C h r i s t a l l e r and Losch) present a model which optimally locates independent, monopolistic producers so that revenue i s maximised. An evenly d i s t r i b u t e d market i s assumed to exist on an isomorphic transportation surface; locations are deter-mined by minimising t o t a l distance from a producer to the market. The theory emphasizes the importance of forward linkages to f i n a l demand, but ignores a l l backward linkages and intermediate forward linkages. As such, i t i s applicable to interdependent i n d u s t r i a l production only i f scale economies are non-existent, and even then the important linkages would tend to be backward ones (as i n the 18th century t e x t i l e industry). The theory does, however, have importance i n explaining the location of other economic a c t i v i t i e s (such as r e t a i l and service functions) which determine the land rent gradients of urban land use models, which are discussed below. The f i r s t attempt to characterize the location of i n d i v i d -4 ual i n d u s t r i a l plants at a regional scale was by Weber. This model i s e n t i r e l y based on production factors; the firm i s seen as locating at a point (and does not compete for space), the market and location of inputs also are points and demand i s given. The optimal location i s the point at which the cost of assembling inputs and transporting output to the market i s minimized. Weber's formulation assumed constant transport rates, constant production functions and the i n a b i l i t y to substitute inputs. These conditions have subsequently been 5 relaxed by Isard, Hoover and Moses, and the analysis has been extended to take into account l o c a l s i t e advantages, such as low-cost labour supply, which would o f f s e t transportation costs. In general, the approach i d e n t i f i e s the location of an industry as the point of lowest production costs, exclus-ive of land costs. As such, i t would seem to emphasize both backward and forward physical linkages as the cost of goods flow between firms could e a s i l y be substituted for the more f a m i l i a r situations of raw material assembly and output for f i n a l demand. The optimal location of primary processors would thus be p a r t l y determined by forward linkages, and that of intermediate processors primarily by backward linkages. 5 As the model i s s t a t i c , there i s no i n d i c a t i o n of a locating sequence. This implies that the r e l a t i v e importance of linkage d i r e c t i o n i s indeterminate. A model of location based on land competition was f i r s t formulated by Von Thunen. 6 Economic a c t i v i t i e s are assigned to concentric rings around a central market through bid-rents, such that intensive land users are able to outbid less inten-7 sive users for s i t e s closer to the centre. Alonso used t h i s model as the basis of his theory of urban land use. He demon-strates that an urban land rent gradient declining monotoni-c a l l y with distance from the centre represents a market equilibrium. The a b i l i t y of a user to successfully compete for land i s again dependent on the i n t e n s i t y of use. Land competition theories predicate the rent gradient on the centralized location of economic a c t i v i t y , insofar as trans-portation costs produce the gradient. But t h i s implies that a l l linkages must occur between the i n d i v i d u a l firm and the c i t y centre; c l e a r l y i f there i s to be any flow between firms, either a l l firms must be located at the centre, or at least the vast majority of linkages must occur there. Agglomeration economies can be postulated as the focusing force causing centralized concentration. However, as the a b i l i t y to compete for land i s dependent upon i n t e n s i t y of use, industries would presumably locate i n an outlying r i n g (as i n the Burgess con-centric zone model). Thus even though i n d u s t r i a l firms may be located close to each other, the cause of proximity i s assigned not to intermediate production l i n k s but to the s i m i l a r i t y of t h e i r reliance on f i n a l demand and service sector linkages and 6 with the s i m i l a r i t y of t h e i r i n t e n s i t y of land use. The model assumes an isomorphic transportation surface. If th i s i s relaxed to allow for r a d i a l transportation routes of r e l a t i v e l y easy access, the land rent surface w i l l have sectoral va r i a t i o n s . This would r e s u l t i n an even greater concentration of i n d u s t r i a l location, although the d i r e c t i o n of linkages would s t i l l be towards the centre. In t h i s way, land competi-t i o n theory can arrive at a complex, functionally segregated model of urban location without allowing for interdependence within functions; the role of i n d u s t r i a l linkages i s not con-sidered to be of s u f f i c i e n t importance to a f f e c t the analysis. If a subsequent stage of the model i s postulated which does allow for l i n k s to e x i s t i n a d i r e c t i o n other than towards the centre, then linkages between firms could r e s u l t i n l o c a l i z e d humps (or subcentres) i n the rent gradient. But the change i n the assumption of c e n t r a l i t y of a c t i v i t y underlines the s t a t i c nature of t h i s model: the two equilibrium states cannot be connected by a growth process. However, i t does present a conceptual method of combining Weberian and land competition models. I f land costs are viewed as an input, a firm would then locate at the point of lowest production costs by minimiz-ing the sum of transportation and land costs; transportation cost minimization could be determined by Weberian analysis while rents would be set by competition between various land users. g This form of synthesis has been used by Fales and Moses to analyze the structure of i n d u s t r i a l location i n mid-19th century Chicago. I t was found that, while aggregated i n d u s t r i a l location could best be explained by Weberian analysis of transportation costs, the location of d i f f e r e n t types of industry could only be understood by also allowing for land competition between industries- In analyzing the ef f e c t s of technological change i n transportation on i n d u s t r i a l location, 9 Moses and Williamson concluded that the introduction of the truck i n the 1920's reduced intraurban goods movement :costs to such an extent that a c c e s s i b i l i t y advantages of central locations were more than o f f s e t by land costs. Hamer has extended th i s argument by suggesting that the present l e v e l of transportation technology has made the costs of intraurban goods movement an i n s i g n i f i c a n t factor i n loca-t i o n decisions, e s p e c i a l l y as l i n e a r production methods, which greatly increase i n d u s t r i a l land requirements, have made land 1 0 costs the c r i t i c a l factor. This implies two stages of e q u i l i -b r i a : the f i r s t (through the impact of r e l a t i v e l y high cost, transportation) results i n a decreasing rent gradient, while the second (given conditions of the e x i s t i n g rent gradient but low transportation costs) results i n the decentralization of a c t i v i t y . While Hamer admits that his model i s more applicable to older c i t i e s , the survival of the rent gradient (even i f i t i s allowed to flatten) seems problematic. In any case, both stages of his model can be derived without any reference to i n d u s t r i a l linkages. We are l e f t with a number of possible interpretations of these theories as they relate to i n d u s t r i a l linkages. Weberian theory emphasizes the importance of linkages, but t h e i r impact on- location i s dependent on t h e i r costs. I f other s i t e related 8 production costs are more s i g n i f i c a n t , as has been suggested i s the case i n the intraurban setting, then linkages may only be of importance i n regional location. On the other hand, as land costs are determined by a c c e s s i b i l i t y , differences i n land costs may be attributable at least i n part to the strength of i n d u s t r i a l linkages. Another consideration i s that the h i s t o r i c a l decline i n transportation costs i s l i k e l y at an end, and a sharp increase i n costs i s probable i n the near future. This would tend to re-emphasize Weberian loca-t i o n factors at the urban scale, as well as to increase the rent gradient slope. Linear production methods make i n d u s t r i a l plants extensive users of land, so i t i s conceivable that i n -dustries w i l l tend to locate i n suburbanized clu s t e r s . Increasing transportation costs also imply a lowering of labour mobility; the a c c e s s i b i l i t y of locations along t r a n s i t l i n e s to labour supplies i s also l i k e l y to increase the tendency to clu s t e r i n g . As i n 19th century c i t i e s , the importance of linkages to location w i l l depend on the r e l a t i v e costs involved i n transporting either goods or people. 2. LINKAGES AT THE LEVEL OF THE FIRM This section w i l l be concerned with i n d i v i d u a l firms. Attention i s focused on two basic assumptions of location theory: the r a t i o n a l i t y of location decisions and the s i m i l a r -i t y of i n d u s t r i a l firms. I w i l l present a characterization of these two aspects with respect to t h e i r impact on i n d u s t r i a l location, again with emphasis on the question of linkages. The assumption of r a t i o n a l economic behaviour, or the theory of r i s k l e s s choice,^"'" i s f u l l y compatible with a s t a t i c 9 equilibrium analysis but runs into problems i n dynamic s i t u a -tions. It i s conceivable that decision-makers are p r o f i t maximizers with f u l l knowledge of a l l relevent production and market factors, s i t e p o s s i b i l i t i e s , etc., i n a given s i t u a t i o n , but the proposition of t h e i r having the a b i l i t y to foresee future conditions i s u n r e a l i s t i c . In other words, a l l loca-t i o n decision involve some r i s k ; the question at hand i s how do r i s k s a f f e c t location decisions and what i s the a b i l i t y of a firm to weigh these risks? In a dynamic economy, the l i n k s a firm has established are constantly subject to change, either through i n t e r n a l events (such as a change i n product l i n e or an increase i n capacity) or external events (relocation of suppliers, increase i n demand for product, e t c . ) . A firm may locate i n such a way as to protect i t s e l f as much as possible from a l l adverse consequences 12 of changes i n the economic environment. As Hoare points out, the p o t e n t i a l linkages of a firm may be as important i n location decisions as are actual linkages; thus a firm may locate with regard to a number of possible suppliers rather than optimally i n regard to the actual supplier, so that reliance on a single source of supply doesn't endanger the firm. This type of "location safety" i s i n fact one aspect of agglo-meration economies. However, i f a firm's inputs and outputs go through a transportation terminal (either to firms within the c i t y or to external firms), i t i s l i k e l y that a location close to the terminal would not be counter-productive should the linkages be altered. Linkages that operate through i n t e r -mediate points are thus more secure, mainly because the 10 l o c a t i o n o f t e r m i n a l s i s r e l a t i v e l y c o n s t a n t . 13 C z a m a n s k i b e l i e v e s t h a t t h e a b i l i t y o f a f i r m t o w e i g h l o c a t i o n r i s k s i s n o t o f f u n d a m e n t a l i n t e r e s t i n an a n a l y s i s o f i n d u s t r i a l d i s t r i b u t i o n , i n t h a t o n l y s u c c e s s f u l f i r m s ( i . e . , b y d e f i n i t i o n t h o s e t h a t e x h i b i t r a t i o n a l b e h a v i o u r ) w i l l s u r v i v e f o r a n y l e n g t h o f t i m e . T h i s p o s i t i o n d o e s n ' t a c c o u n t f o r t h e p o s s i b i l i t y t h a t c h a n c e o r u n f o r e s e e n o c c u r -r e n c e s may a f f e c t t h e r a t i o n a l i t y o f a d e c i s i o n , b u t d o e s s t a t e a v e r y p r a c t i c a l p r i n c i p l e : t h e f u t u r e i s n o t e a s i l y a n a l y z e d . A l t h o u g h much r e s e a r c h h a s b e e n d i r e c t e d t o w a r d s t h e p r o c e s s o f d e c i s i o n - m a k i n g , no t h e o r y h a s y e t b e e n d e r i v e d t o d e t e r m i n e how r i s k s s h o u l d be w e i g h e d . P r o b a b l i s t i c m o d e l s a n d game t h e o r y s h e d some l i g h t on p o s s i b l e methods o f r i s k -t a k i n g b e h a v i o u r , a n d m o d e l s o f d e c i s i o n - m a k i n g show how d e c i s i o n s a r e a r r i v e d a t u n d e r d i f f e r e n t o r g a n i z a t i o n a l c o n -d i t i o n s , b u t t h e a s s u m p t i o n o f r a t i o n a l i t y s t i l l a p p e a r s t o be a g o o d a p p r o x i m a t i o n o f t h e e n d r e s u l t o f t h e d e c i s i o n p r o c e s s . S u b s t i t u t i o n o f o t h e r g o a l s f o r p r o f i t m a x i m i z a t i o n ( s u c h as s a t i s f i c i n g b e h a v i o u r ) d o e s n o t a l t e r t h e a s s u m p t i o n t h a t t h e p r o c e s s i s a b l a c k b o x ; i t o n l y r e d e f i n e s t h e e n d r e s u l t o f t h e p r o c e s s . The s e c o n d a s p e c t r e g a r d i n g l i n k a g e s a t t h e l e v e l o f t h e f i r m i s t h e i m p a c t o f v a r i a t i o n s among i n d u s t r i a l f i r m s . The m o s t o b v i o u s f a c t o r i s t h e a c t u a l t y p e o f p r o d u c t w h i c h t h e f i r m m a n u f a c t u r e s . Much r e s e a r c h h a s b e e n done on t h e l i n k a g e c h a r a c t e r i s t i c s o f i n d i v i d u a l i n d u s t r i e s ; i t h a s b e e n shown t h a t c e r t a i n i n d u s t r i e s ( s u c h as t e x t i l e s , p r i n t i n g a n d f o o d p r o c e s s i n g ) h a v e much s t r o n g e r s p a t i a l l i n k a g e r e q u i r e m e n t s than others. L i t t l e work, however, has been done on the possible influences of variations within industries on linkages. It would appear that differences i n size, i n t e r n a l integration, ownership, production function and age of firms or i n d i v i d u a l plants (in the case of branch operations) could also a f f e c t the s i g n i f i c a n c e of linkages to t h e i r locations. These factors are complex, and t h e i r e f f e c t s on linkage requirements are s t i l l a subject of conjecture. In order to i l l u s t r a t e some of the possible complications involved when intra-industry variations are considered, I w i l l discuss the size factor i n some d e t a i l . Small firms may be more footloose than large firms, as t h e i r lower l e v e l of inputs and outputs would reduce the necess-i t y of r e l y i n g on a single supply source or market outlet. However, large firms are more l i k e l y to operate t h e i r own truck f l e e t s ; these firms thus would tend to locate i n r e l a t i o n to linked firms while small firms would locate with regard to transportation terminal f a c i l i t i e s . On the other hand, large firms are more l i k e l y to operate at a regional or i n t e r -regional scale, either i n terms of markets or sources of supply. As these l i n k s would be maintained by inter-regional c a r r i e r s , locations near terminals (in p a r t i c u l a r port or r a i l f a c i l i t i e s ) would be indicated. The l o c a l i z e d scale of small firms i n -creases the importance of intraurban linkages to t h e i r operations. It has been suggested that small firms are.less s e l f -s u f f i c i e n t than large ones i n terms of infrequent s p e c i a l i z e d service requirements, and so would locate nearer the CBD to gain agglomeration economies. This was not borne out i n a 12 14 study of i n d u s t r i a l location by B l a i r i n Philadelphia. It i s possible that the labour requirements of large firms necessit-ate t h e i r locating i n r e l a t i o n to labour supplies, while small firms are more able to take advantage of cheaper suburban s i t e s . The economies of scale which a large firm can generate may allow for lower transportation costs per unit of output. This would reduce the importance of proximity to linked firms. However, a large firm i s more l i k e l y to be a generator of an i n d u s t r i a l complex, i n that smaller backward or forward linked firms would clu s t e r around i t i n order to reduce the length of t h e i r linkages. The other factors of firm d i f f e r e n t i a t i o n are equally complex, and are i n many cases interconnected. This study w i l l not analyze these factors d i r e c t l y ; a r e l a t i v e l y undifferent-iated categorization of i n d u s t r i a l a c t i v i t y w i l l be u t i l i z e d . This approach i s only p a r t i a l l y dictated by data l i m i t a t i o n s ; a major thrust of the study i s the investigation of the applic-a b i l i t y of location theory to urban i n d u s t r i a l a c t i v i t y as a whole. I t i s anticipated that any s i g n i f i c a n t v a r i a t i o n i n the s p a t i a l consequences of linkage requirements within the i n d u s t r i a l sector w i l l be r e f l e c t e d i n a disagreement between empirical observations and t h e o r e t i c a l expectations. 3. APPROACHES TO LINKAGE RESEARCH Studies concerned with the description and analysis of i n d u s t r i a l linkages t y p i c a l l y approach the question i n one of three ways. The f i r s t , or i n d i r e c t , method i s to investigate i n d u s t r i a l location patterns. The underlying hypothesis of t h i s 13 . approach i s that r e g u l a r i t i e s exhibited by these patterns represent the consequences of linkage e f f e c t s . In other words, firms are assumed to locate at lowest-cost s i t e s ; costs of maintaining necessary backward and forward linkages are thus minimized i n the actual d i s t r i b u t i o n of industries. The strength of areal association between types of industries i s then a r e f l e c t i o n of the importance of linkage association. 15 Examples of t h i s approach are studies by Richter and Streit"'"^. The explanatory power of t h i s method i s l i m i t e d by a reliance on t a u t o l o g i c a l l o g i c and an over-emphasis on .spatial 17 factors of location. I t would appear that the application of t h i s technique at the intra-urban scale would be inappropriate, as the complex set of location factors operative within an urban area cannot be reduced to a single factor based on linkages. The second, or direct, method involves the investigation of actual linkage i n t e r a c t i o n . This may be done either aspat-i a l l y , i n terms of inter-industry input-output flows, or may involve some co r r e l a t i o n of s p a t i a l d i s t r i b u t i o n with flow • * J.- Tt 18 _ , . 19 ^ ^20 , T 21 , information. Hoare , Czamanski, Steed and Lever have undertaken studies i n which inter-industry input-output flows are related to location patterns of various industries to deter-mine those linkages which r e s u l t i n i n d u s t r i a l c l u s t e r i n g . A major problem with t h i s approach i s that the method of c l a s s i -fying industries i s c r i t i c a l to i d e n t i f y i n g linkage r e l a t i o n -ships. The assumption of intra-industry homogeneity may be unfounded i f the categories used are too broad; on the other hand, data requirements of detailed c l a s s i f i c a t i o n schemes are 22 often p r o h i b i t i v e . A related approach which avoids t h i s 14 d i f f i c u l t y involves the use of goods flow data measured between in d i v i d u a l firms rather than i n d u s t r i a l sectors or c l a s s i f i c a -tions. The rel a t i o n s h i p between linkage i n t e n s i t y and s p a t i a l association then can be d i r e c t l y investigated. A poten t i a l drawback i s that the generality of such d i r e c t i n t e r - f i r m research i s lim i t e d . The appropriateness of t h i s method depends on the application of derived r e s u l t s ; i f a nomothetic l e v e l of explanation i s sought, the firms used i n the study must form a representative sample of a l l firms, with a given set of charac-t e r i s t i c s , located within the targeted region. This approach i s used i n the present study. The focus here i s directed towards the investigation of a p a r t i c u l a r linkage s i t u a t i o n (which i s described i n the next section); the generality of the r e s u l t s w i l l be discussed below. The t h i r d approach to linkage research i s related to be-havioural studies of location decisions. The impact of i n t e r -firm linkages on i n d u s t r i a l location are i n f e r r e d by exploring decision-makers 1 perceptions of the r e l a t i v e importance of various location factors. Studies of thi s type include those 23 24 by Cooper and Wood. Many of these studies r e j e c t the basic premises of neo-classical theory of the firm; for example, Wood states that "the p r i n c i p l e s of economic man and perfect competition ignore the very varied structure, goals and con-,25 t r o l s of i n d u s t r i a l organization",; However, i t i s unclear whether the results of behavioural research can be interpreted outside of a t h e o r e t i c a l economic framework. I f a decision-maker perceives a linkage factor to be of great importance to the location decision, t h i s adds to our knowledge of the actual 15 effects of linkages only i f we can impute some form of economic lo g i c to the decision-making process. When undertaken i n con-junction with one of the other types of linkage studies, be-havioural research i s valuable i n establishing the l e v e l of agreement between a firm's observed linkage/location i n t e r -dependency and management's perception of i t s importance. C. RAIL-TERMINAL LINKAGES This study i s based on an analysis of the i n d u s t r i a l linkages of Vancouver's inner core r a i l / t r u c k terminal. Three reasons were paramount i n sel e c t i n g t h i s focus. F i r s t of a l l , r a i l terminal operations are i n t r i n s i c a l l y important i n terms of urban goods movement; they are primary nodes of the trans-portation network and represent s i g n i f i c a n t d i s t r i b u t i o n sink/source locations. Secondly, the economy of Vancouver i s oriented towards regional d i s t r i b u t i o n functions. Trans-shipment f a c i l i t i e s which interface between inter-urban and intra-urban transport modes are thus of great importance. Thirdly, a major urban renewal project (B.C. Place) i s planned for the s i t e of the e x i s t i n g inner c i t y r a i l terminal, and the terminal operation w i l l l i k e l y be relocated i n an outlying area. The e f f e c t of t h i s relocation on the future pattern of i n d u s t r i a l a c t i v i t y w i l l depend on the strength of terminal-industry linkages. 1) TERMINAL LINKAGES Meyburg and Stopher c l a s s i f y urban freight movements into four categories: imports, exports, transient movements and intra-urban movements. They define the l a s t category as 16 "intra-urban c o l l e c t i o n and d i s t r i b u t i o n and l o c a l shipment movement i n which the vehicle, though not necessarily the commodity, has both i t s o r i g i n and destination within the 2 6 same area." Thus shipments between a r a i l terminal and i n d u s t r i a l firms i n the urban area are l o c a l movements even though the commodities involved are i n fact being exported or imported. The intra-urban f r e i g h t flows that are related to a r a i l / truck terminal complex are i l l u s t r a t e d i n Figure 1. The flows shown represent less than truckload (LTL) f r e i g h t that must be trans-shipped at the terminal: piggy-back and bulk-loading operations are not considered. The r a i l terminal serves as the l i n k to other urban centres; linkages between a l o c a l wholesaler and an out-of-town supplier, for example, may operate through the r a i l terminal. The location of the wholesaler within the c i t y i s dependent on the terminal loca-tion to the extent of the importance of the supply linkage. In t h i s way an inter-urban terminal can be considered a surrogate for the actual supplier, and terminal linkages are indistinguishable from d i r e c t i n t e r - f i r m linkages. The flow diagram shows both the d i r e c t and i n d i r e c t net-work of linkages that involve a terminal complex. Direct linkages r e s u l t from f r e i g h t shipments consigned from/to the r a i l terminal. They include shipments that are handled by trucking operations associated with the terminal, providing pick-up, delivery and storage services. Indirect linkages involve the network of goods movement along a production/ d i s t r i b u t i o n chain. A manufacturing firm may locate with 17 FIGURE 1 INTRA-URBAN RAIL-TERMINAL LINKAGES L a r g e M a n u f a c t u r e r I RAIL TERMINAL r S m a l l R e t a i l e r INTRA-URBAN TRUCK TERMINAL S m a l l M a n u f a c t u r e r L a r g e W h o l e s a l e r L a r g e R e t a i l e r S m a l l W h o l e s a l e r < 1 T e r m i n a l - R e l a t e d ( D i r e c t ) F l o w s : I n d i r e c t L i n k a g e s : 18 regard to the terminal location to minimize backward linkages, for example. The wholesaler d i s t r i b u t i n g the manufacturer's products may be located near the manufacturer, or a l t e r n a t i v e l y he may be more concerned with shortening forward linkages to r e t a i l o utlets. In the f i r s t case the influence of the terminal on the wholesaler's location i s clear. In the second case the terminal's influence i s manifested i n the succeeding pattern of fr e i g h t movements between the wholesaler and his customers. The linkage e f f e c t s of a terminal extend to include the f r e i g h t t r a f f i c generated as a r e s u l t of the terminal's location. In t h i s way terminal location i s an important influence not only on the location of d i r e c t l y linked firms but also on the pattern of freight flow throughout the urban area. 2) RAIL TERMINALS AND THE VANCOUVER INDUSTRIAL STRUCTURE The manufacturing structure of Vancouver has been charact-2 7 erized by Steed as "unusual and deceptively narrow". I t \ i s very cl o s e l y t i e d to the requirements of regional resource-based industries, both i n terms of raw-material processing (sawmills, veneer and plywood m i l l s , f i s h processors) and supplying input requirements. "Much of the sheet metal production i s i n the form of cans for the f i s h processing and f r u i t and vegetable plants... Several metropolitan industries have d i r e c t t i e s to the dominant wood products sector, including the truck and t r a i l e r industry through assembly of logging trucks suitable for B r i t i s h Columbia's moun-ta i n conditions; the ship-building industry through construction of newsprint, log and chip barges and tugs; the metal f a b r i c a t i n g industry through produc-ti o n of chains, bolts and wire ropes; and the indust-r i a l machinery and i n d u s t r i a l chemicals industries which service the pulp and paper industry. Much of 19 the expansion and d i v e r s i f i c a t i o n into secondary manufacturing i n Greater Vancouver over recent years has simply involved increasing integration with the p r o v i n c i a l hinterland through the taking up of forward and backward linkages from t h i s primary sector." I n d u s t r i a l a c t i v i t y i n Vancouver t y p i c a l l y involves eit h e r raw material processing for export pr final-product f a b r i c a t i o n , using imported semi-finished inputs, for the regional market. Thus both backward and forward linkages are externally oriented. Figure 2 shows the importance of imports and exports for various i n d u s t r i a l sectors i n terms of the monetary value of flows. The high l e v e l of external linkage suggests that inter-urban terminals represent the source of supply and f i n a l product destination for many of Vancouver's i n d u s t r i a l firms. A s i m i l a r pattern also exists for the wholesaling sector. Vancouver's dominant role as the regional d i s t r i b u t i o n centre emphasizes the importance of trans-shipment operations to the urban economy. 3. THE FALSE-CREEK TERMINAL The r a i l terminal on the north shore of False Creek was o r i g i n a l l y established i n 1886 as the western terminus of the Canadian P a c i f i c transcontinental railway. This terminal be-came the f o c a l point of the development of Vancouver, and, along with the port area a mile to the north, was the centre of manufacturing and warehousing a c t i v i t y i n the pre-automobile period. Areal expansion of the c i t y , induced by the general decrease i n intra-urban transport costs a f t e r World War II, resulted i n a much more dispersed pattern of a c t i v i t y . Hardwick states that: FIGURE 2 IMPORT / EXPORT STRUCTURE OF SELECTED VANCOUVER INDUSTRIES I m p o r t s % o f t o t a l i n p u t t o i n d u s t r y E x p o r t s % of t o t a l o u t p u t from i n d u s t r y E x p o r t s % of t o t a l o u t p u t t o r e s t o f B.C. E x p o r t s % o f t o t a l o u t p u t t o r e s t o f Canada E x p o r t s % o f t o t a l o u t p u t t o U.S. E x p o r t s % o f t o t a l o u t p u t t o r e s t o f W o r l d A g r i c u l t u r e , F o r e s t r y , F i s h i n g , M i n i n g C o n s t r u c t i o n Food and Be v e r a g e s Wood I n d u s t r i e s P a p e r and A l l i e d P r o d u c t s C h e m i c a l and P e t r o l e u m N o n - M e t a l l i c P r o d u c t s M e t a l F a b r i c a t i n g P r i n t i n g and P u b l i s h i n g M a n u f a c t u r i n g NEC Trade and T r a n s p o r t 29 16 38 39 32 82 7 32 34 36 7 60 2 44 84 52 78 44 58 49 83 60 1 7 2 9 6 16 74 39 34 35 34 21 11 0 16 12 36 3 0 21 10 47 1 8 32 0 11 34 0 0 5 2 0 2 12 0 0 9 31 0 0 0 1 4 1 9 ro o S o u r c e : C o m p i l e d from H. C r a i g D a v i s , An I n t e r i n d u s t r y S t u d y o f t h e M e t r o p o l i t a n V a n c o u v e r  Economy, U.B.C. Urban Land'Economics R e p o r t No. 6, 1974. 21 "At the end of the Second World War, most of the wogk places were i n Vancouver's CBD and i t s adjacent waterfront area. This was the centre of commerce, o f f i c e s , d i s t r i b u t i o n and port a c t i v i t i e s , and pro-duction... The location pattern of various economic a c t i v i t i e s has changed. The inner c i t y has seen an expansion of professional, managerial, c l e r i c a l , and service occupations, r e f l e c t i n g the changing role of the central c i t y . Production and labouring occupa-tions have declined... Many i n d u s t r i a l and warehousing a c t i v i t i e s have moved from the central c i t y to the suburbs and most of the new large-scale enterprises have located i n the peripheral zone... The eastern s h i f t of population has made parts of Burnaby most desirable locations for trucking and warehousing a c t i v i t i e s . " 29 The area immediately to the east of the present CBD, which includes the False-Creek terminal, has been characterized as "a zone of t r a n s i t i o n where waterfront, warehouse, and 30 i n d u s t r i a l functions are being phased out." An important issue here i s whether the t r a n s i t i o n away from i n d u s t r i a l functions i n t h i s area i s a r e s u l t of the changing urban structure, or whether the persistence of some of these a c t i v i -t i e s i n the face of transportation cost and i n f r a s t r u c t u r a l changes and zoning pressures indicates the continuing existence 31 of a viable i n d u s t r i a l complex. The decision to relocate the terminal to a suburban s i t e may be either the l o g i c a l end r e s u l t of an h i s t o r i c a l process of i n d u s t r i a l out-migration, character-ized by a lessening of linkage impact on location decisions, or an attempt to hasten the removal of i n d u s t r i a l a c t i v i t y from the urban core to make way for the less i n t r u s i v e o f f i c e , commercial and entertainment functions exemplified by the B.C. Place development proposal. I hope to explore the role of the terminal as a factor i n providing urban agglomeration economies, and thus i t s e f f e c t as an intra-urban "growth pole", through an analysis of i t s linkage pattern. 22 The False-Creek terminal area includes the following transport-related operations: 1. CPR r a i l y a r d f a c i l i t i e s 2. CP Transport trucking operations a. r a i l / t r u c k trans-shipment f a c i l i t i e s (LTL freight) b. container and truck t r a i l e r r a i l / r o a d transfer f a c i l i t i e s . c. CP Express parcel transport operation 3. B.C. Hydro Railway r a i l y a r d a. r a i l f r eight o f f i c e b. bus parcel express depot 4. Trucking and Freight-Forwarding firms: 10 major firms have warehousing f a c i l i t i e s i n the terminal area. 5. Warehousing: 6 independent importers and public warehousing operations are located adjacent to the terminal The entire terminal area i s slated for redevelopment; a l l of these operations w i l l be forced to relocate i n the near future. This rapid change i n the urban-core land-use pattern, and the r e s u l t i n g a l t e r a t i o n of the c i t y ' s intra-urban indust-r i a l linkage network, i s likej*y to have far-reaching effects on the future d i s t r i b u t i o n of economic a c t i v i t y i n the region. I t i s thus necessary to investigate the present pattern of linkages associated with the terminal to f u l l y understand the consequences of i t s relocation. 23 D. FOOTNOTES TO CHAPTER ONE 1. Rogers, A., "Theories of Intra-Urban Spatial Structure: A Dissenting View", Land Economics, Vol. 43, No. 1,1967; page I I I . 2. Kain, J . , "The Di s t r i b u t i o n and Movement of Jobs and Industry", i n Wilson, J.Q. (ed.), The Metropolitan Enigma, Harvard University Press, Cambridge, 1968. 3. a f t e r Nourse, H., Regional Economics, McGraw H i l l , New York, 1968. 4. i b i d . 5. af t e r Hamer, A., Indu s t r i a l Exodus from the Central  City, Heath and Co., Lexington, 1973. 6. Nourse, H., op. c i t . 7. Hamer, A., op. c i t . 8. Fales, R. and Moses, L., "Land-Use Theory and the Spatial Structure of the Nineteenth-Century City", Papers of  the Regional Science Association; Vol. 28, 1970. 9. Moses, L. and Williamson, H., "The Location of Economic A c t i v i t y i n C i t i e s " , American Economic Review, 1967. 10. Hamer, A., op. c i t . o 11. a f t e r Cooper, M.J., The Indu s t r i a l Location Decision-Making Process, University of Birmingham, Centre for Urban and Regional Studies Occassional Paper 34, 1975. 12. Hoare, A., "Three Problems for Industrial Linkage Studies", Area, Vol. 10, No. 3, 1978; page 218. 13. Czamanski, S., Industrial Clusters, Dalhousie University, Halifax, 1974; page 10. 14. B l a i r , J., "Industrial P o l a r i z a t i o n and the Location of New Manufacturing Firms: An Empirical Application", Regional Science Research Ins t i t u t e Discussion Paper Series, No. 89, 1976; page 33. 15. Richter, C.E., "The Impact of In d u s t r i a l Linkages on Georgraphic Association", Journal of Regional Science, Vol.9, 1970. 16. S t r e i t , M.E., "Spatial Associations and Economic Linkages Between Industries", Journal of Regional Science, Vol. 9, 1969. 2k 17. The concept of " s p a t i a l separatism" to denote a neglect of non-spatial factors i s elucidated i n Sack, R.D., "The Spatial Separatist Theme i n Geography", Economic  Geography, Vol. 50, No. 1, 1974. 18. Hoare, A.G., "Linkage Flows, Locational Evaluation, and I n d u s t r i a l Geography: A Case Study of Greater London", Environment and Planning, Vol. A7, 1975. 19. Czamanski, S., Ind u s t r i a l Clusters, op. c i t . 20. Steed, G.P.F., "Commodity Flows and Interindustry Linkages of Northern Ireland's Manufacturing Industries", T i j d s c h r i f t Voor Econ. en Soc. Geografie, Sept./Oct. 1963. 21. Lever, W.F., "Industrial Movement, Spa t i a l Association and Functional Linkages", Regional Studies, Vol. 6, 1972. 22. The question of c l a s s i f i c a t i o n e f f e c t s i s discussed i n d e t a i l i n Czamanski, S., "Some Empirical Evidence of the Strengths of Linkages Between Groups of Related Industries i n Urban-Regional Complexes", Papers of the Regional Science  Association, Vol. 27, 1971. 23. Cooper, M.J., The I n d u s t r i a l Location Decision-Making  Process, op. c i t . 24. Wood, P.A., "Industrial Location and Linkage", Area, Vol. 2, 1969. 25. i b i d ; page 32. 26. Meyburg, A.H. and Stopher, P.R., "A Framework for the Analysis of Demand for Urban Goods Movements", Transportation  Research Record, Vol. 496, 1974; page 69. 27. Steed, G.P.F., "Intrametropolitan Manufacturing: Spatial D i s t r i b u t i o n and Locational Dynamics i n Greater Vancouver", Canadian Geographer, Vol. 17, No. 3, 1973; page 2 34. 28. i b i d ; page 238. 29. Hardwick, W.G., "Vancouver: The Emergence of a Core-Ring Urban Pattern", i n Gentilcore, R.L., Geographical  Approaches to Canadian Problems, Prentice_Hall Ltd., Scarborough, 1971; pages 116 to 117. 30. i b i d ; page 114. 31. Bourne, L.S., "Comments on the Transition Zone Concept", Professional Geographer, Vol. 20, No. 5, 1968; page 315. 2 5 CHAPTER TWO  ANALYSIS OF RAIL TERMINAL LINKAGES A. INTRODUCTION In t h i s chapter, an analysis of urban goods flows i s under-taken to es t a b l i s h the i n d u s t r i a l linkage pattern of the False Creek r a i l / t r u c k terminal. Data gathered in 1975 by the Swan Wooster Engineering Co. for a study e n t i t l e d Evaluation of  Urban Trucking Rationalization i n Vancouver 1 i s used as the basis of the analysis. The data base i s f i r s t described and evaluated. Two techniques are then used to describe terminal linkages. The f i r s t involves a multiple regression approach to determine i f the operations of the two inner area r a i l / t r u c k terminals are si m i l a r . To accomplish t h i s the goods move-ment pattern of one of the terminal areas i s used to predict the pattern of the other terminal area. The second technique employed i s a component analysis of the goods movement data. The results of the two methods are discussed i n the following chapter. B. .. DESCRIPTION OF DATA 1. The Vancouver Urban Trucking Study The Vancouver Urban Trucking (VUT) Study was undertaken to investigate the pattern of urban trucking in the Vancouver region. S p e c i f i c a l l y , the report evaluates the impact of a proposed inner core consolidated terminal on o v e r - a l l trucking e f f i c i e n c y . The authors proceed by f i r s t estimating the demand for intra-urban goods movement. The operating 26 c h a r a c t e r i s t i c s of the e x i s t i n g trucking system in supplying transport services to meet thi s demand are then analyzed. Thirdly, modifications to the system that would r e s u l t from the operations of a consolidated terminal are determined by modelling procedures. The f i n a l stage i s the evaluation of e f f i c i e n c y gains provided by inner area terminal consolidation. The study i s thus focussed primarily on the response of the trucking industry to s p a t i a l aspects of demand for truck-ing services and to a s p e c i f i c nodal change i n the urban transport network. A purely s t a t i c transportation/land-use model i s u t i l i z e d . Demand i s treated as being exogenous to trucking operations, so that considerations of transport e f f i c i e n c y are lim i t e d to t r a f f i c - r e l a t e d factors. The ef f e c t s of trucking operations on land-use patterns, such as the potential of a consolidated terminal to influence the d i s t r i -bution of demand for trucking services, do not enter the analysis. An important aspect of the s t a t i c modelling approach to the present study concerns data requirements and the means used to s a t i s f y them. It was necessary for the authors of the VUT study to estimate both the demand for goods movement services and the actual trucking operations which s a t i s f i e d t h i s demand. Ideally, two independent surveys would be undertaken to gather t h i s information: one to estimate demand and one to estimate supply. However, a s t a t i c transportation model assumes demand to be independent of changes i n e f f i c i e n c y of supply. S p e c i f i c a l l y , both the t o t a l l e v e l and the d i s t r i b u -t i o n of demand would remain constant regardless of the e f f e c t 27 of a consolidated terminal on transport costs. The e l a s t i c i t y of transport demand i s e f f e c t i v e l y set at zero over the cost (or e f f i c i e n c y ) range under consideration. In t h i s way the demand for transport i s always being met; there i s no residual or p o t e n t i a l demand involved. Under this assumption a single survey, can be u t i l i z e d to estimate both supply and demand. For predictive transportation research generally, equating demand with estimated supply i s not overly misleading i f the time horizon i s r e l a t i v e l y short. In the present case, how-ever, equilibrium assumptions may l i m i t the v a l i d i t y of the analysis. Demand for transport services must be equated with t h e o r e t i c a l l y derived surrogate "potential" measures, which may be inaccurate. Such measures are u t i l i z e d i n predictive research (for example, a t t r a c t i o n variables i n regression equations), but t h e i r v a l i d i t y i s based on predictive power or goodness of f i t rather than on i n t e r p r e t a b i l i t y . The VUT study i s based on a survey of trucking operations 2 within the Vancouver region. The target truck population (or "working inventory") was defined as including a l l trucks operating within the region for the purpose of carrying general f r e i g h t . This population was i d e n t i f i e d through the use of B.C. p r o v i n c i a l motor vehicle "registrations and Insurance Corporation of B r i t i s h Columbia records. Of a t o t a l truck r e g i s t r a t i o n i n the region for 1974/75 of 85,264, i t was found that 16,248 s a t i s f i e d a l l conditions for i n c l u s i o n i n the population. S p e c i f i c exclusions include 2,285 bulk c a r r i e r s (dump, cement, garbage and tank trucks) and 1,534 s p e c i a l -use vehicles (owned by large public and private establishments 28 such as B.C. Hydro, B.C. Telephone, e t c . ) . A mail-survey was c a r r i e d out to determine trucking operations, based on a random sample s t r a t i f i e d by f l e e t size and truck type and si z e . The o v e r - a l l sampling rate was 28.4% of the population (4,615 trucks). The response was 35.8% of the sample, or 10.2% of the population. A large number of non-valid responses reduced the e f f e c t i v e sample to 8% of the working inventory. . The survey was composed of a detailed questionnaire to be completed by in d i v i d u a l truck drivers. The questionnaire was designed to reveal the a c t i v i t i e s of the respondent during one s p e c i f i c working day. The respondent indicated the o r i g i n and destination of each t r i p made, t r i p s t a r t i n g and ending times, the type and weight of consignments picked up and delivered, and the type of business carried on at each stop. The Vancouver region was divided into 70 zones, which determin-ed the location code of each stop. The zones were designed to be compatible with the areal aggregations used by the GVRD for land-use data-bases ( i . e . Residential Development Sub-Areas). 2. USE OF VUT DATA TO ESTIMATE DEMAND The data gathered for the VUT study focuses on the oper-ating c h a r a c t e r i s t i c s of the trucking industry. As the primary concerns of the study are t r a f f i c - r e l a t e d e f f i c i e n c y aspects of f r e i g h t transport supply, t h i s approach i s appropriate. It i s also possible to extract information regarding consign-ment flows from t h i s data base. Respondents recorded the type and weight of goods picked-up or delivered at each stop, so that the actual o r i g i n and destination of each shipment can be determined (figure 3). In t h i s way, re a l i z e d demand for goods FIGURE 3: VANCOUVER URBAN TRUCKING STUDY QUESTIONNAIRE EXAMPLE SHEET The following list shows many different types of freight and stops for example purposes only and is not intended to represent the activities ol any one truck. T R U C K Mil F A R E READING: AT S T A R T OF FIRST TRIP 2Z3>00 (miles), AT END OF LAST TRIP 25 ,245 (miles) STOP NUMBER ARRIVAL AND DEPARTURE TIME AT EACH STOP LOCATION OF EACH STOP BY ZONE (SEE MAP) TYPE OF BUSINESS OR ACTIVITY AT EACH STOP TYPE & QUANTITY* OF FREIGHT HANDLED AT EACH STOP LOAD FACTOR FOR TRUCK OR TRAILER AFTER EACH STOP ARRIVAL DEPARTURE PICK-UP DELIVERY START OF 1st TRIP ?& t ruck terminal office* supplied -»iooo\\o foot cdrron<9 Zooo\o •furniture e**>\\e X FREIGHT PRELOADED ON €1 TRUCK AT START T OF 1st TRIP 1st STOP office, buildmg of f i ce * u p p l i e « &*«oib 2nd STOP l O * " * * * private f u r n i t u r e **a«>lb. '/<*> 3rd ,—^  2 2 grocery food c a r t o n * iotx>\\o. e m p t y 4th 17 ra i l terminal rvid&Kinery fe>ttor>* fu l l 5th 37 £K4<t/mill industry lumber l^cx? f bm m a c h i n e r y £>te>n*? 6th 42 icn^truction *ite lumber \^oo fbno empty 7th Cb 3 terminal dutto pert*. 5*=wo II* % 8th <£> fre ight •forwarding •h&r-mincil auto p«rT-^» «?ooolb rool^> 4o»e> |le> e^nptu, 9th &> wholesale, warehouse furniture ^ o o l b . clothing l<?olb. pa i ^ t i<2ooi*> ^4 10th 2 rete l l 1 v t o r e fu rn i tu re *ww lb. clothing roollci paint 11th mariuFacKir* industry <;lec.+T-iccil equipment 12th — 52 tru<ik terminal e\edrr\ca\ equipment 4 ton-> e.rY> (7T(J 30 movement i s estimated. However, t h i s method of working back-wards from supply to demand has many p i t f a l l s , as t h i s caveat from the Executive Summary of the VUT study c l e a r l y implies: "The survey did r e s u l t i n the creation of a data base suitable for the evaluation of o v e r - a l l system-wide e f f e c t s of various options. The data base did not, however, allow micro-level analysis on s p e c i f i c links, or for s p e c i f i c commodities or 4 truck types." It should be noted that the use of t h i s survey even for i t s intended purpose (i . e . , system-wide effects) has met with c r i t i c i s m . The authors of The Economics of Urban Goods Move-ment state that "(the VUT study) i s not very useful i n estimat-ing t o t a l urban flows...The study i s based on survey data that may be suspect. In the present study I am using the VUT data for the risky purpose of micro-level analysis. It i s thus necessary to describe and evaluate the inadequacies of the data i n some d e t a i l so that the appropriate l e v e l of confidence which can be placed on d i f f e r e n t aspects of the analysis can be determined. This discussion i s divided into two sections. The f i r s t deals with the general problems involved with demand estimation from a truck sample. The second section i s concerned with s p e c i f i c aspects of the Swan-Wooster data. i) GENERAL ISSUES It i s important to emphasize that even a complete survey of a l l trucking a c t i v i t i e s w i l l not supply a l l the information needed to understand demand for FGM. F i r s t of a l l , p o t e n t i a l 31 f l o w s a r e i g n o r e d ; o n l y what Hoare terms a c t u a l l i n k a g e s are i n c l u d e d . S e c o n d l y , a c t u a l consignment e n d p o i n t s a r e not n e c e s s a r i l y t h e same as shipment o r i g i n s / d e s t i n a t i o n s . The f i r s t p roblem has been d i s c u s s e d above i n some d e t a i l . I w i l l j u s t n o te t h a t i t i s more s e r i o u s when t h e d a t a i s b e i n g used f o r p r e d i c t i o n t h a n f o r a n a l y s i s , as f e e d back e f f e c t s a r e i g n o r e d . F o r s i m i l a r r e a s o n s , however, a n a l y s i s i n a dynamic framework must a l s o be u n d e r t a k e n w i t h c a r e . Demand e s t i m a t e s g e n e r a t e d from s u p p l y ( t r u c k i n g ) d a t a have no t e m p o r a l d i m e n s i o n o t h e r t h a n t h a t r e s u l t i n g from t h e t i m e -span o f t h e s u r v e y . W h i l e t h i s may seem o b v i o u s , t h e r e l a t i v e s t a b i l i t y o f t h e b u i l t environment can g i v e t h e appearance o f c o n s t a n c y t o demand l e v e l s . The second problem i l l u s t r a t e s t h e need t o m a i n t a i n a s t r i c t d i s t i n c t i o n between t h e concept o f a consignment and t h a t o f a shipment. A shipment r e l a t e s t o s u p p l y , i n t h a t i t i s t h e q u a n t i t y o f goods a c t u a l l y t r a n s p o r t e d between two p o i n t s i n a s i n g l e t r i p ( u s u a l l y by one v e h i c l e ) . A c o n s i g n -ment r e l a t e s t o demand; i t i s t h e u n i t o f goods c o n t r a c t e d f o r d e l i v e r y between two p a r t i e s . I f a consignment i s t r a n s -p o r t e d d i r e c t l y and i n one t r i p from a s u p p l i e r t o a customer, i t i s t h e n i d e n t i c a l t o a shipment. But d i f f i c u l t i e s a r i s e i n s i t u a t i o n s where t r a n s p o r t a t i o n o f t h e consignment i s e i t h e r n o t d i r e c t o r r e q u i r e s more t h a n one v e h i c l e . The f i r s t o f t h e s e p o s s i b i l i t i e s i s b e s t e x p l o r e d by means o f an i l l u s t r a t i o n . 'In f i g u r e 4, a t r u c k r o u t e i s d e s -c r i b e d by t h e s o l i d l i n e and i n d i v i d u a l shipments t r a n s p o r t e d by t h e t r u c k a r e i n d i c a t e d by dashed l i n e s . The t r u c k ' s base FIGURE j RELATIONSHIP BETWEEN TRUCK ROUTES AND CONSIGNMENT FLOWS T r u c k R o u t e : C o n s i g n m e n t : 33 terminal i s located i n zone 5. Shipments e,b and c are pre-loaded on the truck at the terminal, and shipment g i s unloaded at the terminal at the end of the day. From t h i s information the consignment origin/destination zones of a, d and f can be determined. The o r i g i n of e, b and c and the destination of g are unknown. Only i f we have knowledge of the previous and/or subsequent movements of the consignments can t h e i r actual origins/destinations be determined. But t h i s informa-ti o n w i l l r a r e l y be gained through even a comprehensive truck-ing survey. There i s a timing problem, as the transport seg-ments will- l i k e l y occur on d i f f e r e n t days. More importantly (for i t i s possible that the duration of the survey would be s u f f i c i e n t to include shipping dates of a l l t r i p segments), i t i s l i k e l y that the segments w i l l be undertaken by d i f f e r e n t • trucks. Unless demand-related data i s gathered i n the survey (by i n vestigating b i l l s of lading, for example), i t would be impossible to connect shipping links to form the composite consignment l i n k . The d i f f i c u l t y of multi-stage shipments i s tempered somewhat i f terminals can legitimately be considered as con-signment origin/destination endpoints. In the present study emphasis i s placed on the linkage effects of two r a i l terminals. Only the intra-urban transport of goods brought i n or shipped out of the c i t y by r a i l i s of i n t e r e s t ; inter-urban o r i g i n / destination i s not. Sim i l a r l y , wholesale warehouses may represent interim storage or trans-shipment locations, but more commonly are either actual consignment endpoints or can be treated as such. For example, the location of ultimate 34 destination establishments may be more influenced by the warehouse linkage than by the actual supplier linkage. The treatment of certain shipments as consignments i s thus defen-s i b l e when the major concern i s l o c a t i o n a l interdependency of land-use. This i s not the case for questions related to production interdependency. In summary, i f the terminal i n zone 5 i s a r a i l - t r u c k trans-shipment point or a wholesale warehousing operation: a l l shipments i n the figure may r e a l -i s t i c a l l y be treated as consignments for the purposes of t h i s study. On the other hand, i f i t i s simply a trucking company's warehouse, the o r i g i n of shipments e,b,c, and the destination of shipment g (i . e . , zone 5) i s s o l e l y related to the operat-ing c h a r a c t e r i s t i c s of the trucking industry. Treating these shipments as consignments for the purpose of linkage analysis would be misleading. The second source of confusion between consignments and shipments ( i . e . multiple vehicle consignments) i s only common when bulk commodities are involved. As these shipments are of l i t t l e importance for t h i s study (and were s p e c i f i c a l l y exclud-ed from the VUT survey) i t i s unnecessary to discuss them i n d e t a i l . Up to now I have been considering only those problems that would be encountered i f the demand for fr e i g h t transport was being estimated from a complete knowledge of truck movements. The next step i s to investigate t h i s process when the s t a r t i n g point i s a sample of trucking a c t i v i t y . The primary d i f f i c u l t y i s that a random sample of trucks w i l l not provide a random sample of consignments. Truck routes are often s p a t i a l l y concentrated (for' dispatching efficiency),• t h i s s p a t i a l bias w i l l appear i n the sample of consignments. There are two variables complicating t h i s sampling problem: the number of trucks serving d i f f e r e n t parts of the region and the number of consignments delivered by each truck. I f the number of consignments delivered (by trucks of a given size/type category) i s a normally d i s t r i b u t e d variable,, the number of trucks operating on a given route w i l l be proport-ional to the fr e i g h t flow along that route. A random sample of trucks would then provide a representative (but not independent1 sample of shipments c a r r i e d along that route. Furthermore, the number of trucks operating i n various parts of the region would then be proportional to the fr e i g h t flows coming from/going to the sub-areas. The concordance between number of trucks and number of consignments i s dependent on the variance of the shipments per truck variable. Even i f t h i s r e l a t i o n s h i p - i s close, routing e f f e c t s may bias the derived consignment sample. The problem i s i l l u s t r a t e d i n figure 5. Consider the s i m p l i f i e d case of 3 zones served by 3 non-overlapping truck routes. I f there are two trucks on each route and 50% are sampled, the p r o b a b i l i t y of s e l e c t -ing one truck from each route i s . 4 . There i s a.60%-chance that the sample w i l l be t o t a l l y unrepresentative of the actual consignment d i s t r i b u t i o n . Now, 50% of consignments w i l l , have been "sampled", but the sig n i f i c a n c e of the sample d i s t r i b u t i o n w i l l be much lower than would that of an independent sample (assuming that each truck transports more than one consignment). This example demonstrates an "all-or-nothing" aspect of FIGURE 5 TRUCKING OPERATIONS: ROUTE BIAS T r u c k R o u t e s : 37 d i s t r i b u t i o n data gained from truck sampling; determining confidence lev e l s applicable to such data i s extremely d i f f i c u l t . The problem i s most acute where there i s a large imbalance among inter-zonal flows. For example, i f there are three trucks operating between zones 1-2, two between 1-3, and one between 2-3, a 50% sample w i l l give a biased consignment d i s t r i b u t i o n except i n the u n l i k e l y event that one truck from each route i s chosen (probability=.3) and the r a t i o of consign-ments carr i e d by these trucks approaches 3:2:1 (in the proper order). S p a t i a l p a r t i t i o n i n g of shipments into r a t i o n a l i z e d truck routes can occur both between and within trucking firms. For example, a firm may s p e c i a l i z e i n serving a p a r t i c u l a r area of an urban region and also dispatch i t s trucks by subdivid-ing t h i s area into a number of separate routes. Firms operating large f l e e t s are more able to consolidate shipments into e f f i c i e n t routes than are firms with small f l e e t s . Thus a sampling procedure which focuses on a few large firms i s bound to r e s u l t i n biased data unless the sample size i s very large. On the other hand, a smaller across-the-board sample i s much more l i k e l y to be representative of consignment flows. The r e l i a b i l i t y of the data depends on how trucks enter the sample as well as on the t o t a l proportion sampled. The survey procedure used i n the VUT study attempted to remove routing bias i n a number of ways. F i r s t of a l l , the target population was s t r a t i f i e d according to f l e e t s i z e and truck type, and each sub-population was then randomly sampled. However, the v a l i d response rate was much higher 38 for large f l e e t s (defined as more than 30 vehicles) than for small f l e e t s (55% of large f l e e t s as compared to 22% for small 7 f l e e t s ) . This r a t i o worsens when only v a l i d returns for trucks active i n the Vancouver area on the survey date are considered (31% of sampled large f l e e t s , 10% of sampled small f l e e t s ) . 31% of the active v a l i d response consisted of large-f l e e t trucks, while these represent only 14% of the target population. On the other hand, small-f l e e t response was 54% of the t o t a l but represented 76% of the population. In order to account for the si t u a t i o n , the authors of the VUT study state that "there was an extremely favourable response from most of the larger operators whereas the smaller operators g tended to be reluctant to provide information." To compensate for the unbalanced response rates, the VUT study expanded the sample data to the population by weighting s t r a t i f i e d sub-groups according to response. This method was e f f e c t i v e i n terms of reproducing o v e r - a l l truck movements. Manual t r a f f i c counts of trucks crossing selected screenlines were carried out to provide an independent test of the accuracy of the estimation of truck (as opposed to consignment) movements. For trucks with a G.V.W. of between 8,000 and 32,000 pounds, estimated error was 9.35%; for vehicles over 32,000 pounds, estimated error was 14.14%. The authors of the study state that "considering the d i f f i c u l t i e s i n making manual counts ... the errors ... are highly acceptable. The comparison between (expanded survey) assignments and screen-l i n e counts provides an important piece of evidence which supports the contention that the sample survey i s indeed repre-39 9 sentative of vehicle movements on the survey day." In the present study, use of expanded data was deemed inadvisable. The trucking firms operating from the r a i l terminal zones run mainly large f l e e t s ; expansion factors based on o v e r - a l l regional f l e e t size and mix proportions would not be suitable for analysis of these zones. I t i s conceivable that the higher response rate of large operators provides a larger and s i g n i f i c a n t l y more representative consignment sample for the terminal zones than i t does for the region as a whole. S t i l l , the p r e j u d i c i a l e f f e c t s of route r a t i o n a l i z a t i o n on consignment data cannot be t o t a l l y discounted. For example, the three CP Express trucks based i n the terminal zone that are included i n the v a l i d sample have the following route c h a r a c t e r i s t i c s : Truck A: 17 pick-ups i n zone 16; a l l delivered to zone 6 Truck B: 5 d e l i v e r i e s to zone 11 3 " " " 12 13 " " " 13 ] _ II II ' ' ]_4 8 " " " 15; a l l picked-up in:zone 6. Truck C: 28 d e l i v e r i e s to zone 6; a l l picked-up i n zone While t h i s example represents an extreme case of routing bias i t i s not the only case. Most of the large f l e e t s exhibit some degree of route r a t i o n a l i z a t i o n . The d i f f i c u l t y i s that the magnitude and d i r e c t i o n of bias by d e f i n i t i o n remain unknown. The s t a t i s t i c a l conse-quences of possible hidden biases can be major: 40 "The Central Limit Theorem rests on the assumption of random sampling. A random sample i s any sample selected by a chance mechanism with known chances of selection; the chances of selection need not be equal for a l l samples so long as they are known. Once selection b i a s . . . i s allowed to intrude, then the whole apparatus of inference from sample to population f a l l s to the ground, i n that the determination of probable l i m i t s of error i n estimating population parameters ... becomes impossible." Three methods of c o n t r o l l i n g the problem are applied i n the present analysis. F i r s t , trucks dispatched on very l i m i t e d routes and making a large number of pick-ups and d e l i v e r i e s that had l i t t l e or no apparent connection to r a i l terminal operations are eliminated from the sample. The main examples of discarded trucks are four vehicles owned by a uniform-supply company; each truck made between 20 and 30 service stops i n one zone. Secondly, outlying values of inter-zonal consign-ment flows are c l o s e l y investigated to determine whether they can be attributed to routing c h a r a c t e r i s t i c s . Low o u t l i e r s are very d i f f i c u l t to interpret; the absence i n the sample of trucks operating i n a s p e c i f i c zone or group of zones may be suspected, but not tested. Thirdly, various aggregated zone configurations are u t i l i z e d to examine d i s t r i b u t i o n determinants of flow components. The rationale i s that zones aggregated on the basis of a s p e c i f i c independent variable can be treated as homogeneous for the purpose of t e s t i n g the e f f e c t of that variable on consignment d i s t r i b u t i o n . The number of aggregated zones for a given variable must be f a i r l y small to allow for the diminuation of routing d i s t o r t i o n ; the improvement i n r e l i a b i l i t y i s thus achieved at some cost. 41 i i ) ISSUES RELATED TO VUT DATA In th i s section I w i l l outline s p e c i f i c c h a r a c t e r i s t i c s of the VUT data that a f f e c t the r e l i a b i l i t y of the analysis below. The discussion w i l l focus on non-spatial aspects of the sampling procedure, s p a t i a l sampling design (areal aggregation) and the c l a s s i f i c a t i o n methods used for selected c r i t i c a l variables. a) Non-Spatial Aspects of the Sample Any survey of a dynamic process e n t a i l s an i m p l i c i t l y h i e r a r c h i c a l sampling technique. The f i r s t stage requires that the time period(s) sampled be representative of the t o t a l time period over which the analysis i s to be applied. In the case of f r e i g h t transport, goods movement refers to a rate of flow; f r e i g h t shipped during the sampling period must be pro-portional to t o t a l a c t i v i t y occurring during the target time-span. The second stage requires that the flow sample be representative of the t o t a l flow occurring during the survey period. In terms of the f i r s t stage of the sampling hierarchy, the target span for which the analysis i s assumed to be v a l i d i s related to the average "li f e s p a n " of a location decision made by a firm. The r e l a t i v e locations of linked firms determine flows (by d e f i n i t i o n ) ; i t i s reasonable to assume that flows w i l l p e r s i s t as long as r e l a t i v e locations remain constant. I f a firm's present location i s suboptimal but s t i l l r e l a t i v e l y e f f i c i e n t , i t w i l l r e s i s t relocating u n t i l the pot e n t i a l gains of doing so are greater than the discounted value of investment sunk i n the present location. In other words, the expected time-span for persistence of linkage patterns would be some-what less than the time allowed for f u l l plant depreciation -say i n the neighbourhood of 6 to 8 years. The sample period would occur i n the middle of t h i s span, assuming that the i n d u s t r i a l s i t e occupation time i n the urban region i s a normally d i s t r i b u t e d variable. The VUT survey questionnaire covers trucking a c t i v i t i e s for a period of only one day (24 hours). The decision to l i m i t the survey i n t h i s way was due to p r a c t i c a l consideration r e p l i c a t i n g the survey on 2 or more separate days or extending the reporting period (to include a f u l l week, for example) would have involved a prohibitive amount of data handling, and the added inconvenience for truckers l i k e l y would d r a s t i c a l l y reduce the response rate. The c r i t i c a l question then i s whether the f r e i g h t movements taking place on the p a r t i c u l a r day sampled are representative of movements occurring through-out the 6-8 year "population" period. This question has two d i s t i n c t aspects: the f i r s t relates to temporal v a r i a b i l i t y of t o t a l goods flow; the second to the frequency of a c t i v i t y across p a r t i c u l a r linkages. In terms of t o t a l flow, i t i s u n l i k e l y that monthly or seasonal variations i n the i n t e n s i t y of i n t e r a c t i o n are great enough to seriously undermine the a p p l i c a b i l i t y of a one-day sample to.an analysis of i n d u s t r i a l linkages. Construction a c t i v i t y i n Vancouver i s d e f i n i t e l y more intense during the summer months, but shipments of materials to work s i t e s are of minimal importance for linkage studies- Sim i l a r l y , trans-shipment of raw materials i s often seasonal (wheat shipments, for example) but these goods do not enter into the i n t r a -urban flow. The v a r i a t i o n i n frequency of s p e c i f i c linkage movements does, however, present d i f f i c u l t i e s . A one-day sample ri s k s either over or under estimating the importance of infrequent flows, but as the p r o b a b i l i t y of doing so i s inversely rated to flow frequency, the r i s k i s s t a t i s t i c a l l y acceptable. On the other hand, i t has been suggested that the frequency of goods movement i s i n i t s e l f an important determinant of location. A one-day sample does not allow any investigation of t h i s hypothesis. Theoretically, then, a one-day sample should be s u f f i -cient to analyze most aspects of i n d u s t r i a l linkages. Un-fortunately, the date selected for the survey (Thursday, Oct. 16, 1975) "experienced a record r a i n f a l l which continued through-out almost the entire day. This condition resulted i n most of the outside construction industry being i d l e d which correspondingly c o n t r i b u t e d to the high propor-t i o n of i d l e vehicles". At f i r s t glance t h i s would appear to be a catastrophic setback to hopes of obtaining a representative sample. The untimely occurrence of extreme weather conditions (despite i t s low probability) exemplifies the r i s k s inherent i n a one-day sample. While the o v e r - a l l goods flow estimates are undeniably weakened by the large number of i d l e d vehicles, the heavy r a i n f a l l may have actually improved estimates for application i n the present study. As the focus here i s on i n d u s t r i a l 44 linkages, transport of construction materials constitutes "noise" in the data. The proportion of linkage-related shipments postponed due to inclement weather was l i k e l y quite small, so that the trade-off between l o s t v a l i d data and reduction i n unuseful or misleading data i s probably p o s i t i v e . Note that t h i s s i t u a t i o n i s very s i m i l a r to that caused by the decision to exclude -bulk materials from the sample: the v a l i d i t y of the survey i s somewhat weakened for i t s o r i g i n a l l y intended purpose but i s improved i f used for 12 the type of linkage analysis undertaken i n t h i s study. The second aspect of the sampling procedure leads to the question of whether the sample i s representative of t o t a l flows occurring on the day the sample was taken. The main issues concerning the r e l i a b i l i t y of the VUT data are related to the structure of the survey, and have been discussed i n d e t a i l i n the section on generic problems encountered when trucking supply data i s translated into an estimate of goods movement. I might add here that there i s an additional l i m i t a t i o n s p e c i f i c to the VUT survey. The a c t i v i t y question-naire (see figure 3) leaves space for 30 t r i p descriptions. If a truck made more than 30 t r i p s , only the t o t a l number of extra t r i p s i s requested. Many of the incomplete forms-exhibit s i g n i f i c a n t route s t a b i l i t y f or the f i r s t 30 t r i p s ; i t could be argued that the 30 t r i p cut-off acts to diminish o v e r - a l l routing bias problems. But the cut-off also discriminates against a l l short-haul vans, e s p e c i a l l y those d e l i v e r i n g small consignments. I f the o r i g i n and destination of consignment are close together, the time needed to complete delivery i s short; a single truck i s able to supply many more short linkage trip-demands i n one day than long ones. Even when route con-s o l i d a t i o n i s allowed for, t h i s remains true. In figure 6, a truck operating between zones 1 and 2 can make more d e l i v e r -ies i n a given period of time than can a truck of s i m i l a r capacity operating between zones 1 and 3 (assuming consign-ments of equal s i z e ) . A r b i t r a r i l y l i m i t i n g t o t a l d e l i v e r i e s w i l l thus tend to disturb measurement of the e f f e c t of distance on goods flow; s p e c i f i c a l l y , distance decay w i l l appear to be less than i s actually the case. Note that i f the incomplete .records are expanded to allow for missed t r i p s , routing bias w i l l be exacerbated. b) Areal Aggregation The zone map (figure 7) used i n the VUT study describes the areal aggregations that are treated as t r i p o rigin/destina-t i o n points. The design of the map preceded actual sampling; a zone map was included i n each questionnaire package so that respondents could note down the zone number i n which each stop was located rather than the actual firm name or address. It i s thus impossible to disaggregate the resultant flow data from the zone design used i n the VuT study. This method of deciding a p r i o r i on the best aggregation scheme has the p r a c t i c a l advantages of vastly simplifying both questionnaire response and subsequent coding of consignment orig i n s and destinations. I t does not allow for any refinement of the s p a t i a l scale of analysis, but t h i s i s not a serious concern i f the zones are designed so that the mean uncertainty of origin/destination locations i s of the same magnitude as error FIGURE 6 EFFECT OF ROUTE DISTANCE ON PICK-UP AND DELIVERY PATTERNS FIGURE 7 ZONE MAP 48 i n other aspects of the sample. There are two ways v a r i a t i o n i n zone size can be measured. The f i r s t i s based on comparison of zone areas. The zone map i s some what misleading i n t h i s respect, as the outer zones include areas that extend beyond the built-up part of the urban region. I f only the land-uses c l a s s i f i e d as urban (according to the GVRD land-use data base) are included, the e f f e c t i v e size of these zones decreases s u b s t a n t i a l l y . For example, zone 70 (West Vancouver) has a t o t a l area of 25,295 acres of which only 7,706 acres or 30% are c l a s s i f i e d as urban. For the purpose of i d e n t i f y i n g truck stop locations, a r e s t r i c t i o n of the d e f i n i t i o n of zone area to include only land under urban uses i s reasonable p a r t i c u l a r l y on the North Shore, as outlying parts of these zones are composed of very rugged t e r r a i n (zones 62, 63, 64, 66, 68 and 70). On the other hand, western and southern edges of the region are intensively farmed. There i s a p o s s i b i l i t y that some shipments to these zones (especially 46, 49 and 50) would be oriented to a g r i c u l t u r a l a c t i v i t i e s ; i t i s important to recognize t h i s p o t e n t i a l d i s t o r t i o n of t r i p -end estimations. The d i s t r i b u t i o n of zone size shows a strong p o s i t i v e skewness, even i f zones are defined by urban rather than t o t a l area. The variance i s also very large, but the major proportion of i t i s caused by a few outer zones. As can be seen i n the zone map, the size of zones generally increases with distance from the downtown core of the c i t y (zones 3 and 4). Figure 8 tabulates the number of zones and the mean zonal area against distance from zone 6, the location of the CP r a i l / t r u c k terminal. FIGURE 8 MEAN ZONE SIZE BY DISTANCE FROM ZONE 6 A g g r e g a t e T r a v e l - T i m e ( M i n u t e s ) No. of Zones Mean A r e a ( A c r e s ) Cum. No. o f Zones Mean A r e a ( A c r e s ) 0 - 3 1 2^0 1 240 3.1 - 5 5 367 6 346 5.1 - 10 10 524 16 457 10.1 - 12 U 754 30 596 12.1 - 15 12 1053 42 . 726 15.1 - 1 8 10 1374 52 851 18.1 - 22 12 2667 64 1191 22.1 - 30 3 1 096 67 1187 > 3 0 3 13667 70 1 722 O v e r - a l l Mean: 1722 A c r e s S t a n d a r d D e v i a t i o n : 3082 A c r e s C o - e f f i c i e n t o f V a r i a t i o n : 179$ * T h i s r e f e r s t o t h e mean s i z e o f a l l zones w i t h i n t h e o u t e r boundary o f t h e d i s t a n c e c l a s s u nder c o n s i d e r a t i o n . 50 Distance i s measured i n minutes of t r a v e l time. This independ-ent variable i s more accurately referred to as access; the e f f e c t of transportation bottlenecks (bridges, tunnels, etc.) cannot be accommodated i n l i n e a r distance units. (The deriva-t i o n of t r a v e l times i s explained below, and figure 9 i s a map showing the access c l a s s i f i c a t i o n of zones.) The l a s t column i n figure 8 gives the mean area of zones up to the given distance class. In figure 10, the cumulative frequency of area, zones, and origins/destinations of consignments coming to/going from zone 6 are graphed against t r a v e l time. Note that 50% of a l l consignments are shipped a distance of less than 12 minutes t r a v e l time, and the zones containing the receiving/shipping locations contain only a l i t t l e over 11% of the t o t a l urban area. However, 43% of the zones are within 12 minutes of zone 6; i f the o v e r - a l l zonal mean was the same as the mean of these 30 zones ( i . e . , 596 acres), the region as a whole would be divided into 202 zones. Although i r r e g u l a r i t y of areal aggregation causes d i s t o r -13 tions i n the analysis of the data, the benefits of standard-ized areal units (achieved by using a g r i d system, for example) must be weighed against p r a c t i c a l considerations. In the case of the VUT study, i t was clear i n the design stage that ship-ments are negatively correlated to some degree with increasing distance from the urban centre. As i t was necessary to sample the entire urban region, two choices were l e f t : either use a very large number of standardized zones: or allow a wide range of size v a r i a t i o n , with the majority of zones concentrated near the centre of the study area. A large number of zones can FIGURE 9 CLASSIFICATION OF ZONES BY TRAVEL-TIME FROM ZONE 6 T r a v e l -Time i n M i n u t e s 0-3 3.1-5 5.1-10 m 10.1-12 12.1-15 15.1-H 18.1-22 22.1-30 •30 • CP T e r m i n a l FIGURE 10 TOTAL URBAN AREA, NUMBER OF ZONES AND CONSIGNMENTS v s . ACCESS TO ZONE 6 C u m u l a t i v e F r e q u e n c y 10 20 30 AO T r a v e l - T i m e i n M i n u t e s 53 be confusing to respondents and may thus r e s u l t i n a lowered response rate. Irregular zones thus achieve a balance between a reasonably high l e v e l of s p a t i a l resolution i n areas of flow concentration and s i m p l i c i t y of data c o l l e c t i o n and handling. For present purposes, zonal arrangement i s perhaps not as well designed as i t i s for an o v e r - a l l investigation of urban goods movements. I am focussing here on two central zones, and flows between a l l other zones are ignored. As a result, there are proportionally more c e n t r a l l y located shipments i n the relevant sub-samples than there are zones (see figure 10), but the zone design of the VUT study i s s t i l l more suitable for t h i s analysis than a standardized g r i d design (of less than 200 zones). The r e l a t i o n s h i p of consignment origins/destinations and zone arrangement to the d i s t r i b u t i o n of manufacturing a c t i v i t y i s graphed i n figure 11. The derivation of manufacturing concentration categories i s discussed below, and i l l u s t r a t e d i n figure 28. Zones were designed to coincide with land-use boundaries. Manufacturing a c t i v i t i e s are t y p i c a l l y segregated from other uses, so that the 40% of t o t a l manufacturing land that i s most highly concentrated forms d i s t i n c t areas that are contiguous with zones. In d u s t r i a l linkages are c l e a r l y assoc-iated with manufacturing a c t i v i t y ; the arrangement of zones with regard to land-use allows a more detailed examination of t h i s association than would a s p a t i a l l y regular zone arrangement. The second method of measuring zone size v a r i a t i o n i s based on the degree of "shrinkage" entailed i n regarding areas as 14 points. In other words, the amount of uncertainty involved i n locating a trip-end within a zone i s based on the p r o b a b i l i t y FIGURE 11 URBAN AREA, INDUSTRIAL LAND, ZONES AND CONSIGNMENTS (FROM ZONES 6 & 17) v s . INDUSTRIAL CONCENTRATION 1.0 C u m u l a t i v e F r e q u e n c y 20 40 60 I n d u s t r i a l C o n c e n t r a t i o n ( P e r c e n t I n d . Land) 55 d i s t r i b u t i o n of distances between any two points i n that zone. McCarty has developed a measure of shrinkage using the average distance between a l l possible pairs of points by assuming that the zones are square. The average intra-zonal distance, d', i s equal to . 52v/7A, where A i s the area of the zone under consideration; d' i s then the l e v e l of error accepted by using a zone of area A. With reference to the zone map (Figure 7), i t i s not un-reasonable to use equivalent area squares to approximate the majority of zone shapes. While there are a few d e f i n i t e l y elongated zones ( s p e c i f i c a l l y zones 8, 67, 40, 42 and 58) , the use of d' as an approximation of mean intra-zone distance i s s u f f i c i e n t l y accurate for descriptive purposes. The second column of Figure 12 gives the mean d 1 values for zones i n each t r a v e l time category. Each entry i s calcu-lated as the d' for a hypothetical zone equal i n area to the mean of a l l zones i n that category, rather than as the arithmetic mean of i n d i v i d u a l zone d' values. The method chosen always results i n a mean d' equal to or larger than that calculated with the al t e r n a t i v e method. The more con-servative approach i s used because i t s r e l i a b i l i t y i s not as dependent on the actual d i s t r i b u t i o n of sampled trip-ends, and thus can be more confidently used as a test of expected design-based e r r o r . 1 ^ For the Vancouver region as a whole, the VUT zone map allows for an average error of .85 miles by the conservative method against .70 miles by the other method. The actual d' calculated for shipments going to/coming from zone 6 i s .71 miles; t h i s again shows that the flows analyzed FIGURE 12 MEAN INTRA-ZONE DISTANCE BY DISTANCE FROM ZONE 6* T r a v e l - T i m e Mean I n t r a - Z o n e D i s t a n c e C u m u l a t i v e Mean ( M i n u t e s ) ( m i l e s ) I n t r a - Z o n e D i s t a n c e 0 - 3 .32 .32 3.1 - 5 .39 .38 5.1 - 10 .47 • U 10.1 - 12 .56 .50 12.1 - 15 .67 .55 15.1 - 18 .76 .60 1 8.1 - 22 1 .06 .71 22.1 - 30 .68 .71 30 2 . 4 0 .85 Mean I n t r a - Z o n e D i s t a n c e i s c a l c u l a t e d as f o l l o w s : D j ^ = .52^/Area ; i . e . , Dj ^ = t h e a v e r a g e ' I - Z d i s t a n c e o f a h y p o t h e t i c a l zone w i t h a r e a = mean a r e a o f zones i n c l a s s under c o n s i d e r a t i o n . I f c a l c u l a t e d as ~Dj ^ °f a H zones ( i . e . , a r i t h m e t i c mean o f a l l 70 zones o v e r - a l l Dj ^ = -70 m i l e s , w i t h a s t a n d a r d d e v i a t i o n o f .50 m i l e s and a c o - e f f i c i e n t o f v a r i a t i o n o f 71.3?. 57 i n t h i s study are not overly affected by the large s i z e of outlying zones. An index of location uncertainty i s graphed i n Figure 13. This i s a p o t e n t i a l measure, and i s calculated as follows: U = d ' T / d ' ± where U = proportion of uncertainty remaining a f t e r zone effects accounted for d' T = mean intra-zone distance i n t o t a l area, A, which, i s subdivided into n zones. d'^ = d' of a hypothetical zone of areaA/n. The l e v e l of uncertainty for the entire region, as depicted by the value of U for aggregated t r a v e l time categories of up to 3 8 minutes, i s .12 . The d' T value for the region i s 7.1 miles, while the d'^ value (.70 zones) i s .85 miles; a 12% uncertainty l e v e l s t i l l remains. The actual value for zone 6 oriented origin/destination locations i s 10%. Note that U i s a r e l a t i v e measure. Endpoints of shipments t r a v e l i n g 12 minutes or less can be located with an average error of 18%, or .5 miles. The o v e r - a l l U of 12% represents an average error of .85 miles. As d' i s a function of the square root of zone area, a large number of zones would be needed to reduce U s i g n i f i c a n t l y below 10%. For example, a U value of 5% ( i n f e r r i n g a d ' i value of .36 miles for the Vancouver region) requires 402 zones. The diminishing returns of areal d i s -aggregation suggests that the a p r i o r i zone construction necessary f o r general surveys may not be an appropriate method for more s p e c i f i c linkage research. The r e l a t i v e advantages FIGURE 13 LOCATION UNCERTAINTY v s . ACCESS TO ZONE 6 -i 1 , , 10 20 30 40 T r a v e l - T i m e i n M i n u t e s 59 of a l t e r n a t i v e approaches, such as highly focused, detailed small sample surveys, w i l l be discussed i n the f i n a l chapter. The f i n a l problem presented by the areal aggregation scheme of the VUT data involves the i d e n t i f i c a t i o n of terminal related shipments. The CP r a i l / t r u c k terminal i s located p a r t l y i n zone 5 and p a r t l y i n zone 6, although by far the bulk of trucking a c t i v i t i e s associated with the terminal i s located i n zone 6. Unfortunately, zone 6 also includes many other land-uses: over hal f of the t o t a l 2 40 acre area l i e s outside the terminal d i s t r i c t . Much of t h i s area contains older' warehouses and wholesaling operations, and the northern end of the zone consists of the western extension of Chinatown and part of Gastown, with the small r e t a i l shops, restaurants and rooming houses t y p i c a l of these areas. Zone 17, containing the CN terminal, also included a wide range of other a c t i v i t i e s . Clearly, i t i s inaccurate to treat a l l shipments i n these zones as terminal-related consignments. However, disaggregation of the data within zones i s very d i f f i c u l t . As a single record was used to code each consignment, the type of a c t i v i t y c a r r i e d on at origin/destination locations i s available i n automated form for only one trip-end. I t was thus necessary to re-examine v a l i d returned questionnaires to determine which shipments legitimately were associated with the terminal opera-tions. This procedure involved a great deal of guesswork; the majority of respondents did not have the time or i n c l i n a -t i o n to describe stop locations accurately. The technique used to disaggregate terminal-zone shipments into flow components i s described below. 60 A l l i e d to the problem of i d e n t i f y i n g non-terminal flows within terminal zones was the reverse s i t u a t i o n : some consign-ments related to a terminal didn't enter either terminal zone, but Cgine to/went from warehouses in adjacent zones. F i n a l consignment end-points were replaced i n the sample by shipment end-points, for reasons outlined above. I decided to treat intermediate shipments as consignments and ignore intermediate l i n k s ; i n any case, these were impossible to determine due to the short survey period. The implications of t h i s decision on the v a l i d i t y and generality of the analysis w i l l be explored below. c) CLASSIFICATION OF COMMODITY TYPE The most important element i n describing a single consign-ment i s to accurately i d e n t i f y the type or class of commodity involved. In order to use t h i s information to gain an under-standing of the pattern of linkages, the c l a s s i f i c a t i o n scheme on which i d e n t i f i c a t i o n i s based must be s u f f i c i e n t l y s p e c i f i c to allow for disaggregation of consignments with markedly d i f f e r e n t c h a r a c t e r i s t i c s . On the other hand, the l e v e l of disaggregation f e a s i b l e i s inversely related to o v e r - a l l sample s i z e : a highly disaggregated c l a s s i f i c a t i o n scheme applied to a small sample w i l l r e s u l t i n subsets too unreliable for s t a t i s t i c a l purposes. A second consideration i s that inte r p r e t a t i o n of the categories must be both clear and consist-ent. In the VUT survey, commodity c l a s s i f i c a t i o n s were assigned to i n d i v i d u a l consignments from respondents' descriptions. These were often vague. Also, a wide range of d e f i n i t i o n s for commonly used words, such as "parts" may have been i n use 61 by the various truck drivers applying these d e f i n i t i o n s . The c l a s s i f i c a t i o n scheme used i n the VUT study i n i t i a l l y contained 14 categories, and then was reformed into 6 aggregated categories. The data available for t h i s study was coded only i n the aggregated form. The d e f i n i t i o n s of the 6 categories used i n t h i s c l a s s i f i c a t i o n system (based on SCC codes) are: 1. Food Products: 2. Fabricated Materials: 3. General Freight: 4. Machinery: 5. Furniture: 6. Other: Processed Food Products Pulp & Paper & Paper Board Chemicals and Related Products Petroleum and Products Metal Fabricated Products Non-Metallic Basic Products Other Fabricated Products Machinery Parts Transportation & Communication Parts Rubber Tires and Tubes Other Equipment & Tools Apparel & Accessories Personal & Household Goods Medical & Pharmaceutical Products Printed Matter & Of f i c e Supplies Miscellaneous End Products General Freight Machinery Road Motor Vehicles Transportation and Communication Equipment Furniture, Fixings, Refrigerators & Stoves Meat & Meat Preparations Fish & Other Marine Products Dairy Products & Eggs Food Products - Farm Produce Beverages Fodder, Feed Crude Non-Metallic Minerals Waste & Scrap Other Crude Materials (Inedible) Firearms, Weapons & Ammunition explosives Containers & Closures 62 These are evidently very gross categories; the inc l u s i o n of personal and household goods i n the General Freight category, for example, i s unfortunate. However, the sample sizes of flow d i s t r i b u t i o n s used i n the present study are small, and further disaggregation by commodity type greatly decreases the r e l i a -b i l i t y of s t a t i s t i c a l inference. The main problem i s that the vast majority of a l l consignments going to/coming from the two terminal zones i s clustered i n two categories: General Freight and Other. Further disaggregation of these two c l a s s i f i c a t i o n s would be desirable. d) LAND-USE DATA Information available r e l a t i n g to s p a t i a l v a r i a t i o n i n economic a c t i v i t y was r e s t r i c t e d to land-use data gathered by the GVRD. The c l a s s i f i c a t i o n scheme employed i n t h i s data base consists of 19 categories: 1. Total 2. U t i l i t i e s , Transport, Communication 3. Automotive, Wholesale, Outdoor R e t a i l & Commercial, Nurseries 4. Residential Commercial (Motels,Hotels,Trailer Courts) 5. Single Family, Duplex, Conversion 6. Apartment 7. R e t a i l , Personal Service, Indoor Commercial, Recreation (Offices) 8. Schools 9. Churches 10. Hospitals 11. A g r i c u l t u r a l and Vacant 12. Gravel, Peat Bogs, Other Mining 13. Water 14. Vacant 15. Park 16. R a i l 17. Public Open 18. Other 19. Urban 63 e) DISTANCE MEASUREMENT The measure of distance used throughout t h i s study i s average truck t r a v e l time between zones. The t r a v e l time values are based on the estimates made by respondents i n the VUT survey, and have been smoothed to eliminate excessive 17 rounding. Values were only available for t r a v e l time between the terminal zones and 2 3 additional aggregated zones. The remaining times were estimated by in t e r p o l a t i o n . The use of time as a unit of distance i s more suitable for linkage analysis than are l i n e a r measures. The costs of goods movement are d i r e c t l y related to the costs of operating a vehicle of the appropriate size between shipment o r i g i n and destination points. A l l variable costs are more closely associated with t r a v e l time than with distance. For example, gasoline and o i l consumption, driver's wages and vehicle depreciation due to engine wear are a l l d i r e c t l y related to tra v e l time. In addition, the b a r r i e r effects of bridges and tunnels and t r a f f i c - r e l a t e d factors of congestion, road capacity and designated truck routes are a l l accounted for i n the t r a v e l time variable. Travel time i s thus more accurately described as a surrogate for shipment costs than distance. This i s th e o r e t i -c a l l y a t t r a c t i v e , as the importance of linkages to i n d u s t r i a l location patterns are ascribed to the notion of cost trade-offs between s i t e - r e l a t e d and transport-related factors. C. REGRESSION ANALYSIS A regression analysis was applied to compare the d i s t r i b u t i o n of shipments coming from/going to the two terminal zones. The approach used was to generate an equation to explain the d i s t r i b u t i o n of goods movements related to zone 6, and then to use thi s equation to predict the zone 17 flow pattern. The actual flow pattern of zone 17 was then compared with the predicted pattern. I f the variations exhibited between the actual and predicted d i s t r i b u t i o n s are s i g n i f i c a n t , i t would be possible to i n f e r that s i g n i f i c a n t differences e x i s t between the linkage relationships of the terminal zones and the urban region. The i d e n t i f i c a t i o n of the causes of such differences i s made very d i f f i c u l t by the extreme degree of m u l t i c o l l i n e a r i t y between the land-use variables. For example,, the corr e l a t i o n between the density of manufacturing land and that of single family r e s i d e n t i a l land i s -.64; for manufacturing land and auto-related land the value i s .30. Wonnacott and Wonnacott state that: "When the independent variables X and Z are c o l l i n e a r , or nearly so, i t i s c a l l e d the problem of multi-c o l l i n e a r i t y . For prediction purposes, i t does not hurt provided there i s no attempt to predict for values of X and Z removed from t h e i r l i n e of c o l l i n e a r i t y . But s t r u c t u r a l questions cannot be answered - the r e l a t i o n of Y to eithe r X or Z cannot be sensibly investigated. The observed m u l t i c o l l i n e a r i t y of land use variables can be attributed to the effects of zoning regulations, which act to segregate incompatible a c t i v i t i e s , and to i n t e r - s e c t o r a l linkages that r e s u l t i n the agglomeration of i n t e r - r e l a t e d a c t i v i t i e s . For the present purpose of pre d i c t i n g the flows coming to/going from one zone from the observed flows related 65 to another zone, multiple regression technigues are applicable. Analysis of the r e l a t i o n s h i p between flows and independent variables, however cannot be accomplished i n t h i s way. The f i r s t stage of the analysis involved generating a regression equation to model zone 6 flows. A number of variable transformations were applied i n order to f i n d the b e s t - f i t equation. Both dependent and independent variables, except t r a v e l time,, were expressed as density functions (using urban zonal area rather than t o t a l area as the denominator) to decrease the d i s t o r t i o n e f f e c t of large outlying zones. The d i s t r i b u t i o n of a l l variables was strongly lognormal, although t r a v e l time was less so than the others. A natural logarithmic transformation was applied to the variables to s a t i s f y normality assumptions of the regression model. The regression equation i n t h i s form then follows the c l a s s i c gravity model formulation. The transformation of the dependent variable (consignments/urban area) to log form necessited an adjustment to account f o r zones that had no shipments from/to zone 6. Two approaches were used: the f i r s t e ntailed the aggregation of "empty" zones with adjoining "non-empty" zones having s i m i l a r land-use c h a r a c t e r i s t i c s . 12 zones were i n -volved, so that 58 cases were available for analysis. The second approach treats the "empty" cases as missing from the sample; the rationale i s that routing bias would be responsible for low as well as high outlying values. Again, the active size i s 58 cases. The regression of the form: Ln(Consignments/Area)= Ln bo + b, Ln (Land-Use ,/Area)+... 66 + k>TLn (Travel Time) + Ln error explained over 72% of v a r i a t i o n i n both runs, but the residuals were strongly associated with predicted values. This form of equation was thus rejected, and other forms were t r i e d to f i n d an equation with high explanatory value that did not 19 v i o l a t e error independence requirements. The equation that was most suitable i n both these respects was based on a semi-logarithmic transformation u t i l i z i n g 58 aggregated zones. The equation i s i n the following form: Ln (Consignments/Area) = b Q + b^ (Land-use-^/Area)+. . . + b T (Travel time) + E. which translates to . 4. /•» b + b, (Land-Use l)...+b m (Traveltime)+E Consxgnments/Area=e o 1 T ^ • i. /-K bo , e 1 (Land-Use) „ or Consignments/Area= e + — - E A 3 T 6 3 . ' 6 e bT (Travel Time) The beta values derived by t h i s method are l i s t e d i n Figure 14. 2 The explanatory power of the equation was R =.72, which i s s a t i s f a c t o r y considering that no d i r e c t measure of land-use i n t e n s i t y i s included i n the independent variables. In order to account f o r observed c u r v i l i n e a r relationships between the dependent and some independent variables (especially manufactur-ing and auto-related land-uses) quadratic forms of these var-iables were included i n the independent variable l i s t . The predicted consignments/area values are plotted against observed values i n Figure 15. Figure 16 shows the values calculated for zone 17 from the c o - e f f i c i e n t s derived from the 2 zone 6 regression against sampled zone 17 values. The R 67 FIGURE 14 REGRESSION CO-EFFICIENTS Zone 6 Zone 17 T r a v e l - T i m e M a n u f a c t u r i n g M a n u f a c t u r i n g ' R e t a i l A p a r t m e n t s S i n g l e - F a m i l y P a r k s H o s p i t a l s H o t e l s C o n s t a n t B •-.1097 7.2934 -8.1529 12.9345 -7.6029 -2.1207 -1 .4117 -12.5499 -30.801 5 R 2 A d j u s t e d R S t a n d . E r r o r B e t a -.4568 .6541 -.3845 .2661 -.1427 -.2298 -.1084 -.0912 -.1677 -3.3970 .72-..66 1 .017 -.1143 11.9320 •17.4111 4.0569. 3.2997 -1.2696 -1 .9548 3.6184 29.7394 R-2 A d j u s t e d R S t a n d . E r r o r B e t a -.4508 1 .0443 .7757 .0846 .1039 -.1423 -.1835 .0351 .1652 -4.3824 .72 .66 1 .020 FIGURE 15 P r e d i c t e d C o n s i g n m e n t D i s t r i b u t i o n (Ln Cons. / A c r e ) FIGURE 16 P r e d i c t e d C onsignment D i s t r i b u t i o n (Ln Cons. / A c r e ) 70 value of .44 i s s i g n i f i c a n t to 4 decimal places; considering the low r e l i a b i l i t y of some shipments included i n the sample, the v a r i a t i o n between predicted and observed patterns cannot be attributed to differences i n the actual d i s t r i b u t i o n patterns of the two zones. F i n a l l y , a comparable regression analysis on zone 17 flow values was run. The corresponding beta c o - e f f i c i e n t s are shown 2 i n Figure 14. The R value of the derived equation i s also .72, and the pattern of residuals i s s i m i l a r to that of the zone 6 regression. It i s d i f f i c u l t to a t t r i b u t e a great deal of meaning to the differences i n land-use and travel-time c o - e f f i c i e n t s . The i n t e r - r e l a t i o n s h i p of the travel-time and a l l land-use variables i s an example of the problems of i n t e r -preting the regression c o - e f f i c i e n t s . Sayer notes that the distance decay term i s only unambiguous i n the event of zero autocorrelation between variables. This problem i s most acute when urban d i s t r i b u t i o n s are being analyzed: "In dealing with the highly auto-correlated s p a t i a l systems c h a r a c t e r i s t i c of urban and intra-urban situations, i t i s clear that there w i l l be considerable confounding of the e f f e c t s of distance f r i c t i o n . . . a n d map pattern, although, at the inter-urban scale, 20 the problem i s l i k e l y to be much less serious." The regression computations were accomplished by f i r s t entering the t r a v e l time variable to derive the o v e r - a l l distance-decay e f f e c t . Land-use variables were then entered in a stepwise fashion. The very close s i m i l a r i t y of the two t r a v e l time beta values suggests that linkage relationships of the two terminals react i n an i d e n t i c a l fashion to increasing transportation costs. This hypothesis w i l l be tested i n the next section, i n which zonal f r e i g h t flows are disaggregated into terminal-based and non-terminal based components for the purpose of a more detailed analysis. D. COMPONENT ANALYSIS 1. Disaggregation of Flow Distributions There are two main sources of problems i n p o s i t i v e l y i d e n t i f y i n g shipments associated with the two r a i l / t r u c k terminals. The f i r s t i s related to the size of the terminal zones and the inc l u s i o n i n these zones of non-transport a c t i v i t i e s , which has been already mentioned. I would l i k e to add here that smaller or specif i c a l l y - d e s i g n e d terminal zones would not completely solve the i d e n t i f i c a t i o n problem. The land-use structure of i n n e r - c i t y areas i s not highly segregated. Older terminal operations t y p i c a l l y are not separated from other land uses, unlike new terminals located i n outlying 21 s i t e s . The intermixed nature of the older i n d u s t r i a l "zone in t r a n s i t i o n " has been attributed to the lack of zoning regu-latio n s during the period of o r i g i n a l construction; subsequent suburbanization of i n d u s t r i a l a c t i v i t i e s and concommitant take-over of buildings, unsuited to modern horizontal production techniques, by marginal firms and a c t i v i t i e s ; relocation of firms displaced by construction of new o f f i c e buildings and 22 apartments i n the commercial core area, etc. In the case of the CP terminal area, the adjoining warehouse sector has been p a r t l y abandoned by wholesaling firms i n favour of locations with better access to the growing urban region (such as Burnaby and Richmond), and these buildings are now used by 72 firms whose connection with the terminal are at best unclear. A zonal configuration which included these firms i n the "terminal" zone would r i s k i n c l u s i o n of non-terminal flows i n the shipment sample, while a configuration which excluded the warehousing area would r i s k m i s - i d e n t i f i c a t i o n of shipping patterns related to the operations of the trucking industry as true linkages. This example leads to the second source of i d e n t i f i c a t i o n d i f f i c u l t y : the d e f i n i t i o n , or more accurately the interpreta-tion, of terminal-associated flows. This problem has been discussed i n terms of inter-zonal shipments; here intra-zonal aspects w i l l be investigated. Figure 17 i l l u s t r a t e s the i d e n t i f i c a t i o n scheme for terminal zone shipments used i n t h i s study. The diagram i s a representation of zone 6,, but the characterization of shipment types i s equally applicable to zone 17. Both inter-zonal and intra-zonal shipments are represented. Inter-zonal shipments are by f a r the majority of those sampled, but as intra-zonal shipments represent about 10% of the zone 6 t o t a l (although only 5% for zone 17) they merit attention as well. A l l shipments that had at least one end-point i n a terminal zone were c l a s s i f i e d into three types. 1. R a i l Shipments: going to/coming from a t r u c k / r a i l terminal complex. 2. Terminal Oriented Shipments: fr e i g h t movement related to terminal location. 3. Other Shipments: not related to terminal operation. The i n c l u s i o n of the second category serves both a t h e o r e t i c a l and a p r a c t i c a l purpose. I have previously defined a shipment 73 FIGURE 17 SCHEMATIC REPRESENTATION OF SHIPMENTS RELATED TO A RAIL/TRUCK TERMINAL ZONE M i x e d C o m m e r c i a l and I n d u s t r i a l o -W h o l e s a l e Warehouses • o Y o Y o o RAIL TERMINAL S h i p m e n t s - R a i l : T e r m i n a l O r i e n t e d : O t h e r : o E n d p o i n t s - T e r m i n a l :Q R e l a t e d : Q O ther : < J between a w h o l e s a l e warehouse and a r e t a i l o u t l e t ( f o r example) as a consignment, because i t r e p r e s e n t s a complete c o n t r a c t e d f r e i g h t movement. However, i f t h i s consignment had been s h i p p e d t o t h e w h o l e s a l e r v i a r a i l , t h e w h o l e s a l e r - r e t a i l e r d e l i v e r y -l i n k i s ' a s s o c i a t e d t o some e x t e n t w i t h t h e l o c a t i o n o f t h e r a i l t e r m i n a l . The s t r e n g t h o f t h e t e r m i n a l - r e t a i l e r a s s o c -i a t i o n , i n essence t h e s p a t i a l a t t r a c t i o n o f t h e t e r m i n a l on the r e t a i l e r ' s l o c a t i o n d e c i s i o n , i s l i k e l y v e r y minor i n r e l a t i o n t o m a r k e t - a r e a f a c t o r s . On t h e o t h e r hand, t h e c o s t s o f s p a t i a l s e p a r a t i o n a r e r e a l and a r e borne i n some way by everyone i n t h e c h a i n . The w h o l e s a l e r i s m i n i m i z i n g backward l i n k a g e c o s t s by l o c a t i n g near t h e r a i l t e r m i n a l ; t h e e f f e c t i v e a r e a s e r v e d by t h e t e r m i n a l t h e n i s a s s o c i a t e d w i t h t h e whole-s a l e r ' s f o r w a r d l i n k a g e s . I f t h e w h o l e s a l e r l o c a t e d w i t h r e g a r d t o m i n i m i z i n g f o r w a r d l i n k a g e s , t h e t e r m i n a l - w h o l e s a l e r l i n k would i n i t s e l f f u l l y r e f l e c t l o c a t i o n a l i n t e r d e p e n d e n c e (and would be c l a s s e d as a t y p e 1 s h i p m e n t ) . Note t h a t i f t h e r e t a i l e r i n t h i s example i s r e p l a c e d w i t h a secondary manufact-u r e r , t h e a l l o c a t i o n o f i n t e r a c t i o n c o s t s would be more d i r e c t and t h e i n t e r d e p e n d e n c e o f l o c a t i o n s c l e a r . The p r a c t i c a l purpose o f i n c l u d i n g an i n t e r m e d i a t e c a t e g o r y i s t h a t t h e t y p e o f a c t i v i t y c a r r i e d on a t shipment o r i g i n s / d e s t i n a t i o n s was o f t e n d i f f i c u l t t o i n t e r p r e t from t h e completed q u e s t i o n n a i r e s . F o r example, t r u c k s appeared t o be c o n s t a n t l y p i c k i n g up from o r d e l i v e r i n g t o "warehouses"; i t was r a r e l y c l e a r i f t h i s was a t t h e t e r m i n a l , a w h o l e s a l e r , a t r u c k i n g o p e r a t i o n , o r perhaps a g e n e r a l term f o r the b i g room b e h i n d a s t o r e . I t was r a r e l y p o s s i b l e t o i n f e r t r u c k 75 stop location a c t i v i t y from the description of the type of good being delivered, as t h i s required a p r i o r i knowledge of fr e i g h t moving through the terminals. However, i f the stop i n question occurred at the home base of the truck (these were often l i s t e d as "company warehouse") the a c t i v i t y could be determined through i d e n t i f i c a t i o n of the owner or lessee of the truck. In Figure 17 the three types of consignment end-points are denoted by the shapes. In p a r t i c u l a r , the trucking warehouse sector i s treated as part of the terminal complex; the designation of other sectors i s self-explanatory. Inter-zonal f r e i g h t shipments (no. 7 to 14 i n the i l l u s t r a t i o n ) were c l a s s i f i e d according to the location type of the terminal zone end-point. Intra-zonal shipments (1 to 7, 15) were c l a s s i f i e d according to the highest l e v e l of either end-point. For example, shipments represented by type 6 (movements between the terminal and the wholesale sector) are i d e n t i f i e d as r a i l shipments because one end-point i s located at the terminal. Note that only one" i n t r a - s e c t o r a l flow i s shown i n the diagram, and occurs within the mixed commercial and i n d u s t r i a l area (15). The trucking warehouse and r a i l terminal sectors are u n l i k e l y to generate i n t e r n a l flows, while that occurring within the wholesale area would probably be minor. In any case, such flows would be almost impossible to i d e n t i f y accurately. Only 7 of the 60 shipments t r a v e l l i n g within zone 6 are i n t r a - s e c t o r a l . The decision making procedure used i n a l l o c a t i n g consign-ments to categories i s i l l u s t r a t e d i n Figure 18. The propor-t i o n of consignments of each type i s approximated by the area FIGURE 18 TRIPARTITE DIVISION OF TERMINAL ZONE SHIPMENTS TERMINAL ORIENTED OTHER 77 of the c i r c l e s ; t h i s diagram i s constructed from zone 17 data. In zone 6, the proportions of r a i l and terminal consignments i s the reverse of those i l l u s t r a t e d here. The areas of overlap represent uncertainty as to the appropriate category to which a consignment belongs. This uncertainty i s due both to incomplete information and to the d i f f i c u l t y of discerning between s p a t i a l contiguity and l o c a t i o n a l interdependence. The following rules were complied with i n assigning consignments to categories: 1. R a i l - Only shipments with one endpoint location p o s i t i v e l y i d e n t i f i e d as the terminal complex are included (type 1). 2. Terminal Oriented - Included are types 2, 4 and 7. Type 7 are t y p i c a l l y poorly documented shipments (usually of the vague "warehouse" variety) for which there i s some evidence, through commodity type or vehicle ownership, that the assumption of terminal association i s j u s t i f i e d . 3. Other - Included are types 3, 5 and 6. The a l l o c a t i o n of types 5 and 6 to t h i s category s u b s t a n t i a l l y decreased the s i z e of the other categories, but allowed for a higher degree of confidence i n the r e l i a b i l i t y of terminal linkage analysis. In order to i l l u s t r a t e the problems of c l a s s i f i c a t i o n , I w i l l present a few examples. A paint manufacturing firm ships i t s products from i t s own on-site warehouse, located i n zone 17, to a number of other-zone customers. In terms of the c l a s s i f i c a t i o n procedure, these are c l e a r l y type 3 ship-ments. However, i f the firm i s located i n the v i c i n i t y of the terminal i n order to receive inputs and to ship products to other c i t i e s by r a i l , the firm's intra-urban forward linkages picked up i n the VUT survey are greatly influenced by the location of the CN terminal complex. It i s impossible to 78 determine the strength of l o c a t i o n a l dependence through a limi t e d goods movement survey, so t h i s type of i n d i r e c t linkage i s relegated to the "other" category. Another instance of possible l o c a t i o n a l interdependence r e s u l t i n g i n i n d i r e c t linkages i s the operations of an auto-mobile dealer i n zone 17. The majority of shipments going to/ coming from t h i s firm are inter-zone auto parts d e l i v e r i e s . It i s clear that the dealer doesn't receive the majority of his parts supplies through the CN terminal. It i s possible, though, that his location i s clos e l y linked to the terminal i f he i s located with regard to market access, and his c l i e n t e l e consists of employees of the terminal complex and associated firms. This reasoning i s not too far-fetched, because there are many other automobile repair and body shops i n the terminal areas. This i s due, i n part, to zoning regulations and, i n part, to the ease with which the many employees of firms linked to the terminals can leave t h e i r cars during the day for repairs at close-by shops. Shipments involving the auto-mobile dealer and s i m i l a r firms are c l a s s i f i e d as type 3, because the linkage e f f e c t remains uncertain. 2. COMMODITY ANALYSIS The rationale behind s p l i t t i n g the terminal zone shipments into components i s to enable the comparison of the operations of the terminals. The f i r s t element of the analysis i s the make-up of the flows associated with the terminals. If the d i s t r i b u t i o n of commodity types and weights are si m i l a r for both terminals, i t w i l l be possible to conclude that there i s 79 no clear tendency to s p e c i a l i z e i n certain types of shipments, and that differences between the d i s t r i b u t i o n s of terminal shipments can be attributed to external factors. It should be noted, however, that differences i n the commodities handled by the two zones only are i n d i c a t i v e , not proof, of i n t e r n a l s p e c i a l i z a t i o n . The d i s t r i b u t i o n of land-uses r e l a t i v e to the two terminals conceivably could r e s u l t i n one terminal handling the bulk of one commodity, and vice versa. The d i s t r i b u t i o n of commodity types within zone components and between zones was analyzed with a chi square t e s t . As both variables (commodity class and shipment component) are measured 23 on a nominal scale, t h i s method i s appropriate. However, the wide range i n marginal values r e s u l t i n g from the general nature of the commodity c l a s s i f i c a t i o n s tends to increase chi square values. A s i m i l a r e f f e c t i s caused by the s e n s i t i v i t y of t h i s test to sample s i z e . Shipment samples are a b i t large for the size of contingency tables used, because only 5 or 6 degrees 24 of freedom are attained. These two factors make the chi square more sensi t i v e to minor differences i n the sample d i s t r i b u t i o n s than i s j u s t i f i e d by the i n t e r n a l r e l i a b i l i t y of either variable. Furthermore, the v a r i a t i o n i n sample sizes of flow components between and within zones makes comparisons of chi square values "or sig n i f i c a n c e l e v e l s unreliable. Comparisons can be achieved by using weighted indexes based on the chi square, such as Cramer's V or the Contingency 25 C o - e f f i c i e n t . These indexes do not allow for a s i g n i f i c a n c e l e v e l to be attributed to the measured strength of association. A l l flow components were tested, both within and between 80 zones, to determine variations i n commodity types being shipped. The only d i s t r i b u t i o n s tested against each other to obtain a chi square value with a s i g n i f i c a n c e greater than .05 (indicat-ing that the amount of d i s s i m i l a r i t y between the samples would be due to chance more than 1- in- 20 times) were the two R a i l -Terminal components (see Figure 19). The major difference between the two zones occurred i n the other category: zone 17 showed a much higher percentage of "other" goods shipped (43% to 15%). This i s lar g e l y due to the presence of a number of produce warehouses i n zone 17 which may have been over-weighted in the sample. The contingency c o - e f f i c i e n t of the two t o t a l d i s t r i b u t i o n s i s very low (although the chi square value i s s t i l l s i g n i f i c a n t ) , and the only large difference between the two again occurs i n the "other" category. Thus the only s i g n i f i c a n t difference i s commodity types going to/ coming from these zones can be i d e n t i f i e d as the greater number of produce consignments associated with zone 17. The within-zone tests reveal that v a r i a t i o n between R a i l and Other shipments i n zone 6 i s less than that between R a i l and Terminal Oriented shipments. This i s due to the larger number of fabricated goods and produce i n the Terminal Oriented sample. Generally, i t would be expected that the R a i l and Terminal Oriented components would exhibit more conformity with each other than either would with other shipments, as i s the case i n zone 17. The results of the zone 6 intra-zone test series are, consequently, surprising, although the high l e v e l of aggregation used to c l a s s i f y commodity types i s not conducive to r e l i a b l e inference. For example, 78% of both FIGURE 19 COMMODITY TYPE DISTRIBUTION TESTS T e s t D i s t r i b u t i o n s CN - R a i l , Term. Or., O t h e r CP - R a i l , Term. Or., Other CN - R a i l , Term. O r i e n t e d CN - Term. O r i e n t e d , O t h e r CN - R a i l , O t h e r CP - R a i l , Term. O r i e n t e d CP - Term. O r i e n t e d , O t h e r CP - R a i l , O t h e r CN - R a i l - T e r m * O t h e r CP - R a i l - T e r m , O t h e r CN - R a i l - T e r m , CP - R a i l - T e r m CN - O t h e r , CP - Other CN - T o t a l , CP - T o t a l C h i Square U 6 . 6 90.1 27.2 1 08. 9 48.7 73.2 17.5 45.3 116.8 16.1 12.2 59.9 35.0 S i g . .0000 .0000 . 0001 .0000 . 0000 , 0000 .0036 . 0000 , 0000 , 01 31 .0570 , 0000 , 0000 Cramer's V .33971 .28086 .28619 .44571 .35321 .47105 .21753 .32007 .42891 .16801 .13594 .33195 .17026 C o n t i n g e n c y C o - e f f i c i e n t .43304 .36914 .27514 .40710 .33305 .42614 .21256 .30483 .39418 .16568 .13470 .31504 .16785 oo •^Rail-Term = R a i l + T e r m i n a l O r i e n t e d S h i p m e n t s . 8 2 r a i l components are general f r e i g h t shipments. A possible explan-ation i s that the c l a s s i f i c a t i o n of 45% of zone 6 produce ship-ments as a part of the Terminal Oriented component may be inaccurate. The concordance of the two amalgamated R a i l -Terminal components and the f a i r l y close s i m i l a r i t y between the t o t a l commodity type d i s t r i b u t i o n s i s s u f f i c i e n t evidence to j u s t i f y the assumption that the two terminals handle the same type of goods. 3 . ANALYSIS OF COMMODITY SIZE DISTRIBUTIONS The next step i n determining whether the two terminals handle the same type of f r e i g h t i s to examine the size of consignments going through the terminals. I t i s possible that the mix of commodity types handled by the terminals i s similar, but that the size of i n d i v i d u a l consignments going to/coming from one terminal i s consistently smaller than those at the other terminal. For example, i f the CP terminal's service area i s focused towards the smaller firms i n the older core i n d u s t r i a l areas, i t i s possible that consignment sizes would be generally smaller than those going to larger outlying firms v i a the CN terminal. I f there i s no v a r i a t i o n i n consignment size between the two terminals, two conclusions can be reached: 1. there i s no s p e c i a l i z a t i o n i n f r e i g h t services between the two terminals; the terminals therefore compete for the same market. 2 . there i s no systematic v a r i a t i o n i n aggregate demand for transport services between the terminals; production c h a r a c t e r i s t i c s i n terms of size and type of shipped inputs and outputs are uniformly d i s t r i b u t e d among a l l firms, regardless of terminal a f f i l i a t i o n . Analysis of the d i s t r i b u t i o n of consignment size among 8 3 the flow components within zones w i l l also shed l i g h t on d i f f e r -ences i n the c h a r a c t e r i s t i c s of inter-urban f r e i g h t (shipped through a terminal) and intra-urban f r e i g h t . This can be accomplished through comparison of the R a i l and Other compon-ents? i t i s l i k e l y that Terminal Oriented shipments have been subject to bulking or de-bulking procedures and would be more representative of intra-urban consignment size s . At t h i s point a digression on s t a t i s t i c a l techniques i s necessary. Up to now data has been analyzed using i n t e r v a l -l e v e l methods (regression) and nominal-level methods (chi-square and related i n d i c e s ) . In t h i s and subsequent sections data i s organized at an ordinal scale and a number of non-parametric tests are employed to support i n f e r e n t i a l arguments. I w i l l outline the reasons for t h i s approach and the general r e l i a b i l i t y of the s p e c i f i c techniques i n terms of the a v a i l -able data. The variables used for analysis were coll e c t e d at an i n t e r v a l scale. For example, the consignment weight was d i r e c t l y recorded on the questionnaires, and the time required to drive from one stop to another was estimated by truck drivers for each t r i p . However, there are three problems inherent i n the data which make acceptance of the i n t e r v a l scale and application of powerful parametric methods misleading. The f i r s t problem results from the method of data c o l l e c t i o n , and the suspicion that the respondents' use of s p e c i f i c numbers to answer questions creates an i l l u s i o n of accuracy. The weights assigned to consignments were often estimated, as t h i s informa-ti o n i s not always written on b i l l s of lading. Truck drivers 84 do not a l l have the same a b i l i t y to estimate weight, and the assumption of random error seems u n r e a l i s t i c . The average percentage error of estimating a 10,000 l b . shipment may well be s i m i l a r to that of estimating a 100 l b . shipment, but the numerical error would be much greater. Thus a few large estimation errors could have a great e f f e c t on subsequent analysis. Another aspect of respondent error involves the tendency of estimates to centre on standard break points; for example, shipment weights are often given to the nearest 100 or 1,000 lbs. A s i m i l a r tendency f o r t r a v e l time estimates was recognized i n the VUT study, and corrected through the application of smoothing techniques. I t seems reasonable, however, to accept the respondents' estimates as belonging to natural ordinal scales rather than to force an i n t e r v a l scaling on them. The second problem i n the data i s concerned with the non-random sampling procedure discussed above. In regard to the consignment weight variable, shipment size i s c l e a r l y related to truck s i z e . On the one hand, the sample must include the same proportion of trucks i n each size category as there are i n the t o t a l population. Adjustments can be made by using a series of expansion factors i f actual f l e e t mix r a t i o s are known. This can be done for the region as a whole but not for s p e c i f i c zones, as in t r a - r e g i o n a l v a r i a t i o n i n f l e e t mix i s very d i f f i c u l t to estimate. On the other hand, routing bias w i l l a f f e c t a l l consignment-related variables. The use of an ordinal scale does not eradicate these problems, but i t does reduce the influence of questionable outlying values. 85 The t h i r d d i f f i c u l t y i s related not to the inadequacies of the sample, but to the nature of the underlying population d i s t r i b u t i o n s of the variables. Parametric techniques require that the variables be normally d i s t r i b u t e d . Transformation of the data to meet normality assumptions i s not always possible, and i n any case, i s a source of confusion i n the interpretation n a t u r a l of test r e s u l t s . In the present case," log transformation of the consignment weight variable does not normalize the d i s -t r i b u t i o n (for example, see the cumulative frequency d i s t r i b u -tions i n Figures 22 to 2 4); a more powerful and less i n t e r p r e t -able logarithmic transformation i s necessary. The problems confronted when using parametric techniques on t r a v e l time and land-use variables have already been discussed. Non-parametric techniques are both more appropriate and more e a s i l y interpretable for analysis of these variables. The tests that I w i l l most commonly employ are Kendall's Tau C non-parametric correlation c o - e f f i c i e n t and the Kolmogorov-Smirnov paired sample test for d i s t r i b u t i o n s i m i l a r i t y . Kendall's Tau C tests the association of two o r d i n a l -2 S l e v e l variables. The test s t a t i s t i c (Tau C) ranges i n value from +1 to -1; a value of +1 i s obtained i f the variables are d i r e c t l y (positively) correlated and a value of -1 represents inverse (negative) c o r r e l a t i o n . A value of 0 implies that no systematic rel a t i o n s h i p exists between the two variables. Consider two variables, A and B (see figure 20). For purposes of i l l u s t r a t i o n , l e t A=consignment weight and B represent a dummy d i s t r i b u t i o n variable, such that B^= R a i l and B2= Other. B i s a dichotomous variable, but i t can also be interpreted as FIGURE 20 INTERPRETATION OF KENDALL'S TAU C B B B 1 2 1 2 1 2 2 4 6 A 87 27 an ordinal variable containing two categories. A Kendall's Tau C t e s t i s performed on the data to discover whether there i s a difference i n consignment weight d i s t r i b u t i o n between the R a i l and Other flow components, and i f so, what the d i r e c t i o n of difference i s . In Figure 20a the Tau C i s positive, imply-ing that there i s a difference between the two components, and s p e c i f i c a l l y that other shipments tend to be heavier than R a i l shipments. The reverse s i t u a t i o n i s shown i n Figure 20b. The s i g n i f i c a n c e l e v e l of the Tau C c o r r e l a t i o n c o - e f f i c i e n t can be calculated; the d i s t r i b u t i o n of the t e s t s t a t i s t i c i s based on t o t a l sample s i z e . I f there i s a large v a r i a t i o n i n marginal values, the i n t e r p r e t a t i o n of the Tau C becomes vague. The s i g n i f i c a n c e l e v e l i s based on the t o t a l sample siz e regardless of i t s d i s t r i b u t i o n i n the contingency table; thus a test be-tween two equally-sized flow component samples would have the same sign i f i c a n c e as a test i n which one component had a sample size 10 times the other. Imbalance of the A variable marginals i s related to the establishment of categories f o r t h i s variable. Generally, the more evenly d i s t r i b u t e d the values are among categories (for each B sample d i s t r i b u t i o n ) , the more r e l i a b l e i s the Tau C value. Fig . 20c i l l u s t r a t e s one i n t r a c t a b l e problem i n using t h i s technique for present purposes. In the example, R a i l shipments are either small or heavy, while other shipments are a l l of medium weight. The corresponding Tau C value indicates that there i s no r e l a t i o n s h i p between variables A and B. The usual conclusion would be that the two flow components have s i m i l a r consignment weight d i s t r i b u t i o n s , but t h i s i s c l e a r l y a Type II 88 error. In other words, the Tau C shows no systematic difference between the two d i s t r i b u t i o n s , but i s unable to discern i n t e r n -a l differences that don't extend throughout the range of A values. In these cases, an alternative technique i s the Kolmogorov-2 8 Smirnov two-sample t e s t . I t i s generally more conservative than Kendall's Tau C, as i t " i s sensitive to any kind of difference i n the di s t r i b u t i o n s from which the two samples were drawn-differences i n location (central tendency), i n dispersion, 29 i n skewness, etc." The K-S test s t a t i s t i c D i s defined as the maximum difference between the cumulative frequency d i s -t r i b u t i o n s of the two samples. D can then be compared against a test D value calculated for the sizes of the samples being examined, at a desired alpha l e v e l , i n order to esta b l i s h the sign i f i c a n c e of v a r i a t i o n i n the d i s t r i b u t i o n s . Figure 20d applies t h i s test to the same data used i n Figure 20c; the di s t r i b u t i o n s are found to d i f f e r s i g n i f i c a n t l y . The K-S test does not give the d i r e c t i o n of difference or a measure of association as does the Kendall's Tau C. I t i s primarily useful as a conservative check on samples for which the Tau C signi f i c a n c e i s i n question. The r e s u l t s of tests made on the between-sample v a r i a t i o n i n commodity size d i s t r i b u t i o n are l i s t e d i n Figure 21. Note that the only disagreement between the two techniques occurs for the zone 6 R a i l and Terminal Oriented d i s t r i b u t i o n t e s t s . Inspection of the data reveals that the R a i l shipments are generally on the l i g h t and heavy ends of the scale, whereas the Terminal-Oriented shipments tend toward the centre. This FIGURE 21 COMMODITY WEIGHT DISTRIBUTION TESTS T e s t D i s t r i b u t i o n s CN - R a i l , Term. Or., O t h e r • CP - R a i l , Term. Or., O t h e r CN - R a i l , T e r m i n a l O r i e n t e d CN - T e r m i n a l O r i e n t e d , O t h e r CN - R a i l , O t h e r CP - R a i l , T e r m i n a l O r i e n t e d CP - T e r m i n a l O r i e n t e d , O t h e r CP - R a i l , Other CN - R a i l - T e r m ? O t h e r CP - R a i l - T e r m , O t h e r CN - R a i l - T e r m , CP - R a i l - T e r m CN - O t h e r , CP - Other CN - T o t a l , CP - T o t a l K e n d a l l ' s Tau C -.1 5188 -.02766 -.02137 - . 1 8203 -.16197 +.07137 -.10921 -.02450 -.19666 -.06064 +.05127 +.19376 +.12331 S i g . .0000 .4764 .7088 .0002 .0007 .2511 .0548 .6587 . 0000 .2027 .2520 .0001 .0002 Kolmogorov - S m i r n o v i s i g . v a r i a t i o n n o t a p p l i c a b l e n o t a p p l i c a b l e s i g n i f i c a n t s i g n i f i c a n t s i g n i f i c a n t s i g n i f i c a n t s i g n i f i c a n t s i g n i f i c a n t co •^Rail-Term = R a i l + T e r m i n a l O r i e n t e d S h i p m e n t s . 90 s i t u a t i o n i s very s i m i l a r to the example i n Figure 20c, and further i l l u s t r a t e s the care required to interpret Tau C r e s u l t s . The data shows that R a i l and Terminal Oriented shipments are both s i g n i f i c a n t l y larger than Other shipments,in the CN zone. The median shipment sizes are: R a i l : 220 lbs. ) ) Rail-Term: 245 lbs. T.O.: 270 lbs ) Other : 100 lbs. The CP zone flow components show much less v a r i a t i o n . The R a i l and Terminal Oriented shipments are also larger than Other shipments, but neither difference i s s i g n i f i c a n t . The median shipment sizes for t h i s zone are: R a i l : 160 lbs. ) ) 210 lbs. T.O. : 300 lbs. ) Other : 245 l b s . The low median value of the R a i l component suggests that the CP terminal d i r e c t l y serves more small firms than does the CN terminal. Figure 22 shows that the CP R a i l shipment pattern i s bimodal: small ( 100 lbs.) and large ( 5,000 lbs.) ship-ments predominate. This d i s t r i b u t i o n i s s i g n i f i c a n t l y d i f f e r e n t from the CN R a i l d i s t r i b u t i o n , in which 100-250 l b . shipments are most frequent. However, when the two Rail-Terminal d i s t r i b u t i o n s are compared (Figure 23), these variations are dampened by the addition of Terminal-Oriented shipments to the extent that s t a t i s t i c a l s i g n i f i c a n c e can not be assigned to them. It i s possible to account for t h i s s i t u a t i o n by view-ing the CP R a i l component as an i n d i c a t i o n of a basic difference in the operations of the two terminals. In t h i s scenario the FIGURE 22 COMMODITY WEIGHT - RAIL COMPONENTS Ln Commodity Weight (1,000 l b s . ) FIGURE 23 COMMODITY WEIGHT - RAIL-TERMINAL COMPONENTS C u m u l a t i v e F r e q u e n c y Commodity Weight (1,000 l b s . ) d i r e c t services of the CP terminal are used by a combination of smaller firms i n older i n d u s t r i a l areas such as False Creek, and by large firms accustomed to dealing with either CP or BCHR because they are located near trackage owned by these railways. The small v a r i a t i o n of Terminal Oriented components i s due to s i m i l a r i t y i n the wholesale operations located adjacent to the terminals. Yet, t h i s hypothesis i s on shaky ground, because the small sample size of the CN R a i l component may contain enough anomolies to make i t t o t a l l y unrepresenta-t i v e . In addition, the in c l u s i o n of bus parcel shipments and large mail d e l i v e r i e s i n the CP sample, although small i n number (both represent less than 5% of the t o t a l ) , may d i s t o r t the consignment size d i s t r i b u t i o n . The difference between the two Other d i s t r i b u t i o n s i s more e a s i l y explained (Figure 24). 14% of the CP Other shipments are fabricated materials, against 5% for the CN zone. Consignments of these commodities are t y p i c a l l y much larger than other consignment types. The CP zone contains more i n d u s t r i a l firms than the CN zone; the operations of a large paper wholesaler i n zone 6 also adds to the number of heavy consignments. Disaggregation of commodities by type reveals that the s i m i l a r i t y of t o t a l Rail-Term weight d i s t r i b u t i o n s does not hide i n t e r n a l differences. Both general f r e i g h t (Figure 25) and fabricated material (Figure 26) shipment size d i s t r i b u -tions are within the bounds of s i g n i f i c a n t v a r i a t i o n . The larger number of small fabricated material shipments related to the CP terminal supports the hypothesis suggested above, but FIGURE 2k Commodity Weight (1,000 l b s . ) FIGURE 25 FIGURE 26 COMMODITY WEIGHT - RAIL-TERMINAL FABRICATED MATERIAL CONSIGNMENTS Commodity Weight (1,000 l b s . ) 9 7 further investigation i s hampered by the small sample size s . The CN R a i l component includes only two fabricated materials consignments, against 14 for the CP R a i l component. In summary, Rail-Term shipment size d i s t r i b u t i o n s are markedly s i m i l a r while Other d i s t r i b u t i o n s vary s i g n i f i c a n t l y between zones. R a i l d i s t r i b u t i o n s show i n t e r e s t i n g differences, but the small sample size of the CN R a i l component makes firm conclusions regarding terminal operating c h a r a c t e r i s t i c s impossible. 4.. SPATIAL ANALYSIS i) INTRODUCTION In t h i s section I w i l l analyze the s p a t i a l pattern of the flow components of the terminal zones i n a generalized way. Two factors w i l l be examined for t h e i r e f f ects on the various f r e i g h t d i s t r i b u t i o n s : i n d u s t r i a l a c t i v i t y i n the urban region and distance decay (or s p a t i a l f r i c t i o n ) . Description of s p e c i f i c linkages w i l l be kept to a minimum as they tend to be overly t a u t o l o g i c a l . In other words, defining i n t e n s i t y of linkage between a terminal and a given zone as the proportion of t o t a l consignments shipped between the two locations does not a i d i n understanding the reasons underlying the linkage, only i n i d e n t i f y i n g i t . Emphasis w i l l be placed on nomothetic considerations. The rationale of the analysis i s based on a gravity or potent i a l approach. The gravity concept used here i s analytic rather than deterministic; I am investigating the influence of "a t t r a c t i v e " and "repulsive" factors on goods movement rather 98 than attempting to ca l i b r a t e a predictive model based on these factors. The s i m p l i f i e d gravity model used here i s based on Hoare's d i s t i n c t i o n between potential and actual linkages. Actual linkages are defined as goods flows, while p o t e n t i a l linkages are "the contacts with which (a firm) could have been fe a s i b l y 30 linked"'. The r e a l e f f e c t of distance on a firm's linkage pattern i s then given by I (linkage index), defined as: _ A , _ . di = ( dA /dP ) I = p' s u c h t h a t do { d3/d5 ) where A= actual contacts/unit area P= pote n t i a l contacts/unit area D= a distance measure If A and P are related to distance i n the following way: A= K D a , and P= K D b P where and Kp are some constants, then Kp D b Now, gravity models of the form • T.. = Mi M. D.. i j J i j can be rewritten as Tij= k Mj Dij when in t e r a c t i o n between only zone land other zones 1 to j are being considered. By d e f i n i t i o n , ' Tij= A i j , and Mj. = Pj , so that A= Pj D c , or 99 c Thus I=D , which i n f e r s that linkage i n t e n s i t y with respect to distance i s i d e n t i c a l to the calibrated "distance decay" variable. Substituting for I, DC= D ( a" b ) . This i s a s i g n i f i c a n t r e s u l t i n that i t i l l u s t r a t e s the fundamental dual function of the distance variable generated by c a l i b r a t i n g gravity models. Actual t r i p d i s t r i b u t i o n can be explained by gravity model c a l i b r a t i o n only i f the p o t e n t i a l for i n t e r a c t i o n i s constant ( i . e . , i f b=0) or i f the measure used for p o t e n t i a l i s adjusted i n such a way as to ensure that b=0. For example, the o r i g i n a l mechanistic interpretation of the gravity model suggested that a distance exponent of -2 i s appropriate. In the present context, t h i s implies that i n t e r a c t i o n p o t e n t i a l i s s p a t i a l l y constant and that actual contacts are inversely related to the area of a c i r c l e with radius D centred on the point of t r i p generation. I t can be argued that t h i s i nterpretation i s t h e o r e t i c a l l y sound. I f the p o t e n t i a l for i n t e r a c t i o n i s d i r e c t l y related to a general, i n c l u s i v e measure of a c t i v i t y , i t i s not unreason-able to expect an urban area to exhibit a constant l e v e l of i n t e r a c t i o n p o t e n t i a l r e l a t i v e to some function of distance from a central point. In Figure 27, a cross-section of an i d e a l i z e d urban area i s represented. The distance scale i s i n transport cost units; costs are thus used to create a map transformation of actual l i n e a r distance. Following the l o g i c 31 of bid-rent curve arguments , i t can be demonstrated that a declining density gradient w i l l e x i s t that i s l i n e a r l y related to the transport cost function. The density gradient i s shown FIGURE 27 25 20 15 10 5 5 10 15 20 25 D i s t a n c e i n M i n u t e s i i 1 1 1 1 1 1 1 1 1 10 5 3 2 1 . 5 . 5 1 2 3 5 10 D i s t a n c e i n M i l e s 101 i n the diagram, centred at O. If the pot e n t i a l measure includes a l l types of economic a c t i v i t y , and zoning regulations and other sources of v a r i a t i o n are disregarded, the t o t a l p o t e n t i a l in a c i r c l e of radius R cost units w i l l be equal to that i n a l l succeeding concentric c i r c l e s with the same radius. This argument f a l l s apart i f the location i n question i s not situated at the urban centre. For a l l non-central points in t e r a c t i o n potential w i l l vary with distance (as measured by transport costs), and the rel a t i o n s h i p i s unlikely to be either monotonically declining or increasing. The analysis that follows begins with the i d e n t i f i c a t i o n of a p o t e n t i a l measure. Variations between the various flow components, r e l a t i v e to the s p a t i a l d i s t r i b u t i o n of linkage att r a c t i o n , are examined. The e f f e c t of distance on consign-ment d i s t r i b u t i o n s i s investigated, and the 'a' exponents are estimated. The rel a t i o n s h i p of p o t e n t i a l and the location of the two terminals i s then analyzed, i n order to determine values for the 'b' exponents. F i n a l l y , the linkage i n t e n s i t y indices for the flow components are compared. i i ) ANALYSIS OF ATTRACTION The emphasis i n t h i s section i s on the r e l a t i o n s h i p be-tween the flow components of the two terminal zones and the s p a t i a l d i s t r i b u t i o n of i n d u s t r i a l a c t i v i t y . Other types of linkage potential, such as r e t a i l firms, households, etc., w i l l be excluded from consideration. The rationale behind defining a t t r a c t i o n to include only i n d u s t r i a l a c t i v i t i e s i s that the focus of the study i s s p e c i f i c a l l y on the i n d u s t r i a l FIGURE 28 DENSITY OF INDUSTRIAL LAND-USE P e r c e n t I n d u s t r i a l Land 0-1 • 5.1-10 m z-z-z-ZKi^-z-z-z-z-q 20.1-30 30.1-40 40.1-50 50.1-60 >60 103 sector; i n c l u s i o n of other possible linkage types at the s p a t i a l scale of the available data only confuses the issue. The measure of a t t r a c t i o n used i s c r i t i c a l to subsequent analysis. There are generally two ways to define a t t r a c t i o n . The f i r s t i s based on c u r v e - f i t t i n g procedures, such that the variable or combination of variables which best f i t the actual flow d i s t r i b u t i o n i s by d e f i n i t i o n the a t t r a c t i o n function. The second requires an a p r i o r i d e f i n i t i o n that can be then tested against actual linkage patterns to discover the strength of the rela t i o n s h i p between the two. The f i r s t method i s aimed at prediction, and w i l l always provide a closer approximation to the data; however, the derived a t t r a c t i o n function may be d i f f i c u l t to interpret. The second method i s better for a n a l y t i c a l purposes, but the necessarily large error involved may lessen the r e l i a b i l i t y of the analysis. Three i n t e r - r e l a t e d a t t r a c t i o n measures are used i n t h i s analysis: land area used for manufacturing a c t i v i t y ; r e l a t i v e concentration of manufacturing land; and a variable derived by combining the two. The GVRD land-use c l a s s i f i c a t i o n i s not s u f f i c i e n t l y disaggregated to test s p e c i f i c i n d u s t r i a l types; instead, a general c l a s s i f i c a t i o n was constructed to investigate the r e l a t i o n s h i p of a l l i n d u s t r i a l land-uses with the two terminals. Two categories - manufacturing land-use and auto-related land-use - were amalgamated to form the generalized i n d u s t r i a l a t t r a c t i o n variable. The land-uses included are: Man.: Manufacturing Warehousing Major Repairs 10k Auto: Automotive Wholesale Out-door R e t a i l and Commercial Recreational Nurseries The amalgamation of the two classes was deemed to be necessary because land used for wholesale a c t i v i t i e s i s not i n the same c l a s s i f i c a t i o n as land used for warehousing a c t i v i t i e s . This appears to be a questionable d i v i s i o n , e s p e c i a l l y as manufactur-ing land i s included with warehousing land. The aggregated i n d u s t r i a l variable then represents two d i s t i n c t a c t i v i t i e s , manufacturing and d i s t r i b u t i n g , that are major users of i n t r a -urban trucking services. The importance of Vancouver as a supply centre for the rest of the province lends credence to the i n c l u s i o n of wholesaling a c t i v i t i e s i n the a t t r a c t i o n measure. The land-use data compiled by the GVRD does not contain any reference to v a r i a t i o n i n i n t e n s i t y of use i n s p e c i f i c categories. Data related to use intensity, such as employment figures disaggregated s p a t i a l l y and by type of industry, or s i m i l a r l y structured data based on floor-space, were not a v a i l -able for t h i s study. The lack of a land-use i n t e n s i t y measure i s not as disastrous for research on i n d u s t r i a l issues as i t o would be for issues concerning o f f i c e or commercial a c t i v i t i e s , as the range i n use-intensity i s not as large. However, the difference i n the per acre p o t e n t i a l of goods movement demand between the Vancouver dockyard warehouses and the wholesaling operations i n Richmond, that are surrounded by large parking areas, i s s i g n i f i c a n t . In order to f u l l y investigate the 105 r e l a t i o n s h i p of i n d u s t r i a l a c t i v i t y and terminal shipment patterns, some measure of the i n t e n s i t y of a c t i v i t y i s needed. The approach presented here i s to use the density of manufacturing land per t o t a l urban zone area as a surrogate for the i n t e n s i t y of use. Density i s b a s i c a l l y a s t a t i s t i c a l a r t i f a c t i n that the c a l c u l a t i o n of density values i s as dependent on the size and shape of areal aggregation units as i t i s on the actual concentration of a given land-use type. In the case of the VUT data, zones were designed to conform to the actual land-use pattern, so that the bulk of land i n each zone i s of either one primary category or a few related categories. This i s p a r t i c u l a r l y true of i n d u s t r i a l and commercial land, both because these a c t i v i t i e s tend to be s p a t i a l l y concentrated (due to zoning regulations as well as agglomeration economies) and because the VUT zone design recognizes the importance of these a c t i v i t i e s to trucking operations. As a result, the actual areal extent of indus-t r i a l clusters i s represented r e l a t i v e l y well by the zone arrangement used. The 70 zones were aggregated into 9 classes based on the density of manufacturing plus auto-related land r e l a t i v e to t o t a l urban land i n each zone. The derived map of i n d u s t r i a l concentration i s shown i n Figure 28. The results of non-parametric s i g n i f i c a n c e tests applied to various combinations of flow component d i s t r i b u t i o n s are l i s t e d i n Figure 29. The only main conclusion that can be drawn from these results i s that the CN terminal Rail-Term and Other components are the same, while the CP terminal Rail-Term shipments are s i g n i f i -FIGURE 29 FLOW COMPONENT DISTRIBUTION TESTS: CLASSES AGGREGATED BY INDUSTRIAL CONCENTRATION T e s t D i s t r i b u t i o n s K e n d a l l ' s Tau C CN - R a i l , Terra. Or., O t h e r CP - R a i l , Term. Or., O t h e r CN - R a i l , T e r m i n a l O r i e n t e d CN - T e r m i n a l O r i e n t e d , O t h e r CN - R a i l , O t h e r CP - R a i l , T e r m i n a l O r i e n t e d CP - T e r m i n a l O r i e n t e d , O t h e r CP - R a i l , O t h e r CN - R a i l - T e r m ; O t h e r CP - R a i l - T e r m , O t h e r CN - R a i l - T e r m , CP -' R a i l - T e r m CN - O t h e r , CP - O t h e r CN - T o t a l , CP - T o t a l -.05167 -.22702 -.2^223 +.05031 -.10737 -.32365 -.01984 -.31054 -.00478 -.19525 +.22894 +.08437 +.14824 * R a i l - T e r m = R a i l + T e r m i n a l O r i e n t e d Shipments. S i g . K o lmogorov - S m i r n o v ; s i g . v a r i a t i o n .1429 . 0000 . 0000 .2904 .0232 .0000 .7337 .0000 .921 8 . 0000 , 0000 , 081 0 ,0000 n o t a p p l i c a b l e n o t a p p l i c a b l e s i g n i f i c a n t s i g n i f i c a n t s i g n i f i c a n t s i g n i f i c a n t s i g n i f i c a n t s i g n i f i c a n t s i g n i f i c a n t s i g n i f i c a n t s i g n i f i c a n t s i g n i f i c a n t M O C7\ 107 c a n t l y more r e l a t e d t o m a n u f a c t u r i n g c o n c e n t r a t i o n t h a n a r e t h e O t h e r s h i p m e n t s i n z o n e 6. F u r t h e r m o r e , K e n d a l l ' s Tau C i n d i c a t e s t h a t t h e two O t h e r c o m p o n e n t s a r e n o t s i g n i f i c a n t l y d i f f e r e n t , w h i l e t h e C P R a i l - T e r m d i s t r i b u t i o n i s shown t o be s i g n i f i c a n t l y more r e l a t e d t o m a n u f a c t u r i n g c o n c e n t r a t i o n t h a n t h e c o r r e s p o n d i n g C N z o n e component by b o t h t e s t s . I t i s c l e a r t h a t t h e R a i l c o m p o n e n t s a r e t h e m o s t a s s o c i a t e d w i t h manu-f a c t u r i n g c o n c e n t r a t i o n . F i g u r e s 30 a n d 31 i l l u s t r a t e t h e r e l a t i o n s h i p among t h e f l o w c o m p o n e n t s a n d b e t w e e n m a n u f a c t u r i n g l a n d d i s t r i b u t i o n a n d t h e p a t t e r n s o f g o o d s movement. The two C P t e r m i n a l c o m p o n e n t s a r e c l e a r l y more r e l a t e d t o m a n u f a c t u r i n g c o n c e n t r a -t i o n t h a n a r e t h e c o m p a r a b l e z o n e 17 c o m p o n e n t s . F i g u r e 32 p r o v i d e s i n f o r m a t i o n r e l a t e d t o z o n e 6 d e l i v e r i e s ; t h e s e d i s t r i b u t i o n s a r e i n d i s t i n g u i s h a b l e f r o m t h e t o t a l c o n s i g n m e n t p a t t e r n s i n F i g u r e 30; t h e c l o s e s i m i l a r i t y o f p i c k - u p a n d d e l i v e r y d i s t r i b u t i o n s j u s t i f i e s t h e u s e o f t h e c o m b i n e d d a t a . F i g u r e 33 i l l u s t r a t e s t h e t w o c o m b i n e d R a i l - T e r m d i s t r i b u t i o n s . The s i m i l a r i t y i n t h e g e n e r a l p a t t e r n o f t h e c u r v e s i s due t o t h e n a t u r e o f t e r m i n a l t r u c k i n g o p e r a t i o n s i n r e l a t i o n t o o t h e r t r a n s p o r t modes. The l a r g e number o f d e l i v e r i e s c o m i n g f r o m / g o i n g t o z o n e s w i t h v e r y l o w m a n u f a c t -u r i n g c o n c e n t r a t i o n s a r e r e l a t e d t o o t h e r l a n d - u s e s (most l i k e l y r e t a i l / c o m m e r c i a l , e s p e c i a l l y r e g a r d i n g l i n k s w i t h f i r m s l o c a t e d i n t h e C B D ) . The c u r v e s a r e a p p r o x i m a t e l y n o r m a l l y d i s t r i b u t e d a b o v e t h e 10% d e n s i t y c a t e g o r y ( F i g u r e 3 4 a ) . The s t r e n g t h o f l i n k a g e s b e t w e e n t h e t e r m i n a l s a n d z o n e s o f medium c o n c e n t r a t i o n o f i n d u s t r i a l a c t i v i t y a n d t h e r e l a t i v e l y FIGURE 30 ZONE 6 COMPONENT DISTRIBUTIONS vs. INDUSTRIAL CONCENTRATION 1 .0 C u m u l a t i v e F r e q u e n c y ( c o n s i g n m e n t s ) Industrial Concentration (Percent Ind. Land) FIGURE 31 ZONE 17 COMPONENT DISTRIBUTIONS v s . INDUSTRIAL CONCENTRATION FIGURE 32 I l l FIGURE 33 RAIL-TERMINAL COMPONENTS v s . INDUSTRIAL CONCENTRATION 1 . 0 , C u m u l a t i v e F r e q u e n c y ( c o n s i g n m e n t s ) 20 AO 60 I n d u s t r i a l C o n c e n t r a t i o n ( P e r c e n t I n d . Land) 112 FIGURE 34 ADJUSTED COMPONENT DISTRIBUTIONS v s . INDUSTRIAL CONCENTRATION C u m u l a t i v e F r e q u e n c y P e r c e n t I n d u s t r i a l Land C u m u l a t i v e F r e q u e n c y 20 40 60 P e r c e n t I n d u s t r i a l Land 1 1 3 weak linkages with zones of high i n d u s t r i a l concentrations can be attributed to competition from other transport modes to s a t i s f y goods movement demand. I t i s l i k e l y that firms i n the zones of high manufacturing concentration either u t i l i z e d i r e c t r a i l and water transport, or operate t h e i r own i n t e r -urban truck f l e e t s , and so use terminal-related services less than do smaller firms i n less concentrated areas. These di s t r i b u t i o n s are consistent with flow patterns predicted by modal-split models; the usual l o g i s t i c - c u r v e formulation of these models i s derived from the r a t i o of two mode- s p e c i f i c flow patterns, both of which are normally d i s t r i b u t e d . The d i s t r i b u t i o n of the Other components i s shown i n Figure 35. The zone 6 shipments are closely related to the d i s t r i b u t i o n of manufacturing land (although a one-sample K-S te s t just f a i l s to be s i g n i f i c a n t at the 5% l e v e l ) , while zone 17 Other shipments are s p l i t between very low and f a i r l y high concentration zones. The zone 6 shipments exhibit no modal-split c h a r a c t e r i s t i c s , either through the entire range of manufacturing concentration categories or for the s p e c i f i c -a l l y i n d u s t r i a l zones (see Figure 34b). Of the 50% of zone 17 Other shipments going to/coming from these zones, 2 8% were small produce consignments delivered to one zone. These were removed from the sample to test the f u l l extent of possible d i s t o r t i o n (Figure 34b-adjusted d i s t r i b u t i o n ) . The correction normalizes the d i s t r i b u t i o n so that i t becomes sim i l a r to the CP Rail-Term d i s t r i b u t i o n , yet i t i s s t i l l s i g n i f i c a n t l y d i f f e r e n t from either the comparable CP Other 114 1 1 5 or the CN Rail-Term d i s t r i b u t i o n s . The f i n a l part of the analysis of a t t r a c t i o n was to further disaggregate the Rail-Term d i s t r i b u t i o n s according to commodity type. There was found to be no s i g n i f i c a n t variations within the CN Rail-Term samples related to type of good shipped (Figure 36). On the other hand, a l l the CP Rail-Term d i s -aggregated samples showed s i g n i f i c a n t variations from each other (Figure 37). However, even though the sample sizes become very small at t h i s l e v e l of disaggregation, a l l commod-i t y types have s i g n i f i c a n t l y d i f f e r e n t d i s t r i b u t i o n s when tested between terminal zones. This evidence supports the finding that the CP terminal i s more linked to i n d u s t r i a l a c t i v i t i e s than i s the CN terminal. In summary, the use of manufacturing concentration as the measure of a t t r a c t i o n reveals important variations both within and between zonal commodity flows. The i n d i c a t i o n that between zone v a r i a t i o n i n comparable flow components i s not i n the same dir e c t i o n lends credence to the basic rationale of component analysis: i n t e r n a l v a r i a t i o n hides s i g n i f i c a n t differences between zone d i s t r i b u t i o n s . I i i ) /ANALYSIS OF DISTANCE DECAY The distance measure used i n t h i s analysis i s the average t r a v e l time between zones, representing the costs of goods movement. The 70 zones have been aggregated into two sets of 9 zones according to t r a v e l times from zone 6 and zone 17, respectively. The zone maps are i l l u s t r a t e d i n Figures 9 and FIGURE 36 COMMODITY TYPE DISTRIBUTIONS BY INDUSTRIAL CONCENTRATION: CN RAIL-TERMINAL COMPONENT 20 40 60 I n d u s t r i a l C o n c e n t r a t i o n ( P e r c e n t I n d . Land) FIGURE 37 COMMODITY TYPE DISTRIBUTIONS BY INDUSTRIAL CONCENTRATION: CP RAIL-TERMINAL COMPONENT 1.0 , _ . . 20 40 60 I n d u s t r i a l C o n c e n t r a t i o n ( P e r c e n t I n d . Land) FIGURE 38 CLASSIFICATION OF ZONES B I TRAVEL-TIME FROM ZONE 17 T r a v e l -Time i n M i n u t e s 0-3 3.1-5 5.1-10 10.1-12 12.1-15 15.1-18 18.1-22 22.1-30 Li l •30 • CN T e r m i n a l 119 38. I t can be seen that the zone aggregation i s very gross. The zones were based on respondents' estimates, the only available information for truck t r a v e l times. Adjustment of these estimates would give only the appearance of greater r e l i a b i l i t y . The f i r s t stage of the analysis disregards the a t t r a c t i o n variable, and examines the pure ef f e c t s of distance on consign-ment d i s t r i b u t i o n s . The second stage introduces the a t t r a c t i o n p o t e n t i a l to determine whether the flow variations noted i n the previous section can be attributed to distance e f f e c t s . The results of s t a t i s t i c a l tests to determine the s i g n i -ficance of variations between d i s t r i b u t i o n s based on t r a v e l time are shown i n Figure 39. The two Rail-Term and the two Total d i s t r i b u t i o n s exhibit no s i g n i f i c a n t differences, although i n t e r n a l differences within these components are marked. The two R a i l components are both more affected by distance costs than are Terminal-Oriented components; t h i s r e s u l t i s expected as the R a i l components include only d i r e c t consignments, where-as Terminal-Oriented components include many consignments shipped i n two stages. Figure 40 i l l u s t r a t e s the ef f e c t s of distance on the two Rail-Term components. The median shipping distance i s only 8 minutes; f u l l y 80% of t o t a l shipments occur within an 18 minute radius of the terminal zones. The distance decay i n terms of shipments per unit of area i s shown i n Figures 41 and 42. The e l a s t i c i t y of shipments/acre with respect to distance are equivalent to the 'a' distance exponent i n the linkage i n t e n s i t y equation: FIGURE 39 FLOW COMPONENT DISTRIBUTION TESTS: CLASSES AGGREGATED BY TRAVEL-TIME T e s t D i s t r i b u t i o n s CM - R a i l , Terra. Or., O t h e r CP - R a i l , Term. Or., O t h e r CN - R a i l , T e r m i n a l O r i e n t e d CN - T e r m i n a l O r i e n t e d , O t h e r CN - R a i l , Other CP - R a i l , T e r m i n a l O r i e n t e d CP - T e r m i n a l O r i e n t e d , O t h e r CP - R a i l , Other CN - R a i l - T e r m * O t h e r CP - R a i l - T e r m , O t h e r CN - R a i l - T e r m , CP - R a i l - T e r m CN -. O t h e r . CP - O t h e r CN - T o t a l , CP - T o t a l K e n d a l l ' s Tau C +.06592 + .27434-+.45062 -.13834 +.18136 +.23159 +.11845 +.39725 -.03141 +.28885 -.08326 +.20307 +.05764 S i g . . 0684 . 0000 . 0000 ,0049 .0002 . 0002 ,0364 ,0000 ,4975 ,0000 ,0636 ,0000 0822 Kolmogorov - S m i r n o v ; s i g . v a r i a t i o n n o t a p p l i c a b l e n o t a p p l i c a b l e s i g n i f i c a n t s i g n i f i c a n t s i g n i f i c a n t s i g n i f i c a n t s i g n i f i c a n t s i g n i f i c a n t s i g n i f i c a n t s i g n i f i c a n t s i g n i f i c a n t I—' ro o * R a i l - T e r m = R a i l + T e r m i n a l O r i e n t e d S h i p m e n t s . FIGURE 40 FIGURE 41 RAIL-TERMINAL COMPONENT DISTRIBUTIONS: ELASTICITY WITH RESPECT TO TRAVEL-TIME T r a v e l - T i m e i n M i n u t e s FIGURE 42 OTHER COMPONENT DISTRIBUTIONS: ELASTICITY WITH RESPECT TO TRAVEL-TIME T T 1 5 T r a v e l - T i m e i n M i n u t e s 10 20 -1— 30 124 The comparable components between zones show a reversal of distance decay e f f e c t s , while the d i r e c t i o n of within zone va r i a t i o n i s the same i n both cases. The longer length of Other shipments i s p a r t i a l l y due to t h e i r d e f i n i t i o n : consign-ments between Type 3 a c t i v i t i e s and the l o c a l r a i l terminal complex are c l a s s i f i e d as either R a i l or Terminal-Oriented shipments; i t i s thus l i k e l y that many short-distance shipments going t o / coming from Type 3 locations are not i n the Other category. The proportion of t r i p s l o s t due to c l a s s i f i c a t i o n methods should be about equal for both zones. Comparison be-tween the two Other components should thus s t i l l be r e l i a b l e . The difference i n distance decay between the Rail-Term components aids i n understanding the linkage r e l a t i o n s h i p of the two terminals. It has been seen that the CP terminal has stronger l i n k s to i n d u s t r i a l a c t i v i t y than does the CN terminal. One of two possible factors could underlie t h i s observation: 1. the terminals d i f f e r i n terms of operating character-i s t i c s . 2. the terminal operations are similar, but their s p a t i a l environments d i f f e r . The means of estab l i s h i n g the most l i k e l y explanation i s to derive linkage i n t e n s i t y values for each terminal. I f the in t e n s i t y index i s similar, i t i s reasonable to assume that terminal operations are s i m i l a r . The a t t r a c t i o n measure used here w i l l be manufacturing land weighted by concentration. The weighting i s done to allow for land-use i n t e n s i t y . The method of ca l c u l a t i o n i s as follows: P=M(—,), where P= p o t e n t i a l M= man. and wholesaling land U= urban land The use of a density measure for concentration i s affected by the v a r i a t i o n i n zone size s . To minimize d i s t o r t i o n s , the mean i n d u s t r i a l density of aggregated distance zones was calculated by weighting the density values of the subset of zones within each aggregated zone by the size of the zone i n question. The d i s t r i b u t i o n of manufacturing land i s shown i n Figure 43; t h i s graph indicates that zone 17 i s more ce n t r a l l y located with regard to t o t a l i n d u s t r i a l land than i s zone 6. The wide variations i n amount of i n d u s t r i a l land with regard to distance suggests that land-use i n the region i s arranged to some degree i n concentric rings. The v a r i a t i o n of density with distance i s i l l u s t r a t e d i n Figure 44. The t h e o r e t i c a l r e l a t i o n s h i p of density with distance from the urban centre i s : density = (k) (transport costs)-''", as has been discussed previously. The actual density gradient clo s e l y follows the t h e o r e t i c a l approximation, except for an increase i n density occurring about 26 minutes from the terminals This bulge represents the e f f e c t of the s a t e l l i t e c i t y of New Westminster, which contains a major port operation. The fact that zone 6 i s closer to the urban centre than zone 17 accounts for the much higher density i n the immediate v i c i n i t y of t h i s zone. The d i s t r i b u t i o n of i n d u s t r i a l a c t i v i t y .vis i l l u s t r a t e d i n Figures 45 and 46. Figure -4 5 (a,b) i s based on t o t a l p o tential, not p o t e n t i a l per unit area. The d i s t r i b u t i o n of 126 FIGURE U INDUSTRIAL CONCENTRATION v s . TRAVEL-TIME AO 30 I n d u s t r i a l C o n c e n t r a t i on ( P e r c e n t I n d . Land) 20 1 0 CN T e r m i n a l : GP T e r m i n a l : ( I n d . Cone. = T r a v e l - T i m e " 1 0 20 T r a v e l - T i m e i n M i n u t e s 30 AO 128 FIGURE 15 POTENTIAL INDUSTRIAL CONTACTS v s . TRAVEL-TIME 2 0 0 n a) I n d e x 10 20 30 T r a v e l - T i m e i n M i n u t e s b) C u m u l a t i v e F r e q u e n c y 10 20 30 40 T r a v e l - T i m e • i n M i n u t e s FIGURE 46 ELASTICITY OF POTENTIAL INDUSTRIAL CONTACTS WITH RESPECT TO TRAVEL-TIME ( E l a s t i c i t y = -1.0) I I i I I 1 5 10 20 30 T r a v e l - T i m e i n M i n u t e s 130 a c t i v i t y from the more central CP zone c l e a r l y shows that industries are located either close to the c i t y centre (the location of the harbour i s of basic importance here) or i n outlying areas. The exponents r e l a t i n g increasing t r a v e l time and the measure of p o t e n t i a l contacts per unit area were calculated, and are l i s t e d i n Figure 47. The o v e r - a l l Rail-Terminal linkage i n t e n s i t y values turn out to be v i r t u a l l y i d e n t i c a l (Figure 47). This supports the contention that the i n t e r n a l operations of the terminals are the same, and that differences i n the amount of i n t e r a c t i o n between the terminals and i n d u s t r i a l land-uses i s due to r e l a t i v e location. Disaggregation of the i n t e n s i t y index by distance further i l l u s t r a t e s the strong association of the CP terminal with the surrounding i n d u s t r i a l area (Figures 48 and 49). The linkage i n t e n s i t y values of both terminals have two d i s t i n c t sections: CP Rail-Term: 0 - 1 6 minutes I=D " 16 - 38 " I=D + 2 CN Rail-Term: 0 - 9 minutes I=D :, 9 C. 10 - 38 " l=D~' The small samples make the r e l i a b i l i t y of index subsets questionable. However, the r e l a t i v e focus of the CP terminal on near-by industries, and the opposite tendency shown by the CN terminal flows, i s c l e a r l y demonstrated. The higher con-centration of i n d u s t r i a l a c t i v i t y i n the v i c i n i t y of the CP terminal explains the more rapid distance decay exhibited by the CP Rail-Term flow component. This r e s u l t also supports the hypothesis that i n d u s t r i a l FIGURE 47 DISTANCE DECAY EXPONENT VALUES FOR TERMINAL ZONE DISTRIBUTION COMPONENTS D i s t r i b u t i o n Component a - b CP: R a i l - T e r m i n a l O t h e r •2.15 -1 .1 •1.4 •1.4 -.75 + .3 CN: R a i l - T e r m i n a l O t h e r -1 .77 -1.4 -1 .0 -1 .0 -.77 -.4 I—1 The f o r m u l a f o r t h e exponent v a l u e s i s : I = p t h e r e f o r e I = D A = D P = D b , (a-b) FIGURE 48 LINKAGE INTENSITY OF RAIL-TERMINAL COMPONENT DISTRIBUTIONS WITH RESPECT TO TRAVEL-TIME T r a v e l - T i m e i n M i n u t e s FIGURE 49 LINKAGE INTENSITY OF OTHER COMPONENT DISTRIBUTIONS WITH RESPECT TO TRAVEL-TIME T r a v e l - T i m e i n M i n u t e s 134 firms trade off linkage costs for land costs within an urban area. The more highly concentrated area around zone 6 would be more expensive, but t h i s added expense i s exactly made up by the reduced cost of transporting goods a short distance to the terminal. The very low distance decay of Other shipments going to/coming from zone 6 suggests that the non-terminal linkages of these firms are over-stretched. Many of these firms have relocated since the survey was conducted i n 1975; the linkage i n t e n s i t y exponent of +.3 indicates that, as i n d u s t r i a l suburbanization increases, linkage costs may act-u a l l y decrease with distance from the urban centre. The functional i d e n t i t y of the two distance exponents, (a-b) and (c), implies that regression techniques would be unable to distinguish between the flows of the two terminals even i f allowances for component disaggregation are made. The terminals are s t a t i s t i c a l l y associated with the a t t r a c t i o n and distance decay variables i n the same general way; i t i s the i n t e r - r e l a t i o n s h i p of the variables which i s important. The f i n a l step i s to investigate the r e l a t i o n s h i p of distance and commodity type. Figures 50 and 51 reveal that there are no s i g n i f i c a n t differences between zones of the three commodity d i s t r i b u t i o n s . The only s i g n i f i c a n t within-zone var i a t i o n i s that zone 6 Rail-Terminal Fabricated shipments are more sensitive to distance than are other shipments. This supports the conclusion that the CP terminal i s c l o s e l y linked to the i n d u s t r i a l firms i n the immediate v i c i n i t y . 50% of fabricated goods are shipped a distance of less than 7 minutes; commodities i n t h i s category are the most l i k e l y to be involved i n shipments from/to i n d u s t r i a l firms. FIGURE 50 COMMODITY TYPE v s . TRAVEL-TIME: CP RAIL-TERMINAL COMPONENT 1.0 , 1 0 20 30 AO T r a v e l - T i m e i n M i n u t e s FIGURE 51 COMMODITY TYPE v s . TRAVEL-TIME: CN RAIL-TERMINAL COMPONENT 137 E. FOOTNOTES TO CHAPTER TWO 1. Swan Wooster Engineering Co. Ltd.; Evaluation of Urban  Trucking Rationalization i n Vancouver, prepared for Transport Canada Research and Development Centre, Ministry of State for Urban A f f a i r s , and the City of Vancouver; Vancouver, 1978. 2. The information i n t h i s section from the Phase I Report of the Swan Wooster study. 3. Swan Wooster Engineering Co. Ltd.; op. c i t . , page 4-14. 4. Urban Goods Movement Research - A Framework and Results; Urban Goods Movement Report Series, Vol.1, Montreal, 19 79,-page 162. 5. The Economics of Urban Goods Movement; Urban Goods Movement Report Series, Vol. 10, Montreal, 1979, page 27. 6. This i s d e f i n i t e l y the case for some of the large trucking firms operating out of zone 6. Richmond Transport, for example, did serve primarily the southern parts of the region. 7. Large . l e a s e f l e e t s .-.and.: e x t r a - p r o v i n c i a l f l e e t s a r e omitted as size of operating firm i s unknown. 8. Swan Wooster Engineering Co. Ltd., op. c i t ; page 5-5. 9. i b i d . , page 5-43. 10. Yates, F., Sampling Methods for Censuses and Surveys; London, 1960; quoted i n Haggett, P., et a l . , Locational Methods; London, 1977, page 270. 11. Swan Wooster Engineering Co. Ltd., op. c i t . ; page 4-24. 12. This does not necessarily hold for a l l linkage research, as some bulk commodities (eg. o i l , l i q u i d chemicals) are highly s i g n i f i c a n t i n linkage flows. In the present case, however, LTL fr e i g h t constitutes the major ( i f not only) flow component. 13. Haggett, P., et a l . , op. c i t . ; page 287. 14. i b i d ; page 286. 15. i b i d ; page 286. 16. The alte r n a t i v e method i s accurate only i f t r i p s are equally d i s t r i b u t e d among zones. I f t r i p s are dist r i b u t e d according to zone size, mean d' w i l l be underestimated. 138 17. Swan Wooster Engineering Co. Ltd., op. c i t . ; page 5-20. 18. Wonnacott, R. and Wonnacott, T., Econometrics, Wiley and Sons, Inc., New York, 1970; page 61. 19. The regression equation i s a correct model of the observed relationships between variables only i f the residual errors are unbiased. In order to s a t i s f y t h i s condition, errors must be independent, have a mean of zero and a constant variance, and follow a normal d i s t r i b u t i o n . Discussions of methods of te s t i n g residuals for the a p p l i c a b i l i t y of these assumptions are found i n Draper, N. and Smith, H., Applied  Regression Analysis, Wiley and Sons, Inc., New York, 1966, pages 86 to 95; Anscombe, F.J. and Tukey, J.W., "The Examina-ti o n and Analysis of Residuals", Technometrics, Vol. 5, 1963, pages 141 to 160. 20. Sayer, R.A., "A Critique of Urban Modelling", Progress  i n Planning, Vol. 6, Part 3, 1976; page 20 9. 21. Coughlin, R.E., et a l . , Socio'-Economic Aspects of  Motor Freight Terminal Location, Federal Highway Administration, Department of Transportation, Washington, 1976. 22. G r i f f i t h , D.W. and Preston, R.E., "A Restatement of the Transition Zone Concept", Annals of the Association of American. Geographers, Vol. 56, 1966. In defining the " t r a n s i t i o n zone", the authors state that "usually present are such intensive non-retail a c t i v i t i e s as o f f street parking, warehousing, l i g h t manufacturing, wholesaling with stocks, special profess-i o n a l organizational services, transportation terminals, and multi-family residences" (page 341). This description accurately r e f l e c t s the composition of the CP terminal zone. 23. Siegel, S., Nonparametric S t a t i s t i c s , McGraw-Hill Co., New York, 1956; page 104. 24. Blalock, H.M., Soc i a l S t a t i s t i c s , McGraw-Hill Inc., 19 70; page 2 93. 25. i b i d ; pages 295 to 298. 26. i b i d , pages 421 to 424. 27. i b i d ; page 17. 28. Siegel, S., op. c i t ; pages 127-132. 29. i b i d ; page 12 7. 30. Hoare, A., "Three Problems for In d u s t r i a l Linkage Stud-i e s " , Area, Vol. 10, No. 3, 1978; page 218. 31. Alonso, W., "A Theory of the Urban Land Market", Papers  and Proceedings of the Regional Science Association, Vol. 6, 1960; pages 149 to 158. 139 CHAPTER THREE  CONCLUSION The conclusions drawn from t h i s study can be divided into three subgroups: f i r s t , those related to an understanding of r a i l terminal linkages; secondly, conclusions which a i d i n the characterization of the role of the False-Creek terminal in Vancouver's i n d u s t r i a l structure; and t h i r d l y , those re-lated to methodological considerations. A. RAIL TERMINAL LINKAGES The empirical analysis of the preceding chapter was based on a s t a t i c , cross-sectional study of the goods flow patterns associated with two inner-urban r a i l / t r u c k terminal operations. The terminals are located within one mile of each other, so that t h e i r s p a t i a l orientation with regard to the urban region i s si m i l a r i n each case. The s t a t i c nature of the analysis i s not of c r i t i c a l concern as both terminals pre-date the establishment of almost a l l other land-uses i n Van-couver. The temporal sequence by which a c t i v i t i e s linked to the terminals located i s unknown, but i t i s clear that any chain of causation must run from the terminals to the estab-lishment of other a c t i v i t i e s and not i n the opposite d i r e c t i o n . It was found that no s i g n i f i c a n t differences e x i s t be-tween the types of commodities or the weight d i s t r i b u t i o n of consignments„handled by the two terminals. Thus neither terminal s p e c i a l i z e s i n a p a r t i c u l a r type or range of good. Furthermore, consignments tend to be r e l a t i v e l y l i g h t (over 50% 140 weigh less than 250 l b s . ) , but are s t i l l generally heavier than non-terminal shipments coming from/going to terminal zones (60% of these consignments are under 250 l b s . ) . This i s evid-ence of a s l i g h t difference between the size of LTL i n t e r -urban consignments and intra-urban consignments. Larger i n t e r -urban consignments are d i r e c t l y routed from o r i g i n to destina-tion; the terminal operation (including associated bulking/ debulking f a c i l i t i e s ) competes d i r e c t l y with line-haul truck-ing firms for inter-urban consignments too small to be e f f i c i e n t l y transported i n a single haul. The d i s t r i b u t i o n of terminal related consignments with respect to the pattern of i n d u s t r i a l a c t i v i t y i n the urban region exhibits two s i g n i f i c a n t features. F i r s t , approximately 30% of a l l consignments come from/go to areas with l i t t l e or no manufacturing, wholesaling or warehousing a c t i v i t y . The majority of these consignments represent terminal linkages with r e t a i l a c t i v i t i e s . Secondly, the 70% of consignments coming from/going to i n d u s t r i a l areas are most strongly associated with zones of medium land-use i n t e n s i t y . Highly concentrated i n d u s t r i a l areas are located with regard to the a v a i l a b i l i t y of d i r e c t inter-urban transport f a c i l i t i e s , e ither r a i l - s i d i n g s or port access. The modal s p l i t between dir e c t and trans-shipped consignments, based on consignment size, thus i s r e f l e c t e d i n terminal-industry linkages. Large firms with access to sp e c i a l i z e d transport f a c i l i t i e s or who operate t h e i r own truck f l e e t s are not closely linked to the terminal. , While smaller firms with i r r e g u l a r f r e i g h t transport-ation requirements exhibit strong terminal linkages. 141 The length of terminal linkages generally are quite short. Over 50% of terminal consignments t r a v e l a distance taking less than 10 minutes, and only 10% require more than 20 minutes delivery dr i v i n g time. This distance decay i s not only due to the d i s t r i b u t i o n of i n d u s t r i a l a c t i v i t y i n r e l a t i o n to terminal location. The f r e i g h t flows of both terminals exhibit a distance decay exponent of approximately -.75 a f t e r the e f f e c t of potential a t t r a c t i o n has been accounted for. The close s i m i l a r i t y of the terminals' linkage i n t e n s i t y exponents implies that r a i l terminals have a strong l o c a t i o n a l e f f e c t on linked industries, and that t h i s e f f e c t i s constant regardless of a terminal's location i n r e l a t i o n to o v e r - a l l i n d u s t r i a l d i s t r i b u t i o n . The strong linkage connections between the terminals and adjacent industries suggest that inner c i t y i n d u s t r i a l areas form viable i n d u s t r i a l complexes. The urban agglo^ meration economies offered by central locations may not be as important as they once were, but such economies s t i l l i n f l u -ence the location pattern of a certain range of firms. Bourne points out that the distinguishing feature of inner core areas characterized as "zones i n t r a n s i t i o n " i s that they are often f u n c t i o n a l l y very stable; the areas of redevelopment and functional change are i n fact the true t r a n s i t i o n a l parts of the city."'" This c r i t i c i s m of urban growth theory i s supported by the persistence of i n d u s t r i a l a c t i v i t y i n the terminal zones which exhibits a s i g n i f i c a n t l e v e l of i n t e r n a l linkage i n t e r a c t i o n . 142 B". FALSE-CREEK' TERMINAL LINKAGES The i n d u s t r i a l linkages of the False-Creek Terminal exhibit a distance decay which i s more pronounced i n the immediate v i c i n i t y of the terminal than i s the corresponding CN terminal measure. In addition, fabricated materials are the commodity types most sensitive to distance; 70% are shipped a distance of less than 10 minutes (compared to 56% of general freight and 46% of other consignments). The comparable figures for the CN terminal are 5 0% fabricated materials, 5 4% general fr e i g h t shipments and 69% other shipments. Thus i t i s possible to assert that the CP (False-Creek) terminal i s more closely linked to manufacturing a c t i v i t i e s than i s the CN terminal, while the CN terminal i s more associated with wholesaling and d i s t r i b u t i o n a c t i v i t i e s . In addition, the l o c a t i o n a l e f f e c t of these manufacturing linkages i s more pronounced than are the e f f e c t s of d i s t r i b u t i o n linkages. The out-migration of inner-core industries which has occurred since the VUT survey was undertaken i n 1975 was prim-a r i l y due to government-initiated urban redevelopment a c t i v i t y . However, the low distance decay of non-terminal zone 6 ( i . e . , the CP terminal zone) linkages suggests that, by 1975, firms i n the v i c i n i t y of the terminal were not well located i n regard to other i n d u s t r i a l a c t i v i t i e s . The lack of an intra-urban 2 freeway system has resulted i n a "core-ring" urban structure , and the large land requirements of modern one-storey i n d u s t r i a l plants has made the outer r i n g more a t t r a c t i v e for industry. Firms s t i l l located i n the inner core are those most strongly linked to the terminal; relocation of the terminal w i l l l i k e l y 143 e n t a i l a migration of these plants to locations e i t h e r closer to the CN terminal, or i n the v i c i n i t y of the new CP terminal to be b u i l t i n an outlying area. The inner c i t y location of the False-Creek terminal has been a major attractor of i n d u s t r i a l a c t i v i t y i n t h i s area. Redevelopment of the terminal s i t e w i l l reduce the a t t r a c t i v e -ness of inner c i t y locations for i n d u s t r i a l firms; Vancouver w i l l become functionally segregated to a higher degree, with the core area becoming increasingly devoted to commercial a c t i v i t i e s and the suburban ri n g becoming more i n d u s t r i a l i z e d . The only i n d u s t r i a l a c t i v i t i e s to remain i n the core w i l l be those c l o s e l y linked to port f a c i l i t i e s . C. METHODOLOGY This study has made use of a comprehensive urban goods movement data base, c o l l e c t e d for the purposes of truck transportation modelling. The application of th i s data to an analysis of a s p e c i f i c linkage question has been accomplish-ed, at the cost of substantial loss of d e t a i l . The l e v e l of analysis at which s i g n i f i c a n t relationships could be derived i s f a i r l y general; indications of the importance of firm size and type and commodity class to terminal linkage structure, f o r example, cannot be f u l l y investigated. Comprehensive studies supply data which can be best used to delimit data requirements for more highly focused goods movement studies. Daniels and Warnes state that "more modest and progressively refined studies of s p e c i f i c aspects of urban f r e i g h t move-ments may be more appropriate and productive than comprehensive 144 .3 surveys and analyses".' However, analysis of city-wide freight movement data does f a c i l i t a t e the i d e n t i f i c a t i o n of data requirements f o r more narrowly defined research. The present study suggests that r a i l terminals have a s i g n i f i c a n t l o c a t i o n a l e f f e c t on smaller i n d u s t r i a l firms, e s p e c i a l l y i n the manufacturing sector. Further studies, using d i r e c t methods to measure land-use intensity, could be under-taken to test the e f f e c t of the relocated CP terminal on i n d u s t r i a l d i s t r i b u t i o n . A time series approach, coupled with a more disaggregated c l a s s i f i c a t i o n of commodity type, would be most e f f e c t i v e for t h i s purpose. Furthermore, consignments measured d i r e c t l y , by use of b i l l s of lading, would allow for the i d e n t i f i c a t i o n of origin/destination a c t i v i t y type. This method of sampling i s also more r e l i a b l e , as s p a t i a l bias caused by routing factors i s eliminated. 145 D. FOOTNOTES TO CHAPTER THREE 1. Bourne, L.S., "Comments on the Transition Zone Concept", Professional Geographer, Vol. 20, No. 5, 1968; page 315. 2. Hardwick, W.G., "Vancouver: The Emergence of a Core-Ring Urban Pattern", i n Gentilcore, R.L., Geographical  Approaches to Canadian Problems, Prentice-Hall, Scarborough, 1971; page 114. 3. Daniels, P.W. and Warnes, A.M., Movement i n C i t i e s , Methuen and Co. Ltd., London, 1980; page 112. 146 BIBLIOGRAPHY Alonso, W., "A Theory of the Urban Land Market", Papers and Proceedings of the Regional Science Association, Vol. 6, 1960. .Anscombe, F.J. and Tukey, J.W., "The Examination and Analysis of Residuals", Technometrics, Vol. 5, 1963. B l a i r , J., Industrial P o l a r i z a t i o n and the Location of New Manufacturing Firms: An Empirical Application, Regional Science Research Institute Discussion Paper Series, No. 89, 1976. Blalock, H.M., Social S t a t i s t i c s , McGraw-Hill Inc., New York, 1970. Bourne, L. S., "Comments on the Transition Zone Concept", Professional Geographer, Vol. 20, No. 5, 1968. Cameron, G.C., "Intraurban Location and the New Plant", Papers  and Proceedings of the Regional Science Association, Vol. 31, 1973. Cooper, M.J., The Industrial Location Decision-Making Process, Centre f o r Urban and Regional Studies Occassional Paper 34, University of Birmingham, 1975. Coughlin, R.E., et a l . , The Regional Science Research Institute, Socio-Economic Aspects of Motor Freight Terminal Location, Federal Highway Administration, Department of Transporta-t i o n , Washington, 1976. Czamanski, S- I n d u s t r i a l Clusters, Dalhousie University, Halifax, 1976. Czamanski, S. I n d u s t r i a l Complexes, Dalhousie University, Halifax, 1974. ' Czamanski, S. "Some Empirical Evidence of the Strengths of Linkages Between Groups of Related Industries i n Urban-Regional Complexes", Papers and Proceedings of the Regional Science Association, Vol. 27, 1971. Czamanski, S», "Industrial Location and Urban Growth", The  Town' Planning Review, Vol. 37, 1965. Daniels, P.W. and Warnes, A.M., Movement i n C i t i e s : Spatial  Perspectives on Urban Transport and Travel, Methuen and Co., London, 1980. Davis, H.C., An Interindustry Study of the Metropolitan Vancouver Economy, Urban Land Economics Report No. 6, University of B r i t i s h Columbia, 1973. 147 Draper, N. and Smith, H., Applied Regression Analysis, Wiley and Sons Inc., New York, 1966. Fales, R. and Moses, L., "Land-Use Theory and the Spatial Structure of the Nineteenth-Century City". Papers  and Proceedings of the Regional Science Association, Vol. 28, 1972. French, A. and Watson, P., "Demand Forecasting and Development of Framework for Analysis of Urban Commodity Flow", i n Urban Commodity Flow, Highway Research Board Special Report 120, Washington, 1971. Gannon, CA. "Intra-Urban Ind u s t r i a l Location Theory and Inter-Establishment Linkages", Geographical Analysis, Vol. 5, 1973. Goldberg, M.A., "An Economic Model of Intrametropolitan In d u s t r i a l Location", Journal of Regional Science, Vol. 10, 1970. Goldstein, G.S. and Moses, L.N., "Interdependence and the Location of Economic A c t i v i t i e s " , Journal of Urban  Economics, Vol. 2, 1975. G r i f f i t h , D.W. and Preston, R.E., "A Restatement of the Trans-i t i o n Zone Concept", Annals of the Association of American  Geographers, Vol. 56, 1966. Groves, P.A., Towards a Typology of Intrametropolitan Manu- facturing Location, Occassional Papers i n Geography No. 16, University of Hu l l , 1971. Haggett, P., et. a l . , Locational Methods, Methuen and Co., London, 1977. Hamer, A., Industrial Exodus from the Central City, Heath and Co, Lexington, 1973. Hardwick, W. G., '^Vancouver: The Emergence of a Core-Ring Urban Pattern", i n Gentilcore, R.L., Geographi cal  Approaches to Canadian Problems, Prentice-Hall Ltd., Scarborough, 1971. Hay, A., "The Geographical Explanation of Commodity Flow", Progress i n Human Geography, Vol. 3, No. 2, 1979. Hoare, A.G., "Three Problems for Ind u s t r i a l Linkage Studies", Area, Vol 10, No. 3, 1978. Hoare, A.G., "Linkage Flows, Locational Evaluation, and In-d u s t r i a l Geography: A Case Study of Greater London", Environment and Planning, Series A, Vol. 7, 1975. 148 Hoare, A.G., "International Airports as Growth Poles: A Case Study of Heathrow Airport", Transactions of the I n s t i - tute of B r i t i s h Geographers, Vol. 63, 1974. Hoover, E.M., "The Evolving Form and Organization of the Metropolis", i n P e r l o f f , H.S. and Wingo, L., ed., Issues  i n Urban Economics, Resources for the Future, Inc., Johns Hopkins Press, Washington, 1968. House, R.K. and Associates Ltd. and Clayton, Sparks and Associ-ates Ltd., The Economics of Urban Goods Movement, Urban Goods Movement Report Series Vol. 10, Urban Transportation Research Branch, Transport Canada, Montreal, 197 9. Kain, J., "The D i s t r i b u t i o n and Movement of Jobs and Industry", i n Wilson, J.Q., ed., The Metropolitan Enigma, Harvard University Press, Cambridge, 1968. Lever, W.F., "Industrial Movement, Spa t i a l Association and Functional Linkages", Regional Studies, Vol. 6, 1972. Meyburg, A.H. and Stopher, P.v\R. "A Framework for the Analysis of Demand for Urban Goods Movements", Transportation  Research Record, Vol. 496, 1974. M i l l e r , E.W. and M i l l e r , R.M., Industrial Location: A Bib- liography , International Geographical Union Commission on In d u s t r i a l Systems, Clark University Press, 1978. M i l l e r , M.S., Motor Carr i e r Terminal Location i n the Chicago  Region, Chicago Area Transportation Study, No. 341-11, 1974. Moses, L. and Williamson, H., "The Location of Economic A c t i v i t y i n C i t i e s " , American Economic Review, Vol. 57, 1967. Nourse, H., Regional Economics, McGraw H i l l , New York, 1968. Richter, C.E., "The Impact of I n d u s t r i a l Linkages on Geographic Association", Journal of Regional Science, Vol. 9, 1970. Rogers, A., "Theories of Intra-Urban Spatial Structure: A Dissenting View", Land Economics, Vol. 43, No. 1, 1967. Sackr; R.D., "The Spatial Separatist Theme in Geography", Economic  Geography, Vol. 50, No. 1, 1974. Sayer, R. A., "A Critique of Urban Modelling", Progress i n  Planning, Vol. 6, Part 3, 1976. Siegel, S., Nonparametric S t a t i s t i c s , McGraw H i l l Co., New York, 1956. 149 Steed, G.P.F., "Centrality and Location Change: Printing, Publishing and Clothing i n Montreal and Toronto", Economic Georgraphy, Vol. 52, 1976. Steed, G.P.F., "Intrametropolitan Manufacturing: Spatial D i s t -ribut i o n and Locational Dynamics in> .'Greater Vancouver", Canadian Geographer, Vol. 17, No. 3, 19 73. Steed, G.P.F., "Commodity Flows and Interindustry Linkages of Northern Ireland's Manufacturing Industries",Ti j d s c h r i f t Voor Econ. en Soc. Geografie, Sept./Oct., 1968. Stevens, B.H. and Brackett, C.A., Indu s t r i a l Location,Regional Science Research I n s t i t u t e Bibliography Series No. 3, Philadelphia, 1967. S t r e i t , M.E., "Spatial Associations and Economic Linkages Between Industries", Journal of Regional Science, Vol. 9, 1969. Struyk, R.J. and James, F.J., Intrametropolitan I n d u s t r i a l  Location: The Pattern and Process of Change, Heath and Co., Lexington, 1976. Swan Wooster Engineering Co. Ltd., Evaluation of Urban Truck-ing Rationalization i n Vancouver, 2 Volumes, Vancouver, 1978. Tennant, L.G., Inter-Industry Commodity Links: A Study of Selected Industrial Sectors, City of Vancouver Dept. of Planning and C i v i c Development, Information and S t a t i s t i c s Report No. 11, 1974. Tocalis, T.R., "Changing Theoretical Foundations of the Gravity Concept of Human Interaction", i n Berry, B.J.L., ed., The  Nature of Change i n Geographical Ideas, Perspectives i n Geography Vol. 3, Northern I l l i n o i s University Press, 1975. Townroe, P.M., Industrial Movement, Saxon House, Teakfield, Ltd., Westmead, England, 1979. Wilson, A.G. Urban and Regional Models i n Geography and Plan-ning, John Wiley and Sons, London, 1974. Wonnacott, R. and Wonnacott, T., Econometrics, Wiley and Sons Inc., New York, 1970. Wood, P.A., "Industrial Location and Linkage", Area, Vol.2, 1969. Wood, R.T., "Structure and Economics of Intraurban Goods Movement", i n Urban Commodity Flow, Highway Research Board Special Report 120, Washington, 1971. Wood, W.G., Systems Approach Consultants Ltd., Urban Goods Move-ment Research - A Framework and Results, Urban Goods Movement Report Series Vol. 1, Urban Transportation Research Branch, Transport Canada, Montreal, 1979. 

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:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0095538/manifest

Comment

Related Items